| The expression and possible role of manganese superoxide dismutase in malignant pleural mesothelioma | ||
|---|---|---|
| Prev | Chapter 2. Review of the literature | Next |
Carcinogenesis is a multistep process associated with accumulated genetic alterations in somatic cells (Caldas 1998). Tumour initiation begins through carcinogen-induced mutations, and the initiated cells may acquire a selective growth advantage as a result of mutations in genes that control cellular proliferation and death. During the tumour promotion phase, the initiated cells further expand clonally, and additional genetic damage develops through several mutations. These mutations may include the activation of proto-oncogenes or the inactivation of tumour suppressor genes (Shields & Harris 1993, Caldas 1998). Progressive phenotypic changes and genomic instability continue during the phases of tumour progression and malignant conversion (Shields & Harris 1993). It has been estimated that a normal cell must acquire at least five or six mutations to become a cancer cell (Caldas 1998). In the case of mesothelioma, several mutations are probably required for the malignant conversion, because the latency period between the asbestos exposure and the appearance of the tumour is usually very long (Bell & Taste 1997). ROS may participate in the multistage carcinogenesis from initiation to malignant conversion by causing oxidative DNA damage and mutations in proto-oncogenes and tumour suppressor genes, and by activating signal transduction pathways (Kensler et al. 1983, Guyton & Kensler 1993, Cerutti 1994, Irani et al. 1997).
Disturbances in the regulation of cell proliferation and apoptosis are important during tumourigenesis. Disruption of the apoptotic pathway in tumour cells confers a survival advantage and probably contributes to treatment resistance (Fisher 1994). The p53 tumour suppressor gene and the bcl-2 proto-oncogene are the most extensively studied oncogenes known to modulate cancer-related apoptosis. Wild-type p53 acts as a gatekeeper capable of inducing cell cycle arrest and apoptosis in response to DNA damage (Levine 1991). Mutations in the p53 gene may result in uncontrolled proliferation and decreased apoptosis of mutated cells (Amundson et al. 1998). p53-deficient mice are susceptible to asbestos-induced DNA damage (Marsella et al. 1997). However, p53 mutations are infrequent in mesothelioma (detected in about 29% of the cases), compared to many other malignancies (Metcalf et al. 1992, Ramael et al. 1992a, Bell & Testa 1997). Bcl-2 family proteins are also critical in regulating apoptosis, acting as either inhibitors or promoters of cell death. Bcl-2 proto-oncogene prevents apoptosis and is suggested to function as an antioxidant, probably by preventing early mitochondrial apoptotic events and subsequent superoxide release (Hockenbery et al. 1993, Kroemer 1997, Cai & Jones 1998). Mesothelioma cells were recently found to be resistant to apoptosis, but the resistance could not be explained by bcl-2 overexpression (Narasimhan et al. 1998). No specific chromosomal abnormalities have been found in mesothelioma. However, several DNA copy number changes have been detected, involving 1q, 1p, 3p, 4q, 6q, 9p, 13q, 14q, and 22, and indicating that these sites could include important tumour suppressor genes for mesothelioma (Björkqvist et al. 1997, Bell & Testa 1997).
Most human cancers are resistant to chemotherapeutic drugs at presentation or become resistant after an initial partial response (Harrison 1995). There are several possible molecular mechanisms for drug resistance which may co-exist in vivo: exclusion of the drug from the cell, increased detoxification of the drug, inability to enter apoptosis, enhanced repair of drug-induced DNA damage, failure to activate drug, and changes in drug target. P-glycoprotein, which confers drug resistance on tumour cells, is found in the majority of mesothelioma cases but not in non-malignant mesothelium, indicating a role for this efflux protein in the drug resistance of mesothelioma (Ramael et al. 1992b). However, the mechanisms of resistance to drugs and apoptosis in mesothelioma are still poorly understood.
For the past 20 years, MnSOD activity has been postulated to be low in malignant tumours (Oberley & Buettner 1979, Sun 1990). The lowered antioxidant capacity and the oxidant-antioxidant imbalance have been considered to play a role in multistage carcinogenesis (Oberley & Oberley 1997). However, most of the early studies were conducted on cell lines derived from experimental animals (Oberley & Buettner 1979). Conclusions have also been drawn from activity measurements taken from blood samples, which do not reflect the enzyme level of the tumour cells. Furthermore, in many studies enzyme activities were measured from tissue homogenates, which often contain various cell types and do not represent a specific cancer cell type (reviewed by Sun 1990). Reliable comparisons include an assessment of the enzyme in a human cancer cell or tissue and in its non-malignant counterpart, which is the parental cell or tissue from which the tumour originates. Several studies showing lower MnSOD levels in cancer cells have been performed using in vitro transformed cell lines rather than human tumour-derived cells (Oberley & Oberley 1997). Table 1 summarises the studies on MnSOD expression in cancer cells that have been conducted on human tumours by comparing the specific cancer cell type with a relevant control cell type. According to these studies, MnSOD expression is variable but often high in human tumours compared to their normal counterparts.
The levels of the other antioxidant enzymes have been highly variable, but the CuZnSOD and catalase activities have often been low (Sun 1990). Most of these studies, however, have been conducted on homogenised tumour tissues, and only a few reliable comparisons to the parental cell type are available. CuZnSOD activity was higher in Wilms´ tumour tissue compared to adjacent normal tissue (Gajewska et al. 1996), but lower in hepatocellular carcinoma than in normal liver cells (Liaw et al. 1997). Punnonen et al. (1994) investigated cancerous and non-cancerous tissue samples from 23 patients with breast cancer and found that the CuZnSOD and GPx activities were higher in cancer tissue, whereas catalase activity was lower. In a study by Jaruga et al. (1994), the total SOD, catalase and GPx activities were lower in five samples of lung cancer tissue compared to the surrounding normal tissue. In lung cancer, especially in adenocarcinoma, GPx activity has been higher than in normal tissue (Di Ilio et al. 1987, Carmichael et al. 1988). Coursin et al. (1996) investigated 19 samples of human lung cancer and 4 samples of normal human lung and found low or negative immunoreactivity of GPx in all cancer types, including adenocarcinoma. Catalase was negative for the neoplastic cells of all tumour samples, whereas CuZnSOD showed variable immunostaining. GST-π has been up-regulated in most of the malignancies thus far analysed (Tew 1994). GPx or combined GPx and SOD overexpression in transgenic mice has been shown to increase rather than decrease skin susceptibility to carcinogen-induced tumours (Lu et al. 1997).
Table 1. A summary of the studies on MnSOD expression in human tumours with a relevant control cell. Studies with less than five tumour samples were excluded.
| Reference | Tumour studied/ control | MnSOD detection method | Main result | Conclusion/ notes |
|---|---|---|---|---|
| Nishida et al. 1992 | Thyroid tumours/ normal thyroid tissue | IH*, ELISA | MnSOD protein higher in papillary and follicular (but not in anaplastic) carcinoma compared to adjacent normal tissue | MnSOD variable according to the differentiation degree |
| Oberley et al. 1994 | Renal cell carcinoma/ normal kidney proximal tubule | Immunogold, activity, WB# | Clear cell type expressed lower and granular cell type higher MnSOD level than normal proximal tubule | MnSOD expression variable depending on the cell type of the tumour |
| Coursin et al. 1996 | Lung cancer tissue/ normal lung tissue | IH | Adenocarcinoma variable (from negative to highly positive), squamous cell, small cell and large cell carcinomas negative | MnSOD mostly negative in lung cancer, but adenocarcinoma variable |
| Cobbs et al. 1996 | Brain tumour/ normal brain tissue | IH | MnSOD negative in normal brain, strongly positive in high grade tumours | MnSOD high in brain tumours compared to normal brain |
| Landriscina et al. 1996 | Brain tumours/ normal brain tissue | IH, WB, NB**, activity | High MnSOD associated with more aggressive disease | MnSOD higher in brain tumours |
| Van Driel et al. 1997 | Colorectal cancer/ adjacent normal mucosa | IH, activity | MnSOD expression lower in cancer tissue | MnSOD low in colorectal cancer |
| Baker et al. 1997 | Prostatic adenocarcinoma/ normal prostate | IH | Variable expression | Malignant prostate epithelium may have lowered expression of antioxidant enzymes |
| Izutani et al. 1998 | Oesophageal and gastric cancer | IH, PCR | MnSOD staining intensity higher in cancer tissue than in non-cancer tissue | Oesophageal and gastric cancers express high MnSOD |
| Janssen et al. 1998 | Colorectal cancer; 163 patients / normal mucosa | IH | Higher MnSOD (but not CuZnSOD) in colorectal cancer than in non-cancer tissue | High MnSOD in colorectal cancer; the high level associated with poor prognosis |
| * IH; immunohistochemistry, # WB; Western blot, **NB; Northern blot | ||||
In conclusion, the results on the expression of antioxidant enzyme expression in cancer are highly variable, and particularly the role of MnSOD may be more complicated than previously suggested.
Tumour cells generate superoxide and other ROS, and this generation, if it also occurs in vivo, might have effects on tumour cell proliferation and invasion (Bize et al. 1980, Shaughnessy et al. 1989, Szatrowski & Nathan 1991). It has been hypothesised that the production of ROS combined to a decreased antioxidant enzyme level may be characteristic of tumour cells (Toyokuni et al. 1995, Oberley & Oberley 1997).
There is a lot of experimental data to suggest that the malignant phenotype of a cancer cell can be suppressed by raising the MnSOD level of the cell (Oberley & Oberley 1997), and it has been hypothesised that the MnSOD gene is a tumour suppressor (Bravard et al. 1992, Li et al. 1995a). Tumour suppressor gene products are negative growth regulators, and the loss of their function results in the expression of a transformed phenotype (Marshall 1991). The data on the possible role of MnSOD as a tumour suppressor is based on several studies with cells transfected with the MnSOD gene to increase MnSOD activity (Oberley & Oberley 1997). Overexpression of MnSOD by gene transfection has been shown to suppress radiation-induced neoplastic transformation of mouse embryonic fibroblasts (St Clair et al. 1992), the metastatic capacity of mouse fibrosarcoma cells (Safford et al. 1994) and human breast cancer cells (Li et al. 1995a), and the growth rate of human melanoma cells, human transformed fibroblasts, rat glioma cells, hamster cheek pouch carcinoma cells, and human prostate carcinoma cells (Church et al. 1993, Yan et al. 1996, Zhong et al. 1996, Lam et al. 1997, Li et al. 1998a). Inhibition of tumour growth has been attributed to decreased cell division rather than increased cell death (Oberley & Oberley 1997). One explanation for these growth-suppressive effects of MnSOD transfection could be the resulting imbalance of antioxidant enzymes that favours H2O2 accumulation (Zhong et al. 1996, Li et al. 1998b). Other possible mechanisms for the suggested tumour-suppressive activity of MnSOD overexpression could be modulation of the specific oncogenes (Kiningham & St Clair 1998), up-regulation of protease inhibitors (Li et al. 1998c), and inhibition of the transcription factors AP-1 and NF-κ B (Li et al. 1998b).
There are only a few studies on the structure of the MnSOD gene in cancers. Human colon carcinoma-derived MnSOD complementary DNA (cDNA) showed no defects in the translated regions (St Clair & Holland 1991). Ambrosone et al. (1999) were the first to show the polymorphism of the signal sequence of MnSOD to be a risk factor for human breast cancer. Impaired mitochondrial targeting of MnSOD in tumour cells could lead to decreased activity of this enzyme. Mesothelioma, similarly to many other tumours, has been reported to exhibit frequent genomic losses involving chromosome arm 6q (Bell et al. 1997, Bell & Testa 1997, Björkqvist et al. 1997). It has also been suggested that tumour suppressor genes might be located in this region. The MnSOD gene is located in 6q25, but there are no previous studies on the possible role of MnSOD in the pathogenesis of mesothelioma.
The role of MnSOD as a tumour suppressor is still controversial. As previously reviewed, MnSOD expression is high in many human cancers, the level of MnSOD directly correlates with the grade of some tumours (Landriscina et al. 1996, Janssen et al. 1998), and high MnSOD is associated with poor prognosis in some tumours (Nakano et al. 1996, Janssen et al. 1998). In addition, if MnSOD is really anti-apoptotic, as it has been observed (Manna et al. 1998, Kinscherf et al. 1998), MnSOD overexpression in tumours could offer a survival advantage to a tumour cell and lead to treatment resistance.
Several anticancer drugs lead to the generation of ROS during enzymatic activation, which partly explains the cytotoxic effects of cancer therapy (Sinha & Mimnaugh 1990). Anthracyclines are antibiotic anticancer drugs, which are redox-active because of their quinone-hydroquinone structure. (Bachur et al. 1978, Powis 1989). Daunorubicin, doxorubicin and epirubicin are the most common anthracycline drugs. Menadione, mitomycin C and mitoxantrone are also anticancer quinones and capable of inducing ROS formation through redox cycling reactions (Powis 1989). Anthracyclines are widely used to treat various malignancies, including mesothelioma, but their clinical efficacy and use are often limited by the development of drug resistance and side-effects, cardiac toxicity being the most severe one (Maeda et al. 1994, Yen et al. 1996). The toxicity of these drugs is usually attributed to the formation of free radicals, and drug resistance may also be partly due to the enhanced antioxidant capacity of tumour cells (Sinha & Mimnaugh 1990).
Anthracycline-induced cytotoxicity is ameliorated in vitro by adding SOD and catalase or glutathione peroxidase to the culture medium (Doroshow 1986, Sinha et al. 1987). Yen et al. (1996) demonstrated that MnSOD overexpression prevents adriamycin-induced cardiac toxicity in transgenic mice and particularly mitochondrial damage, which is probably critical in anthracycline-induced cardiac injury. MnSOD overexpression may also protect tumour cells against cytotoxicity induced by doxorubicin, TNF-α and IL-1 (Hirose et al. 1993). Resistance to TNF is probably important in the development and growth of malignant tumours, and induction of MnSOD is thought to play a critical role in the development of TNF resistance (Wong et al. 1989). It has been shown that adriamycin may induce TNF resistance at least partly by overexpression of endogenous TNF and MnSOD genes (Maeda et al. 1994), while endogenous TNF may protect tumour cells against adriamycin cytotoxicity by increasing the MnSOD level of the cell (Zyad et al. 1994).
Anticancer drugs induce the activity of glutathione-related enzymes (GST, GPx, glutathione reductase, gammaglutamylcysteine synthetase) and catalase, and the activation of these different mechanisms is associated with drug resistance (deVries et al. 1989, Hao et al. 1994, Cheng et al. 1997). Apart from the antioxidant function, the importance of glutathione and related enzymes, especially the GST family, lies in the detoxification of several drugs (Tew 1994). The depletion of cellular GPx or glutathione increases sensitivity to adriamycin (Taylor et al. 1993), and the increase in GPx and GST expression is associated with resistance to adriamycin (Sinha & Mimnaugh 1990). Positive GST-π expression predicts a poor response to cisplatin-based chemotherapy of patients with non-small cell lung carcinoma (Bai et al. 1996). In a study by Ogretmen et al. (1998) co-ordinated overexpression of the multidrug resistance-associated protein (MRP) and γ -glutamylcysteine synthetase genes correlated with the doxorubicin resistance of human mesothelioma cell lines. The importance of multiple antioxidant pathways was demonstrated by Tanaka-Kagawa et al. (1999), who showed that overexpression of CuZnSOD or catalase alone did not affect the sensitivity of HeLa cells to cisplatinum, whereas simultaneous overexpression of these enzymes reduced this sensitivity. Human lung tumour cell lines with acquired doxorubicin resistance did not show any changes in their antioxidant enzyme activities, whereas P-glycoprotein expression was increased compared to the parental cell line (Keizer et al. 1989). Neither the tumour glutathione content nor GST isoenzyme expression was associated with the response to carboplatin in ovarian cancer patients (Ghazal-Aswald et al. 1996).
Ionising irradiation enhances ROS production in a variety of cells, and this effect is associated with radiation-induced cytotoxicity (Sun et al. 1998). Overexpression of MnSOD by transfection protects cells against irradiation-induced toxicity in vitro, whereas the transfection of CuZnSOD or GPx does not (Sun et al. 1998). In addition, radiation-resistant mice have shown higher SOD and catalase activities than radiation-sensitive ones (Hardmeier et al. 1997). Human breast cancer cells transfected to overexpress GPx were not more resistant to radiation or to doxorubicin (Liebmann et al. 1995).
In conclusion, a lot of experimental data exists indicating that ROS are involved in the cytotoxicity of certain anticancer drugs and irradiation, and that increased antioxidant capacity may partly explain the resistance to these treatment modalities.