Cerebral Venous Collagen Remodeling in a Modified White Matter Lesions Animal Model.

To mimic the expected pathological changes of white matter lesions (WMLs) and increase the stability, we applied modified two-vessel occlusion (modified 2VO) (1-week interval bilateral carotid artery occlusion) in stroke-prone renovascular hypertensive rats (RHRSP) and established a modified WMLs model (RHRSP/modified 2VO) that compared their phenotypes with RHRSP and sham-operated rats. In addition, we tried to differentiate small veins from small arteries through the presence of smooth muscle to study the pathological changes of small veins detailed in the model. RHRSP/modified 2VO rats showed higher stability and more extensive white matter damage without an obvious increase in mortality rate at 12 weeks after the modified 2VO operation compared to RHRSP rats. RHRSP/modified 2VO rats showed more severe small venous collagen deposition than RHRSP rats, and the majority of the deposition was collagen I and IV rather than collagen III. In addition, RHRSP/modified 2VO rats possessed cognitive impairment, mild wall thickness and blood-brain barrier disruption. Our findings suggest that the modified WMLs model (RHRSP/modified 2VO) mimics cognitive impairment and small vessel pathological changes similar to WMLs in humans. Differentiating small veins from small arteries through smooth muscle is feasible, and marked small venous deposition may play an important role in the hypertensive white matter lesions.

In vivo Imaging of the Cerebral Endothelial Glycocalyx in Mice.

BACKGROUND/AIMS: Endothelial glycocalyx refers to the proteoglycan or glycoprotein layer of vessel walls and has critical physiological functions. Cerebral glycocalyx may have additional functions considering the blood-brain barrier and other features. However, the assessment of it has only been performed ex vivo, which includes processes presumably damaging the glycocalyx layer. Here we visualize and characterize the cerebral endothelial glycocalyx in vivo. METHODS: We visualized and quantified the cerebral endothelial glycocalyx in vivo under a 2-photon microscope by tagging glycocalyx and vessel lumen with wheat germ agglutinin lectin and dextran, respectively. The radial intensity was analyzed to measure the thickness of the cerebral endothelial glycocalyx in each vessel type. RESULTS: Cerebral arteries and capillaries have an intact endothelial glycocalyx, but veins and venules do not. The thickness of the glycocalyx layer in pial arteries, penetrating arteries, and capillaries was different; however, it was not correlated with the vessel diameter within each vessel type. CONCLUSION: We characterized the distribution of the cerebral endothelial glycocalyx in vivo. Compared to the results from ex vivo studies, the layer is thicker, indicating that the layer may be damaged in ex vivo systems. We also observed an inhomogeneous cerebral endothelial glycocalyx distribution that might reflect the functional heterogeneity of the vessel type.

Cordocytes-stem cells cooperation in the human brain with emphasis on pivotal role of cordocytes in perivascular areas of broken and thrombosed vessels.

This study is based on data analysis by light and transmission electron microscopy of the surgical cases in cerebral tumors, cerebrovascular malformations, thromboses in the carotid system, and other injuries such as perivascular hemorrhage. We examined cortical arteries and veins, perivascular areas with old hematic masses, vasculogenic foci, and broken large vessels. We identified, characterized, and compared both undifferentiated cells and well-differentiated cordocytes within periadventitial areas where these cells cooperate very well with precursor/stem cells to perform vital functions for cerebral vasculature with immediate effect on brain parenchyma. This useful cellular cooperation was observed by serial sections pointing out the main role of cordocytes during the entire process of collateral vessel formation after thrombosis and, respectively, in vascular wall repair after ruptures. This is the first cytohistopathological study which illustrates and explains some facets of cordocytes-stem cells cooperation around the vessels of human brain with emphasis on the fundamental role of cordocytes in response to vascular injuries. Our pioneering study will be completed for both basic science and modern medical care by further studies.

[The cerebral perivascular spaces: review of the literature on diffuse or focal expansion]. / Gli spazi perivascolari encefalici: revisione della letteratura sulle dilatazioni focali o diffuse.

Virchow-Robin spaces (VRS) are pial-lined, interstitial fluid-filled structures that do not directly communicate with the subarachnoid space, accompany penetrating arteries and veins and can be visualized on magnetic resonance imaging. This article reviews the imageology characteristics, the functions, the causes and the relation with neurological disorders of VRS.

[Enlargment of Virchow-Robin spaces in cranial trauma: literature review]. / La dilatazione degli spazi di Virchow-Robin nel trauma cranico. Aspetti caratteristici in RM e revisione della letteratura.

In this review it is discussed the role of mild traumatic brain injury as a cause of Virchow-Robin spaces (VRS) pathological enlargement. Anatomy and physiology of normal VRS, and their immunological role are described. Special attention is given to magnetic resonance imaging findings of both normal and enlarged perivascular spaces.

Junction between the great cerebral vein and the straight sinus: an anatomical, immunohistochemical, and ultrastructural study on 25 human brain cadaveric dissections.

The cerebral venous system is poorly understood, and best appreciated under macroscopic anatomical considerations. We present an anatomical and immunohistochemical studies to better define the morphological characteristics of the junction between the great cerebral vein and the straight sinus. Twenty-five cadaveric specimens from the anatomy laboratory of the University Victor Segalen of Bordeaux were studied. The observation of the venous junctions with the straight sinus was performed under an operating microscope. The smooth muscular actin immunohistochemical staining was performed for 18 veno-sinosal junctions. Five venous junctions were observed using an electron microscope. We observed 3 different anatomic aspects: type 1 was a junction with a small elevation in its floor and a posterior thickening (14 cases); type 2 was a junction with an outgrowth on the floor like a cornice (7 cases); and type 3 was a junction presenting a nodule. Microscopic study of type 1 and 2 junctions showed a positive coloration to orceine attesting the presence of elastic fibers. Immunohistochemistry revealed the presence of smooth muscular actin and S 100 protein attesting the presence of smooth muscular fibers and nervous fibers. We observed in the ultrastructural study, a morphological progression of the endothelium. The venous orifice of the great cerebral vein into the straight sinus could be anatomically assimilated as a true "sphincter." Its function in the regulation of the cerebral blood flow needs further exploration.

The sphenoparietal sinus of breschet: does it exist? An anatomic study.

BACKGROUND AND PURPOSE: The termination of the superficial middle cerebral vein is classically assimilated to the sphenoid portion of the sphenoparietal sinus. This notion has, however, been challenged in a sometimes confusing literature. The purpose of the present study was to evaluate the actual anatomic relationship existing between the sphenoparietal sinus and the superficial middle cerebral vein. METHODS: The cranial venous system of 15 nonfixed human specimens was evaluated by the corrosion cast technique (12 cases) and by classic anatomic dissection (three cases). Angiographic correlation was provided by use of the digital subtraction technique. RESULTS: The parietal portion of the sphenoparietal sinus was found to correspond to the parietal portion of the anterior branch of the middle meningeal veins. The sphenoid portion of the sphenoparietal sinus was found to be an independent venous sinus coursing under the lesser sphenoid wing, the sinus of the lesser sphenoid wing, which was connected medially to the cavernous sinus and laterally to the anterior middle meningeal veins. The superficial middle cerebral vein drained into a paracavernous sinus, a laterocavernous sinus, or a cavernous sinus but was never connected to the sphenoparietal sinus. All these venous structures were demonstrated angiographically. CONCLUSION: The sphenoparietal sinus corresponds to the artificial combination of two venous structures, the parietal portion of the anterior branch of the middle meningeal veins and a dural channel located under the lesser sphenoid wing, the sinus of the lesser sphenoid wing. The classic notion that the superficial middle cerebral vein drains into or is partially equivalent to the sphenoparietal sinus is erroneous. Our study showed these structures to be independent of each other; we found no instance in which the superficial middle cerebral vein was connected to the anterior branch of the middle meningeal veins or the sinus of the lesser sphenoid wing. The clinical implications of these anatomic findings are discussed in relation to dural arteriovenous fistulas in the region of the lesser sphenoid wing.

Vascularization of the brains of the Atlantic and Pacific hagfishes, Myxine glutinosa and Eptatretus stouti: a scanning electron microscope study of vascular corrosion casts.

The microvascularization of the brains of the hagfishes, Myxine glutinosa L. and Eptatretus stouti, were studied by scanning electron microscopy (SEM) of microvascular corrosion casts. Sections of these casts were used to determine the vascular territories of defined brain areas. Histological serial sections (10 microm) of the brains served for correlation of findings. Analysis of the microvascular casts of both species revealed that the blood supply to and from these brains arose ventrally and dorsally, respectively. Neither species possesses an arterial circle (Circulus Willisi) and both have similar microvascular patterns. The only difference between Myxine and Eptatretus was that the posterior cerebral artery in Myxine divides into mesencephalic and rhombencephalic branches, and in Eptatretus a third branch, termed telencephalic branch, arises from the posterior cerebral artery. 3D-morphometry revealed that luminal diameters of: 1) intracerebral arteries and arterioles range from 35.11 +/- 5.66 microm (mean +/- SEM) in the hypothalamus to 92.69 +/- 14.48 microm in the thalamus; 2) capillaries range from 17.8 +/- 0.44 microm in the olfactory bulb to 21.70 +/- 0.87 microm in the basal ganglia; and 3) intracerebral venules and veins range from 49.38 +/- 4.17 microm in the hypothalamus to 75.58 +/- 6.59 microm in the rhombencephalon. Interbranching distances of arteries and arterioles range from 179.19 +/- 11.32 microm in the optic tectum to 235.19 +/- 94.64 microm in the hypothalamus. Capillaries range from 91.07 +/- 6.22 microm in the hypothalamus to 116.15 +/- 9.45 microm in the thalamus, and venules and veins range from 137.30 +/- 18.11 microm in the hypothalamus to 189.83 +/- 17.47 microm in the optic tectum. Intervascular distances range from 70.58 +/- 3.58 microm in the olfactory bulb to 89.52 +/- 5.74 microm in the optic tectum. Branching angles of arteries and arterioles range from 38.39 +/- 10.9 degrees in the olfactory bulb to 100.73 +/- 9.4 degrees in the optic tectum, and the branching angles of capillaries range from 74.40 +/- 5.42 degrees in the optic tectum to 90.24 +/- 4.66 degrees in the olfactory bulb. Finally, the branching angles of the venules and veins range from 67.84 +/- 6.83 degrees in the tegmentum of the mesencephalon to 92.30 +/- 6.35 degrees in the optic tectum.

Scanning electron microscopy of arteriovenous malformations.

Although arteriovenous malformations (AVMs) have been known to have direct communications between arteries and veins without interposing capillaries, the exact location of arterial and venous junctions have not been defined. Utilizing microscopic and endoscopic observations, Yamada and associates identified shunting arterioles (50 mu-250 mu) directly connected to the AVM core vessels. While dissecting the AVMs in functional areas of the brain, shunting arterioles were sectioned to interrupt the arterial blood supply. This technique allowed cleavage formation between the core vessels and surrounding brain, thus avoiding brain tissue removal and preserving microcirculation to functionally critical brain. We demonstrate histologically for the first time by scanning electron microscopy shunting arterioles and communicating venules (20 mu-200 mu).

Nitric oxide is the predominant mediator for neurogenic vasodilation in porcine pial veins.

The innervation pattern and the vasomotor response of the potential transmitters in the porcine pial veins were investigated morphologically and pharmacologically. The porcine pial veins were more densely innervated by vasoactive intestinal polypeptide (VIP)- and neuropeptide Y-immunoreactive (I) fibers than were calcitonin gene-related peptide (CGRP)-I, choline acetyltransferase-I, Substance P (SP)-I, and NADPH diaphorase fibers. Serotonin (5-HT)-I fibers, which were not detected in normal control pial veins, were observed in isolated pial veins after incubation with 5-HT (1 microM). 5-HT-I fibers, however, were not observed when incubation with 5-HT was performed in the presence of guanethidine (1 microM), suggesting that 5-HT was taken up into the sympathetic nerves. In vitro tissue bath studies demonstrated that porcine pial veins in the presence of active muscle tone relaxed on applications of exogenous 5-HT, CGRP, SP, VIP, and sodium nitroprusside, whereas exogenous norepinephrine and neuropeptide Y induced only constrictions. Transmural nerve stimulation (TNS) did not elicit any response in pial veins in the absence of active muscle tone. However, in the presence of active muscle tone, pial veins relaxed exclusively on TNS. This tetrodotoxin-sensitive relaxation was not affected by receptor antagonists for VIP, CGRP, 5-HT, or SP but was blocked by L-glutamine (1 mM) and abolished by Nomega-nitro-L-arginine (10 microM) and Nomega-nitro-L-arginine methyl ester (10 microM). The inhibition by L-glutamine, Nomega-nitro-L-arginine, and Nomega-nitro-L-arginine methyl ester was reversed by L-arginine and L-citrulline but not by their D-enantiomers. These results demonstrate that the vasomotor effect of all potential transmitters except 5-HT in the pial veins examined resembles that in cerebral arteries. Although porcine pial veins receive vasodilator and constrictor nerves, a lack of constriction on TNS suggests that the dilator nerves that release nitric oxide may play a predominant role in regulating porcine pial venous tone.

Is the pial microvessel a good model for blood-brain barrier studies?

Pial microvessels have commonly been used as model systems for studying blood-brain barrier (BBB) properties instead of cerebral cortical microvessels. Since pial microvessels are relatively accessible they have been especially employed in electrophysiological and pharmacological studies. Measurements of electrical resistance across endothelial cells (EC) as a measure of their barrier properties have been made exclusively from pial microvessels in in vivo BBB studies. Similarly the observed responses of microvessels to the application of pharmacological agents have commonly been made on pial microvessels as representative of BBB vasculature. In this review the properties of pial and cerebral microvessels are compared to determine whether the use of the pial microvessel as a model for BBB studies is valid. Similarities are described in their ultrastructural features, permeability to electron dense tracers and molecular characteristics. Measurements of electrical resistance from pial microvessels are compared with measurements from cerebral EC monolayers in tissue culture and indirect determinations for cerebral microvessels in situ. Two notable differences between pial and cerebral microvessels are described in the adult nervous system. Tight junctions between cerebral EC appear to consist of a uniform population. In pial microvessels however tight junctions consist of two populations in one the inter-EC tight junctions resemble those between cerebral EC, with fusion of adjacent EC membranes. In the second population the inter-EC tight junctions differ with a discernible gap between adjacent EC membranes. The distribution of the endothelial barrier antigen (EBA) is uniform between EC of cerebral microvessels. By contrast EC of pial microvessels from a heterogeneous population for EBA expression which is related to the proximity of the EC to the astrocytic glia limitans. The role of astrocytes in the induction and maintenance of the BBB characteristics is briefly reviewed. The possible significance of the lack of an astrocytic ensheathment of pial microvessels is assessed. In summary, caution is urged in employing pial microvessels in BBB studies and the need for more information on possible pial microvessel heterogeneity is stressed.

The origin of amyloid in cerebral vessels of aged dogs.

Our morphometric study of 30 dogs, mongrels, from 6.5 to 26.5 years of age, shows amyloid angiopathy in cortical and leptomeningeal vessels of all dogs older than 13.2 years of age, and the increase in the numerical density of amyloid-positive vessels correlated with age. Cluster analysis distinguished the group of six dogs (25%) to be relatively less affected, a large group of 13 animals (54%) to have moderate pathology, and five dogs (21%) to have severe amyloid angiopathy. Amyloid accumulation starts in large vessels, particularly in the tunica media of large arteries. Amyloid deposition appears to be associated with smooth muscle cells. Ultrastructural studies of samples from nine dogs are in agreement with in vitro studies suggesting that smooth muscle cells are the source of soluble amyloid beta. beta-protein polymerizes in the basal lamina of the tunica media. Muscle cells in the area of amyloid-beta accumulation degenerate and die. Thioflavin-positivity of only 24% of cortical and 66% of leptomeningeal beta-protein-positive vessels suggests that thioflavin-negative deposits contain soluble, not yet fibrillized protein and/or partially degraded and depolymerized amyloid.

The peptidergic innervation of human coronary and cerebral vessels.

It is now well established that in addition to nerves containing classical transmitters, the mammalian vascular system is also supplied by nerve fibre subpopulations containing several vasoactive peptides. The precise function of these peptides (neuropeptide Y, calcitonin gene-related peptide, vasoactive intestinal polypeptide, somatostatin and the tachykinins) is still unknown, however, their widespread occurrence in perivascular nerves indicates that they are likely candidates for a role in the neurogenic regulation of the vascular system. It has been suggested that they may exert a direct vasomotor action via their own receptors and/or modulate the release and action of other vascular transmitters. Recently, several studies have focused on the supply of nerve fibres storing neuropeptides in the coronary and cerebral vasculature of laboratory animals, however, little is known on the distribution of these putative transmitters in human coronary and cerebral vessels. In this paper, the immunocytochemical evidence that several neuropeptides are localized in subpopulations of afferent and efferent nerve fibres supplying the human coronary and cerebral vasculature is focused.

Early ultrastructural changes in blood-brain barrier vessels of the rat embryo.

The blood-brain barrier (BBB) in fetal rat brain has been shown by others to be more permeable to a variety of blood-borne solutes than the BBB in adults. We used ultrastructural morphometric methods to measured the density of putative vascular pores between the ages of embryonic day (E) 11 and birth to determine the structural basis for this relatively high permeability. We found that fenestrations, that are frequent at E11, declined rapidly and were last seen at E13 in intraparenchymal vessels and at E17 in pial vessels. Interendothelial junctions in fetal brain contained expanded clefts suggestive of paracellular channels at all ages examined, although they disappear after birth. Both of these features likely contribute to high fetal BBB permeability, but endothelial vesicles probably do not. The central nervous system is vascularized by ingrowth of capillary sprouts from the perineurial vascular plexus. Invading capillaries express BBB features in response to inductive signals from the surrounding neural tissue. We compared early ultrastructural changes in perineurial vessels, which are separated from neural tissue by a sizeable perivascular space, with those in intraneural vessels, which are totally enveloped by neural tissue, to determine whether the inductive interaction requires close cellular contact. For the most part, the perineurial and intraneural vessels matured in parallel. Furthermore, cerebellar vessels developed in parallel with cerebral vessels, even though they did not invade neural tissue until a comparatively late stage. These results suggest that intimate contact between neural tissue and vessel walls is not a requirement for BBB expression.

Aberrations in cerebral vascular functions due to Plasmodium yoelii nigeriensis infection in mice.

Plasmodium yoelii nigeriensis infection in mice caused an increase in uptake of 125I-labeled bovine serum albumin, 51Cr-labeled erythrocytes and Evans blue dye from peripheral circulation into the brain. Isolated cerebral microvessels which were characterized in terms of their morphology under scanning electron microscope and enhancement of the specific activities of biochemical markers, viz. alkaline phosphatase, gamma-glutamyl transpeptidase, and monoamine oxidase, showed significant decrease in these activities due to P. yoelii nigeriensis infection. On the other hand, relatively minor (statistically insignificant) changes occurred in the first two enzyme specific activities in the cerebral cortex and monoamine oxidase registered an increase in this tissue due to infection. Histological examination of the cerebral tissue of infected animals by light and electron microscopy showed broken blood vessel walls and leakage of erythrocytes into extravascular space, some of which contained intraerythrocytic malarial parasite in a state of cell division.

Venous microvasculature of the pineal body and choroid plexus in the rat.

Rats were perfused from aorta with Ringer's solution and with Karnovsky's fixative, and injected from the right atrium with Mercox resin. Specimens were properly taken, observed under LM, TEM, SEM, and stereo-photographed. Fenestrated endothelial cells of the pineal capillary were observed to contain plenty of microtubules running parallel to the long axis of the vessel. The endothelial basal lamina appeared anchored by microfibrils onto fine collagen fibrils and onto the basal lamina of the perivascular cells. These findings indicate resistance of the capillary endothelium against compression by perivascular hydrostatic pressure, and strongly suggest that the perivascular cerebrospinal fluid is absorbed into the capillary lumen. Resin cast of the pineal venous system looked like a glomerule with two longer collecting veins on the ventral surface and with shorter ones on the dorsal side all emptying into the straight sinus and the confluence of sinuses. Resin cast of the choroid plexus venous system tended to lobulate and looked like a vine with a spiral collecting vein emptying into the great vein. Two veins were found connecting the great vein to the confluence and to the transverse sinuses. These could be effective collaterals in case of occlusion of Galen's vein.

Vasculature of the paraphysis cerebri of the frog.

The paraphysis cerebri is a glandular structure found in the third ventricle of lower vertebrates. It is well-developed in amphibians and reptiles. The function of the gland is not substantiated, but may play a role in calcium metabolism. To further elucidate its possible endocrine role, the paraphyseal vasculature was examined using casting techniques as well as transmission electron microscopy. Perfusion-fixed paraphyses of Rana pipiens and Rana catesbeiana were either: a) cast with Microfil (with subsequent dehydration, clearing, and macroscopic examination) or Batson's compound (followed by tissue digestion and examination by scanning electron microscopy); or b) processed for transmission electron microscopy. The paraphyseal capillary bed consists of a sinusoidal portal system which receives afferent blood from its associated choroid plexus. The choroid plexus of the third ventricle receives its blood supply via arterioles from the posterior telencephalic artery. These arterioles traverse in the periphery of the paraphysis to branch and supply the choroid plexus. Numerous venules exit from the choroid plexus and drain into sinusoids of the paraphysis. The sinusoid venules appear to empty into a midline venous structure which passes tangentially through the paraphysis. The sinusoids consist of fenestrated endothelium which is indicative of transport vessels. Nerve fibers were observed in the paraphysis, however, histofluorescence revealed no monoaminergic sympathetic innervation of the paraphyseal vasculature.

Scanning electron microscopic study of periendothelial cells of the rat cerebral vessels revealed by a combined method of corrosion casting and KOH digestion.

A combined method of corrosion casting and KOH digestion was used for the scanning electron microscopic (SEM) observation of the external walls of the blood vessels running on the surface and in the parenchyma of the rat cerebral cortex. The plastic resin which was preliminarily injected into the cerebral vessels reinforced the preservation of the vessel walls during the chemical digestion. Using the present method, the walls of blood vessels running on the surface and in the parenchyma of the cerebral cortex were clearly seen with SEM. The arterioles showed a rich arrangement of smooth muscle cells embracing the external surface of the endothelia, and the venules had sparse, flat smooth muscle cells. The capillaries were surrounded by many pericytes which were extremely similar to those seen in the other organs of the animals. There were observed two different kinds of pericyte. One type of the pericytes in the present specimens extended very long cytoplasmic processes parallel to the long axis of the capillaries. Another type of the pericytes had relatively short, broad and flat cytoplasmic processes radiated to embrace the capillary walls. The former was frequently seen in the capillaries on brain surface and in the parenchyma, whereas the latter was relatively rare. The availability of the present method was also discussed in comparison with the previous descriptions.

On the haemodynamics of arteriovenous fistulas in the rat.

Direct and indirect arteriovenous fistulas were applied to the cervical vessels of 19 rats in order to study the haemodynamic parameters of angioma-like, rapid blood flow in small vessels. Flow was measured electromagnetically and intraoperatively using the Doppler sonography, and both methods were compared. Resultant alterations in vessel walls were examined under the electron microscope. Following fistula application, the flow rates increased by a factor of ten. At the same time, the flow pattern profile and stream resistance also changed. At present, the Doppler sonography device employed here is the only one commercially available, yet it could not detect rapid flow rates (greater than 85 cm/sec). The abnormal haemodynamic strain on the venous walls led to morphologically and angiographically detectable alterations.

Ultrastructure of the meninges at the site of penetration of veins through the dura mater, with particular reference to Pacchionian granulations. Investigations in the rat and two species of New-World monkeys (Cebus apella, Callitrix jacchus).

At the sites where a vein penetrates through the dura mater, two aspects deserve particular attention: (i) The delineation of the perivascular cleft, a space belonging to the interstitial cerebrospinal fluid (CSF) compartment, toward the interior hemal milieu of the dura mater. (ii) The relationship between the perivascular arachnoid layer and the subdural neurothelium at the point of vascular penetration. These problems were investigated in the rat and in two species of New-World monkeys (Cebus apella, Callitrix jacchus). Concerning the first aspect, tight appositions of meningeal cells to the vessel wall, the basal lamina of which is widened and enriched with microfibrils, prevent communication between the interstitial CSF in the perivascular cleft and the hemal milieu in the dura mater. With reference to the second aspect, the perivascular arachnoid cells are transformed into neurothelial cells at the point where they become exposed to the hemal milieu of the dura mater and subsequently continuous with the subdural neurothelium. Leptomeningeal protrusions encompassing outer CSF space can penetrate into the dura mater. These protrusions may expand and branch repeatedly, forming along the wall of the dural sinus Pacchionian granulations. At these sites, however, the structural integrity of the sinus wall and the Pacchionian granulation is not lost. Numerous vesiculations not only in the sinus and vascular walls, but also in the cellular arrays of the Pacchionian granulations or paravascular leptomeningeal protrusions indicate mechanisms of transcellular fluid transport. Moreover, the texture of the leptomeningeal protrusions favors an additional function of these structures as a "volume" buffer.