| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Physiology, Institute of Medical Sciences, Hebrew University–Hadassah Faculty of Medicine, Jerusalem, Israel
Submitted 19 December 2005; accepted in final form 22 February 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
), including CA1 pyramidal cells (Halliwell and Adams 1982). Because of its particular properties, it has been assumed since its discovery that IM serves to limit sustained neuronal excitation (Madison and Nicoll 1984). IM is suppressed by muscarinic agonists, although many other neurotransmitters converge to up- or downregulate it by a variety of second-messenger cascades (Marrion 1997). A large body of evidence now suggests that native neuronal M-channels are formed by heteromeric assemblies of Kv7 (previously designated KCNQ) subunits, particularly Kv7.2 with Kv7.3, and perhaps also with Kv7.5 (Peters et al. 2005; Schroeder et al. 2000; Shah et al. 2002; Wang et al. 1998).
Several drugs that selectively modulate both heterologously expressed and native Kv7/M channels have recently been developed (Brown and Yu 2000). Most notably, the ‘cognition enhancers’ linopirdine and XE991 were shown to block IM (Aiken et al. 1995; Schnee and Brown 1998; Wang et al. 1998), whereas the anticonvulsant drug retigabine was found to increase IM (Rundfeldt and Netzer 2000; Tatulian et al. 2001; Wickenden et al. 2000). The availability of these drugs has permitted a more precise examination of the role of IM in controlling neuronal excitability (Aiken et al. 1995; Hu et al. 2002; Yue and Yaari 2004; Gu et al. 2005). Using these drugs to modify IM in CA1 pyramidal cells, we showed recently that this current normally controls the size and duration of the somatic spike afterdepolarization (ADP) (Yue and Yaari 2004). Blocking IM caused uncontrolled growth of the spike ADP, leading to bursting in most neurons. Conversely, augmenting this current reduced the spike ADP and induced regular firing in bursting neurons. These observations, however, did not disclose the location of the Kv7/M channels that regulate the somatic spike ADP and thus the firing mode of these neurons. Although a recent immunohistochemical study has suggested that Kv7.2 subunits aggregate in the axon initial segment of CA1 pyramidal cells (quite often in conjunction with Kv7.3; Devaux et al. 2004), diffuse staining for Kv7.2, Kv7.3, and Kv7.5 was also found throughout the somatodendritic membrane of these neurons (Devaux et al. 2004; Shah et al. 2002). Yet, whether these subunits assemble to form functional Kv7/M channels in both proximal and distal portions of the neuron has not been clearly determined to date.
Here we used intracellular recordings combined with local application of Kv7/M channel modulators to the axo-somatic region and apical dendrites of adult CA1 pyramidal cells. We sought to determine the presence and characterize the function of Kv7/M channels in these two neuronal compartments. We report that both compartments express functional Kv7/M channels that synergistically control the spike output of these neurons, albeit in different ways.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
All experimental protocols were approved by the Hebrew University Animal Care and Use Committee. Transverse hippocampal slices were prepared from adult Sabra rats (8–12 wk old). Animals were anesthetized with ether or isoflurane (3–4%) and decapitated with a guillotine. The brain was removed and immediately immersed in ice-cold oxygenated (95% O2-5% CO2) dissection artificial cerebrospinal fluid (ACSF). The caudal two thirds of one hemisphere (containing one hippocampus) were glued to the stage of a vibratome (Leica, Wetzlar, Germany). Transverse brain slices (400 µm thick) were cut from the region of the hemisphere containing the anterior hippocampus and the hippocampal portion was dissected out of each brain slice. Routinely, the CA3 region was cut away. This was done to avoid spread of spontaneous burst activity from CA3 to CA1 in experiments using K+ channel blockers, although this problem was also prevented in most experiments by adding blockers of synaptic transmission to the ACSF (see following text). The residual hippocampal slices (consisting mainly of CA1 and dentate gyrus) were transferred to an incubation chamber containing oxygenated saline at room temperature (21–24°C), where they were allowed to recover for 
1 h. The slices were transferred one at a time to an interface slice chamber and perfused from below with oxygenated ACSF at 33.5°C. The upper surface of the slices was exposed to the humidified gas mixture.
The standard ACSF contained (in mM): NaCl 124, KCl 3.5, MgCl2 2, CaCl2 1.6, NaHCO3 26, and D-glucose 10 (pH 7.3). For some experiments 1 mM NiCl2 was added to the ACSF and the concentration of MgCl2 was reduced to 1 mM to maintain a constant concentration of divalent cations. In most experiments, the ACSFs also contained the glutamate receptor antagonists 6-cyano-7-nitro-quinoxaline-2,3-dione (CNQX; 15 µM) and 2-amino-5-phosphono-valeric acid (APV; 50 µM) to block fast excitatory postsynaptic potentials, and the 
-aminobutyric acid type A (GABAA) receptor antagonist picrotoxin (100 µM) to block fast inhibitory postsynaptic potentials. Other drugs were added to the ACSF as indicated.
Electrophysiological methods
Current-clamp recordings from the somata of pyramidal cells in the CA1 pyramidal layer were made using sharp, K+-acetate-filled (4 M) glass microelectrodes (70–100 M
). An active bridge circuit in the amplifier (Axoclamp 2A, Axon Instruments) allowed simultaneous injection of current and measurement of membrane potential. The bridge balance was carefully monitored and adjusted before each measurement. The pyramidal cells accepted for this study had stable resting potentials of at least –60 mV and overshooting action potentials.
Drug applications
Drugs were applied to the whole slice by bath application or to discrete locations in the slice using a puff application system. Because the spread of drugs in slices maintained in an interface chamber is predominantly by diffusion, the onset and offset of drug effects are slow (in the order of tens of minutes). For all bath-applied drugs used in this study, maximal effects were obtained within 20–30 min of adding them to the ACSF.
For focal drug applications we used a puffing system consisting of a pneumatic pump (Picospritzer III, General Valve, Fairfield, NJ) connected to a patch pipette (tip diameter about 10 µm) filled with ACSF and the drugs intended for focal application. The tip of the pipette touched the upper surface of the slice. In control experiments the intensity and duration of the pressure pulses were adjusted to produce a drop that covered a circular area about 50 µm in diameter when ejected onto the surface of the slice (visualized by including fast-green dye in the pipette solution). This yielded pressure pulses of 5 bars for 20 ms, which were used for all applications. For somatic drug applications, the puffing pipette was positioned in stratum pyramidale about 25 µm from the recording microelectrode (lodged in the soma of the neuron). To apply drugs to the distal apical dendrites, the puffing pipette was positioned in stratum radiatum about 200–300 µm away from, and vertical to, the stratum pyramidale (the experimental arrangement is shown in Fig. 1A). The effects of focally applied drugs, when present, appeared usually after 3 min of puffing, and attained maximum within 8–10 min. These delays undoubtedly arise from the slow spread of the drugs from their site of application into the slice. We found that to replicate effects of bath-applied drugs, the optimal drug concentrations in our pressure application pipettes generally should be tenfold higher than those used in bath applications.
|
All chemicals and drugs were purchased from Sigma Chemicals (Rehovot, Israel) except for CNQX (RBI, Natick, MA), XE991 (Tocris Cookson, Bristol, UK), and retigabine (kind gift of Dr. J. B. Jensen, NeuroSearch, Ballerup, Denmark). Stock solutions (10 mM) of linopirdine (in ethanol), XE991, retigabine, and 4
-phorbol 12,13-dibutyrate (PDB) (in dimethylsulfoxide) were stored at –20°C. Ethanol and dimethylsulfoxide added to the standard ACSF to a concentration of 0.001% had no effects on the measured spike parameters.
Data measurement and analysis
The intracellular signals were digitized, stored, and analyzed using a Pentium computer and pCLAMP9 data acquisition system (Axon Instruments). Apparent input resistance (RN) was measured from voltage deflections induced by small (50- to 100-pA) 200-ms negative current pulses. It should be remembered, however, that RN measurements in CA1 pyramidal cells are confounded by the presence of other conductances that are active around the resting potential, particularly the hyperpolarization-activated cation conductance (Ih; Surges et al. 2004).
The size of the spike ADP was measured as the integrated "area under the curve" between the fast afterhyperpolarization (fAHP) and the point at which membrane voltage returned to resting potential. The amplitude of Ca2+ spikes was measured as the voltage difference between the peak of the spike and the resting membrane potential. For measuring changes in RN, spike waveform and neuronal discharge behavior after bath or local drug applications, drug-induced changes in resting membrane potential were counteracted by steady injections of appropriate negative or positive currents. Averaged data are expressed as mean ± SE. The significance of the differences between the measured parameters was evaluated using, as required, Student's unpaired or paired t-test. Multiple comparisons were done with two-way ANOVAs, followed by Bonferroni post hoc tests. The significance level in all tests was set to P < 0.05. Correlation between two data sets was expressed using Pearson's coefficient of correlation (–1 
r 1).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
To characterize the distinct roles of axo-somatic and apical dendritic Kv7/M channels in control of neuronal excitability, we first examined the effects of the Kv7/M channel modulators applied focally to either the distal or proximal parts of CA1 pyramidal cells. In separate experiments we examined the effects of the Kv7/M channel blockers linopirdine and XE991 and of the Kv7/M channel opener retigabine. The application pipette contained either 100 µM linopirdine, 30 µM XE991, or 100 µM retigabine dissolved in standard ACSF. The experimental arrangement is illustrated in Fig. 1A. In these and in later experiments, the effects of the focally applied drugs were monitored for 10 min. If a single puff of a drug yielded no obvious effect within the first 3 min, the drug was puffed twice again in immediate succession at the same region (altogether three puffs). If a clear response was observed within the first 3 min, no more puffs were applied in that region. In six control experiments, puffing ACSF without drugs onto either proximal or distal portions of the neurons had no effects on their passive (resting potential and RN) and active membrane properties (spike threshold, amplitude, and ADP).
Applications of linopirdine to the apical dendrites did not significantly affect resting membrane potential and RN recorded in the soma (Table 1; Fig. 1, Ba and Bb). However, when applied to the axo-somatic region (one puff), linopirdine depolarized most neurons (five of six) by 1 to 6 mV. On average, this effect amounted to 3.3 ± 0.6 mV (P < 0.05; Table 1). Concurrently, linopirdine applied to the axo-soma increased RN (measured from control resting potential; see METHODS) by 37.0 ± 12.6% (P < 0.05; Table 1; Fig. 1, Bb and Bc). Similar results were obtained with XE991. Its application to the distal dendrites affected neither resting membrane potential nor RN (P < 0.05; Table 1). However, when applied to the axo-somatic region (one puff), XE991 depolarized the neurons by 1 to 10 mV. On average, this effect amounted to 4.3 ± 0.8 mV (P < 0.05; Table 1). Concurrently, XE991 increased RN by 30 ± 10.9% (P < 0.05; Table 1).
|
). Our observations suggest that deactivation of IM during hyperpolarization from resting potential also contributes appreciably to these phenomena.
Retigabine applied to the apical dendrites also had no effect on resting potential and RN (P < 0.05; Table 1; Fig. 1, Ca and Cb). When applied to the axo-soma (one puff), however, retigabine hyperpolarized the neurons by 1 to 9 mV; the average hyperpolarization was 5.0 ± 0.8 mV (P < 0.05; Table 1). Concurrently, RN decreased by 49.1 ± 4.7% (P < 0.05; Table 1; Fig. 1, Cb and Cc). In addition to that, retigabine applied to the axo-soma markedly augmented the voltage sags during hyperpolarizations and rebound depolarizations (Fig. 1C, compare traces in c to those of similar size in a and b). These effects are expected given that retigabine shifts the IM activation curve in the hyperpolarizing direction by about 20 mV (Main et al. 2000; Tatulian et al. 2001).
The ineffectiveness of distal applications of Kv7/M channel modulators in altering resting membrane potential and RN conceivably may arise from insufficient drug delivery to the apical dendrites. However, in another series of experiments (described below; Figs. 9 and 10), we found that similar distal applications of linopirdine, XE991, and retigabine modulate dendritic Ca2+ spikes and Ca2+-dependent bursting in the same way they do when applied in the bath in micromolar concentrations. In those experiments also, the drugs did not alter resting membrane potential or RN. Notwithstanding, we repeated the dendritic application experiments (three puffs) with even higher concentrations of linopirdine (500 µM), XE991 (300 µM), and retigabine (500 µM) in the application pipette. Each drug was tested in three neurons. Again, there was no change in resting membrane potential and RN within the 10-min observation period after puffing the drugs onto the apical dendrites.
|
|
). Effects of focally applied Kv7/M channel modulators on somatic spike ADP
We previously showed that linopirdine, XE991, and retigabine strongly modify action potential generation in CA1 pyramidal cells when applied in the bath (Yue and Yaari 2004). Most remarkably, linopirdine (0.3–10 µM) and XE991 (0.1—3 µM) dose-dependently augmented the somatic spike ADP, leading to the generation of spike bursts at the higher concentrations. Retigabine (1–10 µM) exerted the converse effect, reducing the size of the spike ADP, while augmenting a partially overlapping afterhyperpolarization (AHP) of medium duration (mAHP). The effects of these drugs on the spike ADP were not secondary to changes in membrane potential because they also occurred when the membrane potential was manually clamped to its natural resting value (Yue and Yaari 2004). Together, these findings suggested that Kv7/M channels control the spike output of CA1 pyramidal cells by limiting the size of the somatic spike ADP.
To evaluate the contribution of axo-somatic and dendritic Kv7/M channels to the control of the somatic spike ADP, we examined the effects of the three drugs when applied focally to axo-soma or apical dendrites. Solitary spikes were evoked by brief (4-ms) positive current pulses from the resting membrane potential recorded in control conditions. Representative results obtained in two experiments with linopirdine are illustrated in Fig. 2, A and B. In the first experiment (Fig. 2A), linopirdine was first applied repetitively (three puffs) to the apical dendrites and 15 min later to the axo-soma (one puff). Dendritic application of linopirdine had no effect on the spike ADP (Fig. 2, Aa and Ab, and traces a and b in Fig. 2Ae), whereas its application to the axo-soma (one puff) appreciably augmented the spike ADP until it triggered a burst (Fig. 2, Ac and Ad, and traces c and d in Fig. 2Ae). In the second experiment (Fig. 2B) the order of applications was reversed. Again, axo-somatic application (one puff) of linopirdine caused considerable augmentation of the spike ADP and induced bursting (Fig. 2, Ba–Bc, and traces a–c in Fig. 2Be). Dendritic application of linopirdine after recovery of this effect (trace not shown) did not modify the spike ADP (Fig. 2Bd and trace d in Fig. 2Be). In a series of 16 experiments, the spike ADP size (measured by integrating the voltage trajectory over time; see METHODS) was 151.8 ± 13.9 mV · ms in control, 151.0 ± 19.5 mV · ms after dendritic application, and 318.7 ± 16.1 mV · ms (P < 0.05) after axo-somatic application of linopirdine. The average increase in ADP size after axo-somatic linopirdine application (measured before the appearance of bursting) was 141.6 ± 25.5%. In all but two neurons (14 of 16) the augmented ADPs ultimately progressed to full-blown bursts consisting of three to eight spikes. The mean number of intraburst spikes in the 14 bursting neurons was 4.1 ± 0.3.
|
Finally, we tested the effects of focally applied retigabine. Application of retigabine to the apical dendrites (three puffs) had variable effects. In some neurons the spike ADP was not affected (Fig. 2, Da and Db and traces a and b in Fig. 2De), whereas in others it was somewhat reduced. On average, dendritic application of retigabine reduced the spike ADP from 171.5 ± 28.4 to 140.3 ± 24.4 mV · ms (n = 8; P < 0.05). When subsequently applied to the axo-somatic region (one puff), however, retigabine consistently and robustly reduced the spike ADP to 65.1 ± 14.5 mV · ms (n = 8; P < 0.05), converting it to a conspicuous mAHP (Fig. 2, Dc and Dd and traces c and d in Fig. 2De). The average decreases in ADP size after dendritic and axo-somatic applications were 18.5 ± 4.7 and 62.4 ± 6.2%, respectively.
The effects of focally applied Kv7/M modulators on the size of the somatic spike ADP are summarized in Fig. 2E. Together, they suggest that the somatic spike ADP is tightly regulated by Kv7/M channels residing in the axo-somatic region of the neuron. During the spike ADP the activity of these channels slowly increases, thereby facilitating the repolarization of the neuron. The small but significant reduction of somatic spike ADP by distally applied retigabine (Fig. 2, Da, Db, and E) suggests that functional Kv7/M channels are expressed in apical dendrites and may be recruited by the backpropagating spike. These channels, however, do not appear to limit the somatic spike ADP in ordinary conditions, as evident from the complete lack of effect of distally applied linopirdine and XE991 (Fig. 2, Aa, Ab, Ca, Cb, and E).
The increases in spike ADP size and in RN induced by axo-somatic applications of linopirdine or XE991 were not positively correlated with each other. The relationship between these two variables obtained in 20 experiments with either linopirdine (n = 11) or XE991 (n = 9) is provided by the scatterplot in Fig. 2F. Assuming a linear correlation between the two variables yielded a negative and small Pearson correlation coefficient (r = –0.12; P = 0.63). Thus even though both effects of linopirdine are mediated by axo-somatic Kv7/M channel block, ADP growth is not merely the end result of the increase in RN. Given that ADP growth is mediated by a regenerative recruitment of INaP unleashed by blocking Kv7/M channels (Yue et al. 2004, 2005), the variation in the amount of ADP increase across different neurons may reflect more closely the variation in INaP densities.
Effects of focally applied Kv7/M channel modulators on spike output
In normal conditions, CA1 pyramidal cells fire trains of independent spikes when stimulated with prolonged suprathreshold positive current pulses (see, for example, Fig. 3Aa), although a tendency for early spike clustering is evident in some neurons, particularly at high stimulus intensity (Jensen et al. 1994; Sue et al. 2001). The number of evoked spikes increases with stimulus intensity and decreases with time. Here we examined how modulation of Kv7/M channel activity in the axo-soma and apical dendrites affects this spike output/stimulus intensity relationship. Representative experiments in which linopirdine, XE991, and retigabine were focally applied are shown in Fig. 3, A–C. In normal ACSF, increasing stimulus intensity (from 0 to 450 pA in 50-pA steps) caused a gradual and quite linear increase in spike output (Fig. 3, Aa, Ba, and Ca). Application of linopirdine (Fig. 3Ab) or XE991 (Fig. 3Bb) to the apical dendrites (three puffs) had no noticeable effect on this monotonic relationship. In contrast, a single puff of linopirdine (Fig. 3Ac) or XE991 (Fig. 3Bc) to the axo-soma substantially modified this relationship by inducing an all-or-none burst of three to six spikes as the neuron's minimal response to threshold-straddling stimuli. The burst was followed by a conspicuous mAHP, which caused a cessation of firing (Fig. 3, Ac and Bc). With stronger stimuli, however, the burst–mAHP complex was followed by one or more solitary spikes. Figure 1, D and E depicts the averaged plots of spike output (the number of evoked spikes) versus stimulus intensity obtained, respectively, in the experiments with linopirdine (n = 8) and XE991 (n = 9). When delivered to the axo-soma, both drugs significantly augmented spike output at all stimulus intensities (P < 0.05). However, the strongest effects were exerted at stimulus intensities that were only slightly above threshold.
|
These data show that Kv7/M channels in the axo-somatic region of CA1 pyramidal cells impose a graded spike output/stimulus intensity relationship, whereas their dendritic counterparts do not influence this relationship in ordinary conditions.
Effects of focally applied 4-aminopyridine on somatic spike ADP and spike output
For comparison, we also examined the effects of focally applied 4-aminopyridine (4AP) on the somatic spike ADP and discharge behavior of CA1 pyramidal cells. It was previously shown that millimolar 4AP induces intrinsic bursting in these neurons by blocking A-type K+ channels (A channels) in the apical dendrites (Hoffman et al. 1997). Blocking these channels facilitates spike backpropagation into the distal dendrites, thereby initiating a local dendritic Ca2+ spike. This, in turn, spreads back to the axo-soma where it augments the spike ADP to the point of bursting (Magee and Carruth 1999). Because of differences in experimental techniques, we first sought to replicate the finding that 4AP induces somatic bursting by blocking K+ channels in apical dendrites. The pressure application pipette contained 30 mM 4AP. A representative experiment (n = 3) is shown in Fig. 4A. A single puff of 4AP delivered to the apical dendrites consistently augmented the spike ADP and induced bursting (Fig. 4, Aa–Ac). In contrast, somatic application of 4AP after recovery of this effect (trace not shown) did not modify the spike ADP (Fig. 4Ad and trace d in Fig. 4Ae). We also performed similar 4AP applications in the reversed order (n = 3). When applied to the axo-soma, 4AP had no effect on the somatic spike ADP (Fig. 4, Ba and Bb and traces a and b in Fig. 4Be). When applied to the apical dendrites (one puff), however, it consistently induced ADP growth and bursting (Fig. 4, Bc and Bd and traces c and d in Fig. 4Be). The mean number of spikes in 4AP-induced bursts was 3.2 ± 0.8. Applications of 4AP had no detectible effects on resting potential and RN. These findings corroborate the notion that blocking A channels in apical dendrites, but not in the axo-soma, markedly enhances the somatic spike ADP, as shown previously (Magee and Carruth 1999). Thus the somatic spike ADP is ordinarily maintained at subthreshold voltages by the synergistic action of Kv7/M channels in the axo-soma and A channels in the apical dendrites.
|
Effects of bath-applied INaP blockers
We have shown above that Kv7/M channels provide inhibitory feedback at the axo-soma that maintains the somatic spike ADP at subthreshold potentials. But what is the source for the positive feedback underlying ADP growth once these channels are blocked? In adult CA1 pyramidal cells, the depolarizing drive for the spike ADP is thought to be provided mainly by a persistent Na+ current (INaP; Azouz et al. 1996; Su et al. 2001; Yue et al. 2005), generated by noninactivating Na+ channels (French et al. 1990). It is thus likely that blocking axo-somatic Kv7/M channels results in growth of the somatic spike ADP by unleashing the depolarizing action of INaP. This hypothesis predicts that blocking axo-somatic persistent Na+ channels would reverse the action of Kv7/M channel blockers. To test this hypothesis, we first examined the effects of three drugs that effectively block INaP in CA1 pyramidal cells—the anticonvulsant phenytoin, the neuroprotective drug riluzole, and the protein kinase C activator PDB (Yue et al. 2005)—on linopirdine-induced bursting. However, at doses that completely block INaP, these drugs also attenuate the transient Na+ current (by nearly 20%; Yue et al. 2005). Therefore we also monitored their effects on the rate of rise of the spike (which reflects the maximal transient Na+ conductance; Hodgkin and Katz 1949) as well as on other spike parameters.
Representative experiments are shown in Fig. 5, A–C. In all three neurons, 10 µM linopirdine added to the ACSF converted their single spike response to a prominent burst of five to six spikes (Fig. 5, Aa and Ab, Ba and Bb, and Ca and Cb). Bath application of phenytoin (100 µM; Fig. 5Ac), riluzole (10 µM; Fig. 5Bc), and PDB (5 µM; Fig. 5Cc) progressively suppressed the bursts. In all cases the bursts were reduced to a single spike before any changes in rise time or amplitude of the first spike were detected (insets in Fig. 5, Ac, Bc, and Cc), indicating that these drugs suppress bursting without substantially reducing transient Na+ current. The same results were consistently obtained in 17 similar experiments in which phenytoin (n = 8), riluzole (n = 4), and PDB (n = 5) were bath applied after linopirdine.
|
; Stefani et al. 1997; Yaari et al. 1987). If such an effect is exerted in adult CA1 pyramidal cells, then these drugs should also suppress 4AP-induced bursting, which is driven by T/R- and L-type Ca2+ channels in apical dendrites of CA1 pyramidal cells (Magee and Carruth 1999). Therefore we examined how these drugs affect 4AP-induced bursting. Representative results are illustrated in Fig. 5, D–F. Using the same concentrations that readily blocked linopirdine-induced bursting, neither phenytoin (100 µM; Fig. 5D; n = 4), nor riluzole (10 µM; Fig. 5E; n = 5), nor PDB (5 µM; Fig. 5F; n = 5) blocked bursting induced by 3 mM 4AP. Whereas the first two drugs had no detectible effects on this activity, PDB caused prolongation of 4AP-induced bursts (Fig. 5F). The latter effect may stem from suppression by PDB of the Ca2+-dependent slow AHP (sAHP) that underlies spike frequency adaptation (Malenka et al. 1986).
Another possible caveat in interpreting the action of riluzole is that it was shown to open big (BK) and small (SK) conductance Ca2+-activated K+ channels (e.g., Cao et al. 2002; Wu and Li 1999). Such an action, if exerted, might contribute to burst suppression by riluzole. However, riluzole (10 µM) also consistently suppressed linopirdine-induced bursting in conditions of blocked BK and SK channels (achieved by adding 100 nM iberiotoxin and 100 nM apamin to the ACSF; see Yue and Yaari 2004) (n = 3; data not shown).
Together, our findings support the notion that the growth of the spike ADP into a burst response after blockage of Kv7/M channels critically depends on recruitment of INaP.
Effects of focally applied INaP blockers
We further examined the distribution of persistent Na+ channels that mediate bursting when Kv7/M channels are blocked. To that end, we monitored the effects of focally applied riluzole and PDB on this activity. The application pipette contained 100 µM riluzole or 50 µM PDB (phenytoin was not used in these experiments because the tenfold concentration required for local application was beyond its solubility in ACSF). Representative results are shown in Fig. 6. Dendritic applications (three puffs) of riluzole (Fig. 6A) or PDB (Fig. 6B) had no effect on linopirdine-induced bursting (Fig. 6, Aa and Ab and Ba and Bb). Thus the number of intraburst spikes both before and after distal drug applications were 4.3 ± 0.5 (n = 4) for riluzole and 4.7 ± 0.7 (n = 6) for PDB. In contrast, single axo-somatic applications of these drugs consistently diminished the bursts to a single spike (Fig. 6, Ac and Bc; P < 0.05), without modifying the rise time of the first spike (Fig. 6, Ac and Bc, insets).
|
). Applying TTX to the apical dendrite (three puffs) had no effect on bursting induced by linopirdine (Fig. 6, Ca and Cb). The number of intraburst spikes both before and after distal TTX applications was 4.2 ± 0.5 (n = 4). Thus spike backpropagation is not essential for linopirdine-induced bursting. When TTX was similarly applied to the axo-soma, it completely blocked the burst (Fig. 6Cc; P < 0.05) before noticeably affecting the upstroke of the first spike (Fig. 6Cc, inset). Several minutes later the first spike was also depressed by TTX. Figure 6D summarizes the effects of distally versus proximally applied Na+ channel blockers on linopirdine-induced bursting.
|
), TTX applied to the apical dendrites (one puff) blocked bursting without affecting the first spike in all of these experiments (Fig. 6, Ea and Eb). Effects of blocking Ca2+ currents
The above findings clearly show that recruitment of proximal persistent Na+ channels plays a critical role in ADP growth and bursting when Kv7/M channels are inoperative. However, they do not exclude the participation of voltage-activated Ca2+ channels in this process. To assess the contribution of the latter channels, we examined the effects of Ni2+ on linopirdine-induced bursting. As illustrated in Fig. 7A, adding 1 mM Ni2+ to the ACSF (by equimolar substitution with Mg2+) caused progressive augmentation of the burst by reducing the repolarization capacity of the neuron (Fig. 7, Aa–Ac). Ultimately each burst transformed to a plateau depolarization lasting more than 100 ms (Fig. 7Ac, inset). Similar results were obtained in 11 experiments. The Ni2+-induced plateau potentials, as well as the initial bursts, were consistently reduced to a solitary spike by bath-applied phenytoin (100 µM; n = 3), riluzole (10 µM; n = 3), and PDB (5 µM; n = 5). An example for the latter action is shown for phenytoin in Fig. 7Ad. Applying 1 mM Ni2+ to neurons superfused with normal ACSF caused reduction in fAHP and spike broadening (Fig. 7, Ba and Bb) but did not induce bursting, as previously described (Su et al. 2001). However, subsequent addition of 10 µM linopirdine to the Ni2+-containing ACSF induced bursting followed by the development of plateau potentials (Fig. 7, Bc and Bd; n = 3). Again, the burst–plateau potential complexes were suppressed by phenytoin (Fig. 7Be).
|
In contrast to the marked facilitatory effects of Ni2+ on bursting induced by linopirdine, bursting induced by millimolar 4AP was completely suppressed by 1 mM Ni2+ applied in the bath (Fig. 7, Da and Db; n = 5) or by puffing Ni2+ only once onto the apical dendrites (Fig. 7, Ea and Eb; n = 3), as shown previously (Magee and Carruth 1999).
From the experiments with Ni2+ we conclude that recruitment of voltage-activated Ca2+ channels during linopirdine-induced bursts is mandatory for burst termination, but not for burst initiation and growth. The combined block of Kv7/M and Ca2+-activated K+ channels deprives the neuron from major repolarizing currents, leading to the development of plateau potentials driven predominantly by INaP.
Effects of focally applied Kv7/M channel modulators on dendritic Ca2+ spikes
The data gathered thus far indicated that axo-somatic Kv7/M channels critically regulate the spike output of CA1 pyramidal cells, but did not disclose any distinct function of apical dendritic Kv7/M channels. It is possible, however, that the latter channels play an important role in dendritic electrical events that arise only when the dendrites themselves are strongly depolarized, such as Ca2+ spikes (Benardo et al. 1982; Wong et al. 1979). These slow spikes do not propagate actively into the axo-soma, but may cause sufficient depolarization of that region to initiate a spike burst (Golding et al. 1999). Because the results of our experiments with retigabine suggested that apical dendrites may express functional Kv7/M channels (Fig. 2, D and E), we investigated whether these channels modulate dendritic Ca2+ electrogenesis.
Figure 8 illustrates the method we used to evoke slow Ca2+ spikes without intervening fast Na+ spikes in the apical dendrites of CA1 pyramidal cells. Perfusing the slices with ACSF containing 0.5 µM TTX and 3 mM 4AP blocked Na+ electrogenesis. In this condition strong, 180-ms-long, positive current pulses applied at the soma evoked regenerative slow spikes (Fig. 8, Aa and Ab), which were consistently and entirely blocked by adding 1 mM Ni2+ to the ACSF (Fig. 8Ac; n = 6). To validate the notion that these Ca2+ spikes recorded in the soma arise in the apical dendrites, we examined the effects of Ni2+ locally applied from a pipette containing 5 mM Ni2+. Indeed, when applied to the axo-soma, Ni2+ did not affect the Ca2+ spikes (Fig. 8, Ba and Bb), but blocked them entirely when applied (one puff) to the apical dendrites (Fig. 8Bc; n = 5). The dendritic Ca2+ spikes were resistant to 10 µM riluzole (Fig. 8, Ca and Cb). Thus Ca2+ spike amplitudes in control versus 30 min in riluzole were 66.1 ± 12.4 versus 67.8 ± 13.0 (n = 4) mV. These results further support our conclusion that riluzole suppresses linopirdine-induced bursting by blocking persistent Na+ channels. A previous study has shown that PDB also does not block dendritic Ca2+ spikes in CA1 pyramidal cells (Malenka et al. 1986).
We first tested the effects of bath applied Kv7/M channel blockers on dendritic Ca2+ electrogenesis. Representative experiments are illustrated in Fig. 9, A–C. In all these experiments the current intensity required for evoking a maximal Ca2+ spike was first determined in control conditions (i.e., in ACSF containing TTX and 4AP) and then again during exposure to the drug. In the experiment described in Fig. 9A, a stimulus intensity of 600 pA was required to evoke a Ca2+ spike in control conditions (Fig. 9Aa), but only 300 pA was necessary for evoking a Ca2+ spike during exposure to 10 µM linopirdine (Fig. 9Ab). Moreover, stimuli of larger intensity evoked maximally a single Ca2+ spike in control conditions, but two Ca2+ spikes in linopirdine-containing ACSF (Fig. 9, Aa and Ab, topmost traces). Although threshold current values for evoking Ca2+ varied with the RN of the neurons, similar effects of linopirdine were obtained in all experiments (n = 8). Thus the mean threshold current required for evoking a Ca2+ spike decreased from 566.7 ± 49.4 to 383.3 ± 30.7 pA (P < 0.05) by adding 10 µM linopirdine to the ACSF. Likewise, adding 3 µM XE991 to the ACSF reduced the stimulus intensity required for evoking Ca2+ spikes from 500.0 ± 57.7.0 to 333.3 ± 33.3 pA (n = 3; P < 0.05) and facilitated the generation of multiple Ca2+ spikes at high stimulus intensities (Fig. 9B). Conversely, adding 10 µM retigabine to the ACSF consistently suppressed the generation of all-or-none Ca2+ spikes even at the highest stimulus intensities (Fig. 9, Ca and Cb; n = 5).
The simplest explanation for these effects of linopirdine and XE991 is that the drug-induced increase in RN facilitates the depolarization of apical dendrites by currents applied at the soma. The converse action of retigabine is similarly accounted for by the drug-induced reduction in RN. To examine whether dendritic Kv7/M channels themselves regulate the transduction of depolarization to Ca2+ spike generation, we applied the three drugs focally to the apical dendrites. Indeed, as illustrated in Fig. 9D, dendritic applications (one puff) of linopirdine significantly reduced the stimulus intensity required for evoking Ca2+ spikes from 680.0 ± 97.0 to 580.0 ± 120.0 pA (n = 5; P < 0.05) and facilitated the generation of multiple Ca2+ spikes at high stimulus intensities (Fig. 9D). Similar effects were exerted by XE991 (Fig. 9E), which reduced the stimulus intensity required for evoking Ca2+ spikes from 666.7 ± 33.3 to 483.3 ± 60.1 pA (n = 4; P < 0.05). Conversely, applying retigabine to the apical dendrites consistently suppressed Ca2+ spike electrogenesis even at the highest stimulus intensities (Fig. 9F; n = 4). Together, these results show that Kv7/M channels in the apical dendrites elevate the threshold for local Ca2+ spiking.
Finally, we examined the effects of the Kv7/M channel modulators on spike output in slices perfused with ACSF containing 3 mM 4AP. In this condition the spike output/stimulus intensity relationship depends on both somatic and dendritic electrical events (Magee and Carruth 1999). In control conditions (4AP-containing ACSF), stimulating the neurons with 180-ms-long positive current pulses usually evoked a single burst as the minimal neuronal responses to threshold-straddling stimuli (Fig. 10, Aa, Ba, and Ca); increasing stimulus intensity evoked additional solitary spikes (Fig. 10, Aa and Ca) or an extra burst (Fig. 10Ba). Linopirdine focally applied to the apical dendrites (one to three puffs) substantially augmented the number of spikes evoked at all stimulus intensities (n = 5; P < 0.05; Fig. 10Ab). A similar action was exerted by distally applied XE991 (Fig. 10, Ba and Bb; n = 3). Conversely, retigabine puffed onto the apical dendrites (one puff) unfailingly suppressed spike output at all stimulus intensities (Fig. 10C; n = 4). Figure 8, D–F depicts the averaged spike output/stimulus intensity plots obtained in these experiments. The marked effects of distally applied linopirdine, XE991, and retigabine on spike output in 4AP-containing ACSF contrast sharply with the lack of any effect when similarly applied in normal ACSF (Fig. 3, D–F).
In conclusion, our results indicate the presence of functional Kv7/M channels in the apical dendrites. These channels, together with their axo-somatic counterparts, regulate the spike output of CA1 pyramidal cells in conditions of enhanced dendritic excitability.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Kv7/M channels in axo-soma
Using local application of Kv7/M channel modulators, we show here that Kv7/M channels in the axo-soma of CA1 pyramidal cells tightly control the size of the somatic spike ADP. Blocking these channels unleashes a regenerative depolarization that causes growth of the spike ADP until it attains spike threshold and triggers an additional spike. This process repeats itself several times, thereby producing a spike burst. Blocking INaP in the axo-somatic region of the neuron reverses this process, suggesting that the latter current furnishes the main depolarizing drive for the regenerative growth of the ADP. In contrast, blocking Ca2+ currents does not affect this process, indicating that recruitment of Ca2+ channels is not obligatory for burst electrogenesis. On the contrary, blocking Ca2+ currents causes protraction of the burst into a plateau depolarization lasting hundreds of milliseconds, indicating that Ca2+-activated K+ currents are essential for burst termination in condition of blocked axo-somatic Kv7/M channels.
A large body of evidence now suggests that the spike ADP in adult CA1 pyramidal cells is generated by recruitment of axo-somatic persistent Na+ channels (Azouz et al. 1996; Su et al. 2001; Yue et al. 2005). These channels activate quickly at subthreshold potentials and do not inactivate substantially over a period of hundreds of milliseconds (French et al. 1990). Because fast repolarization of the spike is incomplete, ending at a potential nearly 10 mV more depolarized than resting potential (i.e., at about –55 mV; Su et al. 2001), these channels are likely to generate a substantial inward spike aftercurrent that would tend to redepolarize the neuron again to spike threshold (Yue et al. 2005). Here we show that this action of INaP is counteracted by activation of nearby Kv7/M channels that generate a repolarizing current, IM. Other slow K+ channels that may generate an outward spike aftercurrent, such as SK Ca2+-activated K+ channels and D-type K+ channels, do not exert control over the spike ADP, as blocking them selectively does not substantially modify the spike ADP, provided Kv7/M channels are active (Yue and Yaari 2004). In contrast, even a 10% reduction in IM causes a significant increase of the spike ADP. Thus the size of the somatic spike ADP and its propensity to trigger a burst are determined largely by a balance between INaP and IM in the proximal portions of the neuron. An imbalance between these two currents in favor of INaP facilitates the spike ADP and converts the output of the neuron from a solitary spike to a high-frequency spike burst.
Although our technique for local application of Kv7/M channel modulators could not discern between Kv7/M channels in axon initial segment, soma, or proximal dendrites, recent immunohistochemical evidence suggests that these channels may aggregate in the axon initial segment, as well as in the nodes of Ranvier (Devaux et al. 2004). Interestingly, the axon initial segment and nodal membranes in many neurons also express a high density of NaV1.6 channels (Boiko et al. 2003). Of all the 
-subunit isoforms of voltage-activated Na+ channels expressed in brain neurons (NaV1.1, NaV1.2, NaV1.3, and NaV1.6; Goldin 1999), NaV1.6 is the main subunit underlying INaP (Burbidge et al. 2002; Smith et al. 1998). Therefore it is conceivable that the two principal species of ion channels that shape the waveform of the somatic spike ADP; that is, NaV1.6 and Kv7/M channels are coexpressed at high densities in the axon initial segment. Thus the axon initial segment may be a strategic site not only for spike initiation (Spruston et al. 1995; Turner et al. 1991), but also for generating the active spike ADP and associated bursting.
Kv7/M channels in apical dendrites
Although it is well established that dendrites of CNS neurons express multiple types of K+ channels that contribute to the complexity and heterogeneity of neuronal discharge behavior (Chen and Johnston 2004; Golding et al. 1999; Johnston et al. 2000), a functional role for Kv7/M channels in these neuronal processes has not been demonstrated before. The evidence that these channels are expressed in apical dendrites of CA1 pyramidal cells is twofold. First immunohistochemical studies have demonstrated moderate staining for Kv7.2, Kv7.3, and Kv7.5 in apical dendrites of CA1 pyramidal cells (Roche et al. 2002; Shah et al. 2002). Second, recent single-channel recordings from these structures (170–250 µm away from the soma) encountered, albeit infrequently, putative M-type K+ channels (Chen and Johnston 2004). Our findings that in some conditions (e.g., 4AP-containing ACSF) apical dendrites are very responsive to directly applied modulators of Kv7/M channels are congruent with these findings and further show that, despite their scarcity, dendritic Kv7/M channels may be of functional significance.
Distal applications of linopirdine or XE991 to block Kv7/M channels in the apical dendrites did not affect "passive" membrane properties of CA1 pyramidal cells recorded in the soma, nor did they modify the somatic spike ADP or the spike output/stimulus intensity relationship in slices perfused with normal ACSF. These findings suggest that in normal conditions dendritic Kv7/M channels do not tightly regulate the overall intrinsic excitability of these neurons. This may be attributable to a low Kv7/M channel density in the apical dendrites (Chen and Johnston 2004); to the restriction of these channels to dendritic branches of higher order, where they would be electrotonically isolated from the soma; and/or to the overwhelming inhibitory effects of other currents, such as IA or Ih, whose densities in the apical dendrites are severalfold higher than in the soma (Hoffman et al. 1997; Magee 1998). Indeed, blocking IA in the apical dendrites with millimolar 4AP augments the somatic spike ADP and induces somatic bursting, whereas blocking IA in the axo-soma does not produce these effects (Magee and Carruth 1999). Interestingly, we found that once IA was blocked, dendritic hyperexcitability came to be tightly controlled by Kv7/M channels. Perhaps in this condition of enhanced spike invasion into the apical dendrites and Ca2+ spike electrogenesis, distal dendritic Kv7/M channels are more effectively recruited and therefore exercise enhanced inhibitory control over dendritic excitability.
Funct
