"Ultrashort" control of hypothalamic hormones secretion: a brief history.

Some 30 years ago we proposed that the synthesis of the hypothalamic releasing factors (RFs) might be influenced by changes of their own titers in the general circulation and in the hypothalamus. At that time the dominating concept was that there were two different hypothalamic RFs controling respectively follicle-stimulating hormone (FSHRF) and luteinizing hormone (LHRF). The possible existence of the first "ultrashort" feedback system was proposed on the basis of the observation that in male rats castrated, to eliminate the "long" feedback effect of testosterone, and then hypophysectomized, to abolish the possible "short" feedback effect of pituitary hormones on the hypothalamus, the intrahypothalamic levels of FSHRF resulted to be higher than in controls. Following a treatment with an hypothalamic extract, the intrahypothalamic stores of FSHRF returned to pre-operation levels, suggesting that a factor present in the hypothalamic extract might exert an inhibitory influence on the synthesis of FSHRF. In 1987, we confirmed these findings using an in vitro perfusion system of rat hypothalamic tissue, where the amount of gonadotropin-releasing hormone (GnRH) released was measured in the effluent. When a GnRH analog was added to the perfusion medium, both the spontaneous and the stimulated release of GnRH were totally obliterated. The presence of specific binding sites for GnRH has been clearly demonstrated in GnRH neurons. It is becoming consequently attractive the hypothesis that paracrine interactions might occur among GnRH neurons, entertaining an "ultrashort" feedback control of GnRH secretion.

5Alpha-reductase type 2 and androgen receptor expression in gonadotropin releasing hormone GT1-1 cells.

Gonadal steroids are potent modulators of gonadotropin releasing hormone (GnRH) secretion, and androgen binding sites and 5alpha-reductase activity have been found in the immortalized GnRH secreting cell line GT1-1, suggesting the existence of a direct androgenic control of GnRH dynamics. Two isoforms of the 5alpha-reductase have been cloned with very different biochemical/functional properties: 5alpha-reductase type 1 (widely distributed in the body) and 5alpha-reductase type 2 (confined in androgen target structures). We have analysed whether, in GT1-1, androgen binding sites are linked to "classical" androgen receptor, and which 5alpha-reductase isoform is active. Reverse transcriptase-polymerase chain reaction analysis showed that the mRNAs coding for androgen receptor and for the two 5alpha-reductase isoforms are all expressed in GT1-1 cells. However, the 5alpha-reductase enzymatic reaction showed a peak of activity at a narrow pH around 5.5, the optimum for the 5alpha-reductase type 2. The affinity for testosterone, of the enzyme present in GT1-1 cells, was very similar to that observed for the recombinant type 2 isozyme expressed in yeasts. The data indicate that GT1-1 cells (i) express a "classical" androgen receptor and (ii) contain the 5alpha-reductase type 2 isoform, a specific marker of androgen-responsiveness.

Motoneuronal cell death is not correlated with aggregate formation of androgen receptors containing an elongated polyglutamine tract.

Spinal and bulbar muscular atrophy (SBMA) is associated with an abnormal expansion of the (CAG)(n)repeat in the androgen receptor (AR) gene. Similar mutations have been reported in other proteins that cause neurodegenerative disorders. The CAG-coded elongated polyglutamine (polyGln) tracts induce the formation of neuronal intracellular aggregates. We have produced a model to study the effects of potentially 'neurotoxic' aggregates in SBMA using immortalized motoneuronal cells (NSC34) transfected with AR containing polyGln repeats of different sizes [(AR.Q(n = 0, 23 or 46)]. Using chimeras of AR.Q(n) and the green fluorescent protein (GFP), we have shown that aggregate formation occurs when the polyGln tract is elongated and AR is activated by androgens. In NSC34 cells co-expressing the AR with the polyGln of pathological length (AR.Q46) and the GFP we have noted the presence of several dystrophic neurites. Cell viability analyses have shown a reduced growth/survival rate in NSC34 expressing the AR.Q46, whereas testosterone treatment partially counteracted both cell death and the formation of dystrophic neurites. These observations indicate the lack of correlation between aggregate formation and cell survival, and suggest that neuronal degeneration in SBMA might be secondary to axonal/dendritic insults.

Expression of a leptin receptor in immortalized gonadotropin-releasing hormone-secreting neurons.

Leptin is secreted by adipocytes and regulates food intake and energy balance through the activation of specific receptors (OB-R). Recent evidence suggests that it is also involved in the control of reproductive processes, by possibly acting on central and peripheral targets. In particular, it has been shown that leptin may indirectly stimulate GnRH release from hypothalamic fragments by acting on interneurons impinging on GnRH-secreting neurons. The possibility that leptin might additionally modulate the activity of GnRH-secreting neurons in a direct way has been addressed in the present study, by using the immortalized GnRH-secreting cell line GT1-7. The presence of OB-R messenger RNA (mRNA) (long form) was detected by RT-PCR analysis of total RNA from GT1-7 cells. An OB-R protein is also expressed in these cells, as shown by immunocytochemistry and by Western blot analysis. The latter has revealed the presence of a single immunoreactive OB-R with an approximate size of 130 kDa. To study the functionality of these receptors, the effect of leptin treatment on GnRH secretion and gene expression in GT1-7 cells were evaluated. Under static conditions, GnRH release was stimulated by exposure to low concentrations of leptin (10(-12) M after 30 min; 10(-10) M after 60 min). The 10(-12) M dose was selected for studying the effect of leptin on GnRH secretion under dynamic conditions. To this purpose, GT1-7 cells were placed in a perifusion system; treatment with leptin (10(-12) M) for 60 min stimulated GnRH release with no changes of pulse frequency. On the contrary, exposure to leptin (10(-12)-10(-10) M) for 1, 3, 6, and 24 h did not affect GnRH gene expression in GT1-7 cells. The present results indicate that GT1-7 cells possess OB-Rs and that leptin may directly affect their function. Taken together with the available reports, these findings suggest that leptin might participate in the regulation of reproductive processes by acting at multiple levels, both centrally and peripherally.

Effects and metabolism of steroid hormones in human neuroblastoma cells.

The development of the central nervous system is influenced by sex steroids and by their metabolites. However, little information on the possible effects of steroid hormones on neuroblastoma cells is available. Human neuroblastoma cell lines have been used as a model of human neuroblasts in vitro to study the metabolism of steroid hormones; in addition, the effects of steroids and steroid antagonists on neuroblastoma cell growth have also been investigated. The results obtained show that SH-SY5Y human neuroblastoma cells may actively metabolize testosterone and progesterone to their respective 5 alpha-reduced metabolites and that differentiation of neuroblastoma cells is paralleled by a significant increase in expression of the type-1 5 alpha-reductase and of the formation of steroid metabolites. All these data are suggestive of a potential role of steroid 5 alpha-reduced metabolites in the biology of neuroblastoma cells. Studies performed to analyze the role of steroid hormones on neuroblastoma cell proliferation show that progesterone at low doses may induce minor stimulation, and at higher doses, a toxic effect on the neuroblastoma cell line SK-N-SH is seen. Moreover, the antiprogestin 17 beta-hydroxy-11 beta-(4-dimethylamino-phenyl-1)-17-(prop-1-ynyl)estra-4,9-dien+ ++-3-one (RU486) decreases the proliferation of these cells in a dose-dependent manner. The effect of RU486 is not antagonized by either progesterone or dexamethasone, a result that seems to exclude the action of RU486 via classic intracellular steroid hormone receptors.

LHRF, LHRH, GnRH: what controls the secretion of this hormone?

In the present study, the possible direct effects of opioids on the release of LHRH have been studied utilizing the GT1-1 cell model. It has been shown that only opioid agonists which bind to delta receptors bring about a significant inhibition of the release of LHRH, when this is stimulated by forskolin- or PGE2. The effects of delta opioid agonists are dose-dependent, and are reversed by the delta specific antagonist naltrindole. If one assumes that the GT1-1 cells represent a good model for the study of LHRH-secreting neurons, the obvious conclusion from these data is that endogenous opioid peptides may influence LHRH secretion also by acting directly on LHRH-secreting neurons. In another study it has been shown that GT1-1 cells possess high affinity-low capacity binding sites for estradiol and androgens. Moreover, a two-fold induction of androgen-binding sites has been observed after treatment of GT1-1 cells with estradiol, indicating that estrogen receptors are functional. The present observations suggest that gonadal steroids might influence LHRH release acting directly on the cells which manufacture and release the hypothalamic hormone. In order to evaluate the possible existence of a humoral communication between glial cells and LHRH-secreting neurons, two different approaches have been used: (a) GT1-1 cells were co-incubated with purified cultures of type 1 astrocytes; (b) GT1-1 cells were exposed to the conditioned medium (CM, untreated or submitted to different experimental procedures) where type 1 astrocytes were grown for 24 h. In both sets of experiments LHRH was measured by RIA in the incubation media of GT1-1 cells. The data show that short periods of either co-culture or exposure to previously frozen or heated CM significantly increase the release of LHRH from the GT1-1 cells. The stimulatory effect on LHRH release appears to be specific for type 1 astrocytes. Surprisingly, fresh CM proved to be inactive; this suggested that the factor(s) involved needed to be 'activated'. Since it is known that TGF beta may be liberated from its supporting proteins by freezing or heating, the effects of TGF beta and of a TGF beta-neutralizing antibody were analyzed. The stimulatory effect exerted by the frozen CM was completely abolished by the antibody, while TGF beta 1 proved able to significantly increase LHRH release from GT1-1 cells. The present data unequivocally show that type 1 astrocytes in culture release some principle(s), probably TGF beta, able to stimulate LHRH release from the LHRH-producing GT1-1 cell line. To the authors' knowledge, the present data provide the first clear-cut demonstration that the glia may directly intervene in the control of LHRH release via the secretion of humoral factors.

Differential effects of low- and high-dose estrogen treatments on vascular responses in female rats.

In an attempt to study the mechanisms by which estrogens affect vascular responses, we utilized aortic preparations from intact and ovariectomized female rats receiving low- and high-dose subcutaneous estrogen treatments. Oil-treated, as well as male rats, served as controls. In ovariectomized females, low-dose 17-beta-estradiol injections (5 microg/kg daily for two days) affected the basal release of nitric oxide, as evaluated by concentration-related curves to superoxide dismutase and N(G)-Methyl-L-arginine acetate, which was found to be greater in 17-beta-estradiol-treated females compared to oil-treated females or males. Conversely, the nitric oxide-related vascular relaxation evoked by acetylcholine and sodium nitroprusside was unchanged. Prostacyclin production was also evaluated. Aortic rings from ovariectomized 17-beta-estradiol-treated females released significantly more prostacyclin than those from oil-treated females. These results point out a possible role for nitric oxide and prostacyclin in the vascular protection brought about by physiological levels of estrogens. When intact females were treated with high doses of ethynilestradiol (100 microg/Kg daily for one month), a component of contraceptive pills, either the basal release of nitric oxide, or acetylcholine-induced relaxation underwent a significant decrease. Likewise, the relaxant responses to sodium nitroprusside were impaired in the aortic rings obtained from ethynilestradiol-treated animals when compared to controls. Similarly, the amount of prostacyclin released from aortic tissues obtained from ethynilestradiol-treated animals was significantly reduced. These results may provide a possible explanation for the higher incidence of cardiovascular disease in women who take contraceptive preparations containing high doses of estrogens.

Astrocyte-neuron interactions in vitro: role of growth factors and steroids on LHRH dynamics.

The data here reviewed, obtained with in vitro models, indicate that growth factors and steroids play a significant role in astrocyte-neuron interactions. Different designs have been adopted: (1) GT1-1 cells (a cell line derived from a mouse hypothalamic LHRH-producing tumor) were cocultured with type 1 rat astrocytes; and (2) GT1-1 cells were exposed to the conditioned medium (CM) in which type 1 rat astrocytes had been grown for 24 h. LHRH release and mRNA LHRH levels were measured respectively in the medium and in cell homogenates, at different time intervals (LHRH release, by RIA; LHRH mRNA by Northern blot analysis). The data obtained show that type 1 astrocytes secrete in the medium TGFbeta, which is able to modulate the release and the gene expression of LHRH in GT1-1 cells; and that one or more LHRH-degrading enzymes is/are present in the conditioned medium of type 1 astrocytes. A second part of the experiments have indicated that type 1 astrocytes are also able to affect, in different directions, the metabolism of testosterone and progesterone into their 5alpha-reduced metabolites occurring in the GT1-1 cells. In particular, it has been observed that the conversion of testosterone into DHT is decreased by the coculture with type 1 astrocytes, while the conversion of progesterone into DHP is increased by the same coculture conditions. Moreover, type 1 astrocytes are sensitive to steroid hormones, and in particular to the 5alpha-reduced metabolites of progesterone; this has been shown by analyzing the effects exerted by different steroids on the gene expression of the typical astrocyte marker GFAP.

Transforming growth factor-beta and astrocytic conditioned medium influence luteinizing hormone-releasing hormone gene expression in the hypothalamic cell line GT1.

On the basis of our previous observations indicating that a principle [possibly transforming growth factor-beta1 (TGFbeta1)] secreted by type 1 astrocytes may increase the release of LHRH in the GT1 cell line, it was deemed of interest to analyze whether TGFbeta1 might influence LHRH gene expression in addition to LHRH release in GT1-1 neurons. The effects of TGFbeta1 on the levels of LHRH messenger RNA (mRNA) present in GT1-1 cells have been compared to those found after either coculture of these cells with type 1 astrocytes or exposure of GT1-1 cells to the conditioned medium in which type 1 astrocytes were grown for 24 h. The data obtained indicate that 1) TGFbeta1 increases LHRH mRNA levels 1 and 6 h after the beginning of treatment; longer exposures (24 h) bring about a decrease in LHRH gene expression; 2) a significant stimulatory effect of TGFbeta1 (1 and 6 h of exposure) is also evident on LHRH release; 3) the exposure to the conditioned medium of type 1 astrocytes is able to increase LHRH gene expression in GT1-1 cells at 1 h; LHRH mRNA levels show a small decrease after 6 h of exposure, which becomes more evident at 24 h; and 4) the coculture of GT1-1 cells with type 1 astrocytes is not able to modify LHRH mRNA levels at any time considered. The present data support the concept that glial cells are able to control, possibly through the release of TGFbeta, the gene expression of LHRH in hypothalamic neurons.

The anterior pituitary gland as a possible site of action of kainic acid.

The purpose of the present study was to analyze the direct effect of kainic acid (KA), an agonist of L-Glutamate, on the secretion of LH and FSH from anterior pituitary (AP) of male rats perifused in vitro. At low concentrations (1 microM), KA was able to stimulate the release of both gonadotropins from AP of 50-day-old male rats, but the response to subsequent stimuli was markedly impaired. This, however, was not due to a neurotoxic action of KA, but seemed rather suggestive of a down-regulation or desensitization of KA receptors. The stimulatory action of KA on LH and FSH secretion was age-dependent, since the agonist was completely ineffective on the AP of 75-day- and 18-month-old male rats. DNQX (6,7-dinitroquinoxaline-2,3-dione), a specific antagonist of the KA receptor subtype, was able to block the KA-induced gonadotropin secretion; similarly, AP-5 (2-amino-5-phosphonovalerate), a competitive NMDA receptor antagonist, prevented the stimulatory effect of KA on LH and FSH release. An interaction between the opiatergic and the excitatory aminoacid (EAA) systems emerged from the observation that pulses of KA applied to AP of 50-day-old male rats during a continuous perifusion with a medium containing morphine (5 microM) failed to increase gonadotropin secretion. These results indicate that KA can, at low concentrations, directly stimulate LH and FSH secretion by acting at AP level; this effect disappears with progression of age, and might be exerted both through NMDA and non-NMDA receptor subtypes. Finally, the results provide evidence that opioids and excitatory aminoacids might influence gonadotropin secretion from AP by acting in opposite directions.

Excitatory amino acids as modulators of gonadotropin secretion.

The effects of quinolinic acid (QUIN) and quisqualate (QA) on the secretion of GnRH from MBH and LH and FSH from AP of 50 day old male rats have been evaluated by means of an "in vitro" perifusion technique.QUIN (100µM) is able to increase GnRH secretion with an action mediated by an NMDA receptor type, as shown by the inhibitory effect exerted by both a competitive (AP-5) and a non-competitive (MK-801) specific antagonist.QA "per se" at the concentrations tested (1-100µM) does not modify GnRH and gonadotropin secretion, but in the presence of a specific KA/QA receptor antagonist (DNQX) exerts a stimulatory effect at both levels.This observation might indicate that of the two QA receptor subtypes (ionotropic and metabotropic), this agonist binds to the metabotropic one with very low affinity: thus it is likely that a higher dose is required in order to have any effect on gonadotropin secretion. However, in the presence of DNQX, which binds to the ionotropic receptor, all the available QA can bind to the metabotropic one and can exert its action at MBH AP levels.

Antisense oligonucleotide blocks progesterone-induced lordosis behavior in ovariectomized rats.

Antisense oligonucleotides were utilized to interfere with the synthesis of progesterone receptor. A specific aODN was shown to decrease significantly the levels of PR in the T47D cell line known to express the PR gene. The molecule described was proved to be effective in hindering the effect of progesterone on target gene expression in T47D cells. The aODN was then utilized in an in vivo study to test its efficacy on the female rat sex behavior. Its injection in the ventromedial hypothalamus significantly inhibited the estrous responsiveness induced by progesterone. This study provides a clear demonstration of the full involvement of PR in the manifestation of lordosis behavior and demonstrates the activity of an aODN as antiprogestative.

Ageing of the neuroendocrine system in the brain of male rats: receptor mechanisms and steroid metabolism.

The work described in this article gives information on the effects of ageing on the hypothalamo-pituitary-testicular axis in rats. The hypothalami of young and old male rats contain similar amounts of luteinizing-hormone-releasing hormone (LHRH); when perifused in vitro they release comparable amounts of LHRH under basal conditions and in response to K+. The addition of an LHRH analogue to the perifusion medium blocks the release of LHRH induced by K+ from the hypothalami of young and old male rats, indicating that the ultrashort feedback mechanism controlling LHRH release functions normally in aged male rats. Ageing also exerts important effects on the density of mu- and kappa-opioid receptors in the brain. The number of hypothalamic mu-opioid receptors was significantly decreased in aged animals; a replacement treatment with testosterone does not reverse this decrease, indicating that the decline of hypothalamic mu receptors and of serum titres of testosterone in old rats are independent phenomena. The number of kappa-opioid receptors in the brain increases in the amygdala and in the thalamus with ageing. Apparently ageing does not influence the number of delta receptors in any of the brain areas investigated. The number of pituitary LHRH receptors decreases in old animals, which might explain the low serum concentration of gonadotrophins in aged rats caused by an inadequate response of the pituitary to hypothalamic LHRH. The impaired secretion of testosterone in aged male rats is accompanied by an increase in the number of testicular LHRH receptors, indicating that the intratesticular mechanisms controlling testosterone release also undergo significant alterations during ageing. The rate of conversion of testosterone to dihydrotestosterone (DHT) and 5 alpha-androstane-3 alpha, 17 beta-diol (3 alpha-diol) is the same in the hypothalami of young and old rats. However, the yields of DHT obtained from the pituitaries of aged male rats are significantly lower than those recorded in the pituitaries of young animals. These results show that the enzymes necessary for metabolizing testosterone via the 5 alpha-reductase pathway are maintained both in the hypothalamus and in the anterior pituitary of aged male rats. However, the 5 alpha-reductase activity of the anterior pituitary of senescent animals appears to be lower than that in the younger controls.

Sex steroids and the control of LHRH secretion.

Gonadal steroids are important hormonal signals that regulate the activity of LHRH synthesizing and releasing neurons. Aside from a direct effect through the feedback mechanisms exerted at hypothalamic and/or anterior pituitary level, gonadal steroids may modify the rhythmic LHRH release by modulating other systems affecting LHRH neurons. 1. In ovariectomized E2-treated female rats, progesterone is able to evoke LHRH release from the perifused hypothalamus without affecting LH and FSH release. 2. Excitatory amino acids (EAA) and their related analogs (NMDA and kainate) are known to stimulate LH release in young rats. When tested in a perifusion system on hypothalamic and anterior pituitary tissues, they differentially stimulate the release of LHRH (NMDA) and of LH (KA); their effect on both structures is markedly reduced following orchidectomy. It appears that gonadal steroids might exert a facilitatory action on the neurosecretory activity of LHRH neurons as well as a modulatory influence on the effect of EAA.

Ageing does not influence the ultrashort feedback control of GnRH secretion in vitro.

Evidence indicates that long and short feedback systems are altered in the aged male rat. Data also indicate the existence of an ultrashort feedback mechanism controlling GnRH secretion. The present experiments were performed to test whether the ultrashort feedback control of GnRH is operating also in old male rats. Mediobasal hypothalami of 18-month-old male rats were perifused in vitro either in the presence or in the absence of a GnRH agonistic analogue (Buserelin: [D-Ser(TBU)6,Des-Gly10]GnRH ethylamide) and stimulated with 5-min pulses of K+ (for a total of six pulses) in order to test their ability to release GnRH. The hypothalamic fragment was exposed to the GnRH analogue either for a part of the experimental period (at the beginning or at the end) or for the whole duration of the perifusion. In both cases, the presence of the analogue diminished or totally abolished the responses to K+ stimulation. This is in line with the results obtained in young animals. The data suggest that the ultrashort feedback mechanism controlling GnRH release is normally functioning also in aged male rats despite the fact that other types of feedback mechanisms (long and short loop) are substantially altered.

Antihormonal activities of 5 alpha-reductase and aromatase inhibitors.

The problem of developing androgen antagonists has been tackled so far only by synthesizing steroids able to displace testosterone and other androgens from their specific receptor sites. The observation that testosterone has to be converted intracellularly either to 5 alpha-reduced metabolites (DHT, 3 alpha-diol, etc.) or to estrogens, in order to become fully active on androgen-dependent structures (both central and peripheral), has opened the possibility of creating molecules which prevent these conversions, and which could then block the actions of testosterone. The availability of these new compounds has allowed a better understanding of the selective physiological role of each of the metabolites of testosterone, and to provide the basis for the development of new hormone antagonists to be used in those clinical conditions for which an inhibition of the actions of testosterone is foreseen. The usefulness of these enzyme inhibitors is underlined by some examples described in this paper. The results obtained may permit the formulation of the following conclusions: (1) The conversion of testosterone to its 5 alpha-reduced metabolites occurring in the neuroendocrine structures may represent an essential step for the appearance of the inhibitory feedback effect testosterone exerts on LH secretion; (2) Testosterone exhibits its negative feedback effect on FSH secretion as such and not following the local aromatization to estrogens; (3) Testosterone exerts its effect on the intrahypothalamic stores of LHRH acting as such and not following its local conversion either to 5 alpha-reduced metabolites or to estrogenic molecules; (4) Some of the new enzyme inhibitors (e.g. 4-OH-A) may represent an interesting tool for the treatment and/or the prevention of BPH and possibly of other androgen-dependent diseases (prostate carcinoma, acne etc.), as shown by their ability to prevent the in vitro conversion of testosterone to its 5 alpha-reduced metabolites both in the normal prostate of the rat and in the human BPH tissue.

Testosterone and the control of hypothalamic GnRH.

The aim of the present study was to test the hypothesis that the restoring effect of testosterone on hypothalamic GnRH stores in orchidectomized rats might be linked to the intrahypothalamic transformation of the hormone into estrogenic metabolites. To this purpose, advantage has been taken of the availability of the potent antiestrogen, tamoxifen (TMX). Different groups of adult male rats castrated since 4 weeks were submitted to a 6-day treatment with testosterone propionate (TP, 2 mg/rat daily); TMX (50 or 200 micrograms/rat daily); or TP (2 mg/rat daily) plus TMX (50 or 200 micrograms/rat daily). The animals were sacrificed 24 h after the last injection, and hypothalamic GnRH content was measured by radioimmunoassay. The results have confirmed the ability of TP to counteract the decreasing effect of orchidectomy on hypothalamic GnRH stores, and have shown that TMX does not have any intrinsic activity on this parameter. Furthermore, TMX at either dose used in the present experiments, was found not to be able to abolish the restoring effect of TP on MBH-GnRH stores. It is concluded that the action of testosterone on hypothalamic GnRH does not require the conversion of the hormone into estrogens.

Ultrashort feedback control of luteinizing hormone-releasing hormone secretion in vitro.

The present experiments were performed to clarify whether LHRH might inhibit its own secretion via an ultrashort feedback mechanism acting directly on the hypothalamus. Using an in vitro system, mediobasal hypothalami (MBHs) of adult male rats were perifused in either the presence or absence of a LHRH agonistic analog [D-Ser(TBU)6,Des-Gly10] LHRH ethylamide shown not to cross-react in the LHRH RIA. In the first series of experiments, six MBHs per chamber were initially perifused with control medium and submitted to two K+ stimulations (110 mM) for 5 min every 30 min; the control medium was then replaced by medium containing the LHRH analog (5 microM), and three additional K+ pulses were applied. In the second series of experiments, a single MBH per chamber was exposed for the duration of the experiments to either control medium or medium containing the LHRH analog (5 microM). In both cases, pulses of K+ were applied to the tissue. The amounts of endogenous LHRH released both under basal conditions and after K+ stimulation were measured in the effluent (1 ml every 5 min) with a specific RIA. The results show that the LHRH analog inhibits basal secretion of endogenous LHRH from the MBH, and diminishes or abolishes the response to K+ stimulations. The specificity of the inhibitory effect exerted by the LHRH analog on LHRH secretion was shown by the inability of TRH to mimic the effect of the LHRH analog. The data are consistent with the hypothesis that LHRH, acting at a hypothalamic level, might participate in the control of its own release via an ultrashort feedback mechanism.