In hippocampal pyramidal neurons, voltage-gated Ca2+ channels open in response to action potentials. This results in elevations in the intracellular concentration of Ca2+ that are maximal in the proximal apical dendrites and decrease rapidly with distance from the soma. The control of these action potential-evoked Ca2+ elevations is critical for the regulation of hippocampal neuronal activity. As part of Ca2+ signaling microdomains, small-conductance Ca2+-activated K+ (SK) channels have been shown to modulate the amplitude and duration of intracellular Ca2+ signals by feedback regulation of synaptically activated Ca2+ sources in small distal dendrites and dendritic spines, thus affecting synaptic plasticity in the hippocampus. In this study, we investigated the effect of the activation of SK channels on Ca2+ transients specifically induced by action potentials in the proximal processes of hippocampal pyramidal neurons. Our results, obtained by using selective SK channel blockers and enhancers, show that SK channels act in a feedback loop, in which their activation by Ca2+ entering mainly through L-type voltage-gated Ca2+ channels leads to a reduction in the subsequent dendritic influx of Ca2+. This underscores a new role of SK channels in the proximal apical dendrite of hippocampal pyramidal neurons.
- SK channel
- pyramidal neuron
- calcium imaging
the afterhyperpolarizing current IAHP is mediated by apamin-sensitive, small-conductance Ca2+-activated K+ (SK) channels that are voltage independent and activated by increases in intracellular Ca2+, thereby linking intracellular Ca2+ elevations to changes in the membrane potential in a variety of neurons (reviewed by Adelman et al. 2012; Pedarzani and Stocker 2008). The constitutive binding of calmodulin to SK channels is responsible for their high sensitivity to Ca2+ (Xia et al. 1998). However, the sources of Ca2+ leading to the activation of SK channels vary in different types of neurons (reviewed by Pedarzani and Stocker 2008; Stocker 2004).
SK channels are part of Ca2+ microdomains (Fakler and Adelman 2008; Lujan et al. 2009; Marrion and Tavalin 1998; Oliver et al. 2000) created by the functional coupling of Ca2+-permeable channels and Ca2+-sensitive channels, and may serve diverse roles depending on their subcellular localization. In the soma of CA1 neurons Ca2+ channels and small-conductance Ca2+-activated K+ channels are found within 50–150 nm of each other (Marrion and Tavalin 1998).
Synaptic NMDA receptors and SK channels are functionally coupled in the dendritic spines of hippocampal, amygdala, and striatal neurons (Bloodgood and Sabatini 2007; Faber et al. 2005; Higley and Sabatini 2010; Lujan et al. 2009; Ngo-Anh et al. 2005). The activation of SK channels in dendritic spines limits the influx of Ca2+ through NMDA receptors and decreases glutamatergic excitatory postsynaptic potentials (Bloodgood and Sabatini 2007; Faber et al. 2005; Ngo-Anh et al. 2005). Moreover, in distal apical dendrites of hippocampal neurons SK channel activation controls the duration of glutamate-induced Ca2+ plateau potentials (Cai et al. 2004). Thus SK channels are part of a negative feedback loop that limits Ca2+ influx through those Ca2+ sources that initially activated them, shaping the amplitude and duration of synaptically evoked Ca2+ transients and modulating glutamatergic synaptic responses.
The regulation by SK channels of Ca2+ sources that are not dependent on synaptic activation, however, has not been explored so far. This may be of particular relevance in the proximal apical dendrite of hippocampal neurons, where SK channels have also been localized (Lin et al. 2008; Lujan et al. 2009; Sailer et al. 2002) and elevations of intracellular Ca2+ induced by the activation of voltage-gated Ca2+ channels by somatic action potentials (APs), which backpropagate to the dendrites, are largest (Callaway and Ross 1995; Christie et al. 1995; Jaffe et al. 1992; Regehr et al. 1989; Regehr and Tank 1994; Spruston et al. 1995). The proximal dendritic compartment is different from the distal compartment in CA1 neurons also because it receives more effective GABAergic innervation (Papp et al. 2001), while most glutamatergic excitatory inputs converge on the distal portion. Indeed, proximal and distal compartments of apical dendrites have different synaptic plasticity thresholds, which may also reflect a different contribution of voltage-gated Ca2+ channels to plasticity induction mechanisms (Parvez et al. 2010).
In view of the feedback regulation of synaptically evoked Ca2+ entry by SK channels shown in distal dendrites and spines, the present study addresses the question as to whether SK channels can modulate AP-induced Ca2+ transients in the proximal apical dendrites of hippocampal pyramidal neurons. We demonstrate that pharmacological modulation of SK channel activity regulates the amplitude and duration of AP-induced intracellular Ca2+ elevations mainly triggered by L-type voltage-gated Ca2+ channels in the proximal neurites of hippocampal neurons.
Tetrodotoxin (TTX) was obtained from Alomone Laboratories (Jerusalem, Israel); apamin from Laxotan (Rosans, France); dl-AP5, NBQX, picrotoxin, and 1-ethyl-2-benzimidazolinone (1-EBIO) from Tocris Cookson (Bristol, UK) or Ascent Scientific (Weston-super-Mare, UK); and tetraethylammonium (TEA), Na2-ATP, Na3-GTP, and 8-(4-chlorophenylthio)adenosine 3′,5′-cyclic monophosphate (CPT-cAMP) from Sigma-Aldrich (Poole, UK); all other salts and chemicals were obtained from Fluka (Sigma-Aldrich).
Rats were handled in accordance with the UK Home Office Animal Procedures Act (1986), and protocols were reviewed and approved by the University College London Animal Ethical Committee. Primary hippocampal neurons were cultured from 0- to 1-day old rats (Goslin and Banker 1991) according to a modified protocol. Briefly, after dissection, the hippocampi were treated with 2.5% trypsin (Invitrogen) and mechanically dissociated with a flame-polished Pasteur pipette. Cells were plated onto poly-d-lysine-coated (0.1 mg/ml) glass or plastic (Nalgene) coverslips at a density of 35,000 cells/cm2 (for electrophysiology and imaging recordings) or 21,000 cells/cm2 (for immunofluorescence staining) in minimum essential medium (Invitrogen, Paisley, UK) supplemented with 10% horse serum, 1 mM pyruvic acid, and 0.59% glucose. After 4–14 h, the medium was substituted with Neurobasal medium supplemented with penicillin (100 U/ml), streptomycin (100 μg/ml), 0.59% glucose, and B27 supplement (Invitrogen). Neurons were kept at 5% CO2, 37°C, and 95% humidity for a variable number of days before the experiments.
For recordings, coverslips were mounted in a custom-built recording chamber placed on the stage of a Nikon E600FN upright microscope. Cells were continuously superfused by a 2.5 ml/min flow of extracellular solution containing (in mM) 140 NaCl, 3.5 KCl, 10 HEPES, 20 glucose, 2.5 CaCl2, and 1.5 MgCl2 (pH 7.4, 300–305 mosmol/kgH2O) (20°C). CaCl2 was reduced to 1.5 mM for current-clamp recordings with trains of four APs. Borosilicate patch pipettes (5–6 MΩ; TW100F-4 glass, World Precision Instruments) were filled with a solution containing (in mM) 135 KMeSO4 (voltage-clamp and current-clamp experiments) or 135 K-gluconate (voltage-clamp experiments), 10 KCl, 10 HEPES, 1 MgCl2, 2 Na2-ATP, 0.4 Na3-GTP (pH 7.2–7.3, 280–290 mosmol/kgH2O). CPT-cAMP (50 μM) was included in the majority of the voltage-clamp experiments to inhibit the slow IAHP (sIAHP).
Voltage-clamp experiments were performed on pyramidal cells at 10–18 days in vitro (DIV). Neurons were clamped at a membrane holding potential of −50 mV and repetitively depolarized to +30 mV for 100–200 ms at a frequency of 0.033 Hz to activate voltage-gated Ca2+ channels. After each depolarization the membrane potential was stepped back to −50 mV, where the apamin-sensitive IAHP was observed as an outward current. Voltage-clamp experiments were conducted in the presence of 0.5 μM TTX and 1 mM TEA to block voltage-gated Na+ channels and some voltage-gated K+ channels. Series resistance (range 15–25 MΩ) was monitored at regular intervals throughout the recording and presented minimal variations (≤15%) in the analyzed cells. Data are reported without corrections for liquid junction potentials.
Data were acquired with an Axopatch 1-D amplifier (Axon Instruments, Foster City, CA) controlled by the Strathclyde Electrophysiology Software WinWCP v.3.2.9 (John Dempster, University of Strathclyde, Glasgow, UK). Data were filtered at 1 kHz and sampled at 3 kHz with a Micro 1401 interface (Cambridge Electronic Design). Data were analyzed by pCLAMP9 Clampfit routine (Axon Instruments) and Origin 7.0 (MicroCal software).
APs were elicited in the presence of glutamate receptor blockers (dl-AP5 25 μM; NBQX 5 μM) in the whole cell current-clamp mode, and data were filtered at 10 kHz and digitized at 20 kHz. Somatic current injections of 10 ms, which evoked single APs, were delivered at a frequency of 20 Hz. Between stimulations the membrane potential of the cells was kept at −60 mV by DC current injection. The resting membrane potential was frequently checked, and only cells with a stable resting potential more hyperpolarized than −50 mV were included in the analysis.
Ca2+ transients were imaged with 20 μM fluo-4 (Molecular Probes, Eugene, OR) dissolved in the intracellular recording solution; 1 mM fluo-4 stock solution was prepared in purified H2O (Super Purity Reagent, Romil) and stored at −20°C (Yasuda et al. 2004). Neurons were filled with fluo-4 via the patch electrode for 10–15 min before imaging to allow dye equilibration in the proximal neurites (Helmchen et al. 1996; Maravall et al. 2000). Drugs were bath applied for 5–10 min. Because of the pharmacological accessibility of cultured neurons, drug effect was already maximal at 5 min.
Two-photon Ca2+ imaging was performed with a Bio-Rad multiphoton microscope based on a 1024 scan head mounted on a Nikon E600FN upright microscope equipped with a Nikon ×60 NA 1.0 water-immersion objective. A Millennia V pump laser coupled to a mode-locked Ti:sapphire infrared laser (Tsunami, Spectra Physics) was used for fluorescence excitation, tuned to 790 nm. For most experiments the laser power at the sample was 3 mW, but for those cells with a weak dye loading a higher power (up to 7 mW) was used to obtain a clear image. The fluorescence emission was collected with an external photomultiplier detector and was not descanned.
Since two-dimensional scans are too slow for accurate determination of the time course and amplitude of the calcium transients, line scans were used. These consisted of successive sweeps at 6-ms intervals across a single line in the field of view. Images were collected with Lasersharp software (Bio-Rad) and analyzed with ImageJ (National Institutes of Health) and Origin 7.0 (MicroCal software). For each recording, background fluorescence was determined from a cell-free area of size comparable to that of the line scan image. After the averaged background signal was subtracted, fluorescence values were taken 300 ms before the triggering of APs and averaged to measure the basal fluorescence (Fbasal). The amplitude of the fluorescence transients at the recording sites was expressed as the fractional change in basal fluorescence, (F − Fbasal)/Fbasal= (ΔF/F), which is approximately proportional to the changes in intracellular Ca2+ (Maravall et al. 2000). Fbasal did not change by more than two times the standard deviation of Fbasal measured under control conditions over the course of the experiment. For data analysis, transients were digitally filtered off-line (adjacent-averaging routine, smoothing factor n = 5; Origin 7). Peak fluorescence was calculated averaging data points 30–60 ms around the maximum. The decay time course of Ca2+ transients was fitted by a single-exponential function.
Data were analyzed with Prism (GraphPad Software, La Jolla, CA), Student's t-test, paired or unpaired as appropriate, or the nonparametric Mann-Whitney test was used for statistical comparisons between two groups (α = 0.05). For comparisons between more than two groups one-way ANOVA or one-way repeated-measures ANOVA followed by the Bonferroni post hoc test was used. All values are expressed as means ± SE.
Hippocampal neurons were fixed in phosphate saline buffer (PBS: 10 mM sodium phosphate, 130 mM NaCl, pH 7.2) containing 4% paraformaldehyde and 4% sucrose for 10 min at room temperature, rinsed twice in PBS, and permeabilized in 0.3% Triton X-100 for 15 min, followed by two more washes in PBS. After a 1-h incubation in 2% H2O2 to block the activity of the endogenous peroxidase, immunodetection was performed with the tyramide signal amplification method (Invitrogen). In short, the fixed, permeabilized, and peroxide-treated neurons were incubated overnight at 4°C with the affinity-purified anti-NSK2 antibody (0.25 μg/ml; see also Cingolani et al. 2002) diluted in blocking buffer. Controls were performed in parallel either by omitting the purified anti-NSK2 or by using the anti-NSK2 preadsorbed to a SK2 fusion protein (TRX-NSK2, 20 μg/ml). After repeated washes with PBS to remove the unbound primary antibodies, a 1-h incubation with the HRP-conjugated anti-rabbit secondary antibody (1:200) in blocking buffer was performed. After the cultures were washed with PBS, the tyramide-Alexa Fluor 488 reagent was added and incubated in the dark for 5 min. Slides were washed, mounted with ProLong Antifade (Invitrogen), and examined with a fluorescence microscope (Axiophot, Zeiss). Pictures were taken with a MicroPublisher camera (QImaging).
Apamin-sensitive IAHP in cultured hippocampal neurons.
Whole cell recordings were performed on morphologically identified hippocampal pyramidal neurons in primary culture. After 10–18 DIV, neurons showed a mean resting potential of −59 ± 1 mV (n = 38); 100- to 200-ms-long somatic depolarizations to +30 mV from a holding potential of −50 mV, delivered in the presence of 0.5 μM TTX and 1 mM TEA, activated voltage-gated Ca2+ currents, followed by an outward current. The observed currents decayed with either a time constant (τ) of 222 ± 17 ms when elicited by a 100-ms-long depolarization (n = 8), or a τ of 485 ± 46 ms for 200-ms-long depolarizations (n = 17). The mean amplitude of the outward current was 83 ± 20 pA (n = 8) in response to 100-ms-long depolarizing pulses and 140 ± 23 pA (n = 17) after 200-ms-long pulses.
d-Tubocurarine (dTC, 100 μM), which blocks SK channels in a reversible manner, inhibited the outward current (Fig. 1, A and B). Similarly, the specific SK channel blocker apamin (5 nM) produced a strong suppression of the outward current (Fig. 1, A and B), demonstrating that it is mediated by SK channels.
To further validate the molecular identity of the outward current and complete its pharmacological characterization, we tested the SK channel enhancer 1-EBIO (250 μM) (Pedarzani et al. 2001), which increased the IAHP peak amplitude (Fig. 1, C and E). A similar increase in IAHP amplitude was observed with the more specific and potent SK channel enhancer 6,7-dichloro-1H-indole-2,3-dione 3-oxime (NS309) at 5 μM (Fig. 1, D and F) (Pedarzani et al. 2005). In addition, both 1-EBIO (Fig. 1, C and E) and NS309 (Fig. 1, D and F) prolonged the decay τ of the IAHP.
Taken together, these results demonstrate that the IAHP is expressed in cultured primary hippocampal neurons and has properties similar to the SK-mediated IAHP recorded in pyramidal neurons from acute hippocampal slices (Sailer et al. 2002; Stocker et al. 1999).
SK2 channel expression in hippocampal neurons.
Evidence obtained in pharmacological and biochemical studies, and work performed on genetically modified animals, point to the SK2 (KCa2.2) subunit as a main contributor to the formation of SK channels mediating the IAHP in hippocampal neurons (Bond et al. 2004; Sailer et al. 2002; Stocker et al. 1999).
Therefore, the distribution of SK2 channel subunits in postnatal hippocampal neurons was investigated with a specific antibody (anti-NSK2) raised against the NH2-terminal region of the SK2 protein (Cingolani et al. 2002). Clear SK2 immunostaining was observed in the soma of the neurons from DIV 2 to DIV 13 (Fig. 2, B–E). The staining increased progressively from DIV 2. Maximal expression was seen at DIV 10–13 (Fig. 2E). For this reason, neurons between DIV 10 and DIV 18 were used for electrophysiological and imaging experiments. SK2 immunoreactivity was visible in the soma of the hippocampal pyramidal neurons and in the proximal and distal portions of neurites (Fig. 2E). Preadsorption of the anti-NSK2 antibodies (Fig. 2F at DIV 13) and omission of the primary antibody (data not shown) resulted in the lack of fluorescent staining at all developmental stages, confirming the specificity of the detected signal.
Imaging and modulation of Ca2+ transients elicited by action potentials backpropagating to proximal dendrite of hippocampal neurons.
In hippocampal pyramidal neurons there is plentiful evidence that SK channels can be activated by Ca2+ entering through voltage-gated Ca2+ channels opening during trains of APs (Cai et al. 2007; Empson and Jefferys 2001; Fernandez de Sevilla et al. 2006; Kaczorowski et al. 2007; Kramar et al. 2004; Oh et al. 2000; Shah et al. 2006; Stocker et al. 1999; but see Gu et al. 2005). This opens the possibility that the activation of SK channels regulates local, voltage-gated Ca2+ channel-mediated calcium signals.
Simultaneous two-photon Ca2+ imaging and whole cell current-clamp recordings allow the testing of this hypothesis. Recent measurements of Ca2+ transients evoked by single APs in small secondary to quaternary dendritic branches of CA1 pyramidal neurons demonstrated that the calcium-sensitive fluorophore fluo-4, with a Kd ranging from 340 nM (34°C) to 800 nM (24°C) (Yasuda et al. 2004) and its large dynamic range, is the dye of choice when used at low concentrations (Sabatini et al. 2002). A general concern with Ca2+-sensitive processes is the potential interference exerted by Ca2+-sensitive dyes because of their concentration and binding properties. In the case of SK channels activated by Ca2+ entering the cell via voltage-gated Ca2+ channels, the dye might act as an exogenous Ca2+ buffer that captures the incoming Ca2+, thereby reducing the activation of the SK channels.
The amplitude and the time course of decay of the IAHP in the absence and presence of fluo-4 were first tested in a subset of voltage-clamp experiments. Fluo-4 at 20 μM did not significantly alter the amplitude (control: 140 ± 23 pA, n = 17; fluo-4: 98 ± 13 pA, n = 14) and the time constant of decay (control: 458 ± 46 ms, n = 17; fluo-4: 574 ± 50 ms, n = 14) of the IAHP elicited by a 200-ms-long pulse (P = 0.1).
Consequently, this concentration of fluo-4 was used to study the effect of SK channels on the dynamics of intracellular free Ca2+ concentration in response to APs in the proximal dendrite. A typical filled pyramidal cell with two neuronal processes (p1, p2) in the field of vision is shown in Fig. 3A. For visualization, cells were subjected to a prolonged depolarization at the end of the experiment to increase the intracellular Ca2+ concentration. This was necessary because under resting conditions the low level of fluo-4 was too dim to reliably detect neuronal processes. APs were elicited by somatic current injections of 10-ms-long depolarizing pulses at 20 Hz (Fig. 3B, top), a firing frequency in the range observed in CA1 place cells when the animals are in proximity of their place field center (Dragoi and Buzsaki 2006; O'Keefe and Dostrovsky 1971). The two APs induced an increase in fluorescence, indicative of Ca2+ entry, in the proximal neurites of the cell (<50 μm from soma; Fig. 3B). The fluorescence signal was collected by line scans along the dashed line in Fig. 3A with a two-photon imaging system with a photomultiplier as a detector to measure the dynamics of intracellular free Ca2+ concentration (Fig. 3B, middle). The time course of decay of the evoked fluorescence transients (Fig. 3B, bottom) showed cell-to-cell variability with an average τ of 443 ± 34 ms (range: 221–987 ms in 28 processes of 25 cells). To test whether increasing the AP duration (Fig. 3C) saturates the fluo-4 fluorescence signal elicited by two APs at 20 Hz, the K+ channel blocker TEA was applied by bath perfusion. TEA at 10 mM reversibly prolonged the duration of the APs (Fig. 3C) and doubled the amplitude of the fluorescence transients (198 ± 46%, n = 4, Mann-Whitney test, P < 0.04; Fig. 3D). TEA also prolonged the decay times of the transients. The prolongation was quite variable, and in the presence of TEA one cell displayed a fluorescence transient that developed into a plateau and did not decay back to baseline values over the duration of the recording. In the cells where the fluorescence transient returned to baseline, the decay time of the transients was increased [τControl: 316–674 ms (n = 4); τTEA: 464–4,800 ms (n = 3)]. The changes in the amplitude and duration of the Ca2+ transients in the proximal dendrite of hippocampal pyramidal neurons are comparable with the effect of TEA on the AP-induced Ca2+ transients in the dendrites of neocortical neurons (Markram et al. 1995).
The effect of TEA on AP duration (Fig. 3C) and the amplitude and time course of the fluorescence transients was reversible (Fig. 3D). The reversibility of the effect of TEA demonstrates the stability of the signals observed under our recording and imaging conditions.
SK channels regulate action potential-induced Ca2+ influx.
TEA at 10 mM blocks several voltage-gated K+ channels, including members of the Kv1, Kv2, Kv3, and Kv7 families, voltage- and Ca2+-activated large-conductance K+ (BK) channels (Coetzee et al. 1999), and, to some extent, SK channels (reviewed in Pedarzani and Stocker 2008). Consequently, the effect observed in the presence of TEA on the Ca2+ transient (Fig. 3, C and D) is due to the block of different types of K+ currents that contribute to the AP repolarization and afterhyperpolarization phases in hippocampal pyramidal neurons. SK channels have been shown to terminate glutamate-evoked Ca2+ plateau potentials in distal apical dendrites (Cai et al. 2004) and regulate the Ca2+ influx through NMDA receptors through a negative feedback mechanism in spines of hippocampal pyramidal neurons (Bloodgood and Sabatini 2007; Ngo-Anh et al. 2005). This makes SK channels good candidates to regulate AP-induced Ca2+ signals in proximal dendritic regions. To investigate whether and to what extent SK channels specifically contribute to the regulation of Ca2+ transients in the proximal dendrites of hippocampal neurons, SK channels were inhibited by apamin, a selective inhibitor. Apamin increased the amplitude of the fluorescence transient (Fig. 4, A and B) elicited by two APs at 20 Hz in the proximal dendrite to 115 ± 4% (n = 5, P < 0.05; Fig. 4C). Apamin also slowed the decay of the fluorescence transients by 18 ± 5% (Fig. 4, B and D; control: τ = 298 ± 11 ms, apamin: τ = 353 ± 18 ms; n = 5, P < 0.05). As a result of the increase in amplitude and prolongation of τ, the amount of Ca2+ entering the cell, characterized by the area under the curve of the fluorescence transient, was significantly increased in the presence of apamin (Fig. 4, B and E; 138 ± 13%; n = 5, P < 0.05).
To test whether the observed effects were a consequence of apamin acting specifically on Ca2+ influx triggered by APs rather than caused by the direct action of the depolarizing current injection, we measured the impact of the SK channel inhibitor after TTX application. TTX strongly attenuated the fluorescent transients (Fig. 4, F–H; n = 4), confirming that, under our experimental conditions, Ca2+ signals in the proximal dendrite mainly arise from the backpropagation of APs. In the presence of TTX, apamin failed to increase Ca2+ influx (Fig. 4, F–H; F3,2 = 27, P = 0.001; TTX vs. control P < 0.05; TTX+apamin vs. control P < 0.05; TTX+apamin vs. TTX P > 0.05; 1-way repeated-measures ANOVA with Bonferroni post hoc test). This indicates that AP-induced activation of calcium channels is necessary for the SK-mediated modulation of Ca2+ signals.
To corroborate the result obtained with apamin, we tested the effect of a structurally unrelated small organic SK channel blocker, dTC, on the Ca2+ transients in the proximal dendrite of hippocampal pyramidal neurons. Although less specific than apamin, dTC has the advantage that it blocks SK channels in a reversible manner. Application of 100 μM dTC led to an increase in the fluorescence transient amplitude to 113 ± 3% (Fig. 5, A and B), similar to the change observed with apamin. The effect of dTC on the amplitude of the fluorescence transients was reversed in five of six cells (Fig. 5, A and B). The reversibility of the dTC effect on the amplitude of the Ca2+ transients rules out the possibility that the observed increase might be a consequence of dye loading. Blocking SK channels with dTC also resulted in an increase of the area of the fluorescence transients to 120 ± 4% (Fig. 5, A and C) and in a prolongation of the time course of decay, which is reflected by an increase of τ by 20 ± 5%.
Several SK channel enhancers have recently been characterized (reviewed in Pedarzani and Stocker 2008) and act by increasing the apparent Ca2+ sensitivity of SK channels (Pedarzani et al. 2001). If the increase in the Ca2+ transients observed upon application of apamin and dTC is due to the inhibition of SK channels acting as negative feedback regulators of Ca2+ influx triggered by APs, then enhancement of SK channel activity should lead to a reduction in the Ca2+ transients. To test this hypothesis the SK channel enhancer NS309 was used (Pedarzani et al. 2005). NS309 (5 μM) reduced both the amplitude of the fluorescence transients by 25 ± 5% (Fig. 5, D and E) and their area by 30 ± 6% (Fig. 5, D and F) in all cells tested (n = 6). In the presence of NS309 the time constant of decay of the fluorescence transients was also shortened by 21 ± 6% (n = 6, Fig. 5D). The overall reduction of the Ca2+ transients observed in the presence of NS309 and the observed opposite effect obtained with SK channel blockers further support the hypothesis that SK channels regulate Ca2+ transients elicited by APs in the proximal dendrite of hippocampal pyramidal neurons by a negative feedback mechanism.
If SK channels, activated by Ca2+ entering through voltage-gated Ca2+ channels, modulate the Ca2+ transients generated by APs, then increasing the number of APs should lead to an enhanced Ca2+ influx, and therefore a stronger SK channel recruitment and a greater effect on the Ca2+ transients. The fluorescence transients measured in the proximal dendrites of pyramidal neurons in response to four APs were 21% larger than those observed in response to two APs at 20 Hz (4 APs: Fig. 6, A and B, ΔF/F = 1.72 ± 0.11, n = 18; 2 APs: Figs. 3⇑–5, ΔF/F = 1.42 ± 0.12, n = 28; P = 0.02, Mann-Whitney test). The inhibition of SK channels by dTC (100 μM) resulted in an increase of amplitude of the fluorescence transients to 130 ± 8% (Fig. 6, A–C) and of area to 144 ± 9% (Fig. 6, B and E; n = 4, P < 0.05). Additionally, dTC caused a prolongation of the fluorescence transients by 21 ± 5% (Fig. 6D; n = 4, P < 0.05). The relative increases in amplitude and area of the fluorescence transients were significantly larger than those observed with two APs after application of dTC (compare Fig. 5, B and C, with Fig. 6, C and E; P < 0.05). To test whether the relative increase in the amplitude of Ca2+ transients observed in response to four APs in the presence of dTC was limited by dye saturation, we applied 10 mM TEA in the presence of dTC. As expected (see also Fig. 3, C and D), TEA substantially increased the influx of Ca2+ by prolonging the duration of the APs (Fig. 6A, bottom), thereby causing a further large increase of the fluorescence transients (Fig. 6B; amplitude 149 ± 15%, area 213 ± 22%; n = 3). The effect of TEA in this context confirms that the increase of the fluorescence transients observed after dTC application was not limited by fluo-4 saturation following four APs at 20 Hz.
In the proximal dendrites of CA1 neurons, different subtypes of voltage-gated Ca2+ channels are activated by the backpropagation of APs and contribute to local Ca2+ elevations, with a predominant role played by Cav1 (L type) channels (Christie et al. 1995). Moreover, L-type Ca2+ channels have been shown to be physically colocalized and selectively coupled to the activation of small-conductance Ca2+-activated K+ channels in somatic patches from acutely dissociated CA1 pyramidal neurons (Marrion and Tavalin 1998). L-type Ca2+ channels were shown to be the main contributors to the activation of the SK-mediated IAHP also in whole cell recordings from CA1 neurons, with a reduction of IAHP by ∼48% upon application of the L-type Ca2+ channel blocker nifedipine (Bosurgi and Pedarzani 2006). We therefore tested whether the regulatory effect of SK channels on AP-induced Ca2+ transients is triggered by Ca2+ influx through L-type Ca2+ channels. First the contribution of L-type Ca2+ channels to the Ca2+ transients in the proximal dendrite was assessed. In all cells tested, application of the L-type channel blocker nifedipine at 10 μM markedly decreased the amplitude and the area of the fluorescence transients induced by four APs at 20 Hz (Fig. 7; n = 5). As expected, τ of the fluorescence transients was not affected by nifedipine (Fig. 7A; τcontrol = 414 ± 39 ms, τNifedipine = 423 ± 64 ms; n = 5), because τ is mainly determined by Ca2+ extrusion. The subsequent application of dTC in the presence of nifedipine did not increase the amplitude of the fluorescence transients significantly (Fig. 7; F4,2 = 61, P < 0.0001; nifedipine vs. control 69.5 ± 1.2%, P < 0.05; nifedipine+dTC vs. control 77.9 ± 1%, P < 0.05; nifedipine vs. nifedipine+dTC, P > 0.05; 1-way repeated-measures ANOVA with Bonferroni post hoc test), and similar results were obtained for the area of the Ca2+ transients (F4,2 = 18.47, P < 0.001; nifedipine vs. control 81 ± 2%, P < 0.01; nifedipine+dTC vs. control 85.2 ± 4.1%, P < 0.01; nifedipine vs. nifedipine+dTC, P > 0.05; 1-way repeated-measures ANOVA with Bonferroni post hoc test). Similarly, no significant difference for the time constant of decay of the fluorescence transients was observed when comparing controls, nifedipine alone, and the dTC-nifedipine coapplication (Fig. 7A; F4,2 = 0.14, P = 0.9; 1-way repeated-measures ANOVA). When applied to neurons in the presence of nifedipine, the effect of dTC on both the amplitude and the duration of the fluorescence transients was therefore strongly attenuated compared with the results obtained in the absence of the L-type Ca2+ channel blocker (Fig. 6, B–D, and Fig. 7A). We performed additional experiments by first applying dTC, followed by the addition of nifedipine. When applied in the presence of dTC, nifedipine reduced the amplitude (F14,2 = 22, P < 0.0001; dTC vs. control 123 ± 5%, P < 0.01; dTC+nifedipine vs. control 78 ± 7%, P < 0.05; dTC vs. dTC+nifedipine, P < 0.001; 1-way ANOVA with Bonferroni post hoc test) and area (F14,2 = 21, P < 0.0001; dTC vs. control 133 ± 7%, P < 0.01; dTC+nifedipine vs. control 76 ± 6%, P > 0.05; dTC vs. dTC+nifedipine, P < 0.001; 1-way ANOVA with Bonferroni post hoc test) of the Ca2+ transients to values similar or below the control values. These results suggest that the activation of SK channels by APs is at least in part due to the activation of L-type Ca2+ channels and in turn regulates Ca2+ influx in the proximal dendrite.
The dynamic response of pyramidal neurons in the hippocampus is modulated by Ca2+ transients that result from influx through voltage-gated Ca2+ channels. In particular, Ca2+ elevations triggered by backpropagating APs show a maximal amplitude in the proximal dendrites and decrease rapidly with distance from the soma (Callaway and Ross 1995; Christie et al. 1995; Jaffe et al. 1992; Regehr et al. 1989; Regehr and Tank 1994; Spruston et al. 1995). In this study, we have investigated how the activation of SK channels affects AP-induced changes in intracellular Ca2+ levels in proximal processes of hippocampal pyramidal neurons. We have found that in this cellular compartment SK channels limit the amplitude and duration of AP-induced Ca2+ transients.
SK channels have been shown to modulate the amplitude and duration of intracellular Ca2+ signals by feedback regulation of the relevant Ca2+ sources in different dendritic compartments, thus affecting dendritic signal integration and synaptic plasticity. In organotypic hippocampal cultures, SK channels are responsible for the repolarization of local dendritic plateau potentials triggered by focal glutamate application to distal apical dendrites of CA1 pyramidal neurons (Cai et al. 2004). In acute hippocampal slices, synaptic stimulation activates glutamate receptors, leading to the activation of SK channels located on the spine heads, which in turn reduces Ca2+ influx through the NMDA receptors (Bloodgood and Sabatini 2007; Lujan et al. 2009; Ngo-Anh et al. 2005). Both types of feedback regulation were shown to occur in response to synaptically evoked processes and on dendritic compartments receiving primarily glutamatergic excitatory synaptic inputs. The physiological role of the second- and higher-order dendrites, where the excitatory inputs predominate, is fundamentally different from that of the proximal apical dendrite, which mainly receives inhibitory synaptic inputs from GABAergic interneurons (Papp et al. 2001). In the proximal dendrite our results show feedback regulation of the amplitude and duration of AP-induced Ca2+ transients by SK channels. This modulation of transient Ca2+ elevations by SK channels could affect the communication between the soma and the distal apical dendritic tree at the single-cell level and shift the balance between excitation and inhibition at the network level.
Given their high sensitivity to Ca2+ (EC50 ∼300 nM; Xia et al. 1998) and relatively fast time constant of activation (∼5 ms at saturating Ca2+ concentrations; Pedarzani et al. 2001; Xia et al. 1998), SK channels are well suited to take part in a feedback loop to regulate Ca2+ influx in proximal apical dendrites of CA1 neurons, where a single AP leads to Ca2+ elevations on the order of ∼300 nM lasting 70–90 ms, while higher and longer-lasting free Ca2+ concentrations are reached in response to trains of APs (Helmchen et al. 1996; Maravall et al. 2000).
Our results suggest that SK channels act in a negative feedback loop by reducing Ca2+ influx through the Ca2+ channels that activate them after APs. This role of SK channels is supported by the effects of specific SK channel blockers (apamin, dTC) and enhancers (NS309) on the magnitude of Ca2+ transients. The presence of apamin or dTC boosted the Ca2+ influx induced by a train of two APs. Consistent with a negative feedback mechanism, application of the SK channel enhancer NS309 resulted in a reduction of the Ca2+ transient.
The regulatory effect of SK channels was particularly evident when a train of four APs was used to elicit a larger Ca2+ influx, leading to a stronger recruitment of SK channels. The increases in the amplitude of Ca2+ transients following application of SK channel blockers were larger compared with the stimulation with two APs and consistently observed in every cell tested.
Ca2+ buffering by Ca2+-sensitive dyes could in principle mimic the effect of SK channel inhibition. However, this is unlikely in our case because we used a low concentration of Ca2+-sensitive dye and did not observe significant changes in the baseline fluorescence over the course of our experiments. If anything, the added buffer capacity would result in gradual decrease in the Ca2+ transients (Helmchen et al. 1996; Maravall et al. 2000) and lead to an underestimation of the effects of SK channel inhibitors on the amplitude of Ca2+ transients. The validity of the increase of the Ca2+ transients induced by SK blockers is further supported by the reversibility achieved on washout of dTC.
In CA1 pyramidal neurons, APs induce Ca2+ transients that are largest in the proximal dendrites (Callaway and Ross 1995; Christie et al. 1995; Spruston et al. 1995), where they are mediated by different subtypes of high-voltage-activated Ca2+ channels (Bloodgood and Sabatini 2007; Christie et al. 1995, 1996). In particular, L-type Ca2+ channels are highly expressed in the somato-dendritic compartment of pyramidal cells in sections (Leitch et al. 2009; Tippens et al. 2008; Westenbroek et al. 1990) and primary cultures (Pravettoni et al. 2000). Additionally, experiments on the specific high-voltage-activated Ca2+ channel subtypes coupled to the activation of the SK-mediated IAHP in hippocampal pyramidal neurons have revealed that L-type Ca2+ channels are important contributors to the activation of IAHP, which was reduced by ∼48% by the L-type Ca2+ channel blocker nifedipine (Bosurgi and Pedarzani 2006). Application of nifedipine showed a contribution of ∼30% by L-type Ca2+ channels to the total AP-induced Ca2+ elevation in the proximal dendrites of cultured hippocampal pyramidal neurons (Fig. 7). This is in line with a previous report on the relative contributions of different voltage-gated Ca2+ channel subtypes to spike-induced Ca2+ influx in hippocampal pyramidal neurons in brain slices (Christie et al. 1995). L-type Ca2+ channel inhibition prevents the increase in the amplitude of the Ca2+ transients by SK channel blockers (Fig. 7) or reverses it back to control values or below when nifedipine is applied after dTC. This is evidence that L-type Ca2+ channels are implicated in the AP-induced Ca2+ influx leading to SK channel activation in the proximal dendrite of hippocampal pyramidal neurons. We cannot, however, exclude the contribution of other Ca2+ channel subtypes (see also Jones and Stuart 2012).
How do SK channels regulate transient Ca2+ elevations triggered by APs in dendrites? Voltage-gated Ca2+ channels open during the repolarizing phase of APs. While inhibition of BK and voltage-dependent K+ channels by TEA leads to broader APs and increased Ca2+ influx (Fig. 3, C and D), SK channel inhibition does not affect the duration of somatic APs (Fig. 6A). However, we cannot exclude the possibility that SK channels might contribute to shaping the waveform of dendritic APs, which have a lower amplitude and a longer duration in CA1 dendrites (Johnston et al. 2000; Spruston et al. 1995). While the waveform of the somatic AP is not directly affected by SK channel activation, the SK channel effect on AP-induced Ca2+ entry in the proximal dendrite might result from the functional interaction of these channels with other dendritic conductances. Thus two K+ currents, IA and ID, are expressed in CA1 dendrites (Golding et al. 1999; Hoffman et al. 1997) and are inactivated at depolarized potentials. The voltage-dependent inactivation properties of A-type K+ channels enable modest levels of membrane depolarization to decrease the size of the available A channel population and likewise increase dendritic AP amplitude and duration (Hoffman et al. 1997). By hyperpolarizing the membrane potential, SK channels could affect the availability and activation state of these conductances in AP trains, with SK channel inhibition and corresponding membrane depolarization favoring the inactivation of IA and ID. Upon inhibition of IA and ID, large-amplitude, backpropagating APs have been shown to activate dendritic Ca2+ channels or favor the dendritic initiation of Ca2+-dependent potentials, resulting in a massive influx of Ca2+ into the dendrites (Golding et al. 1999; Hoffman et al. 1997; Magee and Carruth 1999). This supports the possibility of a potential interaction between SK channels and A- and/or D-type K+ channels that may underlie the increase in Ca2+ influx observed upon SK channel inhibitions in our recordings. The relatively small, albeit significant, effect exerted by SK channel inhibition on Ca2+ influx in proximal hippocampal pyramidal cell dendrites might well match the gradient of A-type K+ channel density, with fewer channels in the proximal compared with the distal dendritic compartment.
A second potential mechanism for the SK-mediated enhancement of Ca2+ transients in the proximal dendritic compartment is linked to the coexistence of L-type Ca2+ channels with different gating behaviors in neurons, which are thought to give rise to distinct intracellular calcium signals in response to neuronal activity (Forti and Pietrobon 1993; Kavalali and Plummer 1994; Koschak et al. 2007). Thus, in addition to cardiac-like L-type channels, hippocampal neurons display L-type Ca2+ channels with anomalous gating properties, characterized by long channel reopenings after repolarization following strong depolarizations, such as bursts or trains of APs (Kavalali and Plummer 1994; Schjott and Plummer 2000). One potential mechanism by which SK channels might modulate Ca2+ influx would therefore be by reducing the activity of L-type Ca2+ channels in their “anomalous gating” phase. This hypothesis is supported by our results showing that strong depolarization caused by APs is essential to generate the SK-mediated feedback on Ca2+ influx, because this was absent in response to pure electrotonic spread when Na+ channels were blocked by TTX (Fig. 4, F–H).
In addition to anomalous gating properties, L-type Ca2+ channels are subject to various mechanisms of channel inactivation that contribute to the control of Ca2+ entry during ongoing neuronal electrical activity. These include Ca2+-dependent inactivation and fast and slow voltage-dependent inactivation (Budde et al. 2002). The inactivation kinetics of L-type Ca2+ channels are generally described as slow, but they vary in different cell types, possibly because of molecular diversity of the channels (splice variants of the pore-forming subunit; interaction with other Ca2+ channel subunits and modulatory proteins) (Budde et al. 2002). The inactivation profile of L-type Ca2+ channels in hippocampal pyramidal neurons has not been specifically characterized. We cannot exclude that SK channels could interfere in some indirect manner with the inactivation process of Ca2+ channels in proximal dendritic processes, contributing to the increase in Ca2+ influx we have observed on SK channel inhibition.
The pharmacological manipulation of SK channel activity not only increased or decreased the amplitude but also affected the area under the curve of the Ca2+ transients elicited by APs in the proximal dendrite of hippocampal pyramidal neurons (Figs. 4⇑–6). SK channel inhibition led also to a significant prolongation of the Ca2+ transients (Figs. 4⇑–6). Interestingly, the L-type Ca2+ channel blocker nifedipine prevented the effect of SK channel inhibitors on the time constant of decay of Ca2+ transients (Fig. 7). The time course of decay of AP-induced Ca2+ transients in dendrites directly reflects the rate of Ca2+ clearance (Scheuss et al. 2006). Sarco(endo)plasmic reticulum Ca2+-ATPases (Mainen et al. 1999; Majewska et al. 2000; Sabatini et al. 2002), plasma membrane Ca2+-ATPases, and Na+/Ca2+ exchangers (Lorincz et al. 2007; Scheuss et al. 2006) are responsible for the Ca2+ clearance from the cytosol of dendrites and spines in CA1 pyramidal neurons. Notably, both plasma membrane Ca2+-ATPases and Na+/Ca2+ exchangers are expressed in the dendrites of primary hippocampal neurons (Kiedrowski 2004; Kip et al. 2006). The increased duration of Ca2+ signals might simply reflect the longer time needed to clear the augmented Ca2+ after SK channel inhibition (Regehr and Tank 1992). Alternatively, SK channels might affect extrusion mechanisms in different ways. In the thalamus, for example, SK channels and sarco(endo)plasmic reticulum Ca2+-ATPases compete for available Ca2+ and shape Ca2+ transients in an interactive manner (Cueni et al. 2008). Another possibility is that the larger Ca2+ accumulations due to SK channel inhibition attenuated Ca2+ extrusion by plasma membrane Ca2+ ATPases and Na+/Ca2+ exchangers, whose function is reduced in a Ca2+-dependent manner (Scheuss et al. 2006), thereby leading to the observed prolongation of the duration of Ca2+ signals.
Our results demonstrate for the first time that the activity of SK channels can regulate the duration of Ca2+ transient decays in the proximal dendrite of hippocampal neurons. This may affect temporal summation of Ca2+ signals, potentially leading to changes in spike timing-dependent plasticity (Caporale and Dan 2008), as we have recently shown in another brain region, the striatum (Nazzaro et al. 2012). Here SK channels take part in the regulation of Ca2+-dependent release of endocannabinoids and plasticity, through a functional coupling with L-type voltage-gated Ca2+ channels activated by trains of APs (Nazzaro et al. 2012). SK-mediated modulation of intracellular Ca2+ dynamics may similarly be relevant for the activation of Ca2+-dependent signaling cascades to induce different forms of plasticity also in the hippocampal region (Cummings et al. 1996).
This work was supported by an EMBO short-term fellowship (EMBO ASTF 137.00-03) and a Human Frontier Science Program (HFSP) short-term fellowship (ST00323/2002-C) to R. Tonini; a Career Establishment Grant from the UK Medical Research Council to P. Pedarzani (CEG G0100066); a Wellcome Trust Student Prize fellowship to T. Ferraro (068583/Z/02/Z); an MRC-DTA PhD fellowship to M. Sampedro-Castañeda; and a Wellcome Trust Senior Research fellowship to M. Stocker (061198/Z/00A). P. Pedarzani and M. Stocker acknowledge support by the ENI-Net. P. Pedarzani acknowledges support by the HFSP (RGP0013/2010).
No conflicts of interest, financial or otherwise, are declared by the author(s).
Author contributions: R.T., T.F., M.S.-C., A.C., C.D.R., and P.P. performed experiments; R.T., T.F., M.S.-C., A.C., and C.D.R. analyzed data; R.T., T.F., M.S., C.D.R., and P.P. interpreted results of experiments; R.T., T.F., and M.S. prepared figures; R.T., T.F., M.S.-C., M.S., C.D.R., and P.P. edited and revised manuscript; R.T., M.S., C.D.R., and P.P. approved final version of manuscript; M.S., C.D.R., and P.P. conception and design of research; P.P. drafted manuscript.
The authors gratefully acknowledge J. Dempster for supplying the Strathclyde Electrophysiology Software and Dr. D. DiGregorio for useful advice and discussion.
- Copyright © 2013 the American Physiological Society