Discordant levels of serum bioactive LH in man as measured in different in-vitro bioassay systems using rat and mouse interstitial cells and human granulosa-luteal cells

Serum samples from healthy men (n = 6), perimenopausal women (n = 9) and patients with polycystic ovarian disease (n = 4) were analysed for LH bioactivity by the two widely used in-vitro bioassay systems: the mouse and rat interstitial cell testosterone-formation assays. The results were compared with an assay employing LH-stimulated cyclic AMP production by cultured human granulosa-luteal cells. Average bioactive LH levels in the mouse cell assay were 1.6-fold higher than those measured using the rat assay. A good correlation (r = 0.89, P less than 0.01) was observed between the bioactive LH levels measured by these two assays. No significant difference was found between the sensitivities of the two assays: 0.009 +/- 0.003 IU/l (mean +/- S.E.M., n = 10) with rat cells and 0.011 +/- 0.003 IU/l (n = 10) with mouse cells. The LH level resulting in half-maximal stimulation of testosterone production in the mouse model was twofold higher than that in the rat model (0.185 +/- 0.020 vs 0.083 +/- 0.022 IU/l, P less than 0.01). The bioactive LH levels measured by the human granulosa-luteal cell assay averaged 12% higher than those obtained with the rat assay (r = 0.84, P less than 0.01), but 58% lower than levels measured with the mouse assay (r = 0.86, P less than 0.01). The data indicate that the target cell model used in the in-vitro bioassay of LH contributes to the documented discrepancies in reports on serum levels of bioactive LH.(ABSTRACT TRUNCATED AT 250 WORDS)