N-Methyl-d -aspartate receptor involvement in dexamethasone and stress-induced hypothalamic somatostatin release in rats

ELSEVIER

Neuroscience Letters 219 (1996) 203-206

NEUROSCIENC[ LETTERS

N-Methyl-D-aspartate receptor involvement in dexamethasone and stressinduced hypothalamic somatostatin release in rats Corinne

E s t u p i n a a, J o r g e A b a r c a b, S a n d o r A r a n c i b i a a'*, J o r g e B e l m a r b

aLaboratoire 'Plasticite" et Adaptation Cellulaires', ERS 5644 du CNRS, Universitc~ Montpellier 2, Place EugOne Bamillon, 34095 Montpellier Cedex 5, France bLal~oratory of Biochemical Pharmacology, Catholic" University of Chile, Santiago, Chile

Received 22 July 1996, revised version received 21 October 1996; accepted 28 Oclober 1996

Abstract

The median eminence (ME) push-pull perfusion technique was used in this work and the results clearly showed that i.p. administration of MK-801 (4 mg/kg), a specific N-methyl-D-aspartate (NMDA) receptor antagonist, totally abolished dexamethasone (Dex) (300 k~g/ 100 g i.p. injected) and immobilization stress-induced hypothalamic somatostatin release in adult male rats. We also observed that glutamate from median eminence-hypothalamic medio basal (ME-MBH) complex, measured by high performance liquid chromatography (HPLC), exhibited a conspicuous secretory pattern, with the total amount released not modified by Dex administration. This indicates that Dex and stress-induced somatostatin (SS) secretion is not mediated by endogenous glutamate variations but likely by activation of NMDA receptors. Keywords." Glutamate; N-Methyl-D-aspartate receptor; MK-801; Somatostatin; Dexamethasone; Stress; Hypothalamus

Various kinds of stress are able to increase hypothalamic somatostatin (SS) secretion [1,7]. In particular, we have shown that hypothalamic SS release from the median eminence (ME) is rapidly and sharply increased after application of acute immobilization (IMO) stress [4] to adult male rats. W e also recently described that acute administration of dexamethasone (Dex), a potent glucocorticoid (GC) analog, induces a strong peak of SS secretion in the ME [9]. Both stress- and GCs-induced SS secretion could be linked events since synthetic GCs can be administrated to mimic the level of plasma adrenocortical steroids occurring during stress [20]. In long-term treatments, studies have shown that GCs induce variations in hypothalamic SS release and content [16]. In vivo, acute administration of Dex increased SS secretion from the ME but, in vitro, SS release from hypothalamic blocks was decreased [9]. These contrasting results suggest that extrahypothalamic neurotransmitters are involved in Dex-induced SS secretion. Glutamate (Glu), the major central nervous system (CNS) excitatory amino * Corresponding author. Tel.: +33 467144249; fax: +33 467144251.

acid, appears to be a good candidate since it positively controls hypothalamic SS release through N-methyl-Daspartate (NMDA) receptors [5,18]. Immunocytochemical studies have highlighted the presence of large amounts of Glu in presynaptic boutons in ventromedian (VMN) and arcuate (NA) nuclei [21], where SS neurons are also present. Furthermore, the presence of Glu has been observed in both nuclei [15], and N M D A receptor m R N A have been detected in the N A [22]. Moreover, in several CNS regions, Glu levels and N M D A receptors binding are rapidly altered after stress [2,15]. The aim of the present work was to determine whether N M D A receptors are involved in both stress- and Dexinduced SS release. We thus i.p. injected the specific N M D A receptor antagonist, MK-801, prior to application of either IMO stress or acute administration of Dex to freemoving adult male rats whose hypothalamic SS secretion was monitored by the push-pull perfusion (PPP) technique [3]. In parallel, with the same technique associated with high performance liquid chromatography (HPLC) [ 13], we measured Glu release under basal and experimental conditions to determine whether or not hypothalamic medio

0304-3940/96/$12.00 © 1996 Elsevier Science Ireland Ltd. All rights reserved Pll S0304-3940(96)13211-0

204

C. Estupina et al./ Neuroscience Letters 219 (1996) 203 206 80

80 -

MK-801 Dex

0

3'0

6'0

9'0 12 0 Time (min)

0

30

60

90

120 150 180 T i m e (min)

Fig. 1. Individual representations of SS secretory profiles from the MEMBH complex in male rats. In the left panel, animal was either IMO stressed (full line) or injected with MK-801 before stress (dashed line). In the right panel, animal received only Dex injection (full line) or consecutively Dex and MK-801 injections (dashed line). Arrows indicate the time of i.p. injection or the IMO stress period.

basal (MBH) Glu secretion could be involved in Dexinduced hypothalamic SS secretion. Male S p r a g u e - D a w l e y rats (Iffa-Credo, Lyon, France) weighing 2 4 0 - 2 8 0 g were used. Upon arrival they were caged under conditions of controlled temperature (22 _+ I°C) and light/dark cycles (light on from 0 7 0 0 1900 h), and were given water and laboratory chow ad libitum. A PPP cannula was implanted in their ME [3], where most hypothalamic SS is secreted, and PPP was performed 8 days later with artificial cerebrospinal fluid at a regular flow rate of 17 #l/min. In view of the slight diffusion expected during PPP, it is likely that ME and MBH secretion had collected in our perfusate [111. Therefore, we prefer to use, as other authors [11] the term MEMBH complex. Samples (270/_d) were collected every 15 min for 150-180 min and aliquoted to test either SS or Glu release in the perfusate. SS was extracted by adding 10% acetic acid 10 N, and glutamate was protected by adding 10% perchloride acid 10 N. Samples were then frozen and kept at - 4 0 ° C until assay. At the end of the experiments, all animals were sacrificed, and their brains removed and fixed for histological examination. Glutamate was measured by HPLC [13], and immunoreactive SS was determined by a sensitive RIA from Patel and Reichlin, as reported elsewhere [4]. Intra- and interassay coefficients of variation were 6.2% (n = 24) and 10.6% (n = 16), respectively. By convention, values equal to or lower than 5 pg/15 min (limit of detection of the assay) were referenced as 5 pg/15 rain in the statistical analysis. After the first 60 min of perfusion (basal conditions), a specific N M D A receptor antagonist, MK-801 (RBI, Natik, USA) diluted in phosphate-buffered saline (PBS) with NaOH 0.1 N was i.p. administered (4 mg/kg) to some rats which were 30 rain later subjected to IMO stress for 60 min. This stress consisted of taping the limbs to metal mounts attached to a wooden board, as reported elsewhere [4]. Another group of rats received a single i.p. dose (300 ~g/100 g) of Dex (Sigma, St. Louis, USA), a potent GC analog, which has already been shown to be effective on SS secretion [9], diluted in 300 >1 of saline 30 min after injection of MK-801. A third group was only Dex i.p.

injected after 60 min perfusion in basal conditions. As control, a fourth group was injected with saline alone. Since we previously showed that Dex [9] and stress [4] effects occur within 45 min, SS release values obtained within 45 min after stress or Dex administration, preceded by MK-801 injection, were calculated for each animal. The mean value was compared to mean basal values. A Student's t-test was performed for the data analysis. P < 0.05 were considered significant. Increased SS values were considered in each animal as a response to Dex administration when the release raised 4fold the value of intra-assay coefficient of variation, as reported elsewhere [9], i.e. 25% in the present case versus basal values. Glu peaks were characterized by the modified method of Santen and Bardin [19]. Since it was not possible in our short-term experimental conditions to determine the precise frequency of glutamatergic secretory peaks, we compared the total amount of Glu secreted 45 min before Dex injection with the total amount secreted 45 min post injection. This time was chosen because we have previously shown that increased secretion of SS occurs 15-45 min alter Dex injection [9]. A Student's paired t-test was applied to these data. P _< 0.05 were considered as significant. Prior injection of MK-801 totally canceled both Dex and stress-induced SS hypothalamic release (Fig. 1). Indeed, IMO stress increased SS release (251.0 + 10.0%; n = 2; P < 0.01 versus basal release), as previously observed [4], but no variation in SS secretion occurred when MK-801 was preinjected (84.7 _ 11.8% of basal release; n = 5: P > 0.05 versus basal release) (Fig. 2). In addition, Dex administration preceded by MK-801 injection did not boost SS secretion (96.8 +_ 19.3% of basal release; n = 7, P > 0.05 versus basal release), whereas it did when injected alone (353.2 _+ 55.2%; n = 12; P < 0.01 versus basal release) [91. (Fig. 2). In vivo SS release was not modified when N M D A specific antagonist was injected alone. The pattern of Glu secretion obtained from the MEMBH complex is characterized by irregular peaks which 450

300 250

i

200

350 300 250

150

7 loo ~

"~

~ 1so ,oo

0

Basal

Stress

MK-801 +Stress

0

Basal

Dex

MK-801 + Dex

Fig. 2. Statistical representation of variations in SS release expressed as percent ( +_ SEM) of basal release. In the left panel, animals were either 1MO stressed or injected with MK-801 before stress. In the right panel, animals received only Dex injection or consecutively Dex and MK-801 injections. *P < 0.001; P-values versus basal release. Number of ani reals used is given in each column.

C. Estupina et al./ Neuroscience Letters 219 (1996) 203-206

Glu .... n-,-.

SS

80

60

~

40

~,

i 20

•

0 0

_•

,

,

30

60

.

,

90

.

,

.

0

150 120 Time ( m i n )

3-

2,S-

~. 1-

i o,s, ~

e

Bml

Fig. 3. Upper panel, Individual representations of Glu and SS secretory profiles under basal conditions from the ME-MBH complex in male rats. Bottom panel, Statistical representation of total amount of Glu release in the ME-MBH complex 45 min prior or after Dex injection. Number of animals used is given in each column.

generally appeared every 4 5 - 6 0 min (Fig. 3). There was no significant difference (P > 0.05) between the total amount of Glu secretion 45 min before Dex injection (1.97 + 0.88 /~mol/15 min; n = 6) and the total amount secreted 45 min post injection (2.03 + 0.90 /zmol/15 min; n = 6) (Fig. 3). The main finding of this work was that N M D A receptors were involved in the physiological response to stress of SS secreting neurons. Indeed, the peak of SS release induced by IMO stress [4] was totally abolished by prior i.p. injection of MK-801, a specific N M D A receptor antagonist (Figs 1 and 2). Furthermore, the acute stimulatory effect of Dex on in vivo SS release was also mediated by glutamate, since the administration of MK-801 prevented Dexinduced SS release (Figs. 1 and 2). We observed no variations in Glu secretion from the ME-MBH complex after Dex administration (Fig. 3), indicating that the action of Glu on Dex-induced SS secretion was not mediated by variations in its ME-MBH release, but rather by activation of N M D A receptors. We previously suggested that an extrahypothalamic glutamatergic influence could be involved in the Dex stimulatory effect on SS secretion. Indeed, the fact that acute in vivo administration of Dex increases SS secretion, whereas the opposite effect is obtained under in vitro conditions [9] indicates that a stimulatory influence is disrupted in vitro. The present results confirm that Glu plays a stimulatory neurotransmitter role in both Dex and stress-induced SS release since the administration of MK-801 prevented SS increase in the two paradigms (Figs 1 and 2). Our laboratory has demonstrated that in basal conditions

205

glutamate controls, both in vivo and in vitro, hypothalamic SS release through N M D A receptors [5,18]. In vivo, this control seems to occur at the MBH level where immunocytochemical studies have described the presence of large amounts of Glu in presynaptic boutons of V M N and NA [21 ]. Glutamatergic control of S S neurons located in VMN and NA of M B H could thus be involved in the stress- and Dex-induced SS secretion. The stimulatory glutamatergic influence could be the result of a fine interplay between Glu locally produced at hypothalamic level [23] and Glu arising from an extrahypothalamic source, likely the hippocampus, since the deafferentation of fornix fibers leads to a decrease in Glu MBH content [24]. Either stress [15] or Dex [12] treatment increase Glu levels in the hippocampus and adrenalectomy markedly attenuates the stress-induced elevation in hippocampic extracellular glutamate levels [14]. Since the largest amount of SS collected in our perfusate is secreted by nerve endings arising from the periventricular nucleus (PEN), which is not comprised in the MBH, we cannot exclude that Glu also acts on these neurons. In fact, Glu has been detected in the PeN [15] and i.p. injected MK-801 is able to reach this nucleus by crossing the brain blood barrier. We must keep in mind that periventricular SS is known to control stress [4] and GCs [25] responsive somatotrope cells. Another possibility is that periventricular SS release could be indirectly controlled by glutamatergic influences through intrahypothalamic connections established between arcuate SS neurons of the MBH and periventricular ones [10]. While we showed that N M D A receptors were involved in stress- and Dex-induced SS secretion, we did not observe any variation in the total amount of Glu released in the ME from the ME-MBH complex after Dex administration. This finding is in keeping with no change observed in the hypothalamic Glu content after acute stress [15]. In basal conditions, the Glu secretion pattern shows irregular peaks (Fig. 3), as already observed in the preoptic area [8]. However, our short-term paradigm of perfusion did not enable us to detect possible variations in pulsatile patterns of Glu release after Dex injection. The absence of changes in Glu release from the ME-MBH complex, while administration of its antagonist abolished the stress and Dex-increased SS secretion, indicates that Dex, and likely stress, modify the glutamatergic transmission at the receptor level. In fact, it has been shown in mouse forebrain that acute swim stress induces changes in the MK-801 binding [2]. GCs could also indirectly modulate the glutamatergic transmission at membrane level since they can be metabolized into neuroactive steroids. These levels are known to be rapidly increased during stress [17] and some neuroactive steroids are able to induce potentiation of the neuronal N M D A response [6]. In conclusion, our results clearly show that Glu is involved in the Dex- and stress- induced SS release. Indeed, the specific N M D A receptor antagonist, MK-

206

C. Estlq~ina et al.I Neurosciem'e Letters 219 (1996) 203 20{~

801, totally abolished Dex- and stress- induced SS release. Since Dex does not induce variations in Glu release, Dex and likely IMO stress, activate N M D A receptors rather than modify Glu secretion. Stress and GCs action on SS release could be two linked events since stress and Dex, which activate the GCs receptor, seem to have a c o m m o n glutamatergic pathway involved in their effects on SS release. However, further work is necessary for strengthen this hypothesis. This work has been partly supported by INSERM (France) and CONICYT (Chile). Technical assistance of Mr. E. Savary is gratefully appreciated as well as the valuable contribution of Mr. D. Lyonnet. Thanks are given to Dr. A. Armario for critical reading of the manuscript. [1] Aguila, M.C., Pickle, R.L., Yu, W.H. and McCann, S., Role of somatostatin and growth hormone-releasing factor in ether stress inhibition of growth hormone release. Neuroendocrinology, 54 (19911 515 520. [21 Akinci, M.A. and Johnston, G.A.R., Sex differences in acute swiln stress-induced changes in the binding of MK-801 to the NMDA subclass of glutamate receptors in mouse forebrain, J. Neurochem., 61 (19937 2290-2293. [31 Arancibia, S,, La technique de perfusion 'push pull' en Neuroendocrinologie, Ann. Endocrinol. (Paris). 48 (1987)410 418. [4] Benyassi, A., Gavaldh. A., Armario, A. and Arancibia, S., Role of somatostatin in the acute immobilization stress-induced GH decrease in rat, Life Sci., 52 (1993) 361 370. [5] Benyassi, A., Tapia-Arancibia. L. and Arancibia, S., Glutamate peripherally administered exerts somaloslatin releasing action m the conscious rat. J. Neuroendocrinol., 3 (1991) 429 432. [6] Bergeron, R., Montigny. C.[,. and Debonnel, G., Potentiation of neuronal NMDA response induced by dehyroepiandrosterone and iis suppression by progesterone: effects mediated via sigma receptors, J. Neurosci., 16 (1996) 1193 1202. [71 Cataldi, M., Magnan, E., Guillaume, V.. Dutour, A., Sauze, N., Mazzocchi, L., Conte-Devolx, B. and Oliver, C., Acute stress stimulates secretion of GHRH and somatostatin into hypophysial portal blood of conscious sheep, Neurosci. Left., 178 (19941 103 106. [8] Demling, J., Fuchs, E., Baumert, M. and Wuttke, W., Preoptic cathecholamine, GABA, and glutamate release in ovariectomized and ovariectomized estrogen-primed rats utilizing a push-pull cannula technique, Neuroendocrinology, 41 ( 19851 212-218. [91 Estupina, C., Tapia-Arancibia, L., Astier, H., Belmar, J. and Arancibia, S., Rapid and opposite effects of dexamethasone on in rive and in vitro hypothalamic somatostatin release, Exp. Brain Res.. (19961 in press. [10] Fodor, M., Csaba, Z., Kordon. G. and Epelbaum, J., Growth hormone-releasing hormone, somatostalin, galanin, and b-endorphin

[1 I]

[121

[131

[141

1151

1161

[171 [lgl

[ 191

120]

[21]

[22]

[23]

1241

[251

aflerents to the hypothalanms periventficular nucleus. J. Chem. Neuroanal., g (19941 61-73. Jan'y, H.. Spreuger, M. and Wuttke, W., Rates of release of GABA and calecholamines in the mediobasal hypothalamus of owu'iectomized and ovariectomized estrogen-treated rats: correlation with blood prolactin levels, Neuroendocrinology, 44 (1986) 422 428. Kite, S., Miyoshi. R., Shimoyama, M. and Shimada. R., Effect of glucocorticoid on in vivo release of neurotransmitlers from the rul hippocampus, 91h International Congress of Endocrinology, Nice. 1992, Absiracl P-18.04.034, p. 555. Lindroih, P. and Mopper, K., l tigh peiTormancc liquid chronlatographic delei'nlinalion of subpicomole anlounts el alllJno acids by precolumm fluorescence dcrivaiization with o phthaldehyde. Anal. Chem., 51 (1979) 1667 1674. Lowy, M.T.. Gaull, L. and Yamamolo, B.. Adrenulectolny attenuales stress-induced elevations in exlracellular glutamate concenml lion in the hippocampus, J. Neurochem., 61 (19931 1957 1960. Palkovits. M., Lang, T., Patlhy, A. and Elekcs, I., Disiribufion and stress mdnced increase of glutamate and asparlute levels in discrele brain nuclei of rat, Brain Res.,373 (19851252 257. Papachrislou. D.N., Liu, J.L. and Patel, Y.C.. Glucocorlicoids reg ulale somatostatin peptide and steady state messenger rihonucleic acid levels m nomlal rat tissues and in a somatostatin-producing islel tumor cell line (1027Bo, Endocrinology, 134 (1994~ 2259 2266. Paul, S.M. and Purdy, R.H., Neuroactive steroids, FASEB J., 6(1992) 2311 2322. Rage, F., Benyassi. A.. Arancibia, S. and Tapia-Arancibia. L., 7 Aminobulyric acid glutamate interaclion in the control of soinalos talin release from hypolhalamic neurons in primary culture: in viw3 corroboration, Endocrinology, 130 ( 19921 1056 1(162. Richard, J.L., Bringer, J., Daures, J.P., Parer-Richard, C., Jafliol, C. and Mirouze, J., Mdlhodes d'analyse de la pulsalilild hormollale, Ann. Endocrinol., 51 (1990) 181 193. Sapolsky, R.M., Krey. L.C. and McEven, B.S., ]'he neuroendocri nology of stress and aging: the glucocorticnids cascade hypolhesis, Endocrinol. Rev., 7 (19861 284-301. Van den Pol, A.N., Glutamate and aspartate immure)reactivity in bypolhalanfic presyuaptic axons. J. Neurosci., I I (1991) 2087 2101. Vun den Pol. A.N.. Hermans Borgmeyer, 1., Hofer, M., Ghosh, P. and Heinemann. S.. Ionotropic glutamate-receptor gene expression in hypothalamus: localization of AMPA, kainate, and NMDA receptor RNA with in silu hybridization, J. Comp. Ncurol., 343 (1994) 428 444. Van den Pol, A.N. and Trombley, P.Q., Glutamate neurons in hypothalamus regulate excitatory transmission, ,I. Neurosci., 13 (19931 2829 2836. Walaas, 1. and Fonnum, F.. Biochemical evidence for glutamate as a transmitter in hippocampal efl'erents to the basal forebrain and hypothalamus in the rat brain, Neuroscience, 5 (19801 1691 1698. Wehrenberg. W.B., Bergman, P.J., Slagg, L., Ndon. J. and Giustina, A.. Glucocorlicoid inhibition of growth m rats: partial reversal with sonlatostatin antibodies, Endocrinology, 127 (19901 2705-2708.