Synchronous cultures from the baby machine. A model for animal cells

A baby-machine system that produces newborn Escherichia coli cells from cultures immobilised on a membrane was developed many years ago in an attempt to attain optimal synchrony with minimal disturbance of steady-state growth, and a model designed to characterise the nature and quality of the synchrony of such cells in a quantitative manner has been published. The baby machine has now been adapted for animal cells, and the present article is an attempt to modify the model to include these cells as well. The model consists of five elements, giving rise to five adjustable parameters (and a proportionality constant): a major, essentially synchronous group of cells with ages distributed normally about zero; a minor, random component from a steady-state population on the membrane that had undergone only very little age selection during the elution process; a fixed background count, to allow for the signals recorded by the electronic particle counter produced by debris and electronic noise; a time-shift, to account for differences between time of cell division and end of sample collection; and the coefficient of variation of the interdivision-time distribution, taken to be reciprocal-normal. It is this last feature, a reciprocal-normal rather than a Pearson type III interdivision-time distribution, that distinguishes this version of the model from its predecessor. The model is fitted by unconstrained non-linear least-squares to data from three different leukemia cell lines. The standard errors of the parameters are quite small in all cases, making their estimates highly significant; the quality of the fit is striking. The five parameters of the model can be divided into two nuisance parameters, two that are associated with the methodology and one that describes an inherent property of the cell itself; it turns out that both methodology parameters are zero in all three data sets studied. We also discuss the partition of the transition-time dispersion between the age distribution of the newborn cells and the age distribution of dividing cells and show that a reliable estimate of the corresponding parameters requires an experiment that extends over at least two and a half doubling times.