1 3508 Vol. 5, , November 1999 Clinical Cancer Research Apoptosis and Expression of Apoptosis Regulating Proteins bcl-2, mcl-1, bcl-x, and bax in Malignant Mesothelioma 1 Ylermi Soini, 2 Vuokko Kinnula, Riitta Kaarteenaho-Wiik, Esko Kurttila, Kaija Linnainmaa, and Paavo Pääkkö Departments of Pathology and Internal Medicine, University of Oulu and Oulu University Hospital, FIN Oulu, Finland, and Occupational Health Institute, Helsinki, Finland ABSTRACT We investigated apoptosis and the expression of bcl-2, mcl-1, bcl-x, and bax in histological sections from 35 malignant mesotheliomas and 21 metastatic adenocarcinomas. Moreover, the expression of bcl-2, mcl-1, bcl-x, and bax were assessed by Western blotting in nonmalignant human mesothelial cells (Met5A) and seven malignant cell lines. The apoptotic index in mesotheliomas was %. Patients with mesotheliomas showing a high apoptotic index (>0.75%) had a worse prognosis (P 0.008). bcl-2 positivity was observed in only seven cases, but bcl-x, mcl-1, and bax positivity was seen in all of them. In immunoblotting experiments, all mesothelioma cell lines were negative for bcl-2 but positive for bcl-x, mcl-1, and bax. The apoptotic index in bcl-2-negative mesotheliomas was % and in bcl- 2-positive ones, % (P 0.014). The apoptotic index did not significantly associate with bcl-x, mcl-1, or bax expression (P 0.19, P 0.25, and P 0.46, respectively). No significant difference was observed in apoptosis or expression of bcl-2, bcl-x, or bax between malignant mesotheliomas and metastatic adenocarcinomas. The former, however, showed more often weak mcl-1 immunoreactivity (P 0.01). The results show that the extent of apoptosis may influence patient prognosis. bcl-2 is inversely associated with the apoptotic index but is relatively infrequently expressed in malignant mesotheliomas. Widespread expression of bcl-x, mcl-1, and bax suggests that these proteins may also take part in apoptosis regulation in mesotheliomas. Received 2/4/99; revised 8/11/99; accepted 8/20/99. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This study was partly supported by the Finnish Anti-Tuberculosis Association Foundation, Sigrid Juselius Foundation, and the Finnish Cancer Societies. 2 To whom requests for reprints should be addressed, at Department of Pathology, University of Oulu, Kajaanintie 52 D, P. O. Box 5000, FIN Oulu, Finland. Phone: ; Fax: ; INTRODUCTION Malignant mesothelioma is a tumor with a grim prognosis, and most patients die within 1 2 years of the diagnosis (1). The development of malignant mesothelioma is often associated with an occupational exposure to asbestos (2). Its development is especially linked to crocidolite, which has been shown to induce apoptosis in pleural mesothelial cells by inducing reactive oxygen species (3, 4). Apoptosis and its regulation may thus play an important pathogenetic role in the development of mesothelial neoplasias. Little is known, however, about spontaneous apoptosis and its regulation in neoplastic mesothelial cells, and no studies have thus far been published on the extent of apoptosis in mesotheliomas or its clinical significance. Two important groups of proteins orchestrate apoptosis and the fate of the cell: the bcl-2 family and caspases (5, 6). The bcl-2 family of proteins consists of structurally related proteins that can be either pro- or antiapoptotic (5, 7). Antiapoptotic members include bcl-2, mcl-1, bcl-xl and proapoptotic members bax, bcl-xs, and bad (7). An important characteristic for their apoptosis-regulating function is their capability to homoor heterodimerize (5, 7). The dimerization is mediated through three conserved domains in the proteins called BH1, BH2, and BH3 (5). The association between the proteins is formed when the helical BH3 loop of another protein is attached to a cleft formed by all three domains of another protein (5). Contrary to earlier notions, heterodimerization is not required for the antiapoptotic function but may be needed for the proapoptotic function, at least with proteins containing the BH3 domain (5). Some of the proteins, like bcl-2, bcl-xl, and bax, contain a COOH-terminal domain by which they are attached to cell membranes, especially mitochondria, and may influence the mitochondrial membrane potential and permeability (5, 8 11). This may lead to an efflux of caspase-activating substances, such as cytochrome c, procaspase 3, and apoptosis inducing factor from the mitochondria (5, 8 11). bax has been shown to stimulate the release of these substances, whereas bcl-2 and bcl-xl inhibit it (9 11). bcl-xl also inhibits apoptosis by attachment to apaf-1, a mammalian CED-4 homologue, thus inhibiting its association with procaspase 9 and cytochrome c and a subsequent activation of downstream effector caspases (5). On the other hand, proapoptotic members of the bcl-2 family, such as bax or bik, may prevent such neutralization of apaf-1 by associating with bcl-xl, which then results in an enhancement of apoptosis (5). The expression of the bcl-2 family proteins may vary depending on the site and type of the tumor. Breast and endometrial adenocarcinomas, for instance, show a frequent bcl-2 expression, whereas some other tumors, such as lung, pancreatic, or prostate carcinomas, express bcl-2 less often (12 16). The expression of mcl-1, on the other hand, seems to be higher in tumors and tissues with a low bcl-2 expression (15 17). Expression of bcl-xl has been reported to be high in prostate
2 Clinical Cancer Research 3509 Table 1 Mitotic index, extent of necrosis, apoptotic index, immunoreactivity for bcl-2, bax, mcl-1, and bcl-x in malignant mesotheliomas, and clinical data on the mesothelioma patients Case Mitosis (%) Necrosis (%) Apoptosis (%) bcl-2 bax mcl-1 bcl-x Age Sex Type a Survivors M E M E M E M E F E M E M E M E M E M E M E M E M E M E M E M E M E M E M E M E U M E U M E U F S M S F S M S M S N 33 F S M S M S M S U F B M B M B N 57 M B 24 a E, epithelial; S, sarcomatoid; B, biphasic; U, unknown; N, not done due to exhaustion of the block. and gastric carcinomas, but bax immunoreactivity is more ubiquitous and is found in several types of different tumors (15, 16, 18 20). In malignant mesotheliomas, expression of bcl-2 has been reported in a minority of tumors, and in cell line studies, neoplastic mesothelioma cells have been shown to express bax but not bcl-2 (21 23). The significance of the expression of the bcl-2 family of proteins in tumors lies in their capability to modulate apoptosis and thus influence tumor behavior and treatment. bcl-2 in particular has been shown to protect tumor cells from, among other things, chemotherapy and radiation-induced apoptosis (7). The significance of the other members of the bcl-2 family in this respect is less well known. A low bax expression in breast carcinoma has, however, been associated with a higher resistance to chemotherapy and a poor prognosis (24). The objective of our study was to investigate apoptosis and the expression of apoptosis regulating proteins bcl-2, mcl-1, bcl-x, and bax in malignant mesotheliomas and also whether expression of these proteins would influence the extent of apoptosis and survival of the patients. We also determined cell proliferation by calculating the mitotic index and, additionally, assessed the extent of necrosis from available tumor sections. Finally, we also included a set of metastatic pleural adenocarcinomas in the study material. MATERIALS AND METHODS Histological Material. Thirty-five malignant mesotheliomas and 21 pleural metastatic adenocarcinomas were retrieved from the files of the Department of Pathology, Oulu University Hospital between 1976 and All of the material had been fixed in 10% buffered formalin and embedded in paraffin. Malignant mesotheliomas were subclassified into epithelial, sarcomatoid, and biphasic subtypes according to the criteria given by Armed Forces Institute of Pathology (25). The number of different subtypes and clinical data of the tumors are given in Table 1. Malignant mesotheliomas were distinguished from metastatic adenocarcinomas on the basis of the presence of intra- or extracellular hyaluronic acid in mesothelioma, whereas adenocarcinomas usually contained intracellular, usually globular, periodic acid-schiff-positive and diastase-resistant epithelial mucin. In problematic cases, diagnostic electron microscopy was also performed, demonstrating typical long microvilli in malignant mesotheliomas (26). By immunohistochemistry, metastatic adenocarcinomas of the pleura were diagnosed on the
3 3510 Apoptosis in Malignant Mesothelioma basis of carcinoembryonic antigen positivity and contained often intracytoplasmic epithelial membrane antigen positivity, whereas malignant mesotheliomas were carcinoembryonic antigen-negative and could exhibit membrane-bound positivity for epithelial membrane antigen (26, 27). The 21 pleural metastatic adenocarcinomas consisted of 14 cases metastatic from the lungs, 3 from breast, 2 from kidneys, 1 from bile duct, and 1 case with an unknown primary location. Cell Cultures. Apoptosis regulating proteins were also investigated in human mesothelioma cell lines and nonmalignant mesothelial cells in culture. Human mesothelioma cells (M10K, M14K, M24K, M25K, M28K, M33K, and M38K) have been established from mesothelioma patients (28 30). Met5A cells are SV40 transformed nonmalignant human pleural mesothelial cells (31) and were kindly provided by Dr. C. C. Harris (Laboratory of Human Carcinogenesis, National Cancer Institute, Bethesda, MD). All of the cells were grown in RPMI 1640 supplemented with 10% FCS, 100 units/ml penicillin, 100 g/ml streptomycin, and 0.03% L-glutamine (all from Life Technologies, Paisley, United Kingdom) at 37 C in 5% CO 2 atmosphere. OY-AML8 are human leukemic cells, and the lysates of these cells were provided by Dr. Pirjo Koistinen, Department of Internal Medicine, University of Oulu. 3 -End Labeling of DNA in Apoptotic Cells. To detect apoptotic cells, in situ labeling of the 3 -ends of the DNA fragments generated by apoptosis-associated endonucleases was performed using the ApopTag in situ apoptosis detection kit (Oncor, Gaithersburg, MD) as described previously (14, 32). The sections, after being dewaxed in xylene and rehydrated in ethanol, were incubated with 20 g/ml Proteinase K (Boehringer Mannheim, Mannheim, Germany) at room temperature for 15 min. The endogenous peroxidase activity was blocked by incubating the slides in 2% hydrogen peroxide in PBS (ph 7.2). The slides were then treated with terminal transferase enzyme and digoxigenin-labeled nucleotides, after which anti-digoxigenin-peroxidase solution was applied on the slides. The color was developed with diaminobenzidine, after which the slides were lightly counterstained with hematoxylin. For control purposes, we used tissue sections from hyperplastic lymph nodes showing an increased number of apoptotic B cells within germinal centers and a low number of apoptotic T cells in the interfollicular areas. Assessment of the Apoptotic Index. Cells were defined as apoptotic if the whole nuclear area of the cell labeled positively. Apoptotic bodies were defined as small, positively labeled, 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 HPFs, 3 and this figure was divided by the number of tumor cells in the same HPFs. Immunohistochemical Stainings. A monoclonal antibody (clone 124) against bcl-2 oncoprotein was obtained from Dako (Glostrup, Denmark). Polyclonal antibodies to mcl-1 and bax were obtained from PharMingen (San Diego, CA) and bcl-x 3 The abbreviation used is: HPF, high power field. Fig. 1 Apoptosis in epithelial mesothelioma by terminal deoxynucleotidyl transferase-mediated nick end labeling. Intensely stained apoptotic cells and fragmented apoptotic bodies can be seen in tumor cells in this field (3 -end labeling of fragmented DNA, 400). from Transduction Laboratories (Lexington, KY). Before application of the primary antibodies, the sections were heated in a microwave oven in 10 mm citric acid monohydrate (ph 6.0) for 10 min. After a 30-min incubation with the primary antibody (dilution 1:50 for bcl-2, 1:1000 for mcl-1 and bax, and dilution 1:500 for bcl-x), a biotinylated secondary antimouse or antirabbit antibody (all three from Dakopatts, Copenhagen, Denmark) was applied (dilution 1: ), followed by the avidinbiotin-peroxidase complex (Dakopatts). For all of the immunostainings, the color was developed by diaminobenzidine, and the sections were lightly counterstained with hematoxylin and mounted with Eukitt (Kindler, Freiburg, Germany). Negative control stainings were carried out by substituting nonimmune mouse, goat, or rabbit serum for the primary antibodies. As a positive control for the immunostainings, a lymph node with follicular hyperplasia was used. The intensity of the immunostainings with all of the antibodies was evaluated by dividing the staining reaction in four groups: 1, weak cytoplasmic staining intensity; 2, moderate cytoplasmic staining intensity; 3, strong cytoplasmic staining intensity; and 4, very strong cytoplasmic staining intensity. The quantity of the immunostaining was evaluated as follows: 0, no positive immunostaining; 1, 25% of tumor cells showing cytoplasmic positivity; 2, 25 50% of tumor cells showing cytoplasmic positivity; 3, 50 75% of tumor cells showing cytoplasmic positivity; and 4, 75% of tumor cells showing cytoplasmic positivity. A combined score for the immunostaining, based on both qualitative and quantitative immunostaining, was composed by adding both the qualitative and quantitative score, which was then divided in five main groups:, no immunostaining, score 0;, weak immunostaining, scores 1 2;, moderate immunostaining, scores 3 4;, strong immunostaining, scores 5 6; and, very strong immunostaining, scores 7 8. To check the reproducibility of the evaluation system concerning the immunostaining for the bcl-2 proteins, another pa-
4 Clinical Cancer Research 3511 Fig. 2 A, diffuse cytoplasmic positivity for bcl-2 can be seen in a sarcomatoid mesothelioma. Some tumor areas showed stronger staining (upper left) than the others (immunohistochemical stain, 120). B, diffuse and partly strong globular cytoplasmic positivity for mcl-1 can be seen in this epithelial malignant mesothelioma (immunohistochemical stain, 120). C, strong cytoplasmic positivity for bcl-x can be seen in this sarcomatoid mesothelioma (immunohistochemical stain, 120). D, diffuse cytoplasmic positivity for bax can be seen in an epithelial mesothelioma (immunohistochemical stain, 120). thologist who was unaware of the original assessment reevaluated the immunostaining results according to the system given above. According to Fisher s exact test, there was a significant association between the evaluation of the immunoreactivity for all of the proteins (P 0.004, P 0.015, P , and P for bcl-x, mcl-1, bcl-2, and bax, respectively). Mitoses and Necrosis. Cell proliferation was assessed by counting the number of mitotic figures per 10 HPFs, and this figure was divided by the number of tumor cells in the same HPFs, thus giving the percentage of mitotic cells/whole tumor cell population. The extent of necrosis was assessed by a morphometric method described previously (33). Briefly, a Nikon Labophot light microscope was used, connected through a video camera (Panasonic F10) to a TV monitor (Sony) equipped with a net of 100 points adherent to the screen. The total magnification for counting the volume density was 100, and the total number of counts was 2000 in each case. Western Blot Analysis. The cells were mixed with the electrophoresis sample buffer and boiled; 75 g of cell protein was applied to a 12% SDS-polyacrylamide gel (34). The gel was electrophoresed for 2 h (80V) at room temperature and transferred onto Hybond ECL nitrocellulose membranes (Amersham, Arlington Heights, IL) in a Mini-PROTEAN II Cell (Bio-Rad, Hercules, CA). The blotted membrane was incubated with the antibodies to bcl-2 (Dako; 1:5,000), bcl-x (Transduction Laboratories; 1:400), mcl-1 (PharMingen; 1:1,000), or bax (PharMingen; 1:1,000), followed by treatment with secondary antibody conjugated to horseradish peroxidase (Jackson Immunoresearch Laboratories; 1:10,000). The proteins were detected by enhanced chemiluminescence system (ECL; Amersham). -Actin expression of the cells was detected with a monoclonal anti-actin antibody (Sigma Chemical Co.; 1:3,000), followed by sheep antimouse antibody conjugated to horseradish peroxidase (Amersham; 1:3,000). Cell protein was measured using the Bio-Rad method (35). Statistical Analysis. SPSS for Windows (Chicago, IL) was used for statistical analysis. The significance of the associations was determined using Fisher s exact probability test, correlation analysis, and two-tailed t test. The survival analysis was performed by the Kaplan-Meier curve, and the significance of associations was tested by the log-rank, Breslow and Tarone- Ware tests. Probability values P 0.05 were considered statistically significant.
5 3512 Apoptosis in Malignant Mesothelioma RESULTS An inflamed, nonneoplastic, reactive mesothelium adjacent to the tumorous tissue showed an average apoptotic index of %. The mean apoptotic index in mesotheliomas was % (Fig. 1). The apoptotic index in epithelial mesotheliomas did not significantly differ from that of sarcomatoid and biphasic types (P 0.82). The apoptotic index in metastatic adenocarcinomas was higher ( %) than in mesotheliomas, but the association was not statistically significant (P 0.12). Nonneoplastic mesothelial cells showed immunoreactivity for mcl-1, bcl-x, and bax but not for bcl-2. The immunoreactivity was stronger for mcl-1 and bcl-x than for bax. bcl-2 positivity was observed in 7 of 35 mesotheliomas (20%; Fig. 2A). The immunoreactivity was strong in three, moderate in one, and weak in three cases. In contrast, mcl-1 positivity was also seen in all cases, and staining was very strong in 10 of 34 (29%; Fig. 2B). bcl-x positivity was seen in all cases, and the staining was very strong or strong in 33 of 35 (94%; Fig. 2C). bax positivity was also seen in all cases, and strong or very strong immunoreactivity was seen in 12 of 35 (34%) cases (Fig. 2D). Malignant mesotheliomas more often expressed weak mcl-1 immunoreactivity than metastatic adenocarcinomas (P 0.01). There was no difference in the expression of bcl-2, bcl-x, or bax between malignant mesotheliomas and metastatic adenocarcinomas (P 0.21, P 0.35, and P 0.82, respectively). The apoptotic index in mesotheliomas with no bcl-2 expression was significantly higher ( %) than in cases with bcl-2 expression ( %; P 0.014). The apoptotic index did not significantly associate with the expression of mcl-1, bcl-x, or bax (P 0.25, P 0.19, and P 0.46, respectively). The mean mitotic index in malignant mesotheliomas was %, and the mean extent of necrosis %. There was a strong positive correlation between the apoptotic and the mitotic index (r , P 0.001) and between the apoptotic index and necrosis (r 0.425, P 0.011). No correlation was, however, found between mitosis and necrosis (r 0.218, P 0.22). No significant association was found between the expression of bcl-2, mcl-1, bcl-x, or bax and mitosis or necrosis (data not shown). Patients with mesotheliomas showing a high ( 0.75%) apoptotic index had a worse prognosis than patients with a low ( 0.75%) index (P log rank, P Breslow, and P Tarone-Ware; Fig. 3). Interestingly, patients with mesotheliomas showing a high mitotic index ( 0.22%) had a significantly worse prognosis than those with a low ( 0.22%) index (P log rank, P Breslow, and P Tarone-Ware). Patients with cases showing strong necrosis ( 6.4%) tended to have a worse prognosis than cases with weak or no necrosis ( 6.4%; P log rank, P Breslow, and P Tarone-Ware). There was no association with patient survival and the expression of bcl-2 (P 0.63), bax (P 0.27), mcl-1 (P 0.61), or bcl-x (P 0.16) in mesotheliomas. Because tumor cell lines have been used in the studies of tumor biology and in the assessment of apoptosis, additional Fig. 3 The apoptotic index is significantly associated with survival in malignant mesothelioma. Patients with tumors showing 0.75% apoptotic index had a worse prognosis (P log rank, P Breslow, P Tarone Ware). experiments were conducted by using seven malignant mesothelioma cell lines in culture. Met5A mesothelial cells, which were used as a model of nonmalignant mesothelial cells, are nonmalignant and diploid and have been widely used as a model to investigate mesothelial cells in vitro (31). Western blotting indicated that all seven mesothelioma cell lines as well as Met5A cells were negative for bcl-2 (Fig. 4A). In this experiment, human myeloid leukemic cell line (OY-AML8) was used as a positive control. All of the cell lines showed positive staining for mcl-1, bcl-x, and bax, although the reactivity was variable. Met5A cells showed positive immunoreactivity for mcl-1, bcl-x, and bax (Fig. 4, B D). DISCUSSION This study was undertaken to investigate the extent of apoptosis and the expression of apoptosis regulating proteins bcl-2, mcl-1, bcl-x, and bax in malignant mesotheliomas. Additionally, tumor proliferation, as determined by the mitotic index and necrosis, was assessed from the tumor sections. The average extent of apoptosis in mesotheliomas was 1.07%, suggesting a moderate rate of programmed cell death compared with other malignant tumors, which usually show an apoptotic index ranging between % (reviewed in Ref. 36). The frequency of bcl-2 expression in malignant mesotheliomas was low with only seven tumors (20%) showing expression. Furthermore, all seven mesothelioma cell lines were negative in the immunoblotting experiments. This is in line with previous studies on mesotheliomas and also on cell line studies where infrequent or no bcl-2 expression was found (21 23). In other tumor types, variable bcl-2 expression has been found. Estrogen-dependent carcinomas, such as breast or endometrial
6 Clinical Cancer Research 3513 Fig. 4 A, Western blot analysis showing negative bcl-2 immunoreactivity in Met5A mesothelial cells and in all mesothelioma cell lines (M10K, M14K, M24K, M25K, M28K, M33K, and M38K). A human myeloid leukemic cell line (OY-AML8) was used as a positive control for bcl-2. Protein (75 g) was loaded for the electrophoresis, and the loading homogeneity was confirmed by -actin. B, Western blot analysis showing mcl-1 immunoreactivity in Met5A mesothelial cells and in all mesothelioma cell lines (M10K, M14K, M24K, M25K, M28K, M33K, and M38K). Protein (75 g) was loaded for the electrophoresis, and the loading homogeneity was confirmed by -actin. C, Western blot analysis showing bcl-x immunoreactivity in Met5A mesothelial cells and in all mesothelioma cell lines (M10K, M14K, M24K, M25K, M28K, M33K, and M38K). Protein (75 g) was loaded for the electrophoresis, and the loading homogeneity was confirmed by -actin. D, Western blot analysis showing variable bax immunoreactivity in Met5A mesothelial cells and in the mesothelioma cell lines (M10K, M14K, M24K, M25K, M28K, M33K, and M38K). Protein (75 g) was loaded for the electrophoresis, and the loading homogeneity was confirmed by -actin. tumors, display a high bcl-2 expression, and in breast tumors, bcl-2 expression is associated with a positive estrogen receptor status and a low extent of apoptosis (12, 13, 37, 38). On the other hand, tumors of some other sites, such as non-small cell lung carcinomas and liver and pancreatic carcinomas, show a low frequency of bcl-2 expression (14, 16, 32). The difference in bcl-2 expression reflects differences in the expression of this protein in different cell lines from which the tumors originate. Interestingly, tumors and tissues showing a low frequency of bcl-2 expression usually show a high frequency of mcl-1 expression, and there appears to be an inverse association between these two antiapoptotic proteins, both in nonneoplastic and neoplastic tissues (15 17). Similarly, all malignant mesotheliomas and mesothelioma cell lines showed a high frequency of mcl-1 expression, suggesting that mcl-1 might partly contribute to inhibition of apoptosis in them. The other members of the bcl-2 group were abundantly expressed in malignant mesothelioma, and they were also positive in all mesothelioma cell lines investigated. bax is more or less constitutively expressed in various types of tumors and does not show such variability as bcl-2 (15, 16, 18 20). In our immunohistochemical analysis and cell line studies, malignant mesotheliomas were shown to abundantly express bcl-x, suggesting that bcl-xl and its truncated form bcl-xs take part in modulation of apoptosis in malignant mesotheliomas. There are relatively few studies on bcl-x expression in other tumors. Its expression is, however, high in some carcinomas, such as those originating from prostate or stomach (15, 18). When comparing apoptosis with the expression of apoptosis regulating proteins of the bcl-2 group, mesotheliomas showing bcl-2 expression had a significantly lower apoptotic index than other tumors. This is well in line with the antiapoptotic function of bcl-2 and is in accordance with previous studies on some other tumors, such as pancreatic and breast carcinomas and salivary gland tumors, or malignant non-hodgkin s lymphomas, which also show a similar association (12, 16, 19, 20). The intensity of the expression of other proteins of the bcl-2 group did not significantly associate with the apoptotic index which, however, does not implicate that they would not participate in apoptosis regulation in malignant mesotheliomas. A high apoptotic index in malignant mesotheliomas was associated with a worse prognosis of the patients. This is in line
7 3514 Apoptosis in Malignant Mesothelioma with some studies on lung and breast carcinomas showing a similar association between survival and apoptosis (36). The effect is, however, paradoxical, because one would expect that tumors with less apoptosis would have a growth advantage and a worse prognosis. The reason for this result might be attributable to an association between apoptosis and proliferation. In several studies, it has been shown that apoptosis is significantly associated with proliferation, but the mechanistic link between this association is unknown (36). Indeed, in our material, there was a strong correlation between apoptotic and mitotic index and patients with tumors harboring a high mitotic index also had a significantly worse prognosis. Not all studies, however, show a positive association between cell proliferation and apoptosis. In our study on hepatocellular carcinomas, for instance, such an association was not found (32). A common problem in mesothelioma diagnosis is its differentiation from pleural metastatic adenocarcinomas. To study differences in apoptosis and the expression of the proteins of the bcl-2 family, we also included a set of metastatic pleural adenocarcinomas in the investigation. On average, metastatic adenocarcinomas had a higher extent of apoptosis compared with malignant mesotheliomas, but the difference did not reach statistical significance. The difference might, however, reflect differences in apoptosis regulation and expression of apoptosis regulating proteins between mesotheliomas and carcinomas. As an example of this, malignant mesotheliomas showed more often weak mcl-1 immunoreactivity compared with metastatic adenocarcinomas. Expression of bcl-2, bax, mcl-1, or bcl-x did not, however, differ between these two groups in such a way that expression of these proteins might be regarded of any help in their differential diagnosis. In conclusion, the results show that bcl-x, mcl-1, and bax are abundantly expressed in malignant mesotheliomas, malignant mesothelioma cell lines, and nonneoplastic mesothelial epithelium, whereas bcl-2 is expressed only in a minority of the tumors but not in the nonneoplastic mesothelium. When expressed, bcl-2, however, inhibits apoptosis, as shown by a significantly lower level of apoptosis in bcl-2-positive malignant mesotheliomas compared with other cases. Strong apoptosis was associated with a worse prognosis of the patients. Neither apoptosis nor the expression of the bcl-2 family proteins studied here are of any help in the differential diagnosis between malignant mesotheliomas and metastatic pleural adenocarcinomas. ACKNOWLEDGMENTS The skillful technical assistance of Raija Sirviö and Manu Toivonen is greatly acknowledged. REFERENCES 1. Walz, R., and Koch, H. K. Malignant pleural mesothelioma: some aspects of epidemiology, differential diagnosis, and prognosis. Histological and immunohistochemical evaluation and follow-up of mesotheliomas diagnosed from 1964 to January Pathol. Res. Pract., 186: , Mossman, B. T., Kamp, D. W., and Weitzman, S. A. Mechanisms of carcinogenesis and clinical features of asbestos-associated cancers. Cancer Investig., 14: , Broaddus, V. C., Yang, I., Scavo, L. M., Ernst, J. D., and Boylan, A. M. Asbestos induces apoptosis of human and rabbit pleural mesothelial cells via reactive oxygen species. J. Clin. Investig., 98: , Broaddus, V. C., Yang, I., Scavo, L. M., Ernst, J. D., and Boylan, A. M. Crocidolite asbestos induces apoptosis of pleural mesothelial cells: role of reactive oxygen species and poly(adp-ribosyl) polymerase. Environ. Health Perspect., 105: , Adams, J. M., and Cory, S. The Bcl-2 protein family: arbiters of cell survival. Science (Washington DC), 281: , Thornberry, N. A., and Lazebnik, Y. Caspases. Enemies within. Science (Washington DC), 281: , Yang, E., and Korsmeyer, S. J. Molecular thanatopsis: a discourse on the BCL2 family and cell death. Blood, 88: , Green, D. R., and Reed, J. C. Mitochondria and apoptosis. Science (Washington DC), 281: , Yang, J., Liu, X., Bhalla, K., Kim, C. N., Ibrado, A. M., Cai, J., Peng, T-I., Jones, D. P., and Wang, X. Prevention of apoptosis by bcl-2: release of cytochrome c from mitochondria blocked. Science (Washington DC), 275: , Kluck, R. M., Bossy-Wetzel, E., Green, D. R., and Newmeyer, D. D. The release of cytochrome c from mitochondria: a primary site for bcl-2 regulation of apoptosis. Science (Washington DC), 275: , Manon, S., Chaudhuri, B., and Guérin, M. Release of cytochrome c and decrease of cytochrome c oxidase in bax-expressing yeast cells, and prevention of these effects by coexpression of bcl-x L. FEBS Lett., 415: 29 32, Mustonen, M., Raunio, H., Pääkkö, P., and Soini Y. The extent of apoptosis is inversely associated with bcl-2 expression in premalignant and malignant breast lesions. Histopathology, 31: , Saegusa, M., Kamata, Y., Isono, M., and Okayasu, I. Bcl-2 expression is correlated with a low apoptotic index and associated with progesterone receptor immunoreactivity in endometrial carcinomas. J. Pathol., 180: , Törmänen, U., Eerola, A-K., Rainio, P., Vähäkangas, K., Soini, Y., Sormunen, R., Bloigu, R., Lehto, V-P., and Pääkkö, P. Enhanced apoptosis predicts shortened survival in non-small cell lung carcinoma. Cancer Res., 55: , Krajewska, M., Krajewski, S., Epstein, J. I., Shabaik, A., Sauvageot, J., Song, K., Kitada, S., and Reed, J. C. Immunohistochemical analysis of bcl-2, bax, bcl-x and mcl-1 expression in prostate cancers. Am. J. Pathol., 148: , Virkajärvi, N., Pääkkö, P., and Soini, Y. Apoptotic index and apoptosis influencing proteins bcl-2, mcl-1, bax and caspases 3, 6 and 8 in pancreatic carcinoma. Histopathology, 33: , Krajewski, S., Bodrug, S., Krajewska, M., Shabaik, A., Gascoyne, R., Berean, K., and Reed, J. C. Immunohistochemical analysis of Mcl-1 protein in human tissues. Differential regulation of Mcl-1 and Bcl-2 protein production suggests a unique role for Mcl-1 in control of programmed cell death in vivo. Am. J. Pathol., 146: , Krajewska, M., Fenoglio-Preiser, C. M., Krajewski, S., Song, K., MacDonald, J. S., Stemmerman, J. S., and Reed, J. C. Immunohistochemical analysis of bcl-2 family proteins in adenocarcinomas of the stomach. Am. J. Pathol., 149: , Soini, Y., Raunio, H., and Pääkkö, P. High grade malignant non- Hodgkin s lymphomas differ from low grade lymphomas in the extent of apoptosis and their expression of bcl-2, mcl-1, bax and p53. Tumor Biol., 19: , Soini, Y., Törmänen, U., and Pääkkö, P. Apoptosis is inversely related to bcl-2 but not to bax expression in salivary gland tumours. Histopathology, 32: 28 34, Chilosi, M., Facchetti, F., Dei Tos, A. P., Lestani, M., Morassi, M. L., Martignoni, G., Sorio, C., Benedetti, A., Morelli, L., Doglioni, C., Barberis, M., Menestrina, F., and Viale, G. Bcl-2 expression in pleural and extrapleural solitary fibrous tumours. J. Pathol., 181: , Segers, K., Ramael, M., Singh, S. K., Weyler, J., Van Meerbeeck, J., Vermeire, P., and Van Marck, E. Immunoreactivity for bcl-2 protein