To investigate the secretory pattern of somatostatin (SS) from the median eminence (ME) in the female rat, as well as estrogenic influence on this secretion, we measured both SS release and hypothalamic content in cycling, 10-day ovariectomized, and ovariectomized rats treated with estradiol for 3 days before. Animals were stereotaxically implanted with a push-pull cannula into the ME, and 10 days later the hypothalamic structure was perfused with artificial cerebrospinal fluid for 120–150 min at a regular flow rate of 17 µl/min. Secretion peaks were observed in the pattern of SS release, whatever the stage of the estrous cycle. The mean amplitude of SS peaks was similar throughout the cycle: 11.7 ± 4.0, 8.6 ± 1.5 and 10.5 ± 1.3 pg at proestrus, estrus and diestrus, respectively, and it was affected neither by ovariectomy (7.4 ± 1.3 pg) nor by estrogen replacement (5.5 ± 1.0 pg). By contrast, mean SS release levels in the proestrus phase were significantly higher than those measured in the other phases: 21.6 ± 2.1 vs. 17.7 ± 1.2 pg/15 min in diestrus (p < 0.05) and vs. 12.0 ± 0.7 pg/15 min in estrus (p < 0.001). Hypothalamic SS content showed variations quite similar to those observed during its release, i.e. with the highest values corresponding to the proestrus phase (1,170.5 ± 224.9 pg/mg of tissue) and to the diestrus (1,156.5 ± 332.1 pg/mg of tissue) and the lowest values in the estrus (511.5 ± 52.9 pg/mg of tissue; p < 0.05 vs. proestrus and diestrus). In addition, the lowest SS content and secretion values were found in ovariectomized animals: 95.5 ± 5.1 pg/mg of tissue (p < 0.001 compared to the values obtained for each stage of the estrous cycle) and 10.0 ± 0.9 pg/15 min (p < 0.001 vs. proestrus and diestrus), respectively. Patterns of SS release and SS hypothalamic content were not modified by estradiol treatment in ovariectomized animals. Our results suggest that (1) whatever the stage of the estrous cycle, SS release from the ME is not uniform and exhibits irregular peaks; (2) mean SS release levels were subjected to gonadal influence; (3) the occurrence of SS peaks seems to be estrogen-independent, and (4) variations in hypothalamic SS content were generally in good agreement with those of neurohormone release.

We studied the sensitivity to a depolarizing stimulus of hypothalamic fragments dissected from cycling female donor rats exposed or not to 30-min stress at 4 degrees C. The neuronal response was estimated in terms of the ability of tissue to release somatostatin when stimulated with 40 mM K+. The data showed no differences in response to K+, regardless of the ovarian cycle of the female donors, whereas tissues dissected from ovariectomized or pregnant rats responded significantly to K+. However, when donors underwent previous cold stress, significant differences were noted at all stages of the cycle, except diestrus-1, compared with control rats. We tested whether GABA and/or neuroactive steroids could be involved in this phenomenon and observed no GABA inhibition of somatostatin release in vitro, but inhibition occurred in the presence of a neuroactive steroid, THDOC. The effect of GABA in vivo on somatostatin release was estrogen dependent because bicuculline modified the total amount of somatostatin secreted in estrus but not in diestrus II. Finally, in hypothalamic primary cultures, GABA inhibition of somatostatin release was only detected when steroids were present in the media throughout culture. Our results suggest that steroid-GABA-somatostatin interactions could explain the different responses of neurons to depolarization.

We have previously reported the rapid response of hypothalamic somatostatin (SS) neurons to acute stress. Since it is well known that glucocorticoids (GC) are involved in neuroendocrinal stress regulation, we investigate in this study the effects of acute administration of dexamethasone (Dex) on both in vivo and in vitro SS release. Freely moving animals received stereotaxic implant of a push-pull cannula into the median eminence for 10 days, and then they were perfused with artificial cerebrospinal fluid for 120-150 min. An i.p. injection of Dex (200 or 300 micrograms/100 g) induced, 15-30 min later, a mean increase in SS hypothalamic output of 62.6 +/- 6.2% of basal secretion. By contrast, after 15 min incubation of hypothalamic fragments with either 10(-7) or 10(-6) M Dex, SS release decreased abruptly to 57.3 +/- 3.3% (n = 16; P < 0.001 compared with basal release) and 78.0 +/- 9.5% (n = 13; P < 0.05 compared with basal release) of basal release, respectively. Other Dex concentrations induced no variations, giving the dose-effect curve an abrupt "on-off" effect. The inhibitory effect was blocked by picrotoxin (10(-4) M) and was immediately reversed when Dex was removed from the medium. Specificity was tested by using another steroid, estradiol, and another tissue, cortex. The rapid action of GC whatever the model used and in particular the blocking in vitro effect of picrotoxin could suggest that GCs act at the level of the membrane and could operate physiologically in response to stress. In addition, the opposite in vivo and in vitro effects on SS release would indicate that GCs exert two different controls on SS neurons.

We have previously reported that peripherally administered dexamethasone induces a rapid increase in hypothalamic somatostatin release. Here we investigated whether somatostatin synthesis could also be affected by this treatment and the potential involvement of glutamate in this effect. Male rats received a saline or a dexamethasone injection (300 microg/100 g body weight) and were killed 30 min later. Thirty minutes prior to dexamethasone treatment, another group received an i.p. injection of MK-801, a NMDA receptor antagonist. Cells expressing somatostatin mRNA in the periventricular nucleus were analyzed by in situ hybridization using digoxigenin-labeled somatostatin oligonucleotide probe. Dexamethasone decreased the number of digoxigenin-labeled cells expressing somatostatin mRNA in the periventricular nucleus as compared to the same histological sections from control rats. The dexamethasone effect was reversed by pretreatment with MK-801, which alone also decreased the number of cells expressing somatostatin mRNA. In summary, dexamethasone administration induces a significant rapid decrease in periventricular cells expressing somatostatin mRNA and this effect is partly abolished by MK-801.