A computational model for quantitative analysis of cell cycle arrest and its contribution to overall growth inhibition by anticancer agents.

摘要

Most anticancer agents induce cell cycle arrest (cytostatic effect) and cell death (cytotoxic effect), resulting in the inhibition of population growth of cancer cells. When asynchronous cells are to be examined, the currently used flow cytometric method can not provide checkpoint-specific and quantitative information on the drug-induced cell cycle arrest. Hence, despite its significance, no good method to analyze in detail the mechanism of cell cycle arrest and its contribution to overall growth inhibition induced by an anticancer agent has yet been established. We describe in this study the development of a discrete time (Markov model)-based computational model for cell cycle progression / arrest with transition probability (TP(i)) as a model parameter. TP(i) was calculated using model equations that include easily measurable parameters such as the fraction of cells in each cell cycle phase and population doubling time. The TP(i) was then used to analyze checkpoint-specific and quantitative changes in cell cycle progression. We also used TP(i) in a Monte-Carlo simulation to predict growth inhibition caused by cell cycle arrest only. Human SCLC cells (SBC-3) exposed to UCN-01 were used to validate the model. The model-predicted growth curves agreed with the observed data for SBC-3 cells not treated or treated at a cytostatic concentration (0.2 mM) of UCN-01, indicating validity of the present model. The changes in TP(i) indicated that UCN-01 reduced the G(1)-to-S transition rate and increased the S-to-G(2) / M and G(2) / M-to-G(1) transition rates of SBC-3 cells in a concentration- and time-dependent manner. When the model-predicted growth curves were compared with the observed data for cells treated at a cytotoxic concentration (2 mM), they suggested that 22% out of 65% and 32% out of 73% of the growth inhibition could be attributed to the cell cycle arrest effect after 48 h and 72 h exposure, respectively. In conclusion, we report here the establishment of a novel method of analysis that can provide checkpoint-specific and quantitative information about cell cycle arrest induced by an anticancer agent and that can be used to assess the contribution of cell cycle arrest effect to the overall growth inhibition.