Cell cycle perturbations and radiosensitization effects in a human prostate cancer cell line

Abstract

Purpose

To test the hypothesis that radiation-induced, transient G2/M arrest could potentially sensitize tumor cells to a subsequent, well-timed radiation dose.

Methods

PC-3 human prostate cancer cells were treated using either radiotherapy or ¹⁸⁶Re-labeled hydroxyethylidene diphosphonate (¹⁸⁶Re-HEDP) treatment in different combinations. The resulting cell cycle shift and clonogenic cell death were analyzed by DNA flow cytometry and colony forming cell assay, respectively.

Results

Radiation doses of 4 Gy and 8 Gy induced a transient G2/M arrest, with a maximum after approximately 16 h. The presence of 2 mM pentoxifylline effectively abrogated this radiation-induced G2 M arrest, confirming a cell-cycle checkpoint-mediated effect. A second dose of 4 Gy, timed at the height of the G2/M arrest, significantly increased clonogenic cell-kill compared to delivery after a suboptimal interval (10 h, 20 h or 25 h after the first radiation fraction). Moreover, timed second doses of 2 Gy, 3 Gy or 4 Gy yielded improved normalized treatment effects compared to non-pretreated control. Radionuclide treatment of PC-3 cells, using ¹⁸⁶Re-HEDP (0.74 MBq/ml and 1.48 MBq/ml; total dose: 4.1 and 8.2 Gy, respectively) also induced a dose-dependent G2/M accumulation, which sensitized the cells to a subsequent external radiation dose of 2 Gy or 4 Gy. The observed pattern of cell-cycle shift towards a predominance of the G2/M phase is in line with the lack of functional p53 in this cell line.

Conclusions

Radiation-induced cell-cycle shift was shown to effectively confer increased radiosensitivity to prostate tumor cells. Optimally timed combination of radiotherapy and radionuclide therapy could thus significantly increase treatment efficacy.

Introduction

It has been observed for some time that the various cell-cycle phases display a varying, relative radio-sensitivity, with the G2/M phase mostly being considered the most radiosensitive one (Sinclair 1968). In fact, Griffith and Tolmach (Griffith and Tolmach 1976) showed that cells become sensitive right at or near the arrest point in G2/M.

The realization of differential sensitivity during the subsequent phases of the cell cycle has led to a quest for an optimal combination therapy approach using the concept of synchronization. Although the initial optimism has, in general, not resulted in major curative successes, the empirical combinations of various chemo- and radiotherapy approaches have proven to incidentally yield protocols with increased therapeutic gain.

More recently, it has become clear that the cell cycle plays a major role in the decision processes that govern the transition of the cell to either proliferation or to proceed to quiescence, to postmitotic differentiation or to apoptotic cell death (e.g., Elledge 1996; Sherr 1996; Lundberg and Weinberg 1999). Moreover, it has been suggested that the increased efficacy of combination therapies could be due to perturbation of the cancer cell-cycle phase distribution (Hennequin et al. 1996; Kano et al. 1998; van Leeuwen-Stok et al. 1998; Wang et al. 1998; Formenti et al. 1999; Zoli et al. 1999; Koutcher et al. 2000). Experimental proof, in principle, could be supplied by induction of a transient wave of cell-cycle arrest, followed by a secondary radiotherapy fraction timed exactly at the height of cell accumulation in the most radiosensitive cycle phase.

In the present study, the hypothesis is tested that an increased control of tumor cell proliferation can be obtained by aimed and well-timed therapy approaches after modulation of the tumor cell-cycle distribution within an experimentally well-defined in vitro model.

Methods and materials

Tumor cell culture

The human prostate tumor cell line PC-3 was used to measure effects on cell cycle progression and clonogenic tumor cell survival. PC-3 cells were originally obtained from the American Type Culture Collection, Rockville, Md., USA (ATCC# CRL 1435), and used between passage number 60 and 130. Cells were cultured in RPMI-1640 culture medium (Gibco BRL, Life Technologies, Gaithersburg, Md., USA), supplemented with 10% fetal calf serum (Gibco BRL, Life Technologies, E.C.-approved quality), 100 U/ml penicillin/streptomycin (Gibco BRL), 1 mM sodium pyruvate (Gibco BRL), and insulin/transferrin/selenite-medium supplement (Sigma Chemical, St. Louis Mo., USA), at 37 °C in a humidified atmosphere of 5% CO₂/95% air.

Radiotherapy and radio-isotope treatment

Radiotherapy treatment of PC-3 cells was given using a Clinac 6/100 linear accelerator (field width 30×30 cm) using doses of 1 Gy, 2 Gy, 3 Gy, 4 Gy or 8 Gy at 2.5-cm depth in 6-well culture cluster plates (Corning Costar, Cambridge Mass., USA). The culture cluster plates were radiated between Perspex plates (1.5 cm on top and 6 cm underneath) at room temperature. Cells were plated in culture wells 48 h before radiation and the volume within the wells was supplemented with fresh culture medium to reach 10 ml total volume. Control wells (0 Gy) were left in the radiation control room during the radiation session. In one experiment pentoxifylline (methylxanthine) (Sigma-Aldrich Chemie, Zwijndrecht, The Netherlands) was added to the cultures immediately before radiation to a final concentration of 2 mM.

Treatment with ¹⁸⁶Re-labeled hydroxyethylidene diphosphonate (¹⁸⁶Re-HEDP) (Mallinckrodt, Petten, The Netherlands) was performed by adding the desired amount of radioactivity to the cells: 0 MBq/ml, 0.37 MBq/ml, 0.74 MBq/ml, and 1.48 MBq/ml (0 μCi/ml, 10 μCi/ml, 20 μCi/ml, and 40 μCi/ml) reaching a total well-volume of 5 ml. ¹⁸⁶Rhenium is a mainly β-emitting radionuclide with a half-life of 90.6 h (mean β-energy: 0.31 MeV and 0.36 MeV).

Treatment effect on prostate tumor cell clonogenic survival

The effect of radiotherapy and radioisotope treatment on human prostate tumor cell survival was investigated using a clonogenic tumor cell assay. Immediately after treatment, cells were harvested using trypsin, washed with PBS, and counted using an electronic cell counter (Casy-1, Schärfe System, Reutlingen, Germany). Cells were plated homogeneously in 6-well culture clusters in densities of 200 cells per well (n=6) for clonogenic cell assay. In the case of combined 4+4 Gy combination treatment, cells were plated in densities of 2,000 cells per well. In the case of the radio-isotope treatment combined with external radiotherapy, cells were plated at densities between 400 and 2,000 cells per well. After 10 days of culture (at 37 °C in a humidified atmosphere of 5% CO₂/95% air), the resulting colonies were fixed using 4% formaldehyde, stained using Giemsa stain (diluted 1:40), and counted under a stereo microscope at 10× magnification.

The number of colonies grown from treated cultures was compared to the number of colonies grown from control treated cultures and the relative survival of colonies was expressed as surviving fraction (SF) according to the following formula:

in which:

The data shown in the figures are for one experiment, but the results were reproduced and confirmed in at least three identical experiments

Flow cytometric determination of cell-cycle phase distribution

The cells treated were prepared for DNA flow cytometry (n=6 for each treatment) to assess the relative distribution in the respective phases of the cell cycle. After harvesting by trypsinization, the cells were washed twice using PBS, fixed in ethanol 70% and stored at 4 °C. Immediately before flow cytometry, the cells were washed in cold (4 °C) PBS, incubated in Ribonuclease A (Sigma #R5000, 250 µg/ml in 0.1% triton ×100) during 20 min at room temperature and labeled by adding an equal volume of propidium iodide solution (100 µg/ml) and incubated in the dark for 20 min at 0 °C. These samples were measured (20,000 events collected from each) in a FACS scan unit (Becton Dickinson Immunology, Mountain View, Canada) and the cell cycle distribution was analyzed using Cellfit cell cycle analysis software (Becton Dickinson).

The data shown are for one experiment, but the results were reproduced and confirmed in at least three identical experiments

Statistical analysis

Experimental results from both clonogenic tumor cell assay and flow cytometric cell-cycle analysis were analyzed using Fisher's Exact Test for 2×2 tables (Chi² analysis; one-tailed P-values <0.05 were considered to indicate significance) unless indicated otherwise.

Results

Radiotherapy effects on the cell cycle distribution

Monolayers of PC-3 human prostate cancer cells were irradiated in vitro during logarithmic cell growth in 6-well culture clusters, filled with maximal amount of RPMI-1640 cell culture medium (10 ml). Cell cultures irradiated with doses of 0 Gy, 4 Gy, and 8 Gy were harvested and prepared for flow cytometry every 4 h after irradiation.

Figure 1A-C shows the relative proportion of cells from these cultures, respectively, in G0/G1 phase, in S-phase, and in G2/M-phase of the cell cycle. One of the first radiation-induced effects observed in the distribution over the cell-cycle phases is an increase of the proportion of cells in the S-phase coupled with a decrease of the G0/G1 proportion. This effect is transient and of only short duration while subsequently—within the first 24 h after radiation—the vast majority of cells enter and stay in the G2/M phase. Sixteen hours after a dose of 4 Gy, 76% of the cells have accumulated in G2/M phase and after a 8-Gy dose this proportion even increases to 88%. The G2/M accumulation after 4 Gy is only temporary in nature and disappears with time. In fact, the cell-cycle distribution observed after radiation displays an oscillating pattern during the observation period within the first 2 days. The effect of treatment observed is dose-dependent, with the 8 Gy dose resulting in a more profound and also more durable G2/M-accumulation. However, also after a 8 Gy dose, the proportion of cells in S-phase can be seen to reach near-normal values again at 44 h after radiation.

Fig. 1.

Cell-cycle phase distribution of PC-3 human prostate cancer cells after single dose radiation of 0 Gy, 4 Gy or 8 Gy

In the next experiment, PC-3 cells were irradiated at lower doses, and the cell cycle distribution at 16 h and the clonogenic cell survival (at 10 days) were determined. The cell-cycle changes observed in this experiment (Fig. 2A) essentially show the same pattern as in the previous experiment: a dose-dependent increase of the G2/M-proportion of cells at the cost of the other two cell-cycle phases. Even after a dose as low as 1 Gy, a (non-significant) increase in the G2/M compartment is observed. Concomitantly with this change in cell-cycle distribution, a decrease in clonogenic cell survival after radiation can be observed. A dose of 2 Gy results in a significant (P<0.0001) decrease in fractional clonogenic survival to 68%.

Fig. 2.

Cell cycle phase distribution of PC-3 human prostate cancer cells 16 h after a radiation dose of 0 Gy, 1 Gy, 2 Gy, 3 Gy or 4 Gy and the ultimate clonogenic survival (measured after 10 days)

The effect of pentoxifylline (2 mM final concentration in culture well added immediately before radiation) on the emergence of a G2/M arrest 17 h later, was investigated. Figure 3 shows that the radiation-dose dependent G2/M arrest can be abrogated by the presence of pentoxifylline.

Fig. 3.

Effect of pentoxifylline (PTX, 2 mM) on G2/M cell cycle arrest 17 h after radiation of PC-3 human prostate cancer cells

¹⁸⁶Re-HEDP treatment effects on the cell-cycle distribution

Analogous to the experiments using external radiation, the cell-cycle distribution of PC-3 human prostate cancer cells was evaluated also after radio-isotope treatment with ¹⁸⁶Re-HEDP. Figure 4A–C shows the resulting percentages of cells in the respective phases of the cell cycle every 12 h after start of radio-isotope incubation (activity concentrations of 0 MBq/ml, 0.36 MBq/ml, 0.74 MBq/ml, and 1.48 MBq/ml). The G2/M accumulation observed is not so dramatic as that after external radiation and is slower in appearance. Nevertheless, a dose-dependent rise in G2/M fraction to over 47% of cells is observed at 36 h after the start of incubation.

Fig. 4.

Cell cycle phase distribution of PC-3 human prostate cancer cells after incubation with ¹⁸⁶Re-HEDP (activity concentrations of 0 MBq/ml, 0.37 MBq/ml, 0.74 MBq/ml or 1.48 MBq/ml)

In contrast to the situation with radiotherapy, the radiation exposure of cells due to the radioactive decay of ¹⁸⁶Re-HEDP changes continuously with time in a logarithmic way. The initial dose rate was 0.15 Gy/h for 1.48 MBq/ml (and 0.08 Gy/h and 0.04 Gy/h for 0.74 MBq/ml and 0.36 MBq/ml, respectively). The total absorbed doses, totalled for the incubation time are given in Table 1.

Table 1.

Summed absorbed dose (Gy) of PC–3 human prostate tumor cells during incubation with ¹⁸⁶Re–HEDP (activity concentrations of 0.37 MBq/ml, 0.74 MBq/ml or 1.48 MBq/ml)

Incubation time	1.48 MBq/ml	0.74 MBq/ml	0.37 MBq/ml
0 h	0	0	0
12 h	1.7	0.85	0.43
24 h	3.2	1.6	0.8
36 h	4.6	2.3	1.15
48 h	5.9	2.95	1.5
60 h	7.1	3.55	1.8
72 h	8.2	4.1	2.05

Effects of timed fractionation of radiotherapy on clonogenic cell survival

To evaluate the effect of combining two treatment cycles of radiotherapy aimed at modulating cell-cycle distribution, PC-3 prostate cancer cells were radiated as monolayers in 6-well clusters using two subsequent doses of 4 Gy. The interval time between the two radiation doses was chosen according to optimal and suboptimal timing of the G2/M accumulation (induced by the first dose): respectively, 15 h after the first dose of 4 Gy, versus 10 h, 20 h or 25 h after the first dose. The fractional survival of colonies (10 days after plating) was calculated compared to the number of colonies in untreated controls. The results (Table 2) show a significant effect of the length of the treatment interval.

Table 2.

Fractional clonogenic survival of PC–3 human prostate cancer cells treated with two fractions of 4 Gy with variable intervals. The fractional survival of colonies (10 days after plating) from treated cell suspensions was calculated compared to the number of colonies in untreated controls (n=6)

Interval	Plated cell #	Colonies/1,000 cells	% clonogenic survival	P valuea
(Control)	200	590.3±42.0	100±7.12
(1×4 Gy)	200	169.8±28.6	28.8±4.85
10 h	2,000	12.9±2.3	2.19±0.39	0.0081
15 h	2,000	9.3±2.0	1.58±0.34	–
20 h	2,000	16.5±2.0	2.80±0.34	<0.0001
25 h	2,000	24.3±2.9	4.12±0.49	<0.0001

^a Significance tested of the difference with the 15-h interval–value (unpaired, one–tailed t test)

In a following experiment, two radiotherapy fractions of variable doses were combined with an inter-dose interval of 16 h (an interval known to result in approximately maximal G2/M phase accumulation. The results (Fig. 5A–C) show the curves of the fractional survival of PC-3 cells normalized to 100% for each pretreatment. It is clear from these figures that a pretreatment with 2 Gy, 3 Gy or 4 Gy sensitizes the cells for subsequent radiotherapy after this well-timed interval. The normalized, fractional clonogenic survival is decreased significantly (P<0.05) after a subsequent second dose of 2 Gy, 3 Gy or 4 Gy, resulting in a steeper slope of the survival curve.

Fig. 5.

Normalized clonogenic survival of radiated PC-3 human prostate cancer cells pretreated 16 h before by various radiation doses

Effect of combining ¹⁸⁶Re-HEDP treatment with radiotherapy

To evaluate the radio-sensitivity of cells after a radio-isotope pretreatment, aimed at raising the proportion of G2/M phase cells, PC-3 prostate cancer cells were incubated with different concentrations of ¹⁸⁶Re-HEDP for 48 h and then treated with doses of 0 Gy, 2 Gy or 4 Gy external radiation and plated for clonogenic assay. The normalized fractional clonogenic survival is plotted in Fig. 6. In addition, in this case the pretreatment severely and significantly (P<0.00001 for the combinations 2 Gy + 1.48 MBq/ml and 4 Gy + 0.74 or 1.48 MBq/ml) affects the sensitivity of the PC-3 cells for a subsequent dose of radiation. The pretreatment using ¹⁸⁶Re-HEDP resulted in a total absorbed dose of 5.9 Gy and 2.95 Gy, respectively, for the incubations using 1.48 MBq/ml and 0.74 MBq/ml (and yielded survival values by itself of 43.1% and 50.7%, respectively, compared to untreated control).

Fig. 6.

Normalized clonogenic survival of radiated PC-3 human prostate cancer cells after a 48 h pretreatment with ¹⁸⁶Re-HEDP (activity concentrations of 0 MBq/ml, 0.74 MBq/ml, or 1.48 MBq/ml)

Discussion

The present study explores the cytotoxic efficacy of well-timed radiotherapy doses after experimental manipulation of the cell-cycle distribution. In previous experiments, a synergistic interaction was observed between chemotherapy and either radiotherapy or radioisotope treatment using the PC-3 prostate tumor cell line (Geldof et al. 1999a, 1999b). These observations prompted us to consider the contribution of treatment-induced shifts in the cell-cycle distribution in causing supra-additive cytotoxic effects. It is a well-known fact that the various phases of the cell cycle display differential radiosensitivity (Sinclair 1968). Cells in the G2/M phase of the cell cycle have been shown to be far more radiosensitive than cells in other phases of the cell cycle. The induction of cells into such a relatively sensitive cell-cycle phase could thus potentially increase the therapeutic gain of a subsequent treatment dose since sensitization by this mechanism does not occur in late responding normal tissues (Hall 1994). Alternatively, any sub-optimal timing of the next treatment fraction would lead to a potentially dangerous loss of therapeutic efficacy.

Accumulation in the (radiosensitive) G2/M cell-cycle phase was effectively achieved in the present study by radiotherapy doses from 1 Gy to 8 Gy. In addition, treatment with radioisotopes—although less effective—did induce a G2/M arrest. A second radiation dose, delivered at the time of maximal G2/M accumulation resulted in significantly more clonogenic cell death than a suboptimally timed radiation using the same dose. The observed maximal cell kill at and around 15 h interval (Table 2) cannot be explained by lethal damage repair and repopulation since it would not explain the difference with the 10 h interval. Therefore, the difference in clonogenic cell kill has to be explained on the basis of differences in cell-cycle phase distribution. Moreover, further investigations, analyzed using normalized treatment curves clearly showed a significant enhancement of proportional treatment effect in cells, pretreated to reach maximal G2/M accumulation compared to control (not synchronized) cells.

Replicating cells are known to respond to radiation with cell-cycle progression delay through activation of cell cycle "checkpoints" (Hartwell and Kastan 1994). Such checkpoints at the G1-S transition and at the G2-M transition are thought to be involved in the detection and correction of DNA damage. The tumor suppressor protein p53 is a critical regulator of the G1/S checkpoint (Kastan et al. 1991; Kuerbitz et al. 1992; Hartwell and Kastan 1994). Cancer cells containing mutated p53 alleles fail to execute the G1/S checkpoint in response to DNA damage. Hence, such p53-null cells will be expected to arrest only in G2/M phase after radiation-induced damage. While the PC-3 human prostate tumor cell line has been described as lacking functional p53 alleles (Isaacs et al. 1991), the pattern of exclusive arrest in G2/M phase after radiation, as observed in the present study, is in line with this anticipation. The transient rise of cells in S-phase after acute radiation also reflects the lack of a G1/S checkpoint. Furthermore, the observed abrogation of G2/M accumulation by pentoxifylline (methylxanthine) in a p53-null cell line confirms the involvement of the G2/M cell cycle checkpoint (Russell et al. 1996; Bohm et al. 2000).

More recent reports indicate that the G2/M checkpoint may actually consist of a multiplicity of G2/M checkpoints, some of which are also under p53 regulation (Innocente et al. 1999; Wang et al. 1999). The dependency on the nature of the specific DNA damage (inflicted either by temperature, UV, radiation etc) has yet to be established conclusively.

In the literature, some discordant reports have been published concerning the relationship between cell-cycle distribution and radiosensitivity of tumor cells. Gupta et al. (Gupta et al. 1997) observed an increased cytotoxic effect of radiation after cell-cycle perturbation using paclitaxel, but conclude that the cytotoxicity of taxol may be multifactorial and that simply increasing the number of cells in G2/M phase is insufficient in itself to increase the response of cells to a subsequent radiation. Indeed, the underlying mechanism of taxol action includes a stable interaction with cellular microtubuli, thereby inhibiting the dynamic reorganization of the microtubular network (Schiff et al. 1979; Schiff and Horowitz 1980). On the other hand, we have demonstrated a far more efficient G2/M accumulation (to over 85% of cell population) after 24-h taxol treatment of PC-3 prostate tumor (Mastbergen et al. 2000), probably indicating the importance of correct timing and using the right (sensitive) cell system.

Likewise, Deweese et al. (Deweese et al. 1998) could not observe a direct relationship between cell-cycle arrest in human prostate cancer cells (PC-3 among them) and the killing effect of low-dose rate radiation (0.25 Gy/hr during 24 h). However, the effect of timing after treatment was not taken into account. In the present report, we demonstrate that the length of the interval between two doses of irradiation is of extreme importance in sensitizing a population of cells.

Low-dose irradiation has been suggested to result in increased sensitivity of cancer cells to the killing effects of protracted low-dose rate radiation exposure over that seen with equivalent doses fractionated high-dose rate radiation treatment by inducing cells in the radiosensitive G2/M phase (Kal et al. 1975; Mitchell et al. 1979; Bedford et al. 1980; Fowler 1990). In the present study. the effects of continuous incubation of cells in radio-isotope-containing culture medium was evaluated in terms of both cell-cycle distribution and clonogenic survival. It appears that relatively high activity concentrations of ¹⁸⁶Re-HEDP (1.48 MBq/ml) had to be included to reach any effective G2/M accumulation (peaking of 47% of cells, which is equivalent to the effect of a 3–4 Gy single-dose—and high-dose rate—external radiation).

It is concluded that experimental manipulation of the cell-cycle distribution is an effective approach to confer specific radiosensitivity to tumor cells in vitro. Although the situation in in vivo tumors is likely to be further complicated by differential factors such as oxygenation, supply of nutrients, and intrinsic cell proliferation rate, the principle of therapy-driven shifts in the cell-cycle distribution will hold and might be an important factor to take into account while planning the treatment. Differences between normal cells and cancer cells in this regard may add new hope for applying this rationale (e.g., Blagosklonny and Pardee 2001; Van Buul et al. 2001). Moreover, the advent of new in vivo imaging techniques, such as positron emission tomography using thymidine-analogue-based radiotracers, might aid in identifying cell cycle effects in target tissues. Thus, optimal combination therapy, guided by screening the tumor cell-cycle progression, could ultimately lead to increased treatment efficacy in individual patients.