J Neurosurg 72:759-762, 1990
Benzodiazepine receptors and cerebrospinal fluid formation GREGG L. WILLIAMS, M.D., MICHAEL POLLAY, M.D., THOMAS SEALE, PH.D., BRENT HISEY, M.D., AND P. ALEX ROBERTS, PH.D. Neurosurgical Section, University o f Oklahoma Health Sciences Center, O k l a h o m a City, Oklahoma
There is autoradiographic evidence that peripheral-type benzodiazepine ligands bind with high affinity to the membranes of choroid plexus tissue. In this study, the binding of a 4'-chloro analog of diazepam (Ro 54864) to rabbit choroid plexus and cerebral cortex was accomplished utilizing an in vitro radioactive assay method. A kinetic analysis of this binding revealed a relatively high affinity of this ligand (KD) for peripheral binding sites in plexus tissue (KD = 16.1 nM/mg protein). There was a 4.6-fold greater density of binding sites (total receptor density (Bmax) = 2.3 pmol/mg) in choroidal membrane as compared to cortical tissue (Bmax = 0.5 pmol/mg). In 40 rabbits in which a ventricular perfusion system was used, the rate of cerebrospinal fluid (CSF) formation was observed to decrease some 48% in the presence of 10 -4 M Ro 5-4864, although some inhibition of secretory activity was still noted at a CSF concentration of 10-8 M. The choroid plexus tissue levels of adenosine 3',5'cyclic monophosphate (cAMP) and adenosine triphosphatase (ATPase) were not affected by 10-4 M Ro 5-4864. The results of this study support the notion that the specific benzodiazepine peripheral binding sites in choroid plexus serve to modulate CSF formation. The mechanism of action is poorly understood but does not involve the transport ATPase system or the second messenger cAMP. KEY WORDS benzodiazepine
"
choroid plexus rabbit
D
URING the past 15 years, distinct advances have been made in understanding major recognition or receptor sites for neurotransmitters or their modulators within and external to the central nervous system (CNS). 1.2,9,~0.J2.15These advances are in large measure based on the ability to label receptor sites on cell membranes with radioactive ligands. 5,8,14A8 Within the past 10 years, benzodiazepine receptors have been described in some detail and, based on kinetic studies, appear to be of two major types. 2"~2"~3"~7'~8From a neuropharmacological standpoint, the most important of these recognition sites is the central binding receptor through which certain neuropharmacological agents, such as diazepam, produce their pharmacological effect. 9 Experiments utilizing various benzodiazepine ligands revealed a second type of acceptor which was found in peripheral tissues (such as kidney and heart) and in brain areas that were principally but not exclusively outside the blood-brain barrier (such as ventricular ependyma, choroid plexus, olfactory bulb, and certain areas o f the limbic system). ~2.~3.~8 The high degree o f binding of the peripheral type of benzodiazepine ligands (for example, 4'-chloro analog J. Neurosurg. / Volume 7 2 / M a y , 1990
diazepam
9 Ro 5 - 4 8 6 4
9 cerebrospinal fluid
9
of diazepam (Ro 5-4864)) to choroid plexus tissue suggested a possible role o f these receptors in the modulation of cerebrospinal fluid (CSF) formation. In the experiments described here, the characteristics o f R o 5-4864 binding with the peripheral binding sites in rabbit choroid plexus and cerebral cortex were studied as were the pharmacological effect of this ligand on the intraventricular formation o f CSF. The plexus adenosine triphosphatase (ATPase) and adenosine 3',5'cyclic monophosphate (cAMP) activity was also measured in the presence o f this analog o f diazepam. M a t e r i a l s and M e t h o d s
Adult New Zealand White rabbits were sacrificed by decapitation. Choroid plexus (obtained from the lateral and fourth ventricles) and slices of cerebral cortex were obtained for the binding experiments. The pooled tissue was homogenized in 50 volumes o f ice-cold 50.0 m M Tris HC1 buffer (pH 7.4) with a cell disrupter. The tissue was then centrifuged (4"C) at 20,000 G and the pellet so obtained was suspended in 50 volumes of the buffer solution. The binding o f the 3H-labeled Ro 5-4864 759
G. L. Williams, et al. (specific activity of 73.8 C i / m m o l ) to the tissue m e m branes was carried out in a reaction mixture volume of 1.0 ml, which contained 0.1 ml tissue (approximately 100 ug of protein), 0.1 rnl of radioligand, and 0.1 ml buffer or competitor and was brought to volume with the Tris buffer. The binding was carried to equilibrium at 4~ during a 60-minute period. The reaction was terminated by rapid filtration using a filtering manifold and W h a t m a n G F / B filter strips. The tissue suspended on the filter paper was washed twice with 5.0 ml iced buffer solution. This procedure was duplicated over a range of concentrations of Ro 5-4864 from 0.25 to 16 nM.* The specific binding of the radioligand was defined as the difference in binding obtained in the presence or absence o f unlabeled Ro 5-4864 (5 #M). The radioactivity of tissue retained by the filter was measured in a scintillation solution with a liquid scintillation spectrometer. Protein m e a s u r e m e n t s of the tissue were made utilizing the Lowry technique. 7 For this study, 40 adult New Zealand White rabbits of either sex, each weighing 2.5 to 3.5 kg, were used for the ventriculocisternal perfusion experiments. The animals were anesthetized with a combination of ketamine (35 mg/kg) and xylazine (5 mg/kg), which was supplemented as required. Following the induction of anesthesia, a t r a c h e o s t o m y was performed and the animals were mechanically ventilated after being paralyzed with 2.0 nag o f p a n c u r o n i u m bromide. Arterial pressure, blood gases, and body temperature were maintained within n o r m a l limits in all animals. After a postanesthesia period o f stabilization (approximately 30 minutes), the lateral ventricle of each animal was entered with a blunt No. 20 cannula through a twistdrill hole in the skull (6.0 m m caudal to the coronal suture and 6.0 m m lateral to the sagittal suture). The outflow from the ventricular system was by way of a flared polyethylene (PE-160) tube which was wedged and cemented into the fourth ventricle with cyanoacrylate. The inflow cannulas were attached to a constantflow infusion p u m p set to deliver the ventricular perfusate at a rate of 105 ul/min. The perfusion solution was similar in c o m p o s i t i o n to normal CSF and contained 3H-dextran as the nondiffusible radiolabeled marker. The fluid entering the ventricular system was buffered to a p H between 7.35 and 7.40 by gassing the solution with 95% 0 2 / 5 % CO2. The fluid from the fourth ventricular outflow system was collected in a microfractionator at 10-minute intervals during the 5hour perfusion period. The effect o f Ro 5-4864 on CSF formation was studied after a 150-minute control period by the addition o f R o 5-4864 to the basic perfusion solution in a range o f concentrations between 10-4 and 10 -8 M. The rate o f CSF formation before and after the addition o f Ro 5-4864 to the inflowing ventricular fluid
was calculated from the dilution of the radioactive dextran during its passage through the ventricular system using the previously described equation: ~ Vf = V~ (C~ - Co/Co), where Vr and V~ represent the rate of CSF formation and rate of perfusate inflow, respectively, in ml/min, and C~ and Co represent the concentration of 3H-dextran in the fluid entering (C~) and exiting (Co) the ventricular system. The control value of Vt. in each experiment was calculated from the mean of computed values measured at 5-minute intervals during the 60minute period immediately prior to dispensing the inhibitor. Each rabbit served as its own control (pre- vs. posttreatrnent). Samples of the inflow and outflow perfusate were obtained at 10-minute intervals for analysis of radioactivity. These 300 IA aliquots were added to a scintillation counting fluid and counted to a 2-sigma level of less than 0.5%. The total ATPase (Na+/K + and Ca++/Mg ++) activity was measured in pooled choroid plexus from the lateral and fourth ventricles of rabbits. The measurements were made on tissue of about 30 ug dry weight. The fresh tissue was homogenized in deionized water (30 mg wet weight/ml), frozen at -80~ and lyophilized. The tissue was kept frozen until assayed. The final incubation volume of 200 ul consisted of 0.015 M Tris buffer, 120 m M NaC1, 4.0 m M KC1, 20 m M MgC12, 0.84 m M ethylenediaminetetra-acetic acid, and the choroid plexus sample. Optimal ATPase activity was obtained with the above electrolyte solution when the p H was adjusted to 7.9 at a temperature of 37~ The buffer, tissue (with and without Ro 5-4864), and the salts were incubated for 5 minutes at 37~ At time 0, the substrate (Na2-adenosine triphosphatc (ATP)) was added to the tubes at 10-second intervals. Total incubation time was 10 minutes. The reaction was terminated by the addition of the color reagent (3 ml v/v 0.45% malachite green in water and 4.2% a m m o n i u m molybdate in 4 N HC1). The color was quenched at 1 minute with 100 t~l of 34% sodium citrate in H20, thus stabilizing the color for 2 hours. Absorbance was measured at 600 nM. Samples were measured in triplicate. The ATPase activity was expressed as nmols of inorganic phosphate released by hydrolysis of ATP)/30 gg of dry weight of tissue/min. The Ro 5-4864 that was added to the plexus tissue as an inhibitor varied in concentration from 10 -8 to 10 -4 M. The assay of c A M P in choroid plexus was accomplished with a c A M P [~25I] radioimmunoassay (RIA) kit.t The assay is a double-antibody RIA which utilizes a pre-reacted antibody complex. The fresh pooled choroid plexus tissue was incubated for 1 hour in an artificial CSF m e d i u m at 370C in a 95% 02/5% CO2 environment. The experimental solution contained l0 -4 M Ro 5-4864 while the control m e d i u m contained no ligand. The incubated tissue was then homogenized with
* Unlabeled Ro 5-4864 obtained courtesy of HoffmanLaroche, Nutley, New Jersey.
t Radioimmunoassay kit manufactured by Dupont Co., Wilmington, Delaware.
760
J. Neurosurg. / Volume 72 / May, 1990
Benzodiazepine receptors and CSF 6% trichloroacetic acid at 4~ to make a 1 ml 10% w/ v solution. After centrifugation, the material was extracted four times with 5 volumes of water-saturated ether; after lyophilization, the extract was purified by ion-exchange chromatography. The working tracer solution (100 ~1) was added to the standard, control, and Ro 5-4864 incubated tubes. After overnight incubation, the cAMP was precipitated and radioactivity was measured in a g a m m a counter. Results were expressed as pmol/mg protein. Results
The quantity of radiolabeled Ro 5-4864 specifically bound to pooled choroid plexus (uncorrected for volume and protein content) at various concentration levels of this ligand is presented in Fig. 1. The shape of this curve demonstrates saturation of the binding of ligand to the peripheral receptors in plexus tissue. This curve represents both site-specific and nonspecific binding. The linearity of the Scatchard plot (Fig. 1 inset) is compatible with a single class of choroid plexus binding sites. From this plot (x axis intercept), after correcting for volume and tissue protein concentration (expressed per mg), the total receptor density (Bmax) can be calculated. The slope of this line represents -1/KD, where K~ represents the affinity of the ligand. The Bmax and KD values for both cortex and plexus tissue are presented in Table 1. The KD value for the plexus receptor
was relatively high at 16.1 nM. The Bmax value for choroid plexus was 2.3 pmols/mg of choroid plexus membrane protein, which is a 4.6-fold greater density of peripheral binding sites in plexus as compared to cerebral cortex. The inhibitory effect of Ro 5-4864 on CSF formation is presented in Fig. 2. The presence of this ligand in the fluid perfusing the ventricular system maximally inhibited fluid production by 48.53% _ 2.13% at a concentration of 10-4 M, although some effect was still noted at a concentration of 10-8 M. The effectiveness of this ligand in inhibiting the production of CSF appears to level off at a concentration of 10 -5 M. The incubation of choroid plexus tissue with 10-4 M Ro 5-4864 did not affect the level of either cAMP or ATPase as compared to control tissue (Table 2). Discussion
It was initially believed that only one class of benzodiazepine receptors existed in the CNS based on the binding characteristics of such c o m p o u n d s as diazepam and flunitrazepam in brain tissue. 2:2 Shortly after the characterization of the high-affinity central binding receptors for diazepam, peripheral-type binding sites were described in some detail, l'z'9'14,x8 These later receptor sites were shown to have a pharmacological profile distinct from that demonstrated for central binding receptors within the CNS. Generally, the benzodiazepine ligands with high affinity for the central binding
TABLE 1 [3H-labeled]Ro 5-4864 binding to peripheral benzodiazepine receptors in rabbit* Tissue KD (nM) Bmax(pmol/mg) cerebral cortex 15.1 0.5 choroid plexus 16.1 2.3 * KD = affinity of the ligand; B~ax= total receptor density.
FIG. 1. Graph showing the total (specific and nonspecific) binding of Ro 5-4864 to choroid plexus membrane. A Scatchard plot of the ratio of bound/free (B/F) radioligand vs. bound (B) is shown (inset). Values were derived from pooled choroid plexus and not corrected for sample volume or protein. Corrected values used to calculate the affinity of the ligand and total receptor density are presented in Table 1. J. Neurosurg. / Volume 7 2 / M a y , 1990
FIG. 2. Graph showing the effect of intraventricular Ro 54864 on the formation of cerebrospinal fluid (CSF) in the rabbit. The values represent the mean standard error of the mean. Results are shown for Ro 5-4864 at concentrations (cone) between 10-8 M and 10-4 M. 761
G. L. Williams, et al. TABLE 2 Effect of Ro 5-4864 on choroM plexus A TPase and cAMP*
Study Total ATPase cAMP control 3.79 4- 0.8 68.0 + 8.0 Ro 5-4864 3.09 +__0.4 68.0 4- 4.0 * All values represent mean _+ standard error of the mean. ATPase = adenosine triphosphate, expressed as umols/10 min/mg protein; cAMP = adenosine 3',5'cyclic monophosphate, expressed as t~mols/mgprotein.
receptors have a lower affinity for the peripheral binding sites, and the reverse was noted for benzodiazepine ligands that had high affinity for the peripheral binding sites, t'5~s In studies using radiolabeled benzodiazepine analogs (such as 3H-labeled Ro-5-4864), the peripheral binding sites were shown to be located within the CNS in the choroid plexus, ventricular ependyma, pineal gland, olfactory bulb, and certain areas of the limbic system. ~5"~4"~s These receptor sites were also demonstrated in the heart, lung, and kidney. Unlike the benzodiazepine ligands that have high affinity for the central binding receptors, the peripheral acting ligands such as Ro 5-4864 are temperature-sensitive and their binding affinity is not modulated by gamma-aminobutyric acid, barbiturates, or small permeable anions? 7 The physical size of these two classes of benzodiazepine receptors is also quite different. Until recently, the peripheral binding sites were considered to be acceptor sites rather than receptor sites since little was known concerning the pharmacological action of these peripheral type sites when activated? 3 Recently, studies have shown that Ro 5-4864 has a proconvulsant and anxiogenic action. 3"~4~7 There are also brief reports that this agent independently stimulates phospholipid methylation and the release of prolactin from the pituitary gland.6,~6 Modulation of Ro 5-4854 binding (inhibition) has also been recently demonstrated by Mantione, et aL,S by an endogenous polypeptide tigand derived from a methanol extract of several tissues f?om the rat, thus suggesting endogenous control of the peripheral binding sites. In this study, we have shown that Ro 5-4864 has high affinity for peripheral-type binding sites located in rabbit choroid plexus, and that the n u m b e r of these sites in choroid plexus is almost 5 times greater than that demonstrated in cerebral cortex (Table 1). The linear relationship between the ratio of b o u n d and free Ro 5-4864: b o u n d Ro 5-4864 in choroid plexus tissue is compatible with the existence of a single class of peripheral benzodiazepine receptors in the choroid plexus (Fig. 1 inset). The results of the ventriculocisternal perfusion experiments clearly support the notion that CSF formation can be significantly modulated by specific receptors activated by peripheral-type benzodiazepine analog ligands. The mechanism of action is not known but appears not to be associated with a cAMP second messenger system or the choroid plexus ATPase762
controlled sodium p u m p s (Table 2), both of which have been shown to be involved in the pharmacological control of CSF formation. ,0.,, References
1. B~navid~s J, Savaki HE, Malgouris C, et al: Autoradiographic localization of peripheral benzodiazepine binding sites in the cat brain with [3H]PK 11195. Brain Res Bull 13:69-77, 1984 2. Braestrup C, Nielsen M: Multiple benzodiazepine receptors. Trends Neurosei 3:301-303, 1980 3. DeLorenzo RJ, Burdette S, Holderness J: Benzodiazepine inhibition of the calcium-calmodulin protein kinase system in brain membrane. Science 213:546-549, 1981 4. File SE, Lister RG: The anxiogenic action of Ro 5-4864 is reversed by phenytoin. Neurosci Lett 35:93-96, 1983 5. Gehlert DR, Yamamura HI, Wamsley JK: Autoradiographic localization of"peripheral-type" benzodiazepine binding sites in the rat brain, heart and kidney. Arch Pharmacol 328:454-460, 1985 6. Grandison L: Characterization ofbenzodiazepinebinding and action on rat anterior pituitary cells. Neurosci Abstr 7:503, 1981 7. Lowry O, Rosebrough N, Farr A, et al: Protein measurement with folin phenol reagent. J Biol Chem 193: 265-275, 1951 8. Mantione CR, Weissman BA, Goldman ME, et al: Endogenous inhibitions of 4'-[3H]chlorodiazepam (Ro 54864) binding to "peripheral" sites for benzodiazepines. FEBS Lett 176:69-74, 1984 9. Martin IL: Endogenous ligands for benzodiazepine receptors. Trends Neurosei 3:299-301, 1980 10. Nathanson JA:/%adrenergic-sensitive adenylate cyclase in secretory cells of choroid plexus. Science 204:843-844, 1979 11. Pollay M: Formation of cerebral spinal fluid. Relation of studies of isolated choroid plexus to the standing gradient hypothesis. J Neurosurg 42:665-673, t975 12. Richards JG, M6hler H: Benzodiazepine receptors. Neuropharmacology 23:233-242, 1984 13. Richards JG, Mrhler H, Haefely W: Benzodiazepine binding sites: receptors or acceptors? Trends Pharmaeol Sei 3:233-235, 1982 14. Schoemaker H, Boles RG, Horst WD, et al: Specific high affinity binding sites for [3H]Ro 5-4864 in rat brain and kidney. J Pharmacol Exp Ther 225:61-69, 1983 15. Snyder SH: Brain receptors: the emergence of a new pharmacology. Trends Neurosci 9:455-459, 1986 t6. Strittmatter WJ, Hirata F, Axelrod J, et al: Benzodiazepine and /3-adrenergic receptor ligands independently stimulate phospholipid methylation. Nature 282: 857-859, 1979 17. Weissman BA, Cott J, Hommer D, et al: Pharmacological, electrophysiological, and neurochemical actions of the convulsant benzodiazepine Ro 5-4864(4'-chlordiazepam). Adv Biochem Psychopharmaeol 38:139-151, 1983 18. Yamamura HI, Gehlert DR, Gee KW, et al: Specific highaffinity [3H]Ro5-4864 benzodiazepine binding sites in the brain and periphery. Prog Ciin Biol Res 192:187-196, 1985 Manuscript received August 16, 1989.
Address reprint requests to: Gregg L. Williams, M.D.,
Neurosurgical Section, University of Oklahoma Health Sciences Center, P.O. Box 26307, Oklahoma City, Oklahoma 73126. J. Neurosurg. / Volume 7 2 / M a y , 1990
Benzodiazepine receptors and cerebrospinal fluid formation GREGG L. WILLIAMS, M.D., MICHAEL POLLAY, M.D., THOMAS SEALE, PH.D., BRENT HISEY, M.D., AND P. ALEX ROBERTS, PH.D. Neurosurgical Section, University o f Oklahoma Health Sciences Center, O k l a h o m a City, Oklahoma
There is autoradiographic evidence that peripheral-type benzodiazepine ligands bind with high affinity to the membranes of choroid plexus tissue. In this study, the binding of a 4'-chloro analog of diazepam (Ro 54864) to rabbit choroid plexus and cerebral cortex was accomplished utilizing an in vitro radioactive assay method. A kinetic analysis of this binding revealed a relatively high affinity of this ligand (KD) for peripheral binding sites in plexus tissue (KD = 16.1 nM/mg protein). There was a 4.6-fold greater density of binding sites (total receptor density (Bmax) = 2.3 pmol/mg) in choroidal membrane as compared to cortical tissue (Bmax = 0.5 pmol/mg). In 40 rabbits in which a ventricular perfusion system was used, the rate of cerebrospinal fluid (CSF) formation was observed to decrease some 48% in the presence of 10 -4 M Ro 5-4864, although some inhibition of secretory activity was still noted at a CSF concentration of 10-8 M. The choroid plexus tissue levels of adenosine 3',5'cyclic monophosphate (cAMP) and adenosine triphosphatase (ATPase) were not affected by 10-4 M Ro 5-4864. The results of this study support the notion that the specific benzodiazepine peripheral binding sites in choroid plexus serve to modulate CSF formation. The mechanism of action is poorly understood but does not involve the transport ATPase system or the second messenger cAMP. KEY WORDS benzodiazepine
"
choroid plexus rabbit
D
URING the past 15 years, distinct advances have been made in understanding major recognition or receptor sites for neurotransmitters or their modulators within and external to the central nervous system (CNS). 1.2,9,~0.J2.15These advances are in large measure based on the ability to label receptor sites on cell membranes with radioactive ligands. 5,8,14A8 Within the past 10 years, benzodiazepine receptors have been described in some detail and, based on kinetic studies, appear to be of two major types. 2"~2"~3"~7'~8From a neuropharmacological standpoint, the most important of these recognition sites is the central binding receptor through which certain neuropharmacological agents, such as diazepam, produce their pharmacological effect. 9 Experiments utilizing various benzodiazepine ligands revealed a second type of acceptor which was found in peripheral tissues (such as kidney and heart) and in brain areas that were principally but not exclusively outside the blood-brain barrier (such as ventricular ependyma, choroid plexus, olfactory bulb, and certain areas o f the limbic system). ~2.~3.~8 The high degree o f binding of the peripheral type of benzodiazepine ligands (for example, 4'-chloro analog J. Neurosurg. / Volume 7 2 / M a y , 1990
diazepam
9 Ro 5 - 4 8 6 4
9 cerebrospinal fluid
9
of diazepam (Ro 5-4864)) to choroid plexus tissue suggested a possible role o f these receptors in the modulation of cerebrospinal fluid (CSF) formation. In the experiments described here, the characteristics o f R o 5-4864 binding with the peripheral binding sites in rabbit choroid plexus and cerebral cortex were studied as were the pharmacological effect of this ligand on the intraventricular formation o f CSF. The plexus adenosine triphosphatase (ATPase) and adenosine 3',5'cyclic monophosphate (cAMP) activity was also measured in the presence o f this analog o f diazepam. M a t e r i a l s and M e t h o d s
Adult New Zealand White rabbits were sacrificed by decapitation. Choroid plexus (obtained from the lateral and fourth ventricles) and slices of cerebral cortex were obtained for the binding experiments. The pooled tissue was homogenized in 50 volumes o f ice-cold 50.0 m M Tris HC1 buffer (pH 7.4) with a cell disrupter. The tissue was then centrifuged (4"C) at 20,000 G and the pellet so obtained was suspended in 50 volumes of the buffer solution. The binding o f the 3H-labeled Ro 5-4864 759
G. L. Williams, et al. (specific activity of 73.8 C i / m m o l ) to the tissue m e m branes was carried out in a reaction mixture volume of 1.0 ml, which contained 0.1 ml tissue (approximately 100 ug of protein), 0.1 rnl of radioligand, and 0.1 ml buffer or competitor and was brought to volume with the Tris buffer. The binding was carried to equilibrium at 4~ during a 60-minute period. The reaction was terminated by rapid filtration using a filtering manifold and W h a t m a n G F / B filter strips. The tissue suspended on the filter paper was washed twice with 5.0 ml iced buffer solution. This procedure was duplicated over a range of concentrations of Ro 5-4864 from 0.25 to 16 nM.* The specific binding of the radioligand was defined as the difference in binding obtained in the presence or absence o f unlabeled Ro 5-4864 (5 #M). The radioactivity of tissue retained by the filter was measured in a scintillation solution with a liquid scintillation spectrometer. Protein m e a s u r e m e n t s of the tissue were made utilizing the Lowry technique. 7 For this study, 40 adult New Zealand White rabbits of either sex, each weighing 2.5 to 3.5 kg, were used for the ventriculocisternal perfusion experiments. The animals were anesthetized with a combination of ketamine (35 mg/kg) and xylazine (5 mg/kg), which was supplemented as required. Following the induction of anesthesia, a t r a c h e o s t o m y was performed and the animals were mechanically ventilated after being paralyzed with 2.0 nag o f p a n c u r o n i u m bromide. Arterial pressure, blood gases, and body temperature were maintained within n o r m a l limits in all animals. After a postanesthesia period o f stabilization (approximately 30 minutes), the lateral ventricle of each animal was entered with a blunt No. 20 cannula through a twistdrill hole in the skull (6.0 m m caudal to the coronal suture and 6.0 m m lateral to the sagittal suture). The outflow from the ventricular system was by way of a flared polyethylene (PE-160) tube which was wedged and cemented into the fourth ventricle with cyanoacrylate. The inflow cannulas were attached to a constantflow infusion p u m p set to deliver the ventricular perfusate at a rate of 105 ul/min. The perfusion solution was similar in c o m p o s i t i o n to normal CSF and contained 3H-dextran as the nondiffusible radiolabeled marker. The fluid entering the ventricular system was buffered to a p H between 7.35 and 7.40 by gassing the solution with 95% 0 2 / 5 % CO2. The fluid from the fourth ventricular outflow system was collected in a microfractionator at 10-minute intervals during the 5hour perfusion period. The effect o f Ro 5-4864 on CSF formation was studied after a 150-minute control period by the addition o f R o 5-4864 to the basic perfusion solution in a range o f concentrations between 10-4 and 10 -8 M. The rate o f CSF formation before and after the addition o f Ro 5-4864 to the inflowing ventricular fluid
was calculated from the dilution of the radioactive dextran during its passage through the ventricular system using the previously described equation: ~ Vf = V~ (C~ - Co/Co), where Vr and V~ represent the rate of CSF formation and rate of perfusate inflow, respectively, in ml/min, and C~ and Co represent the concentration of 3H-dextran in the fluid entering (C~) and exiting (Co) the ventricular system. The control value of Vt. in each experiment was calculated from the mean of computed values measured at 5-minute intervals during the 60minute period immediately prior to dispensing the inhibitor. Each rabbit served as its own control (pre- vs. posttreatrnent). Samples of the inflow and outflow perfusate were obtained at 10-minute intervals for analysis of radioactivity. These 300 IA aliquots were added to a scintillation counting fluid and counted to a 2-sigma level of less than 0.5%. The total ATPase (Na+/K + and Ca++/Mg ++) activity was measured in pooled choroid plexus from the lateral and fourth ventricles of rabbits. The measurements were made on tissue of about 30 ug dry weight. The fresh tissue was homogenized in deionized water (30 mg wet weight/ml), frozen at -80~ and lyophilized. The tissue was kept frozen until assayed. The final incubation volume of 200 ul consisted of 0.015 M Tris buffer, 120 m M NaC1, 4.0 m M KC1, 20 m M MgC12, 0.84 m M ethylenediaminetetra-acetic acid, and the choroid plexus sample. Optimal ATPase activity was obtained with the above electrolyte solution when the p H was adjusted to 7.9 at a temperature of 37~ The buffer, tissue (with and without Ro 5-4864), and the salts were incubated for 5 minutes at 37~ At time 0, the substrate (Na2-adenosine triphosphatc (ATP)) was added to the tubes at 10-second intervals. Total incubation time was 10 minutes. The reaction was terminated by the addition of the color reagent (3 ml v/v 0.45% malachite green in water and 4.2% a m m o n i u m molybdate in 4 N HC1). The color was quenched at 1 minute with 100 t~l of 34% sodium citrate in H20, thus stabilizing the color for 2 hours. Absorbance was measured at 600 nM. Samples were measured in triplicate. The ATPase activity was expressed as nmols of inorganic phosphate released by hydrolysis of ATP)/30 gg of dry weight of tissue/min. The Ro 5-4864 that was added to the plexus tissue as an inhibitor varied in concentration from 10 -8 to 10 -4 M. The assay of c A M P in choroid plexus was accomplished with a c A M P [~25I] radioimmunoassay (RIA) kit.t The assay is a double-antibody RIA which utilizes a pre-reacted antibody complex. The fresh pooled choroid plexus tissue was incubated for 1 hour in an artificial CSF m e d i u m at 370C in a 95% 02/5% CO2 environment. The experimental solution contained l0 -4 M Ro 5-4864 while the control m e d i u m contained no ligand. The incubated tissue was then homogenized with
* Unlabeled Ro 5-4864 obtained courtesy of HoffmanLaroche, Nutley, New Jersey.
t Radioimmunoassay kit manufactured by Dupont Co., Wilmington, Delaware.
760
J. Neurosurg. / Volume 72 / May, 1990
Benzodiazepine receptors and CSF 6% trichloroacetic acid at 4~ to make a 1 ml 10% w/ v solution. After centrifugation, the material was extracted four times with 5 volumes of water-saturated ether; after lyophilization, the extract was purified by ion-exchange chromatography. The working tracer solution (100 ~1) was added to the standard, control, and Ro 5-4864 incubated tubes. After overnight incubation, the cAMP was precipitated and radioactivity was measured in a g a m m a counter. Results were expressed as pmol/mg protein. Results
The quantity of radiolabeled Ro 5-4864 specifically bound to pooled choroid plexus (uncorrected for volume and protein content) at various concentration levels of this ligand is presented in Fig. 1. The shape of this curve demonstrates saturation of the binding of ligand to the peripheral receptors in plexus tissue. This curve represents both site-specific and nonspecific binding. The linearity of the Scatchard plot (Fig. 1 inset) is compatible with a single class of choroid plexus binding sites. From this plot (x axis intercept), after correcting for volume and tissue protein concentration (expressed per mg), the total receptor density (Bmax) can be calculated. The slope of this line represents -1/KD, where K~ represents the affinity of the ligand. The Bmax and KD values for both cortex and plexus tissue are presented in Table 1. The KD value for the plexus receptor
was relatively high at 16.1 nM. The Bmax value for choroid plexus was 2.3 pmols/mg of choroid plexus membrane protein, which is a 4.6-fold greater density of peripheral binding sites in plexus as compared to cerebral cortex. The inhibitory effect of Ro 5-4864 on CSF formation is presented in Fig. 2. The presence of this ligand in the fluid perfusing the ventricular system maximally inhibited fluid production by 48.53% _ 2.13% at a concentration of 10-4 M, although some effect was still noted at a concentration of 10-8 M. The effectiveness of this ligand in inhibiting the production of CSF appears to level off at a concentration of 10 -5 M. The incubation of choroid plexus tissue with 10-4 M Ro 5-4864 did not affect the level of either cAMP or ATPase as compared to control tissue (Table 2). Discussion
It was initially believed that only one class of benzodiazepine receptors existed in the CNS based on the binding characteristics of such c o m p o u n d s as diazepam and flunitrazepam in brain tissue. 2:2 Shortly after the characterization of the high-affinity central binding receptors for diazepam, peripheral-type binding sites were described in some detail, l'z'9'14,x8 These later receptor sites were shown to have a pharmacological profile distinct from that demonstrated for central binding receptors within the CNS. Generally, the benzodiazepine ligands with high affinity for the central binding
TABLE 1 [3H-labeled]Ro 5-4864 binding to peripheral benzodiazepine receptors in rabbit* Tissue KD (nM) Bmax(pmol/mg) cerebral cortex 15.1 0.5 choroid plexus 16.1 2.3 * KD = affinity of the ligand; B~ax= total receptor density.
FIG. 1. Graph showing the total (specific and nonspecific) binding of Ro 5-4864 to choroid plexus membrane. A Scatchard plot of the ratio of bound/free (B/F) radioligand vs. bound (B) is shown (inset). Values were derived from pooled choroid plexus and not corrected for sample volume or protein. Corrected values used to calculate the affinity of the ligand and total receptor density are presented in Table 1. J. Neurosurg. / Volume 7 2 / M a y , 1990
FIG. 2. Graph showing the effect of intraventricular Ro 54864 on the formation of cerebrospinal fluid (CSF) in the rabbit. The values represent the mean standard error of the mean. Results are shown for Ro 5-4864 at concentrations (cone) between 10-8 M and 10-4 M. 761
G. L. Williams, et al. TABLE 2 Effect of Ro 5-4864 on choroM plexus A TPase and cAMP*
Study Total ATPase cAMP control 3.79 4- 0.8 68.0 + 8.0 Ro 5-4864 3.09 +__0.4 68.0 4- 4.0 * All values represent mean _+ standard error of the mean. ATPase = adenosine triphosphate, expressed as umols/10 min/mg protein; cAMP = adenosine 3',5'cyclic monophosphate, expressed as t~mols/mgprotein.
receptors have a lower affinity for the peripheral binding sites, and the reverse was noted for benzodiazepine ligands that had high affinity for the peripheral binding sites, t'5~s In studies using radiolabeled benzodiazepine analogs (such as 3H-labeled Ro-5-4864), the peripheral binding sites were shown to be located within the CNS in the choroid plexus, ventricular ependyma, pineal gland, olfactory bulb, and certain areas of the limbic system. ~5"~4"~s These receptor sites were also demonstrated in the heart, lung, and kidney. Unlike the benzodiazepine ligands that have high affinity for the central binding receptors, the peripheral acting ligands such as Ro 5-4864 are temperature-sensitive and their binding affinity is not modulated by gamma-aminobutyric acid, barbiturates, or small permeable anions? 7 The physical size of these two classes of benzodiazepine receptors is also quite different. Until recently, the peripheral binding sites were considered to be acceptor sites rather than receptor sites since little was known concerning the pharmacological action of these peripheral type sites when activated? 3 Recently, studies have shown that Ro 5-4864 has a proconvulsant and anxiogenic action. 3"~4~7 There are also brief reports that this agent independently stimulates phospholipid methylation and the release of prolactin from the pituitary gland.6,~6 Modulation of Ro 5-4854 binding (inhibition) has also been recently demonstrated by Mantione, et aL,S by an endogenous polypeptide tigand derived from a methanol extract of several tissues f?om the rat, thus suggesting endogenous control of the peripheral binding sites. In this study, we have shown that Ro 5-4864 has high affinity for peripheral-type binding sites located in rabbit choroid plexus, and that the n u m b e r of these sites in choroid plexus is almost 5 times greater than that demonstrated in cerebral cortex (Table 1). The linear relationship between the ratio of b o u n d and free Ro 5-4864: b o u n d Ro 5-4864 in choroid plexus tissue is compatible with the existence of a single class of peripheral benzodiazepine receptors in the choroid plexus (Fig. 1 inset). The results of the ventriculocisternal perfusion experiments clearly support the notion that CSF formation can be significantly modulated by specific receptors activated by peripheral-type benzodiazepine analog ligands. The mechanism of action is not known but appears not to be associated with a cAMP second messenger system or the choroid plexus ATPase762
controlled sodium p u m p s (Table 2), both of which have been shown to be involved in the pharmacological control of CSF formation. ,0.,, References
1. B~navid~s J, Savaki HE, Malgouris C, et al: Autoradiographic localization of peripheral benzodiazepine binding sites in the cat brain with [3H]PK 11195. Brain Res Bull 13:69-77, 1984 2. Braestrup C, Nielsen M: Multiple benzodiazepine receptors. Trends Neurosei 3:301-303, 1980 3. DeLorenzo RJ, Burdette S, Holderness J: Benzodiazepine inhibition of the calcium-calmodulin protein kinase system in brain membrane. Science 213:546-549, 1981 4. File SE, Lister RG: The anxiogenic action of Ro 5-4864 is reversed by phenytoin. Neurosci Lett 35:93-96, 1983 5. Gehlert DR, Yamamura HI, Wamsley JK: Autoradiographic localization of"peripheral-type" benzodiazepine binding sites in the rat brain, heart and kidney. Arch Pharmacol 328:454-460, 1985 6. Grandison L: Characterization ofbenzodiazepinebinding and action on rat anterior pituitary cells. Neurosci Abstr 7:503, 1981 7. Lowry O, Rosebrough N, Farr A, et al: Protein measurement with folin phenol reagent. J Biol Chem 193: 265-275, 1951 8. Mantione CR, Weissman BA, Goldman ME, et al: Endogenous inhibitions of 4'-[3H]chlorodiazepam (Ro 54864) binding to "peripheral" sites for benzodiazepines. FEBS Lett 176:69-74, 1984 9. Martin IL: Endogenous ligands for benzodiazepine receptors. Trends Neurosei 3:299-301, 1980 10. Nathanson JA:/%adrenergic-sensitive adenylate cyclase in secretory cells of choroid plexus. Science 204:843-844, 1979 11. Pollay M: Formation of cerebral spinal fluid. Relation of studies of isolated choroid plexus to the standing gradient hypothesis. J Neurosurg 42:665-673, t975 12. Richards JG, M6hler H: Benzodiazepine receptors. Neuropharmacology 23:233-242, 1984 13. Richards JG, Mrhler H, Haefely W: Benzodiazepine binding sites: receptors or acceptors? Trends Pharmaeol Sei 3:233-235, 1982 14. Schoemaker H, Boles RG, Horst WD, et al: Specific high affinity binding sites for [3H]Ro 5-4864 in rat brain and kidney. J Pharmacol Exp Ther 225:61-69, 1983 15. Snyder SH: Brain receptors: the emergence of a new pharmacology. Trends Neurosci 9:455-459, 1986 t6. Strittmatter WJ, Hirata F, Axelrod J, et al: Benzodiazepine and /3-adrenergic receptor ligands independently stimulate phospholipid methylation. Nature 282: 857-859, 1979 17. Weissman BA, Cott J, Hommer D, et al: Pharmacological, electrophysiological, and neurochemical actions of the convulsant benzodiazepine Ro 5-4864(4'-chlordiazepam). Adv Biochem Psychopharmaeol 38:139-151, 1983 18. Yamamura HI, Gehlert DR, Gee KW, et al: Specific highaffinity [3H]Ro5-4864 benzodiazepine binding sites in the brain and periphery. Prog Ciin Biol Res 192:187-196, 1985 Manuscript received August 16, 1989.
Address reprint requests to: Gregg L. Williams, M.D.,
Neurosurgical Section, University of Oklahoma Health Sciences Center, P.O. Box 26307, Oklahoma City, Oklahoma 73126. J. Neurosurg. / Volume 7 2 / M a y , 1990
