GnRH suppresses excitability of visual processing neurons in the optic tectum

Abstract

Animals change their behavior in response to sensory cues in the environment as well as their physiological status. For example, it is generally accepted that their sexual behavior is modulated according to seasonal environmental changes or the individual's maturational/reproductive status, and neuropeptides have been suggested to play important roles in this process. Some behavioral modulation arises from neuropeptide modulation of sensory information processing in the central nervous system, but the neural mechanisms still remain unknown. Here we focused on the neural basis of neuropeptide modulation of visual processing in vertebrates. The terminal nerve neurons that contain gonadotropin-releasing hormone 3 (TN-GnRH3 neurons) are suggested to modulate reproductive behavior and have massive projections to the optic tectum (OT), which plays an important role in visual processing. In the present study, to examine whether GnRH3 modulates retino-tectal neurotransmission in the OT, we analyzed the effect of GnRH3 electrophysiologically and morphologically. We found that field potentials evoked by optic tract fiber stimulation, which represent retino-tectal neurotransmission, were modulated postsynaptically by GnRH3. Whole cell recording from postsynaptic neurons in the retino-tectal pathway suggested that GnRH3 activates large-conductance Ca²⁺-activated K⁺ (BK) channels and thereby suppresses membrane excitability. Furthermore, our improved morphological analysis using fluorescently labeled GnRH peptides showed that GnRH receptors are localized mainly around the cell bodies of postsynaptic neurons. Our results indicate that TN-GnRH3 neurons modulate retino-tectal neurotransmission by suppressing the excitability of projection neurons in the OT, which underlies the neuromodulation of behaviorally relevant visual information processing by the neuropeptide GnRH3.

animals modify their behavior in response to their physiological status as well as sensory cues in the surrounding environment. For example, for successful reproduction, during the breeding season sensory information processing in sexually mature animals changes in order to facilitate detection, attraction, or acceptance of their partners and to enable performance of sexual behaviors. Such sexual behavior in many animals is modulated according to seasonal environmental changes or their maturational/reproductive status, and neuropeptides have been suggested to play important roles in such modulation (Argiolas and Melis 2013; Dornan and Malsbury 1989). Fish provide a good model for investigating the mechanisms of neuropeptide action on behaviorally relevant sensory processing systems. For example, male goldfish use not only chemical but also visual cues when chasing females during the breeding season (Thompson et al. 2004), and sailfin mollies show seasonal variation in mate choice preferences based on visual cues (Heubel and Schlupp 2008). Such changes in behaviors are suggested to arise from neuropeptide modulation of sensory information processing in the central nervous system, but the neural mechanisms still remain unknown. Here we focused on the neural basis of peptidergic neuromodulation of visual processing in the optic tectum. In nonmammalian vertebrates including teleosts, the nongeniculate system through the optic tectum plays essential roles in visual information processing (Karten and Shimizu 1989; Kinoshita and Ito 2006; Springer et al. 1977; Yager et al. 1977).

The terminal nerve neurons that contain gonadotropin-releasing hormone 3 (TN-GnRH3 neurons) are suggested to receive information concerning reproductive status (Onuma et al. 2005; Zempo et al. 2013) and modulate reproductive behaviors (Yamamoto et al. 1997). These neurons also have massive projections to the optic tectum in teleosts (Amano et al. 2002; Oka and Matsushima 1993; von Bartheld and Meyer 1986). Therefore, we hypothesized that TN-GnRH3 neurons play an important role in modulating visual information processing according to reproductive status. TN-GnRH3 neurons release the peptide GnRH3, and GnRH3 has been shown to modulate activity in the olfactory bulb (Kawai et al. 2010), in olfactory receptor neurons (Eisthen et al. 2000), and in retinal circuits (Umino and Dowling 1991). Here we examined whether GnRH3 modulates retino-tectal synaptic transmission by analyzing field potentials and membrane excitability of visual information processing neurons in the dwarf gourami, a teleost fish that has frequently been used for the study of neuromodulatory GnRH neurons (Abe and Oka 2002; Ishizaki et al. 2004; Wirsig-Wiechmann and Oka 2002).

Our field potential analysis revealed that GnRH3 modulates retino-tectal neurotransmission postsynaptically. Furthermore, we showed that GnRH3 suppresses excitability of the stratum periventricular (SPV) neurons, which have been reported as postsynaptic neurons for retino-tectal neurotransmission (Kinoshita et al. 2005; Vanegas et al. 1971, 1974a), by activating large-conductance Ca²⁺-activated K⁺ (BK) channels. Finally, we examined the localization of GnRH receptors in the optic tectum morphologically with an improved method that employs GnRH peptides labeled with fluorescence (GnRH-fluoroprobe). In brain slices treated with GnRH-fluoroprobe, the fluorescent signals were observed mainly near the cell bodies of the SPV neurons. The results of the present electrophysiological and morphological analyses suggest that GnRH3 modulates retino-tectal neurotransmission by suppressing the excitability of postsynaptic neurons in the optic tectum.

MATERIALS AND METHODS

Animals.

Adult male dwarf gouramis (Trichogaster lalius), ∼5 cm in standard length, were purchased from a local dealer. Each aquarium containing ∼20 fish was maintained at 27°C on a 14:10-h light-dark cycle. All fish were fed worms daily. All procedures were reviewed and approved by the appropriate The University of Tokyo animal experimentation committee.

Slice preparation.

Fish were anesthetized by chilling in ice. We quickly killed them by decapitation and isolated the whole brain. The brain was put in artificial cerebrospinal fluid (ACSF) consisting of (in mM) 134 NaCl, 2.9 KCl, 1.2 MgCl₂, 2.1 CaCl₂, 10 HEPES, and 10 glucose (pH 7.4 adjusted with NaOH). We embedded the brain in a 3.6% low-melting-temperature agar (agarose type IX-A; Sigma-Aldrich, St. Louis, MO) solution prepared with ACSF. Four hundred-micrometer sagittal sections were cut in ACSF containing choline chloride instead of NaCl with a Vibratome 3000 (Leica Biosystems, Wetzlar, Germany). The brain slices were incubated in ACSF on ice for 1 h and brought back to room temperature before the experiments.

Electrophysiology.

For field potential recordings, a slice was placed in a chamber filled with ACSF containing 0.1% BSA to prevent nonspecific binding of peptides to the plastic and glassware. This ACSF was perfused gravitationally at 1–2 ml/min. Recording electrodes were made of borosilicate glass (GD-1.5; Narishige, Tokyo, Japan) with a Flaming-Brown micropipette puller (P-97; Sutter Instrument, Novato, CA). The borosilicate electrode filled with 0.1% agar in 2 M NaCl was placed 100 μm under the surface of the brain slice for recording in the superficial tectal layers [stratum opticum (SO) and stratum fibrosum et griseum superficiale (SFGS)], where the retino-tectal synapses are formed on the distal dendrites of SPV neurons and the largest synaptic responses were recorded (Vanegas 1983; Vanegas et al. 1971), ∼200 μm away from the stimulating electrode. The field potential was amplified by AVH-11 (2,000×; Nihon Kohden, Tokyo, Japan) and recorded with a Digidata 1322A and pCLAMP 8.2 software (Molecular Devices, Sunnyvale, CA). Electrical stimuli were applied with an electronic stimulator (SEN-3301; Nihon Kohden) through an isolation unit (SS-201J; Nihon Kohden). The bipolar stainless steel electrode consisted of a pair of electrolytically polished stainless steel insect pins, ∼10 μm apart and 1 μm in tip diameter, that were lacquer coated, leaving 20- to 30-μm uncoated bare tips. For stimulation of the optic tract fibers, we placed the bipolar stainless steel stimulating electrode in the superficial tectal layers, the main layers of optic tract fiber projections (SO and SFGS). Electrical pulse stimuli of 0.1-ms duration at 0.5–1 mA were applied every 30 s. Synthetic GnRH3 peptides were purchased from GL Biochem (Shanghai, China). Five minutes after the start of recording, GnRH3 was applied for 4 min. We used the following drugs depending on experiments: 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; Abcam, Cambridge, UK), Antide (Sigma-Aldrich), and iberiotoxin (IBX; Abcam).

For whole cell recording, the internal solution consisted of (in mM) 100 K-gluconate, 30 KCl, 0.05 EGTA, 4 MgCl₂, 10 HEPES, 4 K₂ATP, and 0.5 Na₂GTP. The junction potential was −12 mV, and membrane potentials were adjusted using this value. The tip resistance of patch electrodes in ACSF was 10–15 MΩ. Recordings were performed with an Axopatch 200B patch-clamp amplifier (Molecular Devices). Whole cell current- and voltage-clamp recordings were digitized at 10 kHz and stored on a computer with a Digidata 1322A and pCLAMP 9.2 software (Molecular Devices). For analysis of membrane excitability, we used the minimum response that elicited at least one action potential during current injections of three different intensities (0.01, 0.015, or 0.02 nA).

Luciferase assay.

HEK293T cells (70–80% confluent) were cultured in 100-mm dishes. The culture medium contained 5% FBS (Life Technologies, Carlsbad, CA), 4 mM l-glutamine (Wako Pure Chemical Industries, Osaka, Japan), and 4.5 g of Dulbecco's modified Eagle's medium (Nissui Pharmaceutical, Tokyo, Japan) in 500 ml of water. Thirty microliters of polyethylenimine (PEI, Polyethylenimine “Max”; Polysciences, Warrington, PA) and 1.5 ml Opti-MEM (Life Technologies) were mixed and incubated for 5 min. We then added the following constructs into another 1.5 ml of Opti-MEM: 5.5 μg of each GnRH receptor (gift from Dr. K. Okubo; Okubo et al. 2001, 2003), 9 μg of firefly luciferase vector (pGL4.33[luc2P/SRE/Hygro] vector; Promega, Madison, WI), and 0.5 μg of Renilla luciferase vector (pGL4.74[hRluc/TK] vector; Promega). After incubation in PEI solution, both Opti-MEM solutions were mixed and incubated for 20 min at room temperature. We then replaced the culture medium with mixed transfection solution and incubated the dish at 37°C overnight. The transfected HEK cells were then subcultured into 24-well plates in a culture medium containing 1% FBS. Eighteen hours later, 1 μM GnRH2 or 1 μM fluorescence probe in the culture medium containing 1% FBS (vehicle) or the vehicle only was applied for 6 h. GnRH-fluoroprobes [pGlu-His-Trp-Ser-His-d-Lys(N^ε-FAM)-Trp-Tyr-Pro-Gly-NH₂ (GnRH-FL) and pGlu-His-Trp-Ser-His-d-Lys(N^ε-TAMRA)-Trp-Tyr-Pro-Gly-NH₂ (GnRH-TAMRA)] were designed based on a previous report (Hazum et al. 1980). Probes were synthesized with standard Fmoc-based solid-phase peptide synthesis (Oishi et al. 2008). Cells were collected and the luminescence measured by LUMAT LB9507 (Berthold Technologies, Bad Wildbad, Germany) according to the manual for the Dual-Luciferase Reporter Assay System (Promega).

Analysis of localization of GnRH receptors in HEK cells.

For morphological observation, HEK cells were cultured on poly-l-lysine-coated glass (Matsunami Glass Industries) in 60-mm dishes. HEK293 cells were transfected with GnRH receptor (1 μg/60-mm dish) and EGFP-N1 (0.4 μg/dish; Takara) with PEI (2.88 μl/dish) as described above. After two overnight incubations, each glass covered with cells was put in a 24-well plate filled with serum-free medium. Thirty minutes later, HEK cells were incubated with 1 μM GnRH-fluoroprobe or plain serum-free medium for 30 min at 37°C. We then washed cells with PBS twice and examined them under a LSM-710 confocal laser-scanning microscope (Carl Zeiss, Oberkochen, Germany) at 1-μm optical section.

Analysis of localization of GnRH receptors with fluoroprobes.

Two hundred-micrometer-thick frontal slices were used for the analysis. Brain sections were cut as described above and incubated at room temperature for 1 h after slicing. To block nonspecific binding, slices were treated with 0.1% BSA-ACSF for 30 min and then the fluorescence probe solution or vehicle control was applied for 1 h. The GnRH-fluoroprobe (GnRH-FL) was designed based on a previous report (Hazum et al. 1980). We chose GnRH2 as the backbone of GnRH-fluoroprobe because GnRH2 shows the highest affinity for each GnRH receptor in fish for GnRH1-3 (Illing et al. 1999; Okubo et al. 2001, 2003). The probe was synthesized with standard Fmoc-based solid-phase peptide synthesis (Oishi et al. 2008). After application of the GnRH-fluoroprobe, the brain slices were fixed with 4% PFA-PBS at 4°C overnight. The slices were then incubated for 6–8 h with a mouse antiserum against fluorescein, diluted 1:1,000 in the blocking solution. For analysis of localization of GnRH receptors with fluoroprobes, the signals were intensified by sequentially incubating the slices in biotinylated anti-mouse IgG (Vector Laboratories, Burlingame, CA) for 6–8 h, the ABC kit (Vector Laboratories) in PBST for 30–45 min, and streptavidin-Alexa 555 for 4 h. Nuclei were counterstained with Hoechst 33342 (Life Technologies). The slices were then mounted on slides and coverslipped with CC/Mount (Diagnostic BioSystems, Pleasanton, CA). The slides were examined and photographed with a LSM-710 confocal laser-scanning microscope (Carl Zeiss), and the resulting images were analyzed with ImageJ software with the MBF ImageJ plug-ins (Tony Collins, McMaster University). All images were captured in the same exposure condition, and we used the same brightness and contrast values for ImageJ in all three groups.

Statistics.

Statistical analyses were performed with Kyplot5 software (Kyence, Tokyo, Japan) and IGOR Pro 6 software (WaveMetrics, Lake Oswego, OR) with the Taro tool (an IGOR macro set written by Dr. Taro Ishikawa, Jikei University School of Medicine). Different statistical tests were used for different experiments, as described in results. All data in the present study are presented as means ± SE.

RESULTS

GnRH3 modulates retino-tectal neurotransmission.

In nonmammalian vertebrates, retino-tectal neurotransmission is important for visual processing (Karten and Shimizu 1989; Kinoshita and Ito 2006). It has been proposed that synaptic transmission from optic tract fibers to the deep tectal (SPV) neurons (retino-tectal neurotransmission) is essential for visual processing in the optic tectum (Kinoshita and Ito 2006; Vanegas 1974; Vanegas et al. 1974b). We examined whether GnRH3 modulates retino-tectal neurotransmission by stimulating optic tract fibers and recording the evoked local field potential (LFP) from the synaptic region of SPV neurons (Vanegas et al. 1974b). Because the tectal neurons form distinctive layers and show characteristic LFPs in each layer, field potential recordings have often been used in analyses of retino-tectal neurotransmission (Vanegas 1983; Vanegas et al. 1971). The field potential evoked by electrical stimulation is shown in Fig. 1, A and B. The first large positive peak indicates an artifact produced by stimulation and almost masks the presynaptic responses. The subsequent large negative wave most probably represents the postsynaptic response of the SPV neurons in the optic tectum (Kinoshita and Ito 2006; Vanegas et al. 1974a). This synaptic transmission is likely monosynaptic (Kinoshita and Ito 2006; Vanegas et al. 1974b). Because the negative waves were almost completely blocked by 20 μM CNQX (Fig. 1A), they probably reflect glutamatergic neurotransmission. Thus the negative waves are at least suggested to contain monosynaptic glutamatergic current components of SPV neurons in response to optic tract fiber stimulation. It should be noted, however, that the negative waves have also been suggested to contain regenerative population spike components (see Vanegas 1983 as well as discussion). The evoked LFPs were then recorded and compared with those in the presence of GnRH3. Figure 1B shows representative traces before, during, and after GnRH3 application. GnRH3 application increased the rise time from 10% to 90% of peak amplitude (Fig. 1C) and the negative wave duration measured at half-maximum amplitude (Fig. 1D). These effects were diminished by Antide, which is a GnRH receptor antagonist (Fig. 1C: vehicle: 91.8 ± 3.9%, GnRH3: 134.0 ± 15.9%, GnRH3 + Antide: 93.9 ± 4.2%; Steel's test, P < 0.05, Fig. 1D: vehicle: 95.9 ± 1.4%, GnRH3: 115.8 ± 4.1%, GnRH3 + Antide: 98.1 ± 3.1%; Steel's test, P < 0.01). On the other hand, the peak amplitudes of the field potentials were not significantly affected by GnRH3 application [Fig. 1E; vehicle: 98.8 ± 1.0%, GnRH3: 92.6 ± 6.0%, GnRH3 + Antide: 102.2 ± 7.4%; Steel's test, nonsignificant (n.s.)]. Overall, our results indicated that GnRH3 modulated the shape of LFPs evoked by stimulation of optic fibers, which is suggested to be mediated by the GnRH receptors (see below).

Fig. 1.

Gonadotropin-releasing hormone 3 (GnRH3) modulates the field potential of retino-tectal neurotransmission. A: field potentials evoked by electrical stimulation of optic tract fibers (average of 3 traces) before and during 20 μM CNQX application. The first large positive peak represents an artifact, and the subsequent large negative peak indicates the postsynaptic response of stratum periventricular (SPV) neurons. B: field potentials evoked by electrical stimulation of optic tract (average of 3 traces) before, during, and after (washout) GnRH3 application. C–E: rise time (C), duration measured at half-maximum amplitude (D), and peak amplitude (E) in vehicle, GnRH3 (100 nM), or GnRH3 (100 nM) + Antide (1 μM). All values were normalized by dividing the average of 3 trials during GnRH3 application (2 min after application) by the average of 3 trials measured 2.5 min after the start of recording. In the box plots, bottom and top bars are minimum and maximum of each group, respectively, box is quartile area, and middle horizontal bar shows the median. Data were analyzed with Steel's multiple-comparison test (*P < 0.05, **P < 0.01).

GnRH3 modulates LFP postsynaptically.

We used the paired-pulse ratio (PPR) to determine whether GnRH3-induced modulation is a presynaptic or postsynaptic effect. Comparing the PPR (ratio of amplitude of the responses to the first and second stimulations) before and after treatment allows one to determine whether the effect is pre- or postsynaptic (Manabe et al. 1993); if the PPR changes after treatment the modulation is suggested to be presynaptic, and vice versa. Figure 2A shows the LFP in response to paired stimulation of retino-tectal fibers at an interval of 150 ms (Kawai et al. 2010). Figure 2B shows the time course of the PPR. Given that the PPRs appear not to change before and after GnRH application, the results suggest that GnRH3-induced modulation is mainly postsynaptic.

Fig. 2.

GnRH3 modulates field potentials postsynaptically. A: field potentials in response to paired stimulation of retino-tectal fibers (average of 3 traces) before and during GnRH3. B: time course of the paired-pulse ratio (n = 6). In the box plot, bottom and top bars are minimum and maximum of each group, respectively, box is quartile area, and middle horizontal bar shows the median.

GnRH3 suppresses excitability of visual processing neurons in optic tectum.

In the optic tectum, SPV neurons receive direct synaptic inputs from optic tract fibers and project to the other brain regions. SPV neurons are thought to play a principal role in visual information processing in the teleost optic tectum (Kinoshita et al. 2005; Kinoshita and Ito 2006; Vanegas 1974; Vanegas et al. 1974b). The data presented in Figs. 1 and 2 suggest that GnRH3 modulates these postsynaptic SPV neurons. For further analysis of the postsynaptic mechanism induced by GnRH3, we examined effects of GnRH3 on the excitability of SPV neurons with whole cell patch-clamp recording. We generated action potentials in SPV neurons by applying a series of current steps (50 ms, 0.01–0.02 nA) through a recording whole cell pipette. Figure 3, A–C, show representative action potentials of a SPV neuron before (Fig. 3A), during (Fig. 3B), and after (Fig. 3C) application of GnRH3. As illustrated in Fig. 3B, the number of action potentials decreased during application of GnRH3. The normalized response for each condition is plotted in Fig. 3D; the y-axis indicates the response normalized by the spike number evoked by the minimum current step before GnRH3 application. GnRH3 markedly suppressed the ability of spike generation in the SPV neurons (before: 100%, GnRH3: 45.3 ± 17.3%, washout: 107.1 ± 7.1%; ANOVA and Tukey-Kramer test, n = 7 in each group, P < 0.01). The result clearly demonstrates that GnRH3 suppressed the excitability of SPV neurons in the optic tectum.

Fig. 3.

GnRH3 suppress the excitability of SPV neurons. A–C: action potential responses of SPV neurons evoked by current injection (bottom inset) before (A), during (B), and after (C) application of 1 μM GnRH3. D: normalized response before, during, and after application of 1 μM GnRH3. y-Axis indicates the response normalized by the minimum spike number before GnRH3 application. n = 7 for each group. Data were analyzed with ANOVA and Tukey-Kramer test (**P < 0.01). Error bars indicate SE. Before, 1 min before GnRH3 application; during, 2 min after GnRH3 application; after, 5 min after finishing GnRH3 application.

GnRH3 activates large-conductance Ca²⁺-activated K⁺ channel.

To analyze the ionic mechanisms underlying GnRH3-induced modulation, we compared the current-voltage relationships of SPV neurons with or without GnRH3 application. We applied ramp stimulation (Fig. 4A, inset) and analyzed the current-voltage relationship before, during, and after GnRH3 application under voltage clamp (Fig. 4A). The results indicate that the current increased during GnRH3 application and recovered after washout to the level before application. The reversal potential (calculated by the point of intersection of during and before lines) was −85.6 ± 2.4 mV (n = 7), which was near the calculated reversal potential for potassium, −97 mV. Therefore, this result suggests that GnRH3 modulates one or more K⁺ conductance. Because GnRH receptors are coupled to G_q/11 protein (Naor et al. 1995; Stojilkovic et al. 1994a), GnRH3 is expected to modulate the current-voltage relationships via modulation of K⁺ conductance(s) activated by intracellular Ca²⁺ increase. To examine whether Ca²⁺-activated K⁺ channels are modulated by GnRH3, we analyzed the GnRH3-induced current recorded in plain ACSF or in ACSF containing IBX, a blocker of BK channel (Waring and Turgeon 2009). Figure 4B shows the GnRH3-induced current [(current during GnRH3 application) − (current before GnRH3 application)]. The GnRH3-induced current was almost completely blocked by IBX application. This result suggests that GnRH3 suppresses the excitability of SPV neurons by activating BK channels.

Fig. 4.

GnRH3 modulates the large-conductance Ca²⁺-activated K⁺ channel. A: current-voltage relationship before, during, or after application of 1 μM GnRH3 during the falling phase of the ramp stimulation (dashed rectangle in inset). Inset shows the entire ramp protocol. Inset scale: 50 mV, 1.5 ms. Before, 1 min before GnRH3 application; during, 2 min after GnRH3 application; after, 5 min after finishing GnRH3 application. B: GnRH3-induced current calculated by subtracting the current before GnRH3 application from the current during GnRH3 application. n = 7 for 1 μM GnRH3 n = 4 for 1 μM GnRH3 with 100 nM iberiotoxin (IBX). Vertical bar through each symbol indicates SE.

Next we examined whether blockage of BK channel diminishes the GnRH3-induced modulation of the field potential. We performed the field potential recording as shown in Fig. 1B while blocking BK channels by perfusing IBX during the entire recording (Fig. 5). When GnRH3 was applied in the presence of IBX, the GnRH3-induced modulation of the field potential was nullified (Fig. 5A); as shown in Fig. 5, B and C, the differences in rise time and the duration measured at half-maximum amplitude were not significant (Fig. 5B: vehicle: 100.6 ± 0.2% vs. GnRH3: 98.3 ± 1.9%; Student's t-test, n.s.; Fig. 5C: vehicle: 99.0 ± 2.0% vs. GnRH3: 103.0 ± 3.6%; Student's t-test, n.s.). This result strongly suggests that GnRH3 modulates retino-tectal neurotransmission by activating BK channels of the SPV neurons postsynaptically.

Fig. 5.

GnRH3 modulates retino-tectal neurotransmission via large conductance Ca²⁺-activated K⁺ channel. A: field potentials evoked by electrical stimulation of the optic tract (average of 3 traces) in the presence of 100 nM IBX before and during 100 nM GnRH3. B and C: normalized rise time (B) and duration measured at half-maximum amplitude (C) during application of vehicle or GnRH3 (100 nM) in the presence of IBX. There was no significant statistical difference between vehicle and GnRH3 in B and C (n = 5, Student's t-test). Before, 1 min before GnRH3 application; during, 2 min after GnRH3 application. In the box plot, bottom and top bars are minimum and maximum of each group, respectively, box is quartile area, and middle horizontal bar indicates the median.

Localization of GnRH receptors in optic tectum.

Finally, we examined the localization of GnRH receptors in SPV neurons in brain slices with a slight modification of a method involving fluorescently labeled GnRH peptides (GnRH-fluoroprobe) developed for dissociated cells (Hazum et al. 1980; Hazum and Nimrod 1982; Lloyd and Childs 1988). Because technical difficulties so far have prevented analysis of the localization of GnRH receptors by immunohistochemistry or in situ hybridization in the brain, we here examined the localization of GnRH receptors in the optic tectum by an improved method using GnRH-fluoroprobe. We first demonstrated that our synthesized fluoroprobe selectively binds to GnRH receptors with a luciferase assay and a morphological binding assay (see materials and methods for more information; images not shown). We used fluorescein-labeled GnRH for brain slices because the fluorescein signals can be amplified by immunohistochemical methods. GnRH-fluoroprobe was applied for 1 h and was amplified with the methods described in materials and methods. Figure 6 shows GnRH-fluoroprobe signal (Fig. 6A), Hoechst nuclear counterstain (Fig. 6B), and overlay (Fig 6, C and D) in the optic tectum. As shown in Fig. 6, C and D, the GnRH-fluoroprobe signals were observed to surround the cell bodies of the SPV neurons (Fig. 6, Aa, Ca, and Da). Vehicle application did not result in specific binding (Fig. 6, Ab, Cb, and Db). The binding of the fluoroprobe was competitively inhibited by a high dose (10 μM) of nonlabeled GnRH3 peptide (Fig. 6, Ac, Cc, and Dc). The binding of fluoroprobes on synaptic terminals was hard to detect, indicating that the concentration of GnRH receptors in the terminals is very low (Maruska and Tricas 2007). The results of the GnRH-fluoroprobe experiment provide morphological supports for the hypothesis that GnRH3 modulates postsynaptic neurons, the SPV neurons, indicated by our electrophysiological results.

Fig. 6.

GnRH receptors are mainly localized in SPV neurons. A: fluorescent signals of GnRH-fluoroprobe (GnRH-FL). B: signals of Hoechst nuclear stain. C: overlay images. D: magnified overlay images of the areas indicated in white squares in C, in the presence of 1 μM GnRH-fluoroprobe (a), vehicle (b), and 1 μM GnRH-fluoroprobe + 10 μM GnRH3 (c). Each photograph is representative of 3 trials. In the overlay images, magenta is GnRH-fluoroprobe and green is Hoechst. Scale bars: 20 μm (A), 10 μm (D).

DISCUSSION

In the present study, we showed that GnRH3 modulates retino-tectal neurotransmission postsynaptically. GnRH3 increased the rise time and the duration measured at half-maximum amplitude of the field potentials evoked by optic tract fiber stimulations. GnRH3 also suppressed the excitability of SPV neurons by activating BK channels. Our morphological analysis using a GnRH-fluoroprobe indicated that SPV neurons express GnRH receptors.

GnRH3 suppresses excitability of SPV neurons and modulates LFPs by activating BK channels via GnRH receptors.

We first analyzed field potentials, which represent monosynaptic neurotransmission from optic tract fibers to SPV neurons in the optic tectum (Vanegas et al. 1974a, 1974b). The results of analysis showed that GnRH3 modulates the field potential, and our PPR analysis indicates that the modulatory effects are postsynaptic. We then showed that GnRH3 suppresses the excitability of SPV neurons by activating BK channels. It has been generally accepted that currents that are induced by BK channels decrease the number of action potentials and thereby suppress the excitability of neurons (Arias-Garcia et al. 2013; Sun and Dale 1998).

We next analyzed the distribution of GnRH-fluoroprobe binding sites as a way to examine the localization of GnRH receptors. Previous immunohistochemical studies in teleost brains indicated that TN-GnRH3 neurons have massive projection to the deep layers [stratum album centrale (SAC) and stratum griseum centrale (SGC)] and sparse projection to the surface layers in the optic tectum (Maruska and Tricas 2007; Oka and Ichikawa 1990). On the other hand, the optic tract fibers mainly project to the superficial layers (SO and SFGS, where we recorded LFPs) and also sparsely to SAC and SGC (Maruska and Tricas 2007; von Bartheld and Meyer 1987). Our analysis of the localization of GnRH receptors in the optic tectum with GnRH-fluoroprobe showed that GnRH receptors are expressed in SPV neurons. Taken together, these results suggest that TN-GnRH3 neurons mainly act on the cell bodies of the SPV neurons in the deep layer of the optic tectum.

Figure 7 schematically shows the hypothetical model (Fig. 7A), the GnRH3-induced modulation of the field potential (Fig. 7B), and SPV neuron action potentials (Fig. 7C). We illustrated the schematic diagram in Fig. 7B by referring to Vanegas (1983), who suggested that the long dendrite of the SPV neuron possesses an ability to generate regenerative action potentials, and Rall and Shepherd (1968), who also suggested the presence of active dendrites of the mitral cell. We further assumed that the LFP evoked by optic nerve stimulation represents superimposition of excitatory postsynaptic potentials (EPSPs) and back-propagated dendritic spikes. The dashed and solid lines in Fig. 7B indicate the LFP before and during GnRH3 application, respectively. Although the difference in the anatomical localization of GnRH receptors (somatic region of SPV neurons in deep layer) and recording site of the LFPs (synaptic region in superficial layer) may appear puzzling at first sight, the above-mentioned interpretation of the LFP can reasonably explain the present experimental results. Actually, Vanegas (see p. 59 and 60, Vanegas 1983; Fig. 1, Vanegas et al. 1974a) suggested that the negative component of the LFP evoked by the stimulation of optic nerve in fish optic tectum was the superimposition of extracellularly recorded EPSPs and regenerative population spikes. He also suggested that the rise time and duration of the LFP evoked by optic tract fiber stimulation change as shown in Fig. 7B (present study), when the number of action potentials of SPV neurons was experimentally decreased.

Fig. 7.

Hypothetical model of GnRH3-induced modulation of retino-tectal neurotransmission. A: experimental arrangement and schematic image of GnRH3-induced modulation of retino-tectal neurotransmission. B: schematic diagram of local field potential (LFP) illustrated following Venegas (1983) and Rall and Shepherd (1968). Dashed line, field potential before GnRH3 application; solid line, field potential during GnRH3 application. The negative component of the LFP evoked by the stimulation of optic nerve in the optic tectum is based on the assumption that the LFP is the superimposition of extracellularly recorded excitatory postsynaptic potentials EPSPs and regenerative population spikes. C: illustration of the result of whole cell recording of SPV neurons evoked by current injection showing that the excitability of SPV neurons is suppressed by GnRH3.

In summary, considering our electrophysiological and morphological results, we propose a model of GnRH3-induced modulation of retino-tectal neurotransmission (Fig. 7A). First, GnRH3 activates BK channels in SPV neurons, suppressing the excitability of those neurons (Fig. 7C). Then the number of synaptically driven action potentials in SPV neurons evoked by optic tract fiber stimulation is decreased because GnRH3 suppresses excitability. As a result, the rise time and the duration of the LFP increase (Fig. 7B).

Possible intracellular mechanisms of GnRH3-induced modulation.

Here we discuss possible intracellular mechanisms of GnRH3-induced modulation. Although one previous study in rainbow trout examined GnRH-induced modulation of retino-tectal neurotransmission (Kinoshita et al. 2007), the study did not analyze the mechanisms in detail. The results of the present study suggest the following mechanisms. It has been reported that GnRH receptors are G_q/11 protein-coupled receptors (Naor et al. 1995; Stojilkovic et al. 1994a). The G_q/11-type G proteins are coupled to the phospholipase C-mediated signaling pathway (Naor et al. 1995; Stojilkovic et al. 1994a, 1994b). By activating this pathway, GnRH3 induces intracellular Ca²⁺ increase due to release from IP₃-sensitive Ca²⁺ stores (Abe and Oka 2002; Karigo et al. 2014). This intracellular Ca²⁺ increase is considered to activate BK channels. On the other hand, phospholipase C-mediated signaling pathways also induce diacylglycerol production (Naor et al. 1995). Diacylglycerol can be converted to arachidonic acid, and arachidonic acid in turn activates BK channels in mouse vomeronasal neurons (Zhang et al. 2008). One or both of these pathways, intracellular Ca²⁺ increase and/or arachidonic acid modulation, may operate in SPV neurons and suppress the excitability of SPV neurons via activation of BK channels. In the presence of an intracellular solution with high EGTA (1 mM) and no GTP, GnRH3 did not induce a K⁺ current (data not shown), which strongly supports our hypothesis that GnRH3 activates BK channels by causing an increase in intracellular Ca²⁺. Similar modulation of BK channels by GnRH has also been reported in previous studies of gonadotrophs (Sikdar et al. 1989; Waring and Turgeon 2009). Previous studies of modulation of neuronal excitability by GnRH have focused on facilitation caused by inhibition of the K⁺ M current (Brown 1988). Thus, to our knowledge, the present study is the first to show the inhibition of neuronal excitability by GnRH3.

Expected physiological functions of GnRH3-induced modulation.

Retino-tectal neurotransmission plays a central role in visual processing in nonmammalian vertebrates (Karten and Shimizu 1989; Kinoshita and Ito 2006). Tectal neurons in vertebrates receive visual inputs in the surface layer and somatosensory inputs in the deep layer, integrate them, and use this input to select appropriate behaviors (Butler and Hodos 2005; Finlay et al. 1978; Llinas̀ et al. 1976). Our present study contributes to our understanding of GnRH3-induced modulation of visual processing in the optic tectum. We demonstrated that GnRH3 suppresses the excitability of postsynaptic neurons via BK channels, thereby modulating retino-tectal neurotransmission. This modulation of visual information processing could have several effects. For example, it is possible that habituation of visual information processing (Northmore and Gallagher 2003) is prevented by suppressing the excitability of projection neurons. GnRH3-induced modulation may expand the dynamic range of optic tectum response to visual input, as is the case in modulation of auditory system via BK channels (Kurt et al. 2012). Animals generally need visual cues for the success of reproduction and change visual preferences seasonally (Heubel and Schlupp 2008; Satou et al. 1994; Thompson et al. 2004). Previous studies also suggest that TN-GnRH3 neurons change their firing activity when animals receive visual cues from the opposite sex (Okuyama et al. 2014; Ramakrishnan and Wayne 2009). Therefore, GnRH3-induced modulation of synaptic transmission from optic tract fibers to SPV neurons should play an important role in the neuromodulation of behaviorally relevant visual information processing during the breeding season.

GRANTS

This work was supported by Japan Society for the Promotion of Science (JSPS) Grant-in-Aid for Scientific Research (S) 26221104 to Y. Oka and Grant-in-Aid for JSPS Fellow 25-8601 and 24-6309 to C. Umatani and R. Misu, respectively.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: C.U., R.M., S.O., K.Y., H.A., and Y.O. conception and design of research; C.U. and R.M. performed experiments; C.U., R.M., S.O., and K.Y. analyzed data; C.U., K.Y., H.A., and Y.O. interpreted results of experiments; C.U. prepared figures; C.U. and Y.O. drafted manuscript; C.U., R.M., S.O., K.Y., H.A., and Y.O. edited and revised manuscript; C.U., R.M., S.O., K.Y., H.A., and Y.O. approved final version of manuscript.