Hélène Astier

Neuroendocrine and Autonomous Mechanisms Underlying Thermoregulation in Cold Environment

This review focuses on the central regulation of thermoregulatory responses with special attention to the participation of thyrotropin-releasing hormone (TRH) in both autonomous and endocrine responses to a cold environment. Besides a direct projection of TRH neurons from paraventricular nuclei (PVN) to the median eminence, and the subsequent activation of the thyroid axis, there are direct projections from the PVN to the autonomic preganglionic neurons controlling autonomous responses. There projections convey information to peripheral targets involved in thermogenesis through the dorsal vagal complex and the spinal cord, for parasympathetic and sympathetic neurotransmissions respectively. Furthermore, cold exposure increases TRH mRNA levels in the PVN but also in dorsal motor and caudal raphe nuclei, thus providing strong evidence for a functional link between autonomous and neuroendocrine systems involved in thermoregulation. The review also focuses on neuroendocrine regulation of cold-induced TRH/TSH release associated with modifications in somatostatin release, with special reference to the participation of several central neurotransmitters (catecholamines, serotonin or GABA) or the influence of sex steroids.

Direct Evidence of Short-Term Cold-Induced TRH Release in the Median Eminence of Unanesthetized Rats

IR-TRH release in the median eminence was directly estimated in conscious rats during the first 130 min of exposure to cold (4 degrees C), using a push-pull cannulation. A three-fold increase in IR-TRH release was observed, with a peak of 10.00 +/- 2.19 pg/15 min occurring 40 min after exposure to cold; control rats, left at 24 degrees C, stayed at the baseline secretion rate of 3.40 pg/15 min which was the sensitivity limit of the RIA assay.

Physiological evidence for α1-adrenergic facilitatory control of the cold-induced TRH release in the rat, obtained by push-pull cannulation of the median eminence

The alpha-adrenergic antagonists phentolamine and prazosin were administered to male rats to explore their effects on cold-induced TRH release, measured by a chronic push-pull cannula stereotaxically implanted in the median eminence (ME). Phentolamine was given either i.p. (24 or 40 mg/kg), or locally (10(-5) M) in the ME, whereas prazosin was only applied locally (10(-5) M). Phentolamine significantly decreased the cold response (5 +/- 1 pg/15 min vs 21 +/- 5 pg/15 min; P less than 0.02), whatever the administration mode. Moreover, the blocking effect of prazosin directly perfused into the ME (11 +/- 3 pg/15 min vs 26 +/- 9 pg/15 min; P less than 0.05), indicates the specific involvement of alpha 1-adrenergic receptors in the cold-induced TRH response, and points to the ME as a possible site of facilitatory adrenergic control.

The Anterior Periventricular Hypothalamus Is the Site of Somatostatin Inhibition on Its Own Release: An in vitro and Immunocytochemical Study

The site of action of the inhibitory effect of somatostatin (SRIF) on its own release was studied by: (1) measuring SRIF release in vitro from tissue preparations containing either the proximal (periventricular hypothalamus) or the distal (median eminence) portions of the hypothalamic SRIF neurons, and (2) immunocytochemical investigation of the interconnections occurring between SRIF neuronal elements in these hypothalamic regions. In vitro, a biologically active, but noncross-reacting SRIF analog (D-Trp8 SRIF) in the RIA, inhibited 25 mM K+ induced SRIF release from anterior periventricular hypothalamic tissues. The inhibitory effect of D-Trp8 SRIF was dose-dependent, maximal at 10(-7) M, and restricted to this anterior region, since median-eminence SRIF release was not modified by the presence of D-Trp8 SRIF. Additionally, LHRH release from anterior periventricular hypothalamus was unchanged in the presence of D-Trp8 SRIF. In the periventricular nucleus, perikarya and dendrites of labeled SRIF neurons showed frequent apposition of their limiting membranes. Classical synapses were also observed between SRIF-containing axonal processes and labeled perikarya or dendrites. Although membrane appositions between neighboring SRIF axons frequently occurred in the median eminence, no synaptic-like SRIF-SRIF connections could be detected at this level. The data demonstrate a direct inhibitory action of a SRIF agonist on the anterior periventricular hypothalamic release of the peptide. This effect correlates well with the occurrence of SRIF-SRIF synapses in this region; suggesting that SRIF exerts a negative feedback in the control of its own release through autoreceptors located on the perikarya or dendrites of SRIF-containing neurons.

K+-induced thyrotropin-releasing hormone release from superfused mediobasal hypothalami in rats. Inhibition by somatostatin

Somatostatin (SRIF), in concentration of 10(-6) M, significantly inhibited the depolarization-induced release of immunoreactive thyrotropin-releasing hormone (IR-TRH) from superfused mediobasal hypothalami (MBH) containing mainly the median eminence (ME), without affecting the basal release of TRH. The total amount of K+-induced TRH release was 0.24 +/- 0.02 and 0.61 +/- 0.08 pg/MBH/min, respectively, in the presence and absence of SRIF in the medium. The data are consistent with a role of SRIF as a neuromodulator on TRH release from the ME. In contrast, superfusion with Locke medium containing triiodothyronine (10(-6) M) had no effect on basal and K+-induced IR-TRH release in our system.

Actions of Excitatory Amino Acids on Somatostatin Release from Cortical Neurons in Primary Cultures

Abstract: L‐Glutamate, N‐methyl‐D‐aspartic acid (NMDA), quisqualate, and kainate were found to increase endogenous somatostatin release from primary cultures of rat cortical neurons in a dose‐dependent manner. The rank order of potency calculated from the dose‐response curves was quisqualate > glutamate = NMDA > kainate, with EC50 values of 0.4, 20, and 40 μM, respectively. Alanine, glutamine, and glycine did not modify the release of somatostatin. The stimulation of somatostatin release elicited by L‐glutamate was Ca2+ dependent, was decreased by Mg2+, and was blocked by DL‐amino‐5‐phosphonovaleric acid (APV) and thienyl‐phencyclidine (TCP), two specific antagonists of NMDA receptors. The NMDA stimulatory effect was strongly inhibited by APV in a competitive manner (IC50= 50 μM) and by TCP in a noncompetitive manner (IC50= 90 nM). The release of somatostatin induced by the excitatory amino acid agonists was not blocked by tetrodotoxin (1 μM), a result suggesting that tetrodotoxin‐sensitive, sodium‐dependent action potentials are not involved in the effect. Somatostatin release in response to NMDA was potentiated by glycine, but the inhibitory strychnine‐sensitive glycine receptor did not appear to be involved. Our data suggest that glutamate exerts its stimulatory action on somatostatin release essentially through an NMDA receptor subtype.

Ontogeny of the metencephalic, mesencephalic and diencephalic content of catecholamines as measured by high performance liquid chromatography with electrochemical detection

Developmental changes in norepinephrine (NE), dopamine (DA) and epinephrine (E) contents of the rat metencephalon, mesencephalon and diencephalon, have been measured by high performance liquid chromatography with electrochemical detection, from fetal stages (E15 or E17 to E21) to postnatal days (P0 to P30) and compared to the adult levels. The data show a biphasic pattern in NE changes of the three brain areas, with a first increase in the late prenatal period, followed by a further development from PO to P18, thus reaching the adult levels. A similar pattern of development is found for the mesencephalic and diencephalic DA contents. The E levels of the diencephalon are very low in comparison to the NE and DA concentrations, but present a gradual increase from E17 to P18. The results correlate with the development of catecholamine systems in brain area as measured by other methodological approaches.

A Prepro-TRH Connecting Peptide (Prepro-TRH 160–169) Potentiates TRH-Induced TSH Release from Rat Perifused Pituitaries by Stimulating Dihydropyridine- and Omega-Conotoxin-Sensitive Ca2+ Channels

The stimulation of TSH secretion by TRH involves the phosphatidylinositol second messenger pathway via activation of phospholipase C. This effect is mediated by a GTP-binding protein and leads to a mobilization of intracellular Ca2+ stores and an activation of protein kinase C. However, TRH stimulation also results in an influx of extracellular Ca2+. Since we have previously demonstrated that a non-TRH fragment of the prepro-TRH molecule, the connecting peptide PS4 (prepro-TRH 160-169), was able to potentiate the TRH-induced TSH release in a dose-dependent manner, we attempted to determine whether this potentiation might be due to a Ca(2+)-dependent phenomenon and whether a specific class of voltage-dependent Ca2+ channels, the L type Ca2+ channels, might be involved in the effect of PS4. This was studied by perifusing normal pituitary fragments with medium containing either the Ca2+ ionophore, ionomycin, and Co2+ ions, or organic compounds well known to block L-type Ca2+ channels, and by measuring the TSH response to a pulse of TRH (10 nM) in the presence or absence of PS4 (100 nM).(ABSTRACT TRUNCATED AT 250 WORDS)

Adenylate cyclase activation is not sufficient to stimulate somatostatin release from dispersed cerebral cortical and diencephalic cells in glia-free cultures.

Under conditions in which vasoactive intestinal peptide (VIP) induces somatostatin release from cortical and diencephalic neuronal cultures, VIP causes large increases in intracellular cyclic AMP. Both the release of somatostatin and the increase in cyclic AMP elicited by VIP require exogenous calcium, can be blocked by cobalt ion, and can be qualitatively mimicked by depolarizating concentrations of exogenous potassium ion. Direct activation of adenylate cyclase by forskolin causes large increases in cyclic AMP content but does not induce somatostatin release. In the absence of VIP, the calcium ionophore, ionomycin, and the phorbol ester, phorbol 12-myristate-13-acetate, also stimulate somatostatin release. These results indicate that VIP-stimulation of cyclic AMP formation and VIP-stimulation of somatostatin release are calcium-dependent and that the two phenomena are dissociatable. Cyclic AMP formation is not a necessary condition for VIP-induced somatostatin release. Nucleotide formation may be a sufficient condition for release or, possibly in association with calcium influx, it may be an event unrelated to the release process.

The presence of non-neuronal cells influences somatostatin release from cultured cerebral cortical cells

We examined the effect of non-neuronal cells on somatostatin release from cultured cerebral cortical cells. Three culture models were used: (1) neuron-enriched cultures obtained from cortex of 17-day-old rat embryos and exposed to 10 microM cytosine arabinoside (Ara C) for 48 h between days 3 and 5 after plating; (2) whole cell cultures obtained by using the same protocol but untreated with Ara C; (3) glial primary cultures obtained from newborn rats. We studied: (i) the cellular composition of the cultures by using two astroglial markers: vimentin and glial fibrillary acidic protein (GFAP); (ii) the spontaneous and forskolin-stimulated somatostatin release. In 8-day-old cultures morphological data revealed that Ara C treatment reduced glial cells to 6%. At 7 and 10 days of culture somatostatin spontaneously released from Ara C-treated cells was higher than that measured from untreated cells. On the 17th day of culture, neuron-enriched cultures contained a lower amount of somatostatin than whole cell cultures. Forskolin elicited a dose-dependent release of somatostatin from whole cell cultures, but had no effect on neuron-enriched cultures. Astroglial released media (ARM) from glial primary cultures exposed to forskolin for 20 min induced somatostatin release from neuron-enriched cultures. HPLC analysis of endogenous amino acids of ARM showed that glutamate, glutamine, glycine and alanine were significantly increased after forskolin stimulation. Our results suggest a functional interaction between glial cells and neurons secreting somatostatin.