Human granulosa cell secretion of protein(s) which suppress follicular response to gonadotropins.

Abstract

Recently, we reported the identification of a heat- and trypsin-labile substance recovered directly from the venous drainage of the human preovulatory ovary and pooled follicular fluid which suppressed follicular response to gonadotropins. Since many of the steroidal components of follicular fluid are derived from granulosa cells, we questioned whether granulosa cells secrete a substance that inhibits follicular response to gonadotropins. Follicular fluid was aspirated from regularly menstruating women (25–35 yr old) undergoing clomiphene citrate (150 mg/ day for 5 days, beginning 3–8 days after the onset of spontaneously occurring menses) and hCG (4000 IU, 36 h before aspiration) treatment (n = 7). Follicles were characterized by quantitation of viable granulosa cells and steroid concentration. Individual follicular fluids were precipitated by 10–55% ammonium sulfate, dialyzed with 10,000 mol wt exclusion membranes, lyophylized, and eluted by Orange A dye matrix chromatography and high performance liquid chromatography. The granulosa cells from these follicles were cultured with or without varying doses of LH/FSH from 1–4 days. Extracted follicular fluids and tissue culture media were evaluated in an LH/FSH-challenged, immature, hypophysectomized, diethylstilbesterol treated rat (HIFR-hMG) for inhibition of ovarian weight and trunk serum estradiol concentrations. All of the antral fluid/s steroid concentrations suggested premature luteinization. After extraction, all seven follicular fluids contained inhibitory activity, as evidenced by reduction of both rat ovarian weight (45–85%) and trunk serum estradiol concentrations (85–100%) in the HIFR-hMG bioassay. Bioassay of these follicles' granulosa cell culture media indicated inhibitory activity present during the first 48 h, while no inhibitory activity was noted after 72 h of culture. Spent culture media derived from granulosa cells cultured without additional gonadotropins contained inhibitory activity in the HIFR-hMG bioassay throughout the first 2 days in vitro. Thereafter (days 3 and 4), inhibitory activity of media from unstimulated cultures declined. After coincubation of the granulosa cells with varying doses of LH/ FSH, inhibitory activity was markedly suppressed even after the first 24 h. An inverse correlation was apparent between inhibitory activity in the bioassay and the culture medium progesterone level. Although FSH binding of granulosa cells derived from rats used in the HIFR-hMG bioassay was similar with or without injection of test fractions, their aromatase activity was markedly inhibited by treatment with human granulosa cell culture medium. These data suggest that although no marked inhibition in rat granulosa cell FSH binding was induced by the human granulosa cell culture medium, a clear disruption of aromatase activity occurred, which may account for the decreased rat ovarian weight and serum estradiol concentrations found in the bioassay. Here, we have identified from aspirates of individual human follicles obtained during clomiphene- and hCG-stimulated ovarian cycles as well as from culture media derived from these follicle's granulosa cells a substance that actively suppresses follicular responsivity to gonadotropins. When taken together, these data suggest that the human granulosa cell secretes a protein that inhibits follicular response to gonadotropins.