A stochastic model of pulmonary platelet production.

Normally, cells reproduce by mitosis. In mammals, a system has evolved that is unique to cell biology. The circulating blood cells called platelets are produced from parent cells called megakaryocytes, but the mechanism of platelet production is not mitosis. At present, both the site and mechanism of production are still debated. This article describes a production mechanism based on a sequence of random binary divisions of the megakaryocyte cytoplasm as it traverses the pulmonary microcirculation. Using the measured megakaryocyte cytoplasmic volumes circulating in the blood of three different mammals, rat, rabbit, and man, a computer simulation of this process is developed. The simulation depends on two separate stochastic processes. The first involves the probability that a division of a cytoplasmic particle of a specified size will occur. The second relates to the relative sizes of the two particles produced by the binary division. These two processes are controlled by three parameters only: a lower threshold L on the platelet volume, below which the probability of a binary division is zero; a parameter lambda which defines the probability of division for volumes larger than this lower threshold; and the standard deviation S of the Gaussian distribution of possible volumes created by the binary division. A minimization scheme is used to establish those values of L, lambda, and S which provide the best fit to the experimental data obtained from the different mammals.