Tests of the Double-Strand Break, Lethal--Potentially Lethal and Repair--Misrepair Models for Mammalian Cell Survival Using Data for Survival as a Function of Delayed-Plating Interval for Log-Phase Chinese Hamster V79 Cells

Our data (Reddy et al., Radiat. Res. 141, 252-258, 1995) on the kinetics of the repair of potentially lethal damage in log-phase Chinese hamster V79 cells are used to test some predictions which arise from the different assumptions of the repair-misrepair (RMR) (C. A. Tobias, Radiat. Res. 104, S77-S95, 1985), lethal-potentially lethal (LPL) (S. B. Curtis, Radiat. Res. 106, 252-270, 1986) and double-strand break (DSB) (J. Y. Ostashevsky, Radiat. Res. 118, 437-466, 1989) models. The LPL model defines the time available for repair of PLD (<tex-math>$t_{{\rm rep}}$</tex-math>) as the time taken to reach maximal survival in a delayed-plating recovery experiment. Those data show that after this time has elapsed, contrary to the expectation of the LPL model, survival can be increased by changing the medium used for delayed plating from fresh growth medium to conditioned medium. According to the RMR model, all potentially lethal lesions should also be committed by that time and be unavailable for repair in the new medium. Only the DSB model correctly predicted that PLD (= DSBs) would still be available for repair after that time. Second, data for split-dose recovery are used to predict the first-order kinetics time constant for DSB repair (<tex-math>$\tau _{{\rm DSBR}}$</tex-math>) using the DSB model (24 ± 1.5 min). This value is nearly identical to the value of 27 ± 1 min determined from the data obtained by Cheong et al. using pulsed-field gel electrophoresis (PFGE) (Mutat. Res. 274, 111-122, 1992). The value based on PFGE is used to calculate the value of <tex-math>$t_{{\rm rep}}$</tex-math> predicted by the DSB model (2.6 ± 0.1 h), which agrees with the value determined experimentally as the time when changing the delayed-plating medium from growth medium to conditioned medium no longer gives the full recovery seen with delayed plating in conditioned medium (2.5 h). However, some recovery was seen for a change in the medium (growth medium to conditioned medium) up to 5-6 h postirradiation. Reanalysis of the original data on DSB repair shows that they are consistent with two first-order repair rates (18 ± 7 min and about 52 min). These results are consistent with two pools of DSBs (or cells), each with their own <tex-math>$t_{{\rm rep}}$</tex-math>. The early <tex-math>$t_{{\rm rep}}$</tex-math>, associated with <tex-math>$\tau _{\text{fast}}$</tex-math>, is predicted to be 1.7 ± 0.7 h, and the late <tex-math>$t_{{\rm rep}}$</tex-math>, associated with τ_slow, is predicted to be about 5 h. Both values are in excellent agreement with the times at which changing from growth medium to conditioned medium no longer gives the full recovery seen in conditioned medium only (the early <tex-math>$t_{{\rm rep}}$</tex-math>), and the time when changing from growth medium to conditioned medium produces no further increase in survival (the late <tex-math>$t_{{\rm rep}}$</tex-math>), respectively. It is noted that attempts to correlate radiosensitivity with the rates of DSB repair, rather than using an explicit model such as the DSB model, are unlikely to be productive since survival depends on both <tex-math>$\tau _{{\rm DSBR}}$</tex-math> and <tex-math>$t_{{\rm rep}}$</tex-math> (as defined in the DSB model) and the latter may be the more important determinant of radiosensitivity (as it appears to be for ataxia telangiectasia cells compared to normal fibroblasts and for irs compared to V79 cells).