Physiological attributes of endogenous bioactive luteinizing hormone secretory bursts in man

The exact nature of glandular secretory events is difficult to discern in vivo, sinced underlying patterns of hormone release are confounded by metabolic clearance. Moreover, conventional immunoassay does not always agree with bioassay. Here, we have used the rat interstitial cell testosterone in vitro bioassay of luteinizing hormone (LH) to measure serial LH concentrations in the normal adult human male. The resultant bioactive LH time series were analyzed by a deconvolution model in which circulating hormone concentrations are controlled by the operation of four finite, nonzero, and determinable parameters: 1) the location(s); 2) the amplitudes(s), and 3) the half-duration(s) of underlying secretory bursts, acted on by 4) endogenous exponential clearance kinetics. A macroscopic secretory burst was modeled in the algebraic form of a Gaussian distribution of instantaneous molecular secretory rates. The physiological implications of this model were examined by analyzing bioactive LH pulsatility in eight men sampled every 10 min for 6 h. Multiple-parameter deconvolution disclosed endogenous bioactive LH half-lives of 53 ± 5.4 min (range 38-76 min), in agreement with earlier independent estimates of 65 ± 4.9 (42-87) min in four LH-deficient men given a 35 μg iv bolus of human LH. Calculated endogenous production rates of bioactive LH (0.48 ± 0.06 mIU · min^-1 · ml^-1) were also in accord with values estimated previously from steady-state infusions of LH. Deconvolution further indicated that underlying bioactive LH secretory bursts had half-durations of only 12.2 ± 1.5 min, occurred at intervals of 56 ± 1.3 min, and achieved amplitudes of 2.1 ± 0.26 mIU · min^-1 · ml^-1. Linear regression analysis revealed that bioactive LH secretory burst amplitudes were positively correlated with the duration of subsequent interpulse intervals in seven of eight men (P = 00006). We conclude that the endogenous daily production rate of bioactive LH and the half time of bioactive LH clearance can be estimated in vivo by deconvolution modeling and in vitro bioassay of LH.