Glutamate released in the preoptic area during sexual behavior controls local estrogen synthesis in male quail

Abstract

Estrogens are known to act rapidly, probably via membrane estrogen receptors, to induce fast effects on physiological and behavioral processes. Engaging in some of these behaviors, such as sexual behavior, results in an acute modulation of the production of estrogens in the brain by regulating the efficiency of the estrogen synthase enzyme, aromatase. Specifically, we recently demonstrated that aromatase activity (AA) in the male quail brain is rapidly inhibited in discrete brain regions including the medial preoptic nucleus (POM) following exposure to a female. Evidence from in vitro studies point to glutamate release as one of the mechanisms controlling these rapid regulations of the aromatase enzyme. Here, we show that (a) the acute injection of the glutamatergic agonist kainate into the POM of anesthetized male quail inhibits AA and (b) glutamate is released in the POM during copulation. These results provide the first set of in vivo data demonstrating a role for glutamate release in the rapid control of AA in the context of sexual behavior.

1. Introduction

In addition to their effects in the control of reproduction, estrogens synthesized in the brain also modulate a host of other behavioral or physiological traits including aggression, nociception, cognition, and neuroprotection (as reviewed in (Balthazart and Ball, 2013)). These effects on neuronal activity are largely mediated by the interaction of estrogens with their nuclear receptors (Bean et al., 2014; Ervin et al., 2015) but also through membrane-initiated events (Vasudevan and Pfaff, 2007). In the latter case, they result in acute and transient modulations of physiological and behavioral responses (Cornil and Charlier, 2010; Roepke et al., 2011). Rapid effects of estrogens have been reported on sexual and aggressive behavior, song processing and discrimination, social communication, locomotion, nociception, learning and cognitive processes in multiple species including fish, birds and rodents (reviewed in (Cornil et al., 2012; Laredo et al., 2014)).

Aromatase, the enzyme that catalyzes the transformation of androgens into estrogens, is expressed in the brain of all classes of vertebrates including humans (Azcoitia et al., 2011; Callard et al., 1978). As a result of higher expression of the protein, the avian brain expresses a higher aromatase activity (AA) as compared to rodents (Callard et al., 1978) and thus studies in birds have facilitated fundamental progress in the study of the function of brain estrogens and of the mechanisms that regulate their synthesis. In particular, it was found that AA can be rapidly regulated in specific brain regions by external stimuli such as the presence of a female (Cornil et al., 2005; de Bournonville et al., 2013) or an acute stress (Dickens et al., 2012; Dickens et al., 2011; Dickens et al., 2014). In parallel, estradiol concentration is also rapidly regulated in different brain regions in response to environmental stimuli such as the presence of a female (Remage-Healey et al., 2008), an aggressive interaction (Charlier et al., 2011b) or the acute exposure to a stressor (Dickens et al., 2014). These rapid changes in local synthesis of estradiol seem to be essential for the short-term control of several behaviors.

In Japanese quail (Coturnix japonica) in particular it was shown that male sexual behavior is rapidly modulated following systemic (Cornil et al., 2006a; Cornil et al., 2006b) or intracerebroventricular (Seredynski et al., 2013) injections of estrogens (stimulation), anti-estrogens or aromatase inhibitors (inhibition). Interestingly, these manipulations of estradiol bioavailability affect within minutes sexual motivation without changing sexual performance. It was also demonstrated that AA in quail is significantly decreased in the medial preoptic/hypothalamic area after 5 minutes of sexual interaction with a female (Cornil et al., 2005). These changes occur specifically in the medial preoptic nucleus (POM) and in the tuberal hypothalamus (de Bournonville et al., 2013) and similar decreases in AA occur after only viewing a female (Cornil et al., 2005; de Bournonville et al., 2013). Together, these results strongly suggest that rapid changes in the synthesis and action of estradiol are involved in the short-term regulation of male sexual motivation, while sexual performance would be activated by longer term, presumably genomic, actions of estrogens.

In vitro aromatase activity is rapidly (within 5 min) and transiently down-regulated by calcium-dependent phosphorylations (Balthazart et al., 2001 ; Balthazart et al., 2003 ; Charlier et al., 2011a). Similar transient inhibitions are observed following exposure of the medial preoptic-hypothalamic area (HPOA) explants maintained in vitro to the glutamatergic agonists AMPA (α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid) or kainate (Balthazart et al., 2006). In this paper, we therefore wanted to investigate whether the rapid changes in AA observed after copulation are mediated by endogenous glutamate release. This hypothesis was supported by the following lines of evidence: (1) glutamate receptors are expressed in regions of the avian brain containing aromatase cells (Cornil et al., 2000 ; Saldanha et al., 2004), (2) quail preoptic aromatase cells are sensitive to AMPA, kainate and NMDA in vitro (Cornil et al., 2004), (3) glutamate is released in rat medial preoptic area (MPOA) during sexual behavior (Dominguez et al., 2006) and (4) one study in zebra finches demonstrated that infusion of glutamate in an auditory telencephalic region containing aromatase leads to a transient decrease of estradiol concentration (Remage-Healey et al., 2008).

To this end, two groups of experiments were performed. Firstly, we tested in vivo whether AA in the POM is inhibited by a local injection of the glutamatergic agonist kainate. Secondly, we measured by in vivo microdialysis changes in glutamate concentration in the POM of males while they interacted with a female. These experiments identified a significant decrease of AA following kainate injection in the POM as well as a significant increase in preoptic glutamate extracellular concentration following copulation.

2. Materials and Methods

2.1. Subjects

Gonadally intact male Japanese quail (Coturnix japonica) derived from our breeding colony (University of Liege, Belgium) were used as subjects in these studies. Two sexually experienced females obtained from the same colony and selected for their high receptivity served as stimuli during the behavioral tests. All subjects were raised in groups until adulthood (approximately 8 weeks) and were housed individually from that time on. Subjects were maintained on a long day photoperiod (16h light and 8h dark) and provided with food and water ad libitum. All Experimental procedures were in agreement with Belgian laws on the “Protection and Welfare of animals” and on the “Protection of experimental animals” and were approved by the Ethics Committee for the Use of Animals at the University of Liege (Protocol #1235).

2.2. Experiment 1: Effects of kainate on preoptic aromatase activity

2.2.1. Stereotaxic injections

Anesthetized sexually naïve males were injected with kainate in the POM and AA was measured in this brain region immediately after the injections. Birds were anesthetized with Isoflurane and placed in a stereotaxic frame (David Kopf Instruments, Tujunga CA). The skull was exposed and a small hole was drilled. A glass pipette fixed on the arm of the stereotaxic frame and aiming at the POM was then lowered inside the brain (coordinates : 1.6 mm anterior (X) and 2.8 mm dorsal (Y) to the zero reference point (center of the interaural axis), 0.7 mm lateral to the sagittal midline (Z)). These coordinates were determined based on the quail atlas (Baylé et al., 1974) and adjusted to the specific size of our subjects in preliminary experiments. The glass pipette was connected to a 25 μl Hamilton syringe with EP50 tubing. A KDS pump (KDS 220 P #2004249) was used to inject the drug continuously at a speed of 0.2μl/min for 20 minutes (total injected volume of 4μl).

In experiment 1A, kainate was infused unilaterally in the left or right POM of 6 males, while 6 other subjects were similarly injected unilaterally with the vehicle. The injection side was counterbalanced between animals. In experiment 1B, bilateral injections were performed in the left and right POM at the same time in 15 subjects. One side was injected with kainate and the other one with the vehicle solution. The kainate side was counterbalanced between animals. In both experiments, animals were injected for 20 minutes and killed immediately after the end of the injection. The brain was quickly removed from the skull and directly frozen on dry ice (~ 1 min between the sacrifice and the freezing of the brain). Brains were stored at −80°C until microdissections.

Kainate (Tocris, Bristol, UK) was diluted at a final concentration of 100 μM in artificial cerebrospinal fluid (aCSF; 199 mM NaCl, 26.2 mM NaHCO₃, 2.5 mM KCl, 1 mM NaH₂PO₄, 1.3 mM MgSO₄, 2.5 mM CaCl, 11 mM glucose, pH = 7.4). This concentration was based on previous work of our laboratory demonstrating that in these conditions, AA is rapidly (within 5 min) inhibited in preoptic blocks maintained in vitro and this inhibition is rapidly reversed after wash-out of the drug (Balthazart et al., 2006). Fast green (1:1000) was added to the aCSF in order to check the location of the injection after sacrifice. Control injections were made of aCSF including fast green. Preliminary experiments had shown that at this concentration fast green does not have any effect of AA.

2.2.2. Microdissections

Frozen brains were sliced with a cryostat into 200 μm-thick coronal sections. While cutting, the experimenter checked whether fast green coloration was at the expected location i.e. the POM, based on anatomical landmarks, (primarily fiber tracts such as the tractus septopallio-mesencephalicus and the commissura anterior), that are visible in frozen sections in the absence of any histological coloration. Target brain nuclei containing aromatase cells were microdissected using the Palkovits punch method (Palkovits, 1973) adapted for quail (Dickens et al., 2014). Briefly, punches were dissected using a specialized 0.96 mm diameter brain punch apparatus (Leica Biosystems #39443001RM). From each brain, we collected the POM, the bed nucleus of the stria terminalis (BST) and the mediobasal hypothalamus (MBH). Since the POM and BST constitute a continuous aromatase cell population (Foidart et al 1995) and AA has been shown to be rapidly regulated after sexual behavior in both nuclei (de Bournonville et al., 2013), we decided to pool these two nuclei for the assays (POM+BST). For each animal, both POM+BST and MBH were collected separately on the left and right side of the brain in order to compare AA of the kainate-injected hemisphere with the control one (non-injected hemisphere in Exp. 1A or vehicle-injected hemisphere in Exp. 1B). Samples were kept frozen throughout and stored at −80°C until processing.

2.2.3. Aromatase activity assay

Micropunches were homogenized in 120 μl of ice cold TEK buffer (150 mM KCl, 1 mM Na-EDTA, and 10 mM Tris-HCl; pH 7.2) using a glass pestle designed for 1.5 ml Eppendorf tubes. Aromatase activity was determined in these samples by measuring the production of tritiated water during the conversion of [1β-³H]-androstenedione into estrone (Roselli and Resko, 1991) as previously described and validated for quail brain (Baillien and Balthazart, 1997) and adapted for micropunches (Dickens et al., 2014).

Briefly, 50 μl of each of the homogenates was incubated in duplicate in the presence of ³H-androstenedione (final concentration 25 nM, specific activity 20.7 Ci/mmol; Perkin-Elmer), NADPH (4.8 nM) and TEK buffer at 37°C for 15 min. This reaction was stopped by adding 2% charcoal in 10% trichloroacetic acid. ³H-Water was filtered through Dowex cation exchange columns. 10 mL of Optiphase Highsafe (Perkin Elmer) were added to the column eluates and vials were counted for 3 min on a Wallac Winspectral 1414 Liquid Scintillation Counter. Intra-assay coefficient of variation was less than 6.64 %. Protein content of each sample was measured using the commercial Coomassie Plus Protein Assay Reagent (Pierce, Rockford, IL).

Aromatase activity values were corrected for quenching, recovery, blank values and percentage of tritium in β-position in the substrate and finally, the value was normalized for the protein concentration in each sample (see (Baillien and Balthazart, 1997) for detail). With all of these corrections, AA was finally expressed as pmol of aromatization product synthesized per hour and per mg protein.

2.3. Experiment 2: Changes in preoptic glutamate during sexual interactions

2.3.1. Stereotaxic placement of the guide cannula

For experiment 2, a guide cannula was stereotaxically lowered under gas anesthesia to a position aimed to be just adjacent to the POM of adult male quail. The guide cannula was a 22 gauge plastic tubing, that was lowered to its target (X = + 1.6 ; Y = + 2.5 ; Z = ± 0.5) and secured to the skull with dental cement. The guide cannula was then obturated with a dummy insert that was 300 μm longer than the cannula until the microdialysis experiment began. Animals were allowed a 2 weeks recovery period after surgery.

2.3.2. Microdialysis procedure

Males were pre-tested several times for copulatory behavior until they had gained sexual experience and were habituated to the empty microdialysis chamber. On the morning of the microdialysis testing, the probe was perfused with aCSF during several minutes at a speed of 2μl/min. After that time, the probe was inserted into the guide cannula and the bird was placed in a glass chamber (60 cm × 30 cm × 35 cm aquarium). He was then perfused with aCSF at a speed of 1μl/min during at least 4 hours before the beginning of the experimental test in order to habituate to the environment and the perfusion. A tether encased the in- and out-flow tubing. The tubing was attached to a counter-balanced lever arm that prevented addition of extra weight on the head of the animal and to a swivel to avoid torsion of the tubing. The dialysis membrane (Microbiotech 4.15.2.PES) had an outer diameter of 0.2 mm, an active dialyzing length of 1.8 mm and a 6 kDa cutoff.

Two groups of males (n=9 and 11 respectively) were tested with slightly different protocols in experiments 2A and 2B. In experiment 2A, the male first remained alone for the first 30 min of the test during which samples of dialysate were collected every 3 min. The first 5 samples were discarded because preliminary results showed that glutamate outflow was initially heavily affected, in some subjects at least, by the collection procedure. The 5 samples collected during the end of these 30 minutes constituted the pre-experimental baseline period (PRE period). A receptive female was then introduced in the arena and the male was allowed to interact with her during another 15 minutes (INTERACTION period). Finally, the female was removed and the male stayed alone for the final 15 minutes (POST period). Consummatory sexual behavior was scored by direct observation during the 15 minutes of female presence. The latency and frequency of the stereotyped male sexual behaviors (namely neck grabs (NG), mount attempts (MA), mounts (M), and cloacal contact movements (CCM)) were recorded (for a detailed description, see (Adkins and Adler, 1972; Hutchison, 1978)).

In experiment 2B, the animal was submitted to the same experimental conditions as in experiment 2A (i.e., PRE-INTERACTION-POST periods) but before the Interaction period, the male was allowed to see the female during 9 minutes while he was physically separated from her by a glass partition (VIEW period).

In both experiments, samples were collected for future glutamate assay every 3 minutes, providing 5 samples during the PRE, 3 samples during the VIEW (only in Exp. 2B), 5 samples during the INTERACTION and 5 samples during the POST periods. Brains were collected during the following days to determine the probe location. Before brain collection, birds were deeply anesthetized, an electrical lesion was performed to mark the brain tissue located at the tip of the probe and ink was injected in the guide cannula in order to facilitate the localization of the probe. Brains were then dissected out of the skull, frozen on dry ice and stored at −80°C until used for histological analyses. They were then cut on a cryostat in 50 μm-thick coronal sections that were stained with Toluidine blue by standard procedures (see (Aste et al., 1994)). The location of the tip of the probes was then determined under the microscope and plotted on schematic drawings of the quail brain prepared in our laboratory.

It is important to note that in contrast to rodents, female quail exhibit a moderate level of receptivity at any time during their ovarian cycle. Although female quail usually show higher receptivity in the early afternoon (Delville et al., 1986), an exact time-window for hormonal treatment in order to prime females for sexual receptivity has not be determined. Therefore we used this period of the day to perform the experiment to insure a maximal receptivity of the females. Also, we specifically selected stimulus females that showed a pattern of high receptivity (based on a few pre-test sessions) in order to avoid having to deal with an extra factor (i.e. differences of receptivity between females) in the analysis.

2.3.3. Glutamate assay

Glutamate analysis of the microdialysates was performed by a reversed phase narrow bore liquid chromatography assay with gradient elution and fluorescence detection after precolumn derivatization with o-phtalaldehyde and β-mercaptoethanol. Samples were first diluted 1:4 with aCSF and the product was then diluted 1:1 with distilled water before derivatization and injection. The chromatographic system consisted of a Gilson Sampling Injector 231 XL (Villiers-le-Bel, France) and the samples were eluted using a gradient program as described earlier (Van Hemelrijck et al., 2005). Mobile phases (mobile phase A consisted of a 0.025 M sodium phosphate solution pH 9.0 and 1% tetrahydrofuran; mobile phase B contained 90% methanol and 10% water) were delivered at a flow rate of 0.17 mL/min using two Shimadzu pumps (model LC-10AD) and a Shimadzu system controller SCL-10A VP (Shimadzu Benelux, ’s-Hertogenbosch, the Netherlands). A BioAnalyticalSystems vacuum degasser (model LC 26) was used (West Lafayette, IN). The temperature of the autosampler tray was set at 4°C and the injection volume was 10 μL on a C18 narrow bore column (5 μm particle size, 250 × 2 mm, Capcell Pak MG®, Shiseido). Detection was performed with a RF 10A XL fluorescence detector (Shimadzu) at excitation and emission wavelengths of 340 and 450 nm, respectively. Data acquisition was carried out with the integration computer program Clarity (DataApex, Antec, Zoeterwoude, the Netherlands). All analyses were performed by a scientist blinded to the nature of the samples.

2.4. Statistics

All data were analyzed with Statistica 10.0 and 13.0 (Statsoft Inc.) software. AA (in pmol _* h⁻¹ _* mg protein⁻¹) from the kainate-injected hemisphere was compared to AA from the control hemisphere (non-injected in Exp. 1A or VEH-injected in Exp. 1B) using paired comparison Student T-tests. Data were additionally plotted in histograms in which AA in the kainate-injected and control hemispheres were expressed as a percentage of the average AA in the control samples to facilitate comparisons between experiments.

In experiment 2, changes in glutamate concentration expressed in percentage of the average pre-experimental values were analyzed by non-parametric tests because normality conditions were not fulfilled. Friedman analyses of variance for matched samples were used to assess the changes of glutamate over time in a given experimental group. Wilcoxon paired tests were then used when appropriate to compare the different periods, two by two. Mann-Whitney U tests or Kruskal-Wallis analyses of variance were used to analyze the differences between independent groups. Kruskal-Wallis tests were followed when appropriate by Mann-Whitney U tests if more than 2 groups were compared. The p values were multiplied by the total number of comparisons to avoid type I error (Bonferroni correction). Data were considered significant for a corrected p<0.05. All values are expressed as means ± SEM.

3. Results

3.1. Experiment 1A : Unilateral kainate injections inhibit AA compared to the uninjected side

Kainate injection in the POM induced a significant decrease of AA in the injected side of the POM+BST compared to the non-injected hemisphere (contralateral side: 46.2±6.7 pmol _* h⁻¹ _* mg protein⁻¹, kainate side: 32.1±3.7; t₅ = 2.64, p = 0.046; 31% decrease see Fig. 1A) but not in the MBH (contralateral side: 16.8 ±2.7, kainate side: 16.7±2.1, t₅ = 0.60, p = 0.575). By contrast, no significant change in AA was detected in the control group between the vehicle-injected and the non-injected sides regardless of the region (POM+BST: Contralateral side: 31.9±4.0; Vehicle side: 28.5±2.5, t₅ = 1.80, p = 0.133; Fig. 1B; MBH: Contralateral side: 20.5 ±2.5, Vehicle side: 19.1±3.9; t₅ = 0.48, p = 0.650). The marked unilateral decrease in AA induced by kainate was clearly present in all subjects but one, while no or little change was observed after vehicle injection as shown in the bottom panels of Figure 1.

Fig. 1.

Acute effects of an injection of kainate (A, C) or its vehicle (B, D) in POM on aromatase activity (AA) in the male quail POM+BST. Panels A-B present average data (means± SEM) expressed as a percentage of the control hemisphere that was not injected while panel C-D show individual data (in pmol _* h⁻¹ _* mg prot⁻¹) illustrating the general nature of the responses. Kainate significantly inhibited AA but this effect was not present after the vehicle injection. * p < 0.05.

Experiment 1B : Unilateral kainate injections inhibit AA compared to the vehicle-injected side

In order to use each animal as its own control, we replicated the first experiment performing bilateral injections. Kainate was injected in the POM of one side, while the vehicle was infused in the POM of the other side. Inspection of brains during sectioning revealed that 6 subjects had been injected outside POM. Data from these subjects were thus analyzed separately.

Kainate injected in the POM induced again a significant decrease of AA in the ipsilateral POM+BST, compared to vehicle-injected contralateral side (control side: 22.9±1.8 pmol _* h⁻¹ _* mg protein⁻¹, kainate side: 16.7±2.0; t₈ = 7.50, p < 0.001; 27.2% decrease see Fig. 2A), while no difference between sides was detected in the MBH (contralateral side: 21.1±5.2 pmol _* h⁻¹ _* mg protein⁻¹, kainate side: 18.1±4.1; t₅ = 0.95, p = 0.387 ; 14.6% decrease). When the injection was outside POM, no effect was observed in POM+BST (contralateral side: 21.3±3.8 pmol * h⁻¹ * mg protein⁻¹, kainate side: 20.0±3.5; t₅ = 1.11, p = 0.318; 5.7% decrease see Fig. 2B) nor in the MBH (contralateral side: 15.3±4.1 pmol _* h⁻¹ _* mg protein⁻¹, kainate side: 13.7±2.2; t₃ = 0.71, p = 0.526; 10.5% decrease). Like in experiment 1A, the inspection of individual AA values reveals a decrease in enzymatic activity induced by kainate in all subjects in which the injection reached the POM, while no such dramatic effect was found in subjects where the pipette missed the POM (see Fig 2C–D).

Fig. 2.

Acute effect on aromatase activity (AA) in the POM+BST of male quail of an injection of kainate as compared to its vehicle injected in the contra-lateral hemisphere. Panels A-B present average data (means± SEM) expressed as a percentage of the control hemisphere that was injected with vehicle for birds that had an injection side inside (A) or outside (B) POM. Panel C-D show individual data (in pmol _* h⁻¹ _* mg prot⁻¹) illustrating the general nature of the responses. Kainate induced a significant reduction of AA compared to vehicle when injection was inside but not outside POM. *** p < 0.001.

3.2. Experiment 2: Glutamate changes in the POM during sexual interactions with a female

3.2.1. Effects of stressful events

During the microdialysis experiments, we noted the occurrence of marked unexpected events that could have potentially affected the experimental subjects such as the occurrence of a loud noise or the fall of a collection tube inside the test chamber which had to be collected by hand while the male was still under perfusion. Scanning through the glutamate assay results, we noticed the presence of pronounced and unexpected peaks of glutamate concentration and wondered whether these peaks related to such unexpected events. We therefore compared the glutamate concentrations from all samples collected while one of these unexpected events had taken place (a total of 6 events concerning 6 different subjects) to the concentration measured in the preceding sample. A Wilcoxon paired test revealed that glutamate had significantly increased during these events as compared to before (Before stress: 4.2±1.3 μM; During stress: 7.6±2.1; Z = 2.20, p = 0.028; 2.1 fold increase). These specific samples (6 out of 333 in total) were removed from the final analysis to avoid a potential contamination coming from these uncontrollable external stimuli.

3.2.2. Location of the microdialysis probes

Histological analysis indicated that the microdialysis probe was located within the POM for 11 subjects but was outside the boundaries of this nucleus for 5 birds (Fig. 3). In 4 other subjects the center of the probe was rostral to the cytoarchitectonic boundaries of the POM but its caudal edge seemed to enter the most rostral part of the POM. Qualitative examination of the data indicated that the pattern of glutamate changes was similar in these 4 subjects and in the 5 birds with a probe completely outside POM (no reaction to the female) and results of all these 9 birds were therefore pooled in the analyses as an OUT of POM group as opposed to the IN POM group.

Fig. 3.

Neuroanatomical localization of microdialysis probes in the preoptic area for experiment 2. Brain sections were stained with toluidine blue to illustrate the higher density of cells from the POM as well as the position of the center of the electrical lesions. The insert at the top right shows a photomicrograph illustrating the track of the probe and the electrical lesion on the right side as well as the boundaries of the POM as indicated by arrows on the left side. The position of the center of each probe has been plotted on a series of coronal sections through the quail brain that are arranged in a rostral to caudal order from top to bottom. Black circles represent males with a probe inside (or immediately adjacent in one case) the caudal part of the POM located at the level of the anterior commissure (CA) or one section (200 μm) more rostrally (CA-1). Blank circles represent males with a probe inside the rostral part of the POM (CA-2 and CA-3 i.e., 2 and 3 sections or 400 and 600 μm rostral to CA) or outside the nucleus (in the tractus septopallio-mesencephalicus or below the POM). Crossed circles represent males performing the full copulatory sequence including CCM while non-crossed circles represent males that did not reach the CCM. CO, Optical chiasma.

3.2.3. Effects of visual interaction with a female

Effects of visual exposure to a female was tested only during experiment 2B, which included 6 birds who had a probe in POM (IN) and 6 birds with a probe outside POM (OUT). The qualitative analysis of the samples collected while the males were seeing the female suggested that no change in glutamate concentration was observed during this period and this was confirmed by a quantitative analysis.

Average glutamate concentration (% of PRE) during the 9 min of visual exposure to the female was not different in the IN and OUT groups (IN: 107.0±9.4 % of PRE; OUT: 102.4±6.5; U = 12.0, p = 0.662) and these values were very close to the baseline (100%). Even if we considered separately the birds in the IN group that were subsequently going to copulate in the next phase of the experiment (2 out of 6) and compared them to the OUT group, there was still no difference between these 3 groups of subjects (H₂ = 2.85, p = 0.240) and their glutamate concentrations remained very close to the baseline (see Fig. S1B).

3.2.4. Effects of complete sexual interactions

To assess the effect of copulation on glutamate fluctuation, we decided to pool the data from experiments 2A and 2B for the analysis of the three other periods (PRE, INTERACTION and POST) in order to increase statistical power. This decision was based on (a) the absence of effect of the VIEW period in our data and (b) the fluctuation pattern of glutamate which was almost identical in the two experiments despite slight differences in their experimental design (addition of the VIEW condition). More specifically, in both experiments, animals with a probe located in the POM who displayed the full behavior (IN-CCM group) showed a non-significant increased concentration of glutamate in the dialysate during the interaction with the female as compared to the PRE period (Fig.S1A and B; Friedman test, p = 0.125 in Exp. 2A and p = 0.167 in Exp. 2B). However the analysis of the glutamate changes occurring immediately after a CCM (Fig.S1C and D) showed in both experiments a significant increase when compared to the PRE samples (Friedman, p = 0.042 in exp. 2A and p = 0.042 in 2B) and this concentration also tended to be higher than in the control group (i.e. subjects that displayed CCM but with a cannula outside of POM ; p = 0.057 in exp.2A ; p = 0.095 in exp.2B). The amplitude of the peak was very similar in both experiments (i.e. 150% during the interaction with the female and around 200% at the time of a CCM). Thus, even if the subjects of experiment 2B saw the female before copulating during the VIEW period, their glutamate response during copulation seemed identical to the response of subjects who did not see the female before the interaction. It was therefore decided to pool data of the two experiments.

The analysis of the pooled data included 4 experimental groups: birds with a cannula out of POM who did (OUT-CCM, n = 8) or did not copulate (OUT-NoCCM, n = 1) and, birds with a cannula within POM who did (IN-CCM, n = 6) or did not (IN-NoCCM, n = 5) copulate. One single bird in the OUT group did not copulate. The variation in extracellular glutamate concentration measured in this subject (OUT-no CCM) across time was plotted separately from the data obtained from the birds that copulated (OUT-CCM). Statistical analyses were performed both after including this animal in the OUT-CCM group (comparison of all OUT animals regardless of whether they copulated or not vs. IN-NoCCM vs. IN-CCM) or after excluding it from this group (comparison of OUT-CCM vs. IN-NoCCM vs. IN-CCM). Since the results of both analyses were the same, we only present here the analysis including all birds whose probe was outside of the POM regardless of their behavior (OUT group) even if the data of the OUT subjects who did not copulate are plotted separately in figure 4 so that the reader can access the full information.

Fig. 4.

Changes in glutamate concentrations observed by in vivo dialysis in 4 groups of males that had a probe out of POM and did not (white, OUT-NoCCM) or did copulate (Light gray, OUT-CCM) or had a probe in POM and did not (Dark gray, IN-NoCCM) or did copulate (Black, IN-CCM). (A) Glutamate changes occurring every 3 minutes during the baseline period (PRE), during the interaction with the female (INT) and during the recovery period (POST) in these four groups. (B) Glutamate changes averaged during the baseline period (PRE), during the interaction with the female (INTERACTION) and during the recovery period (POST). Sample size (number of subjects) is noted in the histograms of the PRE period. Since only one bird has a cannula outside POM and did not copulate (OUT-NoCCM), the data of this subject are displayed in the graph but for the statistical analyses this bird was included as part of an overall “OUT group”. (B1) Average glutamate changes immediately following copulation in males who copulated (OUT-CCM and IN-CCM). (B2) Individual data of males who copulated with a cannula in POM (IN-CCM) or out of it (OUT-CCM) during the PRE period, immediately following CCMs and during the POST period. $ p < 0.05 compared to IN-NoCCM ; + p < 0.05 compared to OUT-CCM and OUT-NoCCM ; ++ p < 0.01 compared to OUT-CCM and OUT-NoCCM ; * p < 0.05, ** p < 0.05.

Average changes in glutamate concentrations (% of PRE) observed in all 3 minute-samples in these 3 groups of subjects are presented in Fig. 4A.

A clear increase in glutamate concentration was observed during the INTERACTION period in the IN-CCM group but not in the others with the exception of a moderate and transient average increase in the IN-noCCM birds (Fig. 4A). Kruskal Wallis analyses performed for each time point separately indicated that significant group differences were present during the fourth sample of the PRE period (H₂ = 6.88 ; p = 0.032), during the first and third sample of the INTERACTION period (H₂ = 6.31 ; p = 0.042 ; H₂ = 12.18 ; p = 0.002 respectively) and during the first sample of the POST period (H₂ = 10.90 ; p = 0.004). All other time points showed no difference between groups (H₂ ≤ 4.83 ; p ≥ 0.089). At these 4 specific time points, pair-wise comparisons performed by Mann-Whitney U tests with a Bonferroni correction mainly showed that glutamate concentration in the IN-CCM group was higher than in the other groups only at the specific time points of the INTERACTION and POST periods but not in the PRE period.

In order to buffer short-term fluctuations introducing noise in the data, glutamate concentrations (% of PRE) were then averaged for all 3 minutes-samples in each time period (Fig. 4B). Comparisons between groups within each experimental period (Kruskal-Wallis analyses) revealed significant differences between sub-groups during the INTERACTION period (H₂ = 8.5, p = 0.014) while no difference was observed during the PRE and POST periods (H₂ = 0.8, p = 0.683 ; H₂ = 5.8, p = 0.054 respectively). Mann-Whitney U post-hoc tests used to compare all groups two- by-two demonstrated that glutamate concentrations were higher in the IN-CCM than in OUT group during the INTERACTION period (see Figure 4B).

We also noted a sharp peak in glutamate concentration immediately after males performed a complete copulatory sequence including CCM. To characterize this observation in a quantitative manner, we calculated the average concentrations (% of PRE) of the samples that followed each CCM for the IN and OUT of POM groups (see data in the insert of Fig. 4B1). In these samples, there was indeed an even stronger increase in glutamate (203% of PRE) than when the comparisons included the entire interaction period (159% of PRE). This increase following CCM was observed in the IN but not in the OUT group and these two groups were statistically different (UN=9,N=6 = 0, p = 0.005). Note also that this increase in glutamate concentration after copulation was present in a systematic fashion in all males of the IN group but not in the subjects of the OUT group (see Insert B2).

4. Discussion

This study was primarily designed to determine whether the rapid inhibition of aromatase activity observed in the POM during/after expression of male sexual behavior is mediated by a local glutamate release. The two experiments demonstrated that (1) an acute injection of the glutamatergic agonist kainate in the POM of males induces a significant decrease of about 30 % of aromatase activity in this region and (2) glutamate is released in the POM of males specifically in association with the occurrence of copulation. We will discuss these brain events separately and, will subsequently propose an integrated view of the neuroendocrine and neurochemical changes observed during sexual interactions and their potential role in the control of the different aspects of sexual behavior.

4.1. Rapid inhibition of aromatase activity by glutamate

We show here that AA is decreased in the intact POM/BST region of live animals within 20 minutes after an injection of kainate in the POM. These results are consistent with previous data indicating that kainate inhibits AA in quail preoptic explants maintained in vitro (Balthazart et al., 2006) and that glutamate retrodialysis decreases local estradiol concentration in vivo (Remage-Healey et al., 2008). These data thus extend to a more physiological situation (i.e. the POM integrated in its complex circuitry in a live animal) the conclusions of previous in vitro studies and also provide a finer spatial resolution showing that the inhibition of AA specifically involves regulation in the POM as opposed to other aromatase-positive areas in the preoptic-hypothalamic blocks that had been analyzed in vitro. The mechanism by which the activation of glutamate receptors by kainate leads to this decreased AA presumably involves calcium-dependent phosphorylations of the enzyme that have been shown to also induce a transient and rapid inhibition of AA in vitro (Balthazart et al., 2001; Balthazart et al., 2003; Charlier et al., 2011a).

4.2. Glutamate release in POM during copulation

The present data reveal that the extracellular concentration of glutamate increases in the POM of males during copulation. This effect is specific to the POM, as males with a probe outside of this nucleus did not show similar changes. This increase is also specific to copulation, as (1) the highest glutamate level is observed immediately after CCM (IN-CCM group), (2) males who did not copulate (IN-NoCCM group) exhibited only small non-significant changes in glutamate (even though they did not differ statistically from IN-CCM males) and (3) no obvious changes are detected when males only have visual access to the female without being able to interact with her. There was a non-significant 20% increase in glutamate in the POM during the VIEW period in the two birds of the IN group who actually copulated during the interaction period suggesting that more work should be done to confirm or infirm this negative result. Similarly, we observed a slight but non-significant increase of glutamate occurring in males from the IN-NoCCM group when they began to interact with the female. It is possible that glutamate release increases in response to other components of the sexual behavior sequence besides CCM (such as the attempts to mount the female) but if this is the case, this rise has a smaller amplitude and is not significant in our study. Further investigation would be needed to confirm this observation.

In rodents, a release of glutamate in the preoptic area during performance of male sexual behavior has previously been demonstrated (Dominguez et al., 2006; Dominguez and Hull, 2010). In these studies, the highest peak of glutamate occurred during ejaculation, which is consistent with the present data measuring the highest concentration of glutamate in the 3 minutes period immediately following CCM. In rodents, this increase in glutamate in the preoptic area occurring during or after copulation has been shown to mediate the female-dependent dopamine release in the preoptic area that facilitates male sexual behavior (Hull and Dominguez, 2006).

4.3. Glutamate control of AA vs. sexual behavior

This facilitatory effect of glutamate on sexual behavior and its inhibitory effect on AA might at first sight appear contradictory since aromatase produces estrogens that acutely stimulate behavior expression. However, it is unlikely that both effects are mediated by the same mechanism and occur within the same time window. Indeed, facilitative effects of glutamate on copulation based on studies in rats appear to be mediated mainly through the activation of NMDA receptors (Dominguez et al., 2007; Vigdorchik et al., 2012). In contrast, in vitro NMDA only has marginal effects on AA in quail preoptic explants while the glutamatergic agonists, kainate and AMPA, induce a rapid, transient inhibition of the enzymatic activity (Balthazart et al., 2001, 2006). Estradiol concentrations measured by in vivo dialysis in the zebra finch caudomedial nidopallium are decreased by glutamate but not by NMDA retrodialysis (Remage-Healey et al., 2008).

In addition, effects of glutamate on behavior are likely to take place quite rapidly, consistent with the rapid increase in its concentration observed concomitantly with behavioral expression (Fig. 4). In contrast, kainate effects on AA were observed here within 20 minutes (Fig. 1) and only reached their maximum after 15 min during in vitro experiments (Balthazart et al., 2006).

It is therefore plausible that the glutamate release observed during and after copulation can lead to both (1) a rapid NMDA receptor-dependent improvement of sexual behavior and (2) a slower AMPA or kainate receptor-dependent decrease in AA. The higher affinity of glutamate for NMDA receptors as compared to AMPA and kainate receptors (Hollmann and Heinemann, 1994) could contribute to the differential timing of these two effects with the facilitatory effect on sexual behavior occurring when glutamate concentrations begin to increase and the inhibition of AA taking place when high concentrations of glutamate have accumulated.

4.4. Integration of the neuroendocrine and neurochemical control of sexual motivation vs. performance

Together with previous data, the results of the present experiments provide a complex picture of the neurochemical and neuroendocrine events that take place in the preoptic area and are implicated in the control of the expression of male sexual behavior. More specifically, the following four main facts have been well established. First, acute treatments with estrogens or aromatase inhibitors respectively increase or decrease appetitive male sexual behavior without changing sexual performance which rather depends on long-term genomic effects of both testosterone and estrogens (Seredynski et al., 2013). Second, sexual interaction with a female but also simply seeing her are both sufficient to induce a rapid decrease in AA measured ex vivo in the male POM/BST (Balthazart et al., 2006; Cornil et al., 2005; de Bournonville et al., 2013). Third, the glutamatergic agonist kainate inhibits AA within minutes both in vitro (Balthazart et al., 2001) and in vivo (Experiment 1). Fourth, extracellular concentrations of glutamate increase in the preoptic area during copulation (Experiment 2).

These data are fully consistent with the hypothesis that high baseline concentrations of estrogens would chronically activate sexual motivation leading to sexual interactions that would release glutamate in the preoptic area causing an inhibition of AA, a decrease in estrogen concentrations and cessation of behavior due to the subsequent decreased sexual motivation (Seredynski et al., 2013).

If glutamate release can explain the induced decrease in AA occurring after copulation, it does not explain changes in AA induced only by the view of the female (de Bournonville et al., 2013) since glutamate seems to be released specifically after cloacal contact movements but does not appear to be elevated during the view of the female. This result could suggest that if sexual motivation and sexual performance both inhibit AA in the POM, the induced enzymatic decrease is regulated by independent mechanisms. However, several considerations potentially explain the absence of glutamate rise during the visual exposure to the female.

First, time constraints for the in vivo dialysis related to assay sensitivities could obscure significant changes. Visual interaction with the female was monitored during 3 × 3 minutes and this duration might not be sufficient to detect changes in glutamate release or the sampling interval might be too long to catch very transient changes (glutamate release could vary multiple times within seconds and these changes would not be reflected in the average obtained for a 3 minutes period).

Second, the spatial resolution of the dialysis probes might miss some local neurochemical events that take place in limited parts of the POM. The present data already show that glutamate increases after copulation in the caudal POM while no changes occurred in the rostral part of the nucleus or if the probe was outside POM. Differential behavioral effects of electrolytic lesions in rostral and caudal parts of the POM have also been reported (Balthazart et al., 1998) and these sub-regions are differentially activated, as measured by induction of the c-Fos protein, after performance of different components of sexual behavior (Taziaux et al., 2006). These studies suggest that the caudal POM is responsible for the control of male sexual performance while the rostral part as well as/or another brain nucleus such as the BST or the medial amygdala could be more specifically involved in the control of male sexual motivation (Balthazart et al., 1998).

Third, microdialysis sampling only in the POM obviously ignores behaviorally relevant events taking place elsewhere in the brain. Pharmacological studies have not yet determined the brain areas where rapid actions of estrogens modulate male quail sexual motivation; they only tested effects of peripheral (Cornil et al., 2006a; Cornil et al., 2006b) or intracerebroventricular injections (Seredynski et al., 2015; Seredynski et al., 2013). In rodents, the BNST appears to contribute more to the preparation for mating than to copulation per se (Hull and Dominguez, 2006). In quail rapid changes of AA after visual exposure to a female were observed in the POM+BST region. These two nuclei are in close proximity and form a continuous aromatase cell population (Foidart et al., 1995) but differential modulations of these two regions have already been reported. For example, appetitive male sexual behavior induces neuronal activation of aromatase cells in the BST of chicken while consummatory sexual behavior does not (Xie et al., 2011). The BST is thus a potential candidate for the short-term control of male sexual motivation. Future studies should thus investigate the profile of glutamate release in the BST in behaving males. The medial amygdala could also be involved in these effects. Indeed, lesions of this area in male rats abolish the dopamine increase in response to an estrous female (Dominguez et al., 2001) and this dopamine peak is under the control of glutamate (Dominguez et al., 2004). These data thus suggest that the glutamatergic stimulation leading to dopamine release in the preoptic area comes from neurons located in the amygdala (Hull and Dominguez, 2006). The present lack of glutamate rise in POM when the male is seeing the female whereas AA is nevertheless inhibited in this region could thus be explained by the anatomical specificity in glutamate action.

Finally, it is important to keep in mind that two different techniques have been used in these experiments. The glutamate changes in POM were measured in vivo during the performance of sexual behavior (Experiment 2) whereas measures of AA after a visual or sexual interaction with the female (Cornil et al., 2005; de Bournonville et al., 2013) and after the exposure to glutamatergic drugs (Balthazart et al., 2001, 2006) (Experiment 1) were based on brain samples collected post-mortem. Besides the fact that measures of glutamate and AA are not taken at the same time, the ex vivo measures of AA are also potentially affected by multiple factors resulting in measures that do not exactly reflect the level of enzymatic activity that was present in vivo during behavioral expression. These factors include a) the methods used for brain collection-dissection that could massively release transmitters (e.g., glutamate, dopamine) that could in turn affect AA, b) the homogenization of the samples that potentially brings diverse intracellular components in contact with the enzymatic protein and could affect its activity, c) the optimization of the in vitro assay conditions (pH, temperature, ionic conditions, optimal concentration of co-factors) that maximize the enzymatic activity but possibly result in activities that do not reflect the in vivo situation. For these reasons, the in vivo measurement of AA under the same conditions as was used for glutamate data collection would be optimal but such a technique is not available at present to our knowledge. However, changes in local estradiol concentrations have been measured by in vivo microdialysis coupled with enzymatic immunoassay in the zebra finch telencephalon during exposure to male song (Remage-Healey et al., 2008). Measuring changes of estradiol in POM via this technique would clearly be useful to allow for an enhanced understanding of how glutamate can control rapid fluctuations of estrogens in the brain to orchestrate sexual behavior in a dynamic manner.

5. Conclusion

In conclusion, the present studies suggest the following most likely sequence of brain events related to sexual behavior: the naturally high basal estrogen concentration in the brain of male quail (a species selected during centuries for active reproduction and thus high sexual activity) would chronically activate sexual motivation and bring the partners together resulting in a release of glutamate. Glutamate then probably plays both a facilitatory role on behavior and an inhibitory effect on AA. As it does in rats, glutamate could act directly as a transmitter (via NMDA receptors; (Will et al., 2014)) and/or enhance the dopamine release in the preoptic region that is known to occur in quail before and during copulation (Kleitz-Nelson et al., 2010) to facilitate sexual behavior. Then in a second step, glutamate would inhibit AA and the subsequent decreased production of estradiol would cause the cessation of behavior by shutting down sexual motivation (Seredynski et al., 2013). However, the previously reported rapid inhibition of AA in the POM/BST after simply viewing a female (de Bournonville et al., 2013) is either not mediated by glutamate or mediated by glutamate but in another brain site.

Highlights.

• Performance of sexual behavior inhibits preoptic aromatase activity

• Glutamate is released in the medial preoptic nucleus (POM) of copulating males

• Kainic acid infusion in POM decreases aromatase activity within 20 min

• Glutamate release presumably activates NMDA receptors to facilitate sexual behavior

• Behaviorally –induced glutamate release presumably inhibits aromatase activity