- Email: [email protected]
Overexpression of γ-glutamylcysteine synthetase in human malignant mesothelioma
Overexpression of ␥-Glutamylcysteine Synthetase in Human Malignant Mesothelioma ¨ RVINEN, MD, PHD, YLERMI SOINI, MD, PHD, KRISTIINA JA KATRIINA KAHLOS, MD, PHD, AND VUOKKO L. KINNULA, MD, PHD Mesothelioma is a fatal tumor resistant to all treatment modalities for reasons that are still unresolved. Glutathione (GSH)-associated pathways are induced by oxidants and cytotoxic drugs, and they are also involved in the progression and resistance of some tumor cells in vitro. The rate-limiting enzyme in GSH biosynthesis is ␥-glutamylcysteine synthetase (␥GCS). However, the expression of this enzyme has not been systematically investigated in malignant tumors, and there are no studies of ␥GCS in biopsy specimens of malignant mesothelioma. We investigated the immunohistochemical distribution and expression of both subunits of ␥GCS in healthy pleural mesothelium, pleural mesothelioma tumor biopsy samples (34 cases), and mesothelioma cells in culture (7 cell lines). Nonmalignant mesothelium showed no immunoreactivity for either subunit in any of the cases. The heavy (catalytic) subunit of ␥GCS was highly immunostained in 29 and weakly positive in 5 cases. High-moderate and weak immunoreactivity of the light (regulatory) subunit of ␥GCS was found in 15 and 7 tumors, respectively, whereas 12 cases showed no reactivity. There was no correlation with either catalytic or regulatory subunit expression and patient survival. There was, however, a sig-
nificant correlation between the heavy chain and multidrug resistance protein (MRP) 2 (P ⴝ .048), whereas no correlation was observed between the light chain and MRP1 or MRP2. Treatment of cultured mesothelioma cells with buthionine sulfoximine (BSO), to inhibit ␥GCS, significantly potentiated cisplatin-induced cytotoxicity mainly by nonapoptotic mechanism when assessed by counting the living cells, TUNEL (terminal deoxytransferase-mediated dUTP nick-end labeling) assay, and caspase-3 cleavage. In conclusion, ␥GCS is highly positive in most cases of malignant mesothelioma and may play an important role in the primary drug resistance of this tumor in vivo. HUM PATHOL 33:748-755. Copyright 2002, Elsevier Science (USA). All rights reserved. Key words: antioxidant, buthionine sulfoximine, cisplatin, glutathione, multidrug resistance protein. Abbreviations: MnSOD, manganese superoxide dismutase; GSH, glutathione; ␥GCS, ␥-glutamylcysteine synthetase; ␥GCSh, heavy chain of ␥GCS; ␥GCSl, light chain of ␥GCS; GST, glutathione Stransferase; MRP, multidrug resistance protein; HPFs, high power fields; BSO, buthionine sulfoximine.
Mesothelioma is a fatal tumor caused in 80% of the cases by occupational exposure to asbestos fibers.1 It is a rare disease, but because of its long latency period its incidence is still increasing despite the industrial restriction of asbestos use in the West. Asbestos fibers are known to generate reactive free radicals and induce antioxidant enzymes,2 which may play role both in the pathogenesis and high drug resistance of this disease. Although mesothelioma is a rare tumor, it provides an ideal model of a drug- and oxidant-resistant malignancy for the investigation of antioxidant mechanisms in cancer. In our previous studies, we found that manganese superoxide dismutase (MnSOD), one of the most important antioxidant enzymes, is consistently upregulated in mesothelioma tissue and cultured mesothelioma cells.3,4 We also found that MnSOD alone does not explain oxidant or drug resistance of those cells.5 Given the importance of hydrogen peroxide scavenging
mechanisms and glutathione (GSH)-related pathways in the oxidant and drug resistance of malignant cells,6 we conducted additional studies on these mechanisms in mesothelioma.4,7 Those results suggest that catalase does not have any major significance in the drug resistance of this disease, whereas GSH depletion enhances sensitivity of mesothelioma cells, at least to epirubicin.7,8 The rate-limiting enzyme in GSH synthesis is ␥-glutamylcysteine synthetase (␥GCS).9 This enzyme may play significant role in the primary resistance of mesothelioma because a recent microarray study analyzed the mRNA expression of more than 6,500 genes in cultured mesothelioma cells and found that ␥GCS mRNA was expressed in the malignant cells when compared with nonmalignant cells.10 ␥GCS, which is feedback inhibited by GSH, consists of two subunits: a heavy subunit with catalytic function (␥GCSh) and a light subunit with a regulatory role (␥GCSl).11 The heavy subunit has been studied more because it is considered to be functionally more important than the light subunit. Many studies, however, show that both subunits are induced by oxidants and apparently also by oxidant-generating anticancer drugs and that, at least in some cell lines, the induction correlates with drug resistance.12-18 Other than ␥GCS, additional mechanisms are closely associated with GSH. These pathways include GSH peroxidase in the decomposition of hydrogen peroxide, GSH reductase in converting oxidized GSH to the reduced form, and ␥-glutamyl transpeptidase in the GSH salvage pathway. GSH S-transferases (GSTs)—in
From the Department of Internal Medicine and Department of Pathology, University of Oulu; and the Finnish Institute of Occupational Health, Helsinki, Finland. Accepted for publication May 1, 2002. Supported by the Finnish Antituberculosis Association Foundation, The Cancer Society of Finland, The Juselius Foundation, The Ida Montin Foundation, and The Ahokas Foundation. Address correspondence and reprint requests to Vuokko L. Kinnula, MD, PhD, University of Oulu, Department of Internal Medicine, Box 50, 90014 University of Oulu, Helsinki, Finland. Copyright 2002, Elsevier Science (USA). All rights reserved. 0046-8177/02/3307-0011$35.00/0 doi:10.1053/hupa.2002.126191
748
␥-GLUTAMYLCYSTEINE SYNTHETASE IN MESOTHELIOMA (Ja ¨ rvinen et al)
detoxification of electrophiles—and multidrug resistance proteins, or MRPs (especially MRP1 and MRP2)—in enhancing drug transport from the cells— are probably among the most important pathways in the drug resistance of many malignant tumors.19-21 Our recent study found that the most resistant mesothelioma cell line contained the highest GST activity and increased level of GST.7 Many cell line studies have also suggested an association between the expression of MRP1 and ␥GCS22-24 in resistant cells, but unlike MRP1, MRP2 has been shown to confer resistance against cisplatin.25 However, no studies exist regarding the correlation between ␥GCS and MRP proteins in tumor biopsy specimens. We analyzed the expression and location of the heavy and light subunits of ␥GCS in healthy human pleura and in mesothelioma by immunohistochemistry and assessed the association between ␥GCS immunostaining in mesothelioma tissue and tumor proliferation, apoptosis, patient survival, or MRP1 and MRP2 expression. We also investigated the expression of ␥GCS in 7 mesothelioma cell lines, which had been established from untreated mesothelioma patients.26 In two of these cell lines, with a high or intermediate level of ␥GCS expression, the effects of ␥GCS inhibition were investigated by assessing irreversible cell injury and apoptosis. MATERIALS AND METHODS Patients and Tumor Tissues Mesothelioma tissue samples from 34 patients (28 men and 6 women), as well as from healthy human pleura, were retrieved from the files of the Department of Pathology at Oulu University Hospital. The healthy control tissues were obtained from nonsmokers investigated for nonmalignant lung diseases such as sarcoidosis; in those cases the pleura was nonneoplastic and noninflamed, and the biopsy specimen showed normal histopathology. Four of the mesothelioma patients had received anticancer drugs before undergoing surgery; 1 patient had been treated with combination therapy that included radiation therapy. Malignant mesotheliomas were subclassified according to the criteria from the World Health Organization.27 Three subtypes of mesothelioma were included: epitheloid (n ⫽ 23), sarcomatoid (n ⫽ 7), and biphasic (n ⫽ 4). For survival evaluation, patient records were used to obtain clinical data from the follow-up period.
Cell Cultures and Exposures Seven mesothelioma cell lines (M10K, M14K, M24K, M25K, M28K, M33K, and M38K) were established from the tumor tissue of untreated mesothelioma patients, with the exception of M10K cell line, which had been established from mesothelioma metastasis. MeT5A is a nonmalignant, SV-40 transformed cell line (American Type Culture Collection, Rockville, MD). The cells were cultured in Roswell Park Memorial Institute (RPMI) medium with 10% phosphate buffer solution (PBS), 100 U/mL penicillin, 100 g/mL streptomycin, and 0.03% L-glutamine (all from Life Technologies, Paisley, Scotland) at 37 C° in a 5% carbon dioxide atmosphere. Previous studies have shown that M38K cells are the most resistant cell line and contain markedly higher intracel-
lular GSH concentrations than M14K cells, the latter being also more sensitive against exogenous epirubicin;7 therefore, these two cell lines were selected for further exposures. In selected experiments, they were pretreated with buthionine sulfoximine (BSO) to cause GSH depletion. According to our previous studies, 0.2 mM BSO was selected for M14K cells, and 1 mM BSO was selected for M38K cells.7 However, because of the possible toxicity of BSO alone, incubations also were conducted in the presence of 0.1 mM (M14K cells) and 0.5 mM (M38K cells) BSO. Cisplatin (Platinol; Bristol-Myers Squibb, New York, NY) was chosen because it is widely used in the treatment of mesothelioma and other lung tumors, and increased GSH synthesis may be associated with the resistance of cancer cells to classical alkylating agents, such as cisplatin.28,29 The cells were pretreated with BSO for 16 hours, after which they were exposed to cisplatin (10 M–100 M) for 48 hours. To ensure that GSH depletion occured during the entire incubation period, the same concentration of BSO was added to the incubation medium 24 hours after starting the exposure to cisplatin.
Immunohistochemistry of ␥GCS The biopsy specimens were first fixed in 10% neutral formalin, dehydrated, and embedded in paraffin. The rabbit antihuman antibodies for ␥GCSh and ␥GCSl were kindly provided by Dr. Terrance Kavanagh (University of Washington, Seattle). These antibodies were been previously tested in human lung tissues and cultured cells.5,30 Four-micron–thick sections were cut from a representative paraffin block; the sections were deparaffinized in xylene and rehydrated in descending ethanol series. Endogenous peroxidase was consumed by incubating the sections in 0.1% hydrogen peroxide in absolute methanol for 10 minutes. To enhance immunoreactivity, the sections were incubated in 10 mmol citrate buffer (pH 6.0) and boiled in a microwave oven for 2 minutes at 850 W and 8 minutes in 350 W. The paraffin blocks were then incubated with 2% milk powder to diminish background staining. The sections were incubated with the primary antibodies for ␥GCSh and ␥GCSl with dilutions of 1:1000. Immunostaining was performed using the Histostain-Plus Kit (Zymed, South San Francisco, CA) and the chromogen was aminoethyl carbazole (Zymed) In negative controls, the primary antibodies were substituted with FBS or nonimmune rabbit serum. The quantity of the immunostaining was evaluated by assessing the staining as follows: 0 ⫽ no positive immunostaining; 1 ⫽ ⬍ 33% of tumor cells showing positive staining; 2 ⫽ 33% to 66% of tumor cells showing positive staining; and 3 ⫽ 67% to 100% of tumor cells showing positive staining. Also, the intensity of the immunostaining was evaluated by division into groups as follows: 0 ⫽ no positive immunostaining; 1 ⫽ weak cytoplasmic staining; 2 ⫽ moderate cytoplasmic staining; and 3 ⫽ strong cytoplasmic staining. A combined score for the immunostaining, based on both qualitative and quantitative immunostaining, was composed by adding both the qualitative and quantitative score and then divided in four main groups as follows: ⫺ ⫽ no immunostaining/score 0; ⫹ ⫽ weak immunostaining/scores 1 to 2; ⫹⫹ ⫽ moderate immunostaining/scores 3 to 4; and ⫹⫹⫹ ⫽ strong immunostaining/scores 5– 6. Two authors (KJ, YS) evaluated the immunostaining separately, and the correlation coefficient was established according to Cohen’s kappa statistics as described.31 The correlation was good (heavy, 0.6; light, 0.8).
749
HUMAN PATHOLOGY
Volume 33, No. 7 (July 2002)
gland) (dilution 1:50 000). ␥GCS and caspase 3 were detected by an enhanced chemiluminescense system (Amersham), and the luminol excitation was imaged on x-ray film.
3⬘-End Labelling of DNA in Apoptotic Cells and Assessment of Cell Proliferation in Tumor Biopsy Specimens To detect apoptotic cells, in situ labelling of the 3⬘-ends of the DNA fragments generated by apoptosis-associated endonucleases was performed using the ApopTag in situ apoptosis detection kit (Oncor, Gaithesburg, MD) as previously described.32 Cells were defined as apoptotic if the whole nuclear area of the cell labelled positively. Apoptotic bodies were defined as small, positively labelled, globular bodies in the cytoplasm of the tumor cells that could be found either singly or in groups. To estimate the apoptotic index (the percentage of apoptotic events in a given area), apoptotic cells and bodies were counted in 10 high-power fields (HPFs), and this figure was divided by the number of tumor cells in the same HPFs. Tumor cell proliferation was studied with a monoclonal mouse antihuman Ki-67 antibody (Zymed, San Fransisco, CA). Sections 5m thick were cut from representative tumor blocks. The sections were first deparaffinized in xylene and rehydrated in descending ethanol series. To enhance immunoreactivity, the sections were incubated in 10 mmol citrate buffer (pH 6.0) and boiled for 2 minutes at 850 W and then 8 minutes at 350 W. Endogenous peroxidase activity was eliminated by incubation in 0.1% hydrogen peroxide in absolute methanol for 10 minutes. The concentration of primary antibody for anti-Ki-67 was 1:50. In the immunostaining, the avidin-biotin-peroxidase complex method was used, and the color was developed using 3,3⬘-diaminobenzidine. The sections were lightly counterstained with hematoxylin and mounted with Eukitt (Kindler, Freiburg, Germany). Replacement of the primary antibodies with PBS and serum isotype, recommended by the manufacturers at pH 7.2, were used as negative controls. The results for Ki-67 immunostaining were evaluated by counting the percentage of positive nuclei of the whole tumor cell population. For Ki-67, the staining was evaluated semiquantitatively and divided into the following 4 groups: weak staining ⫽ fewer than 5% of cell nuclei positive; moderate staining ⫽ 5% to 10% of cell nuclei positive; strong staining ⫽ 10% to 50% of cell nuclei positive; and very strong staining ⫽ greater than 50% of cell nuclei positive.
Western Blot Analysis The cells were detached with trypsin, centrifuged and washed with PBS, and then mixed with the electrophoresis sample buffer and boiled for 5 minutes at 95 C°. Cell protein was measured by using the Bio-Rad (Hercules, CA) method, and 70 g cell protein was applied per lane in a 12% sodium dodecyl sulphate-polyacrylamide gel. The gel was electrophoresed for 1.5 h (90V) at room temperature, and the protein was transferred (60 minutes, 100V) onto Hybond ECL nitrocellulose membranes (Amersham, Buckinghamshire, England) in a Mini-PROTEAN II Cell (Bio-Rad Mississagua, Ontario, Canada). The blotted membrane was incubated with rabbit antibodies to ␥GCSh and ␥GCSl (dilution 1:40,000 and 1:20,000, respectively) or to caspase 3 (dilution 1:2000) followed by a donkey antirabbit antibody conjugated to horseradish peroxidase (Amersham Biosciences, Amersham, En-
Measurement of Cytotoxicity Depletion of high-energy nucleotides was used because of the sensitivity of this method in the assessment of cell injury. The cells were preincubated for 16 hours with 0.1 mmol [14C] adenine (specific activity, 51 to 55 mCi mmol⫺1; Amersham) to preload the high-energy nucleotides in intact cells. Prelabelled cells were washed three times and then exposed to cisplatin. After 48 hours, the medium was collected, and the cells were extracted with 0.4 mol/L perchloric acid. Thin-layer chromatography was used to separate the purine nucleotides (adenosine triphosphate, adenosine diphosphate, and adenosine monophosphate) in the cell extract as well as the medium and the nucleotide catabolic products (xanthine, hypoxanthine, and uric acid) in the medium.33 The results are expressed as percent distribution of radioactivity (counts per minute [cpm]) between nucleotides in the cells, nucleotides leaked to the medium, and catabolic products in the medium. In a healthy cell, the plasma membrane is not permeable to high-energy nucleotides, but catabolic products are extracted from the cell. Therefore, the appearance of nucleotides in the extracellular medium is a marker for cell membrane injury and cytotoxicity. Previous and preliminary studies indicate that nucleotide depletion is more sensitive than the markers of irreversible cell injury and that these results correlate with each other.34 Therefore, in additional cell injury experiments, the cells were washed with PBS, trypsinized, and counted under a light microscope. Because cell counting does not reveal anything about the mechanism of cell survival, apoptosis was also evaluated from cultured cells treated with BSO and/or exposed to cisplatin. In addition, caspase 3 cleavage was used as another marker of apoptosis as described above (Western blot analysis).
Assessment of Apoptosis in Cultured Cells For the apoptosis assays, the cells were cultured, pretreated with BSO for 16 hours, and then exposed to cisplatin for 48 hours. The pellets were fixed in 10% neutral formalin overnight, after which the formalin was removed, and melted 2% agar was laid over the pellets. The agar blocks were further embedded in paraffin. Sections 5 m thick were cut from the cell blocks and processed for detection of apoptosis as described above. Apoptosis was determined morphologically by in situ labelling of 3⬘-ends of the DNA fragments (TUNEL). Apoptotic cells with clearly labelled nuclei as well as smaller apoptotic bodies were calculated. The percentages of positive cells were expressed as apoptotic indices.
Statistical Analysis Data are expressed as mean ⫾ standard error. SPSS (7.5) for Windows was used in patient correlation data; the differences between the two groups were analyzed by paired 2-tailed t test. The significance of correlations was determined using Fisher exact probability test. The survival of the patients
™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™3 FIGURE 1. Expression of ␥GCS subunits in healthy pleura and in biopsy specimens of mesothelioma. (A) healthy pleura and ␥GCSh. (B) Healthy pleura and ␥GCSl. (C) Epithelial mesothelioma and ␥GCSh. (D) Epithelial mesothelioma and ␥GCSl. (E) Sarcomatoid mesothelioma and ␥GCSl. (F) Sarcomatoid mesothelioma and ␥GCSl. (G and H) Serum controls. Immunostaining was done using the Histostain-Plus Kit (Zymed), and the chromogen was aminoethyl carbazole (Zymed) Original magnification (A) ⫻ 280 (B–H) ⫻ 180.
750
751
HUMAN PATHOLOGY
Volume 33, No. 7 (July 2002)
TABLE 1. A Combined Score (Qualitative and Quantitative) of the Immunoreactivity of ␥GCSh and ␥GCSl in Malignant Mesothelioma Tissue Samples Patient No.
Histological Type
Age (y)/ Gender
␥GCSh
␥GCSl
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
Sarcomatoid Epithelial Biphasic Sarcomatoid Epithelial Epithelial Epithelial Biphasic Epithelial Biphasic Epithelial Sarcomatoid Sarcomatoid Epithelial Epithelial Epithelial Epithelial Sarcomatoid Epithelial Biphasic Epithelial Sarcomatoid Epithelial Epithelial Epithelial Epithelial Epithelial Epithelial Sarcomatoid Epithelial Epithelial Epithelial Epithelial Epithelial
42/F 56/M 57/F 58/M 74/M 66/M 70/F 70/M 50/M 68/M 63/M 52/M 33/F 73/M 67/M 59/M 54/M 67/M 78/M 57/M 57/M 73/M 78/M 79/M 63/M 66/M 72/M 73/F 68/M 69/M 69/M 77/M 52/F 67/F
⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹ ⫹
⫺ ⫹⫹ ⫹⫹ ⫺ ⫹⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹ ⫹ ⫹⫹ ⫺ ⫹⫹ ⫺ ⫺ ⫺ ⫹⫹⫹ ⫺ ⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹
showing no reactivity. The histological type of mesothelioma had no statistically significant effect on ␥GCS reactivity although the epithelial subtypes tended to stain somewhat more strongly (Fig 1). There was no correlation with either ␥GCSh or ␥GCSl immunoreactivity and patient survival (P ⫽ .07 and .42, respectively). Staining with Ki-67 showed no association with ␥GCS expression and tumor cell proliferation (P ⫽ .42 for ␥GCSh and P ⫽ .25 for ␥GCSl). No correlation was seen between apoptosis and ␥GCS (␥GCSl, P ⫽ .79; ␥GCSh, P ⫽ .55). We recently investigated the expression of MRP1 and MRP2 using this same biopsy material of mesothelioma patients.35 For the present study, we assessed if there were any correlations between ␥GCS reactivities and MRP1 or MRP2. There was a significant correlation between ␥GCSh and MRP2 expression (P ⫽ .048), whereas the corresponding correlation between ␥GCSl and MRP2 was not significant (P ⫽ .35). Neither was any correlation seen between MRP1 and ␥GCS (␥GCSh, P ⫽ .67; ␥GCSl, P ⫽ 0.50). Expression of ␥GCS in Mesothelioma Cell Line Cells In Vitro Western blot studies with 7 mesothelioma cell lines indicated that both ␥GCS subunits were detectable in all mesothelioma cell lines investigated. M38K cells, which represent the most resistant cell line,4,5,7 contained the highest level of ␥GCSh (Fig 2). Two cell lines, M14K and M38K, were selected for further studies on cytoxicity. Exposure to Cisplatin With And Without BSO
Abbreviations: ␥GCSh, heavy chain of ␥GCS; ␥GCSl, light chain of ␥GCS; F, female; M, male; ⫺, no immunostaining, ⫹, weak immunostaining; ⫹⫹, moderate immunostaining; ⫹⫹⫹, strong immunostaining.
The M38K cell line was more resistant than the M14K cell line to cisplatin (Fig 3). The concentrations of cisplatin for the next experiments were chosen according to results from this assay. Pretreatment with BSO to deplete intracellular GSH potentiated the cisplatin-induced toxicity significantly (Fig 4). Caspase 3
in relation to the immunoreactivities was assessed by log-rank test, with P ⬍ .05 being statistically significant.
RESULTS Immunohistochemical Expression of ␥GCS in Mesothelioma Biopsy Specimens Nonmalignant mesothelium showed no reactivity of ␥GCSh or ␥GCSl in any of the cases (Fig 1A and B). Most mesotheliomas (29 of 34) were strongly positive for ␥GCSh (Fig 1C and E). The qualitative and quantitative immunostaining scores were combined as described in the Materials and Methods section, and the summary of those results is shown in Table 1. The staining of ␥GCSh was most positive in malignant cells, and the color appeared in the cytosol. Additional positivity could be observed in the stroma of the tumor as well as inflammatory cells. The ␥GCSl immunoreactivity was considerably less positive (Fig 1D and F), with 15 cases showing strong/ moderate reactivity, 7 showing weak reactivity, and 12
FIGURE 2. Western blot analyses showing expression of (A) ␥GCSh (molecular weight ⬃ 73,000) and (B) ␥GCSl (molecular weight ⬃30,000) in nonmalignant transformed MeT5A mesothelial cells and in 7 malignant mesothelioma cell lines.
752
␥-GLUTAMYLCYSTEINE SYNTHETASE IN MESOTHELIOMA (Ja ¨ rvinen et al)
enhanced by BSO (data not shown). The apoptotic index in M38K cells was always very low (between 2.4% in controls and 4.5% in cisplatin-exposed, BSO-treated cells). DISCUSSION In this study, we found that most malignant mesotheliomas express both subunits of ␥GCS and that healthy mesothelium is negative for ␥GCS immunoreactivity. There was a significant correlation between ␥GCSh and MRP2 reactivity, but neither subunit of ␥GCS showed correlation with tumor proliferation or patient survival. As hypothesized, the mesothelioma cell
FIGURE 3. Depletion of cellular high-energy nucleotides (CNucl), accumulation of nucleotide catabolic products (M-HxUa-X), and leakage of intact nucleotides (M-Nucl) in (A) M14K or (B) M38K cells exposed to varying concentrations of cisplatin. Values are mean ⫾ standard error from three separate experiments. P ⬍ .05 vs unexposed (cont).
cleavage has been used as an indicator of the apoptotic process in various cell types. Preliminary studies showed no cleavage of caspase 3 with BSO alone, but cleavage was induced by cisplatin (data not shown). Additional studies using M14K cells showed caspase 3 cleavage within 48 hours using 50 M cisplatin, which was not potentiated by BSO pretreatment (Fig 5A). M38K cells, the most resistant cell line with the highest ␥GCS content, also were more resistant to apoptosis than were M14K cells. In this cell line, no apoptosis could be seen when assessed by caspase 3 cleavage after cisplatin exposure without or with BSO (Fig 5B). Given that caspase 3 cleavage is not a sensitive indicator of apoptosis, apoptosis also was assessed morphologically after the same exposures. Again, cisplatin caused apoptosis in M14K cells but not in M38K cells; apoptosis was not
FIGURE 4. Viability of M14K cells and M38K cells exposed to the indicated concentrations of cisplatin with and without BSO. M14K cells (A) and M38K cells (B) were treated with 0.1 mM (M14K) or 0.5 mM (M38K) BSO and exposed to cisplatin for 48 hours. Cell cytotoxicity was assessed by calculating living cells after the exposures. Values are mean ⫾ standard error from three separate experiments. P ⬍ .05 not pretreated vs BSO pretreated.
753
HUMAN PATHOLOGY
Volume 33, No. 7 (July 2002)
FIGURE 5. Western blot analyses of caspase 3 cleavage in (A) M14K cells and (B) M38K cells exposed to varying concentrations of cisplatin with and without BSO. Caspase 3 cleavage can be observed in M14K cells exposed to 50 M cisplatin, whereas M38K cells show no apoptosis by 50 or 100 M concentration, the latter concentration of which is toxic to M38K cells. BSO had no effect on caspase 3 cleavage when compared with the effect of cisplatin in either cell type.
line with high GSH content and ␥GCSh expression turned out to be resistant toward apoptosis and cisplatin-induced toxicity. Numerous studies have shown that drug-resistant cancer cell lines contain high GSH concentrations and that cell viability can be modulated by GSH depletion.16,21,36 One study on lung cancer showed that approximately 70% of lung tumors express both subunits of ␥GCS.30 Recent microarray study also showed increased expression of ␥GCS mRNA in one mesothelioma cell line compared with nonmalignant MeT5Atransformed mesotelial cells,10 which is why it is even more important to confirm this finding in the tumor biopsy samples of malignant mesothelioma. One study suggested overexpressions of ␥GCSh mRNA and MRP1 in doxorubicin-resistant mesothelioma cells18 and another the deletion of the ␥GCSl gene in mesothelioma.37 In agreement with the microarray study,10 we found that most mesotheliomas were highly positive for ␥GCSh reactivity. The deletion of the ␥GCS light subunit gene may not be a universal feature in mesotheli-
oma because more than half of our biopsy specimens were positive for ␥GCSl reactivity. In our relatively large group of biopsy specimens, no association was found between either subunit of ␥GCS and MRP1 expression. Instead, significant correlation between ␥GCSh and MRP2 expression may play role in the primary resistance of this disease. We found no significant correlation between the reactivity of ␥GCS and patient survival. This does not rule out the possibility of an association because the high expression levels may be related to the high primary resistance of this disease. Association between the light and the heavy chains was also nonsignificant, which could indicate that some of the ␥GCSh gene product may be protein with no enzyme activity or that the light subunit is not needed for full enzyme activity. In fact, it has been stated that the heavy subunit alone comprises approximately 46% of the activity of the holoenzyme.36 GSH level does not necessarily correlate with ␥GCS activity either because the major controlling factor of intracellular GSH, at least in some cell types, is its transport out of the cell via the plasma membrane pumps, most importantly the proteins of the MRP family.36 In our recent study, 58% of these mesothelioma patients were positive for MRP1 and 33% for MRP2. The result of ␥GCSh and MRP2 correlation is in line with the recent finding that MRP2, but not MRP1, may be associated with primary cisplatin resistance.25 Given that tissue biopsy samples of mesothelioma contain tissue stroma, multiple cell types, and nonmalignant cells, the assessment of active ␥GCS protein and/or GSH content from tumor tissue homogenates is not reliable. Therefore, we analyzed both the expression of ␥GCS and the GSH content in cultured mesothelioma cells and assessed the importance of GSH depletion on the oxidant and drug resistance in these cells. M38K cells, which are the most resistant mesothelioma cells in our laboratory, contained not only the highest ␥GCSh reactivity but also had the highest GSH levels in vitro.5,7 The results with cisplatin-exposed cells also were consistent with previous investigations of other cell lines in which ␥GCS expression and GSH were investigated.15,16,38,39 Our recent5 and present results confirm the importance of ␥GCS and GSH in the resistance of mesothelioma cells to oxidants, to epirubicin,5 and to cisplatin in vitro. They also suggest that GSH protects these cells against nonapoptotic cell injury. In conclusion, our study suggests that the synthesis of GSH by ␥GCS may be of major importance in the primary drug resistance of malignant cells. Acknowledgment. The authors thank Dr. Kaija Linnainmaa for providing the mesothelioma cell lines; Dr Terrance Kavanagh for the ␥GCS antibodies and Dr Paavo Pa¨a¨kko¨ for the excellent research facilities in Oulu. The technical assistance of Mrs Raija Sirvio¨ and Mr Manu Tuovinen is gratefully acknowledged. The authors also thank Proffessor Kari Raivio for his critical comments on the manuscript.
754
␥-GLUTAMYLCYSTEINE SYNTHETASE IN MESOTHELIOMA (Ja ¨ rvinen et al)
REFERENCES 1. Mossman BT, Bignon J, Corn M, et al: Asbestos: Scientific developments and implications for public policy. Science 247:294301, 1990 2. Mossman BT, Hansen K, Marsh, JP et al: Mechanisms of fibre-induced superoxide release from alveolar macrophages and induction of superoxide dismutase in the lungs of rats inhaling crocidolite. IARC Sci Publ 90:81-92, 1989 3. Kinnula VL, Pietarinen-Runtti P, Raivio K, et al: Manganese superoxide dismutase in human pleural mesothelioma cell lines. Free Radic Biol Med 21:527-532, 1996 4. Kahlos K, Anttila S, Asikainen T, et al: Manganese superoxide dismutase in healthy human pleural mesothelium and in malignant pleural mesothelioma. Am J Respir Cell Mol Biol 18:570-580, 1998 5. Ja¨rvinen K, Pietarinen-Runtti P, Linnainmaa K, et al: Antioxidant defense mechanisms of human mesothelioma and lung adenocarcinoma cells. Am J Physiol Lung Cell Mol Physiol 278:L696-L702, 2000 6. Tew KD: Glutathione-associated enzymes in anticancer drug resistance. Cancer Res 54:4313-4320, 1994 7. Kinnula K, Linnainmaa K, Raivio KO, et al: Endogenous antioxidant enzymes and glutathione S-transferase in protection of mesothelioma cells against hydrogen peroxide and epirubicin toxicity. Br J Cancer 77:1097-1102, 1998 8. Kahlos K, Soini Y, Sormunen R, et al: Expression and prognostic significance of catalase in malignant mesothelioma. Cancer 91:1349-1357, 2001 9. Meister A, Anderson ME: Glutathione. Annu Rev Biochem 52:711-60, 1983 10. Rihn BH, Mohr S, McDowell SA, et al: Differential gene expression in mesothelioma. FEBS Lett 480:95-100, 2000 11. Seelig, GF, Simondsen RP, Meister A: Reversible dissociation of ␥GCS into 2 subunits. J Biol Chem 259:9345-9347, 1984 12. Galloway DC, Blake DG, Shepherd AG, et al: Regulation of human gamma-glutamylcysteine synthetase: Co-ordinate induction of the catalytic and regulatory subunits in HepG2 cells. Biochem J 328:99-104, 1997 13. Shi MM, Kugelman A, Iwamoto T, et al: Quinone-induced oxidative stress elevates glutathione and induces gamma-glutamylcysteine synthetase activity in rat lung epithelial L2 cells. J Biol Chem 269:26512-26517, 1994 14. Rahman I, MacNee W: Oxidative stress and regulation of glutathione in lung inflammation. Eur Respir J 16:534-554, 2000 15. Iida T, Mori E, Mori K, et al: Co-expression of gammaglutamylcysteine synthetase sub-units in response to cisplatin and doxorubicin in human cancer cells. Int J Cancer 82:405-411, 1999 16. Zhang K, Mack P, Wong KP: Glutathione-related mechanisms in cellular resistance to anticancer drugs. Int J Oncol 12:871882, 1998 17. Tipnis SR, Blake DG, Shepherd AG, et al: Overexpression of the regulatory subumit of g-glutamylcysteine synthetase in HeLa cells increases g-glutamylcysteine synthetase activity and confers drug resistance. Biochem J 337:559-566, 1999 18. Ogretmen B, Bahadori HR, McCauley MD, et al: Co-ordinated over-expression of the MRP and ␥GCS genes, but not MDR1, correlates with doxorubicin resistance in human malignant mesothelioma cell lines. Int J Cancer 75:757-761, 1998 19. Cole SP, Bhardwaj G, Gerlach JH, et al: Overexpression of a transporter gene in a multidrug-resistant human lung cancer cell line. Science 258:1650-1654, 1992 20. Konig J, Nies AT, Cui Y, et al: Conjugate export pumps of the multidrug resistance protein (MRP) family: Localization, substrate specificity, and MRP2-mediated drug resistance. Biochim Biophys Acta 1461:377-394, 1999
21. O’Brien ML, Tew KD: Glutathione and related enzymes in multidrug resistance. Eur J Cancer 32:967-978, 1996 22. Ishikawa T, Bao JJ, Yamane Y, et al: Coordinated induction of MRP/GS-X pump and gamma-glutamylcysteine synthetase by heavy metals in human leukemia cells. J Biol Chem 271:14981-14988, 1996 23. Kuo MT, Bao JJ, Curley SA, et al: Frequent coordinated overexpression of the MRP/GS-X pump and gamma-glutamylcysteine synthetase genes in human colorectal cancers. Cancer Res 56:36423644, 1996 24. Kigawa J, Minagawa Y, Cheng X, et al: Gamma-glutamyl cysteine synthetase up-regulates glutathione and multidrug resistance-associated protein in patients with chemoresistant epithelial ovarian cancer. Clin Cancer Res 4:1737-1741, 1998 25. Borst P, Evers R, Kool M, et al: A family of drug transporters: The multidrug resistance-associated proteins. J Natl Cancer Inst 92: 1295-1302, 2000 26. Pelin-Enlund K, Husgafvel-Pursiainen K, Tammilehto L, et al: Asbestos-related malignant mesothelioma: Growth, cytology, tumorigenicity and consistent chromosome findings in cell lines from five patients. Carcinogenesis 11:673-681, 1990 27. Travis WD, Colby TV, Corrin B, et al: Histological typing of lung and pleural tumors, in: World Health Organization International Histological Classification of Tumours (ed 3). Berlin, Springer, 1999, pp 51-54 28. Hao XY, Bergh J, Brodin O, et al: Acquired resistance to cisplatin and doxorubicin in a small cell lung cancer cell line is correlated to elevated expression of glutathione-linked detoxification enzymes. Carcinogenesis 15:1167-1673, 1994 29. Sugarbaker DJ, Norberto JJ: Multimodality management of malignant pleural mesothelioma. Chest 113:61S-65S, 1998 (suppl 1) 30. Soini Y, Na¨pa¨nkangas U, Ja¨rvinen K, et al: Expression of gamma-glutamyl cysteine synthetase in nonsmall cell lung carcinoma. Cancer 92:2911-2919, 2001 31. Silcocks PB: Measuring repeatability and validity of histological diagnosis—A brief review with some practical examples. J Clin Pathol 36:1269-1275, 1983 32. Soini Y, Virkaja¨rvi N, Lehto V-P, et al: Hepatocellular carcinomas with a high proliferation index and a low degree of apoptosis and necrosis are associated with a shortened survival. Br J Cancer 73:1025-1030, 1996 33. Aalto TK, Raivio KO: Adenine nucleotide depletion from endothelial cells exposed to xanthine oxidase. Am J Physiol 259:C883C888, 1990 34. Spragg RG, Hinshaw DB, Hyslop PA, et al: Alterations in adenosine triphosphate and energy charge in cultured endothelial and P388D1 cells after oxidant injury. J Clin Invest 76:1471-1476, 1985 35. Soini Y, Ja¨rvinen K, Kaarteenaho-Wiik R, et al: The expression of P-glycoprotein and multidrug-resistance proteins 1 and 2 (MRP1 and MRP2) in human malignant mesothelioma. Ann Oncol 12:1239-1245, 2001 36. Griffith OW: Biologic and pharmacologic regulation of mammalian glutathione synthesis. Free Radic Biol Med 27:922-935, 1999 37. Rozet JM, Gerber S, Perrault I, et al: Structure and refinement of the physical mapping of the gamma-glutamylcysteine ligase regulatory subunit (GLCLR) gene to chromosome 1p22.1 within the critically deleted region of human malignant mesothelioma. Cytogenet Cell Genet 82:91-94, 1998 38. El-akawi Z, Abu-hadid M, Perez R, et al: Altered glutathione metabolism in oxaliplatin resistant ovarian carcinoma cells. Cancer Lett 105:5-14, 1996 39. Yao KS, Godwin AK, Johnson SW, et al: Evidence for altered regulation of gamma-glutamylcystcine synthetase gene expression among cisplatin-sensitive and cisplatin-resistant human ovarian cancer cell lines. Cancer Res 55:4367-4374, 1995
755
nificant correlation between the heavy chain and multidrug resistance protein (MRP) 2 (P ⴝ .048), whereas no correlation was observed between the light chain and MRP1 or MRP2. Treatment of cultured mesothelioma cells with buthionine sulfoximine (BSO), to inhibit ␥GCS, significantly potentiated cisplatin-induced cytotoxicity mainly by nonapoptotic mechanism when assessed by counting the living cells, TUNEL (terminal deoxytransferase-mediated dUTP nick-end labeling) assay, and caspase-3 cleavage. In conclusion, ␥GCS is highly positive in most cases of malignant mesothelioma and may play an important role in the primary drug resistance of this tumor in vivo. HUM PATHOL 33:748-755. Copyright 2002, Elsevier Science (USA). All rights reserved. Key words: antioxidant, buthionine sulfoximine, cisplatin, glutathione, multidrug resistance protein. Abbreviations: MnSOD, manganese superoxide dismutase; GSH, glutathione; ␥GCS, ␥-glutamylcysteine synthetase; ␥GCSh, heavy chain of ␥GCS; ␥GCSl, light chain of ␥GCS; GST, glutathione Stransferase; MRP, multidrug resistance protein; HPFs, high power fields; BSO, buthionine sulfoximine.
Mesothelioma is a fatal tumor caused in 80% of the cases by occupational exposure to asbestos fibers.1 It is a rare disease, but because of its long latency period its incidence is still increasing despite the industrial restriction of asbestos use in the West. Asbestos fibers are known to generate reactive free radicals and induce antioxidant enzymes,2 which may play role both in the pathogenesis and high drug resistance of this disease. Although mesothelioma is a rare tumor, it provides an ideal model of a drug- and oxidant-resistant malignancy for the investigation of antioxidant mechanisms in cancer. In our previous studies, we found that manganese superoxide dismutase (MnSOD), one of the most important antioxidant enzymes, is consistently upregulated in mesothelioma tissue and cultured mesothelioma cells.3,4 We also found that MnSOD alone does not explain oxidant or drug resistance of those cells.5 Given the importance of hydrogen peroxide scavenging
mechanisms and glutathione (GSH)-related pathways in the oxidant and drug resistance of malignant cells,6 we conducted additional studies on these mechanisms in mesothelioma.4,7 Those results suggest that catalase does not have any major significance in the drug resistance of this disease, whereas GSH depletion enhances sensitivity of mesothelioma cells, at least to epirubicin.7,8 The rate-limiting enzyme in GSH synthesis is ␥-glutamylcysteine synthetase (␥GCS).9 This enzyme may play significant role in the primary resistance of mesothelioma because a recent microarray study analyzed the mRNA expression of more than 6,500 genes in cultured mesothelioma cells and found that ␥GCS mRNA was expressed in the malignant cells when compared with nonmalignant cells.10 ␥GCS, which is feedback inhibited by GSH, consists of two subunits: a heavy subunit with catalytic function (␥GCSh) and a light subunit with a regulatory role (␥GCSl).11 The heavy subunit has been studied more because it is considered to be functionally more important than the light subunit. Many studies, however, show that both subunits are induced by oxidants and apparently also by oxidant-generating anticancer drugs and that, at least in some cell lines, the induction correlates with drug resistance.12-18 Other than ␥GCS, additional mechanisms are closely associated with GSH. These pathways include GSH peroxidase in the decomposition of hydrogen peroxide, GSH reductase in converting oxidized GSH to the reduced form, and ␥-glutamyl transpeptidase in the GSH salvage pathway. GSH S-transferases (GSTs)—in
From the Department of Internal Medicine and Department of Pathology, University of Oulu; and the Finnish Institute of Occupational Health, Helsinki, Finland. Accepted for publication May 1, 2002. Supported by the Finnish Antituberculosis Association Foundation, The Cancer Society of Finland, The Juselius Foundation, The Ida Montin Foundation, and The Ahokas Foundation. Address correspondence and reprint requests to Vuokko L. Kinnula, MD, PhD, University of Oulu, Department of Internal Medicine, Box 50, 90014 University of Oulu, Helsinki, Finland. Copyright 2002, Elsevier Science (USA). All rights reserved. 0046-8177/02/3307-0011$35.00/0 doi:10.1053/hupa.2002.126191
748
␥-GLUTAMYLCYSTEINE SYNTHETASE IN MESOTHELIOMA (Ja ¨ rvinen et al)
detoxification of electrophiles—and multidrug resistance proteins, or MRPs (especially MRP1 and MRP2)—in enhancing drug transport from the cells— are probably among the most important pathways in the drug resistance of many malignant tumors.19-21 Our recent study found that the most resistant mesothelioma cell line contained the highest GST activity and increased level of GST.7 Many cell line studies have also suggested an association between the expression of MRP1 and ␥GCS22-24 in resistant cells, but unlike MRP1, MRP2 has been shown to confer resistance against cisplatin.25 However, no studies exist regarding the correlation between ␥GCS and MRP proteins in tumor biopsy specimens. We analyzed the expression and location of the heavy and light subunits of ␥GCS in healthy human pleura and in mesothelioma by immunohistochemistry and assessed the association between ␥GCS immunostaining in mesothelioma tissue and tumor proliferation, apoptosis, patient survival, or MRP1 and MRP2 expression. We also investigated the expression of ␥GCS in 7 mesothelioma cell lines, which had been established from untreated mesothelioma patients.26 In two of these cell lines, with a high or intermediate level of ␥GCS expression, the effects of ␥GCS inhibition were investigated by assessing irreversible cell injury and apoptosis. MATERIALS AND METHODS Patients and Tumor Tissues Mesothelioma tissue samples from 34 patients (28 men and 6 women), as well as from healthy human pleura, were retrieved from the files of the Department of Pathology at Oulu University Hospital. The healthy control tissues were obtained from nonsmokers investigated for nonmalignant lung diseases such as sarcoidosis; in those cases the pleura was nonneoplastic and noninflamed, and the biopsy specimen showed normal histopathology. Four of the mesothelioma patients had received anticancer drugs before undergoing surgery; 1 patient had been treated with combination therapy that included radiation therapy. Malignant mesotheliomas were subclassified according to the criteria from the World Health Organization.27 Three subtypes of mesothelioma were included: epitheloid (n ⫽ 23), sarcomatoid (n ⫽ 7), and biphasic (n ⫽ 4). For survival evaluation, patient records were used to obtain clinical data from the follow-up period.
Cell Cultures and Exposures Seven mesothelioma cell lines (M10K, M14K, M24K, M25K, M28K, M33K, and M38K) were established from the tumor tissue of untreated mesothelioma patients, with the exception of M10K cell line, which had been established from mesothelioma metastasis. MeT5A is a nonmalignant, SV-40 transformed cell line (American Type Culture Collection, Rockville, MD). The cells were cultured in Roswell Park Memorial Institute (RPMI) medium with 10% phosphate buffer solution (PBS), 100 U/mL penicillin, 100 g/mL streptomycin, and 0.03% L-glutamine (all from Life Technologies, Paisley, Scotland) at 37 C° in a 5% carbon dioxide atmosphere. Previous studies have shown that M38K cells are the most resistant cell line and contain markedly higher intracel-
lular GSH concentrations than M14K cells, the latter being also more sensitive against exogenous epirubicin;7 therefore, these two cell lines were selected for further exposures. In selected experiments, they were pretreated with buthionine sulfoximine (BSO) to cause GSH depletion. According to our previous studies, 0.2 mM BSO was selected for M14K cells, and 1 mM BSO was selected for M38K cells.7 However, because of the possible toxicity of BSO alone, incubations also were conducted in the presence of 0.1 mM (M14K cells) and 0.5 mM (M38K cells) BSO. Cisplatin (Platinol; Bristol-Myers Squibb, New York, NY) was chosen because it is widely used in the treatment of mesothelioma and other lung tumors, and increased GSH synthesis may be associated with the resistance of cancer cells to classical alkylating agents, such as cisplatin.28,29 The cells were pretreated with BSO for 16 hours, after which they were exposed to cisplatin (10 M–100 M) for 48 hours. To ensure that GSH depletion occured during the entire incubation period, the same concentration of BSO was added to the incubation medium 24 hours after starting the exposure to cisplatin.
Immunohistochemistry of ␥GCS The biopsy specimens were first fixed in 10% neutral formalin, dehydrated, and embedded in paraffin. The rabbit antihuman antibodies for ␥GCSh and ␥GCSl were kindly provided by Dr. Terrance Kavanagh (University of Washington, Seattle). These antibodies were been previously tested in human lung tissues and cultured cells.5,30 Four-micron–thick sections were cut from a representative paraffin block; the sections were deparaffinized in xylene and rehydrated in descending ethanol series. Endogenous peroxidase was consumed by incubating the sections in 0.1% hydrogen peroxide in absolute methanol for 10 minutes. To enhance immunoreactivity, the sections were incubated in 10 mmol citrate buffer (pH 6.0) and boiled in a microwave oven for 2 minutes at 850 W and 8 minutes in 350 W. The paraffin blocks were then incubated with 2% milk powder to diminish background staining. The sections were incubated with the primary antibodies for ␥GCSh and ␥GCSl with dilutions of 1:1000. Immunostaining was performed using the Histostain-Plus Kit (Zymed, South San Francisco, CA) and the chromogen was aminoethyl carbazole (Zymed) In negative controls, the primary antibodies were substituted with FBS or nonimmune rabbit serum. The quantity of the immunostaining was evaluated by assessing the staining as follows: 0 ⫽ no positive immunostaining; 1 ⫽ ⬍ 33% of tumor cells showing positive staining; 2 ⫽ 33% to 66% of tumor cells showing positive staining; and 3 ⫽ 67% to 100% of tumor cells showing positive staining. Also, the intensity of the immunostaining was evaluated by division into groups as follows: 0 ⫽ no positive immunostaining; 1 ⫽ weak cytoplasmic staining; 2 ⫽ moderate cytoplasmic staining; and 3 ⫽ strong cytoplasmic staining. A combined score for the immunostaining, based on both qualitative and quantitative immunostaining, was composed by adding both the qualitative and quantitative score and then divided in four main groups as follows: ⫺ ⫽ no immunostaining/score 0; ⫹ ⫽ weak immunostaining/scores 1 to 2; ⫹⫹ ⫽ moderate immunostaining/scores 3 to 4; and ⫹⫹⫹ ⫽ strong immunostaining/scores 5– 6. Two authors (KJ, YS) evaluated the immunostaining separately, and the correlation coefficient was established according to Cohen’s kappa statistics as described.31 The correlation was good (heavy, 0.6; light, 0.8).
749
HUMAN PATHOLOGY
Volume 33, No. 7 (July 2002)
gland) (dilution 1:50 000). ␥GCS and caspase 3 were detected by an enhanced chemiluminescense system (Amersham), and the luminol excitation was imaged on x-ray film.
3⬘-End Labelling of DNA in Apoptotic Cells and Assessment of Cell Proliferation in Tumor Biopsy Specimens To detect apoptotic cells, in situ labelling of the 3⬘-ends of the DNA fragments generated by apoptosis-associated endonucleases was performed using the ApopTag in situ apoptosis detection kit (Oncor, Gaithesburg, MD) as previously described.32 Cells were defined as apoptotic if the whole nuclear area of the cell labelled positively. Apoptotic bodies were defined as small, positively labelled, globular bodies in the cytoplasm of the tumor cells that could be found either singly or in groups. To estimate the apoptotic index (the percentage of apoptotic events in a given area), apoptotic cells and bodies were counted in 10 high-power fields (HPFs), and this figure was divided by the number of tumor cells in the same HPFs. Tumor cell proliferation was studied with a monoclonal mouse antihuman Ki-67 antibody (Zymed, San Fransisco, CA). Sections 5m thick were cut from representative tumor blocks. The sections were first deparaffinized in xylene and rehydrated in descending ethanol series. To enhance immunoreactivity, the sections were incubated in 10 mmol citrate buffer (pH 6.0) and boiled for 2 minutes at 850 W and then 8 minutes at 350 W. Endogenous peroxidase activity was eliminated by incubation in 0.1% hydrogen peroxide in absolute methanol for 10 minutes. The concentration of primary antibody for anti-Ki-67 was 1:50. In the immunostaining, the avidin-biotin-peroxidase complex method was used, and the color was developed using 3,3⬘-diaminobenzidine. The sections were lightly counterstained with hematoxylin and mounted with Eukitt (Kindler, Freiburg, Germany). Replacement of the primary antibodies with PBS and serum isotype, recommended by the manufacturers at pH 7.2, were used as negative controls. The results for Ki-67 immunostaining were evaluated by counting the percentage of positive nuclei of the whole tumor cell population. For Ki-67, the staining was evaluated semiquantitatively and divided into the following 4 groups: weak staining ⫽ fewer than 5% of cell nuclei positive; moderate staining ⫽ 5% to 10% of cell nuclei positive; strong staining ⫽ 10% to 50% of cell nuclei positive; and very strong staining ⫽ greater than 50% of cell nuclei positive.
Western Blot Analysis The cells were detached with trypsin, centrifuged and washed with PBS, and then mixed with the electrophoresis sample buffer and boiled for 5 minutes at 95 C°. Cell protein was measured by using the Bio-Rad (Hercules, CA) method, and 70 g cell protein was applied per lane in a 12% sodium dodecyl sulphate-polyacrylamide gel. The gel was electrophoresed for 1.5 h (90V) at room temperature, and the protein was transferred (60 minutes, 100V) onto Hybond ECL nitrocellulose membranes (Amersham, Buckinghamshire, England) in a Mini-PROTEAN II Cell (Bio-Rad Mississagua, Ontario, Canada). The blotted membrane was incubated with rabbit antibodies to ␥GCSh and ␥GCSl (dilution 1:40,000 and 1:20,000, respectively) or to caspase 3 (dilution 1:2000) followed by a donkey antirabbit antibody conjugated to horseradish peroxidase (Amersham Biosciences, Amersham, En-
Measurement of Cytotoxicity Depletion of high-energy nucleotides was used because of the sensitivity of this method in the assessment of cell injury. The cells were preincubated for 16 hours with 0.1 mmol [14C] adenine (specific activity, 51 to 55 mCi mmol⫺1; Amersham) to preload the high-energy nucleotides in intact cells. Prelabelled cells were washed three times and then exposed to cisplatin. After 48 hours, the medium was collected, and the cells were extracted with 0.4 mol/L perchloric acid. Thin-layer chromatography was used to separate the purine nucleotides (adenosine triphosphate, adenosine diphosphate, and adenosine monophosphate) in the cell extract as well as the medium and the nucleotide catabolic products (xanthine, hypoxanthine, and uric acid) in the medium.33 The results are expressed as percent distribution of radioactivity (counts per minute [cpm]) between nucleotides in the cells, nucleotides leaked to the medium, and catabolic products in the medium. In a healthy cell, the plasma membrane is not permeable to high-energy nucleotides, but catabolic products are extracted from the cell. Therefore, the appearance of nucleotides in the extracellular medium is a marker for cell membrane injury and cytotoxicity. Previous and preliminary studies indicate that nucleotide depletion is more sensitive than the markers of irreversible cell injury and that these results correlate with each other.34 Therefore, in additional cell injury experiments, the cells were washed with PBS, trypsinized, and counted under a light microscope. Because cell counting does not reveal anything about the mechanism of cell survival, apoptosis was also evaluated from cultured cells treated with BSO and/or exposed to cisplatin. In addition, caspase 3 cleavage was used as another marker of apoptosis as described above (Western blot analysis).
Assessment of Apoptosis in Cultured Cells For the apoptosis assays, the cells were cultured, pretreated with BSO for 16 hours, and then exposed to cisplatin for 48 hours. The pellets were fixed in 10% neutral formalin overnight, after which the formalin was removed, and melted 2% agar was laid over the pellets. The agar blocks were further embedded in paraffin. Sections 5 m thick were cut from the cell blocks and processed for detection of apoptosis as described above. Apoptosis was determined morphologically by in situ labelling of 3⬘-ends of the DNA fragments (TUNEL). Apoptotic cells with clearly labelled nuclei as well as smaller apoptotic bodies were calculated. The percentages of positive cells were expressed as apoptotic indices.
Statistical Analysis Data are expressed as mean ⫾ standard error. SPSS (7.5) for Windows was used in patient correlation data; the differences between the two groups were analyzed by paired 2-tailed t test. The significance of correlations was determined using Fisher exact probability test. The survival of the patients
™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™3 FIGURE 1. Expression of ␥GCS subunits in healthy pleura and in biopsy specimens of mesothelioma. (A) healthy pleura and ␥GCSh. (B) Healthy pleura and ␥GCSl. (C) Epithelial mesothelioma and ␥GCSh. (D) Epithelial mesothelioma and ␥GCSl. (E) Sarcomatoid mesothelioma and ␥GCSl. (F) Sarcomatoid mesothelioma and ␥GCSl. (G and H) Serum controls. Immunostaining was done using the Histostain-Plus Kit (Zymed), and the chromogen was aminoethyl carbazole (Zymed) Original magnification (A) ⫻ 280 (B–H) ⫻ 180.
750
751
HUMAN PATHOLOGY
Volume 33, No. 7 (July 2002)
TABLE 1. A Combined Score (Qualitative and Quantitative) of the Immunoreactivity of ␥GCSh and ␥GCSl in Malignant Mesothelioma Tissue Samples Patient No.
Histological Type
Age (y)/ Gender
␥GCSh
␥GCSl
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
Sarcomatoid Epithelial Biphasic Sarcomatoid Epithelial Epithelial Epithelial Biphasic Epithelial Biphasic Epithelial Sarcomatoid Sarcomatoid Epithelial Epithelial Epithelial Epithelial Sarcomatoid Epithelial Biphasic Epithelial Sarcomatoid Epithelial Epithelial Epithelial Epithelial Epithelial Epithelial Sarcomatoid Epithelial Epithelial Epithelial Epithelial Epithelial
42/F 56/M 57/F 58/M 74/M 66/M 70/F 70/M 50/M 68/M 63/M 52/M 33/F 73/M 67/M 59/M 54/M 67/M 78/M 57/M 57/M 73/M 78/M 79/M 63/M 66/M 72/M 73/F 68/M 69/M 69/M 77/M 52/F 67/F
⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹ ⫹
⫺ ⫹⫹ ⫹⫹ ⫺ ⫹⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹ ⫹ ⫹⫹ ⫺ ⫹⫹ ⫺ ⫺ ⫺ ⫹⫹⫹ ⫺ ⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹
showing no reactivity. The histological type of mesothelioma had no statistically significant effect on ␥GCS reactivity although the epithelial subtypes tended to stain somewhat more strongly (Fig 1). There was no correlation with either ␥GCSh or ␥GCSl immunoreactivity and patient survival (P ⫽ .07 and .42, respectively). Staining with Ki-67 showed no association with ␥GCS expression and tumor cell proliferation (P ⫽ .42 for ␥GCSh and P ⫽ .25 for ␥GCSl). No correlation was seen between apoptosis and ␥GCS (␥GCSl, P ⫽ .79; ␥GCSh, P ⫽ .55). We recently investigated the expression of MRP1 and MRP2 using this same biopsy material of mesothelioma patients.35 For the present study, we assessed if there were any correlations between ␥GCS reactivities and MRP1 or MRP2. There was a significant correlation between ␥GCSh and MRP2 expression (P ⫽ .048), whereas the corresponding correlation between ␥GCSl and MRP2 was not significant (P ⫽ .35). Neither was any correlation seen between MRP1 and ␥GCS (␥GCSh, P ⫽ .67; ␥GCSl, P ⫽ 0.50). Expression of ␥GCS in Mesothelioma Cell Line Cells In Vitro Western blot studies with 7 mesothelioma cell lines indicated that both ␥GCS subunits were detectable in all mesothelioma cell lines investigated. M38K cells, which represent the most resistant cell line,4,5,7 contained the highest level of ␥GCSh (Fig 2). Two cell lines, M14K and M38K, were selected for further studies on cytoxicity. Exposure to Cisplatin With And Without BSO
Abbreviations: ␥GCSh, heavy chain of ␥GCS; ␥GCSl, light chain of ␥GCS; F, female; M, male; ⫺, no immunostaining, ⫹, weak immunostaining; ⫹⫹, moderate immunostaining; ⫹⫹⫹, strong immunostaining.
The M38K cell line was more resistant than the M14K cell line to cisplatin (Fig 3). The concentrations of cisplatin for the next experiments were chosen according to results from this assay. Pretreatment with BSO to deplete intracellular GSH potentiated the cisplatin-induced toxicity significantly (Fig 4). Caspase 3
in relation to the immunoreactivities was assessed by log-rank test, with P ⬍ .05 being statistically significant.
RESULTS Immunohistochemical Expression of ␥GCS in Mesothelioma Biopsy Specimens Nonmalignant mesothelium showed no reactivity of ␥GCSh or ␥GCSl in any of the cases (Fig 1A and B). Most mesotheliomas (29 of 34) were strongly positive for ␥GCSh (Fig 1C and E). The qualitative and quantitative immunostaining scores were combined as described in the Materials and Methods section, and the summary of those results is shown in Table 1. The staining of ␥GCSh was most positive in malignant cells, and the color appeared in the cytosol. Additional positivity could be observed in the stroma of the tumor as well as inflammatory cells. The ␥GCSl immunoreactivity was considerably less positive (Fig 1D and F), with 15 cases showing strong/ moderate reactivity, 7 showing weak reactivity, and 12
FIGURE 2. Western blot analyses showing expression of (A) ␥GCSh (molecular weight ⬃ 73,000) and (B) ␥GCSl (molecular weight ⬃30,000) in nonmalignant transformed MeT5A mesothelial cells and in 7 malignant mesothelioma cell lines.
752
␥-GLUTAMYLCYSTEINE SYNTHETASE IN MESOTHELIOMA (Ja ¨ rvinen et al)
enhanced by BSO (data not shown). The apoptotic index in M38K cells was always very low (between 2.4% in controls and 4.5% in cisplatin-exposed, BSO-treated cells). DISCUSSION In this study, we found that most malignant mesotheliomas express both subunits of ␥GCS and that healthy mesothelium is negative for ␥GCS immunoreactivity. There was a significant correlation between ␥GCSh and MRP2 reactivity, but neither subunit of ␥GCS showed correlation with tumor proliferation or patient survival. As hypothesized, the mesothelioma cell
FIGURE 3. Depletion of cellular high-energy nucleotides (CNucl), accumulation of nucleotide catabolic products (M-HxUa-X), and leakage of intact nucleotides (M-Nucl) in (A) M14K or (B) M38K cells exposed to varying concentrations of cisplatin. Values are mean ⫾ standard error from three separate experiments. P ⬍ .05 vs unexposed (cont).
cleavage has been used as an indicator of the apoptotic process in various cell types. Preliminary studies showed no cleavage of caspase 3 with BSO alone, but cleavage was induced by cisplatin (data not shown). Additional studies using M14K cells showed caspase 3 cleavage within 48 hours using 50 M cisplatin, which was not potentiated by BSO pretreatment (Fig 5A). M38K cells, the most resistant cell line with the highest ␥GCS content, also were more resistant to apoptosis than were M14K cells. In this cell line, no apoptosis could be seen when assessed by caspase 3 cleavage after cisplatin exposure without or with BSO (Fig 5B). Given that caspase 3 cleavage is not a sensitive indicator of apoptosis, apoptosis also was assessed morphologically after the same exposures. Again, cisplatin caused apoptosis in M14K cells but not in M38K cells; apoptosis was not
FIGURE 4. Viability of M14K cells and M38K cells exposed to the indicated concentrations of cisplatin with and without BSO. M14K cells (A) and M38K cells (B) were treated with 0.1 mM (M14K) or 0.5 mM (M38K) BSO and exposed to cisplatin for 48 hours. Cell cytotoxicity was assessed by calculating living cells after the exposures. Values are mean ⫾ standard error from three separate experiments. P ⬍ .05 not pretreated vs BSO pretreated.
753
HUMAN PATHOLOGY
Volume 33, No. 7 (July 2002)
FIGURE 5. Western blot analyses of caspase 3 cleavage in (A) M14K cells and (B) M38K cells exposed to varying concentrations of cisplatin with and without BSO. Caspase 3 cleavage can be observed in M14K cells exposed to 50 M cisplatin, whereas M38K cells show no apoptosis by 50 or 100 M concentration, the latter concentration of which is toxic to M38K cells. BSO had no effect on caspase 3 cleavage when compared with the effect of cisplatin in either cell type.
line with high GSH content and ␥GCSh expression turned out to be resistant toward apoptosis and cisplatin-induced toxicity. Numerous studies have shown that drug-resistant cancer cell lines contain high GSH concentrations and that cell viability can be modulated by GSH depletion.16,21,36 One study on lung cancer showed that approximately 70% of lung tumors express both subunits of ␥GCS.30 Recent microarray study also showed increased expression of ␥GCS mRNA in one mesothelioma cell line compared with nonmalignant MeT5Atransformed mesotelial cells,10 which is why it is even more important to confirm this finding in the tumor biopsy samples of malignant mesothelioma. One study suggested overexpressions of ␥GCSh mRNA and MRP1 in doxorubicin-resistant mesothelioma cells18 and another the deletion of the ␥GCSl gene in mesothelioma.37 In agreement with the microarray study,10 we found that most mesotheliomas were highly positive for ␥GCSh reactivity. The deletion of the ␥GCS light subunit gene may not be a universal feature in mesotheli-
oma because more than half of our biopsy specimens were positive for ␥GCSl reactivity. In our relatively large group of biopsy specimens, no association was found between either subunit of ␥GCS and MRP1 expression. Instead, significant correlation between ␥GCSh and MRP2 expression may play role in the primary resistance of this disease. We found no significant correlation between the reactivity of ␥GCS and patient survival. This does not rule out the possibility of an association because the high expression levels may be related to the high primary resistance of this disease. Association between the light and the heavy chains was also nonsignificant, which could indicate that some of the ␥GCSh gene product may be protein with no enzyme activity or that the light subunit is not needed for full enzyme activity. In fact, it has been stated that the heavy subunit alone comprises approximately 46% of the activity of the holoenzyme.36 GSH level does not necessarily correlate with ␥GCS activity either because the major controlling factor of intracellular GSH, at least in some cell types, is its transport out of the cell via the plasma membrane pumps, most importantly the proteins of the MRP family.36 In our recent study, 58% of these mesothelioma patients were positive for MRP1 and 33% for MRP2. The result of ␥GCSh and MRP2 correlation is in line with the recent finding that MRP2, but not MRP1, may be associated with primary cisplatin resistance.25 Given that tissue biopsy samples of mesothelioma contain tissue stroma, multiple cell types, and nonmalignant cells, the assessment of active ␥GCS protein and/or GSH content from tumor tissue homogenates is not reliable. Therefore, we analyzed both the expression of ␥GCS and the GSH content in cultured mesothelioma cells and assessed the importance of GSH depletion on the oxidant and drug resistance in these cells. M38K cells, which are the most resistant mesothelioma cells in our laboratory, contained not only the highest ␥GCSh reactivity but also had the highest GSH levels in vitro.5,7 The results with cisplatin-exposed cells also were consistent with previous investigations of other cell lines in which ␥GCS expression and GSH were investigated.15,16,38,39 Our recent5 and present results confirm the importance of ␥GCS and GSH in the resistance of mesothelioma cells to oxidants, to epirubicin,5 and to cisplatin in vitro. They also suggest that GSH protects these cells against nonapoptotic cell injury. In conclusion, our study suggests that the synthesis of GSH by ␥GCS may be of major importance in the primary drug resistance of malignant cells. Acknowledgment. The authors thank Dr. Kaija Linnainmaa for providing the mesothelioma cell lines; Dr Terrance Kavanagh for the ␥GCS antibodies and Dr Paavo Pa¨a¨kko¨ for the excellent research facilities in Oulu. The technical assistance of Mrs Raija Sirvio¨ and Mr Manu Tuovinen is gratefully acknowledged. The authors also thank Proffessor Kari Raivio for his critical comments on the manuscript.
754
␥-GLUTAMYLCYSTEINE SYNTHETASE IN MESOTHELIOMA (Ja ¨ rvinen et al)
REFERENCES 1. Mossman BT, Bignon J, Corn M, et al: Asbestos: Scientific developments and implications for public policy. Science 247:294301, 1990 2. Mossman BT, Hansen K, Marsh, JP et al: Mechanisms of fibre-induced superoxide release from alveolar macrophages and induction of superoxide dismutase in the lungs of rats inhaling crocidolite. IARC Sci Publ 90:81-92, 1989 3. Kinnula VL, Pietarinen-Runtti P, Raivio K, et al: Manganese superoxide dismutase in human pleural mesothelioma cell lines. Free Radic Biol Med 21:527-532, 1996 4. Kahlos K, Anttila S, Asikainen T, et al: Manganese superoxide dismutase in healthy human pleural mesothelium and in malignant pleural mesothelioma. Am J Respir Cell Mol Biol 18:570-580, 1998 5. Ja¨rvinen K, Pietarinen-Runtti P, Linnainmaa K, et al: Antioxidant defense mechanisms of human mesothelioma and lung adenocarcinoma cells. Am J Physiol Lung Cell Mol Physiol 278:L696-L702, 2000 6. Tew KD: Glutathione-associated enzymes in anticancer drug resistance. Cancer Res 54:4313-4320, 1994 7. Kinnula K, Linnainmaa K, Raivio KO, et al: Endogenous antioxidant enzymes and glutathione S-transferase in protection of mesothelioma cells against hydrogen peroxide and epirubicin toxicity. Br J Cancer 77:1097-1102, 1998 8. Kahlos K, Soini Y, Sormunen R, et al: Expression and prognostic significance of catalase in malignant mesothelioma. Cancer 91:1349-1357, 2001 9. Meister A, Anderson ME: Glutathione. Annu Rev Biochem 52:711-60, 1983 10. Rihn BH, Mohr S, McDowell SA, et al: Differential gene expression in mesothelioma. FEBS Lett 480:95-100, 2000 11. Seelig, GF, Simondsen RP, Meister A: Reversible dissociation of ␥GCS into 2 subunits. J Biol Chem 259:9345-9347, 1984 12. Galloway DC, Blake DG, Shepherd AG, et al: Regulation of human gamma-glutamylcysteine synthetase: Co-ordinate induction of the catalytic and regulatory subunits in HepG2 cells. Biochem J 328:99-104, 1997 13. Shi MM, Kugelman A, Iwamoto T, et al: Quinone-induced oxidative stress elevates glutathione and induces gamma-glutamylcysteine synthetase activity in rat lung epithelial L2 cells. J Biol Chem 269:26512-26517, 1994 14. Rahman I, MacNee W: Oxidative stress and regulation of glutathione in lung inflammation. Eur Respir J 16:534-554, 2000 15. Iida T, Mori E, Mori K, et al: Co-expression of gammaglutamylcysteine synthetase sub-units in response to cisplatin and doxorubicin in human cancer cells. Int J Cancer 82:405-411, 1999 16. Zhang K, Mack P, Wong KP: Glutathione-related mechanisms in cellular resistance to anticancer drugs. Int J Oncol 12:871882, 1998 17. Tipnis SR, Blake DG, Shepherd AG, et al: Overexpression of the regulatory subumit of g-glutamylcysteine synthetase in HeLa cells increases g-glutamylcysteine synthetase activity and confers drug resistance. Biochem J 337:559-566, 1999 18. Ogretmen B, Bahadori HR, McCauley MD, et al: Co-ordinated over-expression of the MRP and ␥GCS genes, but not MDR1, correlates with doxorubicin resistance in human malignant mesothelioma cell lines. Int J Cancer 75:757-761, 1998 19. Cole SP, Bhardwaj G, Gerlach JH, et al: Overexpression of a transporter gene in a multidrug-resistant human lung cancer cell line. Science 258:1650-1654, 1992 20. Konig J, Nies AT, Cui Y, et al: Conjugate export pumps of the multidrug resistance protein (MRP) family: Localization, substrate specificity, and MRP2-mediated drug resistance. Biochim Biophys Acta 1461:377-394, 1999
21. O’Brien ML, Tew KD: Glutathione and related enzymes in multidrug resistance. Eur J Cancer 32:967-978, 1996 22. Ishikawa T, Bao JJ, Yamane Y, et al: Coordinated induction of MRP/GS-X pump and gamma-glutamylcysteine synthetase by heavy metals in human leukemia cells. J Biol Chem 271:14981-14988, 1996 23. Kuo MT, Bao JJ, Curley SA, et al: Frequent coordinated overexpression of the MRP/GS-X pump and gamma-glutamylcysteine synthetase genes in human colorectal cancers. Cancer Res 56:36423644, 1996 24. Kigawa J, Minagawa Y, Cheng X, et al: Gamma-glutamyl cysteine synthetase up-regulates glutathione and multidrug resistance-associated protein in patients with chemoresistant epithelial ovarian cancer. Clin Cancer Res 4:1737-1741, 1998 25. Borst P, Evers R, Kool M, et al: A family of drug transporters: The multidrug resistance-associated proteins. J Natl Cancer Inst 92: 1295-1302, 2000 26. Pelin-Enlund K, Husgafvel-Pursiainen K, Tammilehto L, et al: Asbestos-related malignant mesothelioma: Growth, cytology, tumorigenicity and consistent chromosome findings in cell lines from five patients. Carcinogenesis 11:673-681, 1990 27. Travis WD, Colby TV, Corrin B, et al: Histological typing of lung and pleural tumors, in: World Health Organization International Histological Classification of Tumours (ed 3). Berlin, Springer, 1999, pp 51-54 28. Hao XY, Bergh J, Brodin O, et al: Acquired resistance to cisplatin and doxorubicin in a small cell lung cancer cell line is correlated to elevated expression of glutathione-linked detoxification enzymes. Carcinogenesis 15:1167-1673, 1994 29. Sugarbaker DJ, Norberto JJ: Multimodality management of malignant pleural mesothelioma. Chest 113:61S-65S, 1998 (suppl 1) 30. Soini Y, Na¨pa¨nkangas U, Ja¨rvinen K, et al: Expression of gamma-glutamyl cysteine synthetase in nonsmall cell lung carcinoma. Cancer 92:2911-2919, 2001 31. Silcocks PB: Measuring repeatability and validity of histological diagnosis—A brief review with some practical examples. J Clin Pathol 36:1269-1275, 1983 32. Soini Y, Virkaja¨rvi N, Lehto V-P, et al: Hepatocellular carcinomas with a high proliferation index and a low degree of apoptosis and necrosis are associated with a shortened survival. Br J Cancer 73:1025-1030, 1996 33. Aalto TK, Raivio KO: Adenine nucleotide depletion from endothelial cells exposed to xanthine oxidase. Am J Physiol 259:C883C888, 1990 34. Spragg RG, Hinshaw DB, Hyslop PA, et al: Alterations in adenosine triphosphate and energy charge in cultured endothelial and P388D1 cells after oxidant injury. J Clin Invest 76:1471-1476, 1985 35. Soini Y, Ja¨rvinen K, Kaarteenaho-Wiik R, et al: The expression of P-glycoprotein and multidrug-resistance proteins 1 and 2 (MRP1 and MRP2) in human malignant mesothelioma. Ann Oncol 12:1239-1245, 2001 36. Griffith OW: Biologic and pharmacologic regulation of mammalian glutathione synthesis. Free Radic Biol Med 27:922-935, 1999 37. Rozet JM, Gerber S, Perrault I, et al: Structure and refinement of the physical mapping of the gamma-glutamylcysteine ligase regulatory subunit (GLCLR) gene to chromosome 1p22.1 within the critically deleted region of human malignant mesothelioma. Cytogenet Cell Genet 82:91-94, 1998 38. El-akawi Z, Abu-hadid M, Perez R, et al: Altered glutathione metabolism in oxaliplatin resistant ovarian carcinoma cells. Cancer Lett 105:5-14, 1996 39. Yao KS, Godwin AK, Johnson SW, et al: Evidence for altered regulation of gamma-glutamylcystcine synthetase gene expression among cisplatin-sensitive and cisplatin-resistant human ovarian cancer cell lines. Cancer Res 55:4367-4374, 1995
755