Abstract
Kv4.2 subunits, which mediate transient A-type K+ current, are crucial in regulating neuronal excitability and synaptic responses within the hippocampus. While their contribution to activity-dependent regulation of synaptic response is well-established, the impact of Kv4.2 on basal synaptic strength remains elusive. To address this gap, we introduced a Kv4.2-specific antibody (anti-Kv4.2) into hippocampal neurons of mice of both sexes to selectively inhibit postsynaptic Kv4.2, enabling direct examination of its impact on excitatory postsynaptic potentials (EPSPs) and currents (EPSCs) during basal synaptic activity. Our results demonstrated that blocking Kv4.2 significantly enhanced the amplitude of EPSPs. This amplification was proportional to the increase in the amplitude of EPSCs, which, in turn, correlated with the expression level of Kv4.2 in the dendritic regions of the hippocampus. Furthermore, the anti-Kv4.2–induced increase in EPSC amplitude was associated with a decrease in the failure rate of EPSCs evoked by minimal stimulation, suggesting that blocking Kv4.2 facilitates the recruitment of AMPA receptors to both silent and functional synapses to enhance synaptic efficacy. The anti-Kv4.2–induced synaptic potentiation was effectively abolished by intracellular 10 mM BAPTA or by blocking R-type calcium channels (RTCCs) and downstream signaling molecules, including protein kinases A and C. Importantly, Kv4.2 inhibition did not occlude further synaptic potentiation induced by high-frequency stimulation, suggesting that anti-Kv4.2–induced synaptic strengthening involves unique mechanisms that are distinct from long-term potentiation pathways. Taken together, these findings underscore the essential role of Kv4.2 in the regulation of basal synaptic strength, which is mediated by the inhibition of RTCCs.
Significance Statement
Synaptic transmission is mediated primarily by AMPA receptors (AMPARs), and there has been considerable interest in elucidating the mechanisms underlying their recruitment during activity-dependent synaptic strengthening. However, the mechanism by which basal synaptic strength is regulated remains elusive. Here, we show that blocking postsynaptic Kv4.2 enhances AMPAR-mediated currents in hippocampal neurons and that this enhancement is mediated by the signaling mechanisms involving R-type Ca2+ channels, protein kinases A and C. Importantly, Kv4.2 inhibition did not occlude activity-dependent synaptic potentiation, suggesting its specific influence in regulating synaptic AMPARs under basal conditions. Thus, our study highlights the critical function of Kv4.2 in regulating Ca2+ signaling at subthreshold potentials, thereby regulating basal synaptic strength.
Introduction
Voltage-gated K+ channel subunit Kv4.2 is a key mediator of transient A-type K+ currents and plays a crucial role in various aspects of neuronal excitability including the regulation of the resting membrane potentials (RMPs), action potentials (APs; Kim et al., 2005), and synaptic responses (Hoffman et al., 1997; Oulé et al., 2021). The abundant expression of Kv4.2 in dendrites is essential for limiting both the amplitude and propagation of dendritic spikes or back-propagating action potentials (b-APs) in the hippocampus (Hoffman et al., 1997; Migliore et al., 1999; Cai et al., 2004; Yang et al., 2015). Furthermore, Kv4.2 participates in synaptic plasticity by raising the threshold for long-term potentiation (LTP) induction (Ramakers and Storm, 2002; Watanabe et al., 2002; Chen et al., 2006; Kim et al., 2007; Zhao et al., 2011). Extensive studies have revealed that activity-dependent internalization of Kv4.2 through NMDA receptor-mediated Ca2+ signaling is pivotal for boosting synaptic responses during LTP (Frick et al., 2004; Chen et al., 2006; Lei et al., 2008; Kim and Hoffman, 2008; Jung and Hoffman, 2009; Granger and Nicoll, 2014). Despite extensive study, the specific function of Kv4.2 at subthreshold voltage ranges in regulating synaptic responses during basal synaptic transmission remains to be fully understood.
Excitatory synaptic transmission is primarily mediated by AMPA receptors (AMPARs) with their density at the postsynaptic membranes being a critical determinant of synaptic strength. AMPAR density is regulated in an activity-dependent manner, as AMPAR trafficking to synaptic sites is mediated by NMDA receptor-dependent Ca2+ signaling, a process central to the mechanism of LTP (Luscher and Malenka, 2012; Bliss and Collingridge, 2013; Hell, 2023). However, regulation of AMPAR density at basal synaptic activity is not well understood. A previous study observed that inhibition of Kv4.2-mediated currents, either through the dominant negative form of Kv4.2 or via pharmacological blockade, led to an increase in the amplitude of miniature excitatory postsynaptic currents (mEPSCs; Kim et al., 2007). This result was successfully reproduced by a more recent study (Murphy et al., 2022), which also showed the regulation of Kv4.2 by R-type Ca2+ channels (RTCCs). These findings suggest a potential role for Kv4.2 in the regulation of postsynaptic AMPARs during basal synaptic activity, yet it has not been tested. Considering that the increased mEPSC amplitude is attributable to an increase in AMPARs at postsynaptic sites or an increase in the quantal size of synaptic vesicles at presynaptic sites, it is essential to employ strategies that can separate presynaptic from postsynaptic effects of Kv4.2 to elucidate its regulatory mechanisms in synaptic transmission.
To investigate the role of Kv4.2 in synaptic transmission without presynaptic influence, we introduced a Kv4.2-specific antibody (anti-Kv4.2) directly into the cells using a whole-cell patch technique. We found that anti-Kv4.2–induced synaptic potentiation was attributed to an increase in AMPAR-mediated currents, which required calcium influx through RTCCs and the activation of protein kinase involving protein kinases A and C in the lateral perforant pathway (LPP) to GC synapses. Additionally, our observations suggested that Kv4.2 channels may act as an input-specific synaptic regulator depending on their expression in the hippocampus. Collectively, our results position Kv4.2 as a key regulator of synaptic strength at basal activity, potentially contributing to the maintenance of the AMPAR density within the physiological range by suppressing RTCC-dependent Ca2+ signaling.
Materials and Methods
Animals
All experiments were performed using C57BL/6 mice of both sexes. Animals were housed in five mice per cage and maintained under specific pathogen-free (SPF) conditions with food and water freely available. The animal maintenance protocols and all experimental procedures were approved by the Institutional Animal Care and Use Committee at Seoul National University (Approval #: SNU-220407-1-7).
Slice preparation and electrophysiology
Hippocampal slices were obtained from 6- to 8-week-old C57BL/6 mice of both male and female mice. Mice were decapitated rapidly after being fully anesthetized by inhalation with isoflurane (Forane; Abbott), and the whole brain was immediately removed from the skull, and submerged in iced-cold artificial cerebrospinal fluid (aCSF) at 4°C containing the following (in mM): 110 choline chloride, 25 NaHCO3, 2.5 KCl, 1.25 NaH2PO4, 0.5 CaCl2, 7 MgCl2, 10 glucose, 1 Na-pyruvate, and 0.57 ascorbate consistently saturating with carbogen (95% O2 and 5% CO2).
Transverse hippocampal slices (∼300 µm thick) were prepared using a vibratome (VT1200S, Leica Microsystems). Slices were incubated at 36°C for 30 min, and the slices were subsequently maintained at room temperature until the recordings in aCSF containing the following (in mM): 125 NaCl, 25 NaHCO3, 2.5 KCl, 1.25 NaH2PO4, 2 CaCl2, 1 MgCl2, and 10 glucose, consistently bubbled with carbogen (95% O2 and 5% CO2).
Hippocampal neurons were visualized using an upright microscope equipped with differential interference contrast optics (BX51WI, Olympus). Experiments were performed on principal neurons in the hippocampus, including mature granule cells and CA3 and CA1 pyramidal neurons. In DG, mature granule cells (GCs), characterized by their location in the outer granule cell layer and their low input resistance (Rin < 200 MΩ), were selected for observations to avoid potential confounding factors related to the maturation stages (Kerloch et al., 2019). Voltage- and current-clamp recordings were made by the whole-cell patch-clamp technique with EPC 10 USB Double amplitude (HEKA Elektronik) at a sampling rate of 10 or 20 kHz. After membrane break-in, 2 min was given to stabilize the neurons.
Patch pipettes (5–7 MΩ) and monopolar stimulator pipettes (3–4 MΩ) were obtained from borosilicate glass capillaries with a horizontal pipette puller (P-97, Sutter Instrument). The internal pipette solution contained the following (in mM): 143 K-gluconate, 7 KCl, 15 HEPES, 4 MgATP, 0.3 NaGTP, 4 Na-ascorbate, and 0.1 EGTA, with pH adjusted to 7.3 with KOH and with an osmolality of approximately 300 mOsml/L. For IPSP recording, chloride concentration was increased to 20 mM by changing K-gluconate and KCl concentrations to 135 and 20 mM, respectively. For the antibody-blocking experiments, the Kv4.2 antibodies or heat-inactivated Kv4.2 antibodies were included (1 µg/ml) in the intracellular solution. To inactivate the antibodies, they were heated at 100°C for approximately 30 min using a heat block and then included in the pipette solution. Series resistance (Rs) after establishing whole-cell configuration was between 10 and 20 MΩ. Rs was monitored by applying a short (1 s) hyperpolarization (1 mV) pulse during the recording. If Rs of the recorded cells changed 20% of the initial value, the cells were discarded and excluded from further analysis.
For voltage-clamp experiments, all recordings were performed at a holding potential of −70 mV unless otherwise indicated. To measure outward K+ currents, TTX (0.5 µM), CdCl2 (300 µM), and NiCl2 (500 µM) were additionally applied to block Na+ and Ca2+ channels. Outward K+ currents were evoked by injecting a +40 mV or +30 mV pulse for 500 ms or 1 s durations.
For EPSC recordings, the stimulator intensity (100 µs pulse width; 8–18 V) of extracellular stimulation was adjusted to evoke EPSC amplitudes between 50 and 150 pA for the baseline delivered every 10 or 30 s between sweeps. For EPSP recordings, the cells were held at their RMP unless otherwise indicated. A stimulator (A360LA Stimulus Isolator, World Precision Instruments) connected to a monopolar electrode filled with recording aCSF was placed in the outer molecular layer (OML) of the DG to evoke LPP stimulation-induced synaptic responses in the mature granule cells of the hippocampus. Stimulation intensity in voltage-clamp and current-clamp experiments were comparable. Synaptic responses at other synapses are otherwise indicated. Most of the experiments were performed in the presence of GABA blockers (100 µM PTX, 1 µM CGP52432) to block inhibitory synaptic transmission. EPSP and EPSC decay time constants were estimated from a single exponential fit with Igor Pro. All rise times are reported as 10–90% of the peak EPSC amplitude.
For minimal stimulation, stimulation intensity was adjusted to evoke a failure rate of approximately 50% as previously established (Min et al., 1998a,b; Hashimotodani et al., 2017). The miniature EPSCs (mEPSCs) were recorded in the presence of 0.5 µM TTX and GABA blockers. Events exceeding 6–7 pA within a specified interval of three to four digitized points (0.5–0.8 ms) that showed a single exponential decay time course were identified as mEPSCs.
NMDAR-EPSCs were measured in Mg2+-free solution with the presence of AMPA (10 µM CNQX) and GABA blockers (100 µM PTX, 1 µM CGP52432). IPSP amplitudes were measured in the presence of AMPA (10 µM CNQX) and NMDA (50 µM APV) blockers.
The LTP was induced by applying 10 bouts of high-frequency stimulation of perforant path synapses or Schaffer collateral synapses at 5 Hz. Each bout consists of 10 stimuli at 100 Hz under the current-clamp mode. The stimulation intensity was adjusted to evoke a minimum of two to three APs at the first bout for GCs and five APs for CA1 PN. For LTP experiments, baseline EPSCs were measured for 5 min before LTP induction, after stabilization of the anti-Kv4.2 effect, and EPSCs were monitored at least 15 min' post LTP induction to observe the increased magnitude.
Intrinsic properties under current-clamp mode and the following parameters were obtained at 3 min (denoted as T1) and 20 min (denoted as T2) after patch break-in: (1) RMP, (2) input resistance [membrane potential changes at a given hyperpolarizing current input (−30 pA, 500 ms)], (3) F–I curve [firing frequencies (F) against the amplitude of injected currents from 50 to 300 or 400 pA for 500 ms duration with 50 pA increment (I)], (4) AP half-width (measured as the width at half of the spike peak amplitude), and (5) AP threshold (the voltage at which dV/dt exceeded 40 mV/ms). The intrinsic properties were measured either at its RMP or at −90 mV by providing hyperpolarizing current.
Immunohistochemistry analysis
For the immunofluorescence staining, 7-week-old C57BL/6 mice were anesthetized with isoflurane and perfused transcardially with 1× PBS and 4% paraformaldehyde (PFA) for 10–15 min. Brains were removed and cut into 30-µm-thick sections using a vibratome (VT1200S, Leica) and postfixed overnight at 4°C in 4% PFA solution. The slices were washed five times in 1× PBS with 0.3% Triton X-100 (PBS-T) for 5 min. Then the sections were incubated three times in a blocking solution (2.5% donkey serum plus 2.5% goat serum in PBS-T) for 1 h at room temperature. The sections were then incubated overnight at 4°C in blocking solutions containing primary antibodies (1:500 Kv4.2 antibody, APC-023, Alomone Labs). After washing five times in 0.3% PBS-T for 5 min, sections were then incubated with secondary antibodies in blocking solution for at least 1 h at RT. After rinsing with PBS, sections were mounted on glass slides using mounting medium containing DAPI. The immunostained sections were imaged with a confocal laser scanning microscope (FV1200, Olympus) using a 40× oil-immersion objective lens.
Drugs
dʟ-APV (d,ʟ-2-amino-5-phosphonovaleric acid), PTX (picrotoxin), NASPM, and calmidazolium were purchased from Tocris Bioscience. We used Kv4.2 antibody (#APC-023, Alomone Labs) known to bind intracellular C domain of Kv4.2 where phosphorylation sites involved in channel trafficking and gating are located (Anderson et al., 2010). CGP62432 was purchased from Abcam. All other drugs were from Sigma-Aldrich.
Data analysis
All data were presented as mean ± standard error of the mean (SEM). Statistical analysis was performed using Igor Pro (Version 7.08, WaveMetrics) and Prism. Mann–Whitney U test and Wilcoxon signed rank test were used to assess between-group and within-group differences, respectively. One-way ANOVA with Tukey’s post hoc test was also conducted for comparisons between multiple groups. p-values of <0.05 were considered statistically significant. Fluorescence intensity was analyzed using ImageJ.
Results
Regulation of intrinsic properties by Kv4.2-mediated transient K+ current
To investigate the role of Kv4.2 in the hippocampus, Kv4.2 antibodies (anti-Kv4.2) were introduced into mature granule cells (GCs) in the dentate gyrus (DG) through intracellular pipette solution at a concentration of 1 µg/ml, according to our previously established protocol (Kim et al., 2020). This method has also been used in previous studies examining ion channel functions in the dendrites and synapses (Wang et al., 2014; Hikima et al., 2021, 2022), where the distribution of antibodies within dendrites has been demonstrated by post hoc immunostaining. Dialysis of anti-Kv4.2 resulted in a significant reduction in the amplitude of the peak outward current [Ipeak; 4.8 ± 0.18 vs 3.4 ± 0.23 (nA), n = 9, p = 0.003, Wilcoxon signed rank test], without affecting the amplitude of sustained outward current (Isus; Fig. 1A–C). However, no change was observed with heat-inactivated Kv4.2 antibodies (inactive Ab; Fig. 1D–F). The difference between Ipeak and Isus may represent transient outward current (ITO) mediated by Kv4.2 (Rhodes et al., 2004), while Isus may contain delayed rectifier K+ currents, nonselective or leak currents. We observed that Ipeak reduction by anti-Kv4.2 reached a steady state within 15 min, which was interpreted as the time required for anti-Kv4.2 to diffuse throughout the cell (Fig. 1A). We regarded the initial response (Iinitial) at 3 min after patch break-in as control (designated as T1) and the response at 20 min as under Kv4.2 inhibition (designated as T2). A comparison of the subtracted currents (T1–T2) between anti-Kv4.2 and heat-inactivated Ab further reinforced the hypothesis that ITO was selectively reduced by active anti-Kv4.2, as evidenced by the larger amplitude of Ipeak being reduced in comparison with that of heat-inactivated Ab [Fig. 1G; anti-Kv4.2 vs inactive Ab (nA): 2.0 ± 0.16 (n = 9) vs 0.6 ± 0.09 (n = 5), p = 0.007, Mann–Whitney U test]. The specificity of anti-Kv4.2 in reducing Kv4.2-mediated ITO was further validated in HEK 293 cells expressing Kv4.2 (Extended Data Fig. 1-1).
Impact of anti-Kv4.2 on intrinsic properties of dentate GCs. A, D, Left, Top, A schematic image of the experimental setup showing dialysis of anti-Kv4.2 (A, red) or inactive Ab (D, gray) in GCs via recording pipette. Left, Bottom, Average time courses showing normalized Ipeak (circle) and Isus (triangle) amplitudes. Right, Top, Representative traces of outward K+ currents at 3 min (T1) and 20 min (T2) after patch break-in with either anti-Kv4.2 (A) or inactive Ab (D). Right, Bottom, Overlaid (left) and subtracted current traces (right) showing selective reduction of ITO by anti-Kv4.2 (A) but not by inactive Ab (D). Scale bars: 1 nA and 200 ms for the main traces; 1 nA and 10 ms for the superimposed traces; 500 pA and 10 ms for the subtracted current. B, E, Bar graphs of the average values of outward K+ transient (Ipeak) or sustained (Isus) current amplitudes with either anti-Kv4.2 (B) or inactive Ab (E). C, F, Bar graphs of the average values of the normalized outward K+ transient (Ipeak) or sustained (Isus) current amplitudes with either anti-Kv4.2 (C; Ipeak, 0.7 ± 0.03, n = 9, p = 0.004; Isus, 1.0 ± 0.05, n = 9, p = 0.65; Wilcoxon signed rank test) or inactive Ab (F; Ipeak, 0.9 ± 0.04, n = 5, p = 0.25; Isus, 1.06 ± 0.05, n = 5, p = 0.61; Wilcoxon signed rank test). G, Top, Superimposed subtracted currents (T1–T2) for anti-Kv4.2 (red) and inactive Ab (gray). Scale bars: 500 pA and 10 ms. Bottom, Bar graph comparing the average values of ITO amplitude with anti-Kv4.2 and inactive Ab (p = 0.007, Mann–Whitney U test). (see also Extended Data Fig. 1-1 for validation of anti-Kv4.2 effect using HEK293 cells). H, Bar graph of the average values of RMP at T1 and T2 with either anti-Kv4.2 or inactive Ab [T1 vs T2: −76.3 ± 2.93 vs −74.8 ± 0.88 (mV), n = 5, p = 0.63, Wilcoxon signed rank test for inactive Ab]. I, Left, Representative traces of voltage responses to a negative current pulse injection at T1 and T2 with either anti-Kv4.2 (top) or inactive Ab (bottom). Scale bars: 2 mV and 50 ms. Right, Bar graph of the average values of input resistance at T1 and T2 with anti-Kv4.2 or inactive Ab [T1 vs T2: 202.1 ± 10.03 vs 202.79 ± 5.15 (MΩ), n = 5, p = 0.63, Wilcoxon signed rank test for inactive Ab]. J, Left, Graph of the average values of the firing frequency–current curve. Right, Top, Representative traces of AP spikes at RMP at T1 and T2 with anti-Kv4.2. Scale bars: 50 mV and 200 ms. Right, Bottom, Superimposed AP waveform marked with asterisks. Scale bars: 50 mV and 200 ms (top) and 20 mV and 1 ms (bottom). K–M, Bar graphs showing the effect of anti-Kv4.2 on AP parameters including AP threshold (K), AP half-width (L), and AP height (M). N, Left, A schematic image of the experimental setup showing dialysis of anti-Kv4.2 in GCs at a holding potential of −90 mV. Right, Bar graph of the average values of membrane potentials [T1 vs T2: −91.2 ± 0.39 vs −89.8 ± 0.67 (mV), n = 6, p = 0.16, Wilcoxon signed rank test] and input resistance [T1 vs T2: 110.6 ± 4.82 vs 118.3 ± 9.56 (MΩ), n = 6, p = 0.31, Wilcoxon signed rank test] at T1 and T2 with anti-Kv4.2. O, Left, Graph of the average values of the firing frequency–current curve at −90 mV. Right, Representative traces of AP spikes at T1 and T2 with anti-Kv4.2 by holding GCs at −90 mV. Scale bars: 50 pA and 200 ms. Statistical significance was evaluated by the Wilcoxon signed rank test and Mann–Whitney U test. **p < 0.01, ***p < 0.001, ****p < 0.0001 for Wilcoxon signed rank test, ##p < 0.01 for Mann–Whitney U test, N.S., not significant.
Figure 1-1
(related to Figure 1). Efficacy of anti-Kv4.2 on K+ current in HEK 293 cells A. (Left) A schematic image of experimental setup showing dialysis of anti-Kv4.2 in either Kv4.2 (Top) or Kv4.1 (Bottom) expressing HEK 293 cells via recording pipette. (Middle) Representative traces of K+ current with anti-Kv4.2 evoked by depolarizing pulse to + 40 mV in Kv4.2 (Top) or Kv4.1 (Bottom) expressing HEK 293 cell. (Right) Superimposed current traces at time points indicated by 1 and 2 in the time course. (Inset) Selective reduction of Ipeak by anti-Kv4.2 in Kv4.2 expressing HEK 293 cell (Top) but not Kv4.1 expressing HEK 293 cell (Bottom). (Right). Difference current trace obtained by subtracting trace 2 from 1. Scale bar, 2 nA and 100 ms for main traces; 2 nA and 5 ms for inset; 500 pA and 100 ms for subtracted current. B. Comparison of changes in normalized peak current over time in Kv4.1 expressing cell (n = 6) and Kv4.2 expressing cell (n = 5) by anti-Kv4.2. Download Figure 1-1, TIF file.
Following the dialysis of anti-Kv4.2, there was a significant depolarization of the resting membrane potential [RMP; Fig. 1H; −76.2 ± 0.73 vs −68.9 ± 0.92 (mV), n = 82, p < 0.0001, Wilcoxon signed rank test] and increase in input resistance [Fig. 1I; 220.8 ± 7.70 vs 255.0 ± 7.72 (MΩ), n = 64, p < 0.0001, Wilcoxon signed rank test]. These changes were not observed with inactive Ab (Fig. 1H,I). The number of action potentials (APs) was significantly increased by anti-Kv4.2 (Fig. 1J), with no significant changes in AP threshold, AP half-width, or AP height (Fig. 1K–M), suggesting that Kv4.2-mediated ITO contributes to the firing frequency of GCs. Consistently, anti-Kv4.2 did not affect input resistance, membrane potential, and firing frequency when the membrane potential was held at −90 mV where ITO activity would be negligible (Fig. 1N,O). These results support that increased firing rate by anti-Kv4.2 is attributable to RMP depolarization associated with increased input resistance induced by inhibiting Kv4.2-mediated ITO.
The increase in EPSP by anti-Kv4.2 is attributed to the increase in EPSC
To assess the functional significance of Kv4.2 in synaptic responses, we measured EPSPs in GCs by stimulating lateral perforant pathways (LPP) in the outer molecular layer (OML) at 10 s intervals (Fig. 2A). Intracellular dialysis of anti-Kv4.2 gradually and significantly increased the amplitude of EPSPs at LPP-GC synapses. We usually started recording 3 min after patch break-in and changes in EPSPs by anti-Kv4.2 dialysis reached steady state at approximately 20 min that persisted throughout the 25 min of recording duration (Fig. 2A). To evaluate the effect of anti-Kv4.2 on synaptic responses within the same cell, we considered the mean amplitude of synaptic responses obtained at 3–5 min (T1) as control and those at 23–25 min (T2) as Kv4.2-inhibited response. The mean amplitude of EPSPs at T2 was significantly larger than that at T1 [Fig. 2B; 5.7 ± 0.20 vs 9.2 ± 0.34 (mV), n = 77, p < 0.0001, Wilcoxon signed rank test]. In the presence of heat-inactivated Kv4.2 antibodies (inactive Ab, 1 µg/ml), EPSP amplitude remained unchanged during recording time [Fig. 2A–C; anti-Kv4.2 vs inactive Ab, p = 0.0003, Mann–Whitney U test], underscoring the contribution of Kv4.2 to synaptic responses at LPP-GC synapses during basal synaptic activity. We confirmed that enhanced synaptic responses by anti-Kv4.2 are specific to excitatory synapses by showing that inhibitory postsynaptic potentials remained unchanged by anti-Kv4.2 (Extended Data Fig. 2-1).
Kv4.2 inhibition enhanced synaptic transmission at LPP-GC synapses. A, D, Left, Top, A schematic image of the experimental setup showing dialysis of anti-Kv4.2 or inactive Ab in GCs via recording pipette, with stimulation delivered to the outer molecular layer (OML) of the DG. Bottom, Average time courses of normalized EPSP (A) and EPSC (D) amplitudes with anti-Kv4.2 (red), inactive Ab (gray), or no Ab (gray diamond) at LPP-GC synapses. Synaptic responses were normalized to mean synaptic response amplitudes at 3–5 min post patch break-in (T1, shaded gray). The dashed line indicates baseline amplitudes. In each time course, evoked amplitudes were averaged into 1 min bins. Middle, Representative traces of EPSPs (A) and EPSCs (D) at T1 and T2 (23–25 min post patch break-in, shaded red) for anti-Kv4.2 (top) or inactive Ab (bottom). The bold line represents the average synaptic response at T1 or T2, while the shaded lines represent individual synaptic responses. Right, Superimposed traces of average EPSPs (A) and EPSCs (D) showing selective increase in synaptic responses only by anti-Kv4.2 (top), not by inactive Ab (bottom). Scale bars: A, 5 mV and 10 ms; D, 50 pA and 10 ms. B, C, Bar graph illustrating changes in average values EPSP (B) and their fold changes by either anti-Kv4.2 or inactive Ab (anti-Kv4.2 vs inactive Ab: 1.7 ± 0.06 vs 0.8 ± 0.09, p = 0.0003, Mann–Whitney U test). E, Bar graph showing that initial EPSC amplitudes at 3–5 min (denoted as T1) were consistent across different experimental conditions (F(2,84) = 0.10, p = 0.91, one-way ANOVA). At 23–25 min (denoted as T2), significant alterations in EPSC amplitudes were observed only by anti-Kv4.2, indicating a specific effect of active antibody on EPSC amplitude (F(2,84) = 5.99, p = 0.003, one-way ANOVA). F, Bar graph comparing fold changes in EPSC amplitudes under three conditions (p = 0.0007 for anti-Kv4.2 vs inactive Ab; p = 0.0003 for anti-Kv4.2 vs no Ab; Mann–Whitney U test; see also Extended Data Fig. 2-1 for anti-Kv4.2–mediated effect on inhibitory synaptic transmission). G, Bar graph comparing anti-Kv4.2 mediated fold changes in EPSP and EPSC amplitudes. H, Fold changes in EPSCs plotted against fold change in EPSPs to show a strong linear correlation (r2 = 0.78, slope = 1.04). Fitted line of data is shown with red solid line while black dashed line indicates linear relationship between changes in EPSPs and EPSCs. Red circles represent individual cell data. I, Left, Top, Average traces of NMDAR-EPSCs at T1 and T2 with anti-Kv4.2. Scale bars: 50 pA and 10 ms. Average time course (left, bottom) and bar graph (right) of the normalized NMDA-EPSC amplitude with anti-Kv4.2 (1.1 ± 0.11, n = 8, p = 0.38, Wilcoxon signed rank test). J, Bar graph showing anti-Kv4.2–mediated effect on EPSC kinetics: rise time [middle, 2.0 ± 0.13 vs 1.9 ± 0.11 (ms), n = 15, p = 0.28, Wilcoxon signed rank test] and decay time [right, 5.4 ± 0.22 vs 5.6 ± 0.33 (ms), n = 20, p = 0.20, Wilcoxon signed rank test]. Left, Top, Superimposed traces of EPSCs at T1 (light gray) and T2 (red). Scale bars: 50 pA and 10 ms. Left, Bottom, Superimposed EPSC amplitudes scaled to the peak of T1. K, Left, Top, Superimposed traces of EPSP at T1 and T2. Left, Bottom, Superimposed EPSP amplitudes scaled to the peak of T1. Bar graph showing that anti-Kv4.2 significantly prolonged decay time (right, p = 0.0007, Wilcoxon signed rank test), not rise time [middle, 4.2 ± 0.31 vs 4.1 ± 0.29 (ms), n = 8, p = 0.14, Wilcoxon signed rank test]. Scale bars: 4 mV and 20 ms. Statistical significance was evaluated by Wilcoxon signed rank test, Mann–Whitney U test, or one-way ANOVA. *p < 0.05, *** p < 0.001, ****p < 0.0001 for Wilcoxon signed rank test, ###p < 0.001 for Mann–Whitney U test; N.S., not significant.
Figure 2-1
(related to Figure 2). Anti-Kv4.2 induced increase in EPSC is AMPAR-specific A. Effect of anti-Kv4.2 on IPSP amplitude. (Top) Average IPSP traces at T1 and T2, which was obtained at time 3-5 and 23-25 minutes after patch break-in, respectively. Scale bars, 5 mV and 10 ms. (Bottom) Average time courses of the normalized IPSP amplitude with anti-Kv4.2. In time course, evoked IPSP amplitudes were averaged into 1-minute bins. B. Bar graph illustrating anti-Kv4.2 mediated effect on IPSP amplitudes (0.9 ± 0.07, n = 8, p = 0.11, Wilcoxon signed rank test). Statistical significance was evaluated by Wilcoxon signed rank test. N.S. = not significant. Download Figure 2-1, TIF file.
The increase in synaptic responses induced by inhibition of dendritic K+ channels, typically attributed to increased dendritic excitability (Hoffman et al., 1997; Johnston et al., 2000), might also stem from an increase in synaptic current. To test this possibility, we measured EPSCs at LPP-GC synapses with anti-Kv4.2 or inactive Ab (Fig. 2D). The baseline EPSC amplitudes at T1 were not affected by the presence of anti-Kv4.2 (Fig. 2E), confirming that EPSCs at T1 in the presence of anti-Kv4.2 can be regarded as control. Surprisingly, we found that anti-Kv4.2 led to a significant increase in EPSC amplitude at T2 by 1.6-fold (Fig. 2E,F; 94.8 ± 3.32 vs 149.6 ± 6.22 (pA), n = 80, p < 0.0001, Wilcoxon signed rank test), while inactive Ab and no Ab had no significant effect (Fig. 2F; p = 0.0007 for anti-Kv4.2 vs inactive Ab; p = 0.0003 for anti-Kv4.2 vs no Ab; Mann–Whitney U test). The fold increase in EPSP amplitude by anti-Kv4.2 was significantly larger than that in EPSC amplitude, but the difference between the two was small (Fig. 2G, EPSC vs EPSP, 1.6 ± 0.05 vs 1.7 ± 0.06, n = 74, p = 0.03, Wilcoxon signed rank test). Furthermore, the fold increase in EPSP amplitude by anti-Kv4.2 was linearly correlated with that in EPSC amplitude with the slope of near unity (1.04 ± 0.23), implying that the enhanced synaptic current through AMPARs is the primary mechanism underlying the increase in EPSP responses (Fig. 2H). We found that NMDAR-mediated currents remained unchanged by anti-Kv4.2 (Fig. 2I), indicating that the anti-Kv4.2 induced increase in EPSC was attributed specifically to AMPA receptor (AMPAR)-mediated currents. This result also confirmed that enhanced synaptic response by anti-Kv4.2 introduced into postsynaptic cells was indeed mediated by postsynaptic mechanism, not by presynaptic mechanism. Additionally, the unchanged kinetics of EPSCs by anti-Kv4.2 indicate that an increase in postsynaptic AMPAR density, rather than alterations in AMPAR functional properties, accounts for the observed EPSC increases (Fig. 2J). In contrast, anti-Kv4.2 slowed down EPSP decay by 20% [Fig. 2K, 30.9 ± 1.82 vs 35.5 ± 1.56 (ms), n = 32, p = 0.0007, Wilcoxon signed rank test]. The larger effect of anti-Kv4.2 on EPSP compared with EPSC and the slowing of EPSP decay may be attributed to the increased dendritic excitability resulting from the inhibition of Kv4.2 currents. This heightened dendritic excitability may have elevated input resistance and/or NMDAR activation, thereby extending EPSP decay. Given the previous results showing that inhibition of Kv4.2 results in an increase in NMDAR activation (Murphy et al., 2022), the increased NDMAR currents may also contribute to the shaping of EPSPs by anti-Kv4.2.
Kv4.2 inhibition increases unitary synaptic currents and recruits silent synapses
To further elaborate the idea that Kv4.2 inhibition promotes the recruitment of AMPARs to synaptic sites, we examined the effects of anti-Kv4.2 on unitary synaptic transmission (Fig. 3A). Dialysis of anti-Kv4.2 resulted in a marked increase in miniature EPSC (mEPSC) amplitudes [Fig. 3B; T1 vs T2, 16.7 ± 0.45 vs 19.5 ± 0.59 (pA), n = 32, p < 0.0001, Wilcoxon signed rank test], indicating an enhanced AMPAR density in postsynaptic sites. However, the increase in mEPSC amplitude by anti-Kv4.2 was only 1.17-fold, which was not sufficient to explain the increase in EPSC amplitude (1.6-fold), suggesting the involvement of other mechanisms. We found that anti-Kv4.2 also elevated mEPSC frequency [Fig. 3C; T1 vs T2, 1.4 ± 0.13 vs 2.3 ± 0.26 (Hz), n = 33, p < 0.0001, Wilcoxon signed rank test], a parameter typically reflecting presynaptic changes. Nonetheless, postsynaptic mechanisms such as conversion of silent synapses to functional synapses by inserting AMPARs at postsynaptic sites might also increase mEPSC frequency (Isaac et al., 1995; Kerchner and Nicoll, 2008).
Kv4.2 inhibition facilitates AMPAR recruitment at synaptic sites. A, Representative traces showing mEPSC at T1 and T2 conditions to examine changes in mEPSC amplitude and frequency by anti-Kv4.2. Scale bars: 50 pA and 50 ms. B, C, Cumulative probability graph and histogram showing changes in mEPSC amplitude (B) and frequency (C) by anti-Kv4.2. Inset: Bar graph showing significant enhancements in both mEPSC amplitudes (B) and frequency (C) induced by anti-Kv4.2. D, E, Top, Representative traces of minimal stimulation-induced EPSC at indicated time points for T1 and T2 with either no Ab (D) or anti-Kv4.2 (E). Bottom, Representative time course of EPSC in response to minimal stimulation with (E, red) or without (D, black) anti-Kv4.2 in intracellular solution. Dashed line indicates 0. Opened circles indicate EPSC amplitudes while closed circle indicates EPSC failure. In each time course, minimal induced EPSCs were evoked with a 5 s sweep interval in the shaded area. Scale bars: 20 pA and 10 ms. F, Bar graph comparing baseline amplitudes (T1) for mEPSC and minimal stimulation-induced EPSCs in conditions with or without anti-Kv4.2. G, H, Bar graphs depicting changes in potency and failure rate with or without anti-Kv4.2 [G, 18.1 ± 1.19 vs 21.0 ± 1.49 (pA), n = 11, p = 0.60 for no Ab; H, 0.5 ± 0.06 vs 0.4 ± 0.06, n = 11, p = 0.78 for no Ab; Wilcoxon signed rank test]. Statistical significance was evaluated by Wilcoxon signed rank test and one-way ANOVA. ***p < 0.001, ****p < 0.0001, N.S., not significant.
To further assess this possibility, minimal stimulation experiment was performed to examine changes in evoked EPSC amplitude and failure rate. We adjusted the stimulation intensity to elicit EPSCs at a 50% failure rate, which was used in the previous studies (Min et al., 1998a,b; Hashimotodani et al., 2017). The amplitude of EPSCs elicited by minimal stimulation at T1 was not statistically different from the mEPSC amplitude (F(2,57) = 2.51, p = 0.091, one-way ANOVA), indicating that this minimal stimulation protocol mostly activates a single synapse (Fig. 3F). In the absence of anti-Kv4.2, there were no significant alterations in potency (EPSC amplitude without failure) and failure rate (Fig. 3D,G,H). Conversely, following anti-Kv4.2 dialysis, there was a noticeable increase in potency and a decrease in failure rate [Fig. 3E,G,H; potency (pA), 18.7 ± 0.85 vs 26.7 ± 1.81, n = 17, p = 0.0001; failure rate, 0.5 ± 0.06 vs 0.2 ± 0.05, n = 17, p < 0.0001; Wilcoxon signed rank test]. The failure rate can be calculated as (1 − p)N, in which p represents the release probability and N represents the number of functional synapses. Given that p is not affected by postsynaptic Kv4.2, these results suggest that Kv4.2 inhibition not only increases AMPAR density but also promotes AMPAR insertion at silent synapses, leading to an increased number of functional synapses, to strengthen basal excitatory synaptic transmission.
Kv4.2-dependent AMPAR regulation requires the R-type calcium channel
It is well-established that synaptic strengthening by AMPAR insertion at postsynaptic sites is triggered by influx of Ca2+ through NMDAR and VGCC (Malenka et al., 1988; Luscher and Malenka, 2012). We thus investigated whether Kv4.2-dependent increases in AMPARs at LPP-GC synapses are Ca2+-dependent by introducing a fast Ca2+ chelator, 10 mM BAPTA, together with anti-Kv4.2 in the intracellular solution. Coapplication of BAPTA and anti-Kv4.2 effectively abolished the increase in EPSC amplitude (Fig. 4A, p = 0.12, Wilcoxon signed rank test), while anti-Kv4.2-induced RMP depolarization remained unaffected (Fig. 4B, +BAPTA vs −BAPTA, p = 0.59, Mann–Whitney U test). These data underscore the necessity of Ca2+ influx for Kv4.2's regulatory effect on AMPARs, but not of intrinsic properties.
RTCC-mediated Ca2+ signaling is essential for the increase in EPSC amplitude by anti-Kv4.2. A, Left, Average time course of normalized EPSC amplitudes with BAPTA (10 mM) and anti-Kv4.2 in intracellular solution compared with that of anti-Kv4.2 alone (red line). Solid line indicates the presence of BAPTA. Middle, Representative EPSC traces at T1 and T2 conditions, obtained at 3–5 and 23–25 min after patch break-in, respectively. Right, Bar graph illustrating EPSC amplitude with anti-Kv4.2 and BAPTA (p = 0.12, Wilcoxon signed rank test). Scale bars: 50 pA and 10 ms. B, Bar graph comparing changes in RMP with or without BAPTA [+BAPTA vs −BAPTA: −75.5 ± 1.53 vs −68.6 ± 2.62 (mV), n = 11, p = 0.002, Wilcoxon signed rank test]. C–G, Left, Average time courses of the normalized EPSC amplitude showing effect of anti-Kv4.2 in the presence of different calcium sources represented by distinct color: Nimo (C), APV (D), RYR (E), NASPM (F), and SNX482 (G). In each time course plot, colored circles represent normalized EPSC amplitudes with inhibitors; red shaded line indicates anti-Kv4.2 alone. Right, Representative EPSC traces at T1 and T2. Scale bars: 100 pA and 10 ms (see also Extended Data Fig. 4-1 for the use of Ni2+ as RTCC blocker). H, Average time course (left), traces (middle), and bar graph (right) showing changes in EPSC amplitudes by 0.3 µM SNX482 (0.92 ± 0.07, n = 4, p = 0.63, Wilcoxon signed rank test). Scale bars: 100 pA and 10 ms. I, Bar graph summarizing anti-Kv4.2–mediated fold changes in EPSC amplitudes across different treatment conditions (+Nimo, 1.4 0.08, n= 16, p = 0.06; +APV, 1.5 ± 0.11, n = 11, p = 0.50; +RYR, 1.4 ± 0.09, n = 12, p = 0.08; +NASPM, 1.4 ± 0.14, n = 13, p = 0.20; +SNX, 1.0 ± 0.06, n = 11, p < 0.0001; +nickel, 1.1 ± 0.06, n = 13, p < 0.0001; compared with anti-Kv4.2 alone; Extended Data Fig. 4-2). In each time course, evoked EPSC amplitudes were averaged into 1 min bins. Solid line indicates the presence of each inhibitor. Statistical significance was evaluated by Wilcoxon signed rank test and Mann–Whitney U test. ####p < 0.0001 (compared with anti-Kv4.2 alone); N.S., not significant.
Figure 4-1
(related to Figure 4): Concentration dependent effect of Ni2+ on potentiating action of anti-Kv4.2 A. Average time course of normalized EPSC amplitudes with anti-Kv4.2 in the presence of either 30 µM (light green) or 100 µM Nickel (dark green). In each time course, evoked EPSC amplitudes were averaged into 1-minute bins. Solid line indicates the presence of Ni2+. B. Bar graph comparing changes in EPSC amplitudes at T1 and T2 by anti-Kv4.2 in the presence of 30 µM Ni2+ (2.0 ± 0.10, n = 5) or 100 µM Ni2+ (1.1 ± 0.06, n = 13), highlighting a concentration-dependent inhibition of the anti-Kv4.2 effect, with a notable reduction observed only at 100 µM Ni2+ (p = 0.04, Mann Whitney U test). Statistical significance was evaluated by Mann Whitney U test. #p < 0.05. Download Figure 4-1, TIF file.
Figure 4-2
(related to Figure 4): Effect of anti-Kv4.2 and SNX482 on EPSCs induced by minimal stimulation A. Representative time course showing the effect of anti-Kv4.2 on potency (open circle) and EPSC failure (closed circle) induced by minimal stimulation in LPP-GC synapses in the presence of SNX482 as illustrated by diagram in the top. Solid line indicates the presence of SNX482. B and C. Bar graph illustrating changes in potency (B, T1 vs T2 (pA): 23.2 ± 5.83 vs 22.6 ± 5.04, n = 4, p = 0.76, Wilcoxon signed rank test) and failure rate (C, T1 vs T2: 0.63 ± 0.02 vs 0.57 ± 0.16, n = 4, p = 0.76) by anti-Kv4.2 in the presence of SNX482. Statistical significance was evaluated by Wilcoxon signed rank test. N.S. = not significant. Statistical significance was evaluated by Wilcoxon signed rank test. N.S. = not significant. Download Figure 4-2, TIF file.
To identify the Ca2+ sources responsible for this regulation, we assessed the effects of anti-Kv4.2 on EPSC amplitudes after pretreating the slices with inhibitors targeting various postsynaptic Ca2+ sources, such as voltage-gated calcium channels (VGCCs), NMDA receptors, ryanodine receptors (RYR), and calcium-permeable AMPARs. Preincubation with nimodipine (10 µM, LTCC), APV (50 µM, NMDAR), ryanodine (10 µM, RYR), or NASPM (10 µM, CP-AMPAR) did not block anti-Kv4.2 induced EPSC enhancement (Fig. 4C–F,I). However, blockade of R-type calcium channels (RTCC) with SNX482 (0.3 µM) markedly suppressed the potentiating action of anti-Kv4.2 (Fig. 4G; SNX482, 1.0 ± 0.06, n = 11, p = 0.58, Wilcoxon signed rank test). Given that SNX482 is also known to inhibit Kv4.2-mediated currents (Kimm and Bean, 2014), Nickel (Ni2+) was used as an alternative RTCC blocker to assess its effect on the action of anti-Kv4.2. Low concentration of Ni2+ (30 µM), which primarily inhibits T-type calcium channels, failed to prevent the increase in EPSC amplitude induced by anti-Kv4.2. However, a higher concentration of Ni2+ (100 µM), which inhibits both R- and T-type calcium channels, effectively suppressed the potentiating action of anti-Kv4.2 (Extended Data Fig. 4-1; Lee et al., 1999; Zamponi et al., 1996). Additionally, in the presence of SNX482, anti-Kv4.2 did not affect the potency or failure rate during minimal stimulation (Extended Data Fig. 4-2). We also confirmed that RTCC inhibition does not affect EPSC amplitudes in the absence of anti-Kv4.2 (Fig. 4H), indicating that RTCC does not regulate AMPAR density or awaken silent synapse in the basal state on its own but mediate Ca2+ signaling required for the recruitment of AMPARs by Kv4.2 inhibition.
Kv4.2-dependent AMPAR regulation is mediated by protein kinase A and protein kinase C
Given that AMPAR trafficking is regulated by a variety of intracellular signaling cascades (Esteban et al., 2003; Boehm et al., 2006; Lisman et al., 2012), we next sought to elucidate the molecular mechanisms underlying Kv4.2-dependent AMPAR regulation. Pretreatment of slices with either blocker targeting protein kinase A (PKA) or protein kinase C (PKC), which is H89 (10 µM) or GF109203X (10 µM), respectively, significantly inhibited the EPSC increase induced by anti-Kv4.2 (Fig. 5A,B,H; p = 0.009 for H89; p < 0.0001 for GF109203X; Mann–Whitney U test). Moreover, combined application of H89 and GF109203X completely nullified the effect of anti-Kv4.2 (denoted as “mixture” in Fig. 5C,H; 1.0 ± 0.04, n = 9, p = 0.26, Wilcoxon signed rank test). We confirmed that the reduction of ITO by anti-Kv4.2 remained intact in the presence of these kinase blockers (Extended Data Fig. 5-1, p = 0.44, Mann–Whitney U test). Furthermore, U0126 (20 µM), a potent inhibitor of extracellular signal-regulated kinase (ERK) which is downstream target for both PKA and PKC, inhibited the EPSC increase induced by anti-Kv4.2 (Fig. 5D,H; p = 0.04, Mann–Whitney U test). In contrast, pretreatment with blockers targeting phospholipase C (PLC, U73122, 10 µM) and Ca2+/calmodulin-dependent kinase (CaMKII, KN93, 5 µM) or inclusion of calmodulin inhibitory peptide (CaM-IP) in the intracellular solution did not significantly affect the anti-Kv4.2-induced increases in EPSCs (Fig. 5E–H). These data suggest the involvement of postsynaptic PKA, PKC, and ERK, but not PLC or CaMKII, in anti-Kv4.2-induced recruitment of AMPARs.
PKA and PKC signaling pathways are required for anti-Kv4.2–mediated increase in EPSC. A–G, Left, Average time course of normalized EPSC amplitudes showing effect of anti-Kv4.2 in the presence of inhibitors targeting various signaling molecules represented by distinct color: H89 (A), GF109203X (B), mixture (H89 and GF109203X; C), U0126 (D), U73122 (E), KN93 (F), and CaM-IP (G). In each time course plot, colored circles represent normalized EPSC amplitudes with inhibitors; red lines indicate normalized EPSC amplitudes with anti-Kv4.2 alone. Right, Representative traces of EPSC showing the effect of anti-Kv4.2 in the presence of various kinase inhibitors at indicated time points at T1 and T2. T1 and T2 were obtained at 3–5 and 23–25 min after patch break-in, respectively. Scale bars: 100 pA and 10 ms. H, Bar graph summarizing anti-Kv4.2–mediated fold changes in EPSC amplitudes across different treatments (+H89, 1.3 ± 0.099, n = 21, p = 0.009; GF109203X, 1.2 ± 0.06, n = 20, p < 0.0001; mixture, 1.0 ± 0.07, n = 9, p < 0.0001; U0126, 1.15 ± 0.12, n = 5, p = 0.01; +U73122, 1.5 ± 0.16, n = 8, p = 0.56; +KN93, 1.5 ± 0.12, n = 8, p = 0.32; +CaM-IP, 1.7 ± 0.09, n = 3, p = 0.31; compared with anti-Kv4.2 alone; see also Extended Data Fig. 5-1 for anti-Kv4.2 effect on K+ currents in the presence of mixture). In each time course, evoked EPSC amplitudes were averaged into 1 min bins. Solid line indicates the presence of each inhibitor. Statistical significance was evaluated by Mann–Whitney U test. #p < 0.05, ##p < 0.01, ####p < 0.0001 (compared with anti-Kv4.2); N.S., not significant.
Figure 5-1
(related to Figure 5). Inhibition of various signaling kinases does not affect the anti-Kv4.2 induced reductions in transient A-type K+ current A. (Left) Time course showing the effect of anti-Kv4.2 on normalized amplitudes of Ipeak in the presence of mixture (GF1093X and H89). (Right, Top) Representative traces of ourward K+ current induced by a 500 ms depolarization to + 30 mV at time point 1 and 2. Scale bars, 1 nA and 200 ms (Bottom, Left) Superimposed K+ current traces at 1 and 2. Scale bars, 1 nA and 20 ms. (Bottom, Right) Subtracted current (1-2), showing specific reduction of ITO by anti-Kv4.2 even in the presence of mixture. Scale bar, 200 pA and 5 ms. B. Bar graph comparing anti-Kv4.2-mediated changes in normalized Ipeak under two conditions: with (Gray, Left) or without (Red, Right) kinase inhibitors (+ mixture vs – mixture, 0.8 ± 0.05 (n = 8) vs 0.7 ± 0.04 (n = 6), p = 0.44, Mann Whitney U test). Statistical significance was evaluated by Mann Whitney U test. N.S. = not significant. Download Figure 5-1, TIF file.
Kv4.2-dependent AMPAR regulation in the hippocampus correlates with its expression pattern
Next, we investigated whether Kv4.2-dependent AMPAR regulation occurs at other synapses in the hippocampus. The amplitude of EPSCs at synapses in the inner molecular layer (IML) of the DG was increased by anti-Kv4.2 to a similar magnitude as was observed at LPP-GC synapses [Fig. 6A, IML vs OML, 1.7 ± 0.04 (n = 10) vs 1.6 ± 0.05 (n = 80), p = 0.06, Mann–Whitney U test]. Then we tested the effect of anti-Kv4.2 in other pyramidal neurons in the hippocampus, which are anatomically distinct from that of GCs (Krueppel et al., 2011; Ferguson and Skinner, 2022). In CA1 pyramidal neurons, EPSCs evoked by stimulation of Schaffer collateral pathways (SC-CA1) increased significantly following anti-Kv4.2 (Fig. 6B, 2.1 ± 0.25, n = 7, p = 0.008, Wilcoxon signed rank test), which was greater than that observed at LPP-GC synapses. Interestingly, anti-Kv4.2 specifically enhanced EPSCs at associational/commissural synapses to CA3 pyramidal neurons in the stratum radiatum (SR) region (Fig. 6C, 1.5 ± 0.07, n = 6, p = 0.03, Wilcoxon signed rank test), but not at mossy fiber (MF-CA3) synapses in the stratum lucidum (SL) region (Fig. 6C, 1.0 ± 0.05, n = 10, p = 0.92), and the effect was smaller than those observed in GCs or CA1 pyramidal neurons. We also confirmed that the effect of anti-Kv4.2 on EPSP amplitude was in parallel with that on EPSC in these synapses (Fig. 6A–C).
Synaptic strength regulation by Kv4.2 corresponds with its expression in the hippocampus. Aa–Ca, Experimental setup showing dialysis of anti-Kv4.2 in GCs and stimulating electrode positioned in different synaptic pathways: IML in DG (Aa), SR in CA1 to stimulate SC-CA1 synapses (Ba), and either SL or SR in CA3 to stimulate MF or associational/commissural fibers to CA3 synapses (Ca), respectively. Ab–Cb, Left, Superimposed EPSCs (top) and EPSPs (bottom) traces at T1 and T2 and (right) bar graphs demonstrating the antimediated changes in EPSP and EPSC amplitudes at various hippocampal synapses including IML in GC (Ab), SC-CA1 (Bb), A/C-CA3 (Cb), and MF-CA3 (Cb) synapses. Scale bars: 100 pA and 10 ms for EPSCs, 5 mV and 10 ms for EPSPs. D, Top, Representative fluorescence immunostaining images showing the subcellular expression pattern of Kv4.2 in DG, CA1, and CA3 of 7-week-old mice. Scale bar, 100 µm. Bottom, The line scan analysis of Kv4.2 fluorescence intensity along the white solid line traversing the image, with bar graph showing average fluorescence intensity (∼30 µm) of corresponding dendritic within the images (marked with 1, 2, and 3 in different dendritic regions of the image). E, Plot of the fold increase in EPSC amplitude by anti-Kv4.2 versus relative fluorescence intensity of Kv4.2 expressions revealing a strong linear correlation (r2 = 0.59). Fluorescence intensity in each dendritic region was normalized to the somatic expressions of Kv4.2 (designated as 1) of each hippocampal region. Fitted line of data is presented with black solid line. Statistical significance was evaluated by Wilcoxon signed rank test and Mann–Whitney U test. *p < 0.05, **p < 0.01, ***p < 0.001 (compared with T1), #p < 0.05, ##p < 0.01 for Mann–Whitney U test; N.S., not significant.
We hypothesized that the subcellular expression pattern of Kv4.2 may be correlated with the different effects of anti-Kv4.2 across various regions of the hippocampus. The expression of Kv4.2 was minimal in the somatic regions of DG, CA1, and CA3 but robust in the dendritic layers, particularly within the SR of CA1 and the molecular layer of DG (Fig. 6D). Expression patterns did not show a significant gradient along proximal to distal axis in dendritic areas of the DG and CA1. This observation is compatible with the effects of anti-Kv4.2 on EPSCs of DG and CA1 shown above. Nevertheless, there is a considerable gradient of Kv4.2-mediated IA along the proximal to distal axis (Hoffman et al., 1997; Murphy et al., 2022), suggesting that the expression pattern may not necessarily align with its functional gradient. The expression of Kv4.2 in the CA3 exhibited a distinct pattern along the dendritic axis: it was negligible in the SL where MF-CA3 synapses are located, while it was obvious in the SR region but much weaker compared with DG and CA1 (Fig. 6D). A plot of the EPSC increase by anti-Kv4.2 in different synapses against the relative intensity of Kv4.2 expressions in the corresponding dendritic regions suggests a correlation between the two (Fig. 6E, r2 = 0.59). Taken together, the role of Kv4.2 in AMPAR recruitment is not affected by the anatomical differences observed between neurons. Rather, it is limited to the synapses where Kv4.2 is expressed, with the degree of modulation dependent on the density of expression.
Kv4.1 regulates EPSP responses without affecting EPSCs
We have previously shown that Kv4.1, another Kv4 family channel expressed in the DG, exhibited a distinct expression pattern compared with Kv4.2 (Kim et al., 2020). In contrast to the Kv4.2 expression pattern shown in Figure 6D, Kv4.1 was moderately expressed in the granular cell layer of DG, with negligible expression in the dendritic region (Kim et al., 2020), suggesting that Kv4.1 plays a different role from Kv4.2 in the regulation of synaptic transmission. To investigate the relationship between the subcellular localization of K+ channels and their role in regulating synaptic responses, we examined the effect of blocking postsynaptic Kv4.1 on synaptic responses using Kv4.1 antibody (anti-Kv4.1) in the intracellular solution. Contrary to the effects of Kv4.2 inhibitions, Kv4.1 blockade did not significantly affect EPSPs or EPSCs at LPP-GC synapses (Fig. 7B,C). Interestingly, however, when stimulating electrodes were placed in a more proximal part of the dendrites within the IML regions of DG, anti-Kv4.1 significantly increased EPSP amplitude [Fig. 7D,E; 2.7 ± 0.18 vs 4.1 ± 0.41 (mV), n = 6, p = 0.03, Wilcoxon signed rank test], without affecting EPSC amplitude (Fig. 7F, p = 0.31, Wilcoxon signed rank test). These data suggest that Kv4.1 inhibition directly increases EPSPs of GCs by increasing intrinsic excitability, but not by increasing AMPAR density. Consistently, the anti-Kv4.1–mediated increase in EPSPs was not inhibited by BAPTA (Fig. 7G, p = 0.83, Mann–Whitney U test), confirming that the increased intrinsic excitability by anti-Kv4.1 does not require Ca2+ signaling.
Blockade of Kv4.1 enhanced EPSP amplitudes without affecting EPSCs in proximal dendrites of GCs. A, D, Top, Experimental setup showing the dialysis of anti-Kv4.1 in GCs via patch pipette. The stimulating electrode positioned either in the OML (A, blue) or IML (D, purple) of the DG. Right, Representative traces of EPSPs (middle) and EPSCs (bottom) at T1 (gray) and T2 (colored) conditions. Scale bars: 5 mV and 10 ms for EPSPs and 100 pA and 10 ms for EPSCs. B, C, Bar graph showing changes in EPSP (B, 1.0 ± 0.06, n = 5, p = 0.81, Wilcoxon signed rank test) and EPSC (C, 1.1 ± 0.06, n = 5, p = 0.18, Wilcoxon signed rank test) amplitudes by anti-Kv4.1 in distal dendrites of GCs. E, F, Bar graph showing selective enhancement in EPSP (E, 1.5 ± 0.10, n = 6, p = 0.03, Wilcoxon signed rank test), not EPSC (F, 1.1 ± 0.06, n = 8, p = 0.31, Wilcoxon signed rank test) amplitudes by anti-Kv4.1 in proximal dendrites of GCs. G, Average time course (left) and bar graph (right) showing anti-Kv4.1–mediated changes in EPSP amplitudes in proximal dendrites of GCs in conditions with (circle) or without (line) 10 mM BAPTA in intracellular solution. In time course, evoked EPSC amplitudes were averaged into 1 min bins. Statistical significance was evaluated by Wilcoxon signed rank test and Mann–Whitney U test. *p < 0.05, N.S., not significant.
AMPAR recruitment by anti-Kv4.2 does not occlude LTP
Previous studies showed that an increase in intrinsic excitability by Kv4.2 inhibitions, along with AMPA recruitment by LTP stimulation, synergistically induces synaptic potentiation (Chen et al., 2006; Kim et al., 2007; Truchet et al., 2012). Adding to this role of Kv4.2 in synaptic plasticity, we have additionally shown that Kv4.2 acts as strong synaptic strength at basal activity. In fact, the density of AMPARs at the synaptic site increases during NMDAR-dependent LTP (Malinow, 2003, Granger and Nicoll, 2014). To test the possibility that the anti-Kv4.2-induced regulation of basal synaptic strength and activity-dependent synaptic plasticity share the same AMPAR recruitment mechanisms, we next examined whether NMDAR-dependent LTP induced by theta burst stimulation (TBS) at LPP-GC synapses was occluded by the presence of anti-Kv4.2 (Kim et al., 2018). TBS was applied 15 min after patch break-in to ensure that the effects of anti-Kv4.2 were stabilized. Prior to TBS induction, baseline EPSCs were recorded for about 5 min, and EPSCs were subsequently measured for 25 min after LTP induction (Fig. 8A). Remarkably, even in the presence of anti-Kv4.2, TBS still led to a notable increase in EPSC amplitude (Fig. 8A, n = 7, p = 0.016, Wilcoxon signed rank test). This enhancement in EPSC amplitude following LTP induction was comparable to that of the control group without anti-Kv4.2 (Fig. 8B, p = 0.80, Mann–Whitney U test). We repeated the same LTP experiments at SC-CA1 synapses. Consistent with the LPP-GC results, TBS-induced LTP at SC-CA1 was not occluded by prior synaptic strengthening by anti-Kv4.2 (Fig. 8C), and the magnitude of EPSC increase after the LTP induction was similar in both experimental groups (Fig. 8D, p = 0.27, Mann–Whitney U test). These results suggest that AMPAR recruitment by anti-Kv4.2 at basal synaptic transmission is mediated by mechanisms distinct from those involved in LTP.
Kv4.2 inhibition does not occlude expression of LTP at LPP-GC synapses and SC-CA1 synapses. A, C, Average time course of normalized EPSC amplitudes (left) with corresponding bar graphs (right) showing changes in EPSC amplitude at LPP-GC (A, red) and SC-CA1 synapses (C, cyan) by anti-Kv4.2 and subsequently by TBS induction. Dialysis of anti-Kv4.2 in GCs increased EPSC amplitudes significantly, stabilizing at levels indicated by the solid line (p = 0.03 for LPP-GCs; p = 0.04 for SC-CA1s; Wilcoxon signed rank test). Subsequent application of TBS, indicated by arrows, further enhanced EPSC amplitudes in the presence of anti-Kv4.2 at both synapses (p = 0.02 for LPP-GC; p = 0.04 for SC-CA1; Wilcoxon signed rank test). B, D, Time course of normalized EPSC amplitudes (left) and bar graphs (right) at LPP-GC synapses (B, red) and SC-CA1 synapses (D, cyan) comparing the magnitude of LTP with or without (gray) anti-Kv4.2. For conditions including anti-Kv4.2, EPSCs were normalized to post–anti-Kv4.2 effect, indicated by solid line in A and C. TBS was applied at time zero (indicated by arrows). The magnitude of LTP did not significantly differ between conditions with or without anti-Kv4.2 at both LPP-GC (1.9 ± 0.16 vs 1.9 ± 0.26, p = 0.79, Mann–Whitney U test) and SC-CA1 (1.8 ± 0.24 vs 1.8 ± 0.07, p = 0.27, Mann–Whitney U test) synapses. Inset: Representative traces of EPSC amplitudes before (gray) and after (black for inactive Ab and red/cyan for anti-Kv4.2) TBS. Scale bars: 100 pA and 10 ms. E, A schematic diagram summarizing the distinct mechanisms of AMPAR-mediated synaptic strength modulation by Kv4.2 channels at basal conditions versus those activated during synaptic plasticity. In time course, evoked EPSC amplitudes were averaged into 1 min bins. Statistical significance was evaluated by Wilcoxon signed rank test and Mann–Whitney U test. *p < 0.05, N.S., not significant.
Discussion
Activity-dependent changes in synaptic functions are one of the key features of neural activities, and AMPARs are a primary target of this regulatory process (Malinow, 2003; Granger and Nicoll, 2014). However, it is not well understood how basal expression of AMPARs is regulated. In the present study, we show that inhibiting Kv4.2 with anti-Kv4.2 results in AMPAR recruitment at postsynaptic sites, which is mediated by RTCC-dependent Ca2+ signaling that activates PKA/PKC pathways. The results suggest that synaptic strengthening by Kv4.2 inhibition occurs through mechanisms distinct from those driving EPSC increases during LTP, as Kv4.2 inhibition did not occlude TBS-induced LTP. Moreover, our data indicate a correlation between the Kv4.2-dependent regulation of AMPARs in the hippocampus and its expression pattern within dendritic regions. The present results proposed that Kv4.2 plays an essential role in maintaining basal AMPAR density through regulatory mechanisms distinct from those involved in LTP (Fig. 8E).
Distinct mechanisms for the regulation of synaptic transmission by Kv4.2 and Kv4.1
Substantial evidence has established the role of Kv4.2-mediated A-type K+ current in dendritic signaling and synaptic plasticity (Hoffman et al., 1997; Kim et al., 2005; Chen et al., 2006; Jung and Hoffman, 2009; Oulé et al., 2021). It was generally believed that inhibiting Kv4.2-mediated outward currents in dendrites would amplify synaptic depolarization (Hoffman et al., 1997). Furthermore, Oulé et al. (2021) proposed that Kv4.2 does not influence AMPA receptors at the synapse, based on the observation that the EPSP slope-to-amplitude ratio remains unaltered in Kv4.2 knock-out mice. We also observed the effect of Kv4.2 inhibition on shaping EPSPs (Fig. 2G), but this effect is small compared with the effect of Kv4.2 inhibition on EPSCs. Our studies provide evidence that enhanced synaptic depolarization by Kv4.2 inhibition is caused mainly by an increase in AMPAR-mediated currents.
It is noted that previous studies regarding Kv4.2 contributions to synaptic transmission have faced methodological drawbacks for detailed analysis. Commonly used transient A-type K+ channel blockers, such as 4-aminopyridine (4-AP), BaCl2, or phrixotoxins, have nonspecific effects on other K+ channels, potentially complicating the interpretation of data. Furthermore, approaches like genetic ablation of Kv4.2 or pharmacological blockade of Kv4.2 failed to distinguish presynaptic and postsynaptic effects. Additionally, compensatory overexpression of other K+ channels cannot be ruled out in studies using genetic ablation models (Kim et al., 2005; Andrásfalvy et al., 2008; Nerbonne et al., 2008).
To address these challenges and elucidate the role of dendritic Kv4.2, we included specific Kv4.2 antibodies in the intracellular solution for targeted inhibition of Kv4.2-mediated transient currents without affecting sustained currents (Fig. 1) and directly measured their effects on AMPA-mediated currents (Fig. 2), showing that Kv4.2 channels act as a powerful regulator of AMPA currents. On the other hand, we found that Kv4.1, another Kv4 family channel with negligible dendritic expression, regulates EPSP responses without affecting EPSC amplitudes, possibly by regulating intrinsic excitability (Fig. 7). These findings highlight the significance of ion channel localization in determining their functional impact. Future studies are necessary to determine whether the ability to regulate AMPAR recruitment is a unique feature of Kv4.2 or other dendritically located K+ channels can similarly influence AMPAR trafficking.
Regulation of AMPAR expression by Kv4.2 through RTCC-mediated Ca2+ signaling in dendrites
We demonstrated that intracellular BAPTA and preincubation of SNX482 completely blocked the potentiating effect of anti-Kv4.2, indicating that Kv4.2-dependent regulation of AMPARs during basal synaptic activity is Ca2+-dependent. In support of this idea, blockade of Kv4.2 led to an increase in mEPSC amplitude and slowed the decay of spontaneous Ca2+ transients in dendritic spines (Murphy et al., 2022). Since these effects were prevented by Ni2+, an RTCC blocker, it was interpreted that RTCCs regulate Kv4.2 and thus regulate Kv4.2-mediated attenuation of mEPSC. Although the mechanism underlying the attenuation of mEPSCs by Kv4.2 was not further investigated, it was found that RTCCs and Kv4.2 form complexes in dendrites and spines. In addition, RTCC-mediated Ca2+ influx promotes Kv4.2 functional expression in dendrites in a KChIP-dependent manner. Combining these findings with ours, we propose that RTCC-mediated Ca2+ signaling increases both AMPAR density and Kv4.2 surface expression. Given that Kv4.2 inhibits AMPAR trafficking to the membrane as shown in the present study, the RTCC-induced simultaneous increases of both AMPAR and Kv4.2 surface expression would limit uncontrolled increment of AMPAR density through Kv4.2-mediated regulation of AMPAR. Consequently, Kv4.2 and RTCCs may establish a negative feedback loop for synaptic Ca2+ and AMPAR levels during basal synaptic activity.
We found that RTCCs, not L-type calcium channels (LTCCs), mediated Kv4.2-dependent regulation of AMPARs (Fig. 4). These results are consistent with previous reports suggesting that RTCCs are a major source of AP-evoked Ca2+ influx in dendritic spines of CA1 PNs (Sabatini and Svoboda, 2000; Yasuda et al., 2003; Bloodgood and Sabatini, 2008). Also, the proximity of Cav2.3 to Kv4.2 in dendritic spines of pyramidal neurons (Murphy et al., 2022) implied that a specialized role of Kv4.2 in synaptic response regulation through RTCCs. Considering heterogeneity in the expression of VGCC classes across different cell types in dendritic spines, it is essential to explore the regulatory role of Kv4.2 on VGCCs in various cell types in future studies.
In addition to VGCCs, NMDARs are potential calcium sources at subthreshold potentials that can also be regulated by K+ channels (Sobczyk et al., 2005; Higley and Sabatini, 2008). For example, small-conductance calcium-activated K+ channels (SK) in CA1 (Ngo-Anh et al., 2005) and large-conductance calcium-activated K+ channels (BK) channels in GCs (Zhang et al., 2018) regulate NMDAR activity in the dendritic spines. Ca2+ influx through NMDARs activates SK or BK channels, which provide hyperpolarizing current to inhibit NMDAR activation, while inhibiting SK or BK channels enhances NMDAR-mediated currents. However, we did not find any evidence that Kv4.2 inhibition increases NMDAR-mediated currents. This may be due to the higher activation threshold of NMDAR (Zito and Scheuss, 2009) compared with RTCC (Wormuth et al., 2016; Neumaier et al., 2020), so that depolarization induced by Kv4.2 inhibition may not be sufficient to enhance the amplitude of NMDAR-mediated current.
Regulation of synaptic strength by PKA-, PKC-, and ERK-mediated signaling pathways in dendrites
It has been previously shown that activation of PKA and PKC modulates biophysical properties of Kv4.2 channels (Hoffman and Johnston, 1998; Watanabe et al., 2002; Schrader et al., 2006) and that both of these signaling pathways converge on ERK-MAPK in mediating the reduction in dendritic K+ currents (Yuan et al., 2002). We demonstrated that inhibition of either PKA or PKC results in a partial reduction of anti-Kv4.2 mediated increase in EPSC amplitude, whereas combination of these inhibitors (PKA and PKC) or ERK inhibition alone leads to complete suppression of the effect of anti-Kv4.2 (Fig. 5). Our results are consistent with the idea that the mechanism involved in EPSC increase by anti-Kv4.2 are dually regulated by PKA and PKC through a common downstream pathway involving ERK-MAPK. We proposed that RTCC-mediated Ca2+ increase induced by anti-Kv4.2 leads to activation of PKC/PKA/MAPK, resulting in AMPAR recruitment. It is interesting to note that RTCC-mediated Ca2+ signaling was reported to regulate Kv4.2 surface density without affecting their biophysical properties in a KChIP-dependent manner in CA1 pyramidal neurons (Murphy et al., 2022), which is distinct from the regulation of Kv4.2 by agonist-induced activation of PKC/PKA/MAPK (Yuan et al., 2002). On the other hand, biophysical properties of Kv4.2 in cerebellar granule cells are regulated by calcium influx through T-type VGCCs in KChIP-dependent manner (Anderson et al., 2010). These results suggest that molecular mechanisms involved in modulation of Kv4.2 channels are diverse and different across cell types. Taking these studies together, it can be suggested that the PKA and PKC pathways, which converge onto the ERK-MAPK pathway, can be activated not only by agonists but also by voltage-dependent Ca2+ signaling and exert a synergistic effect on enhancing synaptic transmission by modulating both the dendritic K+ channel and AMPARs.
Distinctions between RTCC-mediated regulation and NMDAR-dependent regulation for AMPARs and Kv4.2
Our data revealed that Kv4.2-mediated AMPAR recruitment operates by mechanisms distinct from those involved in synaptic LTP induction (Fig. 8). LTP in PP-GC synapses is regulated by NMDAR-mediated dendritic depolarization (Kim et al., 2018) and LTCC-mediated calcium influx (Lopez-Rojas et al., 2016; Kim et al., 2023), leading to the activation of CaMKII and an increase in AMPAR density in synaptic sites. In contrast, we found that Kv4.2-mediated regulation of basal synaptic strength primarily relied on RTCC-mediated calcium influx and signaling molecules such as PKA, PKC, and ERK, suggesting multiple regulatory mechanisms of AMPARs.
It is interesting to note that not only AMPARs but also Kv4.2 are regulated by both RTCC and NMDARs. NMDAR-mediated Ca2+ influx triggers the internalization of Kv4.2 (Kim and Hoffman, 2008; Jung and Hoffman, 2009), which is mediated through phosphorylation of Kv4.2 by PKA at Ser552 (Hammond et al., 2008). Conversely, RTCC-dependent Ca2+ influx promotes the recruitment of Kv4.2 to the membrane (Murphy et al., 2022). These results suggest that Kv4.2 trafficking to and from the surface membrane is intricately regulated by different Ca2+ signaling pathways. Although detailed molecular mechanisms have not yet been uncovered, the functional significance of this differential regulation for Kv4.2 is evident. AMPAR recruitment in association with Kv4.2 internalization induced by NMDAR-mediated calcium influx during LTP works synergistically to enhance synaptic responses (Kim and Hoffman, 2008; Jung and Hoffman, 2009). Meanwhile, concurrent trafficking of AMPARs and Kv4.2 by RTCC-mediated Ca2+ signaling forms a negative feedback loop that limits synaptic responses. In conclusion, our data provide a comprehensive framework for understanding the physiological roles and underlying mechanisms of Kv4.2-dependent regulation of synaptic transmission. This is particularly important given the broad expression of Kv4.2 in the brain and the potential pathological impacts of Kv4.2 dysfunctions.
Footnotes
This work was supported by the National Research Foundation of Korea (RS-2024-00333669 to SHL). We thank all members of Cell Physiology labs for their helpful comments.
The authors declare no competing financial interests.
- Correspondence should be addressed to Won-Kyung Ho at wonkyung{at}snu.ac.kr or Suk-Ho Lee at leesukho{at}snu.ac.kr.
