Specific non-genomic, membrane-localized binding sites for progesterone in the bovine corpus luteum

Fractionation of bovine corpus luteum (CL) homogenates on continuous sucrose density gradients with and without preincubation with 3H-progesterone demonstrated high levels of tracer binding and high content of endogenous progesterone associated with particulate membrane fractions. Analysis of gradient fractions for a range of luteal plasma membrane and intracellular organelle marker enzyme activities indicated that endogenous progesterone content and 3H-progesterone-binding activity were associated with fractions enriched in luteal plasma membrane markers. This was confirmed by pretreatment of homogenates with the saponin, digitonin, prior to fractionation. Digitonin perturbed the buoyant density of luteal surface membrane markers and 3H-progesterone binding to a similar extent, but did not perturb the buoyant densities of other intracellular markers to the same degree. Interestingly, digitonin pretreatment also increased the proportion of progesterone tracer that entered the gradients. We consistently failed to demonstrate significant binding of 3H-progesterone to membrane fractions incubated with progesterone tracer in vitro. However, when digitonin was included in the in vitro binding assay, we observed a dramatic, dose-dependent stimulation of 3H-progesterone binding by digitonin. Other radiolabeled steroids tested (3H-cortisol, 3H-testosterone) bound poorly in the presence or absence of digitonin. 3H-Progesterone binding in the presence of optimal digitonin concentrations increased linearly with increasing luteal membrane concentration; was dependent on the pH, duration, and temperature of incubation; and low levels of progesterone (68 nM) competed for tracer binding. A range of other steroids tested (androgens, estrogens, corticosteroids, steroid precursors) competed at higher concentrations (10- to 100-fold) or did not compete at all for 3H-progesterone binding. There was no correlation between the hydrophobicity of various steroids and their ability to compete for binding. Moreover, a number of agonists and antagonists specific for the genomic progesterone receptor, agonists of peripheral benzodiazepine receptors, and inhibitors of a range of steroidogenic enzymes did not compete for 3H-progesterone binding. Western blots confirmed that detergent-solubilized progesterone-binding sites could be resolved from cytochrome P450 side-chain cleavage and 3beta-hydroxysteroid dehydrogenase. Moreover, extraction of bound steroid from the binding site and HPLC demonstrated identity to progesterone, suggesting that no metabolism of the progesterone tracer had occurred during incubation. Progesterone binding to membranes of large luteal cells was higher compared with binding to small luteal cells, and levels were similar in membranes prepared from CL at all stages of the luteal phase. We suggest that bovine luteal progesterone-binding sites may play a role either in sequestration of newly synthesized progesterone or in the mediation of autocrine and/or paracrine actions of progesterone in the CL.