On Evidence Supporting a Deterministic Process of Bacterial Growth

Recent evidence supporting a proposed model (Koch & Schaechter, 1962) for the control of bacterial cell division is reviewed. Calculations from the published work of others are presented which show that the standard deviation of length of time between a cell division and Nth cell division does not increase, at least up to N = 9. This finding implies that each cell has an excellent clock, which is handed to the daughters without significant error, and that the observed fluctuations in age of cells at division are largely due to an additional fluctuation associated with cell division, but not timing it. In terms of our model, it is strong support for the deterministic growth of cell constituents and the equipartition of cell constituents at division. An expansion of the original model is considered which accounts for the difference between the original model and the finding regarding the correlation coefficients of the ages of mothers and daughters and that between sisters, and the skewed nature of the age-distribution curve. In the new version of the model, the assumption that there are random fluctuations in the critical size at division is replaced by the assumption that the fluctuations are not completely random but that there is a moderately positive correlation between the sizes of divisions of successive generations.