Investigation of low dose effects using colony size as endpoint

Besides extensive investigations of low dose hypersensitivity after low-LET radiation, more recently similar effects have been also detected after high-LET radiation using low energy He ions [1] or high energy carbon ions [2]. Usually, a cutoff value in colony size is used to distinguish surviving from inactivated cells. However, essential information is lost when using only two categories for the analysis. Therefore, we investigated the effects of low dose irradiation based on the average colony size as a function of dose and time after irradiation, thus taking into account all colonies for the analysis without applying cutoff values. Colony area was taken as a measure for the cell number, where the appropriate calibration was based on measurements for control cells. In this case, we found a much more pronounced effect of hypersensitivity after photon irradiation when compared with the standard survival assay analysis. A decrease of average colony size to about 80% of the control value was found at about 0.2 Gy, followed by a steep increase even slightly above the control level at approx. 0.5 Gy. However, for high energy carbon irradiation the effect was somewhat less pronounced. This could be explained by the stochastic distribution of particle hits in the case of low doses and thus low fluences, leading to significant cell-to-cell variations of damage. As a consequence, the structure of dose-response curves observed after a more homogenous distribution of damages in the case of photon radiation will be blurred in the case of particle radiation. The mechanism behind the oscillating structure as observed after photon radiation remains unclear up to now. In order to exclude any artefacts, we have established other methods for the determination of the cell number per colony. This is important, because e.g. it could be possible that cell cycle delays in G2-phase could be responsible for larger colony areas without higher cell numbers due to the correlation between cell size and cycle stage. Therefore, we have developed methods for absolute cell counting per colony based on nuclear staining with HOECHST 33258, facilitating the identification and counting of individual cells in the colony. The new method allows a more precise cell counting, as has been demonstrated by comparison with manual counting and with the previously established method based on the colony area determination (Fig. 2). First irradiation experiments revealed a good correlation between the mean values determined according to the area method and the absolute cell counting method, indicating that the effects demonstrated in Fig. 1 are not due to artefacts (Fig.3). However, due to the limited number of experiments performed with the new technique up to now, the number of colonies taken in to account for the analysis was considerably lower compared to the data shown in Fig. 1, and the oscillating structure could not yet be fully reproduced. Furthermore, it has to be taken into account here, that conditions had slightly changed, because for technical reasons we had to use a different V79 subline for these experiments. In addition, 250 kV X-rays have been used in the recent experiments in contrast to 6 MV linear accelerator photons in the former experiments. References: 1. Tsoulou, E., et al. (2001), Int. J. Radiat. Biol. 77, 1133-1139 2. Böhrnsen, G., et al. (2002), Int. J. Radiat. Biol. 78, 259-266 3. Spadinger, I., et al., (1993), Radiat. Res. 138, S21-S24 Fig. 1 Detection of hypersensitivity after photon (top) and 100 MeV/u carbon ion irrradiation using average colony size as endpoint.