Abstract
Brain penetration of clonidine, an alpha-2 adrenoceptor agonist, was studied using an in vitro cell culture system consisting of primary cultures of porcine brain capillary endothelial cells. Uptake of clonidine was measured as a function of its concentration in the incubation mixture. Saturation of uptake was apparent and could be described by Michaelis-Menten-type kinetics (K M = 1.34 mM;V max = 0.099 nmol/min/cm2). Saturation was not observed at a low temperature (4°C). Transendothelial transport experiments revealed that translocation of clonidine cannot be attributed solely to paracellular leakage. Uptake was reduced at low extracellular pH or by using an incubation buffer that contained the K+ionophore valinomycin. Time-dependent uptake of clonidine and transendothelial transport were slower than expected considering the high octanol-to-buffer partition coefficient of this compound. On the basis of transendothelial transport experiments, we concluded that the carrier system responsible for active transport of clonidine is located at both the apical and the basolateral membrane domain.
Clonidine has been used for decades as an effective agent in long-term antihypertensive therapy and in the acute management of hypertensive crisis. Clonidine has been demonstrated to produce a large variety of pharmacological effects due to activation of alpha-2 adrenergic receptors present in the central nervous system and other organs (Bloor, 1993). In recent years, clonidine has been increasingly used for the purpose of enhancing systemic or neuraxial anesthesia (Hayashi and Maze, 1993; Kauppila et al., 1991; Reddyet al., 1980; Spaulding et al., 1979; Sullivanet al., 1987; Yaksh, 1985). Clonidine has also been shown to be an efficacious drug to control overshoot of sympathetic activity in patients suffering from withdrawal of alcohol (Yam et al., 1992), opioids (Gold et al., 1980) or benzodiazepines. These beneficial central effects were observed after the application of high doses of intravenous clonidine (Tryba et al., 1993). Thus, clonidine is able to cross the blood-brain barrier after intravenous injection and to interact with alpha-2 adrenergic receptors present in the central nervous system. We therefore evaluated the contribution of the blood-brain barrier in controlling clonidine transfer from blood to its central nervous system receptor sites using an in vitro cell culture system consisting of primary cultures of porcine brain capillary endothelial cells.
Methods
Materials.
Clonidine was from Fluka (Buchs, Switzerland). [3H]Aminoclonidine was from DuPont-New England Nuclear (Boston, MA). Octanol-to-100 mM phosphate buffer, pH 7.0, partition coefficients were determined after equilibration for 24 hr at room temperature using isotope-labeled tracers (Amersham, Buckinghamshire, UK, or DuPont-New England Nuclear).
Cell cultures.
Primary cultures of porcine brain capillary endothelial cells were prepared as described (Audus and Borchardt, 1986) with the following modifications: Cortical gray matter from six freshly obtained porcine brains was minced and incubated in MEM (Sigma Chemical Co., St. Louis, MO) containing 0.5% dispase (Boehringer-Mannheim Biochemica, Mannheim, Germany) for 2 hr. Cerebral microvessels were obtained after centrifugation in MEM containing 13% dextran (Sigma). The microvessels were subsequently incubated in MEM containing 1 mg/ml collagenase-dispase (Boehringer-Mannheim) for 4.25 hr. The resulting cell suspension was supplemented with 10% horse serum and filtered through a 150-μm nylon mesh. Brain capillary endothelial cells were isolated on a continuous 50% Percoll gradient (Pharmacia, Uppsala, Sweden) (centrifugation at 1000 ×g for 10 min). Isolated endothelial cells were filtered through a 35-μm nylon mesh before seeding with a density of 100,000 cells/cm2 onto collagen/fibronectin-coated (Boehringer-Mannheim) 24-well cell culture plates (uptake assays) or a density of 200,000 cells/cm2 onto polycarbonate membranes (transendothelial transport; see below). Cells were cultured under standard cell culture conditions (Audus and Borchardt, 1986) [cell culture medium: 45% MEM, 45% Ham’s F-12, 100 μg/ml streptomycin, 100 μg/ml penicillin G, 100 μg/ml heparin, 13 mM NaHCO3and 20 mM HEPES [all from Sigma] containing 10% heat-inactivated horse serum [GIBCO BRL, Basel, Switzerland]).
Uptake assays.
Uptake assays were performed at 20°C using confluent monolayers of porcine brain capillary endothelial cells at day 10. Cells were grown on 24-well cell culture plates. The surface area was 2 cm2/well. Cells were washed using transport buffer, which consisted of 122 mM NaCl, 3 mM KCl, 1.4 mM CaCl2, 1.2 mM MgSO4, 4 mMd-glucose, 10 mM HEPES, 25 mM NaHCO3 and 0.4 mM K2HPO4, pH 7.4. Where indicated, NaCl was substituted with choline chloride, and d-glucose was adjusted to 0 or 20 mM. The reaction was initiated by the addition of 240 μl of transport buffer containing 0.3 μCi of3H-labeled tracer of the respective substrate, sufficient unlabeled substrate to bring the medium to the desired final concentration and 0.3 μCi of the extracellular marker [14C]sucrose.
As a highly hydrophobic reference substance propranolol was used. Uptake of propranolol was determined in uptake buffer as well as in cell culture medium containing 10% horse serum. There was no difference in uptake excluding the possibility of unspecific binding of propranolol to plastic. For inhibition studies, cells were preincubated for 15 min with the respective inhibitor. Stock solutions of inhibitors that were poorly soluble in buffer were prepared using DMSO or ethanol. In this case, the final concentration of DMSO or ethanol in the assay did not exceed 1% (v/v) or 0.5% (v/v), respectively. Control experiments were performed in absence and presence of the respective solvent. DMSO or ethanol in the concentrations used had no detectable effect on the measured cell parameters. Lactated dehydrogenase release was negligible, and uptakes of the extracellular marker sucrose and of phenylalanine, which was used as marker for carrier-mediated transport, did not change. Incubations were terminated after 5 min by rapid removal of the incubation medium followed by washing the cells with ice-cold transport buffer. Cells were then removed from the wells by incubation for 10 min in trypsin (0.25%) and subsequently transferred to scintillation vials. Scintillation fluid was added, cells were solubilized overnight and the amount of radiolabeled substrate taken up by the cells was determined by scintillation counting.
Kinetic experiments.
Uptake of clonidine was measured as a function of its concentration in the incubation mixture. Incubations were performed at room temperature or 4°C. The range of concentrations used varied from 1 μM to 2 mM. For data representation and to obtain estimates of kinetic parameters, a nonlinear regression program was used (Microcal Origin version 3.5).
Transendothelial transport.
For the study of transendothelial transport, up to six polycarbonate membranes (Snap-Well System, Costar, Cambridge, MA) with a confluent monolayer of porcine brain endothelial cells were mounted in a corresponding number of side-by-side diffusion cells (Costar). Both sides of the diffusion cells were filled with prewarmed transport buffer. The entire system was maintained at a constant temperature (37°C) and was supplied with 5% CO2/95% oxygen. At time t = 0, the isotope-labeled compound to be studied was added to the donor chamber. In defined time intervals, samples were drawn from the acceptor chamber and analyzed by scintillation counting. The acceptor chamber volume was readjusted with assay buffer after each sample was taken, and counts from acceptor side samples were corrected for the amount of radioactivity removed by previous sampling. The initial rate of transport was calculated from a linear regression. Pappvalues were calculated according to Papp = dQ/dt·1/A/C0, where dQ/dt is the rate of translocation, A is the surface of the polycarbonate membrane and C0 is the initial concentration of the labeled drug (cm/min).
Results
Uptake vs. time.
Uptake of clonidine into cultured cerebral capillary endothelial cells was measured at various time points up to 50 min (fig. 1). Propranolol was used as a reference substance. In contrast to uptake of propranolol, uptake of clonidine was nearly linear and did not reach saturation within 50 min. The finding that clonidine penetrated brain capillary endothelial cells to a lesser degree than propranolol was unexpected in view of the high hydrophobicity of clonidine. The octanol-to-100 mM phosphate buffer, pH 7.4, partition coefficients were 0.0011 ± 0.0003 (n = 3) for [14C]sucrose, 12.2091 ± 0.6802 (n = 3) for [3H]propranolol and 86.7262 ± 1.3361 (n = 5) for [3H]clonidine (values represent mean ± S.E.M. ofn experiments).
Carryover of extracellular radioactivity through the washing stages and diffusion of sucrose into the cells were minimal. Uptake of the extracellular marker sucrose was followed over a time period of 50 min (fig. 1); <0.07% of the applied dose of sucrose was recovered. This value did not change with incubation time. We therefore used cell-associated sucrose in this and following experiments as an indicator for the intactness of the cell monolayer.
In many cases, unspecific binding of a hydrophobic drug to the plastic of the cell culture dish can be prevented by the addition of 10% serum to the incubation buffer. This procedure, however, was not necessary for clonidine, the extracellular marker sucrose or the reference substance propranolol. There was no difference between the uptake of clonidine and sucrose using cell culture medium or transport buffer (no difference between results by two-tailed Student’s t test; P > .3, n = 4). In addition, direct binding of clonidine to plastic was marginal; recovery from a cell culture dish that was not precoated with cell culture medium and did not contain cells was 98.4 ± 0.3% (n = 4) for sucrose and 92.9 ± 0.6% (n = 4) for clonidine.
Kinetic experiments.
To determine whether carrier-mediated transport processes were contributing to the uptake of clonidine, kinetic experiments were performed. When the initial rates of uptake were plotted as a function of medium concentration, it became evident that overall uptake (fig. 2A, top curve) consisted of two components: (1) a linear term, which could be attributed to passive diffusion, and (2) a term that indicated Michaelis-Menten-type saturation of uptake (fig. 2B). To obtain estimates of kinetic parameters, a nonlinear regression computer program was used. The estimates for the Michaelis constants revealed aKM value of 1.34 mM and a Vmax value of 0.099 nmol/min/cm2.
To demonstrate further that uptake of clonidine into brain capillary endothelial cells involves a carrier, concentration-dependent uptake of clonidine was measured at 4°C (fig. 2A, bottom curve). At a low temperature, uptake was considerably reduced and no saturation of uptake could be observed. The data could be described by a simple linear regression. The coefficient of regression (R) was .9997.
Mechanism of uptake.
The nature of the clonidine transport mechanism was further investigated through transendothelial transport experiments (table 1). Again, translocation of clonidine was low. Earlier experiments revealed a Papp ratio of propranolol to sucrose of 3.2 (data not shown). There was no difference in transport of clonidine from apical to basolateral or in the reverse direction, suggesting that a putative carrier system is not polarized. In another set of experiments, the concentration of glucose in the transport buffer was elevated. As a result, paracellular leakage of sucrose increased by a factor of ∼2. Under these conditions, Papp values of clonidine and sucrose were not significantly different (table 1). We concluded from this experiment that overall transendothelial transport of clonidine consists of a combination of passive diffusion and an additional transport mechanism that cannot be attributed to paracellular leakage.
Further characterization of the carrier mechanism responsible for transport of clonidine was done using different incubation conditions or specific effectors (table 2). Energy depletion of the cells did not significantly influence uptake of clonidine. Uncouplers of oxidative phosphorylation, such as carbonyl cyanidem-chlorophenyl-hydrazone and 2,4-dinitrophenol, as well as the respiratory chain inhibitor rotenone were without effect. Thus, transport of clonidine is not a direct energy-requiring process. However, uptake was reduced by decreasing the pH and in presence of the K+ ionophore valinomycin. Possible contribution of an Na+/H+ antiporter toward clonidine transport could be excluded using the specific inhibitor cimetidine and through depletion of extracellular Na+ by substitution of NaCl with choline chloride in the transport buffer.
Discussion
It was the aim of the present study to determine the mechanisms of transport of clonidine through the blood-brain barrier. Three lines of evidence suggest that transport of clonidine at the blood-brain barrier occurs via a carrier-mediated transport mechanism. First, uptake of clonidine by brain capillary endothelial cells showed self-inhibition (saturation) and could be described by a Michaelis-Menten-type kinetics (KM = 1.34 mM; V max = .099 nmol/min/cm2). Second, at a low temperature, uptake was considerably reduced and could be attributed to passive diffusion only. Third, transendothelial transport experiments suggest a contribution of active transport toward translocation of clonidine. In this case, transport experiments were performed under physiological glucose concentrations (4 mM) or in presence of 20 mM glucose. High concentrations of glucose have been reported to increase the permeability of the epithelial paracellular pathway through reversible dilatation of the paracellular space within tight junctions (Atisook and Madara, 1991; Fricker and Drewe, 1995;Madara et al., 1993). Under these conditions, Papp values of clonidine and the extracellular marker sucrose were nearly identical. This contrasts with the situation with 4 mM glucose; translocation of clonidine could not be attributed solely to paracellular leakage.
Further characterization of the putative carrier mechanism by transendothelial transport experiments revealed no difference between apical-to-basolateral transport or basolateral-to-apical transport. We therefore concluded that the carrier system responsible for transport of clonidine is localized on the apical as well as the basolateral membrane domain of cultured brain capillary endothelial cells. Uptake of clonidine into brain capillary endothelial cells was not a direct energy-requiring process and was reduced in the presence of the K+ ionophore valinomycin or at low extracellular pH. This marked pH sensitivity could have several causes. First, transport of clonidine could be reduced on protonation of the drug: pKaa values are 8.10 for clonidine hydrochloride and 5.96 for clonidine base (clonidine data sheet by Boehringer Ingelheim). Second, activity of the putative carrier system could be influenced by extracellular pH, and, third, translocation of clonidine could be coupled to translocation of H+ from the inside of the cell to its outside. A carrier system with this characteristic and an affinity for clonidine has been described in human placenta (Ganapathy et al., 1986) and opossum kidney cells (Ramamoorthy et al., 1991). The carrier has been identified as an Na+/H+ antiporter responsible for regulation of intracellular pH in vertebrate cells (Frelin et al., 1988). Clonidine has been found to competitively inhibit Na+ translocation (Ganapathy et al., 1986). To test the hypothesis that clonidine might be a substrate for the Na+/H+ antiporter in brain capillary endothelial cells, uptake of clonidine was determined in the absence of extracellular Na+ and the presence of the specific inhibitor cimetidine (Ganapathy et al., 1986). However, in both cases, uptake of clonidine was not influenced, thus making unlikely a possible contribution of the Na+/H+exchanger toward clonidine transport in brain capillary endothelial cells.
Transport of clonidine was compared with transport of propranolol, a reference substance that is considered to cross the blood-brain barriervia a passive diffusion process (Dehouck et al., 1992). Surprisingly, overall uptake and the Papp of clonidine were lower than expected in view of its high hydrophobicity and the involvement of a carrier system. One explanation for the lower-than-expected rate of transendothelial transport could be accumulation and resultant “trapping” of very hydrophobic substances in the hydrophobic interior of the lipid bilayer of cells (Seelig et al., 1994). This observation is corroborated by clinical data. High-dose intravenous clonidine is needed to achieve central effects; in this case, >20 mg/day of the drug is used (Trybaet al., 1993), whereas ≤∼0.5 mg is needed for neuraxial analgesia (Goudas, 1995). In addition, there is experimental evidence that the bioavailability of clonidine in cerebrospinal fluid after intravenous injection may be almost 3 orders of magnitude lower than that after epidural administration (Castro and Eisenach, 1989). There also is a significant difference between epidural and intravenous clonidine in reducing total electroencephalographic power during general anesthesia (De Kock et al., 1992). Another explanation for the poor transport of clonidine would be the presence of a specific carrier system extruding clonidine from endothelial cells. P-glycoprotein, which confers the multidrug resistance phenotype to tumor cells, is such a transport system (Endicott and Ling, 1989;Roninson, 1992). P-glycoprotein has a broad substrate specificity and was localized at the luminal side of capillary endothelial cells in both gray matter of the brain and primary cultured bovine brain capillary endothelial cells (Tsuji et al., 1992) and the cell culture model used in the present study (Huwyler et al., 1996). However, we could exclude the possibility of an involvement of P-glycoprotein by demonstrating that vinblastine and verapamil, two well-established inhibitors of P-glycoprotein, had no effect on uptake of clonidine (data not shown).
In summary, a carrier system for clonidine has been identified using anin vitro cell culture system consisting of cultured porcine brain capillary endothelial cells. Transport was saturable and sensitive to temperature, pH and extracellular K+. Overall transport of clonidine was lower than expected despite this involvement of a carrier. Our data confirm clinical data indicating that clonidine is able to cross the blood-brain barrier, although relatively high doses of intravenous clonidine are needed to achieve central effects.
Acknowledgments
We would like to thank U. Behrens for technical assistance.
Footnotes
-
Send reprint requests to: Dr. J. Huwyler, University Hospital, Department of Anesthesia and Research, Hebelstrasse 20, CH-4031 Basel, Switzerland. E-mail:Huwylerj{at}ubaclu.unibas.ch
-
↵1 This work was supported by Swiss National Science Foundation Grant 32–42179.94.
- Abbreviations:
- DMSO
- dimethylsulfoxide
- MEM
- minimum essential medium
- Papp
- coefficient of permeability
- HEPES
- 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid
- Received October 15, 1996.
- Accepted March 3, 1997.
- The American Society for Pharmacology and Experimental Therapeutics