A mathematical model of LH release from ovine pituitary cells in perifusion.

The goal of this research was to model the action of the ovine pituitary with respect to the action of two reproductive hormones, gonadotrophin-releasing hormone (GnRH) and luteinizing hormone (LH). GnRH, released by the hypothalamus, stimulates the secretion of LH from the pituitary. If the stimulus pulses are regular (low amplitude and low frequency), the response will follow a similar pattern, roughly replicating input variations in frequency and amplitude. However, during application of GnRH at greater than optimal frequencies or concentrations, or in continuous mode, the pituitary delivers a submaximal release of LH. This alteration in cell sensitivity is termed desensitization. The exact mechanisms of desensitization involved in LH release are unknown. My model consisted of a system of nonlinear differential equations and incorporated two proposed mechanisms to account for the observed behavior, a desensitized receptor and limited, available LH. A perifusion system, which permitted the examination of ovine pituitary cells in vitro, was validated and utilized to provoke desensitization and generate data for the model fit. Early experiments indicated that responsiveness could be retained for 6 days and was maximized at 0.1-1.0 nM GnRH. Desensitization was provoked in experiments involving high frequency GnRH pulses or constant GnRH input. An experiment utilizing four cell chambers examined the effect of alterations in interpulse interval on LH output and provided data for the model. The model was fit to these experimental data using nonlinear least squares regression. The presence of variation in the data undescribed by the model was demonstrated in the analysis of the mean square and substantiated in the residual analysis. Possible sources of error were model assumptions and biological or mechanical errors, e.g., channeling in the cell chamber. Time-varying biological disruptions might account for the observed autocorrelation of the residuals. The high degree of correlation among estimates of four parameters made them difficult to estimate separately. Simulations revealed that various combinations of these parameters could lead to similar release patterns for LH. The half-lives of GnRH binding and desensitization were approximately 1 and 4 mins, respectively, in all chambers, while the half-life of resensitization varied from 5-11 mins among chambers. These estimates suggest 20-32 mins is an appropriate period separating pulses. Smaller models revealed that the desensitized receptor significantly contributed to describing the data. Conclusions regarding the significance of limited LH were more difficult to draw as this mechanism involved a parameter that was difficult to estimate separately.PhDAnimal PhysiologyApplied SciencesBiological SciencesBiomedical engineeringUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/130487/2/9732093.pd