The proliferative response of the coagulating gland of the castrated mouse under continuous androgen stimulation: an experimental and computer simulation model

Abstract

The proliferative response of the coagulating gland of the castrated male mouse has been examined during continuous treatment with testosterone propionate. Fourteen days after castration, s.c. daily injections of testosterone propionate were begun. Mitotic (Im) and labelling (IL) index values were obtained at 3 h intervals for up to 100 h after the initial injection. These showed a biphasic response, in which IL reached a maximum at 30 and 70 h, and Im at approximately 45 and 75 h. Fraction-labelled mitoses (FLM) curves were begun 24, 48, and 72 h after the first androgen injection. In each curve the first wave of labelled mitoses rose to 100% and showed a square form indicating little spread in the durations of the G2 and S phases. Values of 7.5, 1.3 and 0.7 h were obtained for the durations of DNA synthesis (ts), the post-synthetic period (tG2) and of mitosis (tm) respectively. In none of the FLM curves was it possible to demonstrate a second wave of labelled mitoses and direct measurement of the cell cycle time (Tc) was not obtained. Continuous tritiated thymidine labelling indices revealed that after a latent period of 25 h, DNA synthesis began and labelling rose rapidly to 80% by 45 h and then more slowly to 95% by 97 h. Cell population changes during androgen stimulation estimated from measurements of total glandular DNA indicated that the number of cells present in the glands remained constant during the first 30 h after stimulation and thereafter increased to approximately 2-3 times the original value. The data are compared with a mathematical model which assumes that the cell population of castrated mice when stimulated passes from a GO compartment through successive waves of DNA synthesis and mitosis. After each cell division the cells may leave or remain in the proliferative cycle. This model has been subjected to computer simulation using the cell cycle parameters obtained in the kinetic experiments. There was good agreement between the stimulation and experimental results in the Im and IL curves, continuous labelling, and total cell number experiments. The simulation of FLM curves was less successful. Although the first wave of labelled mitoses was clearly seen the model predicts a distinct second wave of labelled mitoses. It is concluded that this does not appear because of variation in the duration of G1.