Abstract
Multidrug resistance-associated protein (MRP) is a transport system that is involved in the elimination of xenobiotics and biologically active endogenous substrates. Recently, the presence of MRP has been demonstrated in cultured brain capillary endothelial cells (BCECs). The time-dependent, functional expression of MRP in porcine BCECs was investigated to assess the value of this cell culture model for drug transport at the blood-brain barrier. Western blot analysis was used to investigate MRP expression in freshly isolated porcine BCECs and compared to MRP expression at days 8 and 10 in culture. Subcellular localization of MRP was investigated by immunocytochemistry with an MRP-specific monoclonal antibody, MRPr1. Functional activity of MRP was assessed by efflux studies with the fluorescent MRP substrate glutathione-methylfluorescein (GS-MF). No significant MRP expression was detected in freshly isolated endothelial cells. However, MRP expression is up-regulated in cell culture in a time-dependent manner. Immunostaining revealed predominantly perinuclear and, to a lesser degree, plasma membrane localization of MRP. At 10°C GS-MF efflux was significantly decreased, indicating the involvement of an energy-dependent transport system. Efflux of GS-MF was apparently inhibited by MK571, a specific inhibitor for MRP. Porcine BCECs demonstrate up-regulation of functional MRP expression during culture, as observed in human tissue, and therefore might serve as a useful in vitro system for studying MRP-mediated blood-brain barrier transport.
The blood-brain barrier (BBB)2 has an important function in controlling the flux of endogenous and xenobiotic substances between the blood and the brain. The importance of multidrug resistance proteins for the drug transport through the BBB is well documented. An important member of the adenosine triphosphate (ATP)-binding cassette transporter family is P-glycoprotein (Gottesman and Pastan, 1993; Tsuji et al., 1993; Schinkel et al., 1994, 1996; Huwyler et al., 1996). Another recently identified member is the multidrug resistance-associated protein (MRP) (Cole et al., 1992). MRP has an apparentMr of 190,000 and functions as an ATP-dependent export pump with broad substrate specificity. Predominant substrates are amphiphilic anions, in particular glutathione conjugates (Müller et al., 1994; Leier et al., 1996). MRP is expressed in various tumors (Leier et al., 1996; Narasaki et al., 1996;Nooter et al., 1997a,b; Welters et al., 1998) and in normal tissue (Flens et al., 1996). Its existence in brain tissue is, however, controversial: when the anti-MRP monoclonal antibodies MRPr-1 and MRPm6 were used, no MRP expression could be detected in human brain tissue sections (Flens et al., 1996). However, by filter hybridization of poly(A+) RNA, small amounts of MRP transcripts were visualized in human brain homogenates (Kruh et al., 1995). Reverse transcriptase-polymerase chain reaction analysis of freshly isolated human brain microvessels revealed no MRP mRNA (Seetharaman et al., 1998). In rat microvessels, only low levels of MRP and MRP mRNA were present (Regina et al., 1998). Interestingly, MRP expression was enhanced considerably in primary cultures of brain capillary endothelial cells (BCECs) as well as in immortalized brain endothelial cell lines (Huai-Yun et al., 1998; Regina et al., 1998; Seetharaman et al., 1998). These studies indicate that MRP expression is very variable. It was therefore the aim of the present study to investigate MRP functional expression and subcellular localization by using an in vitro model of the BBB. This cell culture system consists of readily available primary cultures of porcine BCECs (Huwyler et al., 1996).
Materials and Methods
The MRP-specific inhibitor MK-571 was obtained from Biomol (Plymouth Meeting, PA). Chloromethylfluorescein-diacetate (CMFDA) was purchased from Molecular Probes (Eugene, OR). All other chemicals were obtained from commercial sources of the highest quality available.
Cell Cultures.
The small lung cancer cell line H69, which does not express MRP, and the doxorubicin-selected multidrug variant of H69, H69/ADR, which overexpresses MRP, were used as negative and positive controls, respectively (Mirski et al., 1987; Mirski and Cole, 1989). Cells were cultured in RPMI 1640 medium (Gibco BRL, Basel, Switzerland) supplemented with 10% heat-inactivated fetal calf serum, 4 mMl-glutamine, penicillin (50 U/ml), streptomycin (50 mg/ml), and 1 mM sodium pyruvate. The H69/ADR cell line is maintained in medium containing 0.8 μM doxorubicin. Simian virus 40 large T antigen-immortalized human brain microvascular endothelial cells (SV-HCECs) were generously provided by Dr. D. B. Stanimirovic and Dr. A. Murugarandam from the Institute for Biological Sciences, National Research Council of Canada. SV-HCECs (Muruganandam et al., 1997) were seeded at a density of 10,000 cells/cm2 in 0.5% gelatin-coated T-75 culture flasks. Cells (passage 34) were cultured in medium 199 (Earle’s salts, 148.6 mg/liter GlutaMAX I), 20 mM HEPES, 10% heat-inactivated fetal calf serum, 10 μg/ml heparin, 100 U/ml penicillin, and 100 μg/ml streptomycin (all from Gibco BRL), 5 μg/ml insulin, 5 μg/ml transferrin, and 5 ng/ml selenic acid (ITS-premix; Becton-Dickinson, Bedford, MA). Confluent cells were used at day 7 in culture. The pig BCEC culture model has been described in detail previously (Huwyler et al., 1996). Primary cultures of porcine BCECs were prepared according to the method of Audus and Borchardt (1987) with the following modifications. Cortical gray matter from six fresh porcine brains was minced and incubated in Eagle’s minimum essential medium (MEM; Sigma Chemical Co., St. Louis, MO) containing 0.5% dispase (Gibco BRL) for 2 h. 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 (Gibco BRL) for 4.25 h. The resulting cell suspension was supplemented with 10% horse serum and filtered through a 150-μm nylon mesh. BCECs were isolated at 1000g for 10 min by centrifugation through a continuous 50% Percoll gradient (Pharmacia, Uppsala, Sweden). For uptake experiments, isolated endothelial cells were filtered through a 35-μm nylon mesh before seeding at a density of 100,000 cells/cm2 onto collagen/fibronectin (Boehringer)-coated 24-well cell culture plates. Cells were cultured in medium containing 45% MEM, 45% F12-Nutrient Mixture I-12, 100 μg/ml streptomycin, 100 μg/ml penicillin G, 100 μg/ml heparin, 13 mM NaHCO3, and 20 mM HEPES with 10% heat-inactivated horse serum from Gibco BRL (Audus and Borchardt, 1987).
Western Blot Experiments.
MRP was detected by Western blot analysis with the monoclonal antibody (mAb) MRPr1 (Signet, Bedham, MA). SDS-polyacrylamide gel electrophoresis was performed with a Mini-Protean II apparatus (Bio-Rad, Hercules, CA). Cell homogenates (1.5 mg protein/ml) in SDS-gel sample buffer were loaded onto 10% acrylamide/bisacrylamide gels. After electrophoresis, proteins were transferred electrophoretically (2 h at a constant amperage of 2 mA/cm2) to a 0.45-μm pore size nitrocellulose membrane with a Mini Trans-Blot cell (Bio-Rad). The transfer buffer contained 192 mM glycine, 25 mM Tris, and 20% methanol. The membrane was blocked overnight at 4°C with 5% powdered skimmed milk in PBS (Amimed, Allschwil, Switzerland) containing 0.3% Tween 20 (PBS-T). Washed membranes were incubated with mAb MRPr1 (1 μg/ml) in PBS-T, 0.05% Tween 20, 1% bovine serum albumin, and 1% powdered skimmed milk for 2 h at 37°C. Washed membranes were incubated for 1 h at room temperature with horseradish peroxidase-conjugated rabbit anti-rat IgG or rabbit anti-mouse IgG (1:1000; Dako, Zug, Switzerland) in PBS-T containing 0.05% Tween 20, 1% bovine serum albumin, and 1% milk powder. Membranes were washed in PBS-T, and MRP was visualized by using enhanced chemiluminescence detection (ECL kit by Amersham, Buckinghamshire, UK). Signals were scanned and subjected to densitometric analysis with NIH-image software (National Institutes of Health, Bethesda, MD; NIH-image vl.60).
Immunostaining of MRP.
For immunostaining, isolated brain endothelial cells were seeded onto chamber slides (Nunc, Naperville, IL) at a density of 200,000 cells/cm2. Slides were precoated with poly-d-lysine (Sigma), collagen, and fibronectin. Cells were washed and fixed for 2 min in ice-cold acetone either on day 10 in culture (i.e., 2 days after the cells had reached confluence) or on day 8. Cells were incubated for 2 h at 37°C with anti-MRPr1 mAb (Signet, Dedham, MA; Flens et al., 1994) at a concentration of 20 μg/ml in PBS/3% rabbit serum. Cells were subsequently incubated for 1 h at room temperature with fluorescein isothiocyanate-conjugated rabbit anti-rat IgG (Dako Corp., Santa Barbara, CA) in PBS/3% rabbit serum. To asses the localization of MRP in relation to the nucleus, a 4′,6-diamidino-2-phenylindole (DAPI; Sigma) staining was performed by adding DAPI at a concentration of 1 μg/ml to the secondary antibody. The cells were rinsed in PBS, mounted in FluoroSave (Calbiochem, San Diego, CA), and examined on a Zeiss Axiophot fluorescence microscope equipped with a Zeiss Plan-Neofluar 40× objective.
Efflux Experiments.
For efflux experiments porcine BCECs were grown to confluent monolayers in 24-well cell culture plates for 10 days. The surface area was 2 cm2/well. Cells were washed with Hanks’ balanced salt solution (Gibco BRL). Subsequently the cells were loaded with 10 μmol/liter CMFDA for 60 min at 10°C and washed once. The nonfluorescent CMFDA was used as a prodrug for an MRP. At this temperature CMFDA is well absorbed by the cells, and the fluorescent metabolite glutathione-methylfluorescein (GS-MF) is formed intracellularly by the action of esterase and glutathioneS-transferase (Roelofsen et al., 1997). GS-MF is not excreted by the cells at 10°C, whereas at 37°C GS-MF is actively and specifically excreted by MRP.
Functional activity of MRP was investigated by measuring MRP-mediated transport of GS-MF out of the cells. Time-dependent efflux of the fluorescent glutathione conjugate GS-MF into the medium was determined in the presence and absence of MK-571, a specific inhibitor of MRP (Gekeler et al., 1995). Efflux assays were performed at 37°C and 10°C. After 30 min, cells were lysed and the remaining fluorescence in the cells was measured. All fluorescence measurements were done with a fluorescence enzyme-linked immunoabsorbance assay-plate reader (Fluostar; BMG Labtechnologies, Offenburg, Germany).
Data Analysis.
To obtain estimates of kinetic parameters, a nonlinear regression program was used (Microcal Origin version 5.0). For statistical comparison, data of groups were compared by analysis of variance. The level of significance was p = .05. If this analysis revealed significant differences, pairwise comparisons within groups were performed by two-sided unpaired t tests. Thep values were adjusted by Bonferroni’s correction for multiple comparisons.
Results
Western blot analysis with the anti-MRP mAb MRPr1 demonstrated the presence of MRP as a 190-kDa band in monolayers of cultured BCECs, but this signal was not present in freshly isolated BCECs. However, after 10 days in culture, MRP could be detected (Fig.1). Between days 8 and 10 in culture, the MRP signal increased by a factor of 10. The specificity of the 190-kDa MRP signal was confirmed by control experiments with MRP overexpressing human H69/ADR and MRP nonexpressing parental H69 (Par-H69) cells (Almquist et al., 1995). To compare these results with the MRP expression in human BCECs, expression in SV-HCECs (Muruganandam et al., 1997) was investigated. In this cell line, a clear expression of MRP could be demonstrated. Some known cross-reactivities of the MRPr1 mAb (Flens et al., 1994) in the range of 90 to 100 kDa were observed in human as well as porcine cells.
Immunostaining with MRPr1 antibody revealed that MRP in porcine BCECs on day 8 in culture was predominantly located at intracellular sites (Fig. 2). Visualization of the nucleus by DAPI staining revealed that MRP was located around the nucleus in discrete clusters (Fig. 2, inset). After day 10 in culture, MRP staining was not located solely at the nucleus but could also be observed in the cytosol and at the plasma membrane (Fig.3).
The functional expression of MRP in BCECs was investigated by kinetic assays on day 10 in culture. Time-dependent efflux of GS-MF from porcine BCECs can be quantitated by analysis of cell culture supernatants (Fig. 4). The efflux is temperature-dependent and significantly reduced at 10°C (p < .001) as compared with 37°C. In addition, efflux can be inhibited by MK-571, a specific inhibitor of MRP (statistically significant difference, p < .001; Fig. 4). After 30 min of efflux, the remaining cellular fluorescence was 3-fold higher under conditions of reduced activity of MRP (i.e., at 10°C) as compared with standard conditions (37°C; p< .001). Intracellular fluorescence was also considerably enhanced in the presence of MK-571 (Fig. 5). Thus, MRP is functionally expressed and can be inhibited by lowering the temperature or by the use of a specific MRP inhibitor.
Discussion
MRP, a member of the ATP-binding cassette transporter superfamily, is a transport protein involved in multidrug resistance (Cole et al., 1992). Like P-glycoprotein, a member of the same family of membrane transporters (Gottesman and Pastan, 1993), it reveals a broad, although different, substrate specificity (Müller et al., 1994; Leier et al., 1996). The importance of MRP in multidrug resistance has been widely recognized (Leier et al., 1996; Narasaki et al., 1996; Nooter et al., 1997a,b; Welters et al., 1998), but little is known about the physiological function of MRP. In particular, the importance of MRP for the function of the BBB remains to be established. Studies with normal post-mortem human brain tissue indicate low levels of MRP expression. MRP in brain tissue could not be detected with anti-MRP monoclonal antibodies (Flens et al., 1996; Seetharaman et al., 1998). This could be confirmed in our laboratory (H.G., M.T., J.H., J.D. and G.F., unpublished observation). The same is true for freshly isolated brain microvessels (Regina et al., 1998; Seetharaman et al., 1998). However, porcine BCECs cultured for 10 days revealed a significant up-regulation of MRP at the protein level. Our result is in agreement with recent studies with cultured bovine (Huai-Yun et al., 1998) and rat (Regina et al., 1998) capillary endothelial cells. In these studies the MRPm6 mAb and a polyclonal anti-MRP1 antiserum, respectively, were used. In all species, a protein band of approximately 190 kDa representing MRP1 was detected, whereas the pattern of unspecific cross-reactivities differed. The cross-reactivity at 100 kDa observed in porcine BCECs is very similar to the MRPr1 mAb staining pattern observed in human tissue (Flens et al., 1996).
Because little is known about the localization of MRP in brain microvessel endothelial cells, we investigated the subcellular localization of MRP in porcine BCECs. In nonmalignant tissues, MRP has been reported to be located predominantly at intracellular membranes associated with cytoplasmic structures (Almquist et al., 1995; Flens et al., 1996; Nooter et al., 1997a,b). In tumors showing multidrug-resistant phenotypes, MRP expression was also observed at the plasma membrane (Flens et al., 1996). In porcine BCECs, a significant amount of MRP is located at intracellular sites on day 8. However, staining of the cells at a later stage of confluence (day 10) showed a significantly higher amount of MRP at the plasma membrane. It is tempting to speculate that post-translational processing of MRP is paralleled by a translocation of MRP precursor to the outer cell membrane.
MRP in porcine brain endothelial cells is functional, as shown by efflux experiments. Time-dependent efflux of GS-MF from porcine capillary endothelial cells was temperature-dependent, indicating the involvement of a carrier-mediated process. In addition, efflux could be significantly reduced in the presence of the specific MRP inhibitor MK 571. As a consequence, the remaining intracellular GS-MF was significantly elevated in cells treated with MK 571.
Cultured BCECs show functional MRP expression, although its extent is highly variable. The reasons for this variability are still unknown. Cultured BCECs are separated from their natural environment. Therefore, the absence of humoral and cellular signals under cell culture conditions could be a limiting factor. Nevertheless, this in vitro system allows the study of the effect of MRP on the transport of potential substrates.
Acknowledgments
We thank U. Behrens for excellent technical assistance and Dr. Joyce Baumann for careful reading of the manuscript.
Footnotes
-
Send reprint requests to: Dr. Jürgen Drewe, Division of Gastroenterology and Clinical Pharmacology, University Hospital/Kantonsspital, Petersgraben 4, CH 4031 Basel, Switzerland. E-mail: drewe{at}ubaclu.unibas.ch
-
↵1 Current address: F. Hoffmann-LaRoche Ltd., Pharmaceutical Division, PRPN-C, Basel, Switzerland.
-
This work was supported by the Swiss National Science Foundation Grant 32–052918.97, a research grant from the Sandoz Foundation, the ASTRA Research Fonds of the Department of Internal Medicine of the University Hospital Basel, and a scholarship to M.T. from the Association of Chemical Industries, Basel.
- Abbreviations used are::
- BBB
- blood-brain barrier
- MRP
- multidrug resistance-associated protein
- BCECs
- brain capillary endothelial cells
- CMFDA
- chloromethylfluorescein-diacetate
- DAPI
- 4′,6-diamidino-2-phenylindole
- GS-MF glutathione-methylfluorescein
- mAb, monoclonal antibody
- MEM
- Eagle’s minimum essential medium
- MRPr1
- monoclonal antibody against MRP
- Par-H69 cells
- parental H69 cells
- PBS-T
- PBS containing 0.3% Tween 20
- SV-HCECs
- simian virus 40 large T antigen-immortalized human brain microvascular endothelial cells
- Received June 23, 1998.
- Accepted March 29, 1999.
- The American Society for Pharmacology and Experimental Therapeutics