In order to demonstrate the biological effects of PCBs on cell growth and clonogenic survival, the MCF-10A and RWPE-1 human breast and prostate epithelial cells were incubated continuously with 3 μM PCBs for 5 days. This dose was chosen as representative of the upper limit of doses measured in the circulation of people living in the Anniston, Alabama flood plain which represents a heavily PCB contaminated site in the United States (Dr. Allan Silverstone, State University of New York, personnel communication). Following treatment, cells were trypsinized and cell numbers as well as clonogenic survival were determined. The results in showed that Aroclor 1254 and PCB153 significantly inhibited cell growth whereas 4ClBQ completely halted the cell growth in both breast and prostate epithelial lines. Interestingly in this model system, PCB3 and PCB77 did not significantly alter the rate of cell growth compared to the vehicle control cells (DMSO). Both MCF-10A and RWPE-1 cell doubling times (Td) were calculated by using the equation: Td=0.693t/ln (Nt/N0), where Nt and N0 represent cell number at time t and time zero respectively. The doubling times of MCF-10A and RWPE-1 cell lines treated with various PCBs were as follows (PCB: MCF-10A, RWPE-1). Control: 23 h, 22h; PCB3: 26 h, 23h; PCB77: 30 h, 31h; PCB153: 49 h, 59h; Aroclor 1254: 57 h, 68h. The doubling time in the 4ClBQ group couldn’t be calculated due to the complete suppression of the cell growth. The results of doubling time experiments suggested that, relative to MCF-10A, RWPE-1 cells seem to be more sensitive to the PCB 153 and Aroclor 1254 induced growth inhibition. It also appeared that the RWPE-1 cells treated with PCB153 and Aroclor began to grow slowly in the first 3 days, however, during the last two days, the treatment caused cell growth inhibition and then a decrease the cell number per dish ().
PCB-induced perturbations in MCF-10A cell growth and clonogenic survival
In the next series of experiments the ability of PCB treated MCF-10A or RWPE-1 cells to continue to undergo mitosis was assayed using the clonogenic cell survival assay, which quantifies irreversible cytotoxic responses [24
]. The results of the clonogenic cell survival assays also showed that 5 days exposure to 4ClBQ reduced the plating efficiency of the MCF-10A cells from 30% to 10% () and RWPE-1 cells from 32% to 9% () clearly demonstrating irreversible PCB-induced clonogenic cell killing. Aroclor 1254 and PCB153 also caused significant decreases in clonogenic cell survival in both cell lines but not as dramatic as 4ClBQ ( and ). Furthermore, it again appeared that Aroclor 1254 and PCB153 affected the clonogenic survival of RWPE-1 cells more severely than that of MCF-10A cells, suggesting differential toxicity of various PCBs towards cells from different tissue origins. Consistent with the growth curve results, PCB3 and PCB77 did not show any significant clonogenic cell killing compared to the control cells in both breast and prostate cell lines. Overall, the results in demonstrate that PCBs and their metabolites can disrupt cell growth in human mammary and prostate epithelial cells as well as causing irreversible loss of reproductive integrity.
When MCF-10A and RWPE-1 cells were treated with PCBs as in , steady-state levels of intracellular O2·− (as determined by DHE oxidation) were found to be significantly higher in PCB153 (1.5-fold), Aroclor 1254 (2-fold), and 4ClBQ (4-fold) treated groups, compared to vehicle controls (). Similar increases of the DHE oxidation were also obtained from PCB-exposed RWPE-1 prostate cells (). Since DHE oxidation following PCB exposure was measured in the absence of PCBs and the analysis was done using flow cytometry, where no remaining extra-cellular probe was possible, these data strongly suggest that the increase in DHE oxidation occurred inside the cells. Furthermore, these results were confirmed when DHE oxidation was also found to be significantly higher in PCB153 (8-fold in MCF-10A, 3-fold in RWPE-1), Aroclor (9-fold in MCF-10A, 5-fold in RWPE-1), and 4ClBQ (20-fold in MCF-10A, 9-fold in RWPE-1) treated cells when PEG-SOD inhibitable DHE oxidation was determined, relative to PEG-alone vehicle controls ( and ). Since CuZnSOD is believed to be very specific for O2·−, these results provide compelling evidence that PCB exposure results in significantly increased intracellular steady-state levels of O2·− in human breast and prostate epithelial cells.
Steady-state levels of O2·− in PCB exposed cells as determined by DHE oxidation
Because it was previously suggested that PCBs and their metabolites could affect mitochondrial metabolism [36
], the mitochondrial localization of O2·−
in PCB-treated MCF-10A breast and RWPE-1 prostate cells was determined using an O2·−
sensitive, fluorescent dye that localizes in mitochondria (i.e., MitoSOX™
red). The flow cytometry results in demonstrate that exposures to 3 μM PCBs for 5 days significantly increased intracellular MitoSOX™
red oxidation [PCB153 (1.5-fold), Aroclor (3.5-fold) or 4ClBQ (9-fold)], in MCF-10A cells. In addition, , shows that, relative to MCF-10A, RWPE-1 prostate epithelial cells appeared to have significantly higher steady-state levels of MitoSOX™
red oxidation in PCB153 (6-fold), Aroclor 1254 (10-fold), and 4ClBQ (19-fold) treated groups. To further support the conclusion that these increases in MitoSOX™
red oxidation are truly representative of mitochondrial O2·−
, confocal microscopy was used to visualize the intracellular co-localization of MitoSOX™
red oxidation and MitoTracker Green fluorescence. In this experiment the cells were pre-labeled in the dark with MitoSOX™
red and MitoTracker green for 20 mins to allow the fluorophores to be localized in the mitochondria, the attached cells were then washed with PBS (to remove extra-cellular MitoSOX™
red and MitoTracker green), and the PCBs were added for 1 minute. When this was done with 4ClBQ treated MCF-10A and RWPE-1 cells, the MitoSOX™
red oxidation image ( and ) showed much brighter red color compared to the vehicle control MitoSOX™
red oxidation image ( and ). The overlay of the MitoSOX™
red oxidation image ( and ) with the MitoTracker Green image ( and ) showed nearly exact concordance in staining as evidenced by the yellow to orange color seen in and . In a similar experiment PCB153 and Aroclor also showed mitochondrial localization of MitoSOX™
red oxidation using the same methodology (data not shown) in both cell lines. To identify the signal from MitoSOX™
red oxidation was truly representing mitochondrial superoxide, superoxide dismutase (MnSOD) was over expressed in the cells prior to PCBs exposure using the AdMnSOD vector (A′B′C′ and G′H′I′). and show that the oxidation of MitoSOX™
red oxidation after 4ClBQ exposure, was nearly completely suppressed to control levels by AdMnSOD treatment (compare to and to ). Furthermore, the overlay image ( and ) looked identical to the MitoTracker Green image ( and ) confirming that the increased levels of superoxide being detected did originate from mitochondria (compare to and to ). In a similar experiment PCB153 and Aroclor also showed suppression of MitoSOX™
red oxidation following AdMnSOD treatment (data not shown). These results provide compelling evidence supporting the hypothesis that mitochondria are a major site for increased steady-state levels of O2·−
in exponentially growing MCF-10A and RWPE-1 cells during and following PCB exposure.
Steady-state levels of O2·− as determined by Mitosox oxidation via flow cytometry
Mitochondrial localization of O2·− as determined by Mitosox oxidation and Mitotracker colocalization using confocal microscopy in MCF-10A cells
Once it is generated in the cell, O2·− could either spontaneously (or through the action of superoxide dismutase enzymes) undergo dismutation to form H2O2 which also could adversely affect cell growth and proliferation. To determine if the intracellular steady-state levels of hydroperoxides, particularly H2O2, were also increased in MCF-10A breast and RWPE-1 prostate cells treated with PCBs and their metabolites, cells were labeled with the oxidation-sensitive probe CDCFH2 after 5 days of treatment with 3 μM PCB3, PCB77, PCB153, Aroclor, or 4ClBQ. CDCFH2 crosses cellular membranes and is enzymatically hydrolysed by “intracellular” esterases, and then can be further oxidized to produce a green florescent compound (CDCF) trapped inside of the cell for detection by flow cytometry. In PCB treated MCF-10A cells, there was a significant 3-fold increase in CDCFH2 oxidation () in cells treated with 4ClBQ, however, other PCBs did not show any increase in probe oxidation, compared to the control. When the experiment was repeated using PEG-CAT inhibitable CDCFH2 oxidation as described in the methods section (by determining the differences in fluorescence between PEG-CAT and 18 μM PEG alone treated cells for each treatment group), similar results were obtained clearly demonstrating that most (if not all) of the increased probe oxidation caused by 4ClBQ was mediated by H2O2 (). Interestingly, unlike MCF-10A cells, the steady-state level of hydroperoxides detected using the aforementioned methodology in PCB-exposed RWPE-1 cells were found to be significantly increased in PCB 153 (5-fold), Aroclor 1254 (7-fold) and 4ClBQ (15-fold) treatment groups (). To further confirm PCB-induced increases in steady-state levels of hydroperoxides, the experiments in showed similar increases in PEG-CAT inhibitable CDCFH2 oxidation in PCB exposed RWPE-1 cells. To further confirm that changes in CDCFH2 oxidation were detecting only changes in probe oxidation, the previous experiments were also repeated using the oxidation insensitive analog (CDCF) which measures changes in dye uptake, ester cleavage, and efflux (independent of oxidation). The results showed that there was no significant difference in CDCF labeling between any treatment groups compared to control (DMSO treated) cells (data not shown).
Overall, the data obtained from the studies measuring DHE, MitoSOX™red, and CDCFH2 oxidation (–) clearly demonstrate that treatment with PCBs and their metabolites cause increases in intracellular steady-state levels of ROS (i.e., O2·− and H2O2) in MCF-10A breast and RWPE-1 prostate epithelial cells. In addition, the studies using MitoSOX™red and MnSOD over expression also strongly support the hypothesis that mitochondria represent a major source for increased steady-state levels of O2·− in these PCB-treated human breast and prostate epithelial cell lines.
To determine if PCB-induced increases in steady-state levels of ROS were accompanied by alterations in cellular antioxidants associated with O2·− and H2O2 metabolism, intracellular GSH level, CuZnSOD activity, and MnSOD activity were analyzed. The results in demonstrate significantly decreased levels of total GSH in cells treated daily dose of 3 μM for 3 days with PCB153, Aroclor 1254, or 4ClBQ, relative to vehicle controls. Furthermore, treatment with 3 μM PCB153, Aroclor 1254, or 4ClBQ for 3 days also induced very significant 3–4 fold increases in MnSOD activity (but not CuZnSOD), relative to vehicle controls in MCF-10A cells (). Interestingly, when western blot analysis was performed using antibodies against MnSOD there were no significant changes in immunoreactive protein levels among any treatment or control groups, relative to α-actin loading controls (). However, similar exposures to PCBs were found to slightly decrease the MnSOD activity while inducing CuZnSOD activity (3–5 folds) in RWPE-1 cells ().
Intracellular total GSH is decreased in PCB-exposed human MCF-10A cells.
MnSOD activity is induced in PCB-exposed MCF-10A cells
To determine if the biological effects of PCBs are causally related to PCB-induced increases in steady-state levels of ROS, MCF-10A and RWPE-1 cells were treated with PCBs for 1 hour followed by addition of antioxidant enzymes (i.e. 50 U/mL PEG-SOD and PEG-CAT) or a non-specific thiol antioxidant (5 mM, N-acetylcysteine; NAC) for the next 23 hours. This treatment sequence was repeated for 3 consecutive days using 3 μM PCB153, Aroclor 1254, and 4ClBQ and then the cells were assayed for clonogenic cell survival. The results in show that the combination of PEG-CAT and PEG-SOD (50 U/ml each) added one hour after PCBs during each day of exposure for three days can partially rescue MCF-10A and RWPE-1 cells from the cytotoxicity mediated by PBC153, Aroclor, and 4ClBQ. Since PEG not conjugated to the antioxidant enzymes was incapable of rescuing the cells from PCB-induced toxicity, these results strongly support the hypothesis that PCB-induced increases in steady-state levels of O2·− and H2O2 play a significant role in the loss of cellular reproductive integrity seen during PCB exposure. Furthermore treatment with the thiol antioxidant, NAC added one hour after PCBs during each day of exposure for three days, completely rescued the MCF-10A cells from the toxicity associated with PCB 153 and Aroclor as well as partially rescuing cells from the toxicity of 4ClBQ. In RWPE-1 cells, treatment with NAC after PCB exposure could also significantly rescue the cells from PCB-induced clonogenic inactivation (); however, it was also found that NAC treatment of RWPE-1 decreased the plating efficiency of untreated cells by 50% (data not shown). Overall these results strongly support the hypothesis that antioxidant treatments targeting ROS scavenging and/or enhancing the thiol capacity of the cell following PCB exposure can significantly protect human breast and prostate epithelial cells from PCB-induced cytotoxicity. This novel finding suggests that manipulation of antioxidants (following exposure to PCBs and their metabolites) in breast and prostate epithelial cells could mitigate the deleterious biological effects of PCBs.
MCF-10A cells treated with PEGCAT+PEGSOD or 5 mM NAC 1 hour following PCB exposure are protected from PCB-induced toxicity