Diffusion of [3H]dexamethasone in rat subcutaneous slices after injection measured by digital autoradiography

A relatively simple method for the determination of the diffusion coefficient of a substance that has been injected into tissue is described. We illustrate this method using [3]dexamethasone injected into the subcutaneous tissue of rats. Digital autoradiography was used to measure the distribution of the [3H] dexamethasone within the subcutaneous tissue at 2.5 and 20 min after injection. Measured concentration profiles of the injection were compared to a mathematical model of drug diffusion from an injection. There was good agreement between the experimental data and the mathematical model. The diffusion coefficient found using this simple injection method was (4.01 +/- 2.01) x 10(-10) m2/s. This D value was very close to the value of D = (4.11 +/- 1.77) x 10(-10) m2/s found previously using different mathematical and experimental techniques with osmotic pumps implanted for 6, 24, and 60 h in rats (1). The simple method given here for the determination of the diffusion coefficient is general enough to be applied to other substances and tissues as well.     

Quantitative dual-probe microdialysis: evaluation of [3H]mannitol diffusion in agar and rat striatum

Dual-probe microdialysis was used to study interstitial diffusion in the rat brain. A radiolabelled tracer, (3H]mannitol, was continuously infused at different concentrations via a probe acutely implanted into the striatum of an anaesthetized male rat or into a dilute agar gel. Samples were collected by a second probe placed 1 mm away from the first, and the recovered [3H]mannitol was measured by liquid scintillation counting. In the striatum, the delivery of [3H]mannitol was counteracted by its removal from the extracellular space by passive uptake into cells and clearance into the microcirculation, causing the diffusion profile to approach quasi steady-state levels within 2 h. Diffusion data from brain and agar were analysed using a mathematical model. The apparent (effective) diffusion coefficient for [3H]mannitol was D* = 2.9 x 10(-6) cm2/s, the effective volume fraction alpha* = 0.30 and the clearance rate constant kappa= 2.3 x 10(-5)/s. A tortuosity, lambda = 1.81, and penetration distance r = 4.2 mm, were calculated. We conclude that, using dual-probe microdialysis, parameters reflecting geometric and dynamic tissue properties may be obtained using appropriate mathematical analysis. Quantitative dual-probe microdialysis will be valuable in characterizing interstitial diffusion and the clearance processes underpinning volume transmission in the brain.     

Differential binding of dexamethasone to ammonium sulfate precipitates of human adipose tissue cytosols

Saturation analysis of the binding of [3H]dexamethasone [( 3H]DEX) to ammonium sulfate precipitates (ASPs) confirmed the presence of a limited-capacity, high-affinity binder in human adipose tissue cytosols. Various non-radioactive steroids competed with [3H]DEX for binding to the ASPs in the following sequence: dexamethasone (DEX) approximately equal to triamcinolone acetonide (TA) greater than progesterone (P) much greater than estradiol (E2). The steroid specificity of the binder precipitated by AS was consistent with the specificities reported for glucocorticoid receptors in a number of systems. In order to investigate possible regional differences, glucocorticoid binding to ASPs derived from adipose tissues removed from two different sites in the same subject was quantitated. ASPs of human omental adipose tissue bound significantly more [3H]DEX than did similar preparations of subcutaneous adipose tissue from the abdominal wall (116 +/- 32 vs. 50 +/- 22 fmol/mg protein; mean +/- SD; p less than 0.02). The findings are consistent with reports from other laboratories suggesting that intra-abdominal fat is more responsive to glucocorticoids than is subcutaneous adipose tissue.     

The mechanism of [3H]dexamethasone uptake into prolactin producing rat pituitary cells (GH3 cells) in culture

Glucocorticosteroids stimulate growth hormone (GH) synthesis and inhibit prolactin (PRL) synthesis and cell growth in cultured GH3 cells, a clonal cell strain derived from a rat pituitary tumour. This model system was used to study the mechanism by which glucocorticosteroids enter target cells. The cellular uptake of [3H]dexamethasone was temperature dependent and was further inhibited by addition of an excess amount of cold dexamethasone. Half maximal uptake was obtained after about 5 min at 37 degrees C. The initial rates of [3H]dexamethasone uptake were a linear function of the extracellular hormone concentration. The uptake of [3H]dexamethasone in intact cells studied at different temperatures resulted in linear Arrhenius plots, with a calculated energy of activation of 91.0 kJ x mole-1 x degree-1. Scatchard analysis of specifically cell bound [3H]dexamethasone at equilibrium (0 degrees C) showed a straight line with a calculated dissociation constant (Kd) of 1.6 x 10(-9) M and a maximal uptake of 180 x 10(-15) mole/mg cell protein. Specific binding of [3H]dexamethasone to cytosol proteins could only be demonstrated at 0 degrees C. These results indicate that [3H]dexamethasone diffuses passively into the cell, and binds to specific receptors in an energy dependent way.     

H-Pro-[3H]Leu-Gly-NH2: Plasma Profile and Brain Uptake Following Subcutaneous Injection in the Rat

Following subcutaneous injection of the tripeptide H-Pro-[3H]Leu-Gly-NH2 ([3H]PLG) in rats, the profile of intact peptide and its radioactively labeled metabolites was examined both in plasma and in brain tissue. [3H]PLG and metabolites were determined in trichloroacetic acid extracts by reverse-phase paired-ion HPLC. Maximal plasma levels of unmetabolized PLG were reached 6-8 min after administration, after which they decreased with an elimination half-life of 20 min. The uptake of [3H]PLG in the brain ranged from 0.0013% to 0.0017% of the administered dose per g tissue at 6-30 min following subcutaneous injection. After comparing these results with our previous findings with intravenous injection of [3H]PLG, it seemed likely that the subcutaneous route of administration might be more effective in eliciting CNS effects of PLG than the intravenous route of administration. The metabolite profiles in plasma and brain point to an initial cleavage of PLG at the NH2-terminal side and a very rapid degradation of the peptide intermediate H-Leu-Gly-NH2.     

Concanavalin A for in vivo glucose sensing: a biotoxicity review

Over the last two decades there as has been surging scientific interest in employing the glucose- and mannose-specific lectin Concanavalin A (ConA) in affinity biosensors for in vivo glucose monitoring in diabetics. Numerous research groups have successfully shown in in vitro and in vivo studies that ConA-based affinity sensors can monitor glucose very accurately and reproducibly over many months, making ConA-based sensors an extremely interesting prospect for long-term implantation in humans. Despite this progress, there remains concern over the safety of ConA, which has widely been reported as a toxin in the literature. In this article, we review in vitro and in vivo studies related to ConA toxicity in order to assess the health risks posed by ConA in the context of an implantable biosensor. Based on the wealth of information available and on data from our own studies, we can conclude that the site of implantation (subcutaneous skin tissue) and the small amount of ConA (<10 microg/microl) being used in implantable glucose-sensitive detector devices like those proposed by various research groups would pose little or no health risk to its bearer even in the event of unexpected sensor rupture.     

A hydrolase-related transport system is not required to explain the intestinal uptke of glucose liberated from phlorizin

The fate of [3H]glucose released from a wide range of [3H]phlorizin concentrations by phlorizin hydrolase has been studied under conditions where the Na+-dependent glucose transport system in hamster intestine is profoundly inhibited by the glucoside. At 0.2-2.0 mM phlorizin, the [3H]glucose uptake was a constant 11-12% of that generated by the enzyme and at the highest level, it was reduced to that of passive diffusion. Glucose liberated from 0.2 mM [3H]phlorizin is accumulated at a rate nearly equal to that found for 0.2 mM [14C]glucose when this free sugar uptake is measured in a medium containing 0.2 mM unlabeled phlorizin. Furthermore, without sodium, the accumulation rates of hydrolase-derived or exogenous glucose are both reduced to the rate of [14C]mannitol. Our results indicate that the glucose released from phlorizin enters the tissue via the small fraction of the Na+-dependent glucose carriers which escape phlorizin blockade together with a mannitol-like passive diffusion. It enjoys a kinetic advantage for tissue entry over free glucose in the medum by virtue of the position of the site where it is formed, i.e inside the unstirred water layer and near normal entry portals. No special hydrolase-related transport system, like the one proposed for disaccharides, needs to be considered to account for our findings.     

The function of a hydrogen peroxide-detecting electroenzymatic glucose electrode is markedly impaired in human sub-cutaneous tissue and plasma

Electroenzymatic glucose sensors implanted into sub-cutaneous (s.c.) tissue of human subjects and experimental animals exhibit lower sensitivities to glucose than in buffer solutions before implantation. The mechanism of the decrease of sensitivity is not known. Sensors used in this study were fabricated from platinum wires (diameter 0.125 mm) with covalently bound glucose oxidase at the tip of the wire. After coating the tip with polyurethane, wires were placed into 27 gauge steel needles. Sensors were operated potentiostatically at 700 mV against Ag/AgCl pseudo-reference electrodes. These sensors were implanted s.c. in 6 diabetic patients for 7 h. In 4 patients, sensors were responsive to successive increases of plasma glucose levels. Mean sensitivity to glucose in s.c. tissue was 29% of in vitro sensitivity. In 2 patients there was a sudden decrease of sensor currents, unrelated to glucose, shortly after implantation. Sensors were inhibited in human plasma to a similar extent. When sensors were exposed to native plasma and to plasma ultrafiltrate (mol. wt. <10 kDa) for 10 h, identical decreases of signals were found. Exposure to dialysed plasma (mol. wt. >12 kDa) caused much less decrease of sensor signals. Losses of sensor sensitivities to glucose in s.c. tissue and in plasma were totally reversible upon re-exposure of sensors to buffer solutions. We conclude that sensor inactivation in plasma and possibly in s.c. tissue is caused by low molecular weight substances not retained by the polyurethane membrane.     

A high affinity, highly selective ligand for the delta opioid receptor: [3H]-[D-Pen2, pCl-Phe4, d-Pen5]enkephalin

Binding characteristics of a new, conformationally constrained, halogenated enkephalin analogue, [3H]-[D-penicillamine2, pCl-Phe4, D-penicillamine5]enkephalin ([3H]pCl-DPDPE), were determined using homogenized rat brain tissue. Saturation binding studies at 25 degrees C determined a dissociation constant (Kd) of 328 +/- 27.pM and a receptor density (Bmax) of 87.2 +/- 4.2 fmol/mg protein. Kinetic studies demonstrated biphasic association for [3H]pCl-DPDPE, with association rate constants of 5.05 x 10(8) +/- 2.5 x 10(8) and 0.147 +/- 10(8) +/- 0.014 x 10(8) M-1 min-1. Dissociation was monophasic with a dissociation rate constant of 2.96 x 10(-3) +/- 0.25 x 10(-3) min-1. The average Kd values determined by these kinetic studies were 8.4 +/- 2.7 pM and 201 +/- 4 pM. Competitive inhibition studies demonstrated that [3H]pCl-DPDPE has excellent selectively for the delta opioid receptor. [3H]pCl-DPDPE binding was inhibited by low concentrations of ligands selective for delta opioid receptor relative to the concentrations required by ligands selective for mu and kappa sites. These data show that [3H]pCl-DPDPE is a highly selective, high affinity ligand which should be useful in characterizing the delta opioid receptor.     

First step toward near-infrared continuous glucose monitoring: in?vivo evaluation of antibody coupled biomaterials

Continuous glucose monitoring (CGM) is crucial in diabetic care. Long-term CGM systems however require an accurate sensor as well as a suitable measuring environment. Since large intravenous sensors are not feasible, measuring inside the interstitial fluid is considered the best alternative. This option, unfortunately, has the drawback of a lag time with blood glucose values. A good strategy to circumvent this is to enhance tissue integration and enrich the peri-implant vasculature. Implants of different optically transparent biomaterials (poly(methyl-methacrylate) [PMMA] and poly(dimethylsiloxane) [PDMS]) – enabling glucose monitoring in the near-infrared (NIR) spectrum – were surface-treated and subsequently implanted in goats at various implantation sites for up to 3 months. The overall in vivo biocompatibility, tissue integration, and vascularization at close proximity of the surfaces of these materials were assessed. Histological screening showed similar tissue reactions independent of the implantation site. No significant inflammation reaction was observed. Tissue integration and vascularization correlated, to some extent, with the biomaterial composition. A modification strategy, in which a vascular endothelial-cadherin antibody was coupled to the biomaterials surface through a dopamine layer, showed significantly enhanced vascularization 3 months after subcutaneous implantation. Our results suggest that the developed strategy enables the creation of tissue interactive NIR transparent packaging materials, opening the possibility of continuous glucose monitoring.     

Transport characteristics of a novel local drug delivery system using nordihydroguaiaretic acid (NDGA)-polymerized collagen fibers

The use of nordihydroguaiaretic acid (NDGA)-polymerized collagen fibers as a novel local drug delivery system is introduced. The drug loading of these biocompatible fibers is illustrated with the anti-inflammatory agents dexamethasone and dexamethasone 21-phosphate. Capillary zone electrophoresis was used to measure the amount of drug released from the fibers into phosphate buffered saline with time. From these measurements and the use of a mathematical model, we were able to determine the diffusion coefficients for dexamethasone (D = 1.86 x 10(-14) m2/s) and dexamethasone 21-phosphate (D = 2.36 x 10(-13) m2/s) in the NDGA collagen fibers. These values have not been previously reported. These fibers can be used to load other agents as well. The diffusion coefficient of any agent loaded in these fibers can be determined using the techniques and mathematical method described. The rate of drug release from the fibers can be controlled using a PLGA coating. The overall importance of this paper is the potential broad application of this novel drug delivery system for the treatment of various human diseases.     

Effect of salicyclic acid on gluccorticoid receptor in cultured fibroblasts derived from rat carrageenin granuloma.

The binding activity of [3H]dexamethasone to the specific receptor was studied in the cytoplasmic fraction of a established fibroblast line derived from rat carrageenin granuloma in culture condition. Specific receptor to dexamethasone was demonstrated. Scatchard analysis revealed a single class of binding sites with a dissociation constant for [3H]dexamethasone of 3.64 - 10(-8) M and a concentration of binding sites of 0.825 pmol per mg cytosol protein. The number of cytoplasmic binding sites per cell was calculated at 1.15 - 10(5). Total binding activity to [3H]dexamethasone of the cytoplasmic fraction was enhanced when the cells were cultured in a medium containing salicylic acid was at 37 degrees C. The maximum enhancement was seen at the concentration of 10(-3)M and in 3h treatment of salicylic acid. This enhancement by salicylic acid was lost when cycloheximide was added to the culture medium at the same time. If salicyclic acid was added to the cell free system, it showed no effect on the binding activity. The other non-steroidal anti-inflammatory drugs; phenylbutazone and indomethacin,also enhanced the total binding activity to [3H]dexamethasone of the cytoplasmic fraction at the concentration of 2 - 10(-5) M and 2 - 10(-7) M, respectively.     

Glucocorticoids "receptors" in human gingiva.

Specific binding of [3H]-dexamethasone by the cytoplasmic fraction of normal human gingiva was carried out. Different concentrations of [3H]-dexamethasone were used with or without cold dexamethasone to determine the binding. Binding data by Scatchard plot yielded a straight line indicating a single class of specific receptors with equilibrium dissociation constant of 1.39 x 10(-9) M and B max of 80 fmol/mg protein. The proteins satisfied the high affinity and low capacity requirements of receptor.     

Biosensors for in vivo glucose measurement: can we cross the experimental stage

The development of in vivo working glucose sensors needs two decades, so far. The availability of long term functional implantable biosensors for continuous glucose measurings is a basic prerequisite for the individualized optimum insulin treatment of diabetics. Enzymatic electrochemical sensors are described which realize a functional stability over more than 2 years in vitro, however their function in vivo is limited due to certain bioincompatibility expressed by inflammation of the surrounding tissue, exudates, and immun reactions. The paper reflects an overview concerning different sensor covering materials used as more or less suitable diffusion membranes. From experimental studies in animals and human volunteers conclusions are drawn for further developmental steps of biosensors for in vivo use and for the applicability of glucose sensors for transient diagnostic purposes and as a basis for glucose controlled therapeutic measures. The results demonstrate that further progress aimed at long term biostability of implanted biosensors needs to solve technological problems and the serial production of sensors with really comparable qualities as a prerequisite for clinical trials.     

Long-term Carotid Access in the Goat: Observations on Application of a Totally Implantable Catheter System

The investigation of metabolism of individual organs and tissues in vivo requires simultaneous measurements of arterio-venous concentration differences. Long-term metabolic studies require frequent and convenient access to veins and arteries of the relevant organs or tissues, and if blood samples are obtained by repeated veni — or arteriopuncture, difficulties may arise due to vessel thrombosis and an increasing apprehension of the goat during time. Arterial punctures are considered more painful than venipuncture, according to human patients. In the conscious goat, arteriopuncture may induce bleating, dilatation of the pupil, increased heart rate, blood glucose and free fatty acid levels, which may bias the results of a metabolic study (Linzell 1963). Furthermore, it is important to avoid situations which contribute negatively to the welfare of the goat. Establishment of permanent exteriorized carotid loops does not eliminate these problems and has a disadvantage of being vulnerable to injury. Jha et al. (1961) reported that 3 out of 10 goats died from haemorrhages, due to loop injury. The carotid artery may surgically be relocated to a subcutaneous position, which makes it less vulnerable, but also less accessible for percutaneous puncture (Tavenor 1969). Our experience with such subcutaneous relocated artery loops is that scar tissue formation is so strong after being catheterized 3-4 times, that further catheterization requires specialized assistance. The artery is eventually closed by fibrous tissue. An alternative to the repeated percutaneous puncture of subcutaneously relocated artery loops is the establishment of chronic indwelling arterial catheters, where the tip of the catheter is exteriorized to the skin. This method is well described (Huntington et al. 1989), but requires a dressing of the exteriorisation site, and restrain of the goat, in order to protect the tip from being pulled out. Therefore, we describe a modification of a totally implantable catheter system, developed for human intra-arterial long-term delivery of drugs, for permanent catheterization of the common carotid artery of goats. The entire access system of this catheter is under the skin.     

Location of platelet activating factor binding in rat liver.

Autoradiographs of tissue slices from livers perfused with 1 x 10(-9) M-1-O-[3H]octadecyl-2-acetyl-sn-glycero-3-phosphocholine ([ 3H]18:0-sn-3-AGEPC) indicate that binding of this agonist is localized in the portal venules in anterograde perfused livers, and in the central venules in retrograde perfused livers. The pattern of silver grains in anterograde perfused liver was not affected significantly by prior exposure to 100-fold excesses of unlabelled 16:0- or 18:0-sn-3-AGEPC, 16:0-sn-1-AGEPC, or a 1000-fold excess of U.66985. [3H]18:0-sn-3-lyso-GEPC produced the same pattern of binding as the acetylated analogue. Measurement of glucose release stimulated by 16:0-sn-3-AGEPC demonstrated that the retrograde perfused liver was nearly 1000-fold less sensitive to this compound than the anterograde perfused liver. Exposure of the livers to bovine serum albumin prior to 5 x 10(-11) M-[3H]18:0-sn-3-AGEPC resulted in inhibition of stimulated glucose release, and decreased both the amount of label retained in the livers and the amount of silver grains over the portal sinusoidal cells without affecting the amount of grains seen over all other regions of the liver. Glucose release from primary monolayer cultures of hepatocytes or suspensions of liver slices was not stimulated by 16:0-sn-3-AGEPC. The results suggest that specific binding of [3H]18:0-sn-3-AGEPC is restricted to the portal side of the liver microvasculature, the majority of binding is nonspecific, and the biological response to AGEPC requires an intact and perfused vasculature.     

Evidence for the biosynthesis of DHEA from cholesterol by first-trimester human placental tissue: source of androgens.

With a view to establishing whether first-trimester human placentas possess the ability to synthesize DHEA from cholesterol, homogenates of this tissue obtained from two groups of women undergoing elective termination of normally progressing pregnancy between 10 - 12 weeks gestation (n = 5, age 23 - 29 years and n = 5, age 21 - 27 years) were incubated separately with [26-(14)C]cholesterol for the generation of [14C]isocaproic acid + pregnenolone and [7n-3H]pregnenolone for the biosynthesis of [3H]DHEA. Controls consisted of homogenates heated in a boiling water bath for 10 min. Using the reverse-isotope dilution analysis, desmolase efficiency expressed as mean specific activity of [14C]isocaproic acid varied from 282 to 725 dpm/mmol, while that of 17 alpha-hydroxylase and steroid C-17,20-lyase, catalyzed conversion of [7n-3H]pregnenolone to [3H]DHEA varied from 3498 to 26 258 dpm/mmol. The corresponding efficiencies of enzymicconversion varied between 5.8 x 10( -2) and 1.5 x 10( -1) % for [14C]isocaproic acid, but between 5.5 x 10( -2) and 4.1 x 10( -1) % for [3H]DHEA. No such metabolite was evident in the controls of heat-denatured homogenates. These are the first study results to demonstrate that early placentas are capable of converting cholesterol to pregnenolone to DHEA, contrary to the widely held concept of DHEA production by fetal and maternal adrenal glands. This finding has important physiological implications and could provide a new dimension to the concept of fetoplacental steroidogenesis.     

The glucose transport in retinal pigment epithelium is via passive facilitated diffusion

The glucose transport across the bovine retinal pigment epithelium (RPE) was studied in a modified Ussing chamber. Unidirectional fluxes were recorded with radioactive tracers L-[14C]-glucose (LG) and 3-O-methyl-D-[3H]-glucose (MDG). There was no significant difference between the unidirectional MDG fluxes (retina to choroid, and choroid to retina directions) with or without ouabain. The effects of two glucose transporter inhibitors, phloretin and cytochalasin B, on the glucose fluxes from choroid to retina cells were also investigated. The MDG flux was found to be inhibited by 45.5% by phloretin (10(-4) M) and 87.4% by cytochalasin B (10(-4) M). These inhibitory characteristics resembled the facilitated diffusion mode of glucose transport. The glucose transporter protein in the plasma membrane of RPE was located by means of photolabeling [3H]-cytochalasin B. The labeled plasma membrane enriched fraction was analysed by SDS-PAGE. The glucose transporter of bovine RPE was found to have a molecular weight range of 46-53 kDa. The molecular weight range of this transporter protein agreed with those of facilitated glucose transporters in other tissues indicating a molecular similarity between them. The results indicated that the glucose transport across the RPE is via passive facilitated diffusion.     

Positron emission tomography with [18F]fluorodeoxyglucose to evaluate neutrophil kinetics during acute lung injury

We measured neutrophil glucose uptake with positron emission tomographic imaging and [18F]fluorodeoxyglucose ([18F]FDG-PET) in anesthetized dogs after intravenous oleic acid-induced acute lung injury (ALI; OA group, n = 6) or after low-dose intravenous endotoxin (known to activate neutrophils without causing lung injury) followed by OA (Etx + OA group, n = 7). The following two other groups were studied as controls: one that received no intervention (n = 5) and a group treated with Etx only (n = 6). PET imaging was performed 1.5 h after initiating experimental interventions. The rate of [3H]deoxyglucose ([3H]DG) uptake was also measured in vitro in cells recovered from bronchoalveolar lavage (BAL) performed after PET imaging. Circulating neutrophil counts fell significantly in animals treated with Etx but not in the other two groups. The rate of [18F]FDG uptake, measured by the influx constant Ki, was significantly elevated (P < 0.05) in both Etx-treated groups (7.9 +/- 2.6 x 10(-3) ml blood x ml lung(-1) x min(-1) in the Etx group, 9.3 +/- 4.8 x 10(-3) ml blood x ml lung(-1) x min(-1) in the Etx + OA group) but not in the group treated only with OA (3.4 +/- 0.8 x 10-3 ml blood x ml lung(-1) x min(-1)) when compared with the normal control (1.6 +/- 0.4 x 10(-3) ml blood x ml lung(-1) x min(-1)). [3H]DG uptake was increased (73 +/- 7%) in BAL neutrophils recovered from the Etx + OA group (P < 0.05) but not in the OA group. Ki and [3H]DG uptake rates were linearly correlated (R2 = 0.65). We conclude that the rate of [18F]FDG uptake in the lungs during ALI reflects the state of neutrophil activation. [18F]FDG-PET imaging can detect pulmonary sequestration of activated neutrophils, despite the absence of alveolar neutrophilia. Thus [18F]FDG-PET imaging may be a useful tool to study neutrophil kinetics during ALI.     

Efflux of a suppressive neurotransmitter, GABA, across the blood-brain barrier

In this study, GABA efflux transport from brain to blood was estimated by using the brain efflux index (BEI) method. [3H]GABA microinjected into parietal cortex area 2 (Par2) of the rat brain was eliminated from the brain with an apparent elimination half-life of 16.9 min. The blood-brain barrier (BBB) efflux clearance of [3H]GABA was at least 0.153 mL/min/g brain, which was calculated from the elimination rate constant (7.14 x 10(-2) x min(-1)) and the distribution volume in the brain (2.14 mL/g brain). Direct comparison of the apparent BBB influx clearance [3H]GABA (9.29 microL/min/g brain) and the apparent efflux clearance (153 microL/min/g brain) indicated that the efflux clearance was at least 16-fold greater than the influx clearance. In order to reduce the effect of metabolism in the neuronal cells following intracerebral microinjection, we determined the apparent efflux of [3H]GABA in the presence of nipecotic acid, a GABA transport inhibitor in parenchymal cells, using the BEI method. Under such conditions, the elimination of [3H]GABA across the BBB showed saturation and inhibition by probenecid in the presence of nipecotic acid. Furthermore, the uptake of [3H]GABA by MBEC4 cells was inhibited by GABA, taurine, beta-alanine and nipecotic acid in a concentration-dependent manner. It is likely that GABA inhibits the first step in the abluminal membrane uptake by brain endothelial cells, and that probenecid selectively inhibits the luminal membrane efflux transport process from the brain capillary endothelial cells based on the in vivo and in vitro evidence. The BBB acts as the efflux pump for GABA to reduce the brain interstitial fluid concentration.