An important stage in tumorigenesis is the ability of a precancerous cell to escape natural anticancer signals imposed on it by neighboring cells and its microenvironment. We have previously characterized a system of intercellular induction of apoptosis whereby nontransformed cells selectively remove transformed cells from coculture via cytokine and reactive oxygen/nitrogen species (ROS/RNS) signaling. We report that irradiation of nontransformed cells with low doses of either high linear energy transfer (LET) α-particles or low-LET γ-rays leads to stimulation of intercellular induction of apoptosis. The use of scavengers and inhibitors confirms the involvement of ROS/RNS signaling and of the importance of transformed cell NADPH oxidase in the selectivity of the system. Doses as low as 2-mGy γ-rays and 0.29-mGy α-particles were sufficient to produce an observable increase in transformed cell apoptosis. This radiation-stimulated effect saturates at very low doses (50 mGy for γ-rays and 25 mGy for α-particles). The use of transforming growth factor-β (TGF-β) neutralizing antibody confirms a role for the cytokine in the radiation-induced signaling. The system may represent a natural anticancer mechanism stimulated by extremely low doses of ionizing radiation. [Cancer Res 2007;67(3):1246–53]

The progression of cancer in living organisms is understood to be a multistage process, which can be thought of as a series of capabilities or traits that a cell will obtain in the course of tumorigenesis (1). An important concept in understanding cancer is that cells cannot be considered in isolation when looking at the process of tumorigenesis, and the role of neighboring cells is vital in the progression of transformed cells to tumors (2). The cellular microenvironment is so important because once a cell is precancerous, it becomes a potential target for a variety of natural anticancer defense mechanisms.

Some natural anticancer mechanisms are not dependent on neighboring cells, one of the most important of which is p53-mediated apoptosis or growth arrest to prevent tumorigenesis. Responses modulated by p53 can occur in precancerous cells in response to DNA damage or altered gene expression (3, 4). A number of other natural anticancer mechanisms exist that are dependent on the cellular microenvironment, and therefore signaling from neighboring cells. Cell growth within a population and particularly within a tissue is tightly controlled by a plethora of intercellular signals controlling proliferation under normal conditions. This tightly controlled normal tissue signaling actively inhibits cancer formation (5), acting as another fundamental anticancer mechanism beyond the control, such as cell cycle–dependent apoptosis, which each individual cell possesses to prevent tumorigenesis. Importantly, any external stimuli that might affect this cellular signaling network may in turn have an effect on the risk of tumorigenesis.

Considerable debate is presently ongoing about the observation of nontargeted effects following low doses of ionizing radiation in cells that have not seen radiation but have been affected by intercellular signals from irradiated cells in the same population. Such nontargeted effects may or may not lead to deviations from the “linear no threshold” dose response for ionizing radiation, which extrapolates cancer risk for low doses of ionizing radiation based on risks measured at high doses of radiation or from epidemiologic studies (6, 7). A great number of nontargeted effects have been identified, including micronuclei formation (8), mutations (9), changes in transformation frequency (10), a reduction in clonogenic survival (11), and apoptosis (12). A major challenge remains in characterizing the underlying mechanism for these phenomena. One way to address this problem from a unique angle is to identify a relevant model system that involves intercellular signaling by a systematically well-defined mechanism and that represents a good candidate for perturbation by ionizing radiation. Studying the effect of radiation on such a system would give insight into the mechanism of a radiation-induced nontargeted effect resulting from intercellular signaling.

In this study, we have chosen to use one such system for which a model of intercellular induction of apoptosis is well characterized (13, 14) to explore how low doses of radiation might perturb intercellular signaling. The mechanism for this model has been shown in a number of cell lines, although most detail has been obtained in the 208F rat fibroblast cell line and the 208Fsrc3 transformed derivative (1517). The system represents a potential natural anticancer mechanism where nontransformed 208F cells induce apoptosis selectively in the transformed 208Fsrc3 cells by reactive oxygen species (ROS) signaling, which is stimulated by cytokines including transforming growth factor-β (TGF-β; refs. 18, 19). A number of key proponents of this mechanism of intercellular induction of apoptosis make it a good candidate to study for perturbation by low doses of both high and low linear energy transfer (LET) radiation. Both TGF-β (20, 21) and ROS (8, 22) have previously been implicated in radiation-induced signaling, and membrane-bound NADPH oxidases, characteristic of src transformation and vital for the selectivity of this system (15), have been implicated in radiation-induced signaling at low doses (23). The aim of the present study is to develop a mechanism by which irradiation of 208F cells alters their intercellular signaling thereby affecting their ability to selectively induce apoptosis in nonirradiated transformed 208Fsrc3 cells.

Cells. The 208F and the v-src transformed 208Fsrc3 rat fibroblast cell lines, previously described by Heigold et al. (17), were used between passages 18 and 30.

Cell culture. Cells were kept in Eagle's MEM (Sigma, Poole, United Kingdom) supplemented with 5% FCS (Invitrogen, Carlsbad, CA), 2 mmol/L l-glutamine (Invitrogen, Carlsbad, CA), 50 μg/mL penicillin/streptomycin (Invitrogen), 10 μg/mL neomycin (Sigma), and 10 units/mL nystatin solution (Sigma). Cell culture was carried out in plastic tissue culture flasks (BD Biosciences, Oxford, United Kingdom). Cells were passaged twice weekly and maintained at 37°C in a 5% CO2/air incubator.

Antioxidants and other reagents. All antioxidants and inhibitors were prepared to give stock solutions as indicated in the manufacturer's instructions. All dilutions from stock solutions for use in experiments were made using cell culture media. The HOCl scavenger taurine (Sigma United Kingdom) was used at a final concentration of 25 mmol/L. The peroxynitrite decomposition catalyst 5,10,15,20-tetrakis(4-sulfonatophenyl)porphyrinato iron(III), chloride (FeTPPS; Calbiochem, Nottingham, United Kingdom), the NADPH oxidase inhibitor apocynin (Calbiochem), monoclonal anti–TGF-β1 antibody (R&D Systems, Abingdon, United Kingdom), and the caspase-3 inhibitor z-DEVD-fmk (R&D Systems) were used at final concentrations of 10 μmol/L, 50 μg/mL, 10 μg/mL, and 10 μmol/L, respectively. All final inhibitor concentrations have previously been determined by titration.

Coculture. Two milliliters of 208Fsrc3 cells at a density of 5 × 104/mL were seeded into each well of six-well Falcon plates (BD Biosciences). In parallel, 208F cells were seeded at a density of 1 × 105 in 1 mL of medium in custom inserts. These purpose-built inserts consisted of a 2.5-μm-thick replaceable hostaphan (0.35 mg cm−2 polyethylene terephthalate; Hoechst, Weisbaden, Germany) base on a glass ring with three 1-mm slits cut through it to allow media-borne signaling between the two populations. These slits occupied ∼50% of the ring circumference. A stainless steel collar was glued around the neck of the glass ring to hold the insert 1 mm above the base of a six-well plate.

Both cell populations were allowed to settle and attach in a 5% CO2/air gassed incubator at 37°C for 5 h. Subsequently, an additional 1 mL of medium was added to each well of 208Fsrc3 cells. 208F cell populations were irradiated as detailed below. The custom inserts containing the 208F cells were then placed into the six-well plates containing 208Fsrc3 cells and any antioxidants and inhibitors were added before placing the cells into the incubator. The maximum time between irradiation and incubation of the cells in coculture was 15 min. When FeTPPS and/or taurine was used, it was added 24 h after the beginning of the coculture incubation to prevent toxic effects of FeTPPS that occur after prolonged incubation in culture. The resulting coculture system was incubated at 37°C in a 5% CO2/air gassed incubator for 65 h before scoring for apoptosis as detailed below. First, in negative control assays, 208Fsrc3 cells were seeded in the absence of 208F cells. Second, in sham (0 Gy) controls, 208Fsrc3 cells were cocultured with nonirradiated 208F cells, which had been treated in exactly the same manner as irradiated cells except for the actual exposure to the radiation source.

Low-LET irradiations. 208F cells on custom inserts supported in empty six-well plates were irradiated on a 10-mm Perspex build-up sheet at room temperature with 60Co γ-rays (track averaged LET 0.2 keV/μm). Over the dose range 10 mGy to 0.5 Gy, irradiations were given at a dose rate of either 0.04 or 0.15 Gy/min. Over the dose range of 2 to 10 mGy, a dose rate of 0.014 Gy/min was used. Over the dose range of 0.5 to 2 mGy, irradiations were administered at a dose rate of either 0.4 or 1.47 mGy/min. To expose the samples in the lowest dose range, a lead attenuator was used. Detailed comparisons were made between a 2-mGy dose administered at two different dose rates (0.014 or 0.4 mGy/min) and with and without the lead attenuator to ensure that there is no dose rate effect or effects resulting from an altered spectrum of radiation following attenuation. Dose rate comparisons were also made between 0.04 and 0.15 Gy/min and between 0.4 and 1.47 Gy/min. All exposures times were <5 min. Dosimetry was carried out using a farmer type 2670 dosimeter (NE Technology Ltd., Reading, United Kingdom) with a 0.6-cm3 ionization chamber. Following irradiation, 208F cells in inserts were placed into a coculture system as indicated above.

High-LET irradiations. 208F cells on custom inserts supported on a 0.9-μm mylar-based brass ring attachment were irradiated at room temperature with α-particles using a 238Pu source with the details on the source dosimetry and setup documented previously (24). The incident energy of the α-particles at the cells is 3.0 MeV, which corresponds to a LET of 127 keV/μm. The absorbed dose (D) and therefore dose rates are calculated from the measured particle fluences (F) for a given α-particle energy with LET (L) using D = FL and allowing for the subsequent decay in activity with time (t1/2 = 87.7 years). Inserts awaiting radiation were kept at 37°C. When all irradiations had been completed, the 208F cells in inserts were placed into a coculture system as indicated above.

α-particle track traversal calculations. The mean cellular area and nuclear area were calculated by confocal microscopy to be 746 and 176 μm2, respectively. These figures were used to calculate the average α-particle track traversals using the following equation:

\[\mathrm{n\ =\ (DA)\ /\ (0.16L)}\]

where n, number of α-particle traversals; L, LET (keV/μm); A, nuclear or cellular area (μm2); D, dose (Gy).

Morphologic determination of apoptosis. After 65 h in coculture, 208Fsrc3 cells were scored for morphologic signs of apoptosis based on the criteria of membrane blebbing, nuclear condensation, and nuclear fragmentation determined by phase-contrast microscopy as previously described (25, 26). This end point was chosen for consistency with previous studies analyzing apoptosis in 208Fsrc3 cells, and so apoptosis could be scored in individual cells. The time point of 65 h for scoring was chosen based on having a measurable level of naturally occurring intercellular induction of apoptosis in the sham controls, allowing subsequent distinction between any inhibition or stimulation of apoptosis induction following treatment with radiation and/or chemical inhibitors and scavengers. The percentage of apoptotic cells was determined from at least 200 cells classified per assay. Each assay was repeated in triplicate in parallel within one experiment, and mean percent apoptosis score and SD were calculated for each condition. Each experiment was repeated at least twice. All quantitative data obtained are derived using this method. To ensure that the scoring was objective, experiments were also carried out with blind scoring.

Confirmation of morphologic findings. The morphologic determination of apoptosis in the 208F cell system has been thoroughly studied in the past, with comparison to other methods of scoring (15, 17, 25, 26). In this study, parallel control assays ensured that apoptotic cells characterized using the morphologic criteria detailed above showed a positive terminal deoxyribonucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) reaction, indicative of free 3′ hydroxyl groups of the DNA, which is one of the hallmarks of apoptosis.

DNA strand breaks (free 3′ hydroxyl groups) were detected by the TUNEL reaction using a commercially available detection kit (Roche Diagnostics, Penzburg, Germany). The kit protocol was optimized for use with 208Fsrc3 cells in six-well plates. After 65 h, all media were removed and cells were fixed with 3% paraformaldehyde for 10 min at room temperature. Cells were then washed thrice with nonsupplemented MEM (Sigma) and permeabilized with 100 mmol/L Tris (pH 7.4; 0.61 g/50 mL), 50 mmol/L EDTA (0.1 g/50 mL), and 1% Triton X-100 for 10 min at room temperature. After additional three washes with nonsupplemented MEM, the samples were dried by rolling an absorbent cotton bud around the edge of each well. One hundred microliters of TUNEL reaction mixture were added to each well and spread homogenously across the cell monolayer. Cells were incubated with the reaction mixture for 1 h at 37°C in a humidified atmosphere in the dark. The supernatant was then removed and the cells were washed thrice with nonsupplemented medium. The area around the samples was then dried as before. One hundred microliters of Converter horseradish peroxidase (POD) solution were added to each well and spread homogenously across the cell monolayer. Cells were incubated with the Converter POD solution for 30 min at 37°C in a humidified atmosphere, and then the solution was removed and the cells were washed thrice with nonsupplemented medium. Finally, 250 μL of 3,3′-diaminobenzidine (DAB) substrate (Roche Diagnostics) were added per well and the cells were incubated for 10 min at room temperature. The DAB substrate was then removed and the cells were washed thrice with nonsupplemented medium. Cells were then analyzed in PBS by light microscopy for TUNEL staining.

For further confirmation that the morphologic scoring method was identifying apoptotic cells, parallel experiments were carried out with the caspase-3 inhibitor z-DEVD-fmk (R&D Systems), which inhibits apoptosis, and therefore the appearance of apoptotic morphology.

Radiation of nontransformed cells leads to increased levels of apoptosis in nonirradiated transformed cells. It has been shown that nonirradiated 208F cells selectively induce apoptosis in transformed 208Fsrc3 cells by a process referred to as intercellular induction of apoptosis (for review, see refs. 14, 15). We asked the question about whether irradiation of 208F cells with γ-rays or α-particles would perturb the process of intercellular induction of apoptosis.

Table 1 shows that irradiation of 208F cells with 0.5 Gy of either γ-rays or α-particles before coculture with transformed cells leads to an increase in the percentage of nonirradiated transformed cells scored as apoptotic, above that seen when transformed cells are cocultured with nonirradiated nontransformed cells (0 Gy), which exhibit a basal level of intercellular induction of apoptosis resulting from transformed cell–derived TGF-β (see reviews in refs. 14, 15). The increase represents an approximate doubling in the percent apoptosis determined at 65 h and is independent of the LET of the radiation. Negative controls consisting of transformed 208Fsrc3 cells cultured in the absence of nontransformed 208F cells show a baseline level of apoptosis. Further control experiments, where inserts containing irradiated media only were cocultured with nonirradiated transformed cells, do not cause an increase in apoptosis relative to that for the negative control data (data not shown). When nonirradiated 208F cells are cocultured with irradiated 208F cells, neither population shows an increase in the level of apoptosis above a basal level of ∼5% (data not shown). Some variation in the percentage apoptosis scores was observed between assays not carried out in parallel due to some dependence on the passage number of both cell lines. Figure 1 shows the level of interexperimental variation observed.

Table 1.

Comparison of percent apoptosis scored in nonirradiated src transformed cells following 65-h coculture with irradiated 208F cells

Experimental conditionsDose (Gy)% Apoptosis (±SD)
LET comparison Negative control* 14.52 (±0.90) 
    γ-Rays 0 26.70 (±1.54) 
 0.5 47.15 (±1.66) 
    α-Particles 0 27.24 (±2.47) 
 0.5 48.69 (±2.24) 
Experimental conditionsDose (Gy)% Apoptosis (±SD)
LET comparison Negative control* 14.52 (±0.90) 
    γ-Rays 0 26.70 (±1.54) 
 0.5 47.15 (±1.66) 
    α-Particles 0 27.24 (±2.47) 
 0.5 48.69 (±2.24) 

NOTE: Data represent means taken from triplicate repeats carried out in a single experiment, with SD representing intraexperimental variation. Each experiment was repeated on at least two separate occasions and the same trends were observed.

*

Negative control refers to apoptosis scored in transformed cells in the absence of 208F cells.

0 Gy refers to apoptosis scored in transformed cells cocultured in the presence of nonirradiated 208F cells.

Figure 1.

Percentage apoptosis scored in nonirradiated transformed cells after 65-h coculture with 208F cells irradiated with either γ-rays over a dose range of 0.5 mGy to 0.5 Gy (A) or α-particles over a dose range of 0.04 mGy to 0.5 Gy (B). Points, mean value derived from triplicate repeats carried out at the same time, with the sham control (0 Gy) value carried out in parallel subtracted from each data point. Each different symbol represents data obtained in parallel from the same experiment.

Figure 1.

Percentage apoptosis scored in nonirradiated transformed cells after 65-h coculture with 208F cells irradiated with either γ-rays over a dose range of 0.5 mGy to 0.5 Gy (A) or α-particles over a dose range of 0.04 mGy to 0.5 Gy (B). Points, mean value derived from triplicate repeats carried out at the same time, with the sham control (0 Gy) value carried out in parallel subtracted from each data point. Each different symbol represents data obtained in parallel from the same experiment.

Close modal

A dose response is only observable at very low doses of γ-ray or α-particle irradiation.Figure 1A shows the dose-dependent response for the induction of apoptosis for γ-rays over a dose range of 0.5 mGy to 0.5 Gy above that of the sham-irradiated (0 Gy) controls. No increase above control levels was observed following doses of 0.5 or 1 mGy. The initial increase above the sham control level of apoptosis is seen at 2 mGy, where ∼28% of cocultured, nonirradiated transformed cells show apoptotic morphology relative to sham control values of ∼20%. This effect increases up to 50 mGy, at which dose the level of apoptosis reaches a plateau, with ∼40% of transformed cells showing apoptosis. A number of dose rates were needed to achieve the wide range of dose points reported, while maintaining short exposure times (<5 min). No observable difference in the level of γ-ray–induced effects was observed when different dose rates were compared for a given dose. Figure 1B shows the dose-dependent response for the induction of apoptosis for α-particles over a dose range of 0.04 mGy to 0.5 Gy. For α-particle irradiation, the initial increase above the sham control levels was observed at 0.29 mGy where ∼28% of cocultured nonirradiated transformed cells show apoptotic morphology relative to sham control values of ∼20%. The level of apoptosis reached a plateau at a dose of 25 mGy, with the level of apoptosis approximately twice that of the sham control. This increase in percent apoptosis was not observed when 208F cells were irradiated with a dose of 0.04 mGy.

The increase in transformed cell apoptosis is through stimulation of intercellular induction of apoptosis. To determine whether the apoptosis observed in nonirradiated transformed cells cocultured with irradiated 208F cells occurs by mechanism(s) of intercellular induction of apoptosis previously described (14), a variety of scavengers and inhibitors of ROS were used. Figure 2A and B shows that addition of either the HOCl scavenger taurine or the ONOO decomposition catalyst FeTPPS to the medium leads to a significant reduction in the number of transformed cells undergoing apoptosis after coculture with 208F cells irradiated with either 0.5-Gy γ-rays or 0.5-Gy α-particles, from ∼48% down to around 20%. When taurine and FeTPPS were used in tandem, a greater reduction in the level of apoptosis was observed, with the apoptotic levels reduced to ∼10% to 15%, a level similar to that seen for the sham-irradiated controls incubated with taurine and FeTPPS. In all these cases, incubation with both FeTPPS and HOCl reduced the percent apoptosis to background levels observed in the negative controls, where intercellular induction of apoptosis does not occur due to the absence of 208F cells. Figure 2C shows that the NADPH oxidase inhibitor apocynin also leads to a significant reduction in the level of apoptosis in nonirradiated transformed cells down to ∼15%, approaching the background levels observed in the negative controls. These results with ROS inhibitors are consistent with the model of intercellular induction of apoptosis in the absence of radiation, involving at least two ROS/reactive nitrogen species (RNS) pathways of intercellular induction of apoptosis, and show that the pathways are stimulated by ionizing radiation.

Figure 2.

Percentage apoptosis scored in nonirradiated transformed cells cocultured for 68 h in the presence or absence of 25 mmol/L of taurine and/or 10 μmol/L FeTPPS, with 208F cells irradiated with 0.5 Gy γ-rays (A) or 0.5 Gy α-particles (B). C, results for transformed cells cocultured with irradiated or nonirradiated 208F cells for 65 h in the presence or absence of the NADPH oxidase inhibitor apocynin. Columns, mean taken from triplicate repeats carried out in a single experiment; bars, intraexperimental variation. Each experiment was repeated on at least two separate occasions and the same trends were observed. Negative controls represent apoptosis scored in transformed cells in the absence of 208F cells. Sham controls refer to apoptosis scored in transformed cells cocultured in the presence of nonirradiated 208F cells.

Figure 2.

Percentage apoptosis scored in nonirradiated transformed cells cocultured for 68 h in the presence or absence of 25 mmol/L of taurine and/or 10 μmol/L FeTPPS, with 208F cells irradiated with 0.5 Gy γ-rays (A) or 0.5 Gy α-particles (B). C, results for transformed cells cocultured with irradiated or nonirradiated 208F cells for 65 h in the presence or absence of the NADPH oxidase inhibitor apocynin. Columns, mean taken from triplicate repeats carried out in a single experiment; bars, intraexperimental variation. Each experiment was repeated on at least two separate occasions and the same trends were observed. Negative controls represent apoptosis scored in transformed cells in the absence of 208F cells. Sham controls refer to apoptosis scored in transformed cells cocultured in the presence of nonirradiated 208F cells.

Close modal

Addition of TGF-β neutralizing antibody protects against radiation-induced apoptosis. Based on the involvement of at least two distinct ROS/RNS signaling pathways and a large amount of supporting literature (19, 27), TGF-β was identified as a candidate for induction of apoptosis by radiation. Figure 3A shows that incubation with 10 μg/mL TGF-β neutralizing antibody leads to a substantial reduction in the percent apoptosis scored in nonirradiated transformed cells. When nonirradiated transformed cells were cocultured with 208F cells irradiated with either 0.5-Gy γ-rays or 0.5-Gy α-particles, the level of apoptosis is reduced from ∼55% to ∼27%. In sham controls, when nonirradiated transformed cells were cocultured with nonirradiated 208F cells, a reduction from ∼35% down to ∼20% was observed.

Figure 3.

Apoptosis scored in nonirradiated transformed cells cocultured for 65 h with 208F cells irradiated with 0.5 Gy of either γ-rays or α-particles. A, cells were cocultured in the presence or absence of 10 μg/mL TGF-β neutralizing antibody. B, cells were cocultured in the presence or absence of 10 mmol/L of the caspase-3 inhibitor z-DEVD-fmk. Columns, mean taken from triplicate repeats carried out in a single experiment; bars, intraexperimental variation. Each experiment was repeated on at least two separate occasions and the same trends were observed. Negative control represents apoptosis scored in transformed cells in the absence of 208F cells. Sham controls refer to apoptosis scored in transformed cells cocultured in the presence of nonirradiated 208F cells.

Figure 3.

Apoptosis scored in nonirradiated transformed cells cocultured for 65 h with 208F cells irradiated with 0.5 Gy of either γ-rays or α-particles. A, cells were cocultured in the presence or absence of 10 μg/mL TGF-β neutralizing antibody. B, cells were cocultured in the presence or absence of 10 mmol/L of the caspase-3 inhibitor z-DEVD-fmk. Columns, mean taken from triplicate repeats carried out in a single experiment; bars, intraexperimental variation. Each experiment was repeated on at least two separate occasions and the same trends were observed. Negative control represents apoptosis scored in transformed cells in the absence of 208F cells. Sham controls refer to apoptosis scored in transformed cells cocultured in the presence of nonirradiated 208F cells.

Close modal

Confirmation of morphologic scoring as a measure of apoptosis. All quantitative data on apoptosis reported here have been derived from scoring morphologic signs of apoptosis in nonirradiated transformed cells. The morphologic criteria used for scoring were confirmed as apoptosis by comparison with staining with TUNEL (data not shown). Further quantitative confirmation is given in Fig. 3B where incubation with the caspase-3 inhibitor z-DEVD-fmk leads to almost complete inhibition of the appearance of apoptotic morphology. Whereas the caspase-3 inhibitor was used purely to confirm the inhibition of morphologic features of apoptosis, this result does indicate that there is little or no caspase-independent apoptosis occurring.

In this study, we have shown that irradiation of nontransformed 208F cells with either α-particles or γ-rays leads to an increase in the levels of apoptosis in cocultured nonirradiated transformed 208Fsrc3 cells through intercellular induction of apoptosis involving ROS/NOS and TGF-β, previously characterized in the absence of radiation (14). This radiation-induced apoptosis induction selectively removes transformed cells but not nontransformed or tumorigenic cells (28) and occurs after extremely low levels of ionizing radiation. The levels of apoptosis stimulated by radiation are reported following 65 h of coculture and are representative of a single time point in a continuous kinetic process. The level of apoptosis continues to increase as the cells remain in coculture, and eventually the vast majority of transformed cells are removed from coculture by intercellular induction of apoptosis (data not shown). The kinetics are dependent on the relative cell densities of the two cell lines. Importantly, radiation stimulates the signaling involved in intercellular induction of apoptosis and therefore increases the rate at which transformed cells are removed from coculture, as indicated in Table 1 for the 65-h time point.

Relatively small doses of either α-particles or γ-rays are sufficient to stimulate intercellular induction of apoptosis and the effect saturates at extremely low doses (see Fig. 1). This saturation at low doses is a feature that has been observed in studies looking at nontargeted affects of low-dose ionizing radiation for a variety of end points including clonogenic survival (29), proliferating cell nuclear antigen expression (30), and micronuclei formation (31). The sensitivity of the present system to such low doses of radiation is novel, with relatively few reports showing biological effects at doses in the region of 2 mGy and below, as is the case here. Interestingly, Redpath et al. (32) have reported a reduction in transformation frequency as an adaptive response at doses as low as 1 mGy γ-ray radiation. This latter effect might reflect a reduction of transformation frequency per se at low doses of radiation or, speculatively, in the light of our findings, as an elimination of transformed cells triggered by low-dose radiation. In both scenarios, the outcome would be a decreased number of transformed cells.

The proposed mechanism for radiation-stimulated intercellular induction of apoptosis is shown in Fig. 4. The first step is irradiation of the nontransformed cells with ionizing radiation where small doses of either α-particles or γ-rays are sufficient to stimulate intercellular induction of apoptosis above naturally occurring level, with this stimulation saturating at very low doses.

Figure 4.

Proposed mechanism for radiation-stimulated intercellular induction of apoptosis. 1, 208F cells exposed to ionizing radiation. At low doses of α-particles, this will result in some cells receiving an α-particle track, whereas others remain unirradiated. 2, irradiated cells increase the amount of active TGF-β present in the media either through increased activation of existing latent TGF-β present in the media or through secretion of latent TGF-β, which is subsequently activated. 3 and 4, TGF-β acts in an autocrine fashion to stimulate the irradiated cell to produce PO and •NO, which in turn leads to ROS/RNS signaling culminating in transformed cell apoptosis. 5, the active TGF-β released into the media by the irradiated cells will also stimulate nonirradiated neighbor cells to produce PO and •NO (steps 3 and 4). *, key to the selectivity of the signaling system is superoxide produced by membrane-bound NADPH oxidase expressed constitutively in the transformed cells. The nontransformed cells are not susceptible to apoptosis induction as they do not produce the superoxide required to produce the final apoptosis-inducing ROS (•OH and ONOO).

Figure 4.

Proposed mechanism for radiation-stimulated intercellular induction of apoptosis. 1, 208F cells exposed to ionizing radiation. At low doses of α-particles, this will result in some cells receiving an α-particle track, whereas others remain unirradiated. 2, irradiated cells increase the amount of active TGF-β present in the media either through increased activation of existing latent TGF-β present in the media or through secretion of latent TGF-β, which is subsequently activated. 3 and 4, TGF-β acts in an autocrine fashion to stimulate the irradiated cell to produce PO and •NO, which in turn leads to ROS/RNS signaling culminating in transformed cell apoptosis. 5, the active TGF-β released into the media by the irradiated cells will also stimulate nonirradiated neighbor cells to produce PO and •NO (steps 3 and 4). *, key to the selectivity of the signaling system is superoxide produced by membrane-bound NADPH oxidase expressed constitutively in the transformed cells. The nontransformed cells are not susceptible to apoptosis induction as they do not produce the superoxide required to produce the final apoptosis-inducing ROS (•OH and ONOO).

Close modal

The second step in our proposed mechanism is the induction of TGF-β signaling in irradiated nontransformed cells (step 2) leading to an increase in extracellular levels of active TGF-β. The identification of radiation-induced stimulation of TGF-β as a candidate for the radiation-induced signal leading to stimulation of intercellular induction of apoptosis was confirmed using a TGF-β neutralizing antibody (Fig. 3A) that inhibits intercellular induction of apoptosis under all conditions following irradiation with either γ-rays or α-particles. It had previously been shown that pretreatment of nontransformed cells with TGF-β stimulates ROS signaling involved in intercellular induction of apoptosis in the absence of radiation (19, 33). That TGF-β signaling pathways are saturated was shown through the combined effect of TGF-β pretreatment followed by irradiation. The increased level of intercellular induction of apoptosis is not synergistic, consistent with saturation of the TGF-β signaling pathways (data not shown). Barcellos-Hoff (27) has shown that TGF-β is indeed activated postirradiation. TGF-β has also been implicated specifically in low-dose effects of ionizing radiation including cell cycle gene expression (20) and the adaptive response (34). The data presented here are consistent with postirradiation activation of TGF-β.

Two separate ROS pathways induced postradiation are proposed as indicated in steps 3 and 4 in Fig. 4 because both taurine, effective on the HOCl signaling pathway, and FeTPPS, acting on the NO/peroxynitrite signaling pathway, are inhibitors of apoptosis (see Fig. 2A and B). Steps 3 and 4 are coupled to the induction of TGF-β in step 2. TGF-β acts in an autocrine fashion to stimulate the irradiated cell to produce peroxidase (PO) and •NO, which in turn leads to ROS/RNS signaling culminating in transformed cell apoptosis. The TGF-β activity produced by the irradiated cells also stimulates nonirradiated neighbor cells to produce PO and •NO (steps 3 and 4). The crucial step to the selectivity of the signaling system is superoxide produced by membrane-bound NADPH oxidase expressed constitutively only in the transformed cells. This step leads to the formation of an environment of superoxide around the transformed cells. The nontransformed cells are not susceptible to apoptosis induction because they do not have constitutively expressed NADPH oxidase and therefore do not produce sufficient superoxide required to produce the final apoptosis-inducing ROS (•OH and ONOO). The key role of NADPH oxidase was confirmed through the significant reduction in the level of apoptosis observed in the presence of apocynin, an NADPH oxidase inhibitor. This is consistent with previous work by our group confirming the vital role of NADPH oxidase in the selectivity of the system (15).

The α-particle dose response, in particular (Fig. 1B), supports the idea of radiation-induced TGF-β signaling. Radiation-induced apoptosis induction is observed at a dose as low as 0.29 mGy and saturates at a dose of 25 mGy. Based on the mean cell area for the nontransformed 208F cells, the mean number of α-particle track traversals per cell, or in the case of doses as low as these, the fraction of cells that receive an α-particle track, has been calculated and is shown in Table 2. At these doses, not every nontransformed 208F cell is traversed by an α-particle. For instance, at a dose of 2.5 mGy, <10% of the cells are traversed by an α-particle; the radiation-induced increase in intercellular induction of apoptosis is similar to that observed at a dose of 500 mGy, where each cell is traversed by ∼19 α-particle traversals. This finding strongly suggests that at these doses, the signal produced following radiation once saturated is sufficient to stimulate all nontransformed cells in a population, consistent with the activation of TGF-β post irradiation and the potential of TGF-β to autoinduce its own synthesis and release (35, 36). This autoinduction would lead to the stimulation of all nontransformed cells in the population and thus allow optimal intercellular ROS signaling, as indicated in step 5 in Fig. 4. Finally, the fact that even an excessive dose of radiation, such as 500 mGy α-particles, does not cause a subsequent increase in intercellular induction of apoptosis beyond the early saturation level is also consistent with TGF-β signaling. The precise cellular target for ionization that triggers TGF-β signaling postirradiation is unknown, but the dose threshold for γ-rays suggests that it is quite a ubiquitous target within the cell, as the effect is triggered by as little as 2 mGy where each cell receives, on average, 23 electron tracks.3

3

The event frequency ϕ*(0) (the frequency of events of any size per unit absorbed dose) has been calculated for 60Co γ-rays interacting with 208F cells using the methods explained in Appendix A, ICRU report 36, for convex volumes. The cells were measured, using confocal laser scanning microscopy, effectively to be cylinders with a mean cross-section area of 746 μm 2 and a height of 5 μm. From ICRU 36, ϕ*(0) = k × yFbar where k is a constant equal to 0.6408/S (S is the surface area of the convex volume in square micrometer) and yFbar is the frequency mean lineal energy of the radiation and for 60Co γ-rays is 0.256 keV μm−1. Taking these values gives an event frequency for 60Co γ-rays interacting with 208F cells to be 12,045 events per Gy.

Like many cytokine signaling pathways, TGF-β signaling is tightly controlled (37). This early saturation could be accounted for in a number of ways. It may reflect saturation of TGF-β receptors in the nontransformed cells. Equally, there are known examples of negative feedback loops in TGF-β signaling pathways, including the Smad7 feedback loop (38). Alternatively, the saturation seen may be due to saturation of effector molecule release, such as PO and •NO.

Table 2.

Comparison of calculated average α-particle track traversals, for 208F nuclei and whole cells, and relative levels of radiation-stimulated intercellular induction of apoptosis

α-Particle dose (mGy)Calculated average α-particle traversals per cell (% cells hit)Calculated average α-particle traversals per nucleus (% nuclei hit)Radiation-induced intercellular induction of apoptosis
0.04 0.0015 (0.15%) 0.0004 (0.004%) − 
0.29 0.011 (1.1%) 0.003 (0.03%) 
2.5 0.096 (9.6%) 0.023 (2.3%) +++ 
25 0.96 (96%) 0.23 (23%) ++++ 
500 19.2 (100%) 4.55 (100%) ++++ 
α-Particle dose (mGy)Calculated average α-particle traversals per cell (% cells hit)Calculated average α-particle traversals per nucleus (% nuclei hit)Radiation-induced intercellular induction of apoptosis
0.04 0.0015 (0.15%) 0.0004 (0.004%) − 
0.29 0.011 (1.1%) 0.003 (0.03%) 
2.5 0.096 (9.6%) 0.023 (2.3%) +++ 
25 0.96 (96%) 0.23 (23%) ++++ 
500 19.2 (100%) 4.55 (100%) ++++ 

The involvement of TGF-β is particularly interesting when considered in the context of the work of Barcellos-Hoff (5), who emphasizes the importance of the multitude of cell types within a tissue in producing a tumor and, in particular, their interaction through signaling. The system described here is simpler than a tissue, but the findings are still important when extrapolated in vivo to a tissue or whole organism level. In an in vivo situation, a single transformed cell or a small population of transformed cells arising in a tissue would be subject to intercellular induction of apoptosis signaling from a vast excess of neighboring nontransformed cells. Assuming the same radiation-stimulated signaling occurs in vivo, low-dose irradiation of the tissue would stimulate the signaling, increasing the likelihood of the transformed cells being removed before they can progress to become tumorigenic. Thus, intercellular induction of apoptosis may represent an important natural control mechanism for removing transformed cells before they can proceed to tumorigenesis. This is particularly important in light of the fact that tumor cells do not respond to the process of intercellular induction of apoptosis because they develop signaling that quenches the ROS signaling involved in intercellular induction of apoptosis (28). Therefore, the removal of early transformed cells by their nontransformed neighbors may represent an important window of opportunity in the avoidance of cells becoming tumorigenic. When the observation of radiation-induced intercellular induction of apoptosis is interpreted in this light, even low doses of both high-LET and low-LET ionizing radiation may increase the probability that precancerous cells are removed before they progress through tumorigenesis. This would represent a positive effect of low doses of radiation, by which the traditional linear-no-threshold models of radiation-related cancer risk would overestimate low-dose radiation risk, as described by Brenner et al. (6). Positive effects of low-dose ionizing radiation previously described (10, 39) in terms of radiation hormesis include a reduction in transformation frequency following low doses of radiation. The process of intercellular induction of apoptosis described here may also represent a radiation hormesis effect.

The findings presented confirm that radiation-induced TGF-β signaling occurs at very low doses of high-LET or low-LET ionizing radiation and stimulates the previously well-characterized process of intercellular induction of apoptosis to selectively remove transformed cells from coculture. The selectivity of this system is dependent on the constitutive expression of membrane-bound NADPH oxidase, which is a characteristic of a number of transformed cell lines overexpressing genes in the NADPH oxidase activation pathway, including src, as discussed in this study, and ras (40, 41). The low-dose saturation of radiation-induced apoptosis in pretransformed cells has potential implications for the effect of low doses of ionizing radiation on a naturally occurring anticancer defense mechanism.

Grant support: European Commission RISC-RAD project contract F16-CT-2003-508842 and Medical Research Council studentship (D.I. Portess).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank David Stevens for α-particle irradiations, track traversal calculations, and aid in the design of the custom inserts and Richard Doull for dosimetry and designing of the lead attenuation setup.

1
Hanahan D, Weinberg RA. The hallmarks of cancer.
Cell
2000
;
100
:
57
–70.
2
Park CC, Bissell MJ, Barcellos-Hoff MH. The influence of the microenvironment on the malignant phenotype.
Mol Med Today
2000
;
6
:
324
–9.
3
Lane DP. Cancer. p53, guardian of the genome.
Nature
1992
;
358
:
15
–6.
4
Evan G, Littlewood T. A matter of life and cell death.
Science
1998
;
281
:
1317
–22.
5
Barcellos-Hoff MH. It takes a tissue to make a tumor: epigenetics, cancer and the microenvironment.
J Mammary Gland Biol Neoplasia
2001
;
6
:
213
–21.
6
Brenner DJ, Doll R, Goodhead DT, et al. Cancer risks attributable to low doses of ionizing radiation: Assessing what we really know.
Proc Natl Acad Sci U S A
2003
;
100
:
13761
–6.
7
Hall EJ. Radiation, the two-edged sword: cancer risks at high and low doses.
Cancer J
2000
;
6
:
343
–50.
8
Azzam EI, De Toledo SM, Spitz DR, Little JB. Oxidative metabolism modulates signal transduction and micronucleus formation in bystander cells from α-particle-irradiated normal human fibroblast cultures.
Cancer Res
2002
;
62
:
5436
–42.
9
Zhou H, Randers-Pehrson G, Waldren CA, et al. Induction of a bystander mutagenic effect of α particles in mammalian cells.
Proc Natl Acad Sci U S A
2000
;
97
:
2099
–104.
10
Redpath JL, Short SC, Woodcock M, Johnston PJ. Low-dose reduction in transformation frequency compared to unirradiated controls: the role of hyper-radiosensitivity to cell death.
Radiat Res
2003
;
159
:
433
–6.
11
Mothersill C, Seymour C. Medium from irradiated human epithelial cells but not human fibroblasts reduces the clonogenic survival of unirradiated cells.
Int J Radiat Biol
1997
;
71
:
421
–7.
12
Lyng FM, Seymour CB, Mothersill C. Production of a signal by irradiated cells which leads to a response in unirradiated cells characteristic of initiation of apoptosis.
Br J Cancer
2000
;
83
:
1223
–30.
13
Bauer G. Signaling and proapoptotic functions of transformed cell-derived reactive oxygen species.
Prostaglandins Leukot Essent Fatty Acids
2002
;
66
:
41
–56.
14
Bauer G. Reactive oxygen and nitrogen species: efficient, selective, and interactive signals during intercellular induction of apoptosis.
Anticancer Res
2000
;
20
:
4115
–39.
15
Herdener M, Heigold S, Saran M, Bauer G. Target cell-derived superoxide anions cause efficiency and selectivity of intercellular induction of apoptosis.
Free Radic Biol Med
2000
;
29
:
1260
–71.
16
Haberstroh K, Heigold S, Bauer G. Transformed cell-derived reactive oxygen species support and inhibit nitric oxide-mediated apoptosis induction.
Int J Oncol
2002
;
21
:
145
–51.
17
Heigold S, Sers C, Bechtel W, et al. Nitric oxide mediates apoptosis induction selectively in transformed fibroblasts compared to nontransformed fibroblasts.
Carcinogenesis
2002
;
23
:
929
–41.
18
Langer C, Jurgensmeier JM, Bauer G. Reactive oxygen species act at both TGF-β-dependent and -independent steps during induction of apoptosis of transformed cells by normal cells.
Exp Cell Res
1996
;
222
:
117
–24.
19
Jurgensmeier JM, Schmitt CP, Viesel E, Hofler P, Bauer G. Transforming growth factor β-treated normal fibroblasts eliminate transformed fibroblasts by induction of apoptosis.
Cancer Res
1994
;
54
:
393
–8.
20
Azzam EI, de Toledo SM, Gooding T, Little JB. Intercellular communication is involved in the bystander regulation of gene expression in human cells exposed to very low fluences of α particles.
Radiat Res
1998
;
150
:
497
–504.
21
Park CC, Henshall-Powell RL, Erickson AC, et al. Ionizing radiation induces heritable disruption of epithelial cell interactions.
Proc Natl Acad Sci U S A
2003
;
100
:
10728
–33.
22
Shao C, Furusawa Y, Kobayashi Y, Funayama T, Wada S. Bystander effect induced by counted high-LET particles in confluent human fibroblasts: a mechanistic study.
FASEB J
2003
;
17
:
1422
–7.
23
Little JB, Azzam EI, de Toledo SM, Nagasawa H. Bystander effects: intercellular transmission of radiation damage signals.
Radiat Prot Dosimetry
2002
;
99
:
159
–62.
24
Goodhead DT, Bance DA, Stretch A, Wilkinson RE. A versatile plutonium-238 irradiator for radiobiological studies with α-particles.
Int J Radiat Biol
1991
;
59
:
195
–210.
25
Hipp ML, Bauer G. Intercellular induction of apoptosis in transformed cells does not depend on p53.
Oncogene
1997
;
15
:
791
–7.
26
Beck E, Schafer R, Bauer G. Sensitivity of transformed fibroblasts for intercellular induction of apoptosis is determined by their transformed phenotype.
Exp Cell Res
1997
;
234
:
47
–56.
27
Barcellos-Hoff MH. Latency and activation in the control of TGF-β.
J Mammary Gland Biol Neoplasia
1996
;
1
:
353
–63.
28
Engelmann I, Eichholtz-Wirth H, Bauer G. Ex vivo tumor cell lines are resistant to intercellular induction of apoptosis and independent of exogenous survival factors.
Anticancer Res
2000
;
20
:
2361
–70.
29
Seymour CB, Mothersill C. Relative contribution of bystander and targeted cell killing to the low-dose region of the radiation dose-response curve.
Radiat Res
2000
;
153
:
508
–11.
30
Hill MA, Ford JR, Clapham P, et al. Bound PCNA in nuclei of primary rat tracheal epithelial cells after exposure to very low doses of plutonium-238 α particles.
Radiat Res
2005
;
163
:
36
–44.
31
Shao C, Stewart V, Folkard M, Michael BD, Prise KM. Nitric oxide-mediated signaling in the bystander response of individually targeted glioma cells.
Cancer Res
2003
;
63
:
8437
–42.
32
Redpath JL, Liang D, Taylor TH, Christie C, Elmore E. The shape of the dose-response curve for radiation-induced neoplastic transformation in vitro: evidence for an adaptive response against neoplastic transformation at low doses of low-LET radiation.
Radiat Res
2001
;
156
:
700
–7.
33
Haufel T, Dormann S, Hanusch J, Schwieger A, Bauer G. Three distinct roles for TGF-β during intercellular induction of apoptosis: a review.
Anticancer Res
1999
;
19
:
105
–11.
34
Iyer R, Lehnert BE. Low dose, low-LET ionizing radiation-induced radioadaptation and associated early responses in unirradiated cells.
Mutat Res
2002
;
503
:
1
–9.
35
Van Obberghen-Schilling E, Roche NS, Flanders KC, Sporn MB, Roberts AB. Transforming growth factor β 1 positively regulates its own expression in normal and transformed cells.
J Biol Chem
1988
;
263
:
7741
–6.
36
Kim SJ, Angel P, Lafyatis R, et al. Autoinduction of transforming growth factor β 1 is mediated by the AP-1 complex.
Mol Cell Biol
1990
;
10
:
1492
–7.
37
Wormald S, Hilton DJ. Inhibitors of cytokine signal transduction.
J Biol Chem
2004
;
279
:
821
–4.
38
Nakao A, Afrakhte M, Moren A, et al. Identification of Smad7, a TGFβ-inducible antagonist of TGF-β signaling.
Nature
1997
;
389
:
631
–5.
39
Azzam EI, de Toledo SM, Raaphorst GP, Mitchel RE. Low-dose ionizing radiation decreases the frequency of neoplastic transformation to a level below the spontaneous rate in C3H 10T1/2 cells.
Radiat Res
1996
;
146
:
369
–73.
40
Engelmann I, Dormann S, Saran M, Bauer G. Transformed target cell-derived superoxide anions drive apoptosis induction by myeloperoxidase.
Redox Rep
2000
;
5
:
207
–14.
41
Irani K, Xia Y, Zweier JL, et al. Mitogenic signaling mediated by oxidants in Ras-transformed fibroblasts.
Science
1997
;
275
:
1649
–52.