T-47D cells were gamma-irradiated at acute dose rates by a Co-60 source, and their low-dose response was studied by clonogenic assay. T-47D cells have previously been found to express low-dose hyper-radiosensitivity (HRS) [Edin et al., 2007; 2008a; 2008b; 2008c]. Surprisingly, no HRS could be detected in the present study. The reason for this contradiction was not identified despite repeated investigations, but it was speculated whether exposure to alkaline conditions during plating might be involved.
HRS is thought to reflect the failure of an early G2 checkpoint to arrest G2-phase cells irradiated with less than ~0.3 Gy, and exaggerated HRS responses would be expected when irradiating G2-enriched cell populations [Marples et al., 2003; Short et al., 2003]. The low-dose response of T-47D G2-enriched cell populations was measured using Hoechst 33342-based fluorescence-activated cell sorting (FACS). For comparison, similar experiments were performed on T-47D-P cells. These cells had been given a priming dose of 0.3 Gy at a low dose rate (0.3 Gy/h) 3 years earlier, after which HRS was permanently eliminated [Edin et al., 2007; Edin et al., 2008b]. Consistent with the dose-response measurements on asynchronous cultures, no signs of HRS were observed in the G2-enriched T-47D cell populations. This indicates that the removal of HRS was not caused by an insufficient amount of G2 cells, an observation that was corroborated by flow-cytometric measurements of DNA content in asynchronous cultures at the time of irradiation (~18 hours after trypsinization). The experiments with G2 enrichment also demonstrated that G2 is a relatively radioresistant phase for T-47D cells, i.e., the radiosensitivity in G2 phase is not higher than the average sensitivity in other phases of the cell cycle. The radiosensitivity of T-47D-P cells appeared to be slightly increased in G2 phase, although the dose response of asynchronous T-47D-P cells corresponded well with that of asynchronous unprimed cultures.
In studies by Edin et al. [2007; 2008c; 2008b] and Fenne  the low-dose survival of T-47D-P cells was higher than predicted by the linear-quadratic model, with surviving fractions even exceeding 1 for doses up to ~0.3 Gy. A similar consistent trend of higher cloning efficiencies in low-dose irradiated flasks than in controls was not observed in the present study, but surviving fractions at the lowest doses did tend to exceed 1 when medium was changed weekly. This corresponds well with a hypothesis previously put forward by Fenne , suggesting that challenge-irradiation improves the cellular attachment to the flask surface. However, the data are much too sparse to establish a connection between the elevated survival and change of medium.
Another hypothesis explains the enhanced clonogenicity of irradiated T-47D-P cells by radiation-induced recruitment of G0-phase cells into the cell cycle [Fenne, 2008]. To test this hypothesis FACS was employed to select populations of G0/G1-phase cells that were subsequently incubated for different periods of time, ranging from 14-24 hours. The cell-cycle distributions were then measured by flow cytometry, and the fraction of cells remaining in G0/G1 in unirradiated controls was compared to the corresponding fraction in cell populations given a small radiation dose (0.2 Gy). No stimulation of irradiated G0 cells into the cell cycle could be observed using this assay, but the measurements were not precise enough to exclude the possibility that such recruitment does occur.
T-47D-P cells that had been selected in G0/G1 phase and irradiated with 0.2 Gy showed a delayed exit from G1 phase. This was evident from significantly greater G0/G1 fractions in irradiated than in unirradiated cell populations 14 hours (P = 0.0065) and 15 hours (P = 0.021) after plating, and it was also reflected by delayed entry into G2/M at later times. Thus, it appears that the cells are arrested in G1 in response to very low radiation doses. A rapid p53-independent checkpoint pathway has been reported to block the G1/S transition transiently in response to genotoxic stress [Bartek and Lukas, 2001; Kastan and Bartek, 2004], and it should be investigated whether involvement of this pathway can explain the observed delay.