New experimental model of acute aqueductal blockage in cats: effects on cerebrospinal fluid pressure and the size of brain ventricles.

It is generally assumed that cerebrospinal fluid (CSF) is secreted in the brain ventricles, and so after an acute blockage of the aqueduct of Sylvius an increase in the ventricular CSF pressure and dilation of isolated ventricles may be expected. We have tested this hypothesis in cats. After blocking the aqueduct, we measured the CSF pressure in both isolated ventricles and the cisterna magna, and performed radiographic monitoring of the cross-sectional area of the lateral ventricle. The complete aqueductal blockage was achieved by implanting a plastic cannula into the aqueduct of Sylvius through a small tunnel in the vermis of the cerebellum in the chloralose-anesthetized cats. After the reconstitution of the occipital bone, the CSF pressure was measured in the isolated ventricles via a plastic cannula implanted in the aqueduct of Sylvius and in the cisterna magna via a stainless steel cannula. During the following 2 h, the CSF pressures in the isolated ventricles and cisterna magna were identical to those in control conditions. We also monitored the ventricular cross-sectional area by means of radiography for 2 h after the aqueductal blockage and failed to observe any significant changes. When mock CSF was infused into isolated ventricles to imitate the CSF secretion, the gradient of pressure between the ventricle and cisterna magna developed, and disappeared as soon as the infusion was terminated. However, when mock CSF was infused into the cisterna magna at various rates, the resulting increased subarachnoid CSF pressure was accurately transmitted across the brain parenchyma into the CSF of isolated ventricles. The lack of the increase in the CSF pressure and ventricular dilation during 2 h of aqueductal blockage suggests that aqueductal obstruction by itself does not lead to development of hypertensive acute hydrocephalus in cats.

The investigation of cerebrospinal fluid formation by ventriculo-aqueductal perfusion method in cats.

OBJECTIVES: The perfusion of cerebrospinal fluid (CSF) spaces by artificial CSF (aCSF) containing an indicator, is an indirect method used to calculate CSF formation. To evaluate this method, we have developed a ventriculo-aqueductal perfusion method, which enables a direct measurement of CSF formation in the ventricles. METHODS: In chloralose anaesthetized cats, the aqueduct of Sylvius was cannulated so that the outflow end of the plastic cannula was positioned extracranially. Both lateral ventricles were also cannulated, with one cannula for infusion of aCSF containing blue dextrane and the other for measurement of CSF pressure. RESULTS: During ventriculo-aqueductal perfusion (direct method) under physiological CSF pressure, the outflow rate from aqueductal cannula did not differ significantly from the inflow rate, i.e. no CSF formation was observed. When the indirect method based on dilution of blue dextran in the outflowing perfusate was used, the formation of approximately 5 microl/min of CSF was obtained. CONCLUSION: Results of the direct method indicate that net CSF formation inside brain ventricles does not exist. The opposite results obtained by the indirect method questions this method as a reliable study of CSF formation.

Spinal contribution to CSF pressure lowering effect of mannitol in cats.

OBJECTIVES: After application of hyperosmolar mannitol the cerebrospinal (CSF) pressure is usually lowered within 30 min but this effect cannot be explained either by changes in intracranial blood volume and flow or by changes in brain volume. We assume that this effect of mannitol my be consequence of CSF volume decrease primarily in the spinal CSF due to high compliance of the spinal dura. METHODS: To explore such a possibility we planned to separate spinal and cerebral CSF. In chloralose anaesthetized cats dorsal laminectomy of C2 vertebrae was performed and a plastic semi ring was positioned extradurally separating cranial and spinal CSF. CSF pressures were recorded via cannulas positioned in lateral ventricle and lumbar subarachnoid space at L3 vertebrae, respectively. RESULTS: After intravenous bolus of 20% mannitol (0.5 or 1.0 g/kg/ 3 min) in control animals without cervical stenosis, the fall of both ventricular and lumbar CSF pressures was equal over time. At 15 min after mannitol application in cats with cervical stenosis an slight increase of ventricular and a fall of lumbar CSF pressures were observed, while at 30 min a gradient of these pressures of 5.5 and 7 cm H2O at lower and higher dose of mannitol, respectively, were registered. However, after removal of cervical stenosis these gradients disappeared. CONCLUSION: The observed changes of CSF pressures in spinal and intracranial space indicate that spinal subarachnoid space contributes a great deal to overall fall of CSF pressure and volume in the early period after mannitol application probably due to high compliance of the spinal dura.

Evidence for native NMDA receptor subtype pharmacology as revealed by differential effects on the NMDA-evoked release of striatal neuromodulators: eliprodil, ifenprodil and other native NMDA receptor subtype selective compounds.

NMDA increases the release of [14C]acetylcholine and [3H]spermidine or of [14C]GABA and [3H]dopamine from rat striatal slices. The pharmacology of these responses suggests that release of dopamine and GABA, acetylcholine, and spermidine is mediated, respectively, by three distinct NMDA receptor subtypes. IC50 values of compounds for the inhibition of dopamine and GABA release were closely matched, suggesting mediation by the same subtype. This receptor was generally more sensitive to all NMDA antagonists tested relative to that controlling acetylcholine or spermidine release (channel blockers, glycine antagonists, competitive antagonists and polyamine antagonists). The receptors controlling acetylcholine and spermidine release were characterised by lower antagonist sensitivity in general, and that controlling spermidine release was further defined by a marked insensitivity to ifenprodil, eliprodil, magnesium, dextromethorphan, dextrorphan, memantine, desipramine and polyamine spider toxins. In binding studies in which the displacement of 2 nM [3H]MK801 was studied in membranes prepared from a number of brain regions (in the presence of saturating concentrations of glutamate, glycine and spermidine) small regional differences in IC50 values were observed for a number of channel blockers, but no compound generated biphasic displacement curves that would allow masking of a particular subtype and it was not possible to detect binding components that were insensitive to memantine, dextrorphan dextromethorphan or desipramine. Ifenprodil produced biphasic displacement curves in the 1-day-old rat cortex and midbrain (with IC50 values of approximately 2 and 70 microM) and both ifenprodil and eliprodil displaced a small proportion (18%) of [3H]MK-801 with high affinity in the adult rat spinal cord. Displacement of [3H]MK801 by these compounds in all other adult brain regions (cortex, striatum, hippocampus, thalamus, pons, medulla, cerebellum) was monophasic and of low affinity. In general the subtype selectivity suggested by the release studies was not mirrored in the binding experiments, probably because of excessive heterogeneity of sites in the membrane preparations and to the subtype selectivity of [3H]MK801 itself.

Pharmacology of N-methyl-D-aspartate-evoked [3H]noradrenaline release in adult rat spinal cord.

N-Methyl-D-aspartate (NMDA) produced a concentration-related increase in [3H]noradrenaline release from adult rat cervical spinal cord slices. Its potency was relatively low and the response concentrated in dorsal spinal regions although also observed in ventral slices. NMDA did not increase the release of radiolabelled glutamate, aspartate, gamma-aminobutyric acid (GABA), acetylcholine or serotonin. In comparison with previously characterised NMDA responses in the striatum, (release of dopamine, GABA, acetylcholine or spermidine) the spinal response was particularly sensitive to MK-801 and magnesium and to L-689,560 but not to other glycine receptor antagonists (7-chlorokynurenate, CNQX (6-cyano-7-nitroquinoxaline-2,3-dione), DNQX (6,7-dichloroquinoxaline-2,3-dione), (+)-HA966). Dextrorphan and dextromethorphan produced partial or biphasic inhibition curves suggesting a subdivision of NMDA receptors. NMDA-evoked [3H]noradrenaline release was moderately sensitive to CPP and CGP37849 but insensitive to arcaine. These characteristics distinguish the native spinal NMDA receptor subtype(s) from those so far characterised in the striatum suggesting a unique spinal NMDA receptor subtype.

Effect of active transport on distribution and concentration gradients of [3H]benzylpenicillin in the cerebrospinal fluid.

After application of [3H]benzylpenicillin ([3H]BP) in lateral brain ventricle in dogs, the distribution of [3H]BP to contralateral ventricle and cisterna magna was much higher when its active transport from cerebrospinal fluid (CSF) was blocked by probenecid than under control conditions. Analysis of [3H]BP concentrations in both lateral ventricles and cisterna magna over time indicates that active transport restricts distribution of substances along CSF spaces and contributes to the maintenance of their concentration gradients between CSF compartments. This suggests that biochemical changes in a part of the brain and the adjacent CSF compartment may not be reflected into remote compartments of CSF such as lumbar CSF if substances in question are removed from CSF by active transport.

The pharmacology of native N-methyl-D-aspartate receptor subtypes: different receptors control the release of different striatal and spinal transmitters.

1. N-methyl-D-aspartate (NMDA) increases the release of radiolabelled dopamine, GABA, acetylcholine and spermidine from rat striatal slices and of noradrenaline from the dorsal cervical spinal cord. 2. These five responses show differing sensitivities to NMDA and also to a variety of competitive antagonists, NMDA channel blockers, glycine antagonists and polyamine site antagonists. 3. Inhibitory activity profiles for 20 different antagonists are presented. All compounds tested showed some degree of selectivity with regard to the different responses and each response showed particular characteristics that suggested mediation by a particular native NMDA receptor subtype. 4. Receptors controlling dopamine, GABA and noradrenaline release were generally more sensitive to most antagonists compared to those controlling acetylcholine and spermidine release. 5. Receptors controlling spermidine release were furthermore insensitive to magnesium, argiotoxin, ifenprodil and eliprodil and displayed low sensitivity to memantine, dextrorphan and dextromethorphan. 6. Receptors controlling noradrenaline release could be further discriminated from those controlling dopamine and GABA release by very high sensitivity to magnesium and MK-801 and to the glycine antagonist L-689,560 but not to other glycine antagonists (CNQX, DNQX, 7-Chlorokynurenate, HA-966). 7. Many other individual drug or receptor differences were noted. The different profiles observed suggest a wide diversity of native NMDA receptors with different properties and an unexpectedly rich pharmacopeia of subtype selective antagonists of native NMDA receptors. 8. Matching subtype selectivity to particular behavioural effects may be possible and the design of subtype selective NMDA antagonists for particular clinical applications while avoiding side effect generation seems to be feasible.

Homeostatic role of the active transport in elimination of [3H]benzylpenicillin out of the cerebrospinal fluid system.

Cerebral acidic metabolites and penicillin are organic anions which can be carried by active transport into capillaries of the central nervous system (CNS). However, it is generally believed that these metabolites are mainly delivered from CNS to cerebrospinal fluid (CSF) and eliminated by CSF circulation over cortex and its absorption into dural venous sinuses. To test this hypothesis we studied fate of penicillin ([3H]benzylpenicillin) in the CSF under control conditions and when its active transport was blocked by probenecid. After application of penicillin into cisterna magna of control dogs, it is distributed only in traces to lumbar, ventricular and cortical CSF. However, when active transport of penicillin across capillary wall is blocked by probenecid, its disappearance from cisterna is slowed down and its distribution is greatly enhanced so that at 300 min penicillin concentrations in cisternal, lumbar and cortical CSF approach or equal each other. Disappearance of penicillin from cisternal CSF shows a single exponential course (half-time 30 min) in control, while in probenecid pretreated dogs this is a slow multiexponential process. The results indicate that the active transport across capillary wall in CNS, but not generally postulated unidirectional CSF circulation over cortex and its absorption into dural venous sinuses, is instrumental in elimination of cerebral acidic metabolites and in such a way homeostasis in brain and cerebrospinal fluid is maintained.

Does the secretion and circulation of the cerebrospinal fluid really exist?

The secretion and circulation of cerebrospinal fluid have been studied in anaesthetized cats by means of a plastic cannula introduced into the aqueduct of Sylvius and by inspection of free escape of cerebrospinal fluid out of the end of the cannula. The fact that during the 120-minute period of observation not a single drop of CSF escaped out of the cannula, at physiological pressure, indicates that cerebrospinal fluid does neither secrete nor circulate.

"Compensated hyperosmolarity" of cerebrospinal fluid and the development of hydrocephalus.

Acute osmolar loading of cerebrospinal fluid within one lateral ventricle of dogs was examined as a cause of water extraction from the bloodstream and an increase in intracranial pressure. We have shown that a certain amount of (3)H2O from the bloodstream enters osmotically loaded cerebrospinal fluid significantly faster, hence causing a significant increase in intracranial pressure. The noted phenomenon in which intracranial pressure still significantly increases, but in which the hyperosmolarity of the cerebrospinal fluid is no longer present, was named "compensated hyperosmolarity". In the case of the sub-chronic application of hyperosmolar solutions into cat ventricles, we observed an increase in cerebrospinal fluid volume and a more pronounced development of hydrocephalus in the area of application, but without significant increase in intracranial pressure and without blockage of cerebrospinal fluid pathways. These results support the newly proposed hypothesis of cerebrospinal fluid hydrodynamics and the ability to develop new strategies for the treatment of cerebrospinal fluid-related diseases.

Development of hydrocephalus and classical hypothesis of cerebrospinal fluid hydrodynamics: facts and illusions.

According to the classical hypothesis of the cerebrospinal fluid (CSF) hydrodynamics, CSF is produced inside the brain ventricles, than it circulates like a slow river toward the cortical subarachnoid space, and finally it is absorbed into the venous sinuses. Some pathological conditions, primarily hydrocephalus, have also been interpreted based on this hypothesis. The development of hydrocephalus is explained as an imbalance between CSF formation and absorption, where more CSF is formed than is absorbed, which results in an abnormal increase in the CSF volume inside the cranial CSF spaces. It is believed that the reason for the imbalance is the obstruction of the CSF pathways between the site of CSF formation and the site of its absorption, which diminishes or prevents CSF outflow from the cranium. In spite of the general acceptance of the classical hypothesis, there are a considerable number of experimental results that do not support such a hypothesis and the generally accepted pathophysiology of hydrocephalus. A recently proposed new working hypothesis suggests that osmotic and hydrostatic forces at the central nervous system microvessels are crucial for the regulation of interstial fluid and CSF volume which constitute a functional unit. Based on that hypothesis, the generally accepted mechanisms of hydrocephalus development are not plausible. Therefore, the recent understanding of the correlation between CSF physiology and the development of hydrocephalus has been thoroughly presented, analyzed and evaluated, and new insights into hydrocephalus etiopathology have been proposed, which are in accordance with the experimental data and the new working hypothesis.

The formation of cerebrospinal fluid: nearly a hundred years of interpretations and misinterpretations.

The first scientific and experimental approaches to the study of cerebrospinal fluid (CSF) formation began almost a hundred years ago. Despite researchers being interested for so long, some aspects of CSF formation are still insufficiently understood. Today it is generally believed that CSF formation is an active energy consuming metabolic process which occurs mainly in brain ventricles, in choroid plexuses. CSF formation, together with CSF absorption and circulation, represents the so-called classic hypothesis of CSF hydrodynamics. In spite of the general acceptance of this hypothesis, there is a considerable series of experimental results that do not support the idea of the active nature of CSF formation and the idea that choroid plexuses inside the brain ventricles are the main places of formation. The main goal of this review is to summarize the present understanding of CSF formation and compare this understanding to contradictory experimental results that have been obtained so far. And finally, to try to offer a physiological explanation by which these contradictions could be avoided. We therefore analyzed the main methods that study CSF formation, which enabled such an understanding, and presented their shortcomings, which could also be a reason for the erroneous interpretation of the obtained results. A recent method of direct aqueductal determination of CSF formation is shown in more detail. On the one hand, it provides the possibility of direct insight into CSF formation, and on the other, it clearly indicates that there is no net CSF formation inside the brain ventricles. These results are contradictory to the classic hypothesis and, together with other mentioned contradictory results, strongly support a recently proposed new working hypothesis on the hydrodynamics of CSF. According to this new working hypothesis, CSF is permanently produced and absorbed in the whole CSF system as a consequence of filtration and reabsorption of water volume through the capillary walls into the surrounding brain tissue. The CSF exchange between the entire CSF system and the surrounding tissue depends on (patho)physiological conditions that predominate within those compartments.

Experimental hydrocephalus and hydromyelia: a new insight in mechanism of their development.

One group of cats had an acrylic screw implanted into the adqueduct of Sylvius, while the other group of animals received a solution of kaolin into the cisterna magna. Three weeks later the dye phenolsulphonphthalein was instilled into the lateral ventricle to ascertain communication between CSF compartments, and thereafter the brain was perfused with formalin. As shown by planimetry of brain ventricles both groups of experimental animals developed hydrocephalus, i.e., coronal surface of brain ventricles was about 10 times larger in kaolin and about 3 times in aqueductal screw experiments than in the controls, respectively. In aqueductal screw experiments communication of CSF between lateral ventricle and subarachnoid spaces was not blocked but only restricted, i.e., an aqueductal stenosis was produced. In kaolin experiments communication of CSF between lateral ventricles and spinal subarachnoid space was blocked by thick meningeal adhesions in the upper cervical region, while the central spinal canal was dilated (hydromyelia) with enhanced CSF communication between it and the lumbar subarachnoid space. We assume that during systolic expansion of brain the CSF is displaced from the cranial cavity toward the spinal subarachnoid space which accommodates an additional volume of CSF primarily due to compliance of the spinal dural sac, while during diastole CSF recoils in the opposite direction. Thus, in case of aqueductal stenosis the undisplaced volume of CSF from the ventricles can be accommodated due to diminution of cerebral blood volume and brain parenchyma so that hydrocephalus develops over time. Since the cervical subarachnoid space is blocked in kaolin experiments the systolic brain expansion forces CSF from basal cisterns via the fourth ventricle into the aqueduct and central canal with consequent development of hydrocephalus and hydromyelia.