Introduction

Prostate cancer is the most commonly diagnosed malignancy and the second leading cause of cancer-related mortality in the human male population.17 Sole treatment with conventional external beam doses below 68 Gy has limited curative potential for the locally advanced stages of prostate cancer (LAPC) because, approximately, two-thirds are hypothesized to experience disease relapse every 5 years after a treatment is undertaken.20,46 Molecular mechanisms for LAPC radiation resistance are not well understood. It is known that ionizing radiation (IR) induces cellular toxicity and death through mechanisms involving generation of intracellular reactive oxygen species (ROS), such as superoxide radical (\( {\text{O}}^{{ - \cdot }}_{2} \)) and hydrogen peroxide (H2O2). The fate of irradiated cells is believed to be controlled by the network of signaling elements that lead to different modes of cell death or survival. Many stress-responsive genes are inducible by IR10 but, presumably, only a fraction of these genes, such as elements in cell cycle checkpoints, apoptosis and DNA repair7,34 play a key role in the oxidative stress-tolerance phenotype. Although p53 and p21WAF1/Cip1 are known to be involved in cell cycle and apoptosis regulation,5,44 the p53 gene in PC-3 cells encodes a non-functional p53 protein,16 leaving any potential control to p21WAF1/Cip1. The DNA-binding of nuclear factor-kappaB (NF-κB), a major signaling molecule is constitutively present in the androgen-independent prostate tumor cells PC-3, and correlates with disease progression.39 NF-κB promotes malignant behavior by suppressing apoptosis and stimulating transcription of proteins involved in cell cycle progression.24 Overexpression of antioxidant enzymes (AOEs), superoxide dismutases (SODs), is also believed to influence NF-κB,8,25 and NF-κB-binding sequences were found in the SODs promoter.21,45,47 Metastatic lesions of primary prostate cancer contain higher levels of AOEs, SODs and Cat, but significantly lower one of glutathione peroxidase (GPx) than those of primary tumors, suggesting that AOEs in metastatic and primary tumors may be modulated in different fashions.19,31 Although AOEs are able to scavenge oxidants,11 ROS escaping from the enzymes are still capable of damaging cells to a lesser extent.6 Since different constitutive AOE equipments of the metastatic PC-3 cell line offer the possibilities of investigating the antitumor effects of IR, a therapeutic dose of 2 Gy, as well as cumulative doses of 10 and 20 Gy were exploited throughout the work. The IR-induced changes in cellular parameters, such as cell viability, proliferative index, apoptosis, and cell cycle phases, are herein explored, besides the expression of cellular AOEs and NF-κB. To rationalize the molecular basis of PC-3 radioresistance, our measurements are complemented with a systems biology model.

Materials and Methods

Cell Culture

Androgen-independent (PC-3) human prostate carcinoma cell line was obtained from ATCC American Type Culture Collection (CRL 1435, Rockvill, MD). Cells were grown as monolayer cultures in RPMI 1640 supplemented with 10% fetal bovine serum, 2 mM l-glutamine, 100 IU/mL penicillin/100 μg/mL streptomycin at 37 °C in a humidified atmosphere with 5% CO2, and subcultured once a week.

Cell Irradiation

For the investigation of radiation-induced effects on PC-3, 1.2 × 106 cells were seeded in 75 cm2 culture flasks. Cell IR was performed in the exponential phase of cell growth, with 2, 10, or 20 Gy of gamma-rays from a 60Co source and a fixed dose rate of 20 Gy/h.

Cell Viability, Proliferation, and Death Analyses by Flow Cytometry

The cell viability was determined by trypan blue exclusion (TBE) assay. The ratio between viable cell numbers in IR and control cells was presented as the Viability index (Vi). For cell cycle analysis, cells were fixed with 70% ethanol, treated with 50 μg/mL RNAse-A, stained with 50 μg/mL propidium iodide (PI) and analyzed with a FACS Calibur flow cytometer (Becton Dickinson, San Jose, CA, USA). Cell cycle distribution was determined using ModFIT software (Verity Software House, Inc., Topsham, ME, USA). For cell death analysis 105 PC-3 cells were stained with 5 μg/mL Annexin V-FITC and 5 μg/mL PI (Travigen Inc., Gaithersburg, MD, USA) and multiparameter cytometric measurements were performed using minimum of 20,000 events for each sample. The data were processed with the LYSIS II software (Becton Dickinson).

Cell Extract Preparation and Western Blot Analyses

The trypsinized and washed cells were lysed with buffer containing 0.32 M sucrose, 10 mM Tris HCl pH 7.4, 5 mM MgCl2, and 1% Triton X-100. Aliquots of lysates were denatured, separated by electophoresis on 10% acrylamide gel and transferred to nitrocellulose membranes. The membranes were blocked using 10 mM Tris buffer pH 7.4 supplemented with 150 mM NaCl, 1% BSA and 0.1% Tween-20, and separately incubated with rabbit anti-NF-κB or actin antibody (NF-κB p65, C-20; SC-372, or C-11; SC-1615, Santa Cruz Biotechnology, USA), rabbit anti-human MnSOD or CuZnSOD antibody (SOD-101 or SOD-100, Stressgen Biotechnologies, Victoria, BC, Canada), and rabbit anti-human catalase antibody (Calbiochem, USA). A secondary goat anti-rabbit IgG Horseradish Peroxidase Conjugate SAB-300 (Stressgen Biotechnologies, Victoria, BC, Canada) was used for detection. The quantification of specific antigen bands was done by computer image processing (Image J). The level of detected proteins is expressed in arbitrary units (AU/mg of protein).

Statistical Analysis

Values are presented as mean ± SD or mean ± SEM of two separate experiments with N representing the number of measurements in every single experiment. Statistical significance was evaluated using one-way analysis of variance (ANOVA) and the Tukey post-hoc test. The probability thresholds considered significant throughout this work were ***p < 0.001, **p < 0.01, and *p < 0.05.

Results

Antiproliferative and Cytotoxic Effects of Gamma-Radiation on PC-3 Prostate Cancer Cell

To investigate the antiproliferative and cytotoxic effects of IR, the PC-3 cells were irradiated with 2, 10, and 20 Gy. The effects of IR on cell viability, proliferation, and death were analyzed by TBE assay and flow cytometry after double staining with Annexin V-FITC and PI, 72 h after the IR treatment. As the results indicate cell viability is slightly reduced after exposure to IR in the dose-dependent manner (Table 1). The greatest effect of IR on cell proliferation and cell death was observed at a dose of 20 Gy which was decreased for 76%. Thus, the calculated Vi was 24%. In addition, the IR dose inducing a decrease of cell viability from 100 to 50% (IC50) was 8.4 ± 1.6 Gy. The flow-cytometric analysis revealed a small increase in apoptosis, while an increase in the number of cells being in late apoptosis/necrosis was estimated to be between 8% in the control sample and 20% in the sample irradiated with 20 Gy (Table 1). The overall results indicate that suppression in PC-3 cell growth by IR is associated with cell cycle perturbations.

TABLE 1 The dose-dependent effects of gamma-irradiation on the viability and viability index of PC-3 cells, 72 h after treatment (mean ± SD, N = 4)
Full size table

Cell Cycle Perturbations in Irradiated PC-3 Prostate Cancer Cell

Since the IR-prevented cell growth, flow cytometric analysis of ethanol-fixed cells stained with PI was used to test whether IR had an effect on cell cycle arrest. The results showed a dose-dependent increase in the percentage of cells in the G2/M phase, from 28% in control to 53% in the 20 Gy-irradiated sample (Table 2). This increase was followed by a 34% decrease of cell population in the G0/G1 phase, and a 4% increase of cells in the S phase. The percentage of the subG0/G1 phase cells also increased for about 5% after irradiation. Thus, the dominant effect of IR in PC-3 prostate cancer cell was arrest in the G2/M phase of cell cycle.

TABLE 2 The dose-dependent effects of gamma-irradiation on the cell cycle distribution of the PC-3 cells, 72 h after treatment (mean ± SEM, N = 3)
Full size table

Antioxidant Enzymes and Transcription Factor NF-kB in Irradiated PC-3 Prostate Cancer Cells

To analyze AOE expression as a function of IR, we determined the levels of MnSOD, CuZnSOD, and Cat in cell lysates 72 h after the treatment. Western immunoblotting analysis showed that MnSOD level was unchanged at 2 Gy, but it was twofold increased at 10 and 20 Gy (Fig. 1). The expression of CuZnSOD was significantly increased at all radiation doses. The expression of Cat slightly increased after IR, but these alterations were not statistically significant (Table 3). The expression of transcription factor NF-κB was not altered in PC-3 cells after the IR treatment (Table 3), although its transcriptional activity was indirectly detected through transactivation of both SODs. Therefore, ROS are suggested to activate SODs, while NF-κB is suggested to mediate the antiproliferative and cytotoxic action of IR in PC-3 cells, possibly by both scavenging a part of ROS and attenuating the direct oxidation of cellular targets. The attenuating mechanism may provide the basis of high radioresistance of PC-3 cells reflected in the high IC50 and sustained viability.

FIGURE 1
figure 1

The dose-dependent superoxide dismutase expression in PC-3 cells, 72 h after treatment (mean ± SD, N = 4). The probability thresholds considered significant are ***p < 0.001 and **p < 0.01

Full size image
TABLE 3 The dose-dependent effects of gamma-irradiation on the Mn, CuZnSOD, NF-κB, and Catalase protein expressions, 72 h after treatment (mean ± SD, N = 4)
Full size table

Based on several recent contributions demonstrating that the pro- and anti-apoptotic responses can be controlled by the NF-κB-mediated expression of MnSOD,3,43 our further efforts were dedicated toward the modeling of the mechanism using a systems biology approach, which was shown to be a valuable means of elucidating the fundamental background of the ability of cells to interpret the NF-κB signaling associated with the ROS generation.27,28

Model Description

An outline of considered reaction pathways is shown in Fig. 2a, while the corresponding kinetic reactions with the key molecular components are given in Fig. 2b. Because of both 49 molecules participating in 94 unidirectional reactions and insufficient space in the present article, the Auxiliary Material (AM) containing all of the molecular and kinetic details is available. Chosen initial concentrations of participants and values of kinetic constants are our best interpretation of the literature.

FIGURE 2
figure 2

Outline of the signaling pathways with the main modules (a), and kinetic scheme describing the IR-induced expression of MnSOD (b). The p subscript denotes the “protein” term, the g subscript denotes the “gene” term, NF-κBn means nuclear NF-κB, and Cat abbreviates catalase

Full size image

The NF-κB module is a modified version of that previously reported.27,28 The module is essentially the IκB-NF-κB signaling pathway of Hoffmann et al. 13 modeling the experimental data successfully. In the cytoplasm of non-stimulated cells, NF-κB is held inactive by three IκB isoforms, IκB-α, IκB-β, and IκB-ε. Following an equilibration period of 2000 min, raising the concentration of PI3 K to 0.1 μM was chosen to simulate the stimulation of a cell by IR, as previously reported for IKK by Ihekwaba et al.15 Activating PI3 K complex causes phosphorylation and ubiquitination of the IκB proteins. Translocation of unbound NF-κB to the nucleus activates genes including IκB-α. Synthesizing IκB-β and IκB-ε at a steady rate provides complex temporal control of NF-κB activation with a local negative feedback.13 The first 64 reactions taking place within this module are the irreversible mass action reactions, as given in the AM. The MnSOD module is based on the set of 20 unidirectional reactions (67–86) taken form the literature.4 Because varying levels of the mitochondrial form of SOD, MnSODIII, were considered in the previous work,4 MnSODp shown in Fig. 2b corresponds to MnSODIII. The p subscript denotes the “protein” term. The NF-κB and MnSOD modules are linked by the mass action reaction 65 between nuclear NF-κB (NF-κBn) and MnSODg with the g subscript standing for the “gene” term. MnSODg and MnSODp are connected with the reaction 66 (Fig. 2b). The inhibitory effect of MnSOD on IκB-α degradation and NF-κB activation is associated with the reactions 93 and 94, as previously suggested in the literature.25,32 Besides the IR-induced expression of mitochondrial MnSOD, we initially considered the IR-induced expressions of cytosolic CuZnSOD and Cat. Taking into account the reactions 67–86 for MnSOD, some similar sets of kinetic reactions for CuZnSOD and Cat were not knowable. In this context, the problem was simplified to a certain extent. The CuZnSOD and Cat feedback mechanisms, from NF-κBn to H2O2, were considered to proceed via the 87–88–89 and 90–91–92 reaction pathways, respectively, as shown in Fig. 2b. Since the capacity of the system for removing H2O2 was maintained constant, the inclusion of a primarily peroxisomal enzyme, such as Cat, in the H2O2-removing system would make no difference in the simulation performances of the model.4

The modeling environment GEPASI 3.30, running under the MS-Windows operating system, was employed to carry out the simulations of biochemical pathways.26 The GEPASI 3.30 suite of programs is capable of performing time-course and steady-state simulations. Its scan parameter capability is one feature of the software package that was exploited extensively. It allowed us varying specific parameters, selecting the range, and defining the extent of the variation. In this way, unknown values of the kinetic parameters for the reactions 65, 66, and 87–94 were evaluated. Estimated values of the kinetic parameters given in the AM were of the same order of magnitude as those recommended in the literature.9,41 As we are particularly interested in the IR-induced expression of MnSOD, the initial concentrations of both CuZnSODp and catalase (Catp) were chosen to be fixed, 0 μM. The time-dependent concentrations of H2O2 for various levels of MnSODp, such as 0.7, 1, 4, 7, and 10 μM, were observed. The specific amounts of MnSOD and an initial concentration of 0 μM for H2O2 used throughout this work are those reported in the literature.4 To simulate experimental conditions as close as possible, the initial concentration of NF-κB was kept constant (see supplemental data).

Simulation Results

The time course of [NF-κBn] is shown schematically in Fig. 3, while the compounded plot of those of [H2O2] for various initial levels of MnSODp is depicted in Fig. 4. It is important to note that the steady-state concentration of H2O2 as a function of the level of MnSOD was explored too by using equilibrium constant of 0.01 for the reactions 67 and 68. Since the value is <1 and the reactions 67–86 are a mild variation of those previously reported,4 our conclusions qualitatively reproduced Fig. 4d given in the literature.4 Therefore, the steady-state simulations are not included in the present article.

FIGURE 3
figure 3

Nuclear NF-κB (NF-κBn) concentration vs. time for any initial concentration of MnSOD

Full size image
FIGURE 4
figure 4

The time-dependent concentration of H2O2 as a function of the initial concentration of MnSOD

Full size image

Discussion

Our measurements of viability, proliferation, and cell cycle profile in irradiate (IR) PC-3 cells confirm our previous42, as well as, the results of other authors16,35 indicating high IR resistance of metastatic prostate cancer cell lines. The low apoptotic potential of PC-3 cells was observed in our study, since cells responded to high dose of IR with quite a low apoptosis and an increased necrosis. The increase in percentage of necrotic cells was also low, even at rather high IR doses. Dominant response of PC-3 cells on IR was a dose-dependent inhibition of cell proliferation, which was inhibited for 50% (IC50) at 8 Gy, while 50% cell cytotoxicity (EC50) was not attainable within the IR dose range of 2–20 Gy. Cell cycle analysis confirmed that IR causes a decrease in the percentage of cells in G0/G1 and a concomitant increase in the percentage of cells in G2/M, which was in accordance with other studies.18,33 The high tolerance of PC-3 prostate cells to IR was documented in our study by rather low apoptotic response (<1%) which was in agreement with the data of other authors that documented the re-enterence of PC-3 cells in the cell cycle after a transient IR-arrest in G2.2,23 The IR treatment seems to induce an adaptive response and to generate a tolerance to IR-imposed oxidative stress in surviving PC-3 cells, since our analyses show that the viability of the surviving cells is rather high. In the absence of functional p53, the G2-arrest in PC-3 cells is dependent upon the p21WAF1/Cip1 protein, a member of the Cip/Kip family of cyclin-dependent kinase inhibitors5,37 that is sensitive to H2O2.1

In addition to describing cellular characteristics, our data indicated that the IR did not change the expression of NF-κB in PC-3 cells. This was in agreement with the observations of other authors.38 However, the DNA-binding of NF-κB was previously reported to be constitutively activated in PC-3,39,40 and it was accompanied by the high levels of SOD and Cat indicating that the cells were adapted to oxidative stress during differentiation from the primary tumor.19,23,31,45,47 Such conditions in prostate cancer cells may also be a consequence of non-functional p53 protein.16 Our measurements indicate that, after applying IR to metastatic PC-3 cells, the expression of both SODs was further increased while the Cat level remained unchanged. Although mechanisms by which MnSOD affects IR-induced adaptive responses are currently unknown, MnSOD overexpression was suggested to alter the intracellular redox environment by means of changes in mitochondrial H2O2,22 initiating downstream signaling cascades.36 This positive signaling loop in IR cells leads to an increased production of H2O2 which may be important for understanding the observed IR-induced adaptive response.

Based on the present observations and on the literature data, we designed a systems biology model for the adaptation to oxidative stress/resistance to IR, based on SOD induction by ROS (H2O2). Based upon the model, the 0.1–1.3 μM concentration range of H2O2 generated by the input IR between 2 and 20 Gy corresponds to the 0.6–2.8 μM concentration range of H2O2 that is associated with the MnSOD induction of 0.7–10 μM (Fig. 4). Our measurements qualitatively support consequences of the detoxifying/scavanging activity of MnSOD, which are manifested by both intensive survival of the cells and a low death rate at high-dose IRs.

The inclusion of some inhibitors with an adjuvant effect on SODs might be implicated in the search for an effective IR therapy. The IR-induced signaling elements, interacting with each other and forming a network, may provide a by-pass to keep cell survival. Therefore, potential treatment with new drugs that block MnSOD and CuZnSOD expression should down-regulate a group of the IR-induced genes that increase cell radioresistance.12 For example, fibroblasts derived from CuZnSOD knockout mice proliferate more slowly (75%) than control cells, emphasizing the importance of CuZnSOD in cell growth and survival.14 Besides, aloin, an inhibitory agent in the scope of our other studies,30 was shown to be quite an effective inhibitor of the CuZnSOD activity in HeLa cells.29

Conclusion

The high-ionizing radiation resistance (IR) of human prostate cancer cells (PC-3) illustrated by IC50 ≈ 8.5 Gy and EC50 > 20 Gy may be ascribed to the high constitutive and IR-induced expression of mitochondrial MnSOD and cytosolic CuZnSOD. In the absence of catalase, the expression enhancements of MnSOD and CuZnSOD lead to relatively successful PC-3 cell adaptation to prooxidative conditions induced by IR. In this context, our systems biology model shows that a possible mechanism for the adaptation of prostate cancer cells to IR is associated with an increasing trend of the concentrations of H2O2 due to MnSOD induction. We believe that our results provide an experimental and systems biology framework upon which a future therapeutic approach to the treatment of metastatic prostate cancer should rely.