Specific Enhancement of SK Channel Activity Selectively Potentiates the Afterhyperpolarizing Current IAHP and Modulates the Firing Properties of Hippocampal Pyramidal Neurons*

SK channels are Ca2+-activated K+ channels that underlie after hyperpolarizing (AHP) currents and contribute to the shaping of the firing patterns and regulation of Ca2+ influx in a variety of neurons. The elucidation of SK channel function has recently benefited from the discovery of SK channel enhancers, the prototype of which is 1-EBIO. 1-EBIO exerts profound effects on neuronal excitability but displays a low potency and limited selectivity. This study reports the effects of DCEBIO, an intermediate conductance Ca2+-activated K+ channel modulator, and the effects of the recently identified potent SK channel enhancer NS309 on recombinant SK2 channels, neuronal apamin-sensitive AHP currents, and the excitability of CA1 neurons. NS309 and DCEBIO increased the amplitude and duration of the apamin-sensitive afterhyperpolarizing current without affecting the slow afterhyperpolarizing current in contrast to 1-EBIO. The potentiation by DCEBIO and NS309 was reversed by SK channel blockers. In current clamp experiments, NS309 enhanced the medium afterhyperpolarization (but not the slow afterhyperpolarization sAHP) and profoundly affected excitability by facilitating spike frequency adaptation in a frequency-independent manner. The potent and specific effect of NS309 on the excitability of CA1 pyramidal neurons makes this compound an ideal tool to assess the role of SK channels as possible targets for the treatment of disorders linked to neuronal hyperexcitability.

ized by a time constant of decay of ϳ100 ms and by its sensitivity to the bee venom toxin, apamin, and to the scorpion toxins, scyllatoxin and tamapin (5)(6)(7). sI AHP is characterized by a slower time course (in the range of seconds), by its lack of sensitivity to apamin or any other classical K ϩ channel blocker, and by its modulation by several neurotransmitters (1)(2)(3)8). Based on their kinetic and pharmacological features and on the results obtained from genetically manipulated mice, SK channels mediate I AHP , whereas the molecular correlate for sI AHP is still unknown (2-4, 9, 10).
In addition to the use of selective blockers, an important contribution to the elucidation of the physiological role of SK and IK channels has arisen from the use of a small organic compound that enhances channel activity, the benzimidazolinone 1-EBIO (11)(12)(13)(14)(15). 1-EBIO enhances the activity of SK channels in the presence of the physiological activator, intracellular Ca 2ϩ , by increasing the apparent sensitivity of SK channels to Ca 2ϩ (14). As a consequence, 1-EBIO increases the amplitude of SK-mediated AHP currents and their duration in a variety of neurons, leading to profound changes in neuronal activity and firing patterns (14, 16 -18). Although 1-EBIO has been a useful tool to elucidate the function of SK channels in their native context, it has some important limitations. First, it affects not only the SK channels but also the as yet unidentified Ca 2ϩ -dependent K ϩ channels underlying sI AHP (14). Additionally, prolonged applications of 1-EBIO have been shown to lead to a decrease in Ca 2ϩ currents in hippocampal neurons (14). Finally, and most importantly, 1-EBIO displays a relatively low potency (EC 50 on SK channels ϳ700 M) (14). These limitations of 1-EBIO have prompted the development of novel, more potent SK channel enhancers. DCE-BIO, a dichlorinated analogue of 1-EBIO, has been reported to enhance the activity of intermediate conductance Ca 2ϩ -activated K ϩ channels (IK channels) (19). Moreover, recently 6,7-dichloro-1H-indole-2,3-dione-3-oxime (NS309) (20) has been described as a more potent enhancer of the activity of recombinant SK and IK channels.
In the present study, we provide the first characterization of the effect of DCEBIO on recombinant SK channels and a quantification of the potency differences between DCEBIO, NS309, and 1-EBIO on recombinant SK2 channels, the predominant SK channel subtype in hippocampus. We have furthermore investigated the actions of DCEBIO and NS309 on the native SK channels mediating I AHP , on the distinct Ca 2ϩ -activated K ϩ current sI AHP , and on the firing behavior of CA1 pyramidal neurons in hippocampal slices.

MATERIALS AND METHODS
Electrophysiology on Recombinant hSK2 and Kv7.2/7.3 Channels-HEK293 cells stably expressing human SK2 (20) or co-expressing Kv7.2 and Kv7.3 channel subunits (21) were plated on coverslips 12-24 h prior to the experiments. For each experiment, a coverslip was placed in a 15-l perfusion chamber (flow rate ϳ1 ml/min). All experiments were performed at room temperature (20 -22°C)  mV, 200 ms duration) were applied every 5 s from a holding potential of 0 mV. The Kv7.2/7.3 channels were activated by a 1-s step to Ϫ30 mV (close to the voltage of half-maximal activation), and the deactivation was followed for 1 s at Ϫ60 mV (close to the activation threshold) before stepping to the holding potential of Ϫ90 mV. The protocol was applied every 5 or 10 s.
All neurons included in this study had a resting membrane potential below Ϫ55 mV (Ϫ63 Ϯ 1 mV) and an input resistance of 195 Ϯ 10 megohms. Neurons were voltage-clamped at Ϫ50 mV, and 100-mslong depolarizing pulses to ϩ10 mV were delivered every 30 s to elicit robust and reliable Ca 2ϩ action currents in the presence of 0.5 M tetrodotoxin and 1 mM tetraethylammonium, leading to the activation of the Ca 2ϩ -dependent I AHP and sI AHP (see Refs. 5,23,and 24). The amplitude of I AHP and sI AHP were estimated 50 -80 ms and 1 s after the end of the depolarizing pulse, respectively. In current clamp recordings, tetrodotoxin and tetraethylammonium were omitted, and action potentials were elicited by current injections from the resting membrane potential. Only cells with a stable resting potential throughout the current clamp protocols (Ϯ1 mV) were included in the analysis of the current clamp data. Series resistance (range 15-25 megohms) was monitored at regular intervals throughout the recording. All recordings included in this study presented minimal variations (Յ10%) of the series resistance and of the amplitude and duration of the Ca 2ϩ action current, well within the limits needed to maintain a stable amplitude of I AHP and sI AHP under control conditions. Data are reported without corrections for liquid junction potentials.
Data Acquisition and Analysis-Data were acquired using a patch clamp EPC9 amplifier (HEKA, Lambrecht, Germany) filtered at 0.25-1 kHz, sampled at 1-4 kHz, and stored on a Macintosh G4 or Power PC. Analysis was made using the Pulse and Pulsefit (HEKA, Lambrecht, Germany), Igor Pro (Wave Metrics), SigmaPlot (SPSS, Inc.), InStat (Graphpad), and Excel (Microsoft) software. Values are presented as mean Ϯ S.E. For statistical analysis, the Student's t test was used, and differences were considered statistically significant when p Ͻ 0.05. Concentration-response relationships were fitted to the Hill equation I/I max ϭ [E] n /([E] n ϩ (EC 50 ) n ) to obtain EC 50 values and Hill-coefficients (n).
[E] is the concentration of the enhancer.

RESULTS
DCEBIO and NS309, SK Channel Enhancers More Potent than 1-EBIO-DCEBIO and NS309 have been shown to modulate recombinant IK and SK channels, respectively, at lower concentrations than required for 1-EBIO. To compare their relative potency on recombinant SK channels, we have first characterized the effect of DCEBIO on SK2 channels, in view of the essential role played by the SK2 (K Ca 2.2) subunits in mediating I AHP in hippocampal pyramidal neurons (5,7,9). We then compared the effect of DCEBIO to that of NS309 and 1-EBIO on SK2 channels. The three enhancers augment SK channel activity by increasing the apparent sensitivity to Ca 2ϩ , and concentration-response experiments were therefore performed in the inside-out configuration, which allows full control of the free [Ca 2ϩ ]. Activation curves yielded an EC 50 of 0.42 M for Ca 2ϩ (n ϭ 8) (data not shown) with maximal activation obtained at 10 M free Ca 2ϩ . The concentration-response curves for the three enhancers were performed at 200 nM free Ca 2ϩ , which activated 5% (0.048 Ϯ 0.009) of the maximal SK current. Fig. 1A shows the control currents (Ctrl) as well as the currents obtained in the presence of the enhancers upon application of voltage ramps from Ϫ80 to ϩ80 mV. For all compounds, potentiation was concentration-dependent but not voltage-dependent. The traces shown in Fig. 1A were obtained from experiments similar to the one shown in Fig. 1B, where the current at Ϫ75 mV is depicted as a function of time. The inside-out patch was exposed first to 0.01 M Ca 2ϩ to determine the background current level and subsequently to 10 M Ca 2ϩ to define the maximal current, which was used to normalize the currents. The concentrationresponse for DCEBIO was then determined at the subthreshold Ca 2ϩ concentration of 200 nM, with control of maximal and background currents at the end of the experiment (Fig. 1B). The currents measured at the steady-state level of activation were plotted as a function of the  The effects (measured at the end of the step to Ϫ30 mV) were normalized to the control current, with no effect equal to 100%. E, whole-cell currents measured from a HEK293 cell co-expressing Kv7.2/7.3 channel subunits. Each panel shows the current just before (Control) and after perfusion with DCEBIO (100 M), retigabine (3 M), 1-EBIO (1 mM), or NS309 (10 M). Kv7.2/7.3 currents were activated by a 1-slong step to Ϫ30 mV, and deactivation was followed by a subsequent step to Ϫ60 mV. The holding potential was Ϫ90 mV, and the protocol was applied every 10 s. F, Kv7.2/7.3 current at Ϫ60 mV (circles) and Ϫ30 mV (squares) obtained from voltage steps as in D, depicted as a function of time. The cell was exposed to the compounds during the periods indicated by the bars. 0.2% Me 2 SO was tested as a vehicle for the 1 mM 1-EBIO application.
M; n ϭ 1.6). Furthermore, all three compounds were found to have an efficacy of 100% with respect to saturating [Ca 2ϩ ].
Kv7 channels underlie I M , a current contributing to the generation of the mAHP in hippocampal neurons (26 -28). We have therefore tested the SK channel enhancers on recombinant Kv7.2/7.3 channels. Fig. 1, D-F, illustrates the lack of effect of DCEBIO (100 M), NS309 (10 M), and 1-EBIO (1 mM) on these channels at the highest concentrations used for recordings in brain slices in this study. Fig. 1E shows the Kv7.2/ 7.3 currents activated by a 1-s-long step to Ϫ30 mV before and after the addition of the SK enhancers, as well as the reference Kv7 activator retigabine (3 M). The time course in Fig. 1F depicts the current at the end of the step to Ϫ30 mV (open circles) and at the end of the step to Ϫ60 mV (open boxes). The bar diagram in Fig. 1D summarizes the effects from 5-6 experiments and shows that neither DCEBIO nor NS309 nor 1-EBIO significantly affect the Kv7.2/7.3 current, even at the highest concentrations tested.
DC-EBIO and NS309 Selectively Increase I AHP but Not sI AHP in Hippocampal Pyramidal Neurons-When tested on CA1 pyramidal neurons in acute hippocampal slices, DCEBIO potentiated the SK-mediated Ca 2ϩ -activated K ϩ current I AHP without affecting the sI AHP at all concentrations tested (10 and 50 M; n ϭ 4) (data not shown) (100 M; n ϭ 7) (Fig. 2). In particular, 100 M DCEBIO increased the amplitude of the SK-mediated I AHP by 195 Ϯ 40% and prolonged its duration by 319 Ϯ 42%. sI AHP co-exists with I AHP in CA1 pyramidal neurons but is mediated by channels clearly distinct from the SK channels underlying I AHP and of as yet unknown molecular identity (2)(3)(4)(5)9). Neither the sI AHP amplitude (118 Ϯ 10%) nor its time constant of decay (112 Ϯ 5%) were significantly increased by DCEBIO (Fig. 2). The remarkable potentiation of I AHP led to an increase of the charge transfer, estimated as the integral of the two AHP currents (I AHP ϩ sI AHP ), by 154 Ϯ 16% (n ϭ 7) (Fig. 2C). The specificity of the DCEBIO effect on neuronal SK channels was confirmed by the full block of the enhanced I AHP upon application of the SK channel blocker d-tubocurarine (curare, 100 M) (Fig. 2A). sI AHP was instead identified by its suppression by noradrenaline (1 M) ( Fig. 2A), known to inhibit this current by activating the cAMP/protein kinase A pathway in hippocampal neurons (23,29). These results demonstrate that DCEBIO is a more potent enhancer of both recombinant and neuronal SK currents when compared with 1-EBIO.
The rest of our study focuses on NS309, which displays an enhanced potency on recombinant SK2 channels compared with both 1-EBIO and DCEBIO (Fig. 1C). A crucial question is how effective and selective this compound is on SK channels in their native, neuronal environment. When tested on I AHP and sI AHP in CA1 pyramidal neurons in hippocampal slices, 10 M NS309 induced a marked increase of I AHP amplitude with respect to the control currents recorded prior to application of the compound (Fig. 3A). I AHP was measured in isolation, upon inhibition of sI AHP by the cAMP analogue 8CPT-cAMP (50 M). The relative increase in amplitude of the apamin-sensitive I AHP was 182 Ϯ 22% (n ϭ 8) (Fig. 3, A and C). NS309 had an even more prominent effect on the time constant of deactivation of I AHP , which was slowed by ϳ6-fold, changing from 119 Ϯ 19 ms to 654 Ϯ 77 ms after NS309 application (n ϭ 8). As a consequence, the charge transfer of I AHP measured as the integral of the current was increased by almost 10-fold (949 Ϯ 165%; n ϭ 8) (Fig. 3, A-C). By comparison, the application of the same concentration of DCEBIO (10 M) resulted in an increase of the I AHP amplitude by 158 Ϯ 20% and in an increase of its decay time constant by 149 Ϯ 20% (n ϭ 4) (data not shown). NS309 was applied for several minutes to yield a maximal and stable potentiation of I AHP , as illustrated by the time course of action of this drug (Fig. 3B). The augmentation of I AHP by NS309 was only scarcely reversible, even after prolonged wash out periods (data not shown). The application of NS309 (10 M) did not affect the input resistance of the neurons (n ϭ 18).
At a lower concentration (1 M), NS309 had qualitatively similar but slightly less pronounced effects on both amplitude and charge transfer of I AHP . Thus, I AHP amplitude was increased to 139 Ϯ 10% (n ϭ 3) (Fig.  3, D and F). Similar to what was previously observed at higher concentrations, 1 M NS309 slowed the deactivation of I AHP by ϳ4-fold, increasing its time constant of decay () from 86 Ϯ 9 ms to 361 Ϯ 39 ms (n ϭ 3). As a consequence, the I AHP charge transfer increased by ϳ5-fold (500 Ϯ 18%; n ϭ 3) (Fig. 3, D-F). Also at 1 M, the action of NS309 developed slowly (Fig. 3E). The effect of NS309 can be entirely ascribed to an enhancement in the activity of the SK channels underlying I AHP , as the NS309-enhanced current was fully blocked by the SK channel blockers d-tubocurarine (d-TC, curare; 200 M; n ϭ 8) (Fig. 3,  A and B) and apamin (25 nM; n ϭ 3) (Fig. 3, D and E).
Next, to investigate the effects of NS309 (10 M) on the apamininsensitive AHP current sI AHP , we applied the compound together with 25 nM apamin to block I AHP . NS309 affected neither the amplitude (96 Ϯ 13%; p ϭ 0.63; n ϭ 3) (Fig. 4) nor the charge transfer (108 Ϯ 15%; p ϭ 0.67; n ϭ 3; Fig. 4, A and C) of the sI AHP once it had reached steady state. Furthermore, after NS309 application, sI AHP was fully inhibited by 2.5 M noradrenaline (Fig. 4, A and B). This result underscores the selective nature of NS309 as an SK channel enhancer in contrast to 1-EBIO, which increased both I AHP and sI AHP in CA1 pyramidal neurons (14). Thus, we can conclude that NS309 is a selective enhancer of the SK channels mediating I AHP in hippocampal neurons and more potent than both 1-EBIO and DCEBIO.

NS309 Alters the Firing Pattern of Hippocampal Pyramidal Neurons-
The selective potentiation of I AHP , without any effect on sI AHP , produced by NS309 prompted us to use this compound to test the specific impact of SK channel enhancement on the firing properties of hippocampal neurons. This was investigated in current clamp recordings performed in the absence and presence of NS309 (10 M). Under control conditions, depolarizing current pulses elicited trains of action potentials characterized by early and late spike frequency adaptation (Fig. 5A, left panel). Application of NS309 decreased the firing frequency of all CA1 pyramidal neurons tested. As shown in Fig. 5A, middle panel, in the presence of NS309, most cells fired only 3-4 action potentials in a burst-like fashion followed by a prolonged silent phase. This marked effect of NS309 was fully counteracted upon application of the SK channel blocker apamin (25 nM) (Fig. 5A, right panel). In the same cells, the effect on the firing pattern coincided with a pro-nounced enhancement of the medium AHP (Fig. 5B, mAHP), whereas the slow AHP (Fig. 5B, sAHP) was not affected (n ϭ 3), in accordance with the lack of effect of NS309 on sI AHP observed in voltage clamp recordings (Fig.  4).
The NS309-mediated specific enhancement of I AHP amplitude and duration, which affects in turn the medium and late phases of spike frequency adaptation, reveals a prominent effect of SK channels on neuronal excitability and signal encoding properties.
Does the Potentiation of I AHP by NS309 Depend on Frequency?-The slow time course of action of NS309, with time constants in the range of 6 -8 min to reach the maximal effect on I AHP , raised the question as to whether the effect of this compound on I AHP is frequency-dependent. To test this hypothesis, we designed three experimental paradigms. In the first paradigm, short current injections (5 ms long) of sufficient intensity to elicit action potentials were applied in trains of eight at a frequency of 10 Hz (Fig. 6A) (n ϭ 5) or 33 Hz (Fig. 6B) (n ϭ 5) in current clamp experiments. This stimulation pattern mimics the physiological input from bursting CA3 neurons to CA1 neurons (30,31). Upon appli- cation of NS309 (10 M), the mAHP following the single action potentials and the trains of spikes were increased both in amplitude and duration (Fig. 6, A and B, right panels). The increase in mAHP amplitude was not significantly different at 10 Hz compared with 33 Hz (Fig. 6B, left  panel inset). However, in two of the cells stimulated at 33 Hz, the increase of mAHP after the application of NS309 was large enough to prevent firing in response to some of the current injections within the train (not shown).
In the second paradigm, depolarizing steps (100 ms long to ϩ10 mV) were delivered in the presence of tetrodotoxin and tetraethylammonium to elicit Ca 2ϩ action currents in voltage clamp every 30 s (0.033 Hz; n ϭ 8) (Fig. 6, C-E) or 6 s (0.167 Hz; n ϭ 5) (Fig. 6, F-H). Fig. 6, C and F, shows that at both frequencies, NS309 (10 M) caused a substantial increase in the amplitude and duration of I AHP and that the enhanced current was fully suppressed by the application of curare (Fig.  6, C and D, d-TC) (200 M) or apamin (50 nM) Fig. 6, F and G). Neither the enhancement of I AHP amplitude (Fig. 6E) nor its charge transfer (Fig.  6H) nor the time course of potentiation of I AHP by NS309 (Fig. 6I) were significantly different at 0.033 Hz compared with 0.167 Hz.
Finally, the third paradigm comprised trains of eight short depolarizing steps (5 ms to ϩ10 mV) delivered either at 10 or 33 Hz in voltage clamp from a holding potential of Ϫ40 and Ϫ50 mV to mimic the stimulation pattern provided by bursts of action potentials but in voltage clamp recordings. Upon NS309 (10 M) application, neither the time course of I AHP potentiation (Fig. 6J), nor the enhancement in amplitude and charge transfer of I AHP (not shown) were significantly different at the two frequencies tested. In conclusion, our experimental evidence does not support a frequency-dependent component in the action of NS309 on I AHP in CA1 pyramidal neurons.

DISCUSSION
The results obtained in this study show that, in accordance with their effects on recombinant channels, DCEBIO and even more prominently NS309 are potent and specific enhancers of the activity of neuronal SK channels. In particular, the SK-mediated I AHP was increased in amplitude and even more remarkably in duration by the application of concentrations as low as 1 M NS309 (Fig. 3). NS309 did not influence the currents mediated by recombinant BK channels Kv7.4 (KCNQ4) (20) or Kv7.2/7.3 channels (this study) expressed in HEK293 cells. In particular, the lack of effect of NS309 on Kv7.2/7.3 channels supports the hypothesis that the enhancement of the mAHP observed in this study (Fig. 5B) is not because of an increase of I M , a voltage-dependent current contributing to the generation of the mAHP (26), but solely of I AHP . The selectivity of the effects of DCEBIO and NS309 on neuronal SK channels is further supported by the full suppression of the enhanced I AHP and mAHP by the SK channel blockers apamin and d-tubocurarine (Figs. 2,  3, 5, and 6). These results are in good agreement with the reported enhancement of recombinant SK channels by NS309, whereby the enhanced currents were completely blocked by apamin in a heterolo- The right panel shows a superimposition of sI AHP traces before (black) and after (gray) the application of 10 M NS309, emphasizing the lack of effect of this compound on the peak amplitude and time course of sI AHP . B, lack of effect of NS309 (10 M) on the peak amplitude (ampl.) of sI AHP plotted against time. NS309 was applied for 18 min after the sI AHP amplitude had stabilized and produced no effect on the sI AHP amplitude. The subsequent application of noradrenaline (NA; 2.5 M) induced a complete suppression of sI AHP . The time course is from the same representative cell shown in A. C, bar diagram showing that 10 M NS309 had no effect on the sI AHP peak amplitude and charge transfer. FIGURE 5. NS309 reduces the firing frequency of CA1 pyramidal neurons. A, trains of action potentials elicited by 800 ms current injections (260 pA) from the resting membrane potential (-57 mV; no steady current injected) before and after the application of 10 M NS309. In this representative cell, the number of action potentials elicited by the same current injection was decreased in the presence of NS309, and the late phase of spike frequency adaptation was enhanced. These effects were reversed upon application of apamin (25 nM; right panel), which produced an additional small increase in the number and initial frequency of the action potentials in the train, as previously reported (5). NS309 did not affect the resting membrane potential or the input resistance of the cell. Similar results were obtained in three cells. B, afterhyperpolarizations (medium AHP, mAHP; slow AHP, sAHP) following a burst of action potentials triggered by a 200 ms current injection (500 pA) before and after the application of 10 M NS309. NS309 enhanced the mAHP (superimposed traces in right panel) without affecting the sAHP. The effect of NS309 on mAHP was fully counteracted upon application of apamin (25 nM), in agreement with the results obtained on the underlying currents (I AHP and sI AHP ) in voltage clamp recordings. Action potentials were truncated for better resolution of the after hyperpolarizations following the bursts. Similar results were obtained in three cells. DECEMBER 16, 2005 • VOLUME 280 • NUMBER 50  (Apa; 50 nM) in G, which completely blocked the enhanced I AHP . The plots are from the same representative cells shown in C (plot D) and in F (plot G). E and H, bar diagrams summarizing the effects of 10 M NS309 on the I AHP peak amplitude (E) and charge transfer (H) in eight and five cells stimulated at 0.033 and 0.167 Hz, respectively. No significant differences were observed in the potentiation of I AHP at these two frequencies. I, the time course of the potentiating action of 10 M NS309 (see, for example, plots D and G) on the I AHP peak amplitude (left bars) and charge transfer (right bars) were fitted with exponential functions, and the corresponding time constants were obtained for seven cells stimulated at 0.033 Hz and four cells at 0.167 Hz (same protocol as in C and F). The time required by NS309 to have a full effect on I AHP was not significantly different at the two stimulation frequencies. J, the time course of the potentiating action of 10 M NS309 on the I AHP peak amplitude (left bars) and charge transfer (right bars) were fitted with exponential functions, and the corresponding time constants were obtained. I AHP was elicited using a sequence of eight short (5 ms) pulses to ϩ10 mV (holding potential ϭ Ϫ50 mV) delivered in four cells at 10 Hz and five cells at 33 Hz. The time required by NS309 to have a full effect on I AHP was not significantly different at the two stimulation frequencies.

NS309 and DCEBIO on CA1 Neurons
gous expression system (20). Additionally, we have not observed any effect of DCEBIO and NS309 on the amplitude or time course of sI AHP in CA1 neurons (Figs. 2 and 4). These results further strengthen the notion that sI AHP is mediated by a conductance clearly distinct from the SK channels, which are unlikely to contribute to its generation, as supported by differences in their kinetics, pharmacology, distribution, and by recent data obtained from genetically modified animals missing specific SK channel subunits (1-4, 9, 10). NS309 has been reported to block recombinant hERG channels with a K i value of 1.3 M (20), and hERG channels have been suggested to play a role in spike frequency adaptation in dorsal root ganglion-neuroblastoma hybrid cells (25). However, the blocking of hERG channels by NS309 would produce an effect opposite of what we observed on the firing pattern of CA1 neurons. Additionally, the changes in the firing pattern observed after the application of NS309 were fully reversed by the specific SK channel blocker apamin (Fig. 5), making it unlikely for conductances other than SK to contribute to the observed NS309 effects. Finally, even at high concentrations (10 M), NS309 did not change the membrane resistance of the CA1 neurons (n ϭ 18) (data not shown).
When compared with the first SK channel enhancer tested on neurons, 1-EBIO (14), DCEBIO and NS309 were more potent, as expected from the 17-fold difference in the potency of 1-EBIO and DCEBIO, and the 731-fold difference of 1-EBIO and NS309 on recombinant SK2 channels reported in this study. Additionally, both DCEBIO and NS309 were more selective, as they enhanced I AHP without affecting sI AHP , whereas 1-EBIO produced a small but significant enhancement also of sI AHP (14). It could be argued that the increased selectivity of DCEBIO and NS309 on I AHP is because of their higher potency on SK channels, allowing us to use lower concentrations of these compounds on hippocampal neurons compared with 1-EBIO. However, when 1-EBIO and DCEBIO were tested at the same concentration (100 M) on CA1 pyramidal neurons, 1-EBIO caused a significant increase of sI AHP amplitude (1.3-fold, n ϭ 6), 7 whereas DCEBIO did not affect sI AHP (n ϭ 7) (Fig. 2C). Nonetheless, we cannot exclude that, at concentrations higher than those tested in this study, DCEBIO and NS309 could affect also sI AHP . Finally, we have observed a difference between the actions of NS309 and of 1-EBIO and DCEBIO on I AHP , in that NS309 has a more pronounced effect on the time course rather than the peak amplitude of the current when compared with 1-EBIO and DCEBIO. Although all three enhancers act in a Ca 2ϩ -dependent manner, further experiments are needed to clarify whether the pronounced change in the time course of I AHP is the result of a direct effect of NS309 on the SK single channel kinetics or of changes in the dynamics of intracellular Ca 2ϩ in CA1 neurons. NS309 is approximately equipotent on the different SK channel subtypes (homomeric hSK1, hSK2, and hSK3 channels, as well as rat SK2 and SK3 channels) expressed in HEK293 cells (20). NS309 therefore does not provide indications on the subunit composition of the native hippocampal SK channels, which are likely to be SK2 homomeric channels or heteromers, including the SK2 subunit (5,7,9).
The time course of action of NS309 on I AHP in CA1 neurons is ϳ10fold slower than that of 1-EBIO, whereas the effect of both compounds on recombinant SK2 channels in inside-out patches are very fast (com-parable with DCEBIO as shown in Fig. 1). The results of our experiments using different stimulation frequencies to elicit I AHP (Fig. 6) argue against a use dependence of the NS309 effect on I AHP . The slow time course of action of NS309 when compared with 1-EBIO might instead be due to a slower penetration of NS309 through the slice.
The potent and specific effect of NS309 on the excitability and firing behavior of CA1 pyramidal neurons through the enhancement of SK channel activity makes this compound a better tool than 1-EBIO and DCEBIO to assess the role of these channels as possible targets for the treatment of disorders linked to neuronal hyperexcitability.