The techniques for preparing and maintaining rat cortical slices in vitro were similar to previously described methods (Koyanagi et al., 2010; Yamamoto et al., 2010; Kobayashi et al., 2012). Briefly, vesicular GABA transporter (VGAT)-Venus line A transgenic rats (Uematsu et al., 2008) of either sex, aged from 16 to 32 days old, were deeply anesthetized using sevoflurane (5%, Pfizer, Tokyo, Japan) and decapitated. The tissue blocks, including the IC, were rapidly removed and stored for 3 min in ice-cold modified artificial cerebrospinal fluid (ACSF) composed of the following (in mM): 230 sucrose, 2.5 KCl, 10 MgSO₄, 1.25 NaH₂PO₄, 26 NaHCO₃, 0.5 CaCl₂, and 10 D-glucose. The coronal slice preparations (350 μm thickness) were obtained from the IC 0–1050 μm caudally from the middle cerebral artery using a microslicer (Linearslicer Pro 7, Dosaka EM, Kyoto, Japan) and incubated at 32°C for 40 min in a submersion-type holding chamber containing 50% modified ACSF and 50% normal ACSF (pH 7.35–7.40). Normal ACSF contained the following (in mM): 126 NaCl, 3 KCl, 2 MgSO₄, 1.25 NaH₂PO₄, 26 NaHCO₃, 2 CaCl₂, and 10 D-glucose. Modified and normal ACSF were continuously aerated with 95% O₂/5% CO₂. The slices were subsequently placed in normal ACSF at 32°C for 1 h and maintained at room temperature until further use for recording.

The slices were transferred to a recording chamber continuously perfused with normal ACSF at a rate of 2.0–2.5 ml/min. Dual or triple whole-cell patch-clamp recordings were obtained from Venus-positive fluorescent neurons and pyramidal cells identified in layer V of the granular or dysgranular IC using a fluorescence microscope equipped with Nomarski optics (BX51, Olympus, Tokyo, Japan) and an infrared-sensitive video camera (Hamamatsu Photonics, Hamamatsu, Japan). The distance between Venus-positive and pyramidal cells was <100 μm. The electrical signals were recorded using amplifiers (Multiclamp 700B, Molecular Devices, Sunnyvale, CA, United States), digitized (Digidata 1440A, Molecular Devices) and observed on-line; the information was stored on a computer hard disk using Clampex software (pClamp 10, Molecular Devices).

The pipette solution used for recordings from interneurons and pyramidal cells contained the following (in mM): 70 potassium gluconate, 70 KCl, 10 HEPES, 15 biocytin, 0.5 EGTA, 2 MgCl₂, 2 magnesium ATP, and 0.3 sodium GTP. The pipette solution had a pH of 7.3 and an osmolarity of 300 mOsm. The liquid junction potential for the current- and voltage-clamp recordings was -9 mV, and the voltage was corrected accordingly. Thin-wall borosilicate patch electrodes (2–5 MΩ) were pulled on a Flaming-Brown micropipette puller (P-97, Sutter Instruments, Novato, CA, United States).

The recordings were obtained at 30–31°C. The seal resistance was >5 GΩ, and only data obtained from electrodes with access resistance of 6–20 MΩ and <20% change during recordings were included in this study. The series resistance was 50% compensated. The membrane currents and potentials were low-pass filtered at 2–5 kHz and digitized at 20 kHz.

Before unitary excitatory/inhibitory postsynaptic current (uEPSC/uIPSC) or miniature EPSC (mEPSCs)/miniature IPSCs (mIPSCs) recordings, the voltage responses of pre- and postsynaptic cells were recorded after applying long hyperpolarizing and depolarizing current pulse (300 ms) injections to examine the basic electrophysiological properties, including input resistance, single spike kinetics, voltage-current relationships, repetitive firing patterns, and firing frequency.

Because some cell pairs had mutual or ≥2 connections, most cells were recorded under voltage-clamp conditions (holding potential = -70 mV) during uEPSC/uIPSC recordings. Short depolarizing voltage step pulses (2 ms, 80 mV) were applied to presynaptic cells to induce action currents. Where indicated, RAMH (Sigma–Aldrich, St. Louis, MO, United States) was added directly to the perfusate.

Before starting the successive recording of uEPSCs/uIPSCs, we checked the stability of their amplitude for 5–15 min, and excluded the connections that exhibited a gradual decrease of uEPSC/uIPSC amplitude.

Unitary events were recorded in normal ACSF for 5–10 min; we applied RAMH for 7.5–10 min, followed by washing for 10 min. To examine the blocking effect of H₃ receptors, uEPSCs/uIPSCs were recorded under normal ACSF for at least 10 min, and JNJ 5207852 dihydrochloride (JNJ5207852; Tocris, Bristol, United Kingdom) or thioperamide maleate salt (Sigma–Aldrich), H₃ receptor antagonists, was applied for 7.5 min. Subsequently, RAMH was applied in combination with JNJ5207852 or thioperamide for 7.5 min.

We recorded mEPSCs from layer V GABAergic interneurons under application of 1 μM tetrodotoxin (TTX) and 100 μM picrotoxin using the same internal pipette solution based on potassium gluconate and KCl as described above. The holding potential was -70 mV (Kobayashi et al., 2012).

Miniature IPSCs were recorded from layer V pyramidal cells under application of 1 μM TTX, 50 μM D-(-)-2-amino-5-phosphonopentanoic acid (D-APV), and 20 μM 6,7-dinitroquinoxaline-2,3-dione (DNQX). The internal pipette solution contained the following (in mM): 120 cesium gluconate, 20 biocytin, 10 HEPES, 8 NaCl, 5 N-ethyllidocaine chloride (QX-314), 2 magnesium ATP, 0.3 sodium GTP, and 0.1 BAPTA. The pipette solutions had a pH of 7.3 and an osmolarity of 300 mOsm. The holding potential was corrected for a calculated liquid junction potential of -12 mV and set at 0 mV.

The RAMH application protocol during mEPSC/mIPSC recording was same as that used for unitary event recording.

Clampfit software (pClamp 10, Molecular Devices) was used to analyze the electrophysiological data. Input resistance was measured from slopes of least-squares regression lines fitted to V-I curves measured at the peak voltage deflection (current pulse amplitude up to -100 pA). By application of depolarizing step current pulses (300 ms), the action potential threshold was identified as the minimal potential from which the first action potential was elicited. The amplitude of action potential was measured from the resting membrane potential to the peak. The amplitude of AHP was measured from the negative peak to the action potential threshold. Repetitive firing in response to long (300 ms) depolarizing current pulses was evaluated by measuring the slope of least-squares regression lines in a plot of the number of spikes versus the amplitude of injected current (F-I curve; up to 250–300 pA). The amplitude of uIPSCs/uEPSCs was measured as the difference between the peak postsynaptic and baseline currents obtained during a 2–3 ms time window close to the onset. To measure the 20–80% rise time and decay time constants of uEPSCs/uIPSCs, single action currents were induced at 0.07–0.1 Hz, and 10–30 postsynaptic events were aligned to the peak of the presynaptic action currents and averaged.

To quantify the effect of RAMH on the amplitude, paired-pulse ratio (PPR) of the 2nd to 1st uEPSC/uIPSC amplitude, coefficient of variation (CV) and failure rate, we analyzed 15 events just before RAMH application and 225 s after drug application for control and RAMH data, respectively. The uEPSC/uIPSC amplitudes in the range of synaptic noise were considered as failures.

Miniature EPSCs/mIPSCs were detected at a threshold of three times the standard deviation of the baseline noise amplitude using event detection software (kindly provided by Dr. John Huguenard, Stanford University). To measure the amplitude, interevent interval, 20–80% rise time, half duration, and 80–20% decay time, mEPSCs/mIPSCs were analyzed from continuous 5 min recordings before and after RAMH application. To obtain cumulative plots of the interevent interval and the amplitude of mEPSCs/mIPSCs, >1000 events were sampled per each neuron (n = 11).

Four male Sprague–Dawley rats (350–400 g) were used for immunohistochemistry. The rats were deeply anesthetized using sodium pentobarbital (80 mg/kg, i.p.) and transcardially perfused with 100 ml of heparinized normal saline, followed by 400 ml of a freshly prepared mixture of 4% paraformaldehyde and 0.001% glutaraldehyde in 0.1 M phosphate buffer (PB; pH 7.4). The IC was removed and postfixed in the same fixative for 2 h at 4°C. Transverse sections were cut on a microslicer (Vibratome, St. Louis, MO, United States) at 60 μm and cryoprotected in 30% sucrose in PB overnight at 4°C. The sections were frozen on dry ice for 20 mins and thawed in phosphate buffer saline (PBS; 0.01 M, pH 7.4) to enhance penetration. The sections were pretreated with 1% sodium borohydride for 30 min to quench the glutaraldehyde and blocked with 3% H₂O₂ for 10 mins to suppress endogenous peroxidases and 10% normal donkey serum (Jackson ImmunoResearch, West Grove, PA, United States) for 30 min to mask secondary antibody binding sites.

For double immunostaining for the H₃ receptor and vesicular glutamate transporter 1 (VGLUT1) or glutamic acid decarboxylase (GAD) 65/67, the sections were incubated overnight in a mixture of rabbit anti-H₃ receptor (1:50; AB5660P, Millipore, Billerica, MA, United States) and guinea pig anti-VGLUT1 (1:2,000; #135 304, Synaptic Systems, Göttingen, Germany) or mouse anti-GAD65/67 (1:1,000; MSA-225, Assay designs Stressgen, Ann Arbor, MI, United States) antibodies. After rinsing in PBS, the sections were incubated with a mixture of 1 nm gold-conjugated donkey anti-rabbit (1:50, EMS, Hatfield, PA, United States) and biotinylated donkey anti-guinea pig or biotinylated donkey anti-mouse (1:200, Jackson ImmunoResearch) antibodies for 2–3 h. The sections were postfixed with 1% glutaraldehyde in PB for 10 min, rinsed several times in PBS, incubated for 5 min in silver intensification solution (IntenSE^TM M, Amersham, Arlington Heights, IL, United States), and rinsed in 0.1 M sodium acetate and PB. Subsequently, the sections were incubated with ExtrAvidin peroxidase (1:5,000, Sigma, St. Louis, MO, United States) for 1 h, and the immunoperoxidase was visualized using nickel-intensified 3,3′-diaminobenzidine tetrahydrochloride (DAB). The sections were further rinsed in PB, osmicated (0.5% osmium tetroxide in PB) for 30 min, dehydrated in graded alcohols, flat-embedded in Durcupan ACM (Fluka, Buchs, Switzerland) between strips of Aclar plastic film (EMS), and cured for 48 h at 60°C. Chips containing immunoreactivity for the H₃ receptor and VGLUT1 or GAD65/67 in the dysgranular IC region of the IC were cut from the wafers and glued onto blank resin blocks using cyanoacrylate. Thin sections were collected on Formvar-coated single-slot nickel grids and stained with uranyl acetate and lead citrate.

The grids were examined on a Hitachi H 7500 electron microscope (Hitachi, Tokyo, Japan) at 80 kV accelerating voltage. The images were captured using Digital Montage software driving a MultiScan cooled CCD camera (ES1000W, Gatan, Pleasanton, CA, United States) attached to the microscope and saved as TIFF files. To control for the specificity of primary antibodies, the sections were processed according to the above protocol, except that the primary or secondary antibodies were omitted to completely abolish specific staining.

The method of optical imaging with a voltage-sensitive dye has been described previously (Fujita et al., 2010, 2011, 2012, 2017; Mizoguchi et al., 2011), and only a brief account of the methods employed will be given here. Five-week-old male Sprague–Dawley rats (Sankyo Labo, Tokyo, Japan) weighing 152.9 ± 4.5 g (n = 12) received atropine methyl bromide (5 mg/kg, i.p.) and were anesthetized with urethane (1.5 g/kg, i.p.). Anesthesia was maintained throughout the experiments by injecting additional urethane. The adequacy of anesthesia was gauged by the absence of the toe pinch reflex. Animals received a tracheotomy and intubation, and lidocaine (2% gel) was applied to the incisions to ensure complete analgesia. The anesthetized animal was mounted on a custom-made stereotaxic snout frame (Narishige, Tokyo, Japan), which was then tilted 60° laterally to make the left IC accessible to the CCD camera system and electrodes. The left temporal muscle and zygomatic arch were carefully removed, and a craniotomy was performed to expose the IC and the surrounding cortices.

The voltage-sensitive dye, RH1691 (1 mg/ml, Optical Imaging, New York, NY, United States) in 0.9% saline, was applied to the cortical surface for approximately 1 h, residual dye was rinsed with saline for approximately 30 min, and fluorescent changes of RH1691 were measured by a CCD camera system (MiCAM02, Brainvision, Tokyo, Japan) mounted on a stereomicroscope (Leica Microsystems, Wetzlar, Germany). The cortical surface was illuminated through a 632 nm excitation filter and a dichroic mirror by a tungsten-halogen lamp (CLS150XD, Leica Microsystems), and the fluorescent emission was captured through an absorption filter (λ > 650 nm long-pass, Andover, Salem, NH, United States). The CCD-based camera had a 6.4 mm² × 4.8 mm² imaging area consisting of 184 pixels × 124 pixels. To eliminate signal due to acute bleaching of the dye, the final image was obtained by subtracting an image without stimulation from an image with stimulation. The sampling interval was 4 ms. The acquisition time was set at 500 ms. Sixteen consecutive images in response to stimuli were averaged to reduce noise, including interference from the heartbeat, breathing, and spontaneous neural activity.

A tungsten concentric bipolar electrode (impedance = 500 KΩ; Unique Medical, Tokyo, Japan) was inserted 0.3 mm from the surface of the IC. The duration and intensity of stimulation pulse was set at 100 μs and 7 V. Five train pulses with a 50 ms interstimulus interval were applied. Each train pulse was delivered at 30 s intervals. After optical recording in control, RAMH (10 μM) was applied to the cortical surface for 60 min. Another optical recording was performed during the last 30 min of RAMH application.

The change in the intensity of fluorescence (ΔF) in each pixel, relative to the initial fluorescence intensity (F), was calculated (ΔF/F). For the measurement of amplitude and excitatory area, images of responses were processed with a spatial filter (9 pixels × 9 pixels) to avoid the influence of noise and brain movements on each pixel. A significant response was defined as a signal exceeding seven times the SD of the baseline noise. Optical imaging data were processed and represented using software (Brain Vision Analyzer, Brainvision, Tokyo, Japan). To quantify the spatiotemporal profiles of excitation, we set a region of interest (ROI) in the adjacent region of the stimulation site.

All chemicals, unless otherwise specified, were purchased from Sigma–Aldrich.

The data are presented as the means ± SEM. Comparisons of the uEPSC/uIPSC amplitude and PPR between control and drug application were conducted using the paired t-test. The suppression rates of uEPSCs in pyramidal cell to pyramidal cell (Pyr→Pyr) or Pyr→GABAergic interneuron (Interneuron) connections, and uIPSCs in Interneuron→Pyr connections were analyzed using one-way ANOVA followed by post hoc Bonferroni test. Reductions in the amplitude of uEPSC/uIPSCs induced through five trains or paired-pulses were compared using Student’s t-test. Failure rate was compared between control and RAMH using Wilcoxon signed-rank test. Box-and-whisker plots were made as follows: upper and lower limit of box indicates 75th and 25th percentile, respectively; line within box denotes the median; upper and lower limits of the capped lines indicate 90th and 10th percentile, respectively. The RAMH-induced suppression rates of uIPSC amplitude were compared between the 1st and the 2nd–5th uIPSCs using two-tailed multiple t-tests with Bonferroni correction. The amplitude and interevent interval of mEPSCs/mIPSCs were analyzed through non-parametric statistical analysis (Kolmogorov–Smirnoff test; K-S test) to assess the significance of the shifts in the cumulative probability distributions in control and RAMH application. Paired t-tests were used to compare the mean amplitude and interevent interval of mEPSCs/mIPSCs and the mean amplitude of excitation (ΔF/F) and excitation area in the optical imaging study. A level of P < 0.05 indicated statistical significance.

To examine the H₃ receptor-mediated effects on the excitatory synaptic connections (Pyr→Pyr, Pyr→FS, and Pyr→non-FS), 10 μM RAMH, an H₃ receptor agonist, was applied to the perfusate. The pyramidal cells were identified among Venus-negative pyramidal somata after repetitive firing with spike adaptations (Figure 1A).

In Pyr→Pyr connections, the bath application of RAMH (10 μM) invariably suppressed the 1st uEPSC amplitude by 30.4 ± 4.6% (n = 17; P < 0.05, paired t-test; Figures 1B–D), which was not recovered after a 10 min washout (Figure 1C). The RAMH-induced uEPSC suppression was accompanied by increases in the failure rate of the 1st uEPSCs (n = 17; P < 0.01, Wilcoxon test) and PPR (n = 17; P < 0.01, paired t-test), as shown in Figure 1D. In these Pyr→Pyr connections, the 2nd to 5th uEPSCs were less affected after the application of RAMH (Figure 1E), suggesting that the effect of RAMH is likely to be mediated by a presynaptic mechanism.

In the Pyr→FS connections, 10 μM RAMH invariably suppressed the 1st uEPSC amplitude by 30.5 ± 6.7% (n = 12; P < 0.01, paired t-test; Figures 2A,C). Similarly, Pyr→non-FS connections showed consistent suppression of the 1st uEPSC amplitude by 52.9 ± 12.8% (n = 6; P < 0.05, paired t-test; Figures 2B,D). The RAMH-induced suppression of uEPSCs did not recover after the 10 min washout (Figures 2E,F). As there was no significant difference between the reduction rates in the 1st uEPSC amplitude (P = 0.19, Student’s t-test), we combined the results obtained from Pyr→FS and Pyr→non-FS connections into one category, Pyr→Interneuron. Generally, RAMH suppressed the 1st uEPSC amplitude by 37.9 ± 6.6% (n = 18; P < 0.01, paired t-test), which was comparable to that observed in Pyr→Pyr connections (P = 0.37, Student’s test). The RAMH-induced uEPSC suppression was accompanied by increases in the failure rate of the 1st uEPSCs (n = 18; P < 0.05, Wilcoxon test) and PPR (n = 18; P < 0.01, paired t-test), as shown in Figure 2G.

In contrast to the smaller uEPSC amplitude in response to the 2nd to 5th presynaptic action currents in Pyr→Pyr connections (Figure 1F), Pyr→Interneuron connections frequently showed paired-pulse facilitation. RAMH had little effect on the 2nd to 5th uEPSC amplitude in Pyr→Interneuron connections (n = 18; P > 0.1, paired t-test; Figure 2I).

RAMH did not induce a significant change in input resistance in Pyr (n = 90, P = 0.79, paired t-test), FS (n = 30, P = 0.85, paired t-test), and non-FS neurons (n = 8, P = 0.91, paired t-test).

These results suggest that a decrease in uEPSCs through RAMH might reflect presynaptic modulation via H₃ receptors.

Coefficient of variation analysis is another method to evaluate the synaptic efficacy changes occur at pre- or postsynaptic sites (Faber and Korn, 1991; Arami et al., 2013). According to a binomial model of synaptic transmission, functional changes in presynaptic sites are expected to be accompanied by a change in the CV of synaptic responses: Presynaptic changes result in the plot that CV^-2 values against the efficacy changes are located on or below the diagonal line in the case of low release probability. Our CV analysis show in Figures 1E, 2H supports the hypothesis that the suppressive effect of RAMH on uEPSCS is likely to be mediated via a presynaptic mechanism.

Figure 3 shows a typical example of the effects of RAMH on uIPSCs obtained from FS→Pyr (Figures 3A,C) and non-FS→Pyr (Figures 3B,D) connections. In both pairs, 10 μM RAMH almost consistently suppressed the amplitude of uIPSCs. RAMH reduced the 1st uIPSC amplitude by 20.0 ± 5.3% in FS→Pyr connections (n = 31; P < 0.01, paired t-test) and 35.7 ± 9.6% in non-FS→Pyr connections (n = 8; P < 0.05, paired t-test). Figures 3E,F are the time courses of uIPSC amplitude before, during and after the application of RAMH in FS→Pyr and non-FS→Pyr shown in Figures 3A–D, respectively. The recovery of uIPSC amplitude was not observed after a 10 min washout.

As there was no significant difference between these reduction rates (P = 0.11, Student’s t-test), we combined the results obtained from FS→ and non-FS→Pyr synapses into one category, Interneuron→Pyr. Generally, RAMH suppressed uIPSC amplitude by 23.0 ± 4.7% (n = 39; P < 0.01, paired t-test), which was almost comparable to that of the uEPSC suppression observed in Pyr→Pyr or Pyr→Interneuron connections (P = 0.19, one-way ANOVA).

Similar to Pyr→Interneuron connections, H₃ receptors likely suppress uIPSC amplitudes via a presynaptic mechanism. The RAMH-induced uIPSC suppression was accompanied by increases in the failure rate of the 1st uIPSCs (n = 39; P < 0.05, Wilcoxon test) and PPR (n = 39; P < 0.05, paired t-test), as shown in Figure 3G. The result of CV analysis (Figure 3H) also suggested that the suppressive effect of RAMH on uIPSCS is induced by a presynaptic mechanism. Interneuron→Pyr connections exhibited paired-pulse depression in the control, and the analysis of the 2nd and 5th uEPSC amplitude showed a significant suppression after the application of RAMH (n = 33; P < 0.05, paired t-test; Figure 3I). The RAMH-induced suppression rate of the 1st uIPSC amplitude was significantly larger than those of the 2nd and 5th uIPSCs (P < 0.001; two-tailed multiple t-test with Bonferroni correction), suggesting that the effect of RAMH is likely to be mediated by a presynaptic mechanism.

These results suggest that H₃ receptors suppress the uIPSC amplitude between GABAergic and pyramidal cells in the IC, which is likely mediated via a presynaptic mechanism, i.e., the suppression of GABA release from FS and non-FS cells to pyramidal cells.

To confirm that the suppression of uIPSCs through RAMH is mediated via H₃ receptors, we examined the effects of the pre-application of JNJ5207852, an H₃ receptor antagonist, on the RAMH-induced suppression of uIPSCs. Figures 4A–C showed a typical example of the effects of the pre-application of 10 μM JNJ5207852 followed by 10 μM RAMH on uIPSCs obtained from FS→Pyr connections. Application of JNJ5207852 itself had little effect both on the amplitude of uEPSCs obtained from Pyr→Pyr/Interneuron connections (38.6 ± 8.8 pA to 26.4 ± 5.3 pA, n = 8, P = 0.14, paired t-test) and uIPSCs in Interneuron→Pyr connections (88.8 ± 46.5 pA to 89.3 ± 45.9 pA, n = 13, P = 0.15, paired t-test). RAMH had little effect of the amplitude of Pyr→Pyr/Interneuron connections under application of JNJ5207852 (26.4 ± 5.3 pA to 25.9 ± 5.3 pA, n = 8; Figure 4D). Similarly, in 13 Interneuron→Pyr connections, RAMH (10 μM), in combination with JNJ5207852 (10 μM), had little effect on the amplitude of uIPSCs (89.3 ± 45.9 pA to 93.8 ± 45.7 pA; n = 13, P = 0.66, paired t-test; Figure 4D). These results suggest that the RAMH-induced suppression of uEPSCs and uIPSCs is induced through H₃ receptors.

We also examined the effects of thioperamide, an H₃ receptor antagonist, which is known as an inverse agonist (Morisset et al., 2000). Figures 5A–C showed a typical example of the effects of the pre-application of 10 μM thioperamide followed by 10 μM RAMH on uIPSCs obtained from FS→Pyr connections. Under thioperamide application, RAMH had little effect on the uIPSC amplitude. In 14 Interneuron→Pyr connections, RAMH (10 μM), in combination with thioperamide (10 μM), had little effect on the amplitude of uIPSCs (244.9 ± 59.7 pA to 224.4 ± 38.7 pA; n = 8, P = 0.48, paired t-test; Figure 5D). Similarly, RAMH also showed little effect of the amplitude of Pyr→Pyr connections under application of thioperamide (14.7 ± 6.8 pA to 15.0 ± 7.3 pA, n = 2). These results indicated that the RAMH-induced suppression of uIPSC amplitudes was mediated through H₃ receptors.

In addition to the blockade of RAMH-induced suppression of uIPSCs by thioperamide, we observed that the application of thioperamide alone increased the uIPSC amplitude compared with the control (Figures 5B,C), though thioperamide had no significant effect on uEPSCs in Pyr→Pyr connections (n = 5; P = 0.58, paired t-test; Figure 5E). In 10 Interneuron→Pyr connections, thioperamide (10 μM) significantly increased the amplitude of uIPSCs by 102.3 ± 37.1% (P < 0.05, paired t-test; Figure 5F). The enhancement of uIPSCs through thioperamide application was not accompanied by significant changes in the failure rate of the 1st uIPSCs (n = 10, P = 0.18, Wilcoxon test) and the PPR (n = 10; P = 0.57, paired t-test; Figure 5F), and the facilitative effect of thioperamide was larger in the 1st uIPSCs compared with the 3rd uIPSCs (P < 0.05; two-tailed multiple t-test with Bonferroni correction; Figure 5G).

Thioperamide did not induce a significant change in input resistance in Pyr (n = 37, P = 0.95, paired t-test) and FS (n = 10, P = 0.96, paired t-test).

These results suggest that thioperamide facilitates GABAergic synaptic transmission via postsynaptic mechanisms in the connection of Interneuron→Pyr.

The frequency of synaptic events under the application of the Na⁺ channel blocker, TTX, reflects the release probability of neurotransmitters, including GABA and glutamate. To strengthen the hypothesis that H₃ receptors suppress GABA/glutamate release via presynaptic H₃ heteroreceptors, mEPSCs and mIPSCs were recorded from pyramidal cells and GABAergic interneurons, respectively, under the application of TTX and glutamatergic/GABAergic receptor antagonists.

Figure 6 shows the typical effect of RAMH on mEPSCs obtained from GABAergic interneurons in the layer V IC. In addition, 100 μM of picrotoxin was applied to block GABA_A receptor-dependent currents. The interevent interval of mEPSCs was reduced through RAMH (Figures 6A–C; P < 0.001; K-S test) without changing the amplitude (Figures 6A–C; P = 0.16; K-S test). A total of 11 neurons were analyzed to compare the mean interevent interval and amplitude of mEPSCs recorded under control and RAMH conditions. RAMH significantly increased the interevent interval of mIPSCs (0.40 ± 0.06 s vs. 0.48 ± 0.07 s, n = 11, P < 0.01, paired t-test) without affecting the mIPSC amplitude (11.7 ± 0.9 pA vs. 11.9 ± 0.9 pA, n = 11, P = 0.66, paired t-test; Figure 6D).

We recorded mIPSCs from layer V IC pyramidal cells under the application of TTX (1 μM), D-APV (50 μM), and DNQX (20 μM; Figures 7A–C). Similar to mEPSCs, RAMH (10 μM) reduced the frequency of mIPSCs (Figures 7A,B). A cumulative probability plot obtained from the same neuron indicates that RAMH significantly increased the interevent interval of mIPSCs (P < 0.001; K-S test; Figure 7C). However, RAMH did not affect the amplitude of mIPSCs (P = 0.51; K-S test; Figure 7C). A total of 11 neurons were analyzed to compare the mean interevent interval and amplitude of mIPSCs recorded under control and RAMH conditions. RAMH significantly increased the interevent interval of mIPSCs (0.33 ± 0.03 s vs. 0.45 ± 0.07 s, n = 11, P < 0.05, paired t-test) without affecting the mIPSC amplitude (16.4 ± 0.6 pA vs. 15.9 ± 0.6 pA, n = 11, P = 0.12, paired t-test; Figure 7D).

These miniature recordings support the hypothesis derived from the results of unitary synaptic event recordings, suggesting that H₃ receptors in glutamatergic and GABAergic terminals regulate the release of glutamate and GABA, respectively.

The electrophysiological findings in the present study strongly suggest the presynaptic regulation of glutamate and GABA release through H₃ receptors. Electron microscopy and immunohistochemistry are potent tools for providing direct evidence of the existence of H₃ receptors in glutamatergic or GABAergic terminals. We performed pre-embedding electron microscopy and immunohistochemistry for the H₃ receptor and GAD, a GABA-synthesizing enzyme that catalyzes the decarboxylation of glutamate, or for H₃ receptors and VGLUT1, a suitable marker for immunolabeling glutamate terminals.

Figures 8A,B show two examples of the observed localization of H₃ receptors on the GAD-immunopositive terminals in the dysgranular IC. H₃ receptors and GAD were labeled with silver-gold (arrowheads) and immunoperoxidase (arrows), respectively. Similarly, H₃ receptors labeled with silver-gold (arrowheads) were occasionally observed on VGLUT1-immunopositive terminals, as shown in Figures 8C,D. Most presynaptic GAD/VGLUT1-immunopositive axons with H₃ receptors terminated to middle-sized dendrites.

To elucidate the functional role of H₃ receptors in cortical information processing in vivo, we examined the effects of RAMH on excitatory propagation in the IC by the optical imaging technique using a voltage-sensitive dye. As previously we reported, repetitive electrical stimulation of the DI rostral to the MCA, i.e., the gustatory cortex, consistently evokes excitation that propagates rostrocaudally, which is principally mediated by AMPA and GABA_A receptors (Fujita et al., 2010, 2011, 2012; Mizoguchi et al., 2011). Therefore, we used the same protocol to evaluate the effects of RAMH on glutamatergic and GABAergic synaptic transmission in the IC.

Electrical stimulation of the dysgranular IC (5 train of pulses at 20 Hz, 7 V) evoked excitatory propagation that spread parallel to the rhinal fissure (Figure 9A). The amplitude of excitation was largest around the stimulation electrode (ΔF/F = 0.25 ± 0.03%, n = 12; Figure 9B). Application of RAMH (10 μM) to the cortical surface decreased the amplitude between the bottom to peak of optical signals responding to the 1st stimulus (n = 12; P < 0.001, paired t-test; Figures 9A–C). Although the RAMH-induced suppression of the amplitude was also observed in the responses to the 2nd to 5th stimuli, the 1st responses showed the most prominent suppression by RAMH and the suppression was gradually recovered in the following responses (Figure 9D). In parallel to the RAMH-induced suppression of the excitation amplitude, the area of excitation was decreased in the 1st responses from 11.1 ± 0.8 mm² to 8.3 ± 0.9 mm² (n = 11, P < 0.01, paired t-test), and the latter responses showed less decrease in the excitation area (P = 0.06–0.41, paired t-test; Figure 9D). These results suggest that RAMH suppresses excitatory propagation especially responding to the 1st stimulation, though the RAMH-induced suppression is less in the latter responses.

Histaminergic axons derived from the TMN widely project to the cerebral cortex, including the IC (Haas et al., 2008). Previous autoradiographic studies using the H₃ receptor agonists, [³H]RAMH or N alpha-[³H]methylhistamine, have demonstrated the rostrocaudal gradient of H₃ receptor density (Pollard et al., 1993), and the IC shows the highest abundance of H₃ receptor expression in the cerebral cortex (Pollard et al., 1993; Cumming et al., 1994). Consistently, the in situ hybridization of H₃ receptor mRNA combined with autoradiographic mapping of the H₃ receptor antagonist, [¹²⁵I]iodoproxyfan, demonstrated high concentration of H₃ receptors in the cerebral cortex, particularly in the IC (Pillot et al., 2002). Therefore, the IC is considered as the best target to examine the physiological roles of H₃ receptors in the cerebral cortex.

In contrast to the subcellular distribution patterns of H₃ receptors in the TMN, the distinct spatial patterns of H₃ receptors in cerebrocortical neurons remain unclear. The present study corroborates previous reports of H₃ receptor expression in the IC and further demonstrates that H₃ receptors are expressed in VGLU1- or GAD-immunopositive terminals. Several studies have demonstrated the expression of H₃ heteroreceptors in glutamatergic terminals of the dentate gyrus (Brown and Reymann, 1996; Doreulee et al., 2001) and striatum (Doreulee et al., 2001) or GABAergic terminals in the hypothalamus (Jang et al., 2001). However, these electrophysiological findings only implicate the presynaptic functions of H₃ receptors. Thus, electron microscopy and immunohistochemistry performed in the present study provides a direct anatomical evidence of H₃ heteroreceptors in the cerebral cortex.

An in vitro whole-cell patch-clamp study in the rat dentate gyrus demonstrated that histamine or RAMH reduces the amplitude of evoked EPSCs and increases in PPR (Brown and Reymann, 1996) through a reduction in Ca²⁺ influx (Brown and Haas, 1999). This finding is applicable in vivo, as RAMH decreases evoked EPSP amplitude, which is blocked through thioperamide, in the dentate gyrus (Manahan-Vaughan et al., 1998). Similarly, the H₃ heteroreceptor-mediated suppression of evoked EPSPs/EPSCs has been reported in the striatum (Doreulee et al., 2001) and basolateral amygdala (Jiang et al., 2005). Despite abundant H₃ receptor expression in the cerebral cortex, there is little evidence showing the functional roles of H₃ receptors. Microdialysis in the prefrontal cortex demonstrated that thioperamide does not affect the basal release of glutamate and GABA (Welty and Shoblock, 2009). However, GABA release through strong stimulation, such as high potassium or NMDA application, is suppressed with thioperamide (Dai et al., 2006; Welty and Shoblock, 2009).

There are at least three subtypes of GABAergic interneurons in the IC (Kobayashi et al., 2010; Koyanagi et al., 2010; Yamamoto et al., 2010), and the kinetics of postsynaptic currents are different among interneuron subtypes, e.g., FS cells induce much larger uIPSCs in amplitude than LTS cells (Xiang et al., 2002; Kobayashi et al., 2008; Koyanagi et al., 2010). Therefore, the identification of presynaptic and postsynaptic neuron subtypes provides valuable information regarding EPSCs/uIPSC recordings. We consider that paired whole-cell recording has a large advantage in the precise analysis of glutamate/GABA-mediated synaptic transmission, as presynaptic and postsynaptic neuron subtypes can be identified. The present study demonstrated that the H₃ receptor agonist consistently suppressed the amplitude of uEPSCs/uIPSCs, regardless of presynaptic or postsynaptic neuron subtypes, suggesting that H₃ receptors are widely distributed in the IC, rather than expressed in specific neurons.

H₃ receptors are metabotropic receptors coupled to G_i/o proteins (Clark and Hill, 1996). H₃ receptors reduce intracellular cyclic AMP (cAMP) concentrations through the suppression of adenylyl cyclase (Torrent et al., 2005; Moreno-Delgado et al., 2006). The reduction in the cAMP concentration decreases Ca²⁺ currents via L, N, and P/Q channels (Takeshita et al., 1998), and in turn, H₃ autoreceptor activation reduces histamine release from presynaptic terminals. Therefore, it is reasonable to speculate that similar mechanisms underlie the regulation of glutamate or GABA release through H₃ heteroreceptors in the cerebral cortex.

The findings of RAMH-induced suppression of uEPSCs/uIPSCs imply that adenylyl cyclase, in glutamatergic or GABAergic terminals, is constantly activated to maintain intracellular cAMP concentrations, thereby contributing to the facilitatory effects on neurotransmitter release from these terminals. Although several studies supported the findings that H₃ receptor activation suppresses spontaneous GABA release by inhibiting P/Q channels in the hypothalamus (Jang et al., 2001) and striatum (Jang et al., 2001; Arias-Montano et al., 2007), they suggest an independent mechanism of the adenylyl-cAMP pathway. Further studies are necessary to clarify the second messenger systems involved in H₃ receptor activation.

The effects of thioperamide on synaptic transmission are controversial. As described above, thioperamide is considered a typical H₃ receptor antagonist (Arrang et al., 1987), whereas several studies have reported thioperamide as an inverse agonist (Morisset et al., 2000). The present study demonstrates that the application of thioperamide slightly enhanced the uIPSC amplitude.

There are at least two possibilities that explain the thioperamide-induced enhancement of uIPSCs. If histamine is spontaneously released and H₃ receptors are consistently activated in the slice preparations, then the application of thioperamide might suppress the H₃ receptor-dependent decrease in the uEPSC/uIPSC amplitude. Second, thioperamide behaves as an inverse agonist. Little effect of JNJ5207852 on uEPSCs and uIPSCs suggests that the 1st possibility is unlikely. The contradictory effects of thioperamide on uEPSCs and uIPSCs also supports the second possibility. Furthermore, in a previous study of IC pyramidal cells, we showed that histamine increases spike firing via H₂ receptors and RAMH does not change the frequency, suggesting that spontaneous histamine release is less likely in the IC slice preparations (Takei et al., 2012). Therefore, the facilitative effects of thioperamide on the amplitude of uEPSCs/uIPSCs might indicate a role for this molecule as an inverse agonist.

H₃ receptors are involved in higher brain functions, such as cognition, learning and memory, sleep, appetite, and energy metabolism, and the disruption of H₃ receptors induces related neurological disorders (Haas et al., 2008). Based on this evidence, H₃ receptors have recently been considered as promising targets for the treatment of several neuronal disorders, including hypersomnia (Leurs et al., 2005), eating disorders (Leurs et al., 2005; Steffen et al., 2006), schizophrenia (Ito, 2009), and pain (Cannon et al., 2007). It is interesting to focus on the study that the activation of H₃ receptors in the IC impairs aversive taste memory formation (Purón-Sierra et al., 2010), because the functional roles of the IC have been implicated in gustatory, visceral, and nociceptive information processing (Yamamoto et al., 1981; Yasui et al., 1991; Ogawa and Wang, 2002; Jasmin et al., 2003; Kobayashi, 2011). Our present findings that activation of H₃ suppresses glutamate and GABA release from presynaptic terminals may be involved in this process. In addition, recent studies have indicated additional functions for the IC in attention, reasoning, planning, and decision-making processes associated with smoking and drug abuse (Naqvi et al., 2007; Naqvi and Bechara, 2009). Thus, H₃ antagonists are potential candidates for the improvement of the physiological functions of the IC.