SEARCH

SEARCH BY CITATION

Keywords:

  • astrocytes;
  • brain;
  • glutathione disulfide export;
  • glutathione transport;
  • multidrug resistance proteins;
  • oxidative stress

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Astrocytes play an important role in the glutathione (GSH) metabolism of the brain. To test for an involvement of multidrug resistance protein (Mrp) 1 and 5 in the release of GSH and glutathione disulfide (GSSG) from astrocytes, we used astrocyte cultures from wild-type, Mrp1-deficient [Mrp1(–/–)] and Mrp5-deficient [Mrp5(–/–)] mice. During incubation of wild-type or Mrp5(–/–) astrocytes, GSH accumulated in the medium at a rate of about 3 nmol/(h.mg), whereas the export of GSH from Mrp1(–/–) astrocytes was only one-third of that. In addition, Mrp1(–/–) astrocytes had a 50% higher specific GSH content than wild-type or Mrp5(–/–) cells. The presence of 50 μm of the Mrp inhibitor MK571 inhibited the rate of GSH release from wild-type and Mrp5(–/–) astrocytes by 60%, but stimulated at the low concentration of 1 μm GSH release by 40%. In contrast, both concentrations of MK571 did not affect GSH export from Mrp1(–/–) astrocytes. Moreover, in contrast to wild-type and Mrp5(–/–) cells, GSSG export during H2O2 stress was not observed for Mrp1(–/–) astrocytes. These data demonstrate that in astrocytes Mrp1 mediates 60% of the GSH export, that Mrp1 is exclusively responsible for GSSG export and that Mrp5 does not contribute to these transport processes.

Abbreviations used
CFTR

cystic fibrosis transmembrane transductance regulator

GSH

glutathione

γGT

γ-glutamyl transpeptidase

GSSG

glutathione disulfide

GSx

total glutathione (amount of GSH plus twice the amount of GSSG)

HX

hypoxanthine

LDH

lactate dehydrogenase

MM

minimal medium

Mrp

multidrug resistance protein

OATP

organic anion transporting polypeptide

SOD

superoxide dismutase

XO

xanthine oxidase

The tripeptide glutathione (GSH, γ-l-glutamyl-l-cysteinylglycine) is essential for the cellular defence against reactive oxygen species. A high concentration of GSH protects cells against a variety of reactive oxygen species (Halliwell and Gutteridge 1999; Dringen 2000). Because many cell types and tissues have been reported to release GSH (Ballatori et al. 2005), this compound appears to have essential functions in the extracellular space. In the brain, the presence of extracellular GSH has been detected in microdialysis studies (Han et al. 1999) and extracellular GSH has been considered to participate in the supply of GSH precursors from astrocytes to neurons as well as in the modulation of signal transduction (Dringen et al. 1999; Janaky et al. 1999; Dringen and Hirrlinger 2003). The source of extracellular GSH in the brain is most likely astrocytes, as only astrocytes release substantial amounts of GSH among the different types of neural cells investigated in culture (Hirrlinger et al. 2002a). Astrocytes play a very important role in the metabolism of the brain (Kirchhoff et al. 2001) and in the defence of this organ against oxidative stress (Dringen and Hirrlinger 2003). The GSH released by these cells is processed by ectoenzymes to cysteine (Dringen et al. 1999; Han et al. 1999; Dringen et al. 2001) which is the essential extracellular precursor of neuronal GSH (Dringen and Hirrlinger 2003).

Multidrug resistance proteins (Mrps) are members of the subgroup ABCC of the superfamily of ATP-binding cassette (ABC) transporters (Borst and Oude Elferink 2002; Kruh and Belinsky 2003; Schinkel and Jonker 2003). The genes of nine Mrp family members have been identified in the human genome, whereas in the mouse genome the orthologue of Mrp8 is not present (Kruh and Belinsky 2003). Mrps are large proteins that contain 12–17 transmembrane-spanning helices which are organized in two or three membrane domains. Mrps are ATP-driven export pumps which mediate the cellular export of organic anions (Kruh and Belinsky 2003). Their function was first described in the resistance of tumour cells against chemotherapeutical drugs (Cole et al. 1992). In vivo Mrps fulfil several essential transport functions, depending on the expressing cell type and tissue. Classical Mrp substrates are glutathione-S-conjugates, glutathione disulfide (GSSG), conjugates of glucuronate cyclic nucleotides as well as nucleotide analogues (König et al. 1999; Homolya et al. 2003; Kruh and Belinsky 2003).

Substantial evidence has been presented for the expression of Mrps in brain astrocytes. The presence of Mrp1 protein has been demonstrated for astrocytes of rat brain (Mercier et al. 2004), as well as for cultured astrocytes from rat (Decleves et al. 2000; Hirrlinger et al. 2001; Mercier et al. 2003), human (Marroni et al. 2003) and mouse brain (Gennuso et al. 2004). In contrast, mRNA for Mrp2 was not detected in rat brain, cultured rat astrocytes (Hirrlinger et al. 2001, 2002b), or individual astrocytes in mouse brain (Hirrlinger et al. 2005), nor was Mrp2-immunoreactivity found in human brain (Nies et al. 2004). The presence of mRNAs for Mrps 3, 4 and 5 have been reported for cultured rat astrocytes (Hirrlinger et al. 2001; Ballerini et al. 2002; Hirrlinger et al. 2002b). In addition, individual astrocytes in mouse brain contain mRNAs of Mrp1, Mrp4 and/or Mrp5 (Hirrlinger et al. 2005). Moreover, Mrp4 protein was found in astrocytes in human brain sections (Nies et al. 2004).

The Mrps 1 and 2 have been identified as participating in the co-substrate-independent cellular export of GSSG and GSH (Leier et al. 1996; Paulusma et al. 1999; Ballatori et al. 2005). In addition to these Mrps, other Mrp family members, organic anion transporting polypeptides (OATPs) as well as the cystic fibrosis transmembrane conductance regulator (CFTR) have been reported to contribute to cellular GSH export (for overview see Ballatori et al. 2005). While Mrps and CFTR are ATP-driven exporters, OATPs are anti-porters that use the concentration gradient of GSH to import substrates (Li et al. 1998, 2000).

Although GSH export from cultured astrocytes has been reported from several groups (Yudkoff et al. 1990; Sagara et al. 1996; Dringen et al. 1997; Hirrlinger et al. 2002a; Stewart et al. 2002), the transporter(s) involved in this process have not been identified. However, results of experiments using inhibitors of the Mrp-mediated transport processes suggest that Mrps are involved in the release of GSH from astrocytes under unstressed conditions and in the export of GSSG during oxidative stress (Hirrlinger et al. 2001, 2002a,c).

The availability of genetically modified mouse lines that are deficient of individual Mrps now makes it possible to study the contribution of individual Mrps in substrate transport processes.

In order to test for an involvement of Mrp1 and Mrp5 in GSH and GSSG export from astrocytes in the absence of inhibitors which lack specificity for individual Mrps, we have generated for the first time astrocyte cultures derived from the brains of Mrp1-deficient [Mrp1(–/–)] and Mrp5-deficient [Mrp5(–/–)] mice. Here, we compare these cultures with wild-type astrocyte cultures regarding their ability to export GSH under unstressed conditions and GSSG during peroxide stress.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Materials

Acivicin, allopurinol, bovine serum albumin, 5,5′-dithiobis(2-nitrobenzoic acid), hypoxanthine (HX), indomethacin, ouabain, probenecid, prostaglandin E1, sulfosalicylic acid, and xylenol orange were obtained from Sigma (Deisenhofen, Germany). Catalase, fetal calf serum, GSSG, glutathione reductase, the High-Pure PCR Product Purification Kit, oligo(dT)18 adapter primers, superoxide dismutase (SOD), and xanthine oxidase (XO) were purchased from Roche Diagnostics (Mannheim, Germany). NADPH and NADH were from Applichem (Darmstadt, Germany). Primers were purchased from MWG-Biotech AG (Ebersberg, Germany). The RNAeasy-kit was from Qiagen (Hilden, Germany). Dulbecco's modified Eagle's medium was from Life Technologies (Eggenstein, Germany). MK571 and glibenclamide were obtained from Alexis (Grünberg, Germany). Penicillin G, streptomycin sulfate and Triton X-100 were from Serva (Heidelberg, Germany). All other chemicals were obtained at analytical grade from Merck (Darmstadt, Germany). Sterile plastic material, 24-well culture dishes for cell culture and unsterile 96-well microtitre plates were from Nunc (Wiesbaden, Germany) and Greiner (Frickenhausen, Germany).

Cell cultures

Astroglia-rich primary cultures derived from the brains of neonatal wild-type FVB mice, Mrp1(–/–) mice (Wijnholds et al. 1997) and Mrp5(–/–) mice on an FVB genetic background (J. Wijnholds and P. Borst, unpublished data) were prepared and maintained as described (Hamprecht and Löffler 1985). Three hundred thousand viable cells were seeded per well of a 24-well dish in 2 mL medium consisting of 90% Dulbecco's modified Eagle's medium, 10% fetal calf serum, 20 units/mL of penicillin G and 20 μg/mL of streptomycin sulfate and cultivated in a cell incubator (Heraeus, Hanau, Germany) containing a humidified atmosphere of 10% CO2/90% air. The medium was renewed every seventh day. The results were obtained with 14- to 23-day-old cultures. The absence of Mrp1 and Mrp5 in Mrp1(–/–) and Mrp5(–/–) astrocyte cultures, respectively, was confirmed by western blot analysis of homogenates of the astrocyte cultures using specific antibodies for Mrp1 (antibody MRPr1; Signet Pathology Systems Inc., Dedham, MA, USA) and Mrp5 (antibody NKI-12C5; Wielinga et al. 2002) (data not shown).

Experimental incubation

To study export of GSH, the cells were washed with 2 mL of pre-warmed (37°C) minimal medium (MM; 44 mm NaHCO3, 110 mm NaCl, 1.8 mm CaCl2, 5.4 mm KCl, 0.8 mm MgSO4, 0.92 mm NaH2PO4, 5 mm glucose, adjusted with CO2 to pH 7.4) and incubated in the cell incubator with 1 mL incubation medium (MM with 100 μm of the γ-glutamyl transpeptidase (γGT) -inhibitor acivicin (Dringen et al. 1997) containing MK571 (or other compounds) in the concentrations indicated. For analysis of the total glutathione content (GSx = amount of GSH plus twice the amount of GSSG) and the GSSG content in cells, the medium was collected and the cells were washed with 2 mL phosphate-buffered saline (10 mm potassium phosphate buffer, 150 mm NaCl, pH 7.4) and lysed with 500 μL 1% (w/v) sulfosalicylic acid. For determination of the GSx content in the medium, samples of 10 μL medium were collected after the time periods indicated and mixed with 90 μL of 0.11% (w/v) sulfosalicylic acid in a well of a microtitre plate. For determination of the content of GSSG in the medium, the GSH present was derivatized with 2-vinylpyridine according to the method originally described by Griffith (1980). Medium (100 μL) was mixed with 100 μL 1% sulfosalicylic acid, and 130 μL of this mixture was subjected to derivatization with 2-vinylpyridine as described previously (Dringen and Hamprecht 1996; Dringen et al. 1997).

GSSG export was investigated after application of an H2O2-induced permanent oxidative stress to cultured astrocytes. The cells were washed with 2 mL of pre-warmed (37°C) incubation medium (20 mm HEPES, 145 mm NaCl, 1.8 mm CaCl2, 5.4 mm KCl, 1 mm MgCl2, 0.8 mm Na2HPO4, 5 mm glucose, pH 7.4) and incubated at 37°C in 0.5 mL incubation medium containing XO (in volume activities of twice that indicated in the legends of the figures) and SOD (200 U) for 10 min before 0.5 mL HX (2 mm in incubation medium; 37°C) was added per well to initiate the production of H2O2. The superoxide radical anions generated by the XO-catalyzed oxidation of HX or xanthine are dismutated by SOD to H2O2 and O2. The cells were incubated with the H2O2-generating system on a metal grid warmed to 37°C by a water bath. During the incubation, 10 μL aliquots were harvested to monitor the extracellular concentration of H2O2. For analysis of GSx and GSSG contents, the cells were washed with 2 mL phosphate-buffered saline and lysed with 500 μL 1% (w/v) sulfosalicylic acid. In order to dispose of the H2O2 in the incubation medium before analysis of GSx and GSSG contents and before quantitation of the activity of lactate dehydrogenase (LDH) in the medium as indicator for cell damage, 500 μL medium was mixed with 50 μL incubation medium containing the XO-inhibitor allopurinol (55 μm) and catalase (1100 U/mL) (Hirrlinger et al. 1999, 2001).

Determination of extracellular hydrogen peroxide

During incubation of cultured astrocytes with the peroxide generating system, the extracellular concentration of H2O2 was monitored by the colorimetric method described previously (Dringen et al. 1998).

Determination of GSx and GSSG

The contents of GSx and GSSG in cell lysates and incubation media were determined as described previously (Dringen and Hamprecht 1996; Dringen et al. 1997) in microtitre plates according to the colorimetric method originally described by Tietze (1969).

RT–PCR analysis of mRNA expression

Total RNA was isolated from astroglia-rich primary cultures of wild-type mice using the RNAeasy-kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. One microgram of total RNA was reverse transcribed in 20 μL of transcription buffer (supplied as 5 × reaction buffer by the manufacturer) with 1 μg oligo(dT)18-adapter primer and 200 U RevertAid H Minus M-MuLV reverse transcriptase according to the protocol supplied by the manufacturer. The resulting cDNAs were purified using the High-Pure PCR Product Purification Kit (Roche Diagnostics) according to the manufacturer's instructions. The sequence information for the mouse-specific oligonucleotides used in the following PCRs was obtained from the known mouse cDNA sequences of the transporters analysed (for accession numbers see Table 1). PCRs were performed in a total volume of 50 μL of PCR buffer (supplied as 10 × reaction buffer by the manufacturer) containing 1.5 mm MgCl2, 1.25 U of Taq DNA polymerase, 0.5 μm forward and reverse primers (Table 1) and 5 μL of purified cDNAs. Cycling conditions were: 2 min at 94°C, followed by 35 cycles with 1 min at 94°C, 1 min at the annealing temperature given for the individual cDNAs in Table 1, and 1 min at 72°C, and a final 10 min at 72°C. The integrity of the isolated RNA and the reverse transcription reaction were examined using β-actin specific primers. All PCR products were analysed on a 1% agarose gel and had the size expected from the known cDNA sequences of the respective transporters. Liver mRNA was used as a positive control to demonstrate the detectability of mRNAs of Mrp2 and Mrp6 by the protocol applied (data not shown). The conditions used here for RT–PCR analyses allow the highly sensitive qualitative detection of the presence of an individual mRNA, but not the quantitative comparison of mRNA expression levels of one transporter in various samples.

Table 1.  Primer pairs used for the amplification of cDNAs for mouse transport proteins
 Forward primerReverse primerAccession no.Annealing temperature (°C)Size (bp)
Mrp1TGCGCTTCCCACTCAACATCCCGGGCCAGGCTCACACG NM_008576 50562
Mrp2CTTGCGGTGGTCCAGTGTTTTACATGGCGAATGGCAGACAA NM_013806 50571
Mrp3GCAGAACAAGACCTCCGTAATTCTGGTTGTTGTCCACCTT XM_126648 50537
Mrp4AACTGCGGCTTTCACGGATGCCCAACCACGGCTAACAACTCAC AY415508 58546
Mrp5GTCTCCTTTCCTCTCCCACATCACCTTGATGCAGCCTCCAGATAGC NM_013790 50616
Mrp6AGTCGGGAACCTGCTGAACCCACGCGCTCCACGGCTACCAT NM_018795 58589
Mrp7CTATTCGGGAGAACCTGGACAGGCTCAGCACCCTGAAG AF4066421 58477
Mrp9GAGATGCTCTGGCATGTTTTGGACCTTTACAGTCCAACCTC AF502146 58417
CFTRACAGTCATCAACGGAATCGTCCCAAAATCAACATCAGGAG NM_021050 54443
OATP1 CATTCCGGCACCTGTTTACTCAGGCCAGTTGAAAACTCATT AF223067 54353
OATP2 TTGGCTGTTTCATCTTCTCGCCTCATCACAGCTTAGTTTTCC AB031814 54463
β-actinCTCTAGGCACCAAGGTGTGAGGAGAGCATAGCCCTCGTAG M12481 54404

Determination of cell viability and protein content

Cell viability was analysed by determining the activity of LDH in the incubation medium and in the cells using the microtitre plate assay described previously (Dringen et al. 1998). The protein content per well of a 24-well dish was quantified after solubilization of the cells in 0.5 mL of 0.5 m NaOH (Hirrlinger et al. 1999) according to the method described by Lowry et al. (1951), using bovine serum albumin as a standard.

Presentation of data

The experiments shown in the figures were carried out on at least three independently prepared cultures with comparable results. The results shown are those of one representative experiment presented as mean values ± SD of triplicates obtained from replica wells of one culture. In the figures, the bars have been omitted if they were smaller than the symbols representing the mean values. The results shown in the tables represent mean values ± SD of data obtained from n independently prepared cultures. Statistical analysis was performed using anova followed by Tukey's post-hoc test.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Contents of GSx and GSSG in astrocyte cultures

To test for a contribution of Mrp1 and Mrp5 in the export of GSH from astrocytes, astroglia-rich primary cultures were prepared from the brains of wild-type, Mrp1(–/–) and Mrp5(–/–) mice. Cultures from the three mouse lines did not show any morphological difference, nor any difference in viability during the incubation paradigms used here. Compared with wild-type astrocytes, the specific total glutathione content (GSx) of Mrp1(–/–) astrocytes was significantly increased by 50%, whereas Mrp5(–/–) astrocytes did not differ in their GSx content from wild-type astrocytes (Table 2). In astrocytes from all three investigated mouse lines, GSSG was hardly detectable in the cells and accounted for less than 1% of GSx (Table 2).

Table 2.  Initial intracellular GSx and GSSG contents and GSH export rates of astrocyte cultures derived from wild-type and Mrp-deficient mice
 GSx content (nmol/mg) n GSSG content (nmol/mg) n GSH export rate (nmol/(h.mg))(%/h) n
  1. Astroglia-rich primary cultures were incubated for up to 6 h in minimal medium in the presence of acivicin (100 μm). Specific release rates were calculated from the slope of the increase in extracellular GSx within the incubation period between 15 min and 6 h. The data represent mean values ± SD of n experiments, each performed in triplicate, using n independently prepared astrocyte cultures. The values obtained for Mrp1(–/–) astroglial cultures were significantly different (*p < 0.001) from those obtained for astroglial cultures from wild-type and Mrp5(–/–) mice. No significant differences (p > 0.05) were found between the values obtained from wild-type and Mrp5(–/–) astrocytes.

Wild type31.4 ± 7.070.1 ± 0.252.83 ± 0.679.4 ± 1.55
Mrp1(–/–)47.3 ± 4.3*60.1 ± 0.250.97 ± 0.36*2.0 ± 0.5*5
Mrp5(–/–)27.5 ± 6.350.1 ± 0.142.62 ± 0.3910.4 ± 2.35

Export of GSH from astrocytes

During incubation of wild-type astrocytes, GSx accumulated in the medium (Fig. 1a). Analysis of the oxidation state of the GSx in the medium revealed that, predominantly, GSH was present (Fig. 1a). After incubation for 6 h, GSSG contributed to only 9 ± 8% (n = 4 experiments, each performed in triplicate) of the extracellular GSx determined. As the cells were incubated in the presence of the γGT inhibitor acivicin, consumption of extracellular GSH by γGT was prevented (Dringen et al. 1997). The increase in extracellular GSx was proportional to the incubation time (Fig. 1a, open circles). Therefore, rates of GSH release were calculated over the incubation period between 15 min and 6 h. Wild-type astrocytes released GSH into the medium at an average rate of 2.8 ± 0.7 nmol/(h.mg) (Table 2). After 6 h of incubation, 60 ± 11% (n = 4 experiments, each performed in triplicate) of the initial cellular GSx content was found in the medium, indicating a GSH release of 9% of the initial GSx content per h (Table 2). During the incubation, the cellular GSx content, which accounted almost exclusively for GSH, declined at a rate corresponding to the extracellular accumulation of GSx (Fig. 1a). The GSx content per well (sum of intracellular and extracellular GSx) remained almost unchanged during incubation of the astrocytes for 6 h (Fig. 1a), suggesting that, under the conditions used (no extracellular amino acids present during the incubation), no substantial synthesis of GSH occurred. The viability of the cells during this incubation was not compromised, as demonstrated by the lack of LDH release from the cells (data not shown). Consequently, this experimental system was considered suitable for studying the mechanism of GSH release from viable astrocytes.

image

Figure 1. Time course of the intracellular (filled symbols) and extracellular (open symbols) contents of GSx (circles) and GSSG (triangles) in astroglia-rich primary cultures derived from (a) wild-type, (b) Mrp1(–/–) and (c) Mrp5(–/–) mice. Filled squares represent the sum of intracellular and extracellular GSx. The data are from representative experiments that were reproduced with comparable results on four independently prepared cultures. The initial GSx contents (100% values) of the three cultures were (a) 24.3 ± 1.9 nmol/mg protein, (b) 59.0 ± 5.2 nmol/mg protein and (c) 20.7 ± 10.8 nmol/mg protein. The protein contents of the cultures were (a) 187 ± 31 μg, (b) 113 ± 5 μg and (c) 213 ± 18 μg per well.

Download figure to PowerPoint

Identical experiments to that presented for wild-type astrocytes (Fig. 1a) were also performed on astrocyte cultures derived from Mrp1(–/–) and Mrp5(–/–) mice (Figs 1b and c). Compared with wild-type cells, no difference in the export of GSH was observed for Mrp5(–/–) astrocytes (Table 2; Fig. 1c). In contrast, GSH export from Mrp1(–/–) astrocytes and extracellular accumulation of GSH were only one-third of that observed for wild-type astrocytes (Table 2; Fig. 1b). After 6 h of incubation, 19 ± 4% (n = 3 experiments, each performed in triplicate) and 69 ± 19% (n = 4 experiments, each performed in triplicate) of the initial GSx content of the cells were found in the medium of Mrp1(–/–) and Mrp5(–/–) astrocytes, respectively. Mrp1(–/–) astrocytes released GSH at a rate of 2% of the initial GSx per hour, whereas Mrp5(–/–) astrocytes exported 10% of their initial GSH per h (Table 2).

Expression of putative GSH transporters in mouse astrocytes

Various Mrps, OATPs and CFTR have been connected to GSH transport processes (Ballatori et al. 2005). To test for the expression of such transporters in mouse astrocyte cultures, the presence of transporter mRNAs was investigated by RT–PCR analysis. RNA was isolated from wild-type astrocyte cultures and single-stranded cDNAs were synthesized from the mRNAs by reverse transcription. With primer pairs specific for the detection of the mouse Mrps, OATPs and CFTR (Table 1), cDNA fragments of the expected sizes were amplified for Mrp1, Mrp3, Mrp4, Mrp5, Mrp7, Mrp9, OATP1, OATP2 and CFTR (Fig. 2). In contrast, at best, very weak signals for Mrp2 and Mrp6 were found for these cultures (Fig. 2).

image

Figure 2. RT–PCR analysis of the expression of transporters in wild-type mouse astroglia-rich cultures. Agarose gel separation of the amplification products showed cDNA fragments (indicated by asterisks) of the expected size (Table 1) for Mrp1, Mrp3, Mrp4, Mrp5, Mrp7, Mrp9, CFTR, and OATPs 1 and 2. The integrity of the isolated mRNA and the reverse transcription reaction were tested by using β-actin-specific primers.

Download figure to PowerPoint

Effects of transport modulators on the GSH export from astrocytes

MK571 modulates Mrp1-mediated transport processes (Leier et al. 1994, 1996). Therefore, the export of GSH from astrocytes of the three mouse lines was investigated in the presence of MK571. An incubation of astrocyte cultures with MK571 in concentrations of up to 50 μm did not compromise the viability of the cells and did not alter the sum of intracellular plus extracellular GSx nor the ratio of intracellular or extracellular GSH to GSSG (data not shown).

The concentration of GSH detected in the medium of mouse astrocytes increased linearly with the incubation time in the absence and the presence of MK571 in concentrations of up to 50 μm (data not shown), allowing calculation of release rates for GSH. The release rate of GSH from wild-type and Mrp5(–/–) astrocytes in the presence of MK571 showed an unusual concentration dependency for the inhibitor. Low concentrations of MK571 (< 10 μm) increased the rate of accumulation of GSH in the medium, whereas MK571 in a concentration of 50 μm significantly reduced the extracellular accumulation of GSH (Fig. 3; Table 3). A maximum of the stimulating effect of MK571 on the GSH release rate (about 40% compared with controls – absence of MK571) was found for concentrations of 1–5 μm (Fig. 3). In contrast, MK571 in concentrations above 10 μm reduced the GSH release rate of wild-type and Mrp5(–/–) astrocytes by 60% (Fig. 3; Table 3). In contrast to wild-type and Mrp5(–/–) cells, astrocytes from Mrp1(–/–) mice had a strongly reduced GSH export rate which was not significantly stimulated by MK571 at low concentrations nor inhibited by MK571 at a concentration above 10 μm (Fig. 3; Table 3).

image

Figure 3. Effects of MK571 on the rates of GSH release from astroglia-rich primary cultures derived from wild-type, Mrp1(–/–) and Mrp5(–/–) mice. The cells were incubated for up to 6 h in incubation medium containing MK571 in the concentrations indicated. Release rates were calculated from the slope of the linear increase in extracellular GSx within the incubation period between 15 min and 6 h. The data represent mean values ± SD from triplicates obtained in representative experiments. The initial GSx contents of the three cultures were 37.6 ± 2.6 nmol/mg protein (wild-type), 53.1 ± 1.5 nmol/mg protein [Mrp1(–/–)] and 31.9 ± 2.1 nmol/mg protein [Mrp5(–/–)]. The protein contents of the cultures were 145 ± 8 μg (wild-type), 162 ± 14 μg [Mrp1(–/–)] and 183 ± 18 μg [Mrp5(–/–)] per well.

Download figure to PowerPoint

Table 3.  Modulation of GSH export rates by MK571 in astrocytes
AstrocytesGSH export rate (nmol/(h.mg)) n
0 μm MK5711 μm MK57150 μm MK571
  1. Astroglia-rich primary cultures were incubated for up to 6 h in minimal medium in the presence of acivicin (100 μm) and of MK571 in the concentrations indicated. Specific release rates were calculated from the slope of the increase in extracellular GSx within the incubation period between 15 min and 6 h. The data represent mean values ± SD of n experiments, each performed in triplicate, using n independently prepared astrocyte cultures. GSH export rates of Mrp1(–/–) astroglial cultures were significantly different (*p < 0.001) from those obtained for astroglial cultures from wild-type and Mrp5(–/–) mice. For wild-type and Mrp5(–/–) astrocytes, but not for Mrp1(–/–) cells, the GSH export rates of cultures that were incubated in the absence of MK571 were significantly different (p < 0.01; p < 0.001) from those of cultures incubated in the presence of 1 or 50 μm MK571.

Wild type3.34 ± 1.114.79 ± 1.241.23 ± 0.304
Mrp1(–/–)0.92 ± 0.32*0.96 ± 0.41*1.33 ± 0.484
Mrp5(–/–)2.94 ± 0.284.24 ± 0.611.13 ± 0.333

In addition to Mrp1, a variety of other Mrps, as well as OATPs and CFTR, are expressed in astroglial cultures (Fig. 2). To test for an involvement of such transporters in GSH export, various compounds that have been reported to modulate transport processes mediated by these transporters (Table 4) were applied to cultured Mrp1(–/–) astrocytes in concentrations (Table 4) that have been reported to modulate the respective transport processes. However, compared with controls none of the compounds applied affected the residual GSH export from Mrp1(–/–) astrocyte cultures (Table 4).

Table 4.  Effect of inhibitors of transporters on GSH export from Mrp1(–/–) astrocyte cultures
CompoundTransporter addressedReferencesConcentration usedGSH export (% of control) n
  1. Astroglia-rich primary cultures derived from Mrp1(–/–) mice were incubated for up to 6 h in minimal medium in the presence of acivicin (100 μm) with or without the compounds indicated. Specific release rates were calculated from the slope of the increase in extracellular GSx within the incubation period between 15 min and 6 h. The data given are the results of two independent experiments performed in triplicate or representing mean values ± SD of n experiments, each performed in triplicate, on independently prepared astrocyte cultures. One hundred per cent GSH export corresponds to 0.86 ± 0.33 nmol/(h.mg) (n = 10).

ProbenecidMrp4 van Aubel et al. (2002)  1 mm106 ± 33; 160 ± 612
Mrp5 Wielinga et al. (2003)    
IndomethacinMrp4 Reid et al. (2003) 50 μm109 ± 16; 70 ± 32
Prostaglandin E1Mrp4 Reid et al. (2003) 40 μm104 ± 20; 81 ± 172
GlibenclamideCFTR Egan et al. (2002)  5 mm101 ± 155
OuabainOATPs 1 + 2 Schwab et al. (2001)  1 mm102 ± 20; 63 ± 232

Release of GSSG after application of an H2O2-generating system

The effect of a permanent H2O2-induced oxidative stress on the levels of intra- and extracellular GSSG in cultured brain astrocytes was investigated after application of a H2O2-generating system consisting of XO, HX and SOD (Hirrlinger et al. 1999). This system generated almost identical extracellular concentrations of H2O2 in the three types of astrocyte cultures investigated (Table 5), which were almost constant between 15 and 45 min of incubation and depended on the activity of XO applied (Figs 4a–c). Incubation of the cells with HX and SOD in the absence of XO (filled circles) did not generate detectable amounts of H2O2 in the medium (Figs 4a–c). Compared with control incubations (absence of XO), the presence of the complete H2O2-generating system did not alter the amount of LDH found released from the cells which contributed to less than 10% of total LDH for all conditions and all cultures used (Table 4). In addition, the morphology of the cells and the amount of cellular protein determined after washing of the cells was indistinguishable between control and peroxide-treated cells (data not shown). Therefore, the cells were considered as viable under the conditions used.

Table 5.  Effects of chronic H2O2 stress on the GSSG export from cultured astrocytes
 XO (mU/mL)H2O2 (μM)GSSG content Cells (% of initial GSx)MediumLDH release (% of total LDH) n
  1. The cells were incubated with HX (1 mm) and SOD (100 U) in the absence or the presence of XO in the activities indicated. The extracellular concentration of H2O2, the extracellular activity of LDH as well as the intra- and extracellular contents of GSSG were determined after 30 min of incubation. The data represent mean values ± SD of values obtained from experiments on n independently prepared cultures, each performed in triplicate. The mean values of the initial GSx contents of the three cultures were 34.2 ± 1.9 nmol/mg (wild type), 52.1 ± 4.0 nmol/mg [Mrp1(–/–)] and 29.5 ± 9.3 nmol/mg [Mrp5(–/–)].

Wild type 02 ± 11 ± 12 ± 33 ± 23
1056 ± 1514 ± 420 ± 114 ± 13
20108 ± 1019 ± 929 ± 114 ± 13
30155 ± 1222 ± 837 ± 133 ± 13
Mrp1(–/–)02 ± 20 ± 02 ± 22 ± 24
1054 ± 2349 ± 131 ± 15 ± 64
20108 ± 4359 ± 81 ± 14 ± 54
30144 ± 5765 ± 101 ± 24 ± 24
Mrp5(–/–)01 ± 10 ± 00 ± 05 ± 43
1060 ± 1316 ± 725 ± 93 ± 53
20115 ± 2619 ± 632 ± 73 ± 43
30160 ± 2727 ± 938 ± 103 ± 33
image

Figure 4. Effects of a chronic H2O2-induced oxidative stress on the glutathione content in astroglia-rich primary cultures derived from (a, d, g ) wild-type mice, (b, e, h, k) Mrp1(–/–) mice and (c, f, i, l) Mrp5(–/–) mice. The cells were incubated with HX (1 mm) and SOD (100 U) in the absence (filled circles) or the presence of 10 mU (open circles), 20 mU (filled triangles), or 30 mU (open triangles) of XO. The parameters given were calculated for the cells of one well or the medium (1 mL) per well. (a–c) Concentration of H2O2; (d–f) GSx content in the medium; (g–h) GSSG content in the medium expressed as nmol GSx (1 nmol GSSG corresponds to 2 nmol GSx); (j–l) GSSG content in the cells expressed as a per cent of GSx. The initial GSx contents of the three cultures were 36.2 ± 1.9 nmol/mg protein (wild-type), 50.0 ± 4.5 nmol/mg protein[Mrp1(–/–)] and 31.1 ± 1.2 nmol/mg protein [Mrp5(–/–)]. The protein contents of the cultures were 155 ± 32 μg (wild-type), 117 ± 3 μg [Mrp1(–/–)] and 156 ± 11 μg [Mrp5(–/–)] per well.

Download figure to PowerPoint

During incubation of the cells, with or without the H2O2-generating system, the sum of intracellular plus extracellular GSx was almost constant (data not shown). Incubation of the cells with HX and SOD in the absence of XO (open circles) did not cause GSSG accumulation in the cells (Figs 4j–l) or in the medium (Figs 4g–i). In contrast, immediately after application of the intact H2O2-generating system, the amount of intracellular GSSG increased strongly during the incubation in all three types of astrocytes (Figs 4j–l). Although the extracellular concentrations of H2O2 were almost identical at a given XO activity for all three types of astrocytes (Figs 4a–c), the intracellular ratio of GSSG to GSx established by the peroxide stress varied substantially (Figs 4j–l; Table 4). During incubation of wild-type and Mrp5(–/–) astrocytes with 30 mU/mL XO GSSG accounted for less than 50% of GSx (Figs 4j and l). In contrast, Mrp1(–/–) astrocytes that were incubated under identical conditions accumulated GSSG to up to 70% of GSx (Fig. 4k).

In the presence of the H2O2-generating system, the extracellular content of GSx was quickly elevated in wild-type and Mrp5(–/–) astrocytes (Figs 4d and f) by accumulation of GSSG (Figs 4g and i). Only small amounts of GSH (the difference between the amounts of GSx and GSSG) were found in the medium under these conditions (compare Figs 4d and f with Figs 4g and i, respectively). Consequently, a transfer of GSSG from the cells into the incubation medium was observed for wild-type and Mrp5(–/–) astrocytes, if substantial amounts of GSSG were generated intracellularly by incubation with the H2O2-generating system. In contrast, no accumulation of GSSG was found in the medium of Mrp1(–/–) astrocytes during incubation for up to 45 min with the peroxide generating system (Figs 4e and h; Table 4).

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Up until now, studies of the export of GSH from astrocytes have predominantly used cultures obtained from rat brain (Yudkoff et al. 1990; Sagara et al. 1996; Dringen et al. 1997; Hirrlinger et al. 2002a; Stewart et al. 2002). Here, we demonstrate that astrocyte cultures from mouse brain also release GSH. The rate of GSH export from wild-type mouse astrocytes was 2.8 ± 0.7 nmol/(h.mg) and therefore in the range of 2.1 and 3.2 nmol/(h.mg) previously calculated for rat astrocytes (Sagara et al. 1996; Dringen et al. 1997; Hirrlinger et al. 2002a). Consequently, mouse astrocytes appear to be a useful model to study GSH export from astrocytes. Under unstressed conditions, only small amounts of GSSG were found in the medium of the cells. The origin of this GSSG is most likely extracellular oxidation of GSH in the medium. The observation that, under unstressed conditions, predominantly GSH and not GSSG is released from cultured astrocytes confirms previous reports (Sagara et al. 1996; Hirrlinger et al. 2002a). The reason for the surprising stability of the GSH exported from astrocytes is probably a superoxide dismutase-like factor that is released by astrocytes and that protects GSH against oxidation (Stewart et al. 2002).

In contrast to wild-type astrocytes, GSH export from Mrp1-deficient mouse astrocytes was reduced by 60%, demonstrating that Mrp1 is responsible for the majority of GSH export from mouse astrocytes. This reduction in GSH export of Mrp1(–/–) astrocytes compared with wild-type cells matches perfectly the 60% inhibition of GSH export from wild-type astrocytes in the presence of 50 μm MK571, supporting the view that, among the MK571-sensitive transporters, Mrp1 is exclusively involved in GSH export from wild-type astrocytes. In addition, the reduced GSH export is likely to be responsible for the elevated specific GSx content of Mrp1(–/–) astrocytes compared with wild-type cells. Elevated GSH levels have also been observed in some peripheral organs of Mrp1(–/–) mice compared with the respective Mrp1-expressing tissues of wild-type mice (Lorico et al. 1997), suggesting that Mrp1-mediated GSH export is also an important component in the GSH metabolism of these organs.

The residual GSH export of Mrp1(–/–) astrocytes and MK571-treated wild-type astrocytes suggests that, in addition to the MK571-sensitive Mrp1, one or more MK571-insensitive transporters are involved in GSH release from wild-type astrocytes. Apart from Mrp1, other Mrp family members as well as OATPs and CFTR have been demonstrated to be involved in GSH transport in various cell systems (Ballatori and Rebbeor 1998; Kogan et al. 2003; Ballatori et al. 2005). The presence of mRNAs of the Mrps 3, 4, 5, 7, and 9, of OATPs 1 and 2, as well as of CFTR, suggests that the respective proteins may also be present in cells of these cultures. Thus, these transporters could contribute to the residual Mrp1-independent export from mouse astrocytes. However, because the CFTR inhibitor glibenclamide or the OATP anti-porter substrate ouabain did not affect the residual GSH export from Mrp1(–/–) astrocytes, a contribution of these transporters to GSH export appears to be unlikely under the conditions used here.

Among the various Mrp family members, the Mrps 2, 3, 4 and 5 have been connected with GSH transport (for overview see Ballatori et al. 2005). As Mrp2 is not expressed in mouse brain and individual mouse astrocytes (Hirrlinger et al. 2005), a contribution of this transporter in GSH export can be excluded. In contrast, mRNAs of Mrps 3, 4 and 5 are present in mouse astrocyte cultures. Both Mrp3- and Mrp4-mediated transport processes are modulated by MK571 (Bodo et al. 2003; Rius et al. 2003). Because MK571 (50 μm) reduced the GSH export rate of wild-type astrocytes only to the export rate that was also determined for Mrp1(–/–) astrocytes, it is highly likely that Mrp1 exclusively mediates the MK571-sensitive part of the GSH export from astrocytes, and that Mrp3 and Mrp4 do not contribute to this process. The view that Mrp3 and Mrp4 do not participate in the residual GSH export from Mrp1(–/–) astrocytes is strongly supported by the observations that known inhibitors of Mrp4 did not lower the release of GSH from Mrp1-deficient astrocytes, and by reports showing that over-expression of Mrp3 did not increase GSH export or decrease cellular GSH levels (van der Linden et al. 1999; Zelcer et al. 2001).

Mrp5 has to be considered a candidate GSH transporter in astrocytes, as Mrp5-transfected cells secrete GSH and contain less intracellular GSH (Wijnholds et al. 2000b). Mrp5-mediated transport is not inhibited by MK571 (Jedlitschky et al. 2000). Therefore, this transporter could contribute to the observed MK571-insensitive GSH export. However, because Mrp5-deficient astrocytes did not differ from wild-type astrocytes regarding intracellular GSx content, the export rates for GSH and the effects of MK571 on the GSH export, a contribution from Mrp5 in the export of GSH from mouse astrocytes can be excluded. The high expression level of human Mrp5 in over-expressing cells, as well as differences in the functions of mouse and human Mrp5, could be the reasons that GSH export was observed for human Mrp5 in MDCKII cells (Wijnholds et al. 2000b), whereas physiological levels of Mrp5 in mouse astrocytes do not participate in GSH export.

At a concentration of 50 μm, MK571 strongly inhibited the efflux rate of GSH from cultured mouse (present report) and rat (Hirrlinger et al. 2002a) astrocytes. In contrast, in concentrations below 10 μm, MK571 stimulated GSH export. This bimodal concentration-dependent effect of MK571 on GSH release is in line with reports showing that GSH transport via Mrps is stimulated in the presence of other Mrp substrates at low concentrations (Evers et al. 2000; Loe et al. 2000; Hirrlinger et al. 2002a), whereas high concentrations of these substrates inhibit GSH release (Evers et al. 2000; Hirrlinger et al. 2002a). The absence of any MK571 effect on the GSH export rates of Mrp1(–/–) astrocytes demonstrates that Mrp1 is responsible for both the stimulation of GSH export in the presence of 1 μm MK571 and the inhibition of GSH export by 50 μm MK571.

Mrp1(–/–) mice show several phenotypes, strongly depending on the conditions tested (Wijnholds et al. 1997, 1998, 2000a). In the ear, Mrp1 transports leukotriene-C4 and thereby plays a role in inflammation. In the epithelia of choroid plexus, testis, the oropharynx, and kidney, Mrp1 protects cells by cellular extrusion of toxic compounds such as the anti-cancer drug etoposide. The reason for the lack of an obvious brain phenotype of Mrp1(–/–) mice may be the residual Mrp1-independent GSH export from astrocytes that could be sufficient to supply neurons with a GSH precursor under unstressed conditions. However, under pathological conditions, the lack of Mrp1 could become limiting for the rapid supply of GSH precursors to neurons and thereby would impair GSH synthesis and antioxidative defence in these cells. Thus, wild-type and Mrp1-deficient mice should be compared in models of neurological disorders to investigate the consequences of a reduced supply of GSH precursors from astrocytes to neurons under stress conditions.

GSSG is efficiently transported by Mrp1 with a KM value of 93 μm (Leier et al. 1996). Because GSSG is exported from rat astrocytes during oxidative stress and this transport is strongly affected by Mrp-inhibitors, a contribution of Mrp1 in this process was suggested (Hirrlinger et al. 2001, 2002c). As mouse astrocytes contain, at best, minute amounts of GSSG under unstressed conditions, the export of GSSG from astrocytes was studied during a chronic H2O2-induced oxidative stress. This approach demonstrated that wild-type mouse astrocytes released GSSG during oxidative stress, as do rat astrocytes (Hirrlinger et al. 2001), if a substantial amount of intracellular GSSG had been generated during H2O2 disposal in the glutathione peroxidase reaction. In contrast, extracellular GSSG accumulation following application of the peroxide generating system was not observed for Mrp1(–/–) astrocytes, despite an even higher intracellular ratio of GSSG to GSH in these cells compared with wild-type and Mrp5(–/–) astrocytes. These data demonstrate that the export of GSSG from mouse astrocytes is exclusively mediated by Mrp1 and that Mrp5 does not contribute to GSSG export observed for astrocytes during oxidative stress. The fate of the GSSG exported from astrocytes is not known so far. However, cultured rat astrocytes are able to remove extracellular GSSG from the culture medium in an enzyme-catalyzed process that does not involve γGT activity (C. Ruedig and R. Dringen, unpublished results).

The release of GSSG has been reported for several tissues and cell types and has been discussed as a protective mechanism to avoid cellular damage caused by an altered glutathione–redox equilibrium (Akerboom and Sies 1989; Keppler 1999). During oxidative stress, the export of GSSG may be especially important for astrocytes, as, among the different types of brain cells, astrocytes contain little glutathione reductase in vivo (Knollema et al. 1996) and in culture (Gutterer et al. 1999; Hirrlinger et al. 2002d). Therefore, rapid release of GSSG via Mrp1 might be an essential pathway for astrocytes to remove cellular GSSG quickly in order to maintain a reduced thiol reduction potential under oxidative stress. However, at least the over-expression of human Mrp1 in transfected cells did not improve the resistance of these cells toward oxidative stress (Balcerczyk et al. 2003). Application of the chronic oxidative stress model to wild-type and Mrp1-deficient mouse astrocytes, as described here, may be useful in further investigating the hypothesized protective effect of Mrp1-mediated GSSG export from cells.

In conclusion, the data presented here demonstrate that Mrp1 mediates 60% of the release of GSH and is exclusively responsible for the export of GSSG from mouse astrocytes during oxidative stress. In the brain, astrocytes play an essential role in the defence against oxidative stress and in the supply of GSH precursors to neighbouring neurons that are initiated by GSH export from astrocytes (Dringen 2000; Dringen and Hirrlinger 2003). Because astrocytes from Mrp1-deficient mice have a strongly reduced export rate for GSH and do not export GSSG, Mrp1(–/–) mice could be a good model to study the importance of Mrp1 in the GSH metabolism in brain, and the contribution of this transporter in the defence of this organ against oxidative stress.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

We would like to thank Professor Piet Borst (Amsterdam) for critically reading the manuscript and for valuable comments. RD would like to thank Neurosciences Victoria for a senior research fellowship. This project was financially supported by the Deutsche Forschungsgemeinschaft (grants Schu932/3-2 and DR262/7-1) as well as by the Dutch Cancer Society (grant NKI 2001-2473).

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References