Chapter 32:  Blood—Brain—Cerebrospinal Fluid Barriers

Physical and biological processes determine molecular movement across membranes of the blood—brain— cerebrospinal fluid barriers

The processes that determine molecular movement across membranes are diffusion, pinocytosis, carrier-mediated transport and transcellular transport [5]. The types of carrier-mediated transport are described in Chapter 5.

Diffusion is the process by which molecules in solution move from an area of higher to lower concentration. With this type of transport, the net rate of solute flux is directly proportional to the difference in concentration between the two areas. In biological systems, this process is an important mechanism for the movement of molecules within a fluid compartment; however, diffusion across a lipid membrane, such as the cell membranes of the blood—brain barrier, is possible only when the solute is lipid-soluble or when the membrane contains specialized channels. Diffusion is the primary mechanism for blood—brain exchange of respiratory gases and other highly lipid-soluble compounds.

Pinocytosis is a process by which extracellular fluid is engulfed by invaginating cell membranes, forming a vesicle that then separates from the membrane. This vesicle may move through the cell cytoplasm and release its contents on the other side of the cell layer by means of exocytosis. Under normal conditions, pinocytosis is thought to contribute little to the transport of solutes across the blood—brain barrier. Instead, the few vesicles that are observed within brain capillary endothelial cells are probably destined to fuse with lysosomes. Nevertheless, some proteins may traverse the brain endothelial cell through a process that has been called absorptive-mediated transcytosis [6].

Transcellular transport across a layer of cells requires the presence of carrier or channel molecules on the luminal and antiluminal sides of the cells. Facilitated and active transport are defined in Chapter 5. In transcellular facilitated diffusion, the carriers on opposite sides of the cell are usually similar and solutes are not moved against concentration gradients. Active transport across a cell layer, however, requires a special arrangement of transport proteins within the plasma membranes. The active transport system is found on only one side of the cell and usually is associated with a nonactive transport system on the other side of the cell. With this arrangement, a solute accumulates within the cell by active transport through one membrane and subsequently leaves the cell by a channel or facilitated transport process through the opposite membrane. When plasma membranes of two surfaces of a cell have different properties, that cell is said to be polar. Cellular polarity underlies active transcellular transport and secretion of fluid by epithelial cells in the choroid plexus.

When fluid is secreted at one site and absorbed at another, there is bulk flow of fluid. This means that solutes of various sizes move together with the solvent as a bulk liquid. This process is important in the circulation and absorption of CSF, which is secreted by the choroid plexus, circulated through the ventricular and subarachnoid spaces and absorbed through arachnoid villi into the bloodstream.

Transport processes combine to provide stability for constituents of cerebrospinal fluid and brain extracellular fluid

Bradbury and Stulcova [7] defined stability of the blood—CSF systems as follows. If a substance is present in CSF at concentration C_CSF and in plasma at concentration C_pl, stability occurs when, as a result of a change in plasma concentration, a new steady state is reached so that

At steady state, the flux of this substance from plasma to CSF, J_in, must equal its flux out, J_out, so that for any change in plasma concentration, ΔC_pl stability of CSF will occur when

where J_in and J_out represent transport processes that need not be identical. For instance, one process might be passive and the other active. If the carrier involved in J_in is saturated at the usual plasma concentration, then the ratio ΔJ_in/ΔC_pl will approach zero. Such carrier-mediated transport is probably the most common mechanism controlling the flow of water-soluble substances from the capillary lumen to the brain, but carrier systems also have been found to operate for outward flux. Here, the greatest stability is achieved when the carrier system operates well below saturation so that the ratio ΔJ_out/ΔC_CSF is a positive number. Such asymmetrical carrier mechanisms have been implicated in the maintenance of a stable K⁺ concentration in CSF and may exist for some amino acids and organic acids.

Endothelial cells in brain capillaries are the site of the blood—brain barrier

Studies of Reese and Karnovsky and Brightman and Reese demonstrated that brain endothelial cells differ from endothelial cells in capillaries of other organs in two important ways [8]. First, continuous tight junctions are present between the endothelial cells, which prevent transcapillary movement of polar molecules varying in size from proteins to ions. Second, there are no detectable transendothelial pathways. Thus, there is an absence of transcellular channels and fenestrations as well as a paucity of plasmalemmal and intracellular vesicles. As a result of these special anatomical features, the endothelial cells in brain provide a continuous cellular barrier between the blood and the interstitial fluid (Fig. 32-1). Not all areas of the brain contain capillaries that produce a barrier. In these nonbarrier regions, the morphological features of the capillaries are similar to those of systemic microvascular beds. Thus, the tight junctions are discontinuous, there are more plasmalemmal vesicles and some endothelial cells even exhibit fenestrations. Table 32-2 lists the brain regions that contain capillaries of this type. The absence of a blood—brain barrier in many of these regions may relate to their feedback role in the regulation of peptide hormone release.

Surrounding the capillary endothelial cell is a collagen-containing extracellular matrix. Embedded within this basement membrane are contractile pericytes. These cells may regulate endothelial cell proliferation and, under certain pathological conditions, take on phagocytic functions. Almost the entire outer surface of the basement membrane is covered with foot processes from astrocytes. This close association suggests an interaction between astrocytes and endothelial cells that is important for the function of the blood—brain barrier. Support for this hypothesis is found in brain tumors and nonbarrier regions of the brain, where the absence of intimate astrocyte—endothelial cell contact is associated with the absence of a blood—brain barrier. More direct evidence comes from cell culture studies in which astrocytes or their secreted cell products induce endothelial cells to differentiate morphogenically into vascular structures [9] and to increase their interendothelial tight junctional complexes [10].

Substances with a high lipid solubility may move across the blood—brain barrier by simple diffusion

Diffusion is the major entry mechanism for most psychoactive drugs. As shown in Figure 32-2, the rate of entry of compounds that diffuse into the brain depends on their lipid solubility, as estimated by oil/water partition coefficients. For example, the permeability of very lipid-soluble compounds, such as ethanol, nicotine, iodoantipyrine and diazepam, is so high that they are extracted completely from the blood during a single passage through the brain. Hence, their uptake by the brain is limited only by blood flow, and this provides the basis for use of iodoantipyrine to measure cerebral blood flow rate. In contrast, polar molecules, such as glycine and catecholamines, enter the brain only slowly, thereby isolating the brain from neurotransmitters in the plasma. Uptake in the brain of some compounds, such as phenobarbital and phenytoin, is lower than predicted from their lipid solubility as a result of binding to plasma proteins.

Water readily enters the brain by diffusion. Using intravenously administered deuterium oxide as a tracer, the measured half-time of exchange of brain water varies between 12 and 25 sec, depending on the vascularity of the region studied. Although this rate of exchange is rapid compared with the rate of exchange of most solutes, it is limited both by the permeability of the capillary endothelium and by the rate of cerebral blood flow. In fact, the calculated permeability constant of the cerebral capillary wall to the diffusion of water is about the same as that estimated for its diffusion across lipid membranes (Fig. 32-2).

As a consequence of its high permeability, water moves freely into or out of the brain as the osmolality of the plasma changes. This phenomenon is clinically useful since the intravenous administration of poorly permeable compounds such as mannitol (Fig. 32-2) will osmotically dehydrate the brain and reduce intracranial pressure. For example, when plasma osmolality is raised from 310 to 344 mOsm, a 10% shrinkage of the brain will result, with half of the shrinkage taking place in 12 min.

Gases, such as CO₂, O₂, N₂O and Xe, and volatile anesthetics diffuse rapidly into the brain. As a consequence, the rate at which their concentration in the brain comes into equilibrium with the plasma is limited primarily by the cerebral blood flow rate. Hence, the inert gases, such as N₂O and Xe, can be used to measure cerebral blood flow. An interesting contrast is found between CO₂ and H⁺ with regard to their effects on brain pH. Since the blood—brain barrier permeability of CO₂ greatly exceeds that of H⁺, the pH of the brain interstitial fluid will reflect blood pCO₂ rather than blood pH. Consequently, in a patient with a metabolic acidosis and a compensatory respiratory alkalosis, the brain is alkalotic.

Carrier-mediated transport enables molecules with low lipid solubility to traverse the blood—brain barrier

Although d-glucose and l-glucose are stereoisomers, extraction of d-glucose in the brain is more than 100-fold greater than that of l-glucose (Fig. 32-2). This apparently anomalous relationship also is observed for other metabolically essential compounds (Table 32-3). The high permeability of these polar compounds is mediated by specific transport proteins (see Chap. 5) in the plasma membranes of the endothelial cells (Fig. 32-1).

Glucose is the primary energy substrate of the brain, and its metabolism accounts for nearly all of the oxygen consumption in the brain (see Chap. 31). Since entry of glucose into the brain is critical, mechanisms for glucose transport across the blood—brain barrier have been studied particularly well [11]. Stereospecific, but insulin-independent, GLUT-1 glucose transporters are highly enriched in brain capillary endothelial cells (Fig. 32-3) and mediate the facilitated diffusion of this polar substrate through the blood—brain barrier [12]. The activity of these transporters is sufficient to transport two to three times more glucose than normally is metabolized by the brain.

Diminished blood—brain barrier GLUT-1 expression that is substantial enough to compromise energy substrate delivery to the brain has been found in a series of children with seizures, mental retardation, compromised brain development and low CSF glucose concentrations. GLUT-1 protein concentrations in red blood cells obtained from these patients is approximately 50% of normal, consistent with a heterozygous GLUT-1 genomic mutation. Interestingly, the seizures in these patients are particularly responsive to therapy with a ketogenic diet that replaces glucose with ketone bodies as the principal energy substrate for the brain. This is the first recognized example of disease attributed to a blood—brain barrier transport defect.

The stereospecificity of the glucose-transport system permits d-glucose, but not l-glucose, to enter the brain. Hexoses, such as mannose and maltose, also are transported rapidly into the brain; the uptake of galactose is intermediate, whereas fructose is taken up very slowly. 2-Deoxyglucose is taken up quickly and will competitively inhibit the transport of glucose. Once within neurons and glia, 2-deoxyglucose is phosphorylated but not metabolized further. If 2-deoxyglucose is used in tracer quantities, the amount of the phosphorylated tracer in the brain reflects the rate of glucose metabolism (see Chap. 31).

Monocarboxylic acids, including l-lactate, acetate, pyruvate and ketone bodies, are transported by a separate stereospecific system [13]. The rate of entry of these substances is significantly lower than that of glucose; however, they are important metabolic substrates in neonates and during starvation. The rate and capacity of monocarboxylic acid transport are elevated substantially in suckling neonates, consistent with the unique metabolic requirements of the developing brain and the elevated concentrations of monocarboxylic acids in breast milk.

Neutral l-amino acids have various rates of movement into the brain [13,14]. Phenylalanine, leucine, tyrosine, isoleucine, valine, tryptophan, methionine, histidine and l-dihydroxy- phenylalanine (l-DOPA) may enter as rapidly as glucose. These essential amino acids cannot be synthesized by the brain and, therefore, must be supplied from protein breakdown and diet (see Chap. 33). Several are precursors for neurotransmitters synthesized in the brain (see Chaps. 12–14). Transport of these large neutral amino acids into the brain is inhibited by the synthetic amino acid 2-aminonorbornane-2-carboxylic acid (BCH) but not by 2-(methylamino)-isobutyric acid (MeAIB); hence, the transport system in the blood—brain barrier is similar to the leucine-preferring (L) transport system defined by Christensen (see [5]). Since a single type of transport carrier mediates the transcapillary movement of structurally related amino acids, these compounds compete with each other for entry into the brain. Therefore, an elevation in the plasma concentration of one will inhibit uptake of the others (Fig. 32-4). This may be important in certain metabolic diseases, such as phenylketonuria (PKU), where high concentrations of phenylalanine in plasma reduce brain uptake of other essential amino acids (see Chap. 44).

Small neutral amino acids, such as alanine, glycine, proline and γ-aminobutyric acid (GABA), are markedly restricted in their entry into the brain (for example, glycine in Fig. 32-2). These amino acids are synthesized by the brain, and several are putative neurotransmitters (see Chaps. 16 and 18). This restriction is consistent with the inability of MeAIB to inhibit brain uptake of amino acids and suggests that the alanine-preferring (A) transport system [5] is not present on the luminal surface of the blood—brain barrier. In contrast, these small neutral amino acids appear to be transported out of the brain across the blood—brain barrier, suggesting that the A-system carrier is present on the antiluminal surface of the brain capillary [15]. This may explain why the CSF: plasma concentration ratios are particularly low for these amino acids (Table 32-1). Thus, essential amino acids that serve as precursors for catecholamine and indoleamine synthesis are readily transported into the brain (L system), whereas amino acids synthesized by the brain, including those amino acids that act as neurotransmitters, not only are limited in their entry but also are transported actively out of the brain (A system).

In addition, there are distinct transport systems that facilitate the brain uptake of basic, acidic and β-amino acids (Table 32-3). Lysine and arginine are essential amino acids and, therefore, must be provided from the blood. The acidic amino acids glutamate and aspartate are both important metabolic intermediates as well as neurotransmitters (see Chap. 15). While the brain content of these amino acids is maintained primarily by de novo synthesis, they also can be transported into the brain at a slow rate across the blood—brain barrier. The β-amino acid taurine is present at high concentrations in the brain and is involved in volume regulation.

Choline enters the central nervous system through a carrier-mediated transport process that can be inhibited by molecules such as dimethyl aminoethanol, hemicholinium and tetraethyl ammonium chloride. Since choline cannot be synthesized by the brain, it has been proposed that blood—brain barrier transport may regulate the formation of acetylcholine in the central nervous system [16].

Vitamins are, by definition, substances that cannot be synthesized by mammalian organisms but are required in small amounts to support normal metabolism. Since vitamins cannot be synthesized by the brain, they must be obtained from the blood. Thus, specific transport systems are present in the blood—brain barrier for most vitamins (Table 32-3) [17]. These transport systems generally have a low capacity since the brain requires only small amounts of the vitamins and efficient homeostatic mechanisms preserve brain vitamin content without the need for a rapid influx from the blood. Nevertheless, dietary deficiency of some vitamins can produce neurological disease (see Chaps. 33 and 44).

Metal ions are exchanged between plasma and brain very slowly compared with other tissues

Intravenously administered ⁴²K⁺, for example, exchanges with muscle K⁺ in 1 hr, but K⁺ exchange in brain is only half completed in 24 to 36 hr. Na⁺ exchange is somewhat faster, with half-exchange into brain occurring in 3 to 8 hr. Despite its relatively slow entry into the brain, Na⁺ exchange across the blood—brain barrier appears to occur by mediated transport [18]. This occurs, in part, through brain capillary Na,KATPase (see Chap. 5), which is located primarily on the antiluminal membrane of the endothelial cell (Fig. 32-1). Na,K-ATPase in the brain capillary also may mediate removal of interstitial fluid K⁺ from the brain and thereby may maintain a constant brain K⁺ concentration in the face of fluctuating plasma concentrations. In addition, the antiluminal location of Na,K-ATPase may underlie the proposed role of the brain capillary in secretion of interstitial fluid, an extrachoroidal source of CSF.

Some proteins cross the blood—brain barrier by binding to receptors or by absorption on the endothelial cell membrane

Receptor-mediated transcytosis. Most proteins in the plasma are not able to cross the blood—brain barrier because of their size and hydrophilicity. Consequently, concentrations of plasma proteins in the brain are very low (Table 32-1). However, concentrations of certain proteins, such as insulin and transferrin, vary as the plasma concentrations change, and uptake of these peptides in the brain is greater than expected based on their size and lipid solubility. Furthermore, the brain uptake of some proteins is saturable. These properties suggest the presence of a specific transport process. It is now believed that proteins such as insulin, transferrin, insulin-like growth factors and vasopressin cross the blood—brain barrier by a process called receptor-mediated transcytosis [6]. The brain capillary endothelial cell is highly enriched in receptors for these proteins, and following binding of protein to the receptor, a portion of the membrane containing the protein—receptor complex is endocytosed into the endothelial cell to form a vesicle. Although the subsequent route of passage of the protein through the endothelial cell is not known, there is eventual release of intact protein on the other side of the endothelial cell.

Absorptive-mediated transcytosis. Polycationic proteins and lectins cross the blood—brain barrier by a similar but nonspecific process called absorptive-mediated transcytosis [5]. Rather than binding to specific receptors in the membrane, these proteins absorb to the endothelial cell membrane based on charge or affinity for sugar moieties of membrane glycoproteins. The subsequent transcytotic events are probably similar to receptor-mediated transcytosis; however, the overall capacity of absorptive-mediated transcytosis is greater because it is not limited by the number of receptors present in the membrane. Thus, cationization may provide a mechanism for enhancing brain uptake of almost any protein.

Metabolic processes within the brain capillary endothelial cells are important to blood—brain barrier function

Most neurotransmitters present in the blood do not enter the brain because of their low lipid solubility and lack of specific transport carriers in the luminal membrane of the capillary endothelial cell. This is illustrated for dopamine in Figure 32-2. In contrast, l-DOPA, the precursor for dopamine, has affinity for the large neutral amino acid-transport system and more easily enters the brain from the blood than would be predicted by its lipid solubility (Figs. 32-1 and 32-2). This is why patients with Parkinson's disease are treated with l-DOPA rather than with dopamine (see Chap. 45); however, the penetration of l-DOPA into the brain is limited by the presence of the enzymes l-DOPA decarboxylase and monoamine oxidase within the capillary endothelial cell [19]. This “enzymatic blood—brain barrier” limits transendothelial passage of l-DOPA into the brain and explains the need for large doses of l-DOPA in the treatment of Parkinson's disease. Therapy is enhanced by concurrent treatment with an inhibitor of peripheral l-DOPA decarboxylase.

Intracapillary monoamine oxidase also may play a role in the inactivation of neurotransmitters released by neuronal activity since monoamines are accumulated actively and metabolized by brain capillaries [19]. The fact that monoamines show very little uptake when presented from the luminal side suggests that the uptake systems are present only on the antiluminal membrane of the brain capillary endothelial cell (Fig. 32-1).

The brain capillary contains a variety of other neurotransmitter-metabolizing enzymes, such as cholinesterases, GABA transaminase, aminopeptidases and endopeptidases. In addition, several drug- and toxin-metabolizing enzymes typically found in the liver are also found in brain capillaries [20]. Thus, the “enzymatic blood—brain barrier” protects the brain not only from circulating neurotransmitters but also from many toxins.

Carrier-mediated blood—brain barrier transport protects the brain from blood-borne neurotoxins and drugs

P-glycoproteins are transmembranous, ATP-dependent pumps originally discovered for their ability to confer multidrug resistance to neoplastic cells. P-glycoproteins typically are expressed by epithelial barriers and blood—brain barrier endothelial cells. Humans have only one drug-transporting P-glycoprotein, which is the product of the MDR1 gene. Mice express two drug-transporting P-glycoproteins, mdr1a and mdr1b, the former of which is expressed by the blood—brain barrier. Blood—brain barrier P-glycoprotein is absent in mdr1a (—/—) transgenic mice, which also display substantially enhanced entry into the brain of systemically administered P-glycoprotein substrates, such as ivernectin, vinblastine, cyclosporine A, domperidone and digoxin [21]. Thus, P-glycoprotein can efficiently limit the blood—brain barrier permeability of hydrophobic P-glycoprotein substrates by pumping them from barrier endothelial cells back to the blood.

The blood—brain barrier undergoes development

The development of the cerebral microvasculature and the morphological changes attributed to its expression of the blood—brain barrier have been described in detail by Bar and Wolff and Stewart and colleagues [22]. The brain blood vessels that are destined to express the blood—brain barrier are derived from endothelial cells that originate from a plexus of nonbarrier vessels on the surface of the developing brain. The endothelial cells of these primitive vessels are fenestrated and surrounded by a poorly organized extracellular matrix and relatively undifferentiated mesenchymal cells. As these primitive vessels migrate into the brain, their morphology and function take on blood—brain barrier properties. This includes the ensheathment of endothelial cells by astroglial foot processes, their association with pericytes, the formation of a well-defined periendothelial basement membrane, the rapid loss of fenestrations and the more gradual maturation of interendothelial tight junctional complexes. This process involves the loss of large interendothelial junctional clefts that provide a paracellular diffusion pathway in immature cerebral vessels. In the rat, the loss of junctional clefts best coincides with the 30-fold diminution of paracellular diffusion as vessels achieve their most mature level of differentiation.

Developmental changes also occur in the endothelial cell carrier transport systems that selectively deliver nutrients from blood to the developing brain. Brain uptake of substrates such as β-hydroxybutyrate, tryptophan, adenine and choline is substantially higher in neonates relative to adults. The converse is true for glucose. These changes reflect the relative requirements of the developing and adult brain for specific energy and macromolecular substrates.

These developmental observations taken together with in vivo and in vitro experimental findings support the hypothesis that signals derived from brain parenchyma, more specifically perivascular astrocytes, induce endothelial cell expression of the blood—brain barrier phenotype. Transplantation studies indicate that endothelial cell expression of the blood—brain barrier is not intrinsic to brain endothelial cells since systemic endothelial cells can be induced to express barrier properties by transplanted brain tissue. A requirement for astroglial—endothelial cell interactions is supported by observations that brain endothelial cells continue to express blood—brain barrier proteins when surrounded by tumor cells of astroglial origin in the absence of neurons and stop expressing barrier proteins if surrounded by tumor cells of epithelial origin, such as in carcinoma cells that have metastasized to brain. Furthermore, more reductionist experiments in vitro show that brain endothelial cells lose all barrier properties when isolated from the brain and cultured under routine conditions but express certain barrier properties if cultured in the presence of astrocytes or their conditioned medium.

The choroid plexus epithelial cells and the arachnoid membrane form the blood—cerebrospinal fluid barrier

The choroid plexus is a vascular tissue found in all cerebral ventricles (Fig. 32-5). The functional unit of the choroid plexus, composed of a capillary enveloped by a layer of differentiated ependymal epithelium, is shown in Figure 32-6. Unlike the capillaries that form the blood—brain barrier, choroid plexus capillaries are fenestrated and have no tight junctions. The endothelium, therefore, does not form a barrier to the movement of small molecules. Instead, the blood—CSF barrier at the choroid plexus is formed by the epithelial cells and the tight junctions that link them. The other part of the blood—CSF barrier is the arachnoid membrane, which envelops the brain. The cells of this membrane also are linked by tight junctions.

Cerebrospinal fluid is secreted primarily by the choroid plexus

The major site of CSF formation is the choroid plexus, and from a morphological viewpoint, the epithelial cells of this tissue are similar to other secretory cells. There is also some extrachoroidal secretion, which may result from ion transport by brain capillaries, as discussed above. In humans, the rate of CSF secretion is 0.3 to 0.4 ml/min, about one-third the rate at which urine is formed. The total volume of CSF is estimated to be 100 to 150 ml in normal adults, such that CSF is replaced totally three or four times each day. Several constituents are maintained at concentrations different in CSF from those in plasma (Table 32-1), indicating that CSF is not simply a protein-free ultrafiltrate of plasma. Instead, CSF production by the choroid plexus is driven by active ion transport that results in a net secretion of Na⁺ and Cl⁻, the main ionic constituents of CSF. The exact mechanisms involved have yet to be determined fully, but Figure 32-7 represents one model.

In contrast to most epithelia, Na,K-ATPase is found on the apical, or CSF-facing, microvilli of the choroid plexus [23]. Ouabain, an inhibitor of Na,K-ATPase, reduces CSF secretion. Na,K-ATPase is probably the main transporter of Na⁺ from the epithelium to the CSF. It also provides the electrochemical gradient for basolateral, or blood-facing, Na⁺ entry into the epithelium, which probably occurs via a Na⁺/H⁺ antiport system (see Chap. 5).

Cl⁻ influx into the epithelium is via a Cl⁻/HCO⁻₃ exchanger on the basolateral membrane [24]. This exchanger can be inhibited directly with stilbenes or indirectly using acetazolamide, an inhibitor of carbonic anhydrase which reduces the intracellular production of HCO⁻₃. Cl⁻ efflux from the epithelium to the CSF is primarily via a cotransporter, which is either of the K⁺/Cl⁻ or Na⁺/K⁺/Cl⁻ type [25]. This cotransporter can be inhibited by furosemide and bumetanide. Therapeutically, acetazolamide and furosemide are used to decrease the rate of CSF formation in hydrocephalus. Acetazolamide is generally a more effective agent at decreasing CSF production. This may reflect the involvement of a cotransporter in moving ions from CSF to the epithelium (Fig. 32-7).

The choroid plexus receives a number of different forms of innervation, most notably a sympathetic input from the superior cervical ganglia. It also has many hormone receptors [26]. For example, the choroid plexus epithelium has a tenfold greater density of 5-hydroxytryptamine (5-HT)_2C receptors than any other brain tissue, although it does not appear to receive direct serotonergic innervation. A number of these neuroendocrine mechanisms modify choroid plexus blood flow or solute transport by the epithelium, indicating their potential role in controlling CSF secretion rate or composition.

Cerebrospinal fluid circulates through the ventricles, over the surface of the brain, and is absorbed at the arachnoid villi and at the cranial and spinal nerve root sheaths

The CSF circulation is from the lateral ventricles through the foramina of Monro into the third ventricle, the aqueduct of Sylvius, and then into the fourth ventricle. The fluid passes from the fourth ventricle through the foramina of Luschka and Magendie to the cisterna magna and then circulates into the cerebral and spinal subarachnoid spaces (Fig. 32-5).

There is evidence that absorption of CSF by the arachnoid villi occurs by a valve-like process, permitting the one-way flow of CSF from the subarachnoid spaces into the venous sinuses. CSF absorption does not occur until CSF pressure exceeds the pressure within the sinuses. Once this threshold is reached, the rate of absorption is proportional to the difference between CSF and sinus pressures. A normal human can absorb CSF at a rate up to six times the normal rate of CSF formation with only a moderate increase in intracranial pressure.

If obstructions occur at the foramina between the ventricles, the ventricle upstream from the obstruction will enlarge, producing obstructive hydrocephalus. Occasionally, disease processes affect CSF removal. For example, obliteration of the subarachnoid space by inflammation or thrombosis of the sinuses will prevent clearance of fluid. When this occurs, CSF pressure increases and hydrocephalus develops without obstruction of the ventricular foramina. This is called communicating hydrocephalus.

The choroid plexus is the major route of blood—brain barrier exchange for some compounds

The movement of substances from the blood into the CSF is, in many ways, analogous to that from the blood into the brain, with many of the same transporters present in both tissues. Quantitatively, however, there are major differences between transport by the choroid plexus and by the blood—brain barrier. Thus, in terms of O₂, CO₂, glucose and amino acid entry into the brain, the blood—brain barrier predominates. For some other compounds, however, the choroid plexus is the major site of entry. This is the case for Ca²⁺, where the CSF influx rate constant is tenfold greater than the blood—brain barrier value. This reflects the role of the choroid plexus in not only CSF but also brain Ca²⁺ homeostasis. The choroid plexus also may be involved in the transport of hormones into the CSF or may be a source of those hormones. For example, the choroid plexus may secrete insulin-like growth factor-II into the CSF; it also produces and secretes transthyretin, a carrier of thyroxine and retinol, into the CSF [26].

Other active transport systems in the choroid plexus are linked to the efflux of specific solutes [27]. For example, iodide and thiocyanate are transported from the CSF by saturable carrier mechanisms that can be competitively inhibited by perchlorate. This system must be active because transport can be carried out against unfavorable electrochemical gradients. Another important transport system removes weak organic acids from the CSF. Among the molecules cleared by this mechanism are penicillin and neurotransmitter metabolites, such as homovanillic acid and 5-hydroxyindoleacetic acid. This clearance system, which transports against an unfavorable CSF to blood gradient, is saturable and inhibited by probenecid. Clearance of organic acids by a probenecid-sensitive transport mechanism also may occur across the blood—brain barrier in brain capillaries.

The composition of the CSF when it enters the subarachnoid space may be modified by the arachnoid membrane. This tissue is part of the blood—CSF barrier and, like the choroid plexus and the cerebral capillaries, may not be a purely passive barrier but capable of actively modifying CSF composition.

Cerebrospinal fluid has a number of functions

A number of functions have been ascribed to the CSF [28]. The fluid-filled system around the brain has a buoyancy effect that may protect the brain from injury. Also, the rigidity of the skull means that increases in brain volume, such as those that occur from vasodilation or parenchymal cell swelling, could cause marked rises in intracranial pressure. The fluid in the CSF system, which can be displaced, limits such changes in pressure. Similarly, if the brain is dehydrated, the CSF acts as a source of fluid to rehydrate it.

As described above, the CSF system is the major source of entry of a number of substances into the brain. Why certain substances should enter via the blood—CSF barrier while others enter at the blood—brain barrier is uncertain. It might reflect the difference in the passive permeability of the two barriers since the tight junctions of the blood—CSF barrier are quantitatively leakier than those of the blood—brain barrier. Alternately, it may reflect the requirements of regions adjacent to the ventricular system.

The combination of bulk absorption of solute and solvent by the arachnoid villi and the selective removal of molecules by the choroid plexus means that there can be a concentration gradient for molecules reaching the interstitial fluid of the brain to diffuse into the CSF. Those molecules are then cleared by bulk flow or active transport. This function of the CSF, known as its sink action, helps to maintain the low concentration of many substances in both brain and CSF compared with plasma concentrations.

The CSF system also may play a role in signal transduction. It may provide a route for hormones to move within the brain, but it also may be a route of communication from the brain to the rest of the body. The bulk flow of CSF along the optic and olfactory nerves drains through lymphatic tissue, and antigenic material in the CSF may produce a systemic immune reaction.