| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
NEUROSCIENCE |
10 Hz) of excitatory networks in motor cortex: effects of voltage-dependent ion channel blockers
1 Department of Neurobiology and Anatomy, Drexel University College of Medicine, Philadelphia, PA 19129, USA
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
10 Hz oscillations, are generated by a purely excitatory network of interconnected pyramidal cells because they are robust in the absence of GABAergic transmission. It is likely that specific voltage-dependent currents expressed in those cells contribute to the generation of 10 Hz oscillations. We tested the effects of different drugs known to suppress certain voltage-dependent currents. The results revealed that drugs that suppress the low-threshold calcium current and the hyperpolarization-activated cation current are not critically involved in the generation of 10 Hz oscillations. Interestingly, drugs known to suppress the persistent sodium current abolished 10 Hz oscillations. Furthermore, blockers of K+ channels had significant effects on the oscillations. In particular, blockers of the M-current abolished the oscillations. Also, blockers of both non-inactivating and slowly inactivating voltage-dependent K+ currents abolished 10 Hz oscillations. The results indicate that specific voltage-dependent non-inactivating K+ currents, such as the M-current, and persistent sodium currents are critically involved in generating 10 Hz oscillations of excitatory motor cortex networks.
(Received 10 August 2006;
accepted after revision 30 August 2006;
first published online 31 August 2006)
Corresponding author M. Castro-Alamancos: Department of Neurobiology and Anatomy, Drexel University College of Medicine, 2900 Queen Lane, Philadelphia, PA 19129, USA. Email: manuel.castro{at}drexel.edu
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
10 Hz oscillations in motor cortex in slices and in vivo (Castro-Alamancos, 2000; Castro-Alamancos & Rigas, 2002). In vivo, these oscillations are associated with overt tremors and myoclonus that resemble the motor manifestations of continuous partial epilepsy (Castro-Alamancos, 2006). Similar oscillations also occur in the presence of low [Mg2+]o buffers in neocortex (Silva et al. 1991; Flint & Connors, 1996) and in the hippocampus in slices and in vivo (Miles et al. 1984; Traub et al. 1993a; 1993b' 1996,,; Bragin et al. 1997a, 1997b). The obvious similarities between neocortical 10 Hz oscillations caused by disinhibition and those caused by low [Mg2+]o buffers led us to investigate their properties side by side.
An important question is how are 10 Hz oscillations generated? Because the oscillations are robust when GABA receptors are blocked, the 10 Hz oscillations of agranular neocortex necessarily involve glutamatergic neurons. In fact, 10 Hz oscillations caused by disinhibition are specifically abolished by blocking AMPA receptors, which leaves the initial interictal spike (i.e. paroxysmal depolarizing shift) intact (Castro-Alamancos & Rigas, 2002). Thus, a pure excitatory network of connected pyramidal cells generates 10 Hz oscillations in motor cortex. One possibility is that specific voltage-dependent ion channels present in these cells are critically involved in the generation of 10 Hz oscillations. As an initial approximation, in order to determine which currents may be critically involved in generating 10 Hz oscillations, we tested the effects of different drugs know to suppress specific voltage-dependent currents. Because fast excitatory transmission is critical for these oscillations, we also considered the effects of the drugs on glutamatergic synaptic responses.
Voltage-dependent ion channels are critically involved in generating synchronized network oscillations in many brain areas, such as, for example, the thalamus (e.g. McCormick & Pape, 1990; Huguenard & McCormick, 1992; Bal et al. 1995; Steriade et al. 1997) and inferior olive (e.g. Llinas & Yarom, 1981; Llinas & Yarom, 1986; Bal & McCormick, 1997; Manor et al. 1997). In particular, the low-threshold calcium current (IT) (Huguenard, 1996) and the hyperpolarization-activated cation current (IH) (Pape, 1996) play critical roles in producing rhythmic oscillations in these areas. Another subthreshold current that has been proposed to be involved in the generation of rhythmic oscillations in various brain regions is the persistent sodium current (INap) (e.g. Alonso & Llinas, 1989; Amitai, 1994; Crill, 1996; Takakusaki & Kitai, 1997; Bevan & Wilson, 1999; Butera et al. 1999; Bennett et al. 2000; Agrawal et al. 2001; D'Angelo et al. 2001; Hu et al. 2002; Rybak et al. 2003; Darbon et al. 2004; Paton et al. 2006; Wang et al. 2006).
In the absence of inhibition, intrinsic K+ currents are the counterbalance of excitation. Hence, it is possible that K+ currents either impede the generation of 10 Hz oscillations and/or serve to repolarize the membrane during each wave of the oscillation. CA1 and layer V pyramidal cells express three major types of voltage-dependent K+ currents in the soma and dendrites (Storm, 1988, 1990; Hoffman et al. 1997; Korngreen & Sakmann, 2000; Bekkers, 2000a, 2000b; Bekkers & Delaney, 2001); a transient current that rapidly activates and inactivates (IA), a more slowly inactivating current (ID), and a sustained delayed rectifier (IK). Importantly, these three voltage-dependent K+ current components have well-known sensitivities to K+ channel blockers; low doses of 4-aminopyridine (4-AP; micromolar range) block the slowly inactivating K+ current ID, with little effect on IK and IA. High doses of 4-AP (millimolar range) block IA, while tetraethylammonium (TEA) at high doses (10–30 mM) blocks the sustained delayed rectifier IK (Storm, 1988, 1990; Hoffman et al. 1997; Bekkers & Delaney, 2001). In addition, pyramidal cells express a voltage-dependent K+ current, called the M-current (IM), which activates positive to –60 mV and does not inactivate (Halliwell & Adams, 1982; Storm, 1990). Moreover, this current is blocked by XE991 and linopirdine in pyramidal cells (Aiken et al. 1995; Hu et al. 2002; Yue & Yaari, 2004). The potential involvement of these currents in the generation of 10 Hz oscillations of neocortex is unknown. Interestingly, network modelling has shown that an interplay between INap and IM could account for the generation of synchronized oscillations of neocortex (Golomb et al. 2006).
In the present study, we used a pharmacological approach to begin addressing the mechanisms involved in the generation of 10 Hz oscillations of motor cortex. We tested the effects of drugs that suppress different voltage-dependent currents, namely, INap, IT, IH, and voltage-dependent K+ channels on 10 Hz network oscillations generated by disinhibition or low [Mg2+]o.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
7 weeks) eYFP-H mice taken from our breeding colony, as previously described (Castro-Alamancos & Rigas, 2002). Mice were deeply anaesthetized with sodium pentobarbital (60 mg kg–1) and upon losing all responsiveness to a strong tail pinch the brain was rapidly extracted and placed in ice-cold buffer. Slices (400 µm thick) were cut in the thalamocortical plane (Agmon & Connors, 1991) in ice-cold buffer, using a vibratome, as previously described (Castro-Alamancos & Rigas, 2002; Rigas & Castro-Alamancos, 2004). Experiments were performed in an interface chamber at 32°C. The slices were bathed constantly (1–1.5 ml min–1) with artificial cerebrospinal fluid (ACSF) containing (mM): NaCl (126), KCl (3.5), NaH2Po4 (1.25), NaHCO3 (26), MgSO4 7H2O (1), dextrose (10), CaCl2 2H2O (1). The ACSF, which we will refer to as normal ACSF, was bubbled with 95% O2 and 5% CO2. Field recordings were made using low-impedance pipettes (0.5 M
) filled with ACSF. Blind whole-cell recordings were obtained from layer III using patch electrodes of 4–9 M impedance. The electrodes were filled with internal solution containing (mM): 135 K-gluconate, 4 KCl, 2 NaCl, 0.2 EGTA, 5 Tris-phosphocreatine, 0.3 trisGTP, 10 Hepes, 4 MgATP (290 mosmol l–1). All procedures were reviewed and approved by the Animal Care Committee of Drexel University.
We used two different methods to induce 10 Hz oscillations in neocortex. In both models, GABAA receptors were blocked with bicuculline (BMI; 10 µM). In the disinhibition model, blockade of GABAA receptors with BMI was followed by blockade of GABAB receptors with CGP35348 (500 µM). In the low [Mg2+]o model, blockade of GABAA receptors was followed by lowering [Mg2+]o to 0.1 mM. The drugs used in the present study were all purchased from Sigma-Aldrich or Tocris, except for CGP35348, which was a gift from Novartis, and U-92032, which was a gift from Pfizer. All drugs were dissolved in the ACSF except phenytoin, riluzole and U-92032 which were initially dissolved at 1–10 mM in DMSO. Linopirdine was dissolved in ethanol at 10 mM. The final concentration of DMSO or ethanol was applied first to the control buffer in the bath to test for any potential effects. This was later followed by the drug dissolved in the same concentration of DMSO or ethanol. These experiments revealed that the highest dose of DMSO or ethanol applied to the bath in this study had no significant effect on 10 Hz oscillations. In fact, the highest concentration of DMSO used was with U-92032, which did not abolish 10 Hz oscillations.
Data analyses were performed using Origin software. Power spectrum analyses were derived by calculating fast Fourier transforms (FFT) from the spontaneous discharges. Ten randomly selected discharges were measured per experimental condition in each slice. Measurements were made immediately before and 20–30 min after first application of the drugs. For each event, the large-amplitude interictal spike was excluded from the FFT analysis, and thus only the oscillatory phase of the discharge was considered between 115 and 2700 ms after the peak of the first large-amplitude discharge. Population statistical comparisons were performed by summing the FFT power for the specified frequency range (usually, 5–20 Hz) for each slice. These values were then compared between before and during the presence of the drug, using paired t tests. Results are expressed as mean ± S.D. The percentage change caused by each drug for the specified frequency range was the average ratio calculated by dividing the power in that frequency range during and before the drug for each slice.
Evoked field and, in some cases, intracellular potentials were used to monitor the effects of drugs on synaptic responses in control buffer. This was accomplished by placing a recording electrode in layer III and a stimulating electrode in the adjacent layer III to stimulate horizontal fibres, and/or in the underlying layer (V) to stimulate vertical fibres, as previously shown (Castro-Alamancos et al. 1995). Stimulating and recording electrodes were separated 0.5 mm. The peak amplitude of the evoked response between 3 and 6 ms was measured and used as an index of synaptic efficacy for these pathways.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
0.1 Hz in frontal slices (Rigas & Castro-Alamancos, 2004). Subsequent application of either a GABAB receptor antagonist (CGP35348) or lowering [Mg2+]o to 0.1 mM results in the development of 10 Hz oscillations that follow each interictal spike in the agranular (motor) neocortex both in slices (Castro-Alamancos & Rigas, 2002) and in vivo (Castro-Alamancos, 2000). In a previous study, we found that the 10 Hz oscillations caused by disinhibition were completely abolished by an AMPA receptor antagonist, and we also found that an NMDA receptor antagonist did not abolish the oscillations but transformed them by shifting their frequency to a higher range (Castro-Alamancos & Rigas, 2002). Here, we tested the effect of AMPA (GYKI53655) and NMDA (D-AP5) receptor antagonists on 10 Hz oscillations caused by low [Mg2+]o. Figure 1 shows examples of the effects of GYKI53655 (10 µM) and D-AP5 (50 µM) on 10 Hz oscillations generated by low [Mg2+]o. The AMPA receptor antagonist completely abolished the 10 Hz oscillations caused by low [Mg2+]o, without suppressing the first large-amplitude interictal spike. The NMDA receptor antagonist transformed the oscillations by shifting the frequency towards a higher range but did not abolish them. In every case, D-AP5 reduced the duration of the oscillation and shifted it closer toward to the first large-amplitude interictal spike, so that the oscillation rides on top of the positive wave component of the interictal discharge. These results demonstrate that both the 10 Hz oscillations and those during low [Mg2+]o are critically dependent on AMPA receptors, while NMDA receptors play an important role but are not required for the oscillations to be present.
|
10 Hz oscillations generated by disinhibition or by low [Mg2+]o in slices of the agranular neocortex.
Drugs that suppress INap abolish 10 Hz oscillations
The current flowing through Na+ channels consists of transient and persistent components (Crill, 1996). These can be differentiated by their different time courses of inactivation, one very fast that contributes to the transient component, and one an order of magnitude slower that contributes to the persistent component (Crill, 1996). Thus, it has been known for some time that in addition to the transient Na+ current, cortical pyramidal neurons express a persistent Na+ current, INap, reflected as a subthreshold inward rectification, that is blocked by TTX (Hotson et al. 1979; Connors et al. 1982; Stafstrom et al. 1982, 1984, 1985). In some cells, such as CA1 pyramidal neurons, there seems to be a significant difference in the sensitivity of both Na+ current components to TTX, so that small doses of TTX that block the persistent component do not block the transient component (Hammarstrom & Gage, 1998). However, this difference does not exist in other cells, such as tuberomammillary neurons, in which TTX blocks equally both components (Taddese & Bean, 2002). Thus, at least in some cases, small doses of TTX may be used to test the involvement of INap in the generation of oscillations. In addition, several drugs have been shown to selectively suppress the persistent component of the Na+ current with little effect on the transient component, such as phenytoin (Quandt, 1988; Segal & Douglas, 1997) and riluzole (Urbani & Belluzzi, 2000; Niespodziany et al. 2004; Kononenko et al. 2004). In order to test the potential involvement of INap in the generation of 10 Hz oscillations, we applied low doses of TTX (50–100 nM), phenytoin (70–130 µM) and riluzole (5–20 µM) during 10 Hz oscillations generated by either disinhibition or low [Mg2+]o. Because of the reliance of 10 Hz oscillations on glutamatergic synaptic transmission, we also tested the effect of the same drugs on synaptic field potential responses evoked in layer III by stimulating horizontal and/or vertical pathways during normal ACSF.
Figures 2 and 3 show examples of the effects of a low dose of TTX (50–100 nM), phenytoin (130 µM) and riluzole (10–20 µM) on 10 Hz oscillations generated by disinhibition and low [Mg2+]o, respectively. The low dose of TTX, phenytoin and riluzole had a very similar effect on 10 Hz oscillations generated by either method. These drugs completely abolished the 10 Hz oscillations without suppressing the first large-amplitude interictal spike, which usually increased in amplitude. The FFT power spectrum analyses in Figs 2 and 3, and in all subsequent figures, show population data corresponding to spontaneous events before and during application of each drug. For these analyses, we calculated power spectra from 10 randomly selected spontaneous events per slice (n being the number of slice tested) during the different conditions. For the three drugs there was a significant suppression of the activity in the 5–20 Hz range during low [Mg2+]o (74 ± 8, 94 ± 8 and 88 ± 7% suppression for TTX, phenytoin and riluzole, respectively; t test P < 0.01) and during disinhibition (79 ± 7, 88 ± 9 and 90 ± 9% suppression for TTX, phenytoin and riluzole, respectively; t test P < 0.01).
|
|
10 Hz oscillations, strongly supporting the contention that INap is critically involved in the generation of 10 Hz oscillations in the agranular neocortex. However, one possibility is that these drugs abolished 10 Hz oscillations because they suppressed synaptic neurotransmission. In light of the fact that blocking glutamatergic synaptic transmission blocks 10 Hz oscillations (Castro-Alamancos & Rigas, 2002), it is important to test the effects of these drugs on glutamatergic synaptic responses. Thus, we tested the effects of these drugs on synaptic field potential responses evoked in layer III by stimulating horizontal and vertical pathways during bathing in normal ACSF. Electrical stimulation produces a short-latency negative field potential response in layer III that corresponds to an excitatory postsynaptic potential recorded intracellularly (Castro-Alamancos et al. 1995). Figure 4A and C shows examples and population data of the effects of TTX (100 nM; n
= 5), phenytoin (130 µM; n
= 4) and riluzole (20 µM; n
= 4) on evoked synaptic responses in layer III. These drugs produced a similar effect on the field potential response; usually slightly enhancing the peak amplitude of the field potential response for both horizontal and vertical pathways. As expected, application of a larger dose of TTX (e.g. 1 µM) completely abolished the evoked field potential responses (not shown). The effect of these drugs on the amplitude of the evoked field potential response was somewhat surprising. Thus, we performed whole-cell recordings from layer III cells in order to determine their impact on intracellularly recorded EPSPs. We found that TTX (100 nM; n
= 3 cells) and riluzole (10 µM; n
= 3 cells) suppressed evoked EPSPs, while at the same time enhancing the evoked field potential responses (Fig. 4D and E). The EPSP suppression was small but significant (18.7% and 11.6% reduction in peak amplitude for TTX and riluzole, respectively; P < 0.01). The effects of both drugs on field potential and EPSP responses were accompanied by a small depolarization of the recorded cells (<3 mV) and small increases in input resistance (<5% per cell; input resistance before the drugs was 169 ± 40 M, n
= 6 cells). It is likely that the effects of the drugs on the field potential responses are caused by changes in membrane excitability. Therefore, these results indicate that low TTX, phenytoin and riluzole did not abolish 10 Hz oscillations by abolishing glutamatergic synaptic transmission in the neocortex. Instead, these results support the idea that INap is critically involved in the generation of 10 Hz oscillations in the agranular neocortex. However, small changes in synaptic transmission caused by these drugs may also contribute to the abolition of the 10 Hz oscillations.
|
10 Hz oscillations
Low-threshold T-type channels have been characterized in several cortical pyramidal cells (Huguenard, 1996). These channels have been shown to have a critical role in the generation of oscillations in the 7–14 Hz range in the thalamus (Huguenard & McCormick, 1992; Bal et al. 1995; Steriade et al. 1997) and in the generation of slow intrathalamic rhythms characteristic of absence seizures (Kim et al. 2001; Porcello et al. 2003). Thus, they may well be involved in the generation of 10 Hz oscillations in the neocortex. Although there is no selective drug that blocks IT, there are several drugs that have been shown to suppress this current, such as ethosuximide (Coulter et al. 1989), divalent cations, such as Ni2+ (Ye and Akaike, 1993;Huguenard, 1996), mibefradil (McDonough & Bean, 1998; Lacinova, 2004) and U-92032 (Avery & Johnston, 1997; Porcello et al. 2003). However, most of these drugs are not selective blockers of IT. For example, ethosuximide has been reported to affect other currents, such as INap (Leresche et al. 1998). U-92032 has been reported to affect Na+ currents ((Avery & Johnston, 1997) but see (Porcello et al. 2003)), while divalent cations have well-known effects on cellular excitability and synaptic transmission. Thus, the approach in the present study is to employ all these different drugs to test if the results are consistent with a critical involvement of IT in the generation of 10 Hz oscillations. In order to test the potential involvement of IT in generating 10 Hz oscillations, we applied Ni2+ (100–200 µM), ethosuximide (1–10 mM), mibefradil (5–20 µM) and U-92032 (10 µM) during 10 Hz oscillations generated by either disinhibition (Fig. 5) or low [Mg2+]o (Figs 6 and 7). We also tested the effects of these drugs on evoked synaptic responses in the neocortex during normal ACSF.
|
|
|
10 Hz oscillations generated by disinhibition and low [Mg2+]o, respectively. Interestingly, application of Ni2+ produced different effects in the two models used for generating 10 Hz oscillations. The oscillations generated by disinhibition were not suppressed by Ni2+ (19 ± 12% enhancement of 5–20 Hz FFT power; 100–200 µM; n
= 5; t test, not significant; Fig. 5A), while the same doses suppressed (100 µM; n
= 6) or abolished (80 ± 8% suppression of 5–20 Hz FFT power; 200 µM; n
= 4; t test; P < 0.01) the oscillations generated by low [Mg2+]o (Fig. 6A). As shown in Fig. 4B, the dose of Ni2+(200 µM) that abolished the oscillations generated by low [Mg2+]o had little effect on evoked synaptic responses (6 ± 5% suppression of field potential amplitude; n
= 6), indicating that the effect of Ni2+ was not caused by suppressing synaptic transmission. In a few experiments (n
= 4), we found that large doses of Ni2+ (1 mM) also abolished the oscillations generated by disinhibition (not shown), but this dose produced a strong suppression (56 ± 4% suppression; n
= 4; t test P < 0.01) of evoked synaptic responses (Fig. 4B and C) suggesting that for the large dose of Ni2+ an effect on synaptic transmission was the likely cause of the abolition.
The differential effect of the lower dose of Ni2+ in the two models raises the possibility that 10 Hz oscillations generated by low [Mg2+]o involve IT in their generation, while the oscillations generated by disinhibition do not. Alternatively, the reason why the oscillations are abolished by Ni2+ during low [Mg2+]o is because of the replacement of the divalent cation concentration. That is, Ni+2 is simply re-establishing the concentration of divalent cations in the ACSF. Alternatively, if IT was selectively involved in the generation of 10 Hz oscillations caused by low [Mg2+]o, then other drugs that suppress IT should abolish 10 Hz oscillations caused by low [Mg2+]o.
Figures 5B and 6B show examples of the effect of ethosuximide (1–10 mM) on 10 Hz oscillations generated by disinhibition and low [Mg2+]o, respectively. At doses that are known to suppress IT (1–2 mM) (Coulter et al. 1989; Huguenard, 1996), ethosuximide did not abolish the 10 Hz oscillations generated by either disinhibition or low [Mg2+]o. However, this dose of ethosuximide produced a consistent suppression of the activity in the 5–20 Hz range during disinhibition (18 ± 10% suppression of 5–20 Hz FFT power; 1–2 mM; n
= 6; t test, P < 0.01) or low [Mg2+]o (25 ± 15% suppression of 5–20 Hz FFT power; 1–2 mM; n
= 6; t test P < 0.01). But this suppression did not entail an abolition of the 10 Hz oscillations, which were clearly present. Moreover, at these doses synaptic responses were also not significantly affected by ethosuximide (5 ± 5% suppression; n
= 3; t test, not significant; Fig. 4B and C). At a larger dose, ethosuximide (10 mM) abolished the oscillations generated by either disinhibition or low [Mg2+]o, without abolishing the interictal spikes (Fig. 5B). This effect was reflected in a strong suppression of the activity in the 5–20 Hz frequency range during disinhibition (84 ± 11% suppression of 5–20 Hz FFT power; 5–10 mM; n
= 6; t test P < 0.01) and low [Mg2+]o (80 ± 12% suppression of 5–20 Hz FFT power; 5–10 mM; n
= 6; t test P < 0.01). However, the larger dose of ethosuximide (10 mM) had a strong suppressive effect on evoked synaptic responses (54 ± 6% suppression of the field potential amplitude; n
= 5; t test, P < 0.01; Fig. 4B and C) indicating that this was the likely cause of the abolition.
Figures 5C and 7A show examples of the effect of mibefradil (5–20 µM) on 10 Hz oscillations generated by disinhibition and low [Mg2+]o, respectively. Mibefradil did not suppress 10 Hz oscillations at any of the doses tested in either model. In fact, the larger dose of mibefradil (20 µM), which was tested on oscillations caused by low [Mg2+]o, produced a strong enhancement of the activity in the 5–20 Hz frequency range (255 ± 8% enhancement of 5–20 Hz FFT power; n
= 4). Moreover, mibefradil (20 µM) did not significantly affect synaptic evoked responses (3 ± 5% suppression; n
= 4; t test, not significant; Fig. 4B and C). The effects of mibefradil are inconsistent with a critical role of IT in the generation of 10 Hz oscillations in neocortex.
In an effort to further investigate the potential involvement of IT in the generation of 10 Hz oscillations caused by low [Mg2+]o, we tested the effect of an additional drug that is known to suppress IT, U-92032. Figure 7B shows the effect of application of U-92032 (10 µM) on 10 Hz oscillations generated by low [Mg2+]o. In every experiment, U-92032 did not suppress 10 Hz oscillations generated by low [Mg2+]o (n
= 6). Application of the drug produced a slight enhancement in activity in the 5–20 Hz frequency range (20 ± 15% enhancement of 5–20 Hz FFT power; n
= 6). These results, taken together, do not support a critical role for IT in the generation of 10 Hz oscillations in neocortex.
Drugs that suppress IH do not abolish 10 Hz oscillations
The IH current consists of a slowly developing inward activation upon hyperpolarization of the membrane beyond the resting potential (Pape, 1996; Robinson & Siegelbaum, 2003), which is found in pyramidal neurons of the neocortex (Spain et al. 1987; Solomon et al. 1993; Hutcheon et al. 1996). These channels have been shown to play important roles in rhythmogenesis in the thalamus (McCormick & Pape, 1990; Luthi & McCormick, 1998) and hippocampus (Agmon & Wells, 2003). Thus, IH may well be involved in the generation of similar oscillations in the neocortex. To assess the potential involvement of IH in the generation of 10 Hz oscillations, we tested the effect of two known blockers of IH, Cs+ and ZD7288, on 10 Hz oscillations generated by disinhibition or low [Mg2+]o.
Figure 8A and 9A show examples of the effect of Cs+ (2 mM) on 10 Hz oscillations generated by disinhibition and low [Mg2+]o, respectively. Cs+ had little effect on the 10 Hz oscillations generated by either disinhibition (3 ± 6% enhancement of 5–20 Hz FFT power; 2 mM; n
= 4) or low [Mg2+]o (4 ± 5% suppression of 5–20 Hz FFT power; 2 mM; n
= 6). In general, Cs+ tended to slow the oscillation frequency and to enhance its duration. Figures 8B and 9B and C show examples of the effect of another IH blocker, ZD7288 (50 µM) on 10 Hz oscillations generated by disinhibition and low [Mg2+]o, respectively. This drug increased the oscillations generated by disinhibition (13 ± 8% enhancement of 5–20 Hz FFT power; 50 µM; n
= 6) and low [Mg2+]o (23 ± 14% enhancement of 5–20 Hz FFT power; 50 µM; n
= 6). Like Cs+, ZD8827 usually slowed the frequency of the oscillations and increased their duration, which may be attributed, for example, to a hyperpolarization or other effects caused by the block of IH. These effects were more apparent for longer-lasting oscillations, as shown in Fig. 9C. These results are inconsistent with a critical role for IH in the generation of 10 Hz oscillations. Instead, IH may play some role in stopping stronger long-lasting oscillations.
|
|
Low doses of 4-AP (in the micromolar range) are well known to block ID, the slowly inactivating K+ current, while higher doses (in the millimolar range) block IA, the rapidly inactivating K+ current (Storm, 1988, 1990; Hoffman et al. 1997; Coetzee et al. 1999). High doses of TEA (10 mM) significantly block a good portion of the sustained potassium current IK, with little effects on ID and IA (Storm, 1988, 1990; Hoffman et al. 1997; Bekkers & Delaney, 2001). To test the effect of K+-channel blockers on 10 Hz oscillations, we first placed the slices in a low [Mg2+]o buffer to induce 10 Hz oscillations, and then applied either a low dose of 4-AP (25 µM; n
= 8; Fig. 10A) or TEA (10 mM; n
= 6; Fig. 10B). Each of these drugs alone significantly enhanced 10 Hz oscillations (148 ± 9% and 78 ± 7% enhancement of 5–20 Hz FFT power for 10 mM TEA and 25 µM 4-AP, respectively; t test, P < 0.01), indicating that non-inactivating and slowly inactivating K+ current components generally suppress 10 Hz oscillations. However, addition of a low dose of 4-AP (25 µM) to TEA (10 mM), which would suppress simultaneously both slowly and non-inactivating K+ current components, completely abolished 10 Hz oscillations in every experiment (n
= 8; Fig. 10B; P < 0.01). During these conditions, activity consisted of highly rhythmic and rapidly recurring (1 Hz) interictal spikes that contained no hint of the 10 Hz oscillations (see Fig. 10B). In contrast, a high dose of 4-AP (10 mM), which would suppress mostly rapidly and slowly inactivating K+ current components, also resulted in the generation of highly rhythmic and rapidly recurring (1 Hz) interictal spikes, but these contained a robust short-lasting 10 Hz oscillation on top of each interictal spike (see Fig. 10A). These short-lasting oscillations were significantly higher in frequency (16.7 ± 3 Hz) than in the presence of low [Mg2+]o buffer (10.2 ± 1; P < 0.01), as the peak of the power spectrum shifted to the right in every experiment. These oscillations were abolished by TEA. Thus, addition of TEA (10 mM) to a high dose of 4-AP (10 mM), which would further suppress the non-inactivating K+ current component, left the rapidly recurring interictal spikes intact but abolished the short-lasting 10 Hz oscillations riding on top of them (Fig. 10A).
|
50 ms, and does not inactivate (Storm, 1990). Therefore, this current is well suited to provide the hyperpolarizing drive during each cycle of 10 Hz oscillations. To address this possibility, we tested the effects of two IM blockers on 10 Hz oscillations. Interestingly, as shown in Fig. 11, both of these drugs completely abolished 10 Hz oscillations, while leaving the initial interictal spike discharge intact. To test the effect of IM channel blockers on 10 Hz oscillations, we first placed the slices in a low [Mg2+]o buffer to induce 10 Hz oscillations, and then applied either linopirdine (10–20 µM; n
= 6; Fig. 11) or XE991 (10–20 µM; n
= 6; Fig. 11). Each of these drugs alone abolished 10 Hz oscillations (74 ± 9% and 87 ± 7% suppression of 5–20 Hz FFT power for 10 µM linopirdine and 10 µM XE991, respectively; t test, P < 0.01), indicating that IM is critically involved in the generation of 10 Hz oscillations. In addition, we tested the effect of apamin on 10 Hz oscillations caused by low [Mg2+]o buffer, which is known to suppress one type of after-hyperpolarizing potential (IAHP). However, apamin (0.1–1 µM; n
= 4) had little effects on the 10 Hz oscillations and did not suppress them (not shown).
|
10 Hz oscillations (albeit of short duration and higher frequency) in the absence of the rapidly inactivating K+ current, but not vice versa. That is, the rapidly inactivating K+ current component alone is incapable of sustaining 10 Hz oscillations in the absence of the non-inactivating components. Furthermore, the period of each oscillation bout or interictal spike appears to be critically determined by slowly inactivating K+ currents, because block of this component plus any one of the other two components (non-inactivating or rapidly inactivating) resulted in highly rhythmic, rapidly and continuously recurring interictal spikes. Finally, we found that the subthreshold non-inactivating K+ current, IM, appears to play a critical role in the generation of 10 Hz oscillations of agranular cortex because IM blockers readily abolish these oscillations. |
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
10 Hz oscillations generated in motor cortex by disinhibition or low [Mg2+]o are mechanistically related and are abolished by drugs known to suppress INap (Fig. 12 summarizes the effects of the drugs used in this study on the 10 Hz oscillations). In particular, low TTX, phenytoin and riluzole abolished the 10 Hz oscillations without suppressing the interictal spike that triggers them. Although selective blockers of INap do not exist, the abolition caused by these drugs is unlikely to be due to an effect on glutamatergic synaptic transmission because at the doses used they only slightly suppressed synaptic responses in our conditions. However, it is important to be aware that this and other non-specific unknown effects of these drugs (unrelated to INap) may have contributed to the selective abolition of the 10 Hz oscillations.
|
10 Hz oscillations in the motor cortex, which distinguishes these oscillations from oscillations in the same frequency range that occur in other structures (see Introduction). Briefly, mibefradil and U-92032, two drugs that are known to suppress IT, did not suppress 10 Hz oscillations. In fact, both of these drugs tended to enhance the oscillations. In addition, Ni2+, which also suppresses IT, did not suppress oscillations caused by disinhibition, but abolished oscillations caused by low [Mg2+]o. We interpret the observation that Ni2+ abolished 10 Hz oscillations caused by low [Mg2+]o as due to Ni2+ increasing the divalent cation concentration in the low [Mg2+]o buffer. This is supported by the findings that mibefradil and U-92032 did not suppress 10 Hz oscillations caused by low [Mg2+]o, and Ni2+ did not suppress 10 Hz oscillations caused by disinhibition. Moreover, ethosuximide did not abolish 10 Hz oscillations at low doses (1–2 mM). At higher doses, ethosuximide (5–10 mM) abolished 10 Hz oscillations, but also strongly suppressed glutamatergic synaptic responses. Finally, Cs+ and ZD8827, which suppress IH, did not suppress 10 Hz oscillations caused by disinhibition or low [Mg2+]o.
Why may INap be important in generating 10 Hz oscillations?
INap is a non-inactivating component of the TTX-sensitive sodium current. Although its magnitude is small compared to the transient sodium current, it has functional significance because it is activated about 10 mV negative to the transient sodium current, where few voltage-gated channels are activated and neuron input resistance is high. INap adds to synaptic current, and evidence indicates that it is present in dendrites where relatively small depolarizations will activate it, thereby increasing the effectiveness of distal depolarizing synaptic activity (Crill, 1996). In essence, INap can boost the communication line between the apical dendrite (input) and the axon (output) of layer V pyramidal cells, thus, favouring a positive feedforward loop. This boost may be critical for the generation of 10 Hz oscillations because current source density analysis has shown that 10 Hz oscillations consist of large current sinks in the upper layers (II–III) with corresponding sources in layer V (Castro-Alamancos & Rigas, 2002), which indicates the occurrence of strong localized inward currents in the apical dendrites of layer V pyramidal cells. These active spots may need to reach the output of the cell in order to close the recurrent excitatory circuit and produce the network oscillation, and this may happen with the aid of INap and other inward currents expressed on the dendrites of these cells. INap appears to be important for bursting in pyramidal cells, including those in layer V cells (Franceschetti et al. 1995; Azouz et al. 1996; Brumberg et al. 2000). Interestingly, application of glutamate to the apical dendrites of some layer V pyramidal cells produces repetitive bursting at around 10 Hz at the soma, and this appears to depend on INap (Schwindt & Crill, 1995, 1999; Mittmann et al. 1997). Similarly, direct depolarization of the apical dendrite triggers bursting at the soma (Williams & Stuart, 1999). Moreover, it is clear that in some layer V pyramidal cells, Na+-dependent conductances located on the apical dendrites allow the flow of current between the dendrites and the axon (Stuart et al. 1997; Williams & Stuart, 1999; Hausser et al. 2000). One possibility is that 10 Hz oscillations are generated by this same mechanism, so that the synchronous release of glutamate produced by the initial interictal spike depolarizes the apical dendrites of layer V cells that express INap, which triggers repetitive bursting at the soma at around 10 Hz. Synchronization of this activity via synaptic connections would result in the expression and reinforcement of 10 Hz oscillations in the network. Interestingly, a recent modelling study found that INap is critically important for producing synchronized bursting (similar to 10 Hz oscillations studied here) in networks of the neocortex (Golomb et al. 2006), supporting the pharmacological data of the present study. In this scenario, INap supports bistability in the bifurcation diagram of the fast subsystem of variables, allowing bursting based on the ‘square wave mechanism’ (Rinzel et al. 1998). Thus, in this model, the spatial structure of the cells is not needed to account for the generation of 10 Hz oscillations (Golomb et al. 2006).
What is the role of voltage-dependent K+ currents?
Outward currents such as voltage-dependent K+ currents must be important in generating 10 Hz oscillations of motor cortex because during these oscillations GABAergic inhibition is blocked. Thus, K+ currents must provide the repolarizing drive for each cycle of the oscillations. We found that drugs that suppress either the slowly inactivating (low 4-AP) or the non-inactivating (TEA) components of voltage-dependent K+ currents enhanced 10 Hz oscillations when applied separately. However, suppressing these components together abolished the oscillations, indicating that at least one of these components must be present to sustain the oscillations. In contrast, blocking rapidly inactivating K+ currents (high 4-AP) did not abolish the oscillations, suggesting that this component is not critical for sustaining 10 Hz oscillations. Intriguingly, suppressing the subthreshold non-inactivating current, IM, readily abolished 10 Hz oscillations. This result fits well with the idea that IM serves as the hyperpolarizing current during each cycle of the oscillations. IM is well suited for this role because it activates slowly positive to –60 mV, with a time constant around of 50 ms. As suggested by network modelling, IM may be important for the oscillations because it has the appropriate time scale for terminating the active phase of the network bursting in each cycle (Golomb et al. 2006). In essence, the results presented here indicate that IM and the non-inactivating or slowly inactivating components of voltage-dependent K+ currents may provide the repolarizing drive during each wave of the oscillation.
Are 10 Hz oscillations functionally relevant in the motor cortex in vivo?
Behaving rats produce 10 Hz oscillatory activity in frontal neocortex, including motor cortex, in preparation for certain types of whisker movements, such as twitching (Semba & Komisaruk, 1984), which may actually be an abnormal behavioural state that occurs in certain rodent strains (Shaw, 2004). In humans, discharges at 10 Hz are associated with frontal lobe epilepsies such as epilepsia partialis continua (Chauvel et al. 1992). Activity resembling the 10 Hz oscillations studied here are observed in the motor cortex of humans during cortical myoclonus, which is associated with rhythmical jerks of the affected body parts (Chauvel et al. 1992). Interestingly, 10 Hz oscillations of motor cortex induced in behaving animals by disinhibition produce rhythmical jerks of the contralateral body that resemble a cortical myoclonus (Castro-Alamancos, 2006).
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Agmon A & Wells JE (2003). The role of the hyperpolarizationactivated cationic current I (h) in the timing of interictal bursts in the neonatal hippocampus. J Neurosci 23, 3658–3668.
Agrawal N, Hamam BN, Magistretti J, Alonso A & Ragsdale DS (2001). Persistent sodium channel activity mediates subthreshold membrane potential oscillations and low-threshold spikes in rat entorhinal cortex layer V neurons. Neuroscience 102, 53–64.[CrossRef][Medline]
Aiken SP, Lampe BJ, Murphy PA & Brown BS (1995). Reduction of spike frequency adaptation and blockade of M-current in rat CA1 pyramidal neurones by linopirdine (DuP 996), a neurotransmitter release enhancer. Br J Pharmacol 115, 1163–1168.[Medline]
Alonso A & Llinas RR (1989). Subthreshold Na+-dependent theta-like rhythmicity in stellate cells of entorhinal cortex layer II. Nature 342, 175–177.[CrossRef][Medline]
Amitai Y (1994). Membrane potential oscillations underlying firing patterns in neocortical neurons. Neuroscience 63, 151–161.[CrossRef][Medline]
Avery RB & Johnston D (1997). Ca2+ channel antagonist U-92032 inhibits both T-type Ca2+ channels and Na+ channels in hippocampal CA1 pyramidal neurons. J Neurophysiol 77, 1023–1028.
Azouz R, Jensen MS & Yaari Y (1996). Ionic basis of spike after-depolarization and burst generation in adult rat hippocampal CA1 pyramidal cells. J Physiol 492, 211–223.[Medline]
Bal T & McCormick DA (1997). Synchronized oscillations in the inferior olive are controlled by the hyperpolarization-activated cation current I(h). J Neurophysiol 77, 3145–3156.
Bal T, von Krosigk M & McCormick DA (1995). Synaptic and membrane mechanisms underlying synchronized oscillations in the ferret lateral geniculate nucleus in vitro. J Physiol 483, 641–663.[Medline]
Bekkers JM (2000a). Distribution and activation of voltage-gated potassium channels in cell-attached and outside-out patches from large layer 5 cortical pyramidal neurons of the rat. J Physiol 525 Part 3, 611–620.
Bekkers JM (2000b). Properties of voltage-gated potassium currents in nucleated patches from large layer 5 cortical pyramidal neurons of the rat. J Physiol 525, 593–609.
Bekkers JM & Delaney AJ (2001). Modulation of excitability by alpha-dendrotoxin-sensitive potassium channels in neocortical pyramidal neurons. J Neurosci 21, 6553–6560.
Bennett BD, Callaway JC & Wilson CJ (2000). Intrinsic membrane properties underlying spontaneous tonic firing in neostriatal cholinergic interneurons. J Neurosci 20, 8493–8503.
Bevan MD & Wilson CJ (1999). Mechanisms underlying spontaneous oscillation and rhythmic firing in rat subthalamic neurons. J Neuroscience 19, 7617–7628.
Bragin A, Csicsvari J, Penttonen M & Buzsaki G (1997a). Epileptic afterdischarge in the hippocampal-entorhinal system: current source density and unit studies. Neuroscience 76, 1187–1203.[CrossRef][Medline]
Bragin A, Penttonen M & Buzsaki G (1997b). Termination of epileptic afterdischarge in the hippocampus. J Neurosci 17, 2567–2579.
Brumberg JC, Nowak LG & McCormick DA (2000). Ionic mechanisms underlying repetitive high-frequency burst firing in supragranular cortical neurons. J Neurosci 20, 4829–4843.
Butera RJ Jr, Rinzel J & Smith JC (1999). Models of respiratory rhythm generation in the pre-Botzinger complex. I. Bursting pacemaker neurons. J Neurophysiol 82, 382–397.
Castro-Alamancos MA (2000). Origin of synchronized oscillations induced by neocortical disinhibition in vivo. J Neurosci 20, 9195–9206.
Castro-Alamancos MA (2006). Vibrissa myoclonus (rhythmic retractions) driven by resonance of excitatory networks in motor cortex. J Neurophysiol 96, 1691–1698.
Castro-Alamancos MA, Donoghue JP & Connors BW (1995). Different forms of synaptic plasticity in somatosensory and motor areas of the neocortex. J Neurosci 15, 5324–5333.[Abstract]
Castro-Alamancos MA & Rigas P (2002). Synchronized oscillations caused by disinhibition in rodent neocortex are generated by recurrent synaptic activity mediated by AMPA receptors. J Physiol 542, 567–581.
Chauvel P, Trottier S, Vignal JP & Bancaud J (1992). Somatomotor seizures of frontal lobe origin. Adv Neurol 57, 185–232.[Medline]
Coetzee WA, Amarillo Y, Chiu J, Chow A, Lau D, McCormack T, Moreno H, Nadal MS, Ozaita A, Pountney D, Saganich M, Vega-Saenz d. M & Rudy B (1999). Molecular diversity of K+ channels.
) or horizontal pathways (
). Vertical pathways were activated by placing the stimulating electrode below the recording electrode in layer V, while horizontal pathways were activated by placing the stimulating electrode adjacent in layer III. Stimulation alternated between each electrode every 10 s. The data are the average of the number of experiments indicated (n). The upper traces show an example of the effect of low TTX on synaptic responses evoked before and 20 min after application of the drug. B, effect of IT blockers on evoked field potential responses. Each plot shows the effect of Ni2+, ethosuximide and mibefradil on the amplitude of field potential responses evoked in layer III by stimulating vertical pathways. The different symbols correspond to different doses of the drugs as indicated. Vertical pathways were activated by placing the stimulating electrode below the recording electrode in layer V. Stimulation was delivered every 20 s. The data are the average of the number of experiments indicated (n). The upper traces show an example of the effect of ethosuximide (10 mM) on synaptic responses evoked before and 20 min after application of the drug. C, population data on the effect of INap and IT blockers on the amplitude of evoked field potential responses in vertical pathways of motor cortex measured 30 min after onset of drug application. (*P < 0.01, t test; significant suppression compared to baseline). D, effect of TTX (100 nM) on simultaneously recorded intracellular and field potential responses from layer III evoked by a stimulating electrode in layer V. Each response is the average of 10 response trials before and during the drug application. Note the enhancement of the field potential response amplitude and the slight depression of the simultaneously recorded intracellular EPSP. E, effect of riluzole (10 µM) on simultaneously recorded intracellular and field potential responses from layer III evoked by a stimulating electrode in layer V. Each response is the average of 10 response trials before and during the drug.
