Male Sprague–Dawley (SD) rats (7 weeks old) were used. Animals were housed under standard conditions (23–25°C, 12 h light/dark cycle) with free access to food and water. All animal protocols were approved by the Administrative Panel on Laboratory Animal Care of Hallym University. All possible efforts were taken to avoid animals’ suffering and to minimize the number of animals used during the experiment. All reagents were obtained from Sigma–Aldrich (St. Louis, MO, USA), except as noted.

Under Isoflurane anesthesia (3% induction, 1.5%–2% for surgery and 1.5% maintenance in a 65:35 mixture of N₂O:O₂), animals were stereotaxically inserted a brain infusion kit 1 connected to an Alzet 1007D osmotic pump (Alzet, Cupertino, CA, USA) into the right lateral ventricle: 1 mm posterior to Bregma; 1.5 mm lateral to midline; 3.5 mm depth to the dural membrane. The infusion kit was sealed with dental cement. Using this infusion system, rats were given: (1) control siRNA; (2) TRPC6 siRNA; (3) vehicle; (4) U0126 (a selective ERK1/2 inhibitor, 25 μM); (5) TRPC6 siRNA + U0126; or (6) hyperforin (a TRPC6 activator, 6 μM). The sequence of TRPC6 rat siRNA (Genolution Pharmaceuticals, Inc., Seoul, South Korea) was 5′-GGAAUAUGCUUGACUUUGGAAUGUUUU-3′ (Kim and Kang, 2015; Ko and Kang, 2017). Non-targeting control siRNA sequence was 5′-GCAACUAACUUCGUUAGAAUCGUUAUU-3′. In our previous study (Ko and Kang, 2017) and the present study, 25–50 μM of U0126 inhibited ERK1/2 phosphorylation in the hippocampus by ~50% after 7 day-over infusion. Osmotic pump was placed in a subcutaneous pocket in the interscapular region.

One week after surgery, animals were anesthetized (urethane, 1.5 g/kg, i.p.), removed the infusion kit and osmotic pump, and placed in a stereotaxic frame. A bipolar stimulating electrode was placed in the perforant path (coordinates: 8 mm posterior, 4.4 mm lateral to Bregma, 3.0–3.3 mm depth). A recording electrode was placed in the dentate gyrus and stratum pyramidale of the CA1 region (coordinates: 3.8 mm posterior to Bregma, 2.5 mm lateral to the midline; 2.5 mm (to stratum pyramidale of the CA1 region) or 2.9 mm (to the dentate gyrus) depth). Electrode depths were finally determined by optimizing the evoked response. The reference electrode was placed in the cerebellum. Body temperature was monitored and maintained at 37 ± 0.3°C by thermostat during recording. After establishing a stable baseline for at least 30 min and a control input–output (IO) curve, stimuli were delivered at interstimulus intervals of 30 ms as DC square pulses at 0.5 Hz with pairs of 150-μs constant current stimuli. The stimulus current intensity was adjusted at an intensity yielding 50% of the maximal amplitude of population spike. For sustained stimuli, stimuli were delivered in 9 s long tetanic (30 ms interval) stimulus trains (300 total stimuli). For analysis of the responsiveness of Kv4.3 to 4-aminopyridine (4-AP), animals were injected with saline or 4-AP (20 μM, 1 μl) using a guide-electrode system (C315G-MS303, Plastics One, Roanoke, VA, USA) over a 1-min period using a microinjection pump (1 μl/min, KD Scientific, Hollistone, MA, USA). Signals were recorded with DAM 80 differential amplifier (0.1–3000 Hz bandpass, World Precision Instruments, Sarasota, FL, USA) and data were digitized (20 kHz) and analyzed on LabChart Pro v7 (AD Instruments, Bella Vista, NSW, Australia).

To analyze changes in evoked response, all of the population spike amplitude and field excitatory postsynaptic potential (fEPSP) slope measurements were normalized by the average basement values of the population spike amplitude and fEPSP slope. To measure the efficiency of glutamatergic synaptic transmission in the DG, the excitability ratio was calculated as the population spike amplitude vs. the fEPSP slope in the first response. The ratio of the second population spike amplitude/the first population spike amplitude (population spike amplitude ratio) was also analyzed in order to measure GABA-mediated inhibition in the DG and the CA1 region, respectively (Figure 3A; Andersen et al., 1964, 1971; Lomø, 1971; Buckmaster and Wong, 2002). If multiple population spikes were detected, the first population spike was used for population spike amplitude ratio. After recording, the animal was quickly decapitated for Western blot and immunohistochemistry (Kim et al., 2008, 2016).

The hippocampal was homogenized in lysis buffer. Thereafter, the protein concentration in the supernatant was determined using a Micro BCA Protein Assay Kit (Pierce Chemical, Rockford, IL, USA). To analyze subcellular localization of Kv4.3, we used subcellular Protein Fractionation Kit for Tissues (Thermo Scientific, Waltham, MA, USA), according to the manufacturer’s instructions. Western blotting was performed according to standard procedures. Nitrocellulose transfer membranes were incubated with primary antibodies such as rabbit anti-TRPC6 (1:1000, Millipore, #AB5574), rabbit-anti EKR1/2 (1:1000, Biorbyt, #orb160960), rabbit-anti phospho (p)-ERK1/2 (1:1000, Millipore, #05-797RSP), mouse-anti glutamate decarboxylase 67 (1:1000, GAD67, Millipore, #MAB5406) or rabbit anti-Kv4.3 (1:1000, Alomone labs, #APC-017) antibody. Immunoreactive bands were detected and quantified on ImageQuant LAS4000 system (GE Healthcare, Piscataway, NJ, USA). The rabbit anti-β-actin primary antibody (for cytosolic fraction, 1:6000, Sigma, #A5316) or rabbit anti-N-cadherin (for membrane fraction, 1:1000, Abcam, #ab18203) was used as internal reference.

Rats were anesthetized with under urethane anesthesia (1.5 g/kg, i.p.) and perfused through the left ventricle with 200 ml of saline as a vascular rinse followed by a fixative solution containing 4% paraformaldehyde in 0.1 M sodium phosphate buffer (PB), pH 7.4. The brains were removed, postfixed for for 4 h at 4°C, and then cryoprotected by immersion overnight at 4°C in 0.1 M PB containing 30% sucrose. Brains were frozen and serial 30 μm-thick frontal sections were cut with a cryostat. Consecutive sections were contained in six-well plates containing PBS. Free-floating sections were first incubated with 10% normal goat serum (Vector, Burlingame, CA, USA) in PBS for 30 min at room temperature. Sections were then incubated in the mixture of primary antibodies at room temperature for overnight in a solution containing rabbit anti-TRPC6 antibody (1:100, Millipore), rabbit anti-Kv4.3 antibody (1:100, Alomone labs), rabbit anti-Kv4.3 (1:100, Alomone labs)/mouse anti-GAD67 (1:250) or rabbit anti-Kv4.3 (1:100, Alomone labs)/mouse anti-parvalbumin (anti-PV, an interneuron marker, 1:1000, Millipore, #MAB1572) or rabbit anti-Kv4.3 (1:100, Alomone labs)/mouse anti-calbindin D-28k antisera (CB, a granule cell marker, 1:200, SWANT, #300) in PBS containing 0.3% triton X-100. After washing in PBS, sections were incubated for 1 h in a FITC- or Cy3-conjugated secondary antiserum. For nuclei counterstaining, Vectashield mounting medium with DAPI (Vector, Burlingame, CA, USA) was used as a mountant. The antibody that was preincubated with 1 μg of purified peptide (for TRPC6 and Kv4.3) or mouse pre-immune serum (Sigma, #M5905; for other antibodies) was used as for negative control. As the result of negative control test, no immunoreactive structure was observed. Images were captured using an AxioImage M2 microscope or a confocal laser-scanning microscope (LSM 710, Carl Zeiss Inc., Oberkocken, Germany).

A single data point obtained from each animal was used for analysis. All parameters were tested for normal distribution and homogeneity of variances. For variables that fulfilled both assumptions, statistical differences were determined using Mann–Whitney test or ANOVA to determine statistical significance. Bonferroni’s test was used for post hoc comparisons. Values are presented as mean ± standard error of the mean (SEM). A p-value below 0.05 was considered statistically significant.

Figure 1 shows that TRPC6 siRNA resulted in an approximate 70% reduction of TRPC6 protein level and ERK1/2 phosphorylation in the total extract obtained from hippocampus, as compared with control (non-targeting) siRNA (p < 0.05, respectively; Figures 1A,B and Supplementary Figure S1), although it did not affect Kv4.3, GAD67 and ERK1/2 expression levels. Consistent with our previous studies (Kim et al., 2013; Kim and Kang, 2015; Ko and Kang, 2017), TRPC6 expression was predominantly observed in DGC and the molecular layer of the dentate gyrus (Figure 1C). TRPC6 knockdown decreased TRPC6 expression in the hippocampus (p < 0.05 vs. control siRNA, Figure 1C). Furthermore, TRPC6 knockdown reduced the membrane localization of Kv4.3, but increased the cytosolic Kv4.3 intensity (p < 0.05 vs. control siRNA, Figures 1D–F and Supplementary Figure S1). TRPC6 expression was also detected in GABAergic (GAD67-positive) interneurons in the hilus of the dentate gyrus and the CA1 region (Figure 1G). TRPC6 siRNA effectively reduced TRPC6 expression in DGC and interneurons, although it could not influence on GAD67 expression (Figure 1G). Furthermore, TRPC6 knockdown reduced Kv4.3 clusters in the dendrites of CB-positive DGC and PV-positive interneurons (Figures 2A–E). Taken together, our findings indicate TRPC6 knockdown may abolish the dendritic and membrane Kv4.3 localization of in DGC and interneurons.

To determine whether TRPC6 is involved in neuronal excitability, we tested the effect of TRPC6 siRNA infusion on the evoked responses of the DG and CA1 region in vivo. TRPC6 knockdown did not affect the IO curve in the DG (Figure 3B), indicating no effect of TRPC6 knockdown on the basal neurotransmission. Consistent with our previous study (Kim and Kang, 2015), TRPC6 knockdown increased the DG excitability ratio and a paired-pulse inhibition at 30-ms interstimulus interval (p < 0.05 vs. control siRNA, Figures 3C–E). In contrast to the DG, the perforant path stimulation evoked only fEPSP in the CA1 region. TRPC6 knockdown reduced fEPSP amplitude ratio in the CA1 region (p < 0.05 vs. control siRNA, Figures 3F,G). Excitability ratio is an index of synaptic efficacy, DGC excitability and the threshold of epileptiform discharges. In addition, paired-pulse response represents functional changes of GABAergic inhibition (Kim and Kang, 2015). Therefore, our findings indicate that the silencing of TRPC6 increases excitability and inhibitory transmission in the DGC and the CA1 region.

In the present study, TRPC6 siRNA reduced Kv4.3 clusters in dendrites of DGC and interneurons (Figure 2). Thus, it is likely that TRPC6 may play a critical role in subcellular Kv4.3 localization. Since Kv4.3 promotes stronger inhibitory control of firing during sustained activity (Bourdeau et al., 2011), we investigated in more detail the effect of TRPC6 knockdown on the network behaviors of the DG and the CA1 region following repetitive stimulus train (300 stimuli in 9 S) of the perforant path. Control siRNA-infused animals showed steady fEPSP depression in response to repetitive stimulation of perforant path (Figures 4A,A1,A2,C). However, TRPC6 siRNA-infused animals showed population spike generation during repetitive stimulation (p < 0.05 vs. control siRNA; Figures 4B,B1,B2,C). No stimulus-induced afterdischarge was observed in both groups (data not shown).

Following high-frequency stimulation (HFS), control siRNA-infused animals showed the increases in excitability ratio (1.34-fold of basal level) and paired-pulse inhibition (0.64-fold of basal level) in the DG (p < 0.05 vs. basal level, Figures 4D,F,G). TRPC6 siRNA-infused animals revealed the elevated population spike amplitude ratio (1.61-fold of basal level) and resulted in multiple population spikes in the second pulse without changed excitability ratio (p < 0.05 vs. control siRNA; Figures 4E–G).

In the CA1 region, HFS provoked a population spike in TRPC6-infused animals (p < 0.05 vs. control siRNA; Figures 5A,A1,B,B1,C). Following HFS, fEPSP amplitude ratio in response to paired-pulse stimulation was unaltered in control siRNA-infused animals, but increased in TRPC6-infused animals (p < 0.05 vs. control siRNA; Figures 5D–F). Furthermore, TRPC6 siRNA-infused animals showed a population spike in the CA1 region in response to the second perforant path stimulus (Figures 5D–F). Since population spikes are invariably coupled to synchronous population bursts of GABAergic interneurons (Ylinen et al., 1995), our findings indicate that TRPC6 knockdown-mediated Kv4.3 translocation may decrease GABAergic inhibition in response to sustained repetitive stimuli.

Similar to TRPC6 knockdown in the present study, 4-AP (an A-type K⁺ channel inhibitor) increases field potentials, accompanied by synchronously giant inhibitory postsynaptic potential in DGC and CA3 pyramidal cells (Müller and Misgeld, 1990, 1991; Otis and Mody, 1992). Thus, it is likely that TRPC6 knockdown would influence on the responsiveness to 4-AP, if altered Kv4.3 translocation induced by TRPC6 knockdown affected the electrophysiological properties of the DG and the CA1 region. To elucidate this hypothesis, we investigated the effect of TRPC6 knockdown on the responsiveness to 4-AP (20 μM). In control siRNA-infused animals, 4-AP significantly elevated total power as compared to basal level. However, the 4-AP infusion was less effective in TRPC6 siRNA-infused animals (p < 0.05 vs. control siRNA, Figures 6A,B). With respect to reduced efficacy of 4-AP in TRPC6 knockdown animals in the present study, our findings provide the possibility that TRPC6 may play important roles in the regulation of subcellular Kv4.3 localization.

Next, we investigated how TRPC6 knockdown regulates subcellular Kv4.3 localization. Consistent with our previous study (Ko and Kang, 2017), the present study revealed that TRPC6 siRNA inhibits ERK1/2 phosphorylation in the hippocampus (Figures 1A,B). Furthermore, ERK1/2 plays a key role in the subcellular Kv4.3 localization (Setién et al., 2013). Therefore, we explored whether TRPC6 knockdown-mediated ERK1/2 inhibition is involved in the alterations in neuronal excitability, paired-pulse responses and subcellular Kv4.3 localization in the hippocampus. U0126, an ERK1/2 inhibitor, reduced ERK1/2 phosphorylation, but not expressions of ERK1/2, TRPC6 and Kv4.3 (Figures 7A,B and Supplementary Figure S2). U0126 decreased Kv4.3 clusters in the dendrites of interneurons (Figure 7C). U0126 also reduced Kv4.3 intensity in the membrane fraction, but increased it in the cytosolic fraction (p < 0.05 vs. vehicle, Figures 7D–F and Supplementary Figure S2). U0126 did not affect population spike amplitude ratio, but decreased excitability ratio. However, U0126 generated noteworthy multiple population spikes in both the first and the second pulse in the DG. U0126 did not affect paired-pulse responses in the CA1 regions of both groups (Figures 7F–K). Since the reduced excitability ratio represent the decreases in synaptic efficacy and neuronal excitability (Kim and Kang, 2015), our findings reveal that U0126 may diminish synaptic efficacy and neuronal excitability of DGC. However, multiple population spikes demonstrate the synchrony and the “more excitable” properties of DG (Uruno et al., 1994). Therefore, our findings indicate that U0126 may generate synchronous population bursts in the DGC via diminishing membrane Kv4.3 translocation.

To detail the role of TRPC6-mediated ERK1/2 activation in neuronal excitability, we applied co-treatment of TRPC6 siRNA and U0126, and compared its effect to those of TRPC6 and U0126. Similar to TRPC6 siRNA, co-treatment of TRPC6 siRNA and U0126 significantly reduced TRPC6 expression (Figures 8A,B and Supplementary Figure S3). The effect of TRPC6 siRNA on ERK1/2 phosphorylation, Kv4.3 and subcellular Kv4.3 localization was similar to that of U0126 (Figures 8A–E and Supplementary Figure S3). Co-treatment of TRPC6 siRNA and U0126 decreased ERK1/2 phosphorylation and membrane Kv4.3 localization more than TRPC6 and U0126 (p < 0.05 vs. TRPC6 and U0126; Figures 8A–E and Supplementary Figure S3). Co-treatment slightly increased Kv4.3 level in cytosolic fraction, but it was not statistically significant (Figures 8C–E and Supplementary Figure S3). As compared to TRPC6 knockdown or U0126, co-treatment of TRPC6 and U0126 more increased excitability ratio and generated more multiple population spikes, accompanied by enhanced paired-pulse inhibition (p < 0.05 vs. TRPC6 and U0126; Figures 8F–H). Taken together, our findings suggest TRPC6-mediated ERK1/2 activation may inhibit synchronous population bursts in the DGC via regulating membrane Kv4.3 localization.

The remaining question is whether TRPC6 activation directly increases membrane Kv4.3 localization via enhanced ERK1/2 activity. Since hyperforin, a TRPC6 activator (Leuner et al., 2007), increases ERK1/2 activity in hippocampal neurons (Heiser et al., 2013), we explored the effects of hyperforin on evoked potentials and subcellular Kv4.3 localization. Consistent with a previous study (Heiser et al., 2013), hyperforin increased ERK1/2 phosphorylation (p < 0.05 vs. vehicle; Figures 9A,B and Supplementary Figure S4) without changed TRPC6 and Kv4.3 expressions. Furthermore, hyperforin elevated the dendritic and membrane Kv4.3 localizations (p < 0.05 vs. vehicle; Figures 9C–F and Supplementary Figure S4). Hyperforin also reduced excitability ratio due to the decreased population spike amplitudes (p < 0.05 vs. vehicle; Figures 9G–I), while it did not affect paired-pulse inhibitions. These findings suggest that TRPC6-mediated ERK1/2 activation may play an important role in the maintenances of neuronal excitability and membrane Kv4.3 localization in the DG.