SEARCH

SEARCH BY CITATION

Keywords:

  • Blood-brain barrier;
  • Mouse brain endothelial cells;
  • Glutathione;
  • Transport;
  • Membrane vesicles;
  • Sodium dependence

Abstract

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL PROCEDURES
  4. GSH transport by MBEC-4 cells
  5. Effect of extracellular GSH on intracellular GSH
  6. RNA isolation, oocyte preparation, and transport
  7. RESULTS
  8. GSH and GGT levels in MBEC-4 cells
  9. Na+ dependency of GSH uptake in MBEC-4 cells
  10. Na+-mediated increase in cellular GSH
  11. Expression of GSH uptake in Xenopus laevis oocytes
  12. Characteristics of GSH uptake in luminal membrane vesicles
  13. DISCUSSION
  14. Acknowledgements
  15. References

Abstract: We have previously shown GSH transport across the blood-brain barrier in vivo and expression of transport in Xenopus laevis oocytes injected with bovine brain capillary mRNA. In the present study, we have used MBEC-4, an immortalized mouse brain endothelial cell line, to establish the presence of Na+-dependent and Na+-independent GSH transport and have localized the Na+-dependent transporter using domain-enriched plasma membrane vesicles. In cells depleted of GSH with buthionine sulfoximine, a significant increase of intracellular GSH could be demonstrated only in the presence of Na+. Partial but significant Na+ dependency of [35S]GSH uptake was observed for two GSH concentrations in MBEC-4 cells in which γ-glutamyltranspeptidase and γ-glutamylcysteine synthetase were inhibited to ensure absence of breakdown and resynthesis of GSH. Uniqueness of Na+-dependent uptake in MBEC-4 cells was confirmed with parallel uptake studies with Cos-7 cells that did not show this activity. Molecular form of uptake was varified as predominantly GSH, and very little conversion of [35S]cysteine to GSH occurred under the same incubation conditions. Poly(A)+ RNA from MBEC expressed GSH uptake with significant (∼40-70%) Na+ dependency, whereas uptake expressed by poly(A)+ RNA from HepG2 and Cos-1 cells was Na+ independent. Plasma membrane vesicles from MBEC were separated into three fractions (30, 34, and 38% sucrose, by wt) by density gradient centrifugation. Na+-dependent glucose transport, reported to be localized to the abluminal membrane, was found to be associated with the 38% fraction (abluminal). Na+-dependent GSH transport was present in the 30% fraction, which was identified as the apical (luminal) membrane by localization of P-glycoprotein 170 by western blot analysis. Localization of Na+-dependent GSH transport to the luminal membrane and its ability to drive up intracellular GSH may find application in the delivery of supplemented GSH to the brain in vivo.

The blood-brain barrier (BBB) restricts the movement of certain solutes between the blood and the brain. It is well established that the brain endothelial cells are responsible for this selective barrier function. The tight junctions and low rate of pinocytosis are responsible for the exclusion of proteins by the brain. However, low molecular weight, lipid-soluble solutes can diffuse through cell membranes and lipid-insoluble solutes can be transported through selective carrier-mediated transport systems and channels. For example, such transport systems for hexoses and basic and acidic amino acids are known (Oldendorf, 1971). As in epithelia from kidney tubules and intestinal mucosa, the BBB also exhibits active transcellular transport due to selective or polar distribution of transport proteins between the opposite surfaces of the cells (Betz et al., 1980).

It is well known that glutathione (GSH), an endogenous antioxidant, is important for brain function. GSH deficiency is known to be associated with a number of neurological diseases (Orlowsky and Karnowsky, 1976; Meister and Anderson, 1983). Our laboratory has been interested in understanding the metabolism of GSH in the brain. We have shown that GSH is transported across the BBB in the rat (Kannan et al., 1990) and the guinea pig (Zlokovic et al., 1994). Transport was dissociated from γ-glutamyltranspeptidase (GGT)-mediated hydrolysis and resynthesis, was carrier mediated, and exhibited inhibitor specificity (Kannan et al., 1990, 1992; Zlokovic et al., 1994). Evidence for the expression of Na+-dependent and Na+-independent GSH transporters in bovine brain capillary mRNA-injected oocytes was also presented (Kannan et al., 1996, 1998). We hypothesized that a Na+/GSH co-transporter might be mediating uphill transport of GSH from blood plasma into brain, whereas the Na+-independent GSH transporter might function mainly as a facilitative efflux transporter.

To better understand the biochemical features and regulation of GSH transport across the brain endothelium, we turned to a culture model of the BBB. To this end, we used an immortalized mouse brain endothelial cell line (MBEC-4) that has been shown to be morphologically similar to brain endothelium. This cell line exhibited properties of the BBB, such as restricted permeability, electrical resistance, and polarity, and was characterized by appropriate marker enzymes (Tatsuta et al., 1992). The MBEC cells form a monolayer epithelium and express P-glycoprotein in a polarized fashion. By immunoelectron microscopic analysis, P-glycoprotein was shown to be localized to the apical side of the cells and was responsible for transepithelial transport of vincristine and cyclosporin A (Tatsuta et al., 1992; Shirai et al., 1994). Several clones of MBEC were previously prepared from mouse brain capillaries by Tatsuta et al., (1992). In the present study, we have used MBEC-4 cells to establish the presence of GSH transport and to identify Na+-dependent GSH transport in cultured cells as well as mRNA-injected oocytes. Three other transformed cell lines, namely, Cos-1, Cos-7, and HepG2, which have previously been shown not to exhibit Na+-dependent GSH transport (Lu et al., 1996), were used as controls. Furthermore, evidence that the Na+-dependent GSH transporter is located on the apical (luminal) membrane of MBEC cells was obtained from studies using isolated plasma membrane vesicles.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL PROCEDURES
  4. GSH transport by MBEC-4 cells
  5. Effect of extracellular GSH on intracellular GSH
  6. RNA isolation, oocyte preparation, and transport
  7. RESULTS
  8. GSH and GGT levels in MBEC-4 cells
  9. Na+ dependency of GSH uptake in MBEC-4 cells
  10. Na+-mediated increase in cellular GSH
  11. Expression of GSH uptake in Xenopus laevis oocytes
  12. Characteristics of GSH uptake in luminal membrane vesicles
  13. DISCUSSION
  14. Acknowledgements
  15. References

Materials and chemicals

[35S]GSH (500 Ci/mmol), [35S]cysteine (600 Ci/mmol), and [14C]glucose (40-50 Ci/mmol) were obtained from NEN (Braintree, MA, U.S.A.). The 35S-labeled isotopes were >98% pure as determined by HPLC (Fariss and Reed, 1987). Chemicals, buffer reagents, and protease inhibitor cocktail for mammalian cells were obtained from Sigma Chemical Co. (St. Louis, MO, U.S.A.). Membrane filters (HAWP 0025) were obtained from Millipore Corp. (Bedford, MA, U.S.A.).

Cultured cells

Isolation of the SV40-transformed mouse brain endothelial cell line (MBEC-4) and its biochemical and morphological characterization have been described previously (Tatsuta et al., 1992). The cells were maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum, 100 U/ml penicillin, and 100 μg/ml streptomycin. Cells from passages 7-12 were used in all cellular uptake studies. HepG2, Cos-1, and Cos-7 cells were obtained from American Type Culture Collection (Rockville, MD, U.S.A.) and grown to confluence using growth media and culture conditions as described previously (Lu et al., 1996). Total glutathione content of MBEC-4 cells was measured by recycling assay (Tietze, 1969); GGT activity was determined as described earlier (Sze et al., 1993). Protein estimation was made using the Bio-Rad protein assay (Bio-Rad Laboratories, Hercules, CA, U.S.A.).

GSH transport by MBEC-4 cells

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL PROCEDURES
  4. GSH transport by MBEC-4 cells
  5. Effect of extracellular GSH on intracellular GSH
  6. RNA isolation, oocyte preparation, and transport
  7. RESULTS
  8. GSH and GGT levels in MBEC-4 cells
  9. Na+ dependency of GSH uptake in MBEC-4 cells
  10. Na+-mediated increase in cellular GSH
  11. Expression of GSH uptake in Xenopus laevis oocytes
  12. Characteristics of GSH uptake in luminal membrane vesicles
  13. DISCUSSION
  14. Acknowledgements
  15. References

Confluent MBEC-4 cells (1-2 × 106 cells/well in 12-well plates) were pretreated for 30 min with 0.5 mM acivicin and 10 mM DL-buthionine sulfoximine (BSO) to block degradation and resynthesis of GSH, respectively, before uptake studies were performed. In initial experiments, the optimal incubation time (for deriving linear initial rates) was determined by incubating cells pretreated with BSO and acivicin in NaCl buffer for 5, 10, 15, 30, and 60 min at 37°C (or 4°C) with [35S]GSH containing 0.1, 1, or 2 mM GSH.

Cellular radioactivity was determined after three washes with ice-cold NaCl buffer (100 mM NaCl, 1.2 mM MgCl2, 0.81 mM MgSO4, and 25 mM HEPES/Tris, pH 7.4) containing 5 mM GSH followed by treatment with trypsin/EDTA (0.05 and 0.02%, respectively). The 4°C binding ranged between 10 and 15% of the 37°C values and did not increase with time. As the time course of uptake was linear up to 60 min, all subsequent experiments were carried out in the linear part of the curve for 30 min.

To determine whether uptake exhibited Na+ dependency at varying GSH concentrations, uptake was performed for 30 min at 0.05, 0.1, and 1 mM GSH in 10 mM dithiothreitol (DTT) containing ∼1 μCi of [35S]GSH/well in NaCl buffer or choline chloride buffer. The molecular form of uptake was verified by an HPLC method in which the carboxymethyl derivatives of free thiols in cell homogenates were converted to 2,4-dinitrophenyl derivatives before analysis (Fariss and Reed, 1987). These analyses were performed in samples obtained from studies where cellular GGT was either uninhibited or inhibited by acivicin.

In separate experiments, the effect of various organic anions and GSH analogues on GSH uptake was studied. GSH uptake was measured in NaCl medium at 37°C for 30 min in the presence or absence of 2 mM dibromosulfophthalein (DBSP), 0.1 mM 4,4′-diisothiocyanatostilbene-2,2′-disulfonic acid (DIDS), and 2 mM S-octyl-GSH.

Effect of extracellular GSH on intracellular GSH

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL PROCEDURES
  4. GSH transport by MBEC-4 cells
  5. Effect of extracellular GSH on intracellular GSH
  6. RNA isolation, oocyte preparation, and transport
  7. RESULTS
  8. GSH and GGT levels in MBEC-4 cells
  9. Na+ dependency of GSH uptake in MBEC-4 cells
  10. Na+-mediated increase in cellular GSH
  11. Expression of GSH uptake in Xenopus laevis oocytes
  12. Characteristics of GSH uptake in luminal membrane vesicles
  13. DISCUSSION
  14. Acknowledgements
  15. References

MBEC-4 cells were treated with 10 mM BSO/0.5 mM acivicin for 30 min at 37°C. After a quick wash, they were incubated with 4 mM GSH in NaCl buffer or choline chloride buffer (containing 100 mM choline chloride replacing 100 mM NaCl in the NaCl buffer) for 1 h at 37°C. Viability, as measured by trypan blue exclusion, did not change in either buffer at the end of experiments. Cell diameter was measured using a Coulter channelyzer counter. Cells were washed with the respective buffers, and cellular GSH and protein were determined as described above.

RNA isolation, oocyte preparation, and transport

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL PROCEDURES
  4. GSH transport by MBEC-4 cells
  5. Effect of extracellular GSH on intracellular GSH
  6. RNA isolation, oocyte preparation, and transport
  7. RESULTS
  8. GSH and GGT levels in MBEC-4 cells
  9. Na+ dependency of GSH uptake in MBEC-4 cells
  10. Na+-mediated increase in cellular GSH
  11. Expression of GSH uptake in Xenopus laevis oocytes
  12. Characteristics of GSH uptake in luminal membrane vesicles
  13. DISCUSSION
  14. Acknowledgements
  15. References

Poly(A)+ RNA was isolated from MBEC-4, HepG2, and Cos-1 cells using the Fastrack kit from Invitrogen (San Diego, CA, U.S.A.). The A280 was used to calculate RNA concentration. Isolation of stages 5 and 6 defolliculated oocytes, microinjection of mRNA, and maintenance of injected oocytes in modified Barth’s medium were as described previously (Fernandez-Checa et al., 1993; Kannan et al., 1996). We assessed the functional quality of oocytes by expressing rat sodium taurocholate transport protein and rat organic anion transport protein by injecting oocytes with cRNA of respective clones (made available through the Molecular Biology Core of the USC Center for Liver Diseases). In addition, water-injected controls were included in all experiments to check whether the oocytes were “leaky” or not.

Pilot studies performed to determine the maximal expression of GSH transport revealed that 4 days after poly(A)+ RNA injection, expression was maximal. The incubation time for uptake (1 h) was also established from experiments on the time course of GSH uptake for 15, 30, 60, 90, and 120 min. The net uptake of radioactivity in poly(A)+ RNA-injected oocytes was linear up to 60 min (data not shown).

Oocytes were injected with equal amounts of poly(A)+ RNA (45 ng/oocyte) from MBEC-4, Cos-1, or HepG2 cells. Oocytes were washed three times in either NaCl buffer (NaCl) or choline chloride buffer (Na+-free). NaCl buffer contained 100 mM NaCl, 1.2 mM MgCl2, 0.81 mM MgSO4, and 25 mM HEPES/Tris (pH 7.4), and choline chloride replaced NaCl isosmotically in the Na+-free buffer. Oocytes were resuspended in 100 μl (four to six oocytes) in either buffer in the presence of 1 μCi of [35S]GSH (containing 0.05 or 2 mM GSH) in 10 mM DTT. Before the addition of the tracer, oocytes were pretreated with 0.5 mM acivicin for 30 min to inhibit GGT activity. GGT activity in Stage 6 oocytes was below the detection limit of the assay (Fernandez-Checa et al., 1993). At the end of incubation, oocytes were washed with the respective chilled buffers containing 5 mM GSH. Single oocytes from each group were individually lysed with 1 ml of 10% sodium dodecyl sulfate, and radioactivity in single oocytes was measured in a Beckman scintillation counter after addition of 8 ml of scintillation fluid. Three or four experiments were performed with four to six oocytes per group, and radioactivity in individual oocytes was determined and averaged per group.

Derivation and characterization of luminal and abluminal membrane vesicles

Total plasma membrane vesicles. Total plasma membrane vesicles were prepared by the method of Leier et al., (1994), with modifications described below. Approximately 6-8 × 108 mouse brain endothelial cells were harvested by centrifugation at 1,200 g for 8 min (ravg, Sorvall SS-34 rotor, 4°C). The cells were suspended in 40 ml of hypotonic lysis buffer (1 mM sodium bicarbonate, pH 7.0, supplemented with a cocktail of protease inhibitors as recommended by the supplier) and gently stirred at 4°C for 1 h. The suspension was then centrifuged at 100,000 g (ravg, Beckman SW 45Ti rotor) for 30 min at 4°C. The resulting pellet was suspended in 20 ml of the lysis buffer and homogenized in a Potter-Elvehjem (S818 pestle, 4°C, 900-1,100 rpm, at ∼3 strokes/min). The homogenate was diluted twofold in a buffer containing 520 mM sucrose, 0.4 mM CaCl2, and 10 mM HEPES/Tris (pH 7.4) and subsequently centrifuged at 1,200 g for 8 min (ravg, Sorvall SS-34 rotor, 4°C) to pellet the crude nuclear fraction. The supernatant was diluted twofold with the suspension buffer (260 mM sucrose, 0.2 mM CaCl2, 5 mM HEPES/Tris, pH 7.4) and centrifuged at 100,000 g (ravg, Beckman SW 45Ti rotor) for 45 min at 4°C. This pellet, representing mixed (total) plasma membranes, was suspended in 15 ml of the suspension buffer.

Separation and isolation of three fractions from mixed plasma membranes. The separation of the two membrane domains (luminal and abluminal) was based on a protocol established by Meier et al. (1984a,b) for the isolation of apical and basolateral membrane fractions from hepatocytes. The total plasma membrane suspension was homogenized at 4-8°C in a tight glass-glass Dounce homogenizer by 50 strokes. The suspension (3.5 ml) was then layered over a step gradient of 2.5 ml of 30% sucrose (wt/wt, 1.132 g/ml), 2.5 ml of 34% sucrose (wt/wt, 1.146 g/ml), and 4 ml of 38% sucrose (wt/wt, 1.166 g/ml). The tubes were spun at 195,200 g (ravg) for 2 h at 4°C (Beckman SW41 rotor). The plasma membrane fractions at the interface of the 260 mM sucrose buffer-30% layer, the 30-34% layer, and the 34-38% layer were collected. The approximate density of the fractions was determined in a pilot separation of the mixed plasma membrane suspension in a continuous sucrose gradient of 25-40% (wt/wt). Four broad and predominantly tubid bands were observed, whose densities were estimated by the buoyancy of density marker beads in a parallel tube. The individual fractions were diluted in 40 ml of suspension buffer and centrifuged at 100,000 g (ravg) for 45 min at 4°C. The resulting pellets were suspended in 400-600 μl of the suspension buffer and visiculated by 20 passages through a 25-gauge needle. Then, 100- to 200-μl aliquots of the membrane vesicles were rapidly frozen in liquid nitrogen and stored at -80°C. Typically, the vesicles were stored at a concentration of ≥4 mg/ml for up to 2 months.

Enzymatic characterization of MBEC membrane fractions. The purity and enrichment of the membrane fractions were monitored by the membrane marker enzymes alkaline phosphatase, Na+,K+-ATPase (5 mM ouabain inhibitable), and Mg2+-ATPase. We estimated the contamination of intracellular organelles by measuring the activities of glucose 6-phosphatase (microsomes), acid phosphatase (lysosomes), and succinic dehydrogenase (mitochondria) in each membrane preparation. Mg2+-ATPase and ouabain-inhibitable Na+,K+-ATPase activities were measured by the endpoint release of inorganic phosphate (Fiske and SubbaRow, 1925) from ATP. Alkaline phosphatase was measured by quantifying the amount of p-nitrophenol liberated, using p-nitrophenylphosphate as a substrate (Meier et al., 1984a). Activity of succinic dehydrogenase was determined by the reduction of 2,6-dichlorophenol indophenol by phenazine methosulfate (Hatefi and Stigall, 1978). Glucose 6-phosphatase and acid phosphatase were measured by the release of inorganic phosphate from glucose 6-phosphate and β-glycerophosphate, respectively (Fernandez-Checa et al., 1992).

Identification of luminal and abluminal membrane fractions. We used P-glycoprotein 170 as a marker of the luminal membrane. P-glycoprotein 170 has been immunolocalized to the luminal domain in these mouse brain endothelial cells (Tatsuta et al., 1992). Furthermore, we confirmed the adequate separation of membrane fractions by the absence of immuno-detectable P-glycoprotein 170 in the membrane fraction at the 34-38% sucrose interface. As a Na+-dependent transport pathway for D-glucose has been reported in the abluminal domain of bovine brain endothelial plasma membranes (Lee et al., 1997), we probed the fractions for Na+ dependency of D-glucose uptake (see below) to functionally identify the abluminal fraction.

For western blot analysis, plasma membrane proteins (50 μg) were electrophoresed in an 8% agarose gel (Bio-Rad) and were transferred to nitrocellulose membranes. Immunoblot analysis was first carried out with a 1:1,000 dilution of monoclonal primary antibody C219 (Signet Laboratories, Dedham, MA, U.S.A.) and then with peroxidase-conjugated anti-mouse IgG at 1:1,000 dilution. P-glycoprotein was detected with the enhanced chemiluminescence western blotting detection system (Amersham, U.K.). Protein from rat liver homogenate (50 μg) was used as a positive control in identifying the P-170 band in western blot analysis.

Transport studies in membrane vesicles

The transport of [35S]GSH or D-[14C]glucose was measured by the rapid filtration technique. In brief, frozen membranes were rapidly thawed for 3 min at 37°C in a water bath and vesiculated by 15 passages through a 25-gauge needle. The suspension was diluted (2-3 mg of protein/ml) in an appropriate resuspension buffer and mixed by five more passes through a 25-gauge needle. Uptake of glucose and GSH was measured under a valinomycin-induced (10-12 μg/mg of protein, dissolved in ethanol) K+ gradient of 100 mM > 20 mM, directed outward. To study the effect of Na+ on the uptake of D-glucose and GSH, a gradient of Na+out > Na+in (100 mM > 0 mM) was applied, in addition to the outwardly directed K+ gradient. The incubation buffer in GSH uptake studies had a final composition of [35S]GSH (0.005, 0.030, 0.1, 1, 5, 10, and 30 mM), 100 mM NaCl, 20 mM KCl, 1 mM DTT, 0.2 mM CaCl2, 5 mM MgCl2, 10 mM HEPES/Tris (pH 7.4), and varying amounts of sucrose to maintain isosmolarity. The final composition of the buffer in D-glucose uptake studies that were carried out for various timepoints up to 10 min was identical, except with 0.5 mM D-glucose. The composition of the stop buffer was 100 mM KCl, 60 mM sucrose, 0.2 mM CaCl2, 5 mM MgCl2, and 10 mM HEPES/Tris (pH 7.4). In some studies, the effect of voltage clamping on GSH uptake was studied by resuspending vesicles in 75 mM K+ for 1 h at 25°C in the presence of valinomycin. The incubation buffer and stop buffers also contained 75 mM K+. All membrane fractions were pretreated with 3 mM acivicin for 30 min at 25°C.

Typically, 20 μl of membrane suspension (30-50 μg of protein) was preincubated for 3 min at 37°C in water bath. Uptake of [35S]GSH or [14C]glucose was initiated by adding 80 μl of incubation buffer, containing [35S]GSH and the appropriate ionic gradient. At the desired time, uptake was terminated by adding 1 ml of ice-cold stop buffer. [35S]GSH or D-[14C]glucose transported and bound to membrane vesicles was trapped on 0.45-μm cellulose acetate filters by rapid filtration. The filters were washed twice with 4 ml of ice-cold stop buffer, and the radioactivity retained on the filters was measured in a liquid scintillation counter. Care was taken to ensure uniformly low quenching and high efficiency. “Nonspecific” binding was estimated as the amount of radioactivity bound to membranes at 4°C and zero time, by adding a chilled aliquot of membranes to 1 ml of ice-cold stop solution with the radiolabel.

The osmotic response of the membrane vesicles was determined by the net uptake of [35S]GSH in vesicles of varying volumes. The amount of [35S]GSH transported into intravesicular space was estimated at equilibrium (30 min) in the presence of an inwardly directed Na+ gradient. The intravesicular volumes were varied by changing the osmolarity of the incubation medium using raffinose, a trisaccharide with a high membrane reflection coefficient. Uptake was terminated with stop solutions of corresponding osmolarity.

We analyzed our kinetic data using the SAAM II program (Simulation, Analysis, and Modeling II; SAAM II User’s Guide, SAAM Institute, University of Washington, 1994) for curve fitting and estimation of kinetic parameters as described (Aw et al., 1996; Kannan et al., 1996). As the kinetic data were obtained from two individual preparations, the values were weighted by their standard deviation to yield approximate kinetic parameters. The substrate to transport rate data were analyzed using three models: (1) a Michaelis-Menten hyperbolic relationship, (2) a sigmoidal Hill equation, and (3) a combination of Michaelis-Menten and Hill models to delineate multiple kinetic components in total GSH transport.

GSH and GGT levels in MBEC-4 cells

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL PROCEDURES
  4. GSH transport by MBEC-4 cells
  5. Effect of extracellular GSH on intracellular GSH
  6. RNA isolation, oocyte preparation, and transport
  7. RESULTS
  8. GSH and GGT levels in MBEC-4 cells
  9. Na+ dependency of GSH uptake in MBEC-4 cells
  10. Na+-mediated increase in cellular GSH
  11. Expression of GSH uptake in Xenopus laevis oocytes
  12. Characteristics of GSH uptake in luminal membrane vesicles
  13. DISCUSSION
  14. Acknowledgements
  15. References

Total glutathione level measured by the Tietze assay in MBEC-4 cells from passage 7 through 12 averaged 49.4 ± 11.6 nmol/mg of protein (means ± SD, n = 12). GGT activity was 15 ± 0.4 mU/mg of protein/min, and treatment of MBEC-4 cells with 1 mM acivicin for 30 min decreased the GGT activity to <1.5 mU/mg of protein/min, which is the lowest detectable limit of the assay. GGT activity in the cultured cells is two- to threefold lower than that of freshly isolated mouse brain capillaries, as reported earlier (Tatsuta et al., 1992).

Na+ dependency of GSH uptake in MBEC-4 cells

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL PROCEDURES
  4. GSH transport by MBEC-4 cells
  5. Effect of extracellular GSH on intracellular GSH
  6. RNA isolation, oocyte preparation, and transport
  7. RESULTS
  8. GSH and GGT levels in MBEC-4 cells
  9. Na+ dependency of GSH uptake in MBEC-4 cells
  10. Na+-mediated increase in cellular GSH
  11. Expression of GSH uptake in Xenopus laevis oocytes
  12. Characteristics of GSH uptake in luminal membrane vesicles
  13. DISCUSSION
  14. Acknowledgements
  15. References

Figure 1 shows the effect of Na+ removal on GSH uptake in MBEC-4 cells and Cos-7 cells under conditions of inhibition of GGT and γ-glutamylcysteine synthase (GCS) with acivicin and BSO, respectively. Net GSH uptake, expressed as nanomoles per 106 cells per hour (Fig. 1), was significantly lower (p < 0.01) in the Na+-free (choline chloride) medium than the Na+-containing (NaCl) medium, whereas GSH uptake in Cos-7 cells was Na+ independent, as shown previously in our laboratory (Lu et al., 1996). Partial Na+ dependency of MBEC-4 cells was also seen at 0.05 mM GSH (0.028 ± 0.003 and 0.013 ± 0.002 nmol/106 cells/h in the presence and absence of Na+, respectively).

image

Figure 1. Effect of replacement of Na+ on net GSH uptake by MBEC-4 and Cos-7 cells. GSH uptake was performed for 30 min at 2 mM GSH in either NaCl buffer (filled columns) or choline chloride buffer (hatched columns) at 37 or 4°C. The cells were treated with 10 mM BSO and 1 mM acivicin for 15 min prior to uptake. Uptake data expressed as nmol/106 cells/30 min represent net uptake (uptake at 37°C — uptake at 4°C) and are means ± SEM from four or five preparations each performed in duplicate.

Download figure to PowerPoint

To verify that the uptake of radioactivity in experiments described in Fig. 1 is in the form of intact GSH, the molecular form of uptake of radioactivity after 1-h incubation of MBEC-4 cells pretreated with acivicin and BSO was determined. As shown in Fig. 2A, uptake was predominantly (>92%) in the form of GSH. The HPLC profile was similar when uptake was studied under GGT-uninhibited conditions (not shown).

image

Figure 2. HPLC profile of cell homogenate after 1-h incubation of MBEC-4 cells with 35S-labeled GSH containing 1 mM GSH in NaCl buffer (A) or with [35S]cysteine plus 50 μM cysteine in DTT (B). Pretreatment with BSO and acivicin and derivatization of samples after uptake are as described in EXPERIMENTAL PROCEDURES. Top: UV absorption at 365 nm is shown; bottom: radioactivity (dpm) is shown. The mass peaks for cysteine and GSH are indicated.

Download figure to PowerPoint

To further exclude the possibility that the observed GSH uptake shown in Fig. 2 is from GGT-mediated degradation and subsequent resynthesis, uptake of [35S]cysteine (+50 μM cysteine) was examined by HPLC. Cysteine remained exclusively in its native form as cysteine, and no conversion to GSH occurred during the 1-h incubation (Fig. 2B), indicating that BSO treatment had completely inhibited GSH synthesis.

In experiments designed to assess the effect of potential inhibitors on GSH uptake by MBEC-4 cells, the organic anions DBSP (2 mM) and DIDS (0.2 mM) inhibited GSH uptake significantly, whereas the GSH analogue S-octyl-GSH (2 mM) did not change GSH uptake. Mean ± SD GSH uptake determined without any inhibitor at 1 mM GSH was 0.48 ± 0.06 nmol/106 cells, whereas the uptakes with DBSP, DIDS, and S-octyl-GSH were 0.13 ± 0.04, 0.05 ± 0.01, and 0.53 ± 0.04 nmol/106 cells, respectively.

Na+-mediated increase in cellular GSH

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL PROCEDURES
  4. GSH transport by MBEC-4 cells
  5. Effect of extracellular GSH on intracellular GSH
  6. RNA isolation, oocyte preparation, and transport
  7. RESULTS
  8. GSH and GGT levels in MBEC-4 cells
  9. Na+ dependency of GSH uptake in MBEC-4 cells
  10. Na+-mediated increase in cellular GSH
  11. Expression of GSH uptake in Xenopus laevis oocytes
  12. Characteristics of GSH uptake in luminal membrane vesicles
  13. DISCUSSION
  14. Acknowledgements
  15. References

In an effort to examine whether the Na+-dependent transporter drives up cellular GSH in MBEC-4 cells, we measured GSH in cells after incubation with GSH in NaCl and choline chloride buffers (Table 1). Cells were pretreated with BSO and acivicin.

Table 1. Effect of extracellular GSH on intracellular GSH concentration in MBEC cells in Na+-containing and Na+-free medium
   Cellular GSH
Experiment Medium GSH (mM) Incubation buffer Control (no GSH) (nmol/106 cells) + 4 mM GSH (nmol/106 cells)% increase over control
  1. Data are means ± SEM (n = 3). MBEC-4 cells were pretreated with 10 mM BSO and 1 mM acivicin for 15 min and incubated with 4 mM GSH for 1 h at 37°C in either NaCl or choline chloride buffer. Intracellular GSH was determined by the recycling assay. Cell diameters measured at the end of the incubations were not significantly different in the presence or absence of GSH as well as between NaCl and choline chloride buffers.

I4NaCl12.6 ± 0.217.1 ± 1.036.7 ± 3.2
 4Choline chloride14.8 ± 1.914.5 ± 1.2-2.1 ± 0.8
II4NaCl16.2 ± 1.319.5 ± 2.020.1 ± 1.3
 4Choline chloride17.4 ± 2.218.6 ± 2.36.8 ± 1.1
III4NaCl15.3 ± 0.918.7 ± 1.322.2 ± 1.0
 4Choline chloride14.8 ± 1.114.1 ± 0.9-4.7 ± 0.5

Incubation in the presence of NaCl containing 4 mM GSH caused a significant increase in cellular GSH, shown as a percent increase over control incubation without GSH in the medium. Changes in cell GSH in the absence of Na+ did not show an increase but were similar or lower than in the corresponding control group (Table 1). The cell diameters among the controls and groups incubated with GSH in either Na+-containing and Na+-free buffers did not differ significantly from each other (data not shown). This excludes the possibility that changes in GSH concentration in the two groups are due to changes in cell volume. The viability of cells was also similar in all groups.

Expression of GSH uptake in Xenopus laevis oocytes

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL PROCEDURES
  4. GSH transport by MBEC-4 cells
  5. Effect of extracellular GSH on intracellular GSH
  6. RNA isolation, oocyte preparation, and transport
  7. RESULTS
  8. GSH and GGT levels in MBEC-4 cells
  9. Na+ dependency of GSH uptake in MBEC-4 cells
  10. Na+-mediated increase in cellular GSH
  11. Expression of GSH uptake in Xenopus laevis oocytes
  12. Characteristics of GSH uptake in luminal membrane vesicles
  13. DISCUSSION
  14. Acknowledgements
  15. References

In initial experiments, the time course of expression of GSH in X. laevis oocytes was studied. Oocytes pretreated with acivicin were injected with equal amounts (25 ng/36 nl) of either water or MBEC mRNA. Although water-injected oocytes had negligible uptake, expression of GSH uptake in NaCl medium with MBEC mRNA-injected oocytes increased from day 2 and was maximal on day 4. Uptake on day 4 was 15-fold higher than that of water controls (Fig. 3). All subsequent oocyte expression studies were performed on day 4 after mRNA injection. In separate experiments, the molecular form of uptake was verified to be >93% GSH by HPLC (not shown).

image

Figure 3. Time course of expression of GSH uptake in MBEC mRNA-injected X. laevis oocytes. Oocytes that were pretreated with acivicin were injected with either water (open circles) or MBEC mRNA (25 ng/36 nl; filled circles), and expression was followed as a function of days by studying uptake of 1 mM GSH in NaCl medium at 18°C each day. Data are means ± SE from three oocyte preparations. Four to six oocytes were used in each preparation, and uptake was determined in individual oocytes and is presented as nmol/oocyte/h.

Download figure to PowerPoint

Effect of removal of Na+ on GSH uptake was studied in MBEC mRNA-injected oocytes (Fig. 4). A comparison of GSH uptake in the presence or absence of Na+ in oocytes injected with mRNA from different cell types was made to assess the specificity of Na+ requirement for uptake by MBEC mRNA. These parallel uptake studies were performed with mRNA from HepG2 and Cos-1 cells. The latter cell lines in culture have been shown to exhibit Na+-independent GSH transport (Lu et al., 1996). Uptake was studied for 1 h in acivicin-pretreated oocytes in the presence of 1 mM GSH. Na+ removal caused a significant (∼70%) decrease in uptake expressed as nanomoles per oocyte per hour in MBEC mRNA-injected oocytes (Fig. 4). This decrease varied somewhat from experiment to experiment and ranged from 40 to 70%. The data show that a Na+-independent GSH transport system is also expressed. Partial Na+ dependency of uptake was also observed for 0.05 and 2 mM GSH concentrations in the medium (data not shown). On the other hand, as expected, uptake was Na+ independent in the case of mRNA from HepG2 and Cos-1 cells. The figure also shows that the rate of GSH uptake in MBEC-mRNA was two- to threefold higher than that of Cos-1 and HepG2 mRNA.

image

Figure 4. Comparison of GSH uptake in presence or absence of sodium in oocytes injected with mRNA from different cell types. GSH uptake was performed on day 4 after water or mRNA injection. Equal amounts of mRNA (45 ng/25 nl) from MBEC, HepG2, or Cos-1 cells were injected into acivicin-pretreated oocytes, and uptake was studied with 1 mM GSH at 18°C in Na+-containing (open columns) and Na+-free (hatched columns) buffers. Data are means ± SEM from three or four preparations, each performed with four to six oocytes per condition. Uptake of radioactivity was determined in individual oocytes for each oocyte preparation.

Download figure to PowerPoint

Characterization of luminal and abluminal membrane fractions

The specific activity and enrichment of marker enzymes in the membrane fractions are reported in Table 2. Alkaline phosphatase was asymmetrically distributed and enriched predominantly in the 30% membrane fraction (17-fold). Mg2+-ATPase was enriched uniformly in both the 30% and the 38% fractions (2.5- and 1.6-fold, respectively), whereas low-ouabain affinity Na+,K+-ATPase was selectively enriched in the 30% fraction (16-fold), as reported earlier (Nag, 1990; Sanchez del Pino et al., 1995). With the exception of succinic dehydrogenase (a mitochondrial marker) in the 38% fraction, the subfractions were minimally contaminated by lysosomes (acid phosphatase), endoplasmic reticulum (glucose 6-phosphatase), and mitochondria (30% fraction). These results demonstrate the yield of highly enriched and pure membrane fractions from MBEC-4 cells suitable for functional studies specific to the individual domains.

Table 2. Specific activities and relative enrichments of marker enzymes in MBEC membrane fractions
  Na+,K+-ATPase Mg2+-ATPase Alkaline phosphatase
 HLMAHLMAHLMA
  1. Data are means ± SEM from four independent preparations. Specific activities are expressed as nmol of product formed/min/mg of protein. H, homogenate; L, luminal membrane fraction (30% sucrose); M, mixed membrane fraction; A, abluminal membrane fraction (38% sucrose). Relative enrichment is expressed as the ratio of the specific activity in the homogenate to that in the individual membrane fractions.

Specific activity5.1 ± 1.868 ± 1038 ± 1433 ± 1696 ± 30208 ± 49138 ± 38149 ± 363.6 ± 1.160 ± 1715 ± 4.64.5 ± 1.6
Relative enrichment 16 ± 5.99.2 ± 4.46.6 ± 1.6 2.5 ± 0.61.6 ± 0.41.8 ± 0.6 17 ± 1.34.2 ± 0.51.3 ± 0.2

To identify the domain specificity of the membrane fractions separated from MBEC, we probed for established markers of the luminal and abluminal domains of brain endothelial cells. As P-glycoprotein 170 has been immunolocalized to the luminal membrane domain of intact MBEC cells, we used it to identify the luminal membrane fraction. Similarly, we attempted to identify the abluminal fraction in our separations by the presence of an analogous Na+-dependent D-glucose transport system reported by Lee et al. (1997) in abluminal membrane vesicles from bovine BBB. Figure 5A shows the western blot of MBEC-4 fractions. As previously reported, the liver homogenate, which was used as the positive control, showed additional bands in addition to the two at 170 kDa corresponding to P-glycoprotein and at 190 kDa, respectively (Kamath and Morris, 1998). The observation of two closely occurring bands is consistent with similar findings from other laboratories that suggested that C219 reacts with two proteins of the brain capillaries: P-glycoprotein and an unrelated protein (Jette et al., 1995a,b; Kamath and Morris, 1998). P-glycoprotein 170 was detected predominantly in the 30% fraction, with weak staining in the 34% fraction. This is expected because the membrane fraction buoyant at the 30-34% interface represents a mixture of 30 and 38% membrane fragments. Notably, P-glycoprotein 170 was not detected in the 38% fraction (Fig. 5A), which confirms the purity of the membrane fractions resulting from our separation. We observed substantial Na+-dependent transport of D-glucose exclusively in the 38% membrane fraction, hence presumed to be abluminal membranes (Fig. 5B). Uptake of D-glucose in the presence of inwardly directed Na+ gradient in the 38% fraction was higher at all timepoints studied (not shown). D-Glucose transport in the 30% membrane fraction was Na+ independent. Substantial facilitated (passive) transport of D-glucose was observed in both 30 and 38% fractions. Henceforth, the 30% membrane fraction will be referred to as luminal membrane vesicles and the 38% fraction as abluminal membrane vesicles.

image

Figure 5. Characterization of the luminal and abluminal membrane vesicles: western blot analysis for P-glycoprotein 170 in membrane fractions from MBEC-4 cells (A) and uptake of glucose in luminal (30%) and abluminal (38%) membrane fractions (B). For western blot analysis, plasma membrane proteins (50 μg) were electrophoresed in 8% agarose gel and were transferred to nitrocellulose membranes. A 1:1,000 dilution of a primary monoclonal antibody C219 and a 1:1,000 dilution of peroxidase-conjugated anti-mouse IgG were used in immunoblot analysis. P-glycoprotein 170 was detected by enhanced chemiluminescence as a band at 170 kDa (A). The lanes from left to right are as follows: lane 1, 38% fraction; lane 2, 34% fraction; lane 3, 30% fraction; lane 4, mixed membranes from MBEC-4; lane 5, rat liver homogenate. Equal loading of samples was verified by Coomassie Blue gel stain (not shown). A faint band appearing below 170 kDa is not identified but is known to be present in brain capillaries. Uptake of 0.5 mM D-glucose (B) was carried out in the presence or absence of a 100 mM NaCl inwardly directed gradient for 15 s. Values are means ± SE from duplicate determinations in four independent membrane preparations.

Download figure to PowerPoint

Characteristics of GSH uptake in luminal membrane vesicles

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL PROCEDURES
  4. GSH transport by MBEC-4 cells
  5. Effect of extracellular GSH on intracellular GSH
  6. RNA isolation, oocyte preparation, and transport
  7. RESULTS
  8. GSH and GGT levels in MBEC-4 cells
  9. Na+ dependency of GSH uptake in MBEC-4 cells
  10. Na+-mediated increase in cellular GSH
  11. Expression of GSH uptake in Xenopus laevis oocytes
  12. Characteristics of GSH uptake in luminal membrane vesicles
  13. DISCUSSION
  14. Acknowledgements
  15. References

Demonstration of uptake in total plasma membrane vesicles by osmotic response. To distinguish net GSH uptake into the intravesicular space from nonspecific binding to membrane proteins, the osmolarity of the incubation medium was varied from 280 to 680 mOsmol/L. As the amounts of luminal and abluminal membrane vesicles in each isolation were limited, we performed these studies in total plasma membrane vesicles, which is a mixture of both. GSH uptake at 5 and 100 μM in the presence of an inwardly directed Na+ gradient was inversely proportional to medium osmolarity, demonstrating transport into an osmotically active space (data not shown). Extrapolation to infinite osmolarity by simple linear regression showed a binding component (y-intercept) of 10-25% of total GSH uptake.

Na+-dependent GSH uptake selective to luminal membrane vesicles. Having thus established the validity of the membrane vesicles, we studied the effect of an inwardly directed Na+ gradient on GSH uptake at three concentrations (5 μM, 100 μM, and 2 mM) in luminal (30%) and abluminal (38%) membrane vesicles. In the presence of the Na+ gradient, GSH uptake at all three concentrations was consistently enhanced only in luminal membrane vesicles (Fig. 6). The magnitude of the increase was at least twofold in all four membrane preparations. Under identical conditions, uptake of GSH by abluminal membrane vesicles was not influenced by the Na+ gradient. As our preliminary studies indicated the presence of both Na+-independent and Na+-dependent pathways for GSH in luminal membrane vesicles, we measured GSH uptake in the presence of a valinomycin-induced outward gradient of K+. This enabled us to maximize the detection of the Na+-dependent component. In the presence of valinomycin and a voltage clamp (Kin = Kout), GSH uptake in the luminal membrane vesicles was consistently enhanced by the Na+ gradient, although the magnitude of the increase was ∼50% (data not shown).

image

Figure 6. Uptake of GSH at three GSH concentrations [5 μM (A), 100 μM (B), and 2 mM (C)] in the MBEC-4 plasma membrane fractions in the presence or absence of 100 mM NaCl inwardly directed sodium gradient. Values are shown as means ± SE from four independent membrane preparations.

Download figure to PowerPoint

Substantial Na+-independent uptake was observed in luminal membrane vesicles. However, in the presence or absence of Na+, uptake of GSH in abluminal membrane vesicles occurred at a substantially lower rate (Fig. 6).

Overshoot of Na+-dependent GSH uptake in luminal membrane vesicles. Luminal and abluminal membrane vesicles exhibited time-dependent uptake of GSH, as demonstrated in Fig. 7. Additional evidence for Na+-driven accumulation of GSH in luminal membrane vesicles is provided by the “overshoot” in uptake above equilibrium values in the presence of an Na+ gradient. This overshoot of uptake is a characteristic of many ion-coupled secondary active transporters. The transient accumulation of GSH is indicative of its coupled transfer with Na+ ions. An inwardly directed gradient of Na+ results in an initial electrochemical gradient of Na+ that dissipates rapidly due to ion equilibration. Thus, a transient and concentrative accumulation of GSH occurs, which manifests as an overshoot over the rate measured at an equilibrium point such as 10 min. An overshoot of GSH uptake was not observed in abluminal membrane vesicles (Fig. 7).

image

Figure 7. Time course of 100 μM GSH uptake in luminal (left) and abluminal (right) MBEC-4 membrane vesicles. Uptake was measured under a K+ gradient of 100 mM > 20 mM directed outward. An inwardly directed 100 mM NaCl gradient was also applied in addition to outwardly directed K+ gradient and compared with 100 mM KCl (no K+ gradient). Data represent means from triplicate determinations of a single preparation from luminal and abluminal vesicles. Na+-driven accumulation of GSH in luminal membrane vesicles is seen as an overshoot on the left, whereas overshoot was not observed in abluminal membrane vesicles on the right.

Download figure to PowerPoint

Kinetics of GSH uptake in luminal membrane vesicles. The kinetics of GSH uptake in two independent luminal membrane preparations is shown in Fig. 8. Uptake in the absence of Na+ (dashed curve) was best fit by the sum of a high-affinity (Km∼ 1.7 mM) Michaelis-Menten component and a low-affinity sigmoidal Hill unit (Km∼ 12.5 mM, Hill coefficient = 3). Notably, the kinetic characteristics of the low-affinity component for GSH uptake in the absence of Na+ appear to be similar to those observed in the hepatic canalicular membrane (Ballatori and Dutczak, 1994). In the presence of a Na+ gradient, total uptake of GSH (dotted curve, Fig. 8) was best described by the sum of a high-affinity (Km∼ 67 μM) Michaelis-Menten component and a low-affinity sigmoidal Hill unit (Km∼ 7.5 mM, Hill coefficient = 2). Analysis of Na+-dependent transport of GSH (solid curve, Fig. 8) represented by the difference in uptake in the presence and absence of Na+ also revealed the presence of two kinetic components. However, both components were adequately described by Michaelis-Menten kinetics with Km values of 350 μM and 3.4 mM, respectively. Clearly, both Na+-dependent and Na+-independent pathways for GSH uptake exist in luminal membrane vesicles.

image

Figure 8. Kinetics of GSH uptake by MBEC-4 luminal plasma membrane vesicles. Data were fitted using the SAAM II program as described in EXPERIMENTAL PROCEDURES. Data are means ± SD from duplicate determinations in three independent preparations. Symbols: KCl gradient (□), inward Na+ gradient (○), and Na+-dependent uptake alone (▴). Best fits to uptake in the absence of Na+ (dashed curve), in the presence of Na+ (dotted curve), and Na+-dependent uptake alone (solid curve) are shown for the entire GSH concentration range studied (A) and for 0.05-1 mM (B). Total uptake of GSH was a sum of high affinity (Km∼ 67 μM) and low affinity (Km∼ 7.5 mM) with a Hill coefficient of nH = 2. The Na+-dependent transport alone was adequately described by Michaelis-Menten kinetics with Km of 350 μM and 3.4 mM, respectively.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL PROCEDURES
  4. GSH transport by MBEC-4 cells
  5. Effect of extracellular GSH on intracellular GSH
  6. RNA isolation, oocyte preparation, and transport
  7. RESULTS
  8. GSH and GGT levels in MBEC-4 cells
  9. Na+ dependency of GSH uptake in MBEC-4 cells
  10. Na+-mediated increase in cellular GSH
  11. Expression of GSH uptake in Xenopus laevis oocytes
  12. Characteristics of GSH uptake in luminal membrane vesicles
  13. DISCUSSION
  14. Acknowledgements
  15. References

In the present study, we have provided evidence for the presence of a Na+-dependent GSH transporter in a mouse brain endothelial cell line. This evidence was gathered from studies in cultured MBEC-4 cells, plasma membrane vesicles, and MBEC-mRNA-injected oocytes. In addition, experimental evidence was obtained for the first time to show that the Na+-dependent GSH transporter is localized at the apical (luminal) membrane of mouse brain endothelial cells. An elevation of cellular GSH by uptake of extracellular GSH mediated by the Na+-dependent GSH transporter was also shown. As mentioned in the introductory section, the SV-40-transformed MBEC-4 cells retained the biochemical and morphological properties of the BBB (Tatsuta et al., 1992). Our findings of the presence of Na+-dependent and Na+-independent GSH transport in these cells confirms our previous work on GSH transport in an in situ brain perfusion model and expression in bovine brain capillary mRNA-injected oocytes (Zlokovic et al., 1994; Kannan et al., 1996). As MBEC-4 cells contained a millimolar amount of GSH and significant GGT activity, precautions were taken to inhibit GCS and GGT prior to uptake measurements. The uptake was specific, saturable, and inhibited by organic anions, as in our previous studies (Kannan et al., 1990, 1992). Confirmation of both uptakes was also obtained by expression of transport in MBEC poly(A)+ RNA-injected oocytes.

We also examined whether the Na+/GSH co-transporter affects endothelial cell GSH concentration. Intracellular GSH concentration was raised by incubation with extracellular GSH only in the presence of Na+ in the incubation medium. Supraphysiologic extracellular concentrations were used in these experiments to be able to detect a change in GSH concentrations. The final GSH concentration achieved in the cells in response to Na+/GSH uptake is difficult to predict as numerous factors such as membrane potential, the ability of cells to sustain the Na+ gradient, and the activity of the efflux transporter could influence the results.

Although the uptake studies in MBEC-4 cells show clearly the presence of Na+-dependent GSH transport, they do not address the question of the location of GSH transporters. Isolated membrane vesicles have been used as a model for studying the symmetry/asymmetry of carrier proteins. There are recent reports that describe selective (luminal or abluminal) or nonselective membrane localization of large neutral amino acids and hexose transporters (Sanchez del Pino et al., 1992, 1995; Lee et al., 1997). As MBEC-4 cells were reported to exhibit polarity, it was of interest to investigate the specific domain localization of the Na+-dependent GSH transporter in the plasma membrane vesicles. The presence of Na+-dependent GSH transport in mixed plasma membrane vesicles was established at first. The mixed membranes were then separated into three distinct fractions. Taking advantage of the recently described localization of P-glycoprotein and Na+/glucose co-transporter to opposite poles, that is, to the luminal and abluminal membranes of brain endothelial cells, these were used as “markers” to identify the 30 and 38% sucrose density fractions from our vesicle preparations as luminal and abluminal, respectively. This was also supported by the relative enrichment of marker enzymes in the two membrane domains (Table 2).

Demonstration of Na+-dependent GSH transport only in the luminal membrane (Fig. 6) is consistent with our hypothesis that this transporter may mediate uphill transport of micromolar amounts of GSH from plasma to millimolar amounts of GSH in the brain endothelial cells in vivo (Kannan et al., 1998). The Na+-independent transporter may function as a facilitative transporter and would have the potential for bidirectional operation, as observed in cell culture studies (Lu et al., 1996). It serves as an effluxer under physiological conditions because net transport would be determined by a concentration gradient. However, further work at the molecular level to clone the individual GSH transporters will be needed to assess the relative contributions of these two transporters in the brain. Such studies are underway in our laboratories.

Acknowledgements

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL PROCEDURES
  4. GSH transport by MBEC-4 cells
  5. Effect of extracellular GSH on intracellular GSH
  6. RNA isolation, oocyte preparation, and transport
  7. RESULTS
  8. GSH and GGT levels in MBEC-4 cells
  9. Na+ dependency of GSH uptake in MBEC-4 cells
  10. Na+-mediated increase in cellular GSH
  11. Expression of GSH uptake in Xenopus laevis oocytes
  12. Characteristics of GSH uptake in luminal membrane vesicles
  13. DISCUSSION
  14. Acknowledgements
  15. References

Able technical assistance was provided by Diana Tang and Vyjayanthi Raghunathan. The authors thank the Cell Culture and Molecular Biology Cores of the University of Southern California Research Center for Liver Diseases for maintenance of cell lines and X. laevis oocytes. This work was supported by National Institutes of Health grant GM 53820.

References

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL PROCEDURES
  4. GSH transport by MBEC-4 cells
  5. Effect of extracellular GSH on intracellular GSH
  6. RNA isolation, oocyte preparation, and transport
  7. RESULTS
  8. GSH and GGT levels in MBEC-4 cells
  9. Na+ dependency of GSH uptake in MBEC-4 cells
  10. Na+-mediated increase in cellular GSH
  11. Expression of GSH uptake in Xenopus laevis oocytes
  12. Characteristics of GSH uptake in luminal membrane vesicles
  13. DISCUSSION
  14. Acknowledgements
  15. References
  • 1
    Aw T.Y., Ookhtens M., Ren C., Kaplowitz N. (1986) Kinetics of GSH efflux from isolated rat hepatocytes. Am. J. Physiol. 250 G236G243.
  • 2
    Ballatori N. & Dutczak W.J. (1994) Identification and characterization of high and low affinity transport systems for reduced glutathione in liver cell canalicular membranes. J. Biol. Chem. 269 1973119737.
  • 3
    Betz A.L., Firth J.A., Goldstein G.W. (1980) Polarity of the blood-brain barrier: distribution of enzymes between the luminal and antiluminal membranes of brain capillary endothelial cells. Brain Res. 192 1728.
  • 4
    Fariss M.W. & Reed D.J. (1987) High performance liquid chromatography of thiols and disulfides: dinitrophenol derivatives. Methods Enzymol. 143 101109.
  • 5
    Fernandez-Checa J.C., Takikawa H., Horie T., Ookhtens M., Kaplowitz N. (1992) Canalicular transport of reduced glutathione in normal and mutant Eisai hyperbilirubinemic rats. J. Biol. Chem. 267 16671673.
  • 6
    Fernandez-Checa J.C., Yi J., Garcia-Ruiz C., Knezic Z., Tahara S.M., Kaplowitz N. (1993) Expression of rat liver sinusoidal and canalicular GSH transport systems in Xenopus laevis oocytes. J. Biol. Chem. 268 23242328.
  • 7
    Fiske C.H. & SubbaRow Y. (1925) The colorimetric determination of phosphorus. J. Biol. Chem. 66 375400.
  • 8
    Hatefi Y. & Stigall D.L. (1978) Preparation and properties of succinate: ubiquinone oxidoreductase (complex II). Methods Enzymol. 53 2127.
  • 9
    Jette L., Murphy G.F., LeClerc J., Beliveau R. (1995a) Interactions of drugs with P-glycoprotein in brain capillaries. Biochem. Pharmacol. 50 17011709.
  • 10
    Jette L., Pouliot J., Murphy G.F., Beliveau R. (1995b) Isoform I (mdr3) is the major form of P-glycoprotein expressed in mouse brain capillaries. Evidence for cross-reactivity of antibody with an unrelated protein. Biochem. J. 305 761766.
  • 11
    Kamath A.V. & Morris M.E. (1998) Functional expression of P-glycoprotein in the hepatic canalicular membrane of developing rats. J. Pharm. Sci. 87 300305.
  • 12
    Kannan R., Kuhlenkamp J.F., Jeandidier E., Trinh H., Ookhtens M., Kaplowitz N. (1990) Evidence for carrier-mediated transport of glutathione across the blood-brain barrier in the rat. J. Clin. Invest. 85 20092013.
  • 13
    Kannan R., Kuhlenkamp J.F., Ookhtens M., Kaplowitz N. (1992) Transport of glutathione at the blood-brain barrier in the rat: inhibition and age-dependence. J. Pharmacol. Exp. Ther. 263 964970.
  • 14
    Kannan R., Yi J., Tang D., Li Y., Zlokovic B.V., Kaplowitz N. (1996) Evidence for the existence of a sodium-dependent glutathione (GSH) transporter. Expression of bovine brain capillary mRNA and size fractions in Xenopus laevis oocytes and dissociation from γ-glutamyltranspeptidase and facilitative GSH transporters. J. Biol. Chem. 271 97549758.
  • 15
    Kannan R., Yi J., Zlokovic B.V., Kaplowitz N. (1998) Carrier-mediated GSH transport at the blood-brain barrier and molecular characterization of novel brain GSH transporters, in Glutathione in the Nervous System (Shaw C. A., ed), pp. 4562. Taylor & Francis, Washington, D.C.
  • 16
    Lee W.J., Peterson D.R., Sukowski E.J., Hawkins R.A. (1997) Glucose transport by isolated plasma membranes of the blood-brain barrier. Am. J. Physiol. 41 C1552C1557.
  • 17
    Leier I., Jedlitschky G., Buchholz U., Keppler D. (1994) Characterization of ATP-dependent leukotriene C4 export carrier in mastocytoma cells. Eur. J. Biochem. 220 599606.
  • 18
    Lu S.C., Sun W., Yi J., Ookhtens M., Sze G., Kaplowitz N. (1996) Role of two recently cloned rat liver GSH transporters in the ubiquitous transport of GSH in mammalian cells. J. Clin. Invest. 97 14881496.
  • 19
    Meier P.J., Sztul E.S., Reuben A., Boyer J.L. (1984a) Structural and functional polarity of canalicular and basolateral membrane vesicles isolated in high yield from rat liver. J. Cell Biol. 98 9911000.
  • 20
    Meier P.J., Meier-Abt A.S., Barrett C., Boyer J.L. (1984b) Mechanisms of taurocholate transport in canalicular and basolateral rat liver plasma membrane vesicles. J. Biol. Chem. 259 1061410622.
  • 21
    Meister A. & Anderson M.E. (1983) Glutathione. Annu. Rev. Biochem. 52 711760.
  • 22
    Nag S. (1990) Ultracytochemical localisation of Na+, K+-ATPase in cerebral endothelium in acute hypertension. Acta Neuropathol. 80 711.
  • 23
    Oldendorf W.H. (1971) Brain uptake of labeled amino acids, amines and hexoses after arterial injection. Am. J. Physiol. 221 16291639.
  • 24
    Orlowsky M. & Karnowsky A. (1976) Glutathione metabolism and some possible functions of glutathione in the nervous system. Int. Rev. Neurobiol. 19 7581.
  • 25
    Sanchez del Pino M.M., Hawkins R.A., Peterson D.R. (1992) Neutral amino acid transport by the blood-brain barrier. J. Biol. Chem. 267 2595125957.
  • 26
    Sanchez del Pino M.M., Peterson D.R., Hawkins R.A. (1995) Biochemical discrimination between luminal and abluminal enzyme and transport activities of the blood-brain barrier. J. Biol. Chem. 270 1491314918.
  • 27
    Shirai A., Naito M., Tatsuta T., Dong J., Hanaoka K., Mikami K., Oh-hara T., Tsuruo T. (1994) Transport of cyclosporin A across the brain capillary endothelial cell monolayers by P-glycoprotein. Biochim. Biophys. Acta 1222 400404.
  • 28
    Sze G., Kaplowitz N., Ookhtens M., Lu S.C. (1993) Bidirectional membrane transport of intact glutathione in Hep G2 cells. Am. J. Physiol. 28 G1128G1134.
  • 29
    Tatsuta T., Naito M., Oh-hara T., Sugawara I., Tsuruo T. (1992) Functional involvement of P-glycoprotein in blood-brain barrier. J. Biol. Chem. 267 2038320391.
  • 30
    Tietze F. (1969) Enzymic method for quantitative determination of nanogram amounts of total and oxidized glutathione: applications to mammalian blood and other tissues. Anal. Biochem. 27 502522.
  • 31
    Zlokovic B.V., Mackic J.B., McComb J.G., Weiss M.H., Kaplowitz N., Kannan R. (1994) Evidence for transcapillary transport of reduced glutathione in vascular perfused guinea-pig brain. Biochem. Biophys. Res. Commun. 201 402408.DOI: 10.1006/bbrc.1994.1715