Development of gonadotrophin-binding sites in the immature rat ovary.

There was a temporal relationship between ovarian development and the sites to which 125I-labelled gonadotrophins became bound. Labelled human LH and FSH bound uniformly to 5- and 15-day-old ovaries. At 21 days, heavy FSH and light LH binding was observed over the granulosa cells increased at Day 33 and again at Day 38. Quantitative determination by gamma-ray spectrophotometry of binding to ovarian sections showed that LH binding gradually increased with advancing age, but FSH binding remained relatively constant between Days 5 and 33 with a significant increase between Days 33 and 38.     

The Parkes lecture: controlled ovarian stimulation in women

Recent advances in knowledge of the endocrine and paracrine mechanisms that regulate human ovarian folliculogenesis have been parallelled by the introduction into clinical practice of new drugs that can be used safely and effectively to stimulate ovarian function in infertile women. Most notably, recombinant DNA technology has been applied to the production of molecularly pure forms of the gonadotrophins, FSH and LH, opening the way to the development of improved strategies for manipulating the ovarian paracrine system. The clinical objectives of controlled ovarian stimulation fall into two categories, depending on patient needs: (1) induction of multiple follicles from which mature oocytes can be harvested for use in assisted reproduction protocols such as in vitro fertilization and embryo transfer; or (2) induction of spontaneous ovulation of a single mature follicle so that conception might occur in vivo. This review summarizes the physiological principles upon which the use of gonadotrophins for clinical purposes is based, highlighting new opportunities for improved treatment as a result of the availability of recombinant FSH and LH.     

Ontogeny of pituitary gonadotrophin secretion in the fetal and post-natal pig in response to LHRH in vitro

Monolayer cultures of anterior pituitary cells from male or female pigs of 60, 80, 105 days of fetal life or of 60, 160 and 250 days of post-natal life were prepared and treated with LHRH (1 pM to 10 nM). Dose-related increases of LH were first seen at 80 days of gestation in both sexes, while only female fetuses responded to maximal LHRH at 60 days. Basal and stimulated LH release doubled in cultures from 105-day-old fetuses when compared with those at 80 days. Compared to late fetal stages LH release was 20- to 30-fold higher in cell cultures from 60-day-old (post-natal) donors without further change during the post-natal period. In all pre- and post-natal age groups basal and maximal LH release of pituitary cells from males was lower than that of females. FSH stimulation started in male and female cells at 80 days of gestation only at LHRH concentrations exceeding or equal to 0.1 nM. By 105 days FSH secretion was dose-related and pituitary cells of females responded with higher FSH values than did those of males. In general, post-natal cells released much higher amounts of FSH than did prenatal cells. Basal and maximal release of FSH decreased during post-natal development in both sexes. Basal as well as maximal FSH release of cultures from female donors was higher than that found in cultures from male donors. Determination of total LH and FSH content in fetal pituitary cell cultures indicated that the developmental increase in gonadotrophin release potential is a function of the total gonadotrophin content in vitro. We conclude that (1) the in-vitro release of gonadotrophins to LHRH is dose-, age- and sex-dependent; (2) in the female fetal pig LH responsiveness develops earlier than FSH responsiveness; (3) apparently, these maturational changes mainly reflect alterations in pituitary gonadotrophin content; and (4) there is no simple relationship between in-vitro release and circulating gonadotrophins.     

Multiple regulation of testicular steroidogenesis

One single injection of ethylene dimethane sulfonate (EDS) to mature rats causes specific degeneration of testicular Leydig cells which is complete after 3 days. At this time no steroidogenic activities can be detected, indicating that Leydig cells are the source of steroids. The mechanism of this cytotoxic effect of EDS has been investigated with isolated cells. Extensive protein alkylation has been shown to occur in Leydig cells, Sertoli cells and hepatocytes. Steroid production by Leydig cells is always inhibited by EDS, but cytotoxic effects of EDS could only be demonstrated in Leydig cells from mature rats or tumour tissue and not in Leydig cells from immature rats. A new population of Leydig cells develops during the next 2-5 weeks after EDS treatment. In hypophysectomized rats this repopulation only occurs when hCG is given daily. FSH has no effects. The proliferative activity in the interstitial tissue increases within 2 days after administration of hCG or EDS and there are indications that LH and locally produced factors are involved in the proliferation of Leydig cells or Leydig cell precursor cells. Inhibition of cAMP production with inhibitors of adenylate cyclase results in an enhancement of the LH-stimulated steroid production similar to that observed with an LHRH agonist and phospholipase C (PLC). Since the effects of LHRH and PLC on protein phosphorylation and steroid production are similar and different from LH or active phorbol esters, it is proposed that LHRH and PLC may stimulate steroid production via liberation of calcium from a specific intracellular pool. Sterol carrier protein2 (SCP2) which is specifically localized in Leydig cells and regulated by LH probably plays a role in the delivery of cholesterol to the mitochondria although the mechanism of this carrier function is not clear. The results indicate that regulation of Leydig cell development and the steroidogenic activities by gonadotrophins and locally produced factors occur via different transducing systems and regulatory pathways.     

Changes in circulating levels of immunoreactive follicle stimulating hormone, luteinizing hormone, and testosterone during sexual development in the rhesus monkey, Macaca mulatta

Basal serum levels of follicle stimulating hormone (FSH), luteinizing hormone (LH), and testosterone (T) and the responsiveness of these hormones to a challenge dose of luteinizing hormone releasing hormone (LHRH), were determined in juvenile, pubertal, and adult rhesus monkeys. The monkey gonadotrophins were analyzed using RIA reagents supplied by the World Health Organization (WHO) Special Programme of Human Reproduction. The FSH levels which were near the assay sensitivity in immature monkeys (2.4 +/- 0.8 ng/ml) showed a discernible increase in pubertal animals (6.4 +/- 1.8 ng/ml). Compared to other two age groups, the serum FSH concentration was markedly higher (16.1 +/- 1.8 ng/ml) in adults. Serum LH levels were below the detectable limits of the assay in juvenile monkeys but rose to 16.2 +/- 3.1 ng/ml in pubertal animals. When compared to pubertal animals, a two-fold increase in LH levels paralleled changes in serum LH during the three developmental stages. Response of serum gonadotrophins and T levels to a challenge dose of LHRH (2.5 micrograms; i.v.) was variable in the different age groups. The present data suggest: an asynchronous rise of FSH and LH during the pubertal period and a temporal correlation between the testicular size and FSH concentrations; the challenge dose of LHRH, which induces a significant rise in serum LH and T levels, fails to elicit an FSH response in all the three age groups; and the pubertal as compared to adult monkeys release significantly larger quantities of LH in response to exogenous LHRH.     

Effect of an LHRH Analog on Growth and Hormone Receptor Levels in DMBA-Induced Mammary Tumors in the Rat

Abstract

Dimethylbenz(a)anthracene (DMBA)-induced mammary tumors in the rat are well known to be hormono-dependent. Daily injections of an LHRH agonist, [D-Ala⁶, des-Gly-NH₂¹⁰]LHRH ethylamide (LHRH-A), 1 ug daily, for 38 days results in a 35% decrease in the number of tumors present at the beginning of the experiment compared to a decline of 45% after ovariectomy and of 8% in the control group. This is accompanied by a marked reduction in ovarian LH and FSH receptors. LHRH-A treatment also resulted in reduction in the number of progesterone and prolactin receptors in the tumors. In addition, an increase in plasma LH and FSH and a decline in plasma prolactin (PRL) concentrations are observed. The mechanisms by which the LHRH agonist induces its antitumoral effect probably relate to an ovarian desensitization to LH and FSH with a concomitant decrease in circulating levels of estrogen and prolactin, two well known stimuli for the growth of DMBA-induced mammary tumors.     

Effects of LHRH-analogues on mitogenic signal transduction in cancer cells

The expression of luteinizing hormone-releasing hormone (LHRH) and its receptors has been demonstrated in a number of human malignant tumors, including cancers of the breast, ovary, endometrium and prostate. These findings suggest the presence of an autocrine regulatory system based on LHRH. Recent studies in our laboratory have demonstrated that the function of LHRH produced by ovarian cancer cells is the inhibition of their proliferation. Dose-dependent antiproliferative effects of LHRH-agonists have been observed by several laboratories in cell lines derived from the above cancers. Interestingly, also LHRH-antagonists have marked antiproliferative activity in most of the ovarian, breast and endometrial cancer cell lines tested so far, indicating that the dichotomy of LHRH-agonists/LHRH-antagonists is not valid for the LHRH-system in cancer cells. In addition, our data suggest that the classical LHRH receptor signal transduction mechanisms known from the pituitary (phospholipase-C, protein kinase C, adenylyl cyclase) are not involved in the mediation of LHRH effects in cancer cells. Data obtained by several groups, including ours, rather suggest that LHRH analogs interfere with the signal transduction of growth-factor receptors and related oncogene products associated with tyrosine-kinase activity. The mechanism of action is probably an LHRH-induced activation of a phosphotyrosine phosphatase, counteracting the effects of receptor associated tyrosine kinase. In our hands, LHRH analogs virtually blocked the EGF-induced MAP-kinase activity of ovarian and endometrial cancer cells. The pharmacological exploitation of this mechanism might provide promising new therapies for these cancers.     

Gonadotropin levels in ovarian cyst fluids: a predictor of malignancy?

Gonadotropins can stimulate ovarian cancer growth in cell cultures. Corresponding LH/hCG receptors have been demonstrated in ovarian cancer. However, reduction of elevated serum gonadotropins by GnRH analogs in ovarian cancer patients did not lead to growth restriction, which means that serum levels of gonadotropins may not play the most important role in ovarian cancer. We therefore analyzed the LH and FSH concentrations in cyst fluids of ovarian cancer. Patients with preoperatively diagnosed cystic ovarian tumors were eligible for the study. Serum samples of the patients were obtained during surgery, while the fluids within the cysts were aspirated after surgical removal of the tumor. FSH and LH levels in serum and cyst fluids were measured using single antibody EIA (Boehringer Mannheim GmbH, Germany). Cyst fluids and sera of 108 patients were evaluated. While there were no significant differences in the FSH and LH serum concentrations, highly significant differences in the FSH and LH levels in cyst fluids were found. Only cancer cysts contained FSH and LH, while the corresponding concentrations in benign cysts were always below the measuring range of the assays. This clear division between high gonadotropin levels in cysts of serous ovarian cancer and low or absent concentrations in benign ovarian tumors further supports the hypothesis that FSH and LH may play a role in ovarian cancer; however, explanations for this surprising finding are still lacking.     

Effects of luteinizing hormone releasing hormone on gonadotrophins in the hamster

The changes in serum gonadotrophins in male hamsters following one injection of 15 μg luteinizing hormone releasing hormone (LHRH) (Group A) were compared with those following the last injection of LHRH in animals receiving an injection approximately every 12 hr for 4 days (Group B) or 12 days (Group C). Peak follicle stimulating hormone (FSH) levels (ng/ml) were 1776±218 (Group A), 2904±346 (Group B), and 4336±449 (Group C). Peak luteinizing hormone (LH) values (ng/ml) were 1352±80 (Group A), 410±12 (Group B), and 498±53 (Group C). Serum FSH:LH ratios, calculated from the concentrations measured 16 hr after the last LHRH injections, were higher in Groups B and C than in Group A. Similar injections of LHRH (100 ng or 15 μg/injection) for 6 days elevated the serum FSH:LH ratio in intact males. Five such LHRH injections (100 ng/injection) blunted the rise in serum LH in orchidectomized hamsters. Direct effects of LHRH on gonadotrophin secretory dynamics or altered brain-pituitary-testicular interactions may alter the ratio of FSH to LH in the hamster.     

Adrenal C19-5-ene steroids induce full estrogenic responses in rat pituitary gonadotrophs

Previous studies have shown that the C19 adrenal steroid 5-androstene-3 beta, 17 beta-diol (5-ene-diol), a metabolite of dehydroepiandrosterone (DHEA), can stimulate typical estrogenic responses in target tissues. Since estrogens are known to cause a specific stimulatory effect on LHRH-induced LH release in rat anterior pituitary cells in culture, we have taken advantage of the precision of this system to study the effect of 5-ene-diol or DHEA on this precise estrogen-sensitive parameter. Pretreatment for 48 h with 17 beta-estradiol (E2), 5-ene-diol or DHEA induces a 2.4-, 2.7- and 2.6-fold stimulation of LH release induced by 0.3 nM LHRH, the effect being exerted at respective 50% maximally effective concentrations (ED50 values) of 0.015, 45 and 115 nM. Following a 48-h preincubation with 10 nM E2, 1 microM 5-ene-diol or 1 microM DHEA, the maximal LH and FSH responses to LHRH are increased by approx 50% above control. On the other hand, the sensitivities of the LH and FSH responses to LHRH as assessed by ED50 values of LHRH action are increased by 3.3- to 7.5-fold. As further proof of the estrogenic nature of the effect of 5-ene-diol and DHEA, the effects of E2, 5-ene-diol and DHEA are inhibited competitively by simultaneous incubation with the antiestrogen LY156758 (keoxifene). The 2-fold stimulation of LHRH-induced LH release caused by DHEA-S, at concentrations within the range found in the plasma of women, is also completely blocked by 120 nM LY156758. In direct binding studies, 5-ene-diol and DHEA or DHEA-S have approx 85- and greater than 10,000 lower affinities than E2, respectively, for the estrogen receptor in rat anterior pituitary homogenate and human breast carcinoma cytosol. The present data clearly show that 5-ene-diol, DHEA and DHEA-S can exert full estrogenic activity in rat gonadotrophs, thus supporting the potential estrogenic role of these C19 adrenal steroids in estrogen-dependent processes, especially breast cancer.     

Targeting breast and prostate cancers through their hormone receptors

A targeted treatment that effectively destroys human breast, prostate, ovarian, and testicular cancer cells that express luteinizing hormone/chorionic gonadotropin (LH/CG) receptors has been developed. The treatment consists of a conjugate of a membrane-disrupting lytic peptide (Hecate, Phor14, or Phor21) and a 15-amino acid segment of the beta chain of CG. Because these conjugates act primarily by destroying cell membranes, their effects are independent of cell proliferation. The conjugates are relatively small molecules, are rapidly metabolized, and are not antigenic. In a series of independent experiments conducted in three different laboratories, the validity of the concept has been established, and it has been shown that the LH/CG receptor capacity of the cancer cells is directly related to the sensitivity of the lytic peptide conjugates. Sensitivity to the drugs can be increased by pretreating prostate or breast cancer cells with FSH or estradiol to up-regulate LH/CG receptors. A series of 23 in vivo experiments involving a total of 1630 nude mice bearing xenografts of human prostate or breast cancer cells showed convincingly that all three lytic peptide-betaCG compounds were highly effective in destroying tumors and reducing tumor burden. Hecate-betaCG was less effective in mice bearing ovarian epithelial cancer cell xenografts, but was highly effective in treating granulosa cell tumors in transgenic mice. In addition, Hecate-betaCG and Phor14-betaCG were highly effective in targeting and destroying prostate and breast cancer cell metastases in the presence or absence of the primary tumors. Although effective in vitro, neither Hecate nor Phor14 alone were effective in reducing primary tumor volume or burden in nude mice bearing prostate or breast cancer xenografts.     

Effects of intermittent pulsatile infusion of luteinizing hormone-releasing hormone on dihydrotestosterone-suppressed gonadotropin secretion in castrate rams

The feedback effects of dihydrotestosterone (DHT) on gonadotropin secretion in rams were investigated using DHT-implanted castrate rams (wethers) infused with intermittent pulsatile luteinizing hormone-releasing hormone (LHRH) for 14 days. Castration, as anticipated, reduced both serum testosterone and DHT but elevated serum LH and follicle-stimulating hormone (FSH). Dihydrotestosterone implants raised serum DHT in wethers to intact ram levels and blocked the LH and FSH response to castration. The secretory profile of these individuals failed to show an endogenous LH pulse during any of the scheduled blood sampling periods, but a small LH pulse was observed following a 5-ng/kg LHRH challenge injection. Dihydrotestosterone-implanted wethers given repeated LHRH injections beginning at the time of castration increased serum FSH and yielded LH pulses that were temporally coupled to exogenous LHRH administration. While the frequency of these secretory episodes was comparable to that observed for castrates, amplitudes of the induced LH pulses were blunted relative to those observed for similarly infused, testosterone-implanted castrates. Dihydrotestosterone was also shown to inhibit LH and FSH secretion and serum testosterone concentrations in intact rams. In summary, it appears that DHT may normally participate in feedback regulation of LH and FSH secretion in rams. These data suggest androgen feedback is regulated by deceleration of the hypothalamic LHRH pulse generator and direct actions at the level of the adenohypophysis.     

Functional and morphological changes in the hypothalamo-pituitary-gonadal axis of aged female rats.

Age-related functional and morphological alterations in the hypothalamo-pituitary-gonadal axis were investigated in old recurrently pseudopregnant (RPP) female rats, and these alterations were compared with those in young diestrous rats. LHRH in the median eminence (ME) and mediobasal hypothalamus (MBH) as well as plasma FSH, LH, and progesterone were measured by RIA. LHRH in the lateral ME (LME) and pituitary FSH and LH were evaluated by morphometry and densitometrical immunocytochemistry. Furthermore, by light microscopy, we classified and counted the number of ovarian follicles and corpora lutea. LHRH concentrations in the ME and MBH were similar in old and young rats, whereas in old rats, plasma FSH was markedly increased, LH was moderately increased, and plasma progesterone was unchanged. The number and the total area and immunoreactivity of LHRH-labeled axon cross sections in the LME were reduced in old rats. The number of nucleated FSH-labeled cells and total FSH area and immunoreactivity were almost twice in old compared with young animals. The measurements of LH-labeled cells were not different between the two groups. In old rats, the numbers of ovarian follicles and corpora lutea were reduced and that of atretic follicles increased. In conclusion, age-related morphological impairments of LHRH axons associated with an increased number of FSH gonadotropes and higher plasma FSH in our old RPP rats suggest hypothalamic and pituitary disturbances, which may largely contribute to the complex hormonal disarrangement responsible for the decline of reproductive functions in old female rats.     

Ovotoxic effects of galactose involve attenuation of follicle-stimulating hormone bioactivity and up-regulation of granulosa cell p53 expression

Clinical evidence suggests an association between galactosaemia and premature ovarian insufficiency (POI); however, the mechanism still remains unresolved. Experimental galactose toxicity in rats produces an array of ovarian dysfunction including ovarian development with deficient follicular reserve and follicular resistance to gonadotrophins that characterize the basic tenets of human POI. The present investigation explores if galactose toxicity in rats attenuates the bioactivity of gonadotrophins or interferes with their receptor competency, and accelerates the rate of follicular atresia. Pregnant rats were fed isocaloric food-pellets supplemented with or without 35% D-galactose from day-3 of gestation and continuing through weaning of the litters. The 35-day old female litters were autopsied. Serum galactose-binding capacity, galactosyltransferase (GalTase) activity, and bioactivity of FSH and LH together with their receptor competency were assessed. Ovarian follicular atresia was evaluated in situ by TUNEL. The in vitro effects of galactose were studied in isolated whole follicles in respect of generation of reactive oxygen species (ROS) and expression of caspase 3, and in isolated granulosa cells in respect of mitochondrial membrane potential, expression of p53, and apoptosis. The rats prenatally exposed to galactose exhibited significantly decreased serum GalTase activity and greater degree of galactose-incorporation capacity of sera proteins. LH biopotency and LH-FSH receptor competency were comparable between the control and study population, but the latter group showed significantly attenuated FSH bioactivity and increased rate of follicular atresia. In culture, galactose increased follicular generation of ROS and expression of caspase 3. In isolated granulosa cells, galactose disrupted mitochondrial membrane potential, stimulated p53 expression, and induced apoptosis in vitro; however co-treatment with either FSH or estradiol significantly prevented galactose-induced granulosa cell p53 expression. We conclude that the ovotoxic effects of galactose involves attenuation of FSH bioactivity that renders the ovary resistant to gonadotrophins leading to increased granulosa cell expression of p53 and follicular atresia.     

Autoradiographical localization of luteinizing hormone releasing hormone (LHRH) receptors on rat testicular intertubular cells fractionated on Percoll density gradients

The specific binding of 125I-labelled [D-Ser(tBu)6,des-GlyNH2(10)] LHRH ethylamide (LHRH-A) to testicular intertubular cells fractionated on Percoll density gradients was investigated. The greatest binding per cell occurred in the density region which contained the largest proportion of Leydig cells (sp. gr. 1.0820-1.0585). Autoradiographs of the cells from this region confirmed that silver stains were predominantly located over the Leydig cell, significantly (P less than 0.01) more grains were observed over this cell type in the total binding fractions than in the non-specific binding fractions. However, 5.9% of cells other than Leydig cells (testicular macrophages and indeterminate connective tissue cells) from this region also displayed significant displaceable binding (P less than 0.01). The location of [125I]LHRH-A binding to cells in other density regions, which did not contain identifiable Leydig cells, could not be established by autoradiography. These results confirm that the Leydig cell possesses LHRH receptors, but also indicate that other testicular cells have specific, high-affinity binding sites for LHRH-A, and may either be responsive to direct stimulation by LHRH, or may partially mediate the effects of LHRH and its agonists on Leydig cell function.     

Pituitary response to exogenous LHRH in superovulated women

The response of the pituitary to exogenous LHRH was investigated in 9 normally ovulating women during the late follicular phase of a spontaneous (control) cycle, a cycle during treatment with clomiphene and a cycle during treatment with 'pure' FSH. During clomiphene treatment, basal FSH concentrations increased significantly up to Day 6 of the cycle and then decreased progressively while basal LH values showed a continuous rise. During treatment with FSH, basal LH concentrations decreased significantly. The response of both FSH and LH to LHRH showed a significant and quantitatively similar decrease during clomiphene or FSH administration as compared to the spontaneous cycles. It is suggested that basal secretion of FSH and LH is regulated by two separate mechanisms, and that an ovarian inhibitory factor(s) attenuates the response of both FSH and LH to exogenous LHRH and possibly the endogenous LH surge in superovulated cycles.     

Variability in gonadotrophin preparations as a factor in the superovulatory response

Much of the variability in superovulatory response has been attributed to variation in ovarian response of individual animals. Alternatively, differences in the relative abundance of FSH and LH activity in gonadotrophin preparations may contribute to superovulatory variation. This report presents evidence for variability in LH and FSH activity among equine chorionic gonadotrophin, porcine FSH and human menopausal gonadotrophins. Lower ratios of $\text{FSH}\text{LH}$ activity appeared to reduce ovulatory success in rats, and addition of PLH to FSH-P reduced superovulation in crossbred cows.     

Effect of ovariectomy at two periods of the year on LH and FSH basal concentrations and pituitary response to LHRH in the brown hare (Lepus europaeus)

In the brown hare, fertile mating takes place from the beginning of December to September. Seasonal variations of basal concentrations of LH and FSH, and pituitary response to a monthly i.v. injection of LHRH were studied in intact control females and in females ovariectomized during the seasonal anoestrus (OVX1) or during the breeding season (OVX2). In intact females, both basal and LHRH-stimulated LH levels showed an annual variation, with minimal values during anoestrus. During the breeding season, the LH response to LHRH exhibited a biphasic pattern. In contrast, there was no clear seasonal variation in basal and LHRH-stimulated FSH concentrations. After ovariectomy during anoestrus, basal LH remained low for 2 months and began to increase in December. After ovariectomy during the breeding season, LH basal concentrations increased within a few days after the operation. Thereafter, LH values remained high in both groups of females until September, and decreased significantly as in intact females. The pattern of LH release after LHRH remained monophasic in the two groups of ovariectomized females. In OVX1 females, the LH response increased as early as October, was maximum from December to April and decreased progressively until October. IN OVX2 females, the LH response decreased regularly after ovariectomy to a minimum in October. In the 2 groups of ovariectomized females, basal FSH concentrations and pituitary response to LHRH rose rapidly after ovariectomy and did not vary significantly thereafter. These results showed a direct central effect of season on the regulation of basal concentrations of LH, modulated by a negative feed-back of ovarian secretions during the breeding season. In intact hares, the enhanced LH response after LHRH during the breeding season was related to an acute positive effect of ovarian secretions. The regulation of FSH was less dependent on season and remained under a negative control of the ovary throughout the year.     

Polycystic ovary syndrome: interaction of follicle stimulating hormone and polypeptide growth factors in oestradiol production by human granulosa cells

The mechanism of the ovarian dysfunction in polycystic ovary syndrome, the most common cause of anovulatory infertility, remains obscure. Clinical data suggest that follicle stimulating hormone (FSH) action may be inhibited at the ovarian level by paracrine factors derived, presumably, from interstitial cells. The greater responsiveness to FSH of granulosa cells isolated from polycystic ovaries (PCO) compared with that seen in cells derived from normal ovaries, provides some support for this hypothesis and we present data which suggests that epidermal growth factor, or more likely transforming growth factor alpha, could be a candidate for this inhibitor. It should be emphasized, however, that the cardinal biochemical feature of the PCO is hypersecretion of androgens by interstitial cells. Stromal tissue from the PCO will secrete significant quantities of androstenedione in response to LH, whereas there is a negligible response in stroma from normal ovaries. It remains to be determined whether androgens have a direct inhibitory effect on FSH-induced oestradiol production in the human follicle, or whether they might exert an indirect effect by activating inhibitory polypeptide growth factors.