| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Department of Neurology and Neurosurgery, Montreal Neurological Institute, McGill University, Montreal, Quebec H3A 2B4, Canada; and 2Dipartimento di Scienze Fisiologiche-Farmacologiche Cellulari-Molecolari, Sezione di Fisiologia Generale e Biofisica Cellulare, Università degli Studi di Pavia, 27100 Pavia, Italy
Submitted 12 January 2004; accepted in final form 14 May 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
; Manns et al. 2003b; Scoville and Milner 1957) and brain imaging (Fernandez et al. 1999; Owen et al. 1996) studies in humans as well as neurophysiological studies in humans (Fried et al. 1997) and animals (Suzuki et al. 1997; Young et al. 1997). Significantly, it has been shown that the level of sustained entorhinal neural activity correlates positively with "declarative" memory performance (Fernandez et al. 1999) and during working memory tasks, neurons in the EC may display sustained activity during the delay period and enhanced or suppressed responses to matching stimuli (Hasselmo et al. 2000; Suzuki et al. 1997; Young et al. 1997).
The cholinergic innervation of the EC by the basal forebrain (Mesulam et al. 1983) is very robust and cholinergic fibers profusely innervate EC layer II neurons (Alonso and Amaral 1995). EC layer II funnels polysensory cortical signals into the hippocampus (Insausti et al. 1987) thereby occupying a critical position in the neocortical-hippocampal-neocortical memory circuit. It is well established that the cholinergic system modulates the level and patterns of cortical activation (Celesia and Jasper 1966) and that cholinergic influences contribute to cortical plasticity (Bear and Singer 1986; Dykes 1997; Winkler et al. 1995), state-dependent learning (Shulz et al. 2000) and declarative memory formation (Aigner and Mishkin 1986; Chang and Gold 2003; Hasselmo et al. 1996; Tang et al. 1997). The contribution of the basal forebrain cholinergic system to memory processes is highlighted by the degeneration of cholinergic neurons in Alzheimer's disease (Dunnett and Fibiger 1993), and cholinergic modulation of EC neuronal activity might be critical in the normal memory function of this structure (Hasselmo et al. 2000). Indeed, cholinergic activation of EC promotes the emergence of oscillatory neuronal dynamics (Buzsáki 1996) which are implicated in learning and memory processes (Hasselmo 1999b; Huerta and Lisman 1993; Larson and Lynch 1986).
Principal neurons from layer II of the EC belong to the broad category of regular spiking neurons (Alonso and Klink 1993; Connors and Gutnick 1990) and respond to step depolarizations with slowly adapting trains of action potentials followed by a typical, slow afterhyperpolarization (Alonso and Klink 1993) (Fig. 1A1). Previous current-clamp investigations carried out in EC slices with intracellular sharp electrodes have shown that cholinergic stimulation of these neurons through muscarinic receptors affects their electrophysiological state and properties in at least two ways: on the one hand, it produces a basal, sustained depolarization; on the other hand, it drastically modifies their firing behavior by inducing Ca2+-dependent plateau potentials and bursting activity (Klink and Alonso 1997). In addition, in a recent voltage-clamp analysis based on whole cell, patch-clamp experiments, we showed that the muscarinic-receptor-dependent depolarization of EC layer II neurons largely relies on the activation of a nonspecific cation inward current that we referred to as INCM, and reported preliminary evidence suggesting that INCM could be bidirectionally regulated by Ca2+ influx (Shalinsky et al. 2002). A Ca2+- and activity-dependent regulation of INCM could introduce plasticity in the firing behavior of the cells so as to lead to afterdischarges and make the cellular responses dependent on the previous firing history. The goal of this patch-clamp study was therefore twofold. First, we needed to re-examine, under whole cell recording conditions, the firing repertoire of EC layer II cells undergoing muscarinic stimulation and test the hypothesis that the occurrence of plateau-potential-driven activity is influenced by the previous firing history. Second, by carrying out a voltage-clamp analysis and using intracellular Ca2+ chelators, we aimed to verify the hypothesis that an activity- and Ca2+-dependent regulation of INCM is implicated in patterning spike discharge in EC layer II cells during cholinergic modulation. Our results show that, indeed, INCM appears to be subject to both Ca2+-dependent up- and downregulation and are consistent with an activity-dependent regulation of this current playing an important role in patterning the firing discharge of EC layer II cells during muscarinic modulation. Some of these data has been previously presented in preliminary form (Magistretti et al. 2001).
|
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Brain slices were prepared from male Long-Evans rats (100–250 g, i.e., 30–60 days of age) as previously described (Alonso and Klink 1993). Briefly, animals were quickly decapitated, and the brain was rapidly removed from the cranium and placed in a cold (4°C) Ringer solution containing (in mM): 124 NaCl, 5 KCl, 1.25 NaH2PO4, 2 CaCl2, 2 MgSO4, 26 NaHCO3, and 10 glucose (pH 7.4 by saturation with 95% O2-5% CO2). Horizontal slices of the retrohippocampal region were cut at 350–400 µm on a vibratome (Pelco Series 1000, Redding, CA) and then transferred to an incubation chamber in which they were kept submerged for at least a 1-h period at room temperature (
22°C).
Patch-clamp, whole cell recordings
The recording chamber was mounted on the stage of an upright microscope (see following text). Slices were transferred, one at a time, to the chamber and perfused with one of the extracellular solutions described in Table 1, according to the specific experimental purpose. In current-clamp experiments, extra- and intracellular solutions were always E1 and I1. Patch pipettes were fabricated from thick-wall borosilicate glass capillaries by means of a Sutter P-97 horizontal puller. The solutions used to fill the patch pipettes are also described in Table 1. When filled with one of these solutions, the patch pipettes had a resistance of 3–5 M
. Slices were observed with an Axioskop microscope (Zeiss, Oberkochen, FRG) equipped with a x40 water-immersion objective lens and differential-contrast optics. A near-infrared charge-coupled device (CCD) camera (Sony XC-75) was also connected to the microscope and used to improve cell visualization for identification of neuron types and during the approaching and patching procedures. With this equipment, the principal cells of EC layer II were easily distinguished based on their somato-dendritic shape, size, and position (Dickson et al. 2000). Patch pipettes were brought in close proximity to the selected neurons while manually applying positive pressure inside the pipette. Tight seals (>10 G) and the whole cell configuration were obtained by suction. Series resistance (Rs) was estimated on-line by canceling the fast component of whole cell capacitive transients evoked by –10-mV voltage steps with the amplifier compensation section (with the low-pass filter set at 10 kHz) and reading out the corresponding value, and was on average 16–18 M. Rs was always compensated by 40% with the amplifier's built-in compensation section. Current- and voltage-clamp recordings were performed at room temperature (22°C) using an Axopatch 1D amplifier or an Axopatch 200B amplifier (Axon Instruments, Foster City, CA). All current-clamp experiments were carried out using the Axopatch 200B amplifier in the "I-Clamp fast" mode, which, in the presence of Rs values like those typically obtained in our conditions, allows for reliable recording of membrane-voltage signals (Magistretti et al. 1996, 1998). The low-pass filter (–3 dB) was set at 5 kHz. In voltage-clamp recordings, the general holding potential was –60 mV unless otherwise indicated.
|
Carbamyl choline (carbachol, CCh) was delivered by bath superfusion at 5–30 µmol/l, during synaptic transmission block with kynurenic acid (1 mM) and picrotoxin (100 µM). Stock solutions of nifedipine (20 µM) were prepared using dimethylsulfoxide (DMSO) as the solvent, and stored at –20°C in the dark. Nifedipine aliquots were then re-dissolved in recording solution and applied through a light-shielded perfusion channel. All chemicals and reagents, including those listed in Table 1 and CCh, were purchased from Sigma (St. Louis, MO), except tetrodotoxin (TTX), which was purchased from Alomone Labs (Jerusalem, Israel).
Data acquisition
All recordings were stored by PCL coding on VHS tape (Neurocorder, Neurodata, New York). In voltage-clamp experiments, voltage protocols were commanded and current signals were acquired with a Pentium PC interfaced to an Axon DigiData 1200 interface, using the Clampex program of the pClamp software (V8.0, Axon Instruments). Ramp voltage protocols consisted in 1-s linear depolarizations from –100 to –50 mV (depolarization rate = 50 mV/s), always preceded by a 500-ms fixed step at –100 mV. Data stored on VHS tape was digitized and plotted off-line by sampling at 20 kHz using the Axoscope software (V1.1, Axon Instruments).
Data analysis
Whole cell recordings were analyzed by means of the Clampfit program of the pClamp software (Axon Instruments). Linear regressions were performed using Origin 6.0 (MicroCal Software, Northampton, MA). Average values were expressed as means ± SE. Statistical significance was evaluated, when not otherwise explicitly stated, by means of the two-tail Student's t-test for unpaired data or, in other cases, by applying the one-way ANOVA and/or the Bonferroni multiple comparison test. In current-clamp recordings, neurons were electrophysiologically categorized as stellate cells if they displayed robust time-dependent inward rectification (sag% >40%) (Dickson et al. 2000).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Under current-clamp conditions, we found that in all neurons examined (n = 43) triggering a short train of action potentials during bath application of CCh (5–20 µM) elicited a depolarizing afterpotential (DAP) that could sustain delayed spiking activity. Two exemplary cases are illustrated in Fig. 1, A2 and B. Afterdischarges displayed spike-frequency patterns qualitatively similar in all neurons examined, independently of whether the cells were electrophysiologically categorized as stellate or pyramidal-like neurons (Alonso and Klink 1993; Dickson et al. 2000). However, in response to comparable stimuli (0.1-nA, 1-s current steps, resting potential approximately –60 mV under superfusion with 5-µM CCh), pyramidal-like cells typically displayed afterdischarges with a more pronounced accelerating-decelerating spike-frequency pattern than stellate cells. This behavior was characterized by a significantly higher peak firing frequency during the afterdischarge in pyramidal cells (4.03 ± 0.69 Hz; n = 5) than in stellate cells (2.29 ± 0.26 Hz; n = 8) (P < 0.05) and a shorter (though not statistically significant) afterdischarge duration in pyramidal cells (15.24 ± 0.97 s) than in stellate cells (29.52 ± 7.57 s).
Importantly, in all cases tested (n = 13), we also noted that when a given suprathreshold stimulus was repeatedly applied, the DAPs that developed after each subsequent stimulus were first enhanced and then suppressed back to the initial level (Fig. 2A). This biphasic enhancement-suppression pattern always took place, whether (Fig. 2A1) or not (Fig. 2A2) an afterdischarge developed. However, when one of the stimuli did give rise to an afterdischarge, the responses to subsequent stimuli were always suppressed (Fig. 2A1). The suppression period after afterdischarges is better illustrated by the example shown in Fig. 2B. Note in Fig. 2B1 that a brief depolarizing stimulus led to an afterdischarge that lasted for 20 s and that immediately after this event the same stimulus triggered a spike train with minimal DAP. However, after a recovery period of 1 min, the stimulus was again capable of triggering an afterdischarge. Figure 2B2 illustrates another example of the phenomenon in the same neuron: in this case, the stimulus was re-applied at increasing intervals during the suppression period. Note that a stimulus interval of 27 s was required to elicit a subsequent afterdischarge. Although we did not systematically analyze the duration of the suppression period, the elicitation of two consecutive afterdischarges of equivalent duration required a minimal recovery period of 20 s (n = 5).
|
Voltage-clamp analysis of INCM responses to activating stimuli
Having established that during muscarinic modulation the activity pattern of EC layer II neurons is dependent on the previous firing history, we then tested the hypothesis that a bidirectional activity (Ca2+)-dependent regulation of INCM (Shalinsky et al. 2002) might, at least partially, underlie the observed firing phenomena. With this goal in mind, we first performed a series of voltage-clamp experiments to analyze the behavior of INCM in response to brief step depolarizations to voltage levels expected to trigger Ca2+ influx; 50- to 500-ms depolarizing voltage steps at 0 mV were applied prior to and during bath application of CCh. To maximize the isolation of INCM and improve clamp conditions, these experiments were initially carried out with Cs+ as the main intracellular cation (to completely block K+ conductances) and with 0.5 mM intra-pipette EGTA (intracellular solution I2) (Shalinsky et al. 2002). Results from a representative cell are illustrated in Fig. 3. Whereas 100-ms voltage steps at 0 mV caused no modifications in the holding current under control conditions, during the CCh response they triggered prominent, inward tail-like currents that decayed in seconds (Fig. 3B) and were followed by a transient depression of INCM (Fig. 3A). Qualitatively (see following text and Fig. 4 for quantification), the same result was observed in all cells tested in this way (n = 12).
|
|
), which is unlikely to be produced by voltage-gated Ca2+ channels due their very low conductance; and, finally, in a number of neuronal systems, muscarinic-receptor activation has been shown to produce negative (instead of positive) modulation of high-voltage-activated Ca2+ currents with a G-protein-dependent mechanism (e.g., Wanke et al. 1994), whereas low-voltage-activated, T-type Ca2+ channels, which can be enhanced by muscarinic receptor activation (Fisher and Johnston 1990), are expected to be largely inactivated in the present experimental conditions (holding potential was –60 mV). This notwithstanding, we directly addressed the preceding issue by comparing CCh-dependent Itail induction with CCh effects on VDCC. Depolarizing voltage steps that triggered prominent Itails during CCh delivery also evoked inward VDCCs (identified by their sensitivity to 200 mM Cd2+: not shown) that in all cases (n = 5) were markedly depressed by CCh (by 33% on average; Fig. 3C). These results clearly demonstrate that Itails cannot coincide with tail Ca2+ currents resulting from CCh potentiation of VDCC.
|
Because both Itails and INCM downregulation were triggered by step depolarizations able to activate VDCCs, we next examined whether these phenomena depend on changes in intracellular Ca2+ concentration induced by voltage-dependent Ca2+ entry through voltage-gated Ca2+ channels. A series of experiments was performed in which different concentrations of Ca2+ chelating agents were used in the intra-pipette solution, in the presence of Cs+ as the main intracellular cation (intracellular solutions I2–I5; Fig. 4).
Itail induction and INCM downregulation both proved to be markedly sensitive to intracellular Ca2+-buffering conditions. We first examined the current modifications triggered by 500-ms voltage pulses at 0 mV that were delivered as the first depolarizing stimuli after the full development of a steady INCM. In the presence of 0.5 mM intra-pipette EGTA (Fig. 4A1), Itails evoked by these stimuli were always prominent and fast-developing. At 400 ms from the end of the depolarizing step (a time long enough to allow passive transients to fully subside), Itail amplitude exceeded by 524.7 ± 107.7% (n = 7) that of the preceding basal INCM. Moreover, Itails were always followed by a transient INCM downregulation. At the peak of INCM downregulation, which was reached after 11.4 ± 1.9 s from the end of the depolarizing pulse, current amplitude was decreased on average by 41.0 ± 8.3% (n = 7). With increasing concentrations of intracellular Ca2+ chelators (5 mM EGTA, 5–30 mM BAPTA), Itails' peak amplitude became progressively smaller (Fig. 4A, 2–4), and their onset more slowly developing (Fig. 4A, 2 and 3, insets), such that in these cases Itail peak amplitude could be unequivocally determined. Average data on basal INCM and Itail peak amplitude for various intracellular-Ca2+ buffering conditions are illustrated in Fig. 4B. First of all, it is worth noting that the basal INCM response was not abolished even at the highest Ca2+-buffering capacity tested (30 mM BAPTA), neither was its peak amplitude significantly reduced by increasing concentrations of the Ca2+ chelators (P > 0.05 for the comparisons between 0.5 mM EGTA and all the other conditions; Bonferroni multiple comparison test), indicating that muscarinic-receptor-dependent induction of basal INCM does not have an absolute dependence on intracellular Ca2+. Second, in the presence of 30 mM intra-pipette BAPTA, Itails were almost completely abolished (Fig. 4, A4 and B), indicating that this form of muscarinic-receptor-dependent inward-current potentiation depends on intracellular Ca2+ concentration ([Ca2+]i) increases elicited by the depolarizing pulses applied. Additional information was provided by the intermediate Ca2+-chelator concentrations used. In the presence of 5 mM intracellular EGTA, the Itails evoked by the depolarizing pulses had a peak current amplitude over the basal INCM level (
I) that averaged 900 pA, with an average percent increase (% I) of 700% (Fig. 4C). In the presence of the same concentration of BAPTA, Itails, although still prominent, where significantly smaller with an average peak % I of 300% (Fig. 4C). (The different behaviors observed when using the same concentration of EGTA and BAPTA, which have similar KDs for Ca2+, are likely a consequence of the slower kinetics of EGTA in binding Ca2+ as compared with BAPTA: after abrupt triggering of Ca2+ influx to the cytoplasm, higher [Ca2+]i levels are expected to be transiently reached in the presence of EGTA than in the presence of BAPTA.) (10 mM) BAPTA caused a further decrease of % I at Itail peak (although not statistically significant with respect to 5 mM BAPTA; Fig. 4C). As illustrated in Fig. 4D, the increase in Itail inhibition due to more efficient or stronger Ca2+ buffering was paralleled by a progressive decrease in the speed of Itail onset, as quantified by the measurement of Itail time to peak.
Downregulation of INCM after a depolarizing pulse proved to be even more sensitive than Itails to increases in buffering capacity for intracellular Ca2+, as it was already abolished by 5 mM intra-pipette EGTA (Fig. 4A2). Remarkably, in the presence of 5 mM EGTA and 5 and 10 mM BAPTA, INCM downregulation was replaced by a prominent inward-current plateau (Fig. 4A, 2 and 3) and by an increase in current noise level, indicative of increased channel activity. Plateau amplitude exceeded by 58.7 ± 21.6% that of the preceding basal INCM in the case of 5 mM EGTA (n = 5), by 39.0 ± 15.0% in the case of 5 mM BAPTA (n = 4), and by 60.7 ± 13.8% in the case of 10 mM BAPTA (n = 15). (Plateau amplitude was measured 60–120 s after Itail peak, depending on the time course of Itail development).
The preceding results demonstrate that Itails, plateau-current induction, and INCM downregulation all depend on [Ca2+]i elevations triggered by membrane depolarization, and strongly suggest that these three distinct phenomena are caused by different [Ca2+]i dynamics (i.e., different degrees of [Ca2+]i increase and/or different spatiotemporal distribution of [Ca2+]i) secondarily to voltage-dependent Ca2+ entry. In particular, INCM downregulation would require higher levels of free intracellular Ca2+ than those resulting in induction of Itails and plateau currents, because INCM downregulation was abolished at lower [Ca2+]i chelating levels. This interpretation was also supported by experiments in which 50-, 100-, and 500-ms step depolarizations at 0 mV were delivered in sequence during an INCM response in the presence of 5 mM EGTA or 10 mM BAPTA in the recording pipette (Fig. 5). In the former condition, a 50-ms depolarizing pulse typically induced a small Itail, whereas a 100-ms pulse elicited a more pronounced Itail followed by an evident plateau; a 500-ms pulse, delivered on the plateau phase, evoked an even bigger Itail that was then followed by a sharp, profound INCM downregulation phase (Fig. 5A1). In the presence of 10 mM intra-pipette BAPTA, 50-, 100-, and 500-ms depolarizing pulses induced Itails of increasing amplitude, with the 500-ms pulse only being followed by a prominent plateau (Fig. 5A2). No INCM downregulation was ever observed in this condition. In the presence of 30 mM intra-pipette BAPTA, none of the preceding depolarizing stimuli caused significant modifications of the basal INCM amplitude level (Fig. 5A3). Average data on these effects are illustrated in Fig. 5B, 1 and 2 (peak Itail amplitude), and Fig. 5C (plateau potentiation or downregulation). These results not only indicate that Itails, plateaus and INCM downregulation are all Ca2+-dependent processes, but also strongly suggest that the induction, in sequence, of the three phenomena depends on increasing degrees of [Ca2+]i elevations (either in terms of absolute levels at one particular location or in terms of spatiotemporal distribution) that can be obtained with cumulative voltage-dependent Ca2+ entry.
Potentiation and downregulation of CCh-activated inward current require voltage-dependent Ca2+ entry
The intimate relation between voltage-dependent Ca2+ entry and induction of both Itails and plateaus as well as INCM downregulation during muscarinic stimulation was further confirmed by experiments carried out in the presence of partial block of voltage-gated Ca2+ channels. In EC layer II neurons, a current fraction equal to 45% of total VDCCs is sensitive to the L-type current blocker, nifedipine (L. Castelli and J. Magistretti, unpublished results). Therefore experiments were carried out in the presence of 20 µM nifedipine in the bath and 0.5 mM intra-pipette EGTA (recording solutions: E2, I2). In other control experiments, the extracellular solution was added with the same amount of vehicle (DMSO) used to dissolve nifedipine aliquots (final concentration: 0.1% vol/vol). In the presence of nifedipine, CCh application evoked the usual INCM, but the Itail responses that followed the application of 500-ms voltage pulses at 0 mV were significantly reduced in amplitude as compared with those obtained in control conditions (Fig. 6A). The percent increase in peak postdepolarization current was 151.3 ± 135.0% (n = 4) in the presence of nifedipine versus 605.1 ± 84.7% (n = 3) in control conditions (P < 0.05). INCM downregulation was also markedly reduced (percent decrease in peak current of 10.7 ± 8.4 vs. 40.0 ± 3.4%; P < 0.05).
|
Potentiated currents require the maintenance of the muscarinic stimulus and display the same sensitivity to Co2+ and current-voltage relationship as INCM
As discussed in the preceding text, induction of potentiated, plateau inward currents could be obtained in the presence of 5 mM intra-pipette EGTA or 10 mM intra-pipette BAPTA using appropriate depolarization patterns. In the continuous presence of CCh stimulation, the plateau currents, after an initial, transient decay phase, reached a steady level that showed little or no tendency to decrease further for 
5 min (n = 10) and up to 15 min (n = 2; Fig. 7A1). By contrast, if the continuous application of CCh with the perfusing solution was interrupted after the development of a plateau, the current decayed to the baseline within 4–5 min (Fig. 7A2; n = 3). These results show that plateau currents, like both the basal INCM and Itails, are strictly muscarinic-receptor-dependent and that they can persist for prolonged periods, provided the muscarinic stimulus is continuously present.
|
). Consistent with our previous observations, INCM responses could not be evoked in the presence of 2 mM Co2+ in the perfusing medium (Fig. 7B1; n = 4), suggesting that the underlying ion channels are sensitive to Co2+ block. Remarkably, the application of 2 mM Co2+ after the development of a plateau current evoked in control conditions resulted in a fast inhibition of the same current and a decay of the total CCh-induced inward current to the baseline (Fig. 7B2; n = 2). Hence, both INCM and potentiated currents show similar sensitivity to Co2+. Finally, the current-voltage relationship of both the basal INCM and depolarization-induced potentiated currents was studied using relatively short depolarizing voltage ramps (1 s, 50 mV/s; Fig. 8, A and B; recording solutions: E2, I4). The voltage range explored by these ramps was limited to a window between –100 and –60/–50 mV to prevent further current upregulation resulting from voltage-dependent Ca2+ entry at more positive voltages. The currents recorded during ramp application in control conditions, at the peak of the INCM response, and after the development of a Itail (evoked by a 500-ms pulse at 0 mV) in a representative cell are shown in Fig. 8B. The control ramp current was then subtracted from the INCM ramp current and the INCM ramp current was subtracted from the Itail ramp current to obtain the voltage dependence of the basal INCM and that of the pulse-evoked Itail, respectively. The current-voltage relationships of the subtracted currents are illustrated in Fig. 8C. Both the basal INCM and the upregulated current showed a linear behavior from –100 to –60 mV in all cases studied (n = 9). Moreover, the average, extrapolated reversal potential determined by linear regression of subtracted currents was not significantly different in INCM ramp currents (+13.8 ± 7.6 mV) and Itail ramp currents (+21.8 ± 4.2 mV; P = 0.3, t-test for paired data). The parallel sensitivity to muscarinic receptor stimulation and block by Co2+, together with the similarity in the current-voltage behavior, suggest that INCM, Itails, and plateau currents may represent different modulatory states of the same current.
|
The data presented in the preceding text have shown that Itails, plateau currents, and INCM downregulation can be sequentially induced in a voltage- and Ca2+-dependent manner with increasing degrees of intracellular-Ca2+ accumulation. In an attempt to further clarify how the activity- and Ca2+-dependent regulation of INCM may influence the generation of distinct firing patterns during an afterdischarge (i.e., sustained delayed firing or "bursty" discharge with a pronounced accelerating-decelerating pattern), we carried out further experiments involving trains of step depolarizations. To facilitate comparison with the current-clamp observations (Figs. 1 and 2), this last series of experiments was also performed using K+ as the main intracellular cation and 0.5 mM intra-pipette EGTA (intra-pipette solution I1). However, the extracellular solution contained Cs+ (5 mM) and Ba2+ (1 mM) (extracellular solution E3), a cocktail that blocks Ca2+-dependent slow afterhyperpolarization (sAHP; not shown) and isolates INCM from K+ currents modulated by CCh in a subthreshold voltage range (Shalinsky et al. 2002).
In these conditions, single 500-ms step depolarizations at 0 mV in the continuous presence of CCh (30 µM) always triggered a large Itail that slowly decayed toward the prestep current level (Fig. 9A and inset; n = 10). However, the application of a long train of identical depolarizing steps was followed by a transient decay of INCM to a level lower than that preceding the train. A slow, transient depression of INCM was invariably observed with trains of 200- to 500-ms step depolarizations to 0 mV delivered at 0.3–1 Hz (n = 6). Hence, in the presence of intracellular K+, potentiation can be dissociated from INCM downregulation, with downregulation requiring repetitive depolarizing stimuli to be induced. The observation that, using 0.5 mM intra-pipette EGTA, Itails elicited by equal stimuli are followed by INCM downregulation in the presence of Cs+, but not K+, as the main intracellular cation can be easily explained on the basis of our conclusion that down-regulation requires higher levels of [Ca2+]i increase than potentiation: indeed, higher [Ca2+]i rises, able to induce INCM downregulation, are more likely to occur in an in situ neuron during poorly clamped dendritic Ca2+ events, which are expected to be greatly facilitated by Cs+ block of K+ conductances. Alternatively, VDCC could by differently affected by intracellular K+ versus Cs+ (see Brette et al. 2003).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
). In the present study, we provide evidence that INCM is Ca2+ sensitive and, most importantly, that it can be substantially modified in amplitude in opposite directions by voltage-dependent Ca2+ influx, thus providing a simple mechanism by which cholinergic input can modulate the firing behavior of these cortical neurons in an activity-dependent manner. First, we found that, during INCM induction, step depolarizations able to activate VGCC were followed by transient, slow tail-like inward currents that substantially potentiated the effect of muscarinic stimulation in terms of induction of depolarizing current. Postdepolarization tail-like currents (Itails) were sensitive to increasing concentrations of intracellular Ca2+-chelating agents, although their complete abolishment required very strong intracellular-Ca2+ buffering (30 mM BAPTA). Itails were reduced or suppressed by VGCC block or in the absence of extracellular Ca2+, indicating that they depend on [Ca2+]i elevations secondary to voltage-dependent Ca2+ entry.
Second, under particular conditions, Itails were followed by a marked depression of predepolarization INCM amplitude. Similarly to Itail induction, INCM downregulation depended on [Ca2+]i elevations caused by voltage-dependent Ca2+ entry. However, INCM downregulation showed higher sensitivity to increases in intracellular-Ca2+ buffering than Itail induction, being readily abolished by 5 mM intracellular EGTA. This suggests that INCM downregulation requires larger [Ca2+]i rises (either in absolute terms or in terms of spatiotemporal distribution) than those sufficient to trigger Itails. Consistent with this interpretation, during an INCM response evoked in intracellular K+, trains of step depolarizations expected to cause smaller degrees of Ca2+ influx resulted in sustained "plateau-like" increases in inward-current amplitude, whereas trains of step depolarizations expected to cause larger Ca2+ influx produced a more pronounced, transient current potentiation and were followed by the apparent development of INCM depression. Thus [Ca2+]i elevations can result in either potentiation or depotentiation of the muscarinic induction of inward current, depending on the pattern of depolarization.
Third, recordings carried out in "intermediate" conditions of intracellular-Ca2+ buffering revealed that Itails may be followed by inward-current plateaus. These plateaus were always accompanied by an increase in the current noise level, indicating an increased channel activity, and were strictly dependent on the maintenance of the muscarinic stimulus. Our results indicate that the three phenomena of Itails, current plateaus, and INCM downregulation may be induced in sequence with increasing degrees of intracellular-Ca2+ accumulation.
Our data did not allow us to definitely establish whether the basal INCM and depolarization-induced potentiated currents (which included Itails and plateau currents) are based on the same channel population. However, INCM and potentiated currents shared several common features: 1) strict dependence on muscarinic receptor activation; 2) suppression by 2 mM Co2+; 3) linear current-voltage dependence between –100 and –60/–50 mV; 4) downregulation in a voltage- and Ca2+-dependent manner (see, for instance, Fig. 9B); 5) and association to an increase in current noise, indicative of increased activity of a population of channels characterized by relatively high conductance. We have previously estimated that the channel activity underlying INCM has a single-channel conductance of 14 pS (Shalinsky et al. 2002). The preceding observations are consistent with a homogeneous population of channels being responsible for both the basal INCM and Ca2+-potentiated currents, although they do not exclude the possibility that distinct channels sharing some common features generate INCM, Itails, and plateau currents.
In addition to EC neurons, muscarinic-dependent plateau potentials have been shown in a variety of cortical neuronal populations including hippocampal pyramidal cells (Caeser et al. 1993; Fraser and MacVicar 1996; Young et al. 2003) and interneurons (McQuiston and Madison 1999), and neocortical pyramidal cells (Andrade 1991; Haj-Dahmane and Andrade 1998; Schwindt et al. 1988) and typically suggested to reflect the activation of a Ca2+-dependent nonspecific cation current (Caeser et al. 1993; Haj-Dahmane and Andrade 1998; Schwindt et al. 1988). A noteworthy case is that of prefrontal cortex pyramidal cells in which, similarly to EC neurons, the mechanism of sustained muscarinic depolarization has also been shown to rely on the activation of a nonspecific cation current (Haj-Dahmane and Andrade 1996); and it also was hypothesized that both the basal muscarinic-receptor-induced, sustained inward current and the current generating the depolarizing afterpotential reflect a common underlying mechanism (Haj-Dahmane and Andrade 1998). In fact, in smooth muscle, the muscarinic-receptor-activated cation current, ICAT, displays a very prominent, transient Ca2+-dependent upregulation (Inoue and Isenberg 1990; Kim et al. 1998). As discussed in the preceding text, our data suggest that in EC layer II neurons, a Ca2+-dependent upregulation of the same channel activity responsible for INCM may be the mechanism underlying the generation of potentiated currents.
To our knowledge, no previous study carried out in vertebrate neurons has shown a Ca2+-dependent downregulation of a neurotransmitter-activated inward current operated by metabotropic-type receptors. Nevertheless, Ca2+-dependent inactivation is a phenomenon well known to occur in a variety of cationic channels, including voltage-activated Ca2+ channels (Gutnick et al. 1989), N-methyl-D-aspartate receptors (Legendre et al. 1993), cyclic nucleotide-activated cation channels (Zufall et al. 1991), and the light-dependent channels (trp and trpl) in Drosophila photoreceptors (Hardie and Minke 1994; Ranganathan et al. 1991). It seems noteworthy that in the invertebrate visual cascade, Ca2+ influx mediates both an early facilitation followed by a somewhat slower downregulation of the light-induced current that displays transient inactivation (Hardie and Minke 1995; Scott et al. 1997). Overall, similar properties also characterize the INCM current expressed by EC neurons and, as discussed by Shalinsky et al. (2002), suggest that the channels underlying INCM may belong to the TRP family (Clapham 2003; Harteneck et al. 2000; Voets and Nilius 2003). Indeed, there has been increasing evidence indicating that TRP channels may mediate many metabotropic responses in CNS neurons (Chuang et al. 2002; Clapham et al. 2001; Gee et al. 2003; Strubing et al. 2001), and a recent report has shown that in cerebellar Purkinje cells a metabotropic slow excitatory postsynaptic current is mediated by the TRPC1 cation channel (Kim et al. 2003). Also, a recent report has shown that heterologous expression of human TRPM5 in HEK-293 cells generates a nonselective cation channel that carries Na+, K+, and Cs+ equally well and is regulated by [Ca2+]i, being activated at low and inhibited at high [Ca2+]i levels, thus resulting in a bell-shaped dose-response curve for [Ca2+]i (Prawitt et al. 2003). A similar Ca2+-dependent regulation is also manifested by TRPM4 channels, which are good candidates as mediators of plateau potentials (Nilius et al. 2003).
Interactions between physiological processes involving positive and negative feedback mechanisms allow physiological signals to exhibit emergent properties, most notably plateau potentials and oscillation. The Ca2+-dependent up- and downregulation of muscarinic-receptor-dependent inward current(s) are likely to participate in the emergence of plateau potentials and oscillatory trains of activity in EC layer II cells. Indeed, some key features of such firing behaviors showed clear similarities with the modulatory characteristics of muscarinic inward current(s). In particular, 1) in current clamp (CC), afterpotentials of variable amplitude could be evoked by trains of action potentials; likewise, in voltage clamp (VC), Itails followed depolarizing protocols; 2) in CC, successive afterpotentials could sum up to reach firing threshold, which could result in sustained low-frequency firing; likewise, in VC, repetitive depolarizations could produce summation of Itails and plateau currents; and 3) in CC, afterdischarges superimposed to afterpotentials were self-terminating and followed by a period of suppression; likewise, in VC, high-frequency trains of depolarizing pulses were followed by INCM downregulation and transient depression.
We also noted that some neurons (preferentially the stellate cells) during cholinergic modulation could produce prolonged delayed responses of regular, low-frequency firing, whereas others (preferentially pyramidal-like cells) generated more pronounced self-terminating bursts of activity (Klink and Alonso 1997). The Ca2+-dependent modulatory properties of INCM are consistent with an important role of this current in the generation of these self-terminating responses. On one hand, the ability of elevated [Ca2+]i to produce current potentiation can lead to the generation of activity-dependent, prolonged afterdepolarizations. On the other hand, when [Ca2+]i rises further secondarily to intense spike discharge, then a slower, transient process of INCM inhibition takes place and is followed by refractoriness. The latter properties would necessarily confer hysteresis on the [Ca2+]i dependence of inward-current amplitude that, in turn, may lead to oscillation.
Metabotropic activation of nonspecific cation conductances by neurotransmitters and peptides has been shown in many types of mammalian brain neurons (Chakfe and Bourque 2000; Congar et al. 1997; Li et al. 1999; Shen and North 1992). In many instances, neuromodulatory actions lead to the development of emergent properties such as bursting oscillations and bistable behavior (Andrade 1991). Ca2+-dependent up- and downregulation of a neurotransmitter-operated cation current may be a widespread mechanism to switch the firing modalities of different brain neuronal populations and the dynamics of the networks in which they are embedded (Marder and Calabrese 1996). In the case of the entorhinal network, cholinergically induced changes in neuronal dynamics by INCM may have profound implications in normal memory functioning (Fransen et al. 2002; Hasselmo et al. 2000; Lisman and Idiart 1995). In addition, given the apparent involvement of muscarinic receptors in the generation of epileptic seizures (Turski et al. 1989), INCM may also be a factor contributing to the hyperexcitability of the EC and related temporal lobe structures.
Finally, further cues on the possible role of plasticity in cholinergic-dependent firing patterns are provided by the observation that cholinergic stimuli differentially regulate the intrinsic activity of neurons in superficial and deep layers of the EC. During cholinergic stimulation, EC layer II principal neurons respond to a short stimulus with a delayed, self-terminating response, which can be suppressed by repetitive stimulus application. In contrast, we have recently shown that pyramidal neurons of EC layer V respond to an input with persistent activity that can be stepped up by repetitive application of the same input (Egorov et al. 2002). At the network level, EC layers II and V represent the input and output interface, respectively, of the EC-hippocampal system. Whereas sensory input from multiple cortical areas converges on layer II cells, which then send associative information to the hippocampus via the perforant path, layer V neurons receive the hippocampal output and project back to neocortex (Amaral and Witter 1995). Significantly, both EC layer II and V receive a dense cholinergic innervation (Alonso and Amaral 1995). Both the importance of the EC-hippocampal system in the "declarative" memory processes (Eichenbaum 2000; Scoville and Milner 1957; Squire 1998) and the particular role of acetylcholine in multiple aspects of plasticity (Hasselmo 1999a; Whitehouse et al. 1982) are well established. We suggest that the behavior of EC layer II neurons is more appropriate for the formation of short-term associations of on-going sensory information and that the activity-dependent up- and downregulation of INCM may represent a cellular mechanism partly underlying the enhanced and suppressed responses that are frequently observed in parahippocampal neurons during delayed matching tasks (Brown and Aggleton 2001; Suzuki et al. 1997; Young et al. 1997). This firing behavior could be perhaps utilized for the detection of novelty and recency of sensory events (Brown and Aggleton 2001; Fransen et al. 2002) and contribute to recognition memory (Manns et al. 2003a). On the other hand, the graded persistent firing more characteristic of layer V neurons could constitute the basis of a neural integrator of hippocampal outputs capable of sequentially organizing the distinct elements of a memory trace or, in other words, forming the basis of a tracking device for episodic memories.
|
GRANTS |
|---|
) on which an afterdischarge superimposed. The cells in A and B were electrophysiologically identified as stellate and pyramidal-like, respectively (see the text for details). Calibration bars in A2 and B: same values as in A1. Recording solutions were E1 and I1.
