6.2. Studies on cultured cells
6.2.1. Expression of antioxidant enzymes
6.2.1.1. MnSOD (I-IV)
First, the expression of MnSOD in non-malignant mesothelial cells (Met5A) was compared to the reactivities in two malignant mesothelioma cells (M14K and M38K) (I). In Met5A cells, immunoreactivity for MnSOD was weak or negative, except in a few cell clusters, where it was strong (Fig. 6a,b). Both mesothelioma tumour cell lines showed strong immunoreactivity for MnSOD, and the proportion of positive cells was much higher and the staining more intense than in Met5A cells (Fig. 6c). M38K cells showed an intense, granular staining pattern, which was similar to that of the tissue from which the cell line had originally been established (Fig. 6d). The MnSOD activity was significantly lower in Met5A cells than in M14K cells (p=0.004) or in M38K cells (p=0.001, n=8). Fig. 7 shows a representative Western blot (a), Northern blot (b), and RT-PCR (c) for MnSOD in Met5A, M14K and M38K cells. The MnSOD activity, protein and mRNA levels were highest in mesothelioma M38K cells.
In further experiments comparing the expression of MnSOD in mesothelioma and adenocarcinoma (II) the MnSOD protein expression of seven mesothelioma cell lines was compared to that of non-malignant Met5A cells and A549 lung adenocarcinoma cells. Higher MnSOD expression was observed by Western blot in all mesothelioma cell lines than in Met5A and A549 cells (II).
Figure 6. MnSOD immunostaining of non-malignant mesothelial Met5A cells is low (a), except in the cell clusters occasionally formed during cell culture, where it is strong (b). Mesothelioma cell line M38K shows strong MnSOD immunostaining (c), and a similar strong granular staining is seen in M38 tissue from which the cell line was originally established (d). Magnifications 400x (a,c,d) and 100x (b).
6.2.1.2. Other antioxidant enzymes (I)
The expressions of CuZnSOD, catalase, GPx and GR were assessed in non-malignant Met5A mesothelial cells and in mesothelioma M14K and M38K cells. The specific activity of CuZnSOD was higher in M38K cells than in Met5A cells (53 ± 17 and 30.3 ± 15.7 U/mg protein, respectively, p=0.015, n=8), whereas the activities in Met5A and M14K cells did not differ significantly (33.1 ± 10.6 U/mg protein in M14K cells). The mRNA level of CuZnSOD mRNA was also higher in M38K cells than in Met5A cells (p=0.017, I).
The activity of catalase was higher in mesothelioma M38K cells than in non-malignant Met5A cells (p=0.012), whereas in mesothelioma M14K cells the activity was lower (p=0.036) than in Met5A cells. The catalase mRNA level was also higher in M38K cells than in Met5A cells (p=0.048) or M14K cells (p=0.038). The mRNA level of glutathione peroxidase did not differ significantly between the three cell lines (data not shown). However, there was a tendency for higher GPx and GR activities in M38K cells compared to Met5A cells (Table 6).
6.2.2. MnSOD expression in the mitochondria of mesothelioma cells
6.2.2.1. Location of MnSOD in mitochondria of mesothelioma cells (III, unpublished data)
To verify that high MnSOD in mesothelioma cells is related to an elevated level of this enzyme in the mitochondria, and not to an increased volume or number of mitochondria per cell, the activities of two constitutive mitochondrial enzymes were measured. The activity of citrate synthase was 45.9 ± 5.3 and 32.7 ± 10.7 nmol/min/mg protein in mesothelioma M38K cells and in non-malignant Met5A cells, respectively (p=0.13), the corresponding activity for cytochrome χ oxidase being 3.2 ± 1.0 and 2.4 ± 0.9 nmol/min/mg protein (p=0.34). When the MnSOD activities were related to these mitochondrial enzymes, MnSOD activity was about six-fold in M38K cells compared to Met5A cells (p<0.001), suggesting higher MnSOD activity per mitochondrion in mesothelioma cells.
Since high mitochondrial MnSOD may be associated with mitochondrial changes (Oberley & Oberley 1997), the morphology of these two cell types was compared by transmission electron microscopy. Mitochondria of mesothelioma cells were dense and abundant and the cristae were clumsy or hardly recognisable compared to mesothelial cells (Fig. 8a,b, unpublished data). Endoplasmic reticulum was often abundant in the vicinity of mitochondria. The mitochondria did not differ in size between the two cell lines. MnSOD expression by immunogold staining was clearly more intense in the mitochondria of M38K cells than Met5A cells, and no extramitochondrial staining was seen (Fig. 8c,d, unpublished data).
6.2.2.2. Mitochondrial superoxide levels (III)
The high MnSOD in the mitochondria of mesothelioma M38K cells might affect the oxidant levels in these cell organelles. Superoxide generation from isolated mitochondria was assessed by lucigenin chemiluminescence, which showed significantly lower superoxide production in M38K than Met5A cells. When standardised with xanthine oxidase as described before (Pitkänen & Robinson 1996), mitochondrial superoxide production was 2.37 ± 0.16 and 8.68 ± 0.61 nmol/min/mg protein in M38K and Met5A cells, respectively (p<0.001).
6.2.3. Oxidant levels in intact non-malignant and malignant cells (III)
The oxidant production of intact mesothelial Met5A and mesothelioma M38K cells was next assessed by the DCDHF-DA method. The fluorescence intensity standardised against the cell protein content increased during the culture, and the intensities did not differ between Met5A and M38K cells, indicating that there was no difference in the oxidant stress despite the difference on the mitochondrial level. In order to evaluate the significance of catalase and glutathione in the oxidant production of non-malignant Met5A cells
Figure 8. Transmission electron microscopy showing fairly normal mitochondria with visible cristae in non-malignant Met5A cells (a), whereas in mesothelioma cells M38K (b) mitochondria are dense and crista structure clumsy. Immunoelectron microscopy for MnSOD shows that MnSOD labelling is restricted to mitochondria, and that MnSOD staining per mitochondria is much more intense in M38K cells (d) than in Met5A cells (c) Magnifications 51250 x (a,b), 63750 x (c), 82500 x (d).
and malignant mesothelioma M38K cells, selected experiments were conducted in cells pretreated with ATZ and BSO. Catalase inhibition caused 213% and 244% enhancement of fluorescence in Met5A and M38K cells, respectively. The corresponding increases in DCDHF fluorescence after glutathione depletion were 116% and 157%, respectively. An exogenous superoxide-producing agent, menadione (10 µM), caused only 59% and 68% increases in the fluorescence of Met5A and M38K cells, respectively. The oxidant production was, however, markedly enhanced in both cell types when catalase was inhibited prior to menadione exposure, whereas the effect of glutathione depletion was less intense and only significant in M38K cells.
6.2.4. Cytotoxicity assays (I)
To assess the role of the levels of MnSOD in the oxidant and drug resistance, non-malignant Met5A (low MnSOD expression), mesothelioma M14K (moderate MnSOD expression) and mesothelioma M38K (high MnSOD expression) cells were exposed to different concentrations of menadione and epirubicin. The depletion of high-energy nucleotides as a marker of early cell injury was significantly higher in Met5A cells than in mesothelioma cells, M38K cells showing the highest resistance (Fig. 9a). Menadione at a concentration of 25 µM for 4 h also caused significant lytic cell injury in Met5A and M14K cells, but not in M38K cells, when assessed by LDH release (p<0.05, M38K compared to Met5A cells, I). Mesothelioma cells and non-malignant mesothelial cells were also exposed to epirubicin, an anthracyclin used to treat mesothelioma. Epirubicin (0.1 and 0.2 µg/ml for 48 h) caused a significant cell injury in Met5A cells, but not in M38K cells when assessed by measuring nucleotide depletion or LDH release. (Fig. 9b).
6.2.5. Proliferative and apoptotic responses of mesothelioma cells to epirubicin (IV)
Given that M38K cells, with their highest MnSOD and catalase levels, were found to be more resistant to menadione and epirubicin than M14K cells and Met5A cells with less MnSOD and catalase, this resistance might be related to the resistance of mesothelioma cells to apoptosis. Therefore, the effect of epirubicin on apoptosis and proliferation was also compared in the mesothelioma cells M14K (lower MnSOD) and M38K (higher MnSOD). The higher resistance against epirubicin-induced toxicity of M38K cells was demonstrated by the XTT method (IV). The spontaneous apoptotic indices without exposure were 3.1% and 2.5% in M14K and M38K cells, respectively. Epirubicin induced apoptosis in a dose-dependent manner in M14K cells but not in M38K cells when assayed by the TUNEL-method and by detecting caspase-3 activation (Fig. 10). The proportion of spontaneously proliferating cells was 46% and 51% in M14K and M38K cells, respectively. The extent of proliferation was reduced by epirubicin exposure in M38K cells in a dose-dependent manner, whereas the proliferative activity in M14K cells was increased with low doses of epirubicin (IV).
Figure 10. Epirubicin (0.1-1.0 µg/ml 48 hours) induced apoptosis in M14K mesothelioma cells with low MnSOD and M38K mesothelioma cells with high MnSOD. Apoptosis was induced dose-dependently in M14K cells assessed by TUNEL-method (a) and by caspase 3 activation (b), but not in M38K cells (a,c). Caspase 3 proenzyme (32 kD) is cleaved into 17 kD (detected by the antibody used) and 10 kD fragments when activated.