NEUROENDOCRINOLOGY Somatostatin Inhibition of Gonadotropin-Releasing Hormone Neurons in Female and Male Mice Janardhan P. Bhattarai,* Attila Kasza´ s,* Seon Ah Park, Hua Yin, Soo Joung Park, Allan E. Herbison, Seong Kyu Han,† and Istva´n M. A´ braha´m† Department of Oral Physiology and Institute of Oral Bioscience (J.P.B., S.A.P., H.Y., S.J.P., S.K.H.), School of Dentistry and BK21 Program, Chonbuk National University, Jeonju, 561-756, Republic of Korea; Downloaded from https://academic.oup.com/endo/article/151/7/3258/2456783 by guest on 24 September 2021 Centre for Neuroendocrinology and Department of Physiology (A.E.H., I.M.A´ .), Otago School of Medical Sciences, University of Otago, Dunedin 9054, New Zealand; and Research Group of Neurobiology (A.K., I.M.A´ .), Hungarian Academy of Sciences, Eötvös Loránd University H-1119 Budapest, Hungary Previous studies indicate that somatostatin regulates gonadotropin secretion. We investigated here whether somatostatin has direct effects on GnRH neurons in the adult male and female mice. Dual-labeling immunofluorescence experiments revealed the presence of somatostatin-immuno- reactive fibers adjacent to GnRH neurons, and three-dimensional confocal reconstructions dem- onstrated apparent somatostatin fiber appositions with 50–60% of GnRH neurons located throughout the brain in both male and female mice. Perforated patch-clamp recordings from GnRH-green fluorescent protein neurons revealed that approximately 70% of GnRH neurons re- sponded in a dose-dependent manner to 10–300 nM somatostatin with an acute membrane hy- perpolarization and cessation of firing. This effect persisted in the presence of tetrodotoxin and amino acid receptor antagonists, indicating a direct postsynaptic site of action on the GnRH neu- ron. The identity of the somatostatin receptors underlying this action was assessed using GnRH neuron single-cell RT-PCR. Of the somatostatin receptor subtypes, the sstr2 transcript was the most prevalent and detected in both males and females. The expression of sstr2 by GnRH neurons was confirmed in the sstr2 knockout/LacZ knock-in mouse line. Electrophysiological studies demon- strated that the sstr2-selective agonist seglitide exerted acute hyperpolarizing actions on GnRH neurons identical to those of somatostatin. Together, these studies reveal somatostatin, acting through sstr2, to be one of the most potent inhibitors of electrical excitability of male and female GnRH neurons identified thus far. (Endocrinology 151: 3258–3266, 2010) omatostatin is an important neuromodulator peptide and female rats (3–6). Intracerebroventricular (icv) ad- Sin the central nervous system and alters the activity of ministration of somatostatin has been shown to inhibit LH neurons via five subtypes of G protein-coupled, soma- pulsatility in sheep (7) and depresses LH levels in rats (8). tostatin receptors (sstr1–5) (1). Somatostatin has been Furthermore, data from animals that overexpress or lack shown to modulate the reproductive axis at the level of the GH suggest that somatostatin may also modulate the re- pituitary; in male rats, somatostatin inhibits the GnRH- productive axis within the hypothalamus (9, 10). Indeed, induced release of LH in vitro (2) and decreases plasma LH Van Vugt et al. (8) demonstrated that icv administration concentrations in vivo (3). Moreover, central injections of somatostatin analog octreotide decreases the activation with somatostatin or a somatostatin analog resulted in of hypothalamic GnRH neurons and inhibits LH release in smaller and pyknotic gonadotropic cells and decreased rats. Although these results suggest that somatostatin af- gonadotropic cell numbers in the pituitaries of both male fects GnRH neurons, little if any attention has been given ISSN Print 0013-7227 ISSN Online 1945-7170 Abbreviations: ACSF, Artificial cerebrospinal fluid; AH, anterior hypothalamus; AP-5, l,2- Printed in U.S.A. amino-5-phosphonopentanoic acid; CNQX, 6-cyano-7-nitroquinoxaline-2,3-dione diso- Copyright © 2010 by The Endocrine Society dium salt; GABA, ␥-aminobutyric acid; GFP, green fluorescent protein; icv, intracerebro- doi: 10.1210/en.2010-0148 Received February 4, 2010. Accepted March 30, 2010. ventricular; MS, medial septum; PFA, paraformaldehyde; RMP, resting membrane First Published Online April 21, 2010 potential; rPOA, rostral preoptic area; sstr, somatostatin receptor; TBS, Tris-buffered saline; * J.P.B. and A.K. contributed equally to this work. TTX, tetrodotxin. † S.K.H. and I.M.A´ . share the senior authorship of the work. 3258 endo.endojournals.org Endocrinology, July 2010, 151(7):3258–3266 Endocrinology, July 2010, 151(7):3258–3266 endo.endojournals.org 3259 to the functional analysis of somatostatin inputs to GnRH was 80-␮m, resulting in 0.6-␮m optical thickness. Confocal neurons. stacks were prepared from each GnRH neuron (0.6-␮m steps, ϫ In the present study, we used immunocytochemistry, resolution 512 512 pixels) and analyzed using Imaris 5.0 im- age analysis software (Bitplane AG, Hannover, Germany). Con- single-cell gene profiling, and electrophysiology in a vari- voluted surfaces were created and used to identify the location ety of transgenic mouse models to characterize the effects and number of somatostatin appositions on GnRH neurons in of somatostatin upon GnRH neurons in the mouse. We the three-dimensional space. All somatostatin-GnRH neuron show here that most GnRH neurons are influenced di- connections were analyzed using three-dimensional reconstruc- rectly by somatostatin and that this occurs predominantly tion of the confocal images. GnRH neurons located in the medial septum (MS), rostral preoptic area (rPOA), and anterior hypo- through the sstr2 receptor and results in a profound inhi- thalamus (AH) were examined. Two sections representing each bition of GnRH neuron electrical activity. anatomical level were selected from each animal. The number of GnRH neurons and number of somatostatin appositions on each Downloaded from https://academic.oup.com/endo/article/151/7/3258/2456783 by guest on 24 September 2021 GnRH neuron was determined by an investigator blind to the experimental groupings. The number of GnRH neurons receiv- Materials and Methods ing somatostatin-containing appositions was calculated as per- centage of GnRH neurons at each anatomical level. Animals All experiments were approved by the Ethics Committees of Chonbuk National University Animal Welfare, the Babraham In- Histochemistry of sstr2 expression stitute, or Eo¨ tvo¨ s Lora´nd University. Male and female postnatal Six adult male and six female sstr2ϩ/lacZ mice were perfused 45–60 d GnRH-green fluorescent protein (GFP) (11) or female with 3% PFA and coronal sections through the MS, rPOA, and (48–58 d old) heterozygous sstr2 knockout/lacZ knock-in mice AH were processed for X-gal histochemistry by washing in TBS (sstr2ϩ/lacZ) (12, 13) were used for neuroanatomical studies. followed by X-gal solution [2 mM MgCl2,4mM K3Fe(CN)6,4 ␤ Adult male and female (45–60 d old) transgenic GnRH-GFP-mut5 mM K4Fe(CN)6, and 4 mg/ml 5-bromo-4-chloro-3-indoyl- -D- mice (14) were used for single-cell RT-PCR and electrophysiolog- galactosidase in TBS] overnight at room temperature to reveal ical analysis. The estrous stage of female mice was assessed by vag- lacZ-expressing cells within the brain sections. Sections were then inal smear, and those in diestrus were selected for experimentation. processed for GnRH immunocytochemistry using an LR1 antibody All mice were housed under 12-h light, 12-h dark cycles (lights on and diaminobenzidine chromogen as detailed previously (17). at 0700 h) with ad libitum access to food and water. GnRH neurons located in MS, rPOA, and AH were examined. Two sections from each brain region were selected, and the numbers of single (GnRH) and double-labeled (GnRH plus X-gal staining) neu- Tissue processing for immunohistochemistry and rons were determined. The X-gal expression in GnRH-immunore- histochemistry active neurons was calculated as the percentage of total number of For somatostatin immunohistochemistry and X-gal histo- GnRH-immunoreactive neurons in each region. chemistry, animals were deeply anesthetized with Avertin and transcardially perfused with 4% paraformaldehyde (PFA). In case of sstr2ϩ/lacZ animals, 3% PFA was used to avoid over- Brain slice preparation and electrophysiology fixation of the ␤-galactosidase enzyme. Brains were removed, Brain slices were acutely prepared as described (18). Mice postfixed for 2 h, and placed into 30% sucrose Tris-buffered were decapitated and brains rapidly removed and placed in the saline (TBS) solution overnight at 4 C. The next day, a 1:4 series ice-cold bicarbonate-buffered artificial cerebrospinal fluid of 30-␮m-thick coronal sections were cut through the septum (ACSF) of the following composition (in mM): 126 NaCl, 2.5 and hypothalamus on a sliding microtome. KCl, 2.4 CaCl2, 1.2 MgCl2,11D-glucose, 1.4 NaH2PO4, and 25 NaHCO3 (pH 7.4 when bubbled with 95% O2 and 5% CO2). Brains were blocked and glued with cyanoacrylate to the chilled Immunohistochemistry and analysis of stage of a vibratome (Microm, Walldorf, Germany), and 150- to somatostatinergic appositions 200-␮m-thick coronal slices containing the rPOA were cut. The After TBS wash, the sections were incubated in a primary slices were placed in oxygenated ACSF for at least1hatroom antibody directed against somatostain-14 (T-4103, 1:20,000; temperature. The slices were transferred to the recording cham- Peninsula, St. Helens, UK) for 48 h at 4 C. This was followed by ber, held submerged, and continuously superfused with ACSF at donkey antirabbit Cy5 (1:400 for 4 h; Jackson Laboratories, a rate of 4–5 ml/min. The slices were viewed with an upright Cambridge, UK). The specificity of the somatostatin antibody microscope (BX51WI; Olympus, Tokyo, Japan) and fluorescent has been reported previously in multiple rodent species (15, 16). GnRH neurons identified at ϫ10 and ϫ40 objective magnifica- The omission of the primary antibody in these studies resulted in tion by brief fluorescence illumination and then viewed and a complete absence of immunoreactivity.