INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

It is well known that long-term potentiation (LTP) in the hippocampal CA1 pyramidal cells, which may be crucial for learning and memory, requires postsynaptic calcium increase by activation of N-methyl-D-aspartate (NMDA) receptors (Bliss and Collingridge 1993). In addition, recent pharmacological investigations have shown that, of known subtypes of voltage-dependent calcium channels (VDCCs), at least L-, P/Q-, T-, and R-type VDCCs contribute to the induction of hippocampal LTP (Grover and Teyler 1990; Ito et al. 1995; Magee and Johnston 1997). These subtypes of VDCCs seem to be differently distributed in the hippocampal CA1 pyramidal cell; T- and R-type VDCCs are localized more abundantly at the distal dendrite than at the proximal dendrite and soma, whereas L-type VDCCs are generally found at the proximal dendrite and soma (Magee and Johnston 1995a,b). In fact, the calcium imaging technique revealed that, at the more distal area of the apical dendrite, the greater component of calcium increase induced by action potentials is mediated by T- and/or R-type VDCCs, while other subtypes tend to mediate the calcium influx more greatly at the more proximal area (Christie et al. 1995).

Interestingly, the properties of membrane potential at the distal dendrite are also remarkably different from those at the proximal dendrite and soma. The amplitude of dendritic action potentials becomes smaller as they are backpropagating more distally along the apical dendrite. Moreover, even at the same site of the distal dendrite, the amplitude shows a gradual reduction when they are firing repetitively (Spruston et al. 1995). Our previous study has demonstrated that membrane depolarization during LTP induction is indeed much smaller at the distal apical dendrite than at the soma (Isomura and Kato 2000). Such backpropagating action potentials have been considered to be critical for the induction of hippocampal LTP (Häusser et al. 2000; Linden 1999; Magee and Johnston 1997; Paulsen and Sejnowski 2000). Thus the distribution of VDCCs and the amplitude of dendritic action potentials, both of which may be related to LTP induction, are apparently dependent on the distance from the soma along the apical dendrite. However, little attention has so far been paid to any distance-dependent difference in the induction of LTP at the dendritic synapses. In the present study, we investigated whether the contribution of T- and R-type VDCCs to LTP induction is distance-dependent along the proximal-distal dendritic axis in the rat hippocampal CA1 pyramidal cells, by using low concentration of Ni²⁺ as a T- and R-type VDCC blocker (Avery and Johnston 1996; Magee and Johnston 1995a).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Hippocampal slices (300 µm thick) were prepared from ether-anesthetized Wistar rats (postnatal days 18-30) with a microslicer and allowed to recover in artificial cerebrospinal fluid (ACSF) at room temperature for >1 h (Isomura and Kato 1999, 2000). The ACSF consisted of the following (in mM): 124 NaCl, 2.5 KCl, 1.2 KH₂PO₄, 26 NaHCO₃, 1.2 MgSO₄, 2.5 CaCl₂, and 25 D-glucose and was saturated with 95% O₂-5% CO₂. Each slice was then transferred into a submerged-type recording chamber continuously circulated with the ACSF at 30-32°C. Low concentration of NiCl₂ (25 µM; ~IC₅₀ for T-type VDCC; Avery and Johnston 1996) dissolved in the KH₂PO₄-omitted ACSF was bath-applied to block the Ni²⁺-sensitive VDCCs. Nimodipine (Tocris Cookson, MO) was prepared from a 50 mM stock solution in ethanol and kept in the dark, and bath-applied to block L-type VDCCs (final concentration 10 µM). DL-2-amino-5-phosphonovaleric acid (DL-APV; 100 µM, Tocris) was also applied to block NMDA receptors. Field potentials were recorded by a glass electrode filled with 2.5 M NaCl (2-5 M), which was placed in a "proximal," "middle," or "distal" recording site in the stratum radiatum of the CA1 region. The proximal, middle, and distal recording sites were 50, 150, and 200 µm distant from the stratum pyramidale, respectively. The signals were amplified with an amplifier (Axoclamp 2-B; Axon Instruments, CA), filtered at 3 kHz, and digitized at 5 kHz with an A/D interface (Digidata 1200; Axon Instruments). The Schaffer collaterals were stimulated by a bipolar tungsten stimulating electrode placed near each recording site. To monitor changes of field excitatory postsynaptic potentials (EPSPs), test pulses were delivered every 12 s with the intensity adjusted to evoke field EPSPs ranging from 0.4 to 0.6 mV in amplitude (2-5 V, 200 µs). Five consecutive field EPSP slopes were averaged for each data point. LTP was induced by two trains of theta-burst stimulation (TBS; 10 bursts at 5 Hz, each burst consisting of 4 pulses at 100 Hz) at an interval of 20 s, with the same intensity as test pulses, unless otherwise mentioned. For current-source-density (CSD) analysis, the single electrode was penetrated into the CA1 region sequentially every 50 µm along the proximal-distal dendritic axis, and the spatial distribution of current sink and source was calculated by a homogeneous CSD approximation (Holsheimer 1987). In some experiments, population spikes evoked by the distal Schaffer-collateral stimulation were recorded from the stratum pyramidale where most of the somata are located. To examine after-depolarizing potential (ADP), membrane potentials were intracellularly recorded from the soma in the stratum pyramidale by a sharp glass microelectrode filled with 2.5 M K-acetate (75-100 M). They were amplified with the same amplifier, filtered at 10 kHz, and digitized at 20 kHz (resting membrane potentials, 62 to 70 mV; input resistances, >25 M). The duration of ADP was measured as the time when the potential stayed above one-half of the spike threshold after spike onset. All data were expressed as mean ± SD in the text, and Student's t-test or analysis of variance (ANOVA) was applied for statistical comparisons.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Low concentration of Ni²⁺, a T- and R-type VDCC blocker, is known to inhibit the induction of hippocampal LTP (Ito et al. 1995; Magee and Johnston 1997), but it is not clear at all whether the Ni²⁺-sensitivity of LTP induction was dependent on the distance from the soma, along the apical dendrite, of the CA1 pyramidal cell. To assess the possible distance-dependency, we recorded field potentials from one of the three distinct recording sites (proximal, middle, and distal dendritic sites) in the stratum radiatum of each hippocampal slice, and field EPSPs were evoked by stimulating the Schaffer collaterals near the recording site. The field EPSPs that were evoked and recorded at one recording site seem to be successfully distinguished from those at the other sites, because CSD analysis revealed that the distribution of current sink by synaptic stimulation was usually restricted within approximately 100 µm near the stimulating position (Fig. 1A). Thus field potential recordings made it possible to detect site-specific synaptic responses. LTP examined here was induced by the standard TBS protocol, and NMDA receptor activation was, as broadly believed, truly prerequisite for the induction of this type of LTP, because the NMDA receptor antagonist DL-APV completely blocked the LTP induction at any recording position (proximal, 104.0 ± 14.9%, n = 4; distal, 101.9 ± 17.1%, n = 5; Figs. 1B and 2E).

View larger version (18K): [in this window] [in a new window]   Fig. 1. Long-term potentiation (LTP) induced at the distal and proximal apical dendrites of the hippocampal CA1 pyramidal cells. A: field potential recordings and current-source-density (CSD) analysis in the CA1 region of rat hippocampal slice preparations. The Schaffer collaterals were stimulated with constant intensity by a bipolar stimulating electrode at the same positions (asterisks), and field excitatory postsynaptic potentials (fEPSPs) were recorded at various recording sites along the apical dendrites of the pyramidal cells. Current density at each recording site was calculated from these field potentials at 3.6 ms after the stimulation (left traces and open circles, proximal stimulation; right traces and closed circles, distal stimulation). Note that the current sink is restricted almost within 100 µm near the stimulating position. s.lm., stratum lacunosum-moleculare; s.r., stratum radiatum; s.p., stratum pyramidale; s.o., stratum oriens; alv, alveus. Scale bars: 10 ms and 0.5 mV. B: representative data showing an essential role of N-methyl-D-aspartate (NMDA) receptor activation in the induction of LTP anywhere within the stratum radiatum. The NMDA receptor antagonist, 100 µM DL-2-amino-5-phosphonovaleric acid (DL-APV), completely blocked the induction of LTP by theta-burst stimulation (TBS) at both the distal (top) and proximal (bottom) recording sites (open circles, control; closed circles, DL-APV).

View larger version (22K): [in this window] [in a new window]   Fig. 2. Ni2+-sensitive LTP induction dependent on the distance from the somata along the apical dendrites. A-C: effects of low concentration of Ni2+, a T- and R-type calcium channel blocker, on TBS-induced LTP at the proximal (A), middle (B), and distal (C) recording sites. Top: typical averaged traces of field excitatory postsynaptic potentials (EPSPs) before (thin) and 40 min after (thick) the TBS delivery in the absence () and presence (Ni2+) of 25 µM Ni2+. Scale bars: 10 ms and 0.5 mV. Bottom: summarized TBS-induced LTP in the control (open circles, with error bars indicating SE) and Ni2+-applied (closed circles) conditions. D: normal LTP induction in the presence of 10 µM nimodipine, an L-type VDCC blocker, at the proximal (left) and distal (right) recording sites. E: comparison of the effects of Ni2+, nimodipine, and DL-APV on the magnitude of LTP at the proximal and distal dendritic areas. The number of recorded slices for each group is shown in parentheses.

At the proximal recording site of the apical dendrite, robust TBS-induced LTP was observed similarly in the absence and presence of 25 µM Ni²⁺ [(), 161.5 ± 37.6%, n = 8; Ni²⁺, 172.3 ± 17.4%, n = 6; P > 0.5; Fig. 2A]. By the same concentration of Ni²⁺, however, the induction of LTP was inhibited mildly at the middle dendritic area [(), 158.1 ± 14.0%, n = 6; Ni²⁺, 131.3 ± 20.3%, n = 4; P < 0.04; Fig. 2B], and more strongly at the distal dendritic area [(), 150.1 ± 21.4%, n = 8; Ni²⁺, 117.6 ± 17.9%, n = 8; P < 0.005; Fig. 2C]. Thus although there was no significant difference among the magnitudes of induced LTP at the three distinct distances in the control condition (ANOVA, P > 0.7), those were largely decreased in a distance-dependent manner in the presence of Ni²⁺ (ANOVA, P < 0.001). The absolute slopes of field EPSPs before LTP induction were almost equal in the control and Ni²⁺-applied conditions at any recording site (e.g., 0.60 ± 0.09 V/s vs. 0.59 ± 0.17 V/s at the distal area; P > 0.8), suggesting that this inhibition is not due to the difference in baselines of EPSP amplitude. In contrast to the effect of Ni²⁺, the susceptibility of LTP at neither the proximal nor distal areas was significantly influenced by the potent L-type VDCC blocker, nimodipine (proximal, 158.8 ± 47.5%, n = 8; distal, 147.2 ± 19.7%, n = 6; Fig. 2, D and E). Thus Ni²⁺ and nimodipine failed to suppress LTP completely, unlike DL-APV; other VDCCs such as P/Q-type channels may compensate loss of the calcium component for LTP induction (Ito et al. 1995).

How does Ni²⁺ inhibit the induction of LTP at the distal dendritic synapses? It might primarily be expected that Ni²⁺ would directly block dendritic calcium influx through T- and R-type VDCCs during membrane depolarization, thereby leading to inhibit the induction of distal LTP. Otherwise, Ni²⁺ might affect membrane excitability during LTP induction to change the susceptibility of LTP indirectly. Therefore the following experiments were designed to examine possible modulatory effects of Ni²⁺ on membrane depolarization, such as EPSPs and action potentials, during LTP induction. First, when recorded at the distal synapses, neither field EPSP amplitude [(), 0.448 ± 0.059 mV; Ni²⁺, 0.447 ± 0.059 mV; n = 5; P > 0.8] nor paired-pulse facilitation (PPF) [(), 133.3 ± 15.8%; Ni²⁺, 134.1 ± 14.7%; n = 5; P > 0.4] exhibited significant changes before and during the Ni²⁺ application (Fig. 3A). Since field EPSP amplitude and PPF are good parameters to reflect relative changes, respectively, in postsynaptic responsibility and presynaptic release probability, these results suggest that the basal synaptic transmission itself is unchanged by Ni²⁺, which is consistent with previous observations (Gillessen and Alzheimer 1997; Ito et al. 1995).

View larger version (18K): [in this window] [in a new window]   Fig. 3. Indirect inhibitory effect of Ni2+ on membrane excitability that may affect LTP induction at the distal apical dendrite. A: Ni2+-insensitivity of basal synaptic transmission at the distal dendritic synapses. Left: fEPSP traces during paired-pulse stimulation before () and during (Ni2+) bath-application of 25 µM Ni2+. Scale bars: 10 ms and 0.2 mV. Right: no change in fEPSP amplitude (top) or paired-pulse facilitation (PPF) (bottom) by the Ni2+ application. B: inhibitory effect of Ni2+ on somatic spiking evoked by distal synaptic input. Left: population spikes (PS) elicited by the distal synaptic stimulation before (), during (Ni2+), and after (washout) 25 µM Ni2+ perfusion. Scale bars: 10 ms and 0.2 mV. Right: slight but significant reduction in the amplitude of PS by Ni2+ application. Individual paired squares represent the change in the absolute amplitude of population spikes. Inset: intrasomatically recorded EPSP before (thick) and during (thin) Ni2+ application. Scale bars: 10 ms and 1 mV. C: Ni2+-insensitivity of after-depolarizing potential (ADP). Left: somatic recordings of ADPs following discharge by brief current injection in the absence () and presence (Ni2+) of 25 µM Ni2+. Each horizontal arrow indicates the duration of ADP. Scale bars: 10 ms and 2.5 mV. Right: no difference in the duration of ADP between the control and Ni2+-treated conditions. D: only partial recovery from the Ni2+-dependent inhibition of LTP by the reinforcement of somatic spiking during LTP induction. Left: population spikes evoked by the distal synaptic stimulation; the stimulation with 1.5-fold stronger intensity was sufficient to overcome the reduction in the amplitude of population spikes in the presence of 25 µM Ni2+. Right: pooled data of LTP poorly rescued even by the 1.5-fold stronger TBS (closed diamond, n = 10) at the distal dendritic site in the presence of 25 µM Ni2+. Open and closed circles indicate the same data as shown in Fig. 2C.

Next, by recording population spikes from the stratum pyramidale, we examined whether Ni²⁺ might suppress the somatic discharge evoked by distal synaptic input. Recent studies by local Ni²⁺ application imply that low-threshold, Ni²⁺-sensitive calcium channels may actively amplify EPSPs at the apical dendrite of the hippocampal pyramidal cell, to signal distal synaptic input to the soma (Gillessen and Alzheimer 1997). We observed a Ni²⁺-dependent reduction of intrasomatically recorded EPSPs elicited by the distal synaptic stimulation (Fig. 3B, inset). The amplitude of population spikes was also slightly but significantly reduced to 85.5 ± 10.9% by Ni²⁺ perfusion (n = 7; P < 0.02; Fig. 3B). Once membrane depolarization reached the spike threshold, action potentials were generated almost normally in the presence of Ni²⁺ (data not shown), and the duration of ADPs, which has been implicated in the ability of burst firing (Andreasen and Lambert 1995; Azouz et al. 1996), was not affected irrespective of the presence of Ni²⁺ [(), 12.4 ± 2.7 ms; Ni²⁺, 12.9 ± 3.3 ms; n = 6; P > 0.7; Fig. 3C]. Thus our results provide the possibility that Ni²⁺ may suppress the somatic discharge by reducing dendritic EPSP amplification, leaving the distal field EPSP unchanged.

Finally, we attempted to estimate how much such a slight reducing effect of Ni²⁺ on spike generation actually contributes to the inhibition of LTP induction. The stronger synaptic stimulation is predicted to evoke the larger EPSPs and to generate action potentials at the higher rate. Indeed, the synaptic stimulation with 1.5-fold stronger intensity usually generated much greater population spikes in the presence of Ni²⁺ (Fig. 3D, left). This strong reinforcement will successfully overcome the reduction in spike generation during LTP induction, whereas it will fail to overcome the direct blockade of calcium influx through Ni²⁺-sensitive VDCCs. In the presence of Ni²⁺, inhibited LTP at the distal dendritic site was not significantly recovered even by the reinforced TBS with 1.5-fold stronger intensity, although only small enhancement was observed (129.0 ± 23.0%, n = 10; Fig. 3D, right). Furthermore, a similar reinforcement of spiking by stimulating the axons antidromically also failed to weaken the inhibition (119.5 ± 8.2%, n = 3). These indicate that Ni²⁺ seems to suppress the induction of LTP at distal dendritic synapses predominantly by the direct blockade of calcium influx, rather than by changing membrane excitability.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Here we showed a distance-dependent inhibition of LTP induction by low concentration of Ni²⁺ in the apical dendrites of the hippocampal CA1 pyramidal cells. Our results suggest that this inhibition of LTP induction is primarily due to the direct blockade of calcium influx mediated by Ni²⁺-sensitive VDCCs. It has previously been demonstrated that Ni²⁺ inhibits LTP induction even in the condition that neurons are forced to discharge by intracellular current injection (Magee and Johnston 1997). This finding strongly supports the possible inhibition of LTP induction by the direct blockade of calcium influx. Furthermore, the distance-dependency of Ni²⁺-sensitivity of LTP induction is consistent with the abundant localization of T- and R-type VDCCs at the distal dendrite (Magee and Johnston 1995a). Indeed, calcium imaging studies made it clear that the dendritic calcium increases triggered by subthreshold EPSPs or backpropagating action potentials are highly sensitive to Ni²⁺, especially at the distal dendrite (Christie et al. 1995; Magee et al. 1995). Thus the distinct VDCC distributions may underlie the differential inducibility of LTP along the apical dendrite of the pyramidal cell.

It is also possible to consider that Ni²⁺ occludes dendritic EPSP amplification (Gillessen and Alzheimer 1997; Urban et al. 1998), thereby depressing the somatic generation of backpropagating action potentials, which are necessary for LTP induction (Magee and Johnston 1997). This will be, however, only a minor effect, because a strong reinforcement of spiking hardly enhanced the magnitude of LTP in the presence of Ni²⁺. Alternatively, one can explain that the inhibition of LTP might be due to a Ni²⁺-mediated occlusion of burst firing rather than solitary discharge, given that bursting action potentials are more effective for inducing LTP than single action potentials (Pike et al. 1999; Thomas et al. 1998). In the present study, however, Ni²⁺ did not affect the duration of ADP that has been implicated in burst firing (Andreasen and Lambert 1995; Azouz et al. 1996). Moreover, it is unlikely that Ni²⁺ might modify the spike backpropagation, which has been reported not to be influenced by manipulation of extracellular calcium (Jung et al. 1997; Poolos and Johnston 1999).

As action potentials at the distal dendrite are much smaller than at the proximal dendrite during LTP induction (Isomura and Kato 2000), low-threshold VDCCs such as T-type channels should be activated more readily at the distal dendrite, compared with high-threshold VDCCs, including known R-type channels. Unfortunately, we could not determine which subtype of these two VDCCs actually contributes to the induction of distal LTP, because no specific organic blocker for either subtype has been available yet. Recently, _1G, _1H, and _1I subunits have been identified as specific ₁ subunits with biophysical characteristics of T-type VDCCs (Craig et al. 1999; McRory et al. 2001), and at least the _1G subunit has been shown to exist at the apical dendrites of the hippocampal pyramidal cells (Craig et al. 1999). On the other hand, it has been reported that the only cloned subunit of R-type VDCCs, the _1E subunit, is not distributed at any dendrite in the CA1 region (Yokoyama et al. 1995). We cannot, however, exclude the possibility that other T- or R-type VDCCs with unknown ₁ subunits may mediate the induction of distal LTP. Further molecular characterization of VDCCs is needed to elucidate the distance-dependent induction of LTP by their distinct VDCC subtypes.

What is the functional significance of the distance-dependency of Ni²⁺-sensitive synaptic plasticity? Activation of NMDA receptors is undoubtedly prerequisite for the induction of input-specific hippocampal LTP. It is conceivable, however, that NMDA receptor-mediated calcium component alone may probably be insufficient to induce LTP completely, and, therefore additional calcium ions should be supplied from VDCCs for inducing the NMDA receptor-dependent LTP. Especially at the distal dendrite where action potentials are too small to activate high-threshold VDCCs, the calcium influx through low-threshold, Ni²⁺-sensitive VDCCs might play a critical role in the induction of NMDA receptor-dependent LTP. This will make it possible to keep the LTP inducibility constant at any synaptic site of the apical dendrites. In fact, the magnitude of normal LTP is maintained at similar levels throughout the stratum radiatum, as demonstrated in the present study. Such equal inducibility of input-specific synaptic plasticity at the Schaffer collateral synapses may be advantageous to allow individual inputs to be equally processed, leading to store a large amount of information efficiently in the hippocampus.