| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
NEUROSCIENCE |
1 Department of Neurosurgery, Yale University School of Medicine, New Haven, CT 06520, USA
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
, and were capable of fast-firing. Muscimol-mediated excitatory actions were found in a nominally Ca2+-free/high-Mg2+ bath solution using cell-attached recording. GABAA receptor antagonists inhibited half the recorded neurons and blocked burst firing. Gramicidin perforated-patch recording revealed a GABA reversal potential positive to both the resting membrane potential and spike threshold. Together, these data suggest GABA is excitatory to many NPY cells. NPY and SST consistently hyperpolarized and reduced spike frequency in these neurons. No hyperpolarization of NPY on membrane potential was detected in the presence of tetrodotoxin, AP5, CNQX and bicuculline, supporting an indirect effect. Under similar conditions, SST hyperpolarized the cells, suggesting a direct postsynaptic action. Depolarizing actions of GABA and GABA-dependent burst-firing may synchronize a rapid release of GABA, NPY, and SST, leading to pre- and postsynaptic inhibition of excitatory hippocampal circuits.
(Received 30 August 2006;
accepted after revision 27 December 2006;
first published online 4 January 2007)
Corresponding author A. N. van den Pol: Department of Neurosurgery, Yale University School of Medicine, 333 Cedar St, New Haven, CT 06520, USA. Email: anthony.vandenpol{at}yale.edu
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Neuropeptide Y (NPY) and somatostatin (SST) are modulating neuropeptides in the hippocampus (Colmers et al. 1987, 1991; Boehm & Betz, 1997; Schweitzer et al. 1998). Both peptides show potent inhibitory effects on epileptiform activity in rodent epileptic models (Klapstein & Colmers, 1993, 1997; Tallent & Siggins, 1999; Mazarati & Wasterlain, 2002; Tu et al. 2005). NPY reduces granule cell excitability in chronically epileptic human hippocampal tissue (Patrylo et al. 1999). NPY and SST have anticonvulsant effects in rats (Woldbye et al. 1996, 1997; Mazarati & Telegdy, 1992). Mice deficient in NPY show an increased susceptibility to seizures (Erickson et al. 1996), and intracerebroventricular NPY infusion plays an antiepileptic role in NPY-deficient mice (Baraban et al. 1997; Vezzani et al. 1999), suggesting that NPY may serve as an endogenous anticonvulsant. An enhanced seizure severity has also been demonstrated in SST knockout mice, further supporting an anticonvulsant role of SST (Buckmaster et al. 2002).
In the hilus, a subset of GABA interneurons also contain both NPY and SST (Chan-Palay, 1987; Köhler et al. 1987; Kosaka et al. 1988; Esclapez & Houser, 1995); all three neuroactive substances are generally inhibitory, suggesting these cells may play a strong role in hippocampal inhibition. The neurons in the hilar area are sparsely distributed, and are sensitive to ischaemia and hypoxia (Johansen et al. 1987; Hsu & Buzsaki, 1993), and selective loss has been suggested in animal models of epilepsy (Sloviter, 1987; Sperk et al. 1992; Mitchell et al. 1995, 1997; Schwarzer et al. 1995) and in human epileptic tissue (de Lanerolle et al. 1989; Robbins et al. 1991; Mathern et al. 1995). Loss of these cells may further exacerbate epilepsy (Buckmaster & Dudek, 1997; van Vliet et al. 2004). An important question relates to how these inhibitory neurons respond to GABA; previous work on unidentified hilar cells in guinea pigs suggested that in some cells, GABA might be excitatory, and in others, inhibitory (Michelson & Wong, 1991; Freund and Buzsaki, 1996).
As it has not previously been possible to detect the subset of hilar neurons that colocalize GABA, NPY, and SST, little specific physiological information is available on the GABA responses of these hilar interneurons, despite the high frequency of GABAergic synapses on these cells (Freund & Buzsaki, 1996). In the present paper, whole-cell, cell-attached, and gramicidin recordings were used in hippocampal slices of transgenic mice that express GFP in these NPY neurons. GABAA receptor activation excited, and GABAA receptor antagonists depressed activity, suggesting GABA was excitatory to many, but not all of the NPY cells, even in the adult. NPY and SST consistently inhibited the NPY cells.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We previously reported the genesis of transgenic mice (PrP57) in which the reporter gene coding for enhanced green fluorescent protein (GFP) was driven by a mouse prion promoter (van den Pol et al. 2002). In the present experiments, we used a new line of mice (PrP50) on a Swiss-Webster background with the same mouse prion promoter driving GFP expression, in which a subset of neurons in the hippocampal hilar area expressed GFP. Hippocampal slices were prepared from transgenic mice. To verify the phenotype of neurons that expressed GFP, mice were anaesthetized with an overdose of sodium pentobarbital (100 mg kg–1) and perfused transcardially with physiological saline, followed by 4% paraformaldehyde. Rabbit antisera against NPY (Peninsula Laboratory Inc) and SST (Immunostar) were used. The NPY antiserum was used at a concentration of 1: 2500, and the SST antiserum was used at a concentration of 1: 3000. Secondary antisera conjugated to Texas Red (Molecular Probes) were used at concentrations of 1: 250–1: 500, as described in detail elsewhere (van den Pol et al. 2004). For purposes of corroboration, we also used a different transgenic mouse (kindly provided by Drs T. Horvath and J. Friedman) in which a different reporter gene, sapphire-bound tau, was driven by an NPY promoter from a BAC sequence (Roseberry et al. 2004).
Slice preparation
Experiments were done in transgenic mouse hippocampal brain slices. Slices were generated from 14-day- to 6-week-old-mice, which were maintained in a 12 h/12 h light/dark cycle, with food and water available continuously. The mice were given an overdose of sodium pentobarbital (100 mg kg–1) and the brains were rapidly removed and placed in ice-cold oxygenated (95% O2, 5% CO2) high-sucrose solution which contained (mM): 220 sucrose, 2.5 KCl, 6 MgCl2, 1 CaCl2, 1.23 NaH2PO4, 26 NaHCO3, 10 glucose, pH 7.4 with NaOH. A hippocampal block was prepared, and coronal slices (200 µm thick) were cut on a vibratome. After a 1–2 h recovery period, slices were moved to a recording chamber mounted on an Olympus (Tokyo, Japan) BX51WI upright microscope equipped with video-enhanced infrared-differential interference contrast (DIC) and fluorescence. Slices were perfused with a continuous flow of gassed (95% O2, 5% CO2) artificial cerebrospinal fluid (ACSF) that contained (mM): 124 NaCl, 3 KCl, 2 MgCl2, 2 CaCl2, 1.23 NaH2PO4, 26 NaHCO3, and 10 glucose. Some experiments were conducted in a Hepes buffer-containing solution (mM): 150 NaCl, 3 KCl, 2 MgCl2, 2 CaCl2, 1.23 NaH2PO4, 10 Hepes and 10 glucose, pH adjusted to 7.4 with NaOH. Only one neuron was used per slice, and then the slice was replaced with a fresh slice from an oxygenated holding chamber. Neurons were visualized with blue excitation light and an Olympus 40x water-immersion lens. Use of mice for these experiments was approved by the Yale University Committee on Animal Use.
Patch-clamp recording
Whole-cell current- and voltage-clamp recordings were performed using pipettes with 4–6 M resistance after filling with pipette solution. The pipette was made of borosilicate glass (World Precision Instruments, FL, USA) using a PP-83 vertical puller (Narishige, Tokyo, Japan). For most recordings, the composition of the pipette solution was (mM): 130 KMeSO4 (or KCl for GABA-mediated synaptic activity), 1 MgCl2, 10 Hepes, 1.1 EGTA, 2 Mg-ATP, 0.5 Na2-GTP, and 10 Na2-phosphocreatine, pH 7.3 with KOH. Neurons in the hilus area were recorded under direct visual observation of GFP fluorescence and differential interference contrast. After a gigaohm seal was obtained, a gentle negative pressure was applied to break through to the whole-cell configuration. An EPC10 amplifier and Patchmaster software (HEKA Elektronik, Lambrecht/Pfalz, Germany) were used for data acquisition. Slow and fast capacitance and series resistance were automatically compensated using Patchmaster software. The measured liquid junction potential was +9.5 mV when a KMeSO4 pipette solution was used, which was compensated to the membrane potential values tested. Access resistance was monitored continuously, and only those cells with stable access resistance (change <10%) were used for analysis. The recordings were made at 32°C. After a stable baseline was established, drugs were applied by a large 350 µm diameter flow pipe; when drugs were not being applied, normal buffer ran continuously from the flow pipe to avoid any peristaltic action. Additional details are presented elsewhere (Fu et al. 2004).
Spike amplitude was measured from the difference between threshold and peak of a spontaneous action potential, and the spike duration was measured at the half-amplitude from neurons with spontaneous spikes. The amplitude of the afterhyperpolarization (AHP) after an action potential was determined by measuring the peak downward deflection of the AHP from the baseline. Action potential threshold was determined as the potential at the point that the very sharp slope of the depolarizing phase began. Spike trains were elicited by injections of depolarizing current of 3 s duration. The extent of spike frequency adaptation was expressed as the ratio of the frequency of the last two spikes of the train to the frequency of the first two spikes, and termed the adaptation ratio (Mott et al. 1997). Input resistance (RN) was determined from the slope of voltage–current plots prepared by plotting the steady-state voltage during the current step versus the amplitude of the step.
The membrane time constant (
m) was determined at resting membrane potential from the biexponential curve best fitting the rising phase of the response to a small hyperpolarizing current step (<10 mV from resting potential). The slower component of the biexponential fit was used for m value.
For the cell-attached recordings, the patch pipettes were filled with ACSF, and a loose seal was formed on the identified hilar neurons to measure spontaneous action currents in voltage clamp.
For perforated patch-clamp experiments, gramicidin was dissolved in dimethysulfoxide and then diluted to 50 µg ml–1 in pipette solution, as described elsewhere (Chen et al. 1996; Gao et al. 1998). Both KCl and KMeSO4-based pipette solutions were used for gramicidin recording. To avoid gramicidin spillover during positive pressure approach to the recorded cell, recording pipettes were backfilled with gramicidin, and normal intracellular buffer was used at the tip. In some cases, Alexa Fluor hydrazide 594 (20 µM, Molecular Probes) was included in the gramicidin pipette solution to monitor the integrity of the cell membrane. Achieving successful perforation took up to 40 min, and experiments were started after getting a stable access resistance. The series resistance of perforated patch was 60–80 M, whereas it was 15–30 M in whole-cell recording. GABA reversal potential experiments were repeated in each cell tested, and only stable values for EGABA during this repetition were used.
Pulsefit (HEKA Electronik), Axograph (Axon Instruments, Union City, CA, USA), and Igor Pro (WaveMetrics, Lake Oswego, OR, USA) software were used for analysis. Spontaneous postsynaptic currents were detected with an algorithm in Axograph (Clements & Bekkers, 1997), and only those events with amplitude >5 pA were counted. The frequency of action potentials was measured using Axograph as well. Data are expressed as mean ± S.E.M.; t tests and ANOVA with Bonferroni post hoc test were used for statistical comparison. P < 0.05 was considered statistically significant.
Dye loading and cell morphology
In some slices from transgenic mice, Alexa Fluor Hydrazide 594 (20 µM, Molecular Probes) was used to fill the NPY cells through the patch pipette. Following break-in, the dye was loaded for 1–2 h, with a stable hyperpolarizing current (–0.1 nA) passing into the electrode. Then the slice was immersion-fixed overnight in a phosphate-buffered (pH 7.4) 4% paraformaldehyde at 4°C, and the morphology of the dye-filled cells was studied with an Olympus Fluoview 300 confocal scanning laser microscope (Olympus BX 51).
In five PrP50 mice, the retrograde transported dye cholera toxin subunit B-Alexa 594 (Molecular Probes; 0.5% in sterile saline) was pressure injected through a glass pipette into the septum, to determine if hilar NPY cells that expressed GFP sent axons here.
Chemicals and reagents
Bicuculline methiodiode (BIC), muscimol, DL-2-amino-5-phosphonovaleric acid (AP-5), 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), picrotoxin and gramicidin were purchased from Sigma (St Louis, MO, USA), tetrodotoxin (TTX) was obtained from Tocris Cookson (Ballwin, MO, USA). NPY (human/rat) and SST (SST14, human/mouse) were purchased from Phoenix Pharmaceuticals Inc (Belmont, CA, USA). All chemicals except gramicidin were dissolved in water to make a stock solution, and diluted with ACSF to final concentrations before using; the gramicidin solution was made just before use. NPY and SST were stored dry and rehydrated with ACSF before use.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In the present study, we used a transgenic mouse, PrP50, that expressed GFP under control of a mouse prion promoter to study the electrophysiological properties and responses of hilar NPY neurons. GFP was also expressed in small neurons adjacent to the granule cell layer in the hippocampus (Fig. 1A). These small cells were not included in the present study. In addition, outside the scope of the present study, a number of cortical neurons, mostly interneurons, expressed GFP. We previously reported the distribution of GFP-expressing neurons in a different transgenic mouse, PrP57, made with the same promoter construct (van den Pol et al. 2002). The cells that expressed GFP were different in the two lines of transgenic mice, suggesting a contribution from both the promoter sequence and the site of chromosomal insertion.
|
Immunocytochemistry was used to determine the transmitters contained in the GFP-expressing cells in the hilus. Antisera against NPY and SST, labelled with Texas Red, stained the GFP-positive neurons (Fig. 1B–E). In counts of 100 GFP-expressing cells in the hilus, 90% also expressed NPY and 88% expressed SST. That a small number of neurons did not express peptide antigen could be due to low levels of the peptides that may have been below the detection limits of the immunostaining, or alternatively, could arise from related hilar cells that did not contain these neuropeptides. To test whether the GFP labelled NPY neurons outside the hilus, we examined the cerebral cortex of the same mice. In the cortex, GFP and NPY did not colocalize, suggesting that in this mouse the prion promoter selectivity to NPY cells was restricted to the hilus, and did not generalize to other NPY cells in the CNS. All electrophysiological studies here focused on the GFP cells of the hilus that coexpressed NPY and SST; for simplicity, these cells will be called NPY cells here. Previous immunocytochemical studies have shown these cells also contain the fast-acting amino acid transmitter, GABA (Freund & Buzsaki, 1996). Some non-GFP CA3 pyramidal cells from the same mouse were used for control purposes. As a further control, additional experiments used slices from a different transgenic mouse that expressed sapphire-tau under control of the NPY promoter as described in detail elsewhere (Roseberry et al. 2004). In both lines of transgenic mice, the NPY cells had the same size, shape, and location. Somatostatin cells from other transgenic mice have been studied in other regions of the hippocampus (Oliva et al. 2000), but not in the hilus.
To examine the dendritic arbors of hilar NPY cells, Alexa fluor hydrazide 594 (20 µM) was used in the patch pipette to label the cells. Scanning confocal microscopy was used to study the dendritic trees in the thick hippocampal slices after dye filling (Fig. 2). We found several different types of dendritic arbor. NPY cells had 3–4 primary dendrites, branching one (Fig. 2Ba and b) to three (Fig. 2Aa and b) times. Some hilar NPY cells maintained their dendritic trees within the hilus (Fig. 2B and C). Others had the cell body in the hilus, but the majority of the dendrites travelled into the granule cell layer of the dentate gyrus, and extended into the molecular layer, parallel to the granule cell dendritic trees (Fig. 2A). These data suggest that, morphologically, hilar NPY cells may not be a strictly homogeneous group.
|
Electrophysiological characterization of the hilar NPY neurons
Whole-cell patch-clamp recordings were made to determine the neurophysiological properties of hilar GFP-expressing NPY neurons in acutely prepared in vitro brain slices. In 56 hilar NPY neurons recorded with current–clamp, 27 cells showed spontaneous firing at rest (–61.3 ± 1.4 mV), with a mean firing frequency of 0.6 ± 0.1 Hz (0.02–2.9 Hz); most cells fired irregularly (Fig. 3Aa). 29 cells did not discharge spontaneously at rest (–68.8 ± 1.1 mV, Fig. 3Ac). The average membrane potential of these 56 cells was –65.2 ± 1.0 mV (–53.1 to –80.7 mV), measured in the first 3 min after achieving whole-cell access, and during the interspike interval. The average input resistance was 335 ± 26 M (range 112–727 M, n = 36) (Fig. 3Ba and b). The input resistance of spontaneous firing cells (n = 17) was 354 ± 46 M and was lower for silent cells (n = 19) (318 ± 26 M) (P > 0.05; t test). The time constant (m) was 13.5 ± 0.7(n = 35). The amplitude and duration of spikes was 60.3 ± 1.7 mV (n = 23) and 1.2 ± 0.1 ms (n = 23), respectively. All GFP-expressing NPY neurons displayed large monophasic afterhyperpolarizations (AHPs) after action potentials (Fig. 3Ab). The mean amplitude of the AHP was 15.6 ± 1.0 mV (n = 24), and seven cells with similar resting membrane potentials between –58 mV and –62 mV showed an AHP of 15.3 ± 1.3 mV; AHP was measured in spontaneously spiking cells. The action potential threshold was –50.5 ± 1.0 mV (n = 23).
|
(n = 13), which was not significantly different than for NPY hilar neurons (P > 0.05; t test; n = 49). m was 33.2 ± 5.4 (n = 11). The spike amplitude and duration was 74.9 ± 2.8 mV (n = 5) and 1.4 ± 0.2 ms (n = 5), respectively. The AHP in pyramidal cells was triphasic, with an initial hyperpolarization (fast AHP) followed by a depolarization (ADH) and then a later hyperpolarization (slow AHP), as described (Gulyas et al. 1998) (Fig. 3Cc). The amplitude of the slow AHP was 3.5 ± 0.3 mV (n = 5), much smaller than that of hilar NPY neurons (P < 0.01; t test; n = 29). The action potential threshold of CA3 pyramidal cells was –50.0 ± 2.7 mV (n = 5). Spike frequency adaptation (SFA) and patterns of repetitive firing were studied during depolarizing current injections. After injection of a 3 s positive current, 20 of 21 recorded NPY cells showed repetitive firing continuously or discontinuously, consistent with previous reports on unidentified hilar interneurons (Mott et al. 1997). Figure 3Ea and F show typical responses of hilar NPY neurons to the current injection. When the time interval between each successive spike in a train is plotted against the interval number, it indicates a progressive increase in the interval between action potentials (Fig. 3Eb). SFA was measured in each cell using the adaptation ratio, which was expressed as the ratio of the instantaneous frequency of the last two spikes of the train to the instantaneous frequency of the first two spikes (Mott et al. 1997). The measurement was done using current levels of + 200 pA. Fourteen neurons in our study displayed an adaptation ratio varying from 0.17 to 0.95; the mean adaptation ratio was 0.56 ± 0.06, which corresponds approximately to the central peak of the adaptation ratio histogram in Fig. 2H. Most GFP-hilar neurons could be driven to high-frequency bursts with long-lasting square pulses; spike failure was found only in 1 of the 20 cells at 280 pA (Fig. 3G).
The average number of action potentials discharged by NPY cells in a 3 s depolarizing pulse of 280 pA was 99.9 ± 12.2 (n = 16). In contrast, the mean number of action potentials discharged by pyramidal cells in a 3 s depolarizing pulse of 280 pA was only 17.9 ± 3.8 (n = 11) significantly lower than that of NPY cells (P < 0.01; t test; n = 27). Compared to hilar interneurons, pyramidal neurons showed much stronger SFA (Fig. 3Ia and b) (P < 0.01; t test; n = 30), with a mean adaptation ratio of 0.20 ± 0.03 (range 0.05–0.33, n = 10).
Fast amino acid transmitters
Responses to inotropic glutamate agonists. Under current-clamp whole-cell recording, large-diameter flow pipe application of the glutamate agonists AMPA (25 µM) or N-methyl-aspartate (NMDA, 50 µM) consistently depolarized the tested NPY neurons (Fig. 4A and B). At resting membrane potential, AMPA induced a mean depolarization of 17.8 ± 2.3 mV (n = 4), and NMDA induced a mean depolarization of 5.7 ± 1.1 mV (n = 6).
|
GABA excites NPY neurons
Whole-cell recording.
The actions of GABA and GABAA receptor agonists on hilar NPY neurons were unusual and complicated. With KMeSO4 in the pipette, the GABAA receptor agonist muscimol (25 µM) was applied focally (
1 s) under current-clamp recording; most neurons were hyperpolarized (Fig. 4Ea). Ten of 16 neurons showed a simple monophasic hyperpolarization of –12.1 ± 1.5 mV (Fig. 4Ea), 5 of 16 neurons showed a biphasic effect, that is, a hyperpolarization of –10.4 ± 1.8 mV, followed by a later depolarization of 7.8 ± 0.9 mV (Fig. 4Eb). One cell was depolarized by muscimol with no preceding hyperpolarization (data not shown). In the presence of TTX (0.5 µM), muscimol (25 µM) hyperpolarized three of four cells, and another showed a biphasic effect. Muscimol significantly decreased the input resistance in all four cells tested (65 ± 11% of control; P < 0.05, ANOVA). A decrease in input resistance was observed in both the hyperpolarizing and depolarizing phase in the cells showing a biphasic effect. In contrast to the recordings with KMeSO4 pipette solution, when pipettes contained higher Cl– (30 mM KCl), the responses to muscimol (25 µM) and GABA (100 µM) were consistently depolarizing (n = 3) (Fig. 4Fa and b), as expected.
Cell-attached recording.
As the direction of the muscimol response is dependent on intracellular Cl–, which may be lowered by whole-cell recording with KMeSO4 pipettes as used above, we used cell-attached recording to test the actions of muscimol under a more normal intracellular milieu. Using pipettes filled with ACSF, a loose seal (100–200 M) was formed on identified hilar NPY neurons to measure the spontaneous activity in cell-attached mode. To reduce indirect synaptic actions, we also used nominally Ca2+-free/high-Mg2+ (10 mM) ACSF. Under this condition, muscimol (25 µM) robustly caused a change in the spike frequency. The mean shift in spike frequency for all 11 cells was not statistically significant (increase from 15.5 ± 15.0 to 24.6 ± 33.4 spikes (30 s)–1, but the statistical variance for the change in spike frequency in the presence of muscimol increased from 1.7 spikes to 536.4 spikes, a very large and significant increase in the data dispersal, indicating the possibility of two opposing effects. When we graphed the change in spike frequency from the 30 s before muscimol to the 30 s in muscimol (Fig. 4Ga), we found that neurons were divided into two groups, one group with a strongly increased frequency of action currents as seen in 5 of 11 NPY cells (P16 through adult) (Fig. 4Gb), and another group of NPY cells showing a substantially decreased spike frequency in 6 of the 11 (Fig. 4Ga). These results demonstrate that GABAA receptor activation exerts two different actions – GABA directly excites some NPY hilar neurons when intracellular Cl– is not disturbed, but inhibits others.
Pyramidal cell controls. To determine if the excitatory effect of muscimol would occur in cells other than hilar NPY neurons, we tried it on control CA3 pyramidal neurons in brain slices from 14- to 16-day-old transgenic mice. Muscimol consistently evoked a monophasic hyperpolarization of pyramidal cells recording in current-clamp with KMeSO4 in the pipette (n = 5) (Fig. 4H). In additional experiments, we used cell-attached recording to avoid disturbing the intracellular ion concentrations. To study the effects of muscimol on spike frequency, 5 mM K+ ACSF was used to increase the probability of spontaneous firing. Muscimol consistently inhibited spike frequency in all eight pyramidal cells tested (Fig. 4I). These data indicate that the excitatory effect of GABA found above in hilar NPY neurons was not a general property of all hippocampal neurons with our recording conditions in these brain slices, as no sign of GABA excitation was found in any of the 13 pyramidal cells tested. The data showing inhibitory actions of GABA on mouse pyramidal cells are consistent with previous reports on these cells in the mouse and other species at this stage of development (Swann et al. 1989; Khazipov et al. 2004; Wong et al. 2005). Depolarizing actions of GABA have been reported in neonatal rat pyramidal cells, but only in cells younger than postnatal day 12 (Ben-Ari et al. 1989); all recordings in the present study were made from older animals.
GABAA receptor antagonist reduces spike frequency of NPY cells. GABAergic hilar interneurons synapse with other interneurons and form local circuits (Hajos et al. 1996; Forti & Michelson, 1998). Here we tested if the activity of these hilar NPY neurons would be altered by blocking the GABAA receptor. BIC (30 µM) was applied to 16 cells, 6 of these showed no spikes, and 10 showed spontaneous firing (third week). Interestingly, BIC evoked an inhibitory effect in 12 of 16 cells recorded with KMeSO4 pipettes (Fig. 5Ab and F), hyperpolarizing the membrane potential by 3.2 ± 0.5 mV (P < 0.01; ANOVA), which recovered after washout (0.6 ± 0.6 mV). As noted above for muscimol, the statistical variance for the change in spike frequency in the 1 min prior to BIC was 0.25 Hz, which increased dramatically and significantly to 47.2 Hz in BIC, a variance increase of almost 200-fold. Similar to muscimol, BIC evoked two different effects in cells. The spike frequency was decreased to 13 ± 9% of control, which recovered to 79 ± 10% after washout in seven cells (P < 0.01; n = 7; ANOVA). BIC showed an excitatory effect in 4 of the 16 cells (Fig. 5Aa and F), after 1 min application of BIC, spikes were evoked in one silent cell, and spike frequency was increased by 317% in three spontaneously firing cells; the membrane potential was depolarized by 2.9 ± 1.0 mV in these four excited cells (P > 0.05; ANOVA). Thus in a number of cells studied here, synaptic GABA release appeared to have excitatory actions.
|
In order to maintain the physiological intracellular environment undisturbed, cell-attached recording rather than whole-cell recording was used in both 2–3-week-old and adult animals.
Cells without spontaneous firing were not used in these experiments. In eight cells from 3rd-week mice, five cells were inhibited by BIC (30 µM) (Fig. 5Cb and F), the firing rate was reduced by 54 ± 13% with a full recovery (P < 0.05; ANOVA). Three other cells were excited (Fig. 5Ca and F); the firing rate was increased by 48 ± 18% with a full recovery (P > 0.05; ANOVA). In five cells from adult mice, BIC (30 µM) inhibited two cells (spike frequency was decreased to 8 ± 8% of control) and excited three cells (spike frequency was increased to 680 ± 487% of control) (Fig. 5D and F). Together, these results suggest that the inhibitory response to blocking GABA receptors exists in about half the hilar NPY neurons recorded when the intracellular ionic components were maintained at normal levels by using cell-attached recording.
Pyramidal control neurons are excited by bicuculline. With whole-cell recording, BIC (30 µM) evoked an excitatory effect on all 3rd week pyramidal neurons recorded (spike frequency was increased to 304 ± 54% of control, and membrane potential was depolarized by 1.2 ± 0.6 mV (P < 0.05; n = 6; ANOVA) (Fig. 5E and F). This is consistent with the normally inhibitory actions of GABA on pyramidal neurons.
Picrotoxin and bicuculline show similar actions. It has been reported that BIC might block small-conductance calcium-activated potassium (SK) channels (Debarbieux et al. 1998; Khawaled et al. 1999) in addition to blocking GABAA receptors. To demonstrate that the inhibition by BIC on NPY interneurons was related to the GABAA receptor, we tested another GABAA receptor antagonist – picrotoxin (PIC) (Constanti, 1978; Barker et al. 1983; Yoon et al. 1993). PIC (80 µM) showed two types of effects on hilar NPY neurons studied with KMeSO4 whole-cell recording, as well as with cell-attached recording. With whole-cell recording, three of four cells were inhibited (spike frequency was decreased to 42 ± 21% of control) by picrotoxin, whereas one was excited (spike frequency was increased to 125% of control). With cell-attached recording, two of four cells were inhibited (spike frequency was decreased to 38 ± 21% of control), and two were excited (spike frequency was increased to 154 ± 19% of control; Fig. 5G). These results further demonstrate that GABAA receptor-mediated excitatory activity existed in hilar NPY neurons, and that some of the hilar NPY neurons were tonically excited by GABAergic synaptic transmission. The effect of GABAA agonist and antagonists on these hilar NPY neurons was further studied using a nominally Ca2+-free/high-Mg2+ (10 mM) ACSF to decrease synaptic transmission. With whole-cell recording in the presence of TTX 0.5 µM to block spontaneous firing, neither BIC nor PIC exerted any effect on the membrane potential (three cells tested) (Fig. 5Ha and b) or input resistance (103 ± 2% of control; P > 0.05; n = 7; ANOVA). These data indicate that BIC and PIC do not act directly on the hilar NPY neurons in the absence of synaptic GABA activity.
Gramicidin perforated-patch recording. Gramicidin generates small holes in the plasma membrane that are impermeable to Cl– (Akaike, 1996; Chen et al. 1996; Wang et al. 2001). This allows a relatively unbiased approach to determining whether GABA's depolarizing actions might be based on a mechanism where the GABA reversal potential is positive to the resting membrane potential. To test the hypothesis that the GABA reversal potential would cause GABA to be excitatory, gramicidin-based recordings were made in slices from mice aged between 15 and 21 days. In five of seven NPY neurons, the reversal potential of the muscimol-induced current was positive to the resting membrane potential, as determined by the inward or outward current that was evoked by flow pipe application of muscimol (25 µM) (Fig. 6A). In these five neurons, the mean reversal potential was –35.0 ± 4.2 mV (Fig. 6B). A GABA reversal potential that was only slightly positive to the resting membrane potential could still be inhibitory due to shunting of excitatory synaptic activity, but here the reversal potential is even positive to the spike threshold, suggesting GABA would be excitatory. In two of seven cells, the reversal potential was near the resting membrane potential, with a mean of –75.1 mV. Based on these data, GABA evoked an inward current, and a substantial depolarization at rest in the majority, but not all, of the neurons tested. In three cells in which the GABA reversal potential was positive to the resting membrane potential, muscimol and BIC were also used under current-clamp mode. Muscimol (25 µM) evoked a depolarization, whereas blocking synaptic GABA actions with BIC (30 µM) evoked a hyperpolarization (Fig. 6C and D). Both results are consistent with a depolarization mediated by the GABAA receptor.
|
BIC inhibits the burst firing evoked by 4-AP. To investigate the glutamate-independent activity of GABAA receptor in hilar NPY neurons activity, the convulsant compound 4-aminopyridine (4-AP) (50 µM) was used to excite these cells in the presence of glutamate receptor blockers AP-5 (50 µM) and CNQX (10 µM). Consistent with previous studies on unidentified hilar interneurons (Michelson & Wong., 1991), at resting membrane potential, NPY neurons were excited by 4-AP with robust bursting firing as studied with whole-cell recording; after washout of the 4-AP, spike frequency returned to normal levels. Bursts of spikes were interrupted by a silent period of membrane potential hyperpolarization (10–15 mV) lasting 2–5 s, at intervals of 6–20 s (Fig. 7A). In contrast, in the presence of BIC (30 µM) together with AP-5 and CNQX, 4-AP failed to evoke burst firing in these neurons (n = 5) (Fig. 7Ba), although a rhythmical hyperpolarization was still found in some cells (Fig. 7Bb). In some cells, 4-AP evoked an increase in random firing in the presence of GABA and glutamate receptor antagonists, but bursting was not found. These results are similar to those of previous studies (Michelson & Wong, 1991; Forti & Michelson, 1998), and suggest that GABAA receptor activity in the hilar interneuron network is necessary for the maintenance of burst firing in these NPY neurons.
|
Numerous NPY- and SST-immunoreactive axons are found in the dentate hilus (Deller & Leranth, 1990; Leranth et al. 1990), which raises the possibility that NPY and SST may exert an effect on hilar NPY neurons. To test this hypothesis, we evaluated the actions of NPY and SST on spontaneous firing and action potentials in identified GFP-hilar neurons in hippocampal slices under current-clamp. Flow pipe application of NPY (1 µM) for 1 min consistently hyperpolarized the membrane potential and depressed the spontaneous firing of GFP-hilar neurons in a reversible manner (Fig. 8A and B). In eight spontaneously firing cells tested, NPY depressed the spike frequency by 79.4%, which subsequently recovered (Fig. 8B). This effect was highly significant (P < 0.01; n = 8; ANOVA;). The membrane potential was reversibly hyperpolarized (P < 0.01; n = 13; ANOVA; Fig. 8D).
|
In the presence of 0.5 µM TTX, 50 µM AP-5, 10 µM CNQX and 30 µM BIC, SST (1 µM) hyperpolarized the membrane potential (P < 0.01; n = 6; ANOVA) (Fig. 8E and F). Together, the blockers caused no statistically significant change in mean membrane potential. The extent of SST-mediated hyperpolarization was smaller than that without TTX and glutamate and GABA antagonists (P < 0.05, n = 19, t test; Fig. 8F), suggesting that the inhibition of SST on GFP neurons was mediated partially by an indirect, and partially by a direct, mechanism. SST (1 µM) showed a tendency to decrease the input resistance of these neurons (by 5 ± 2%; P = 0.07; t test; n = 5), consistent with a direct effect. Another possibility underlying the inhibitory effects of SST might be related to the reduction of voltage-activated Ca2+ currents (Wang et al. 1990). Under this condition of blocked synaptic actions, no hyperpolarization was detected with application of NPY (1 µM) (Fig. 8E and F) (mean effect: –0.26 ± 0.25 mV; P > 0.05; n = 7; ANOVA), nor was a change in input resistance found (101 ± 3% of control; n = 4; P > 0.05; t test) suggesting an indirect effect of NPY on NPY neurons. No direct effect of NPY was detected. A previous report in unidentified hilar neurons suggested about half of the hilar cells tested were directly inhibited by NPY via a mechanism dependent on activation of a G protein-coupled inwardly rectifying potassium current (Paredes et al. 2003).
In order to further clarify if inhibition of glutamatergic neurotransmitter actions mediated the indirect inhibition of NPY on hilar NPY neurons, AP-5 and CNQX were applied in the bath solution, and the effect of NPY in the presence of these glutamate receptor antagonists was investigated. In the presence of these glutamate receptor blockers, NPY evoked no significant hyperpolarization (1.5 ± 1.1 mV; P > 0.05; n = 5; ANOVA), suggesting that the inhibition of NPY was induced mainly by reducing glutamate release. This is similar to the previous reported inhibition by NPY of excitatory synaptic transmission on hippocampal excitatory neurons (Colmers et al. 1987).
Sapphire-tau NPY transgenic mice
To test whether the transmitter actions we observed might have been due to some oddity in the PrP50 transgenic mouse we used, we also used slices from an unrelated transgenic mouse on a different genetic background, in which a large NPY BAC sequence drove GFP expression (Roseberry et al. 2004). We did whole-cell recording with KMeSO4-containing pipettes on sapphire-expressing hilar interneurons in brain slices of 4- to 6-week-old mice. In slices of these NPY-GFP transgenic mice, the NPY interneurons showed a similar firing pattern to that in brain slices from Prp50 mice. In 11 cells recorded, 5 fired spontaneously with an average frequency of 0.5 ± 0.2 Hz, six cells were silent (data not shown). The average membrane potential was –64.5 ± 2.2 mV (–50.2 mV to –75.4 mV, n = 11). The mean input resistance was 314 ± 40 M (162–536 M, n = 10), similar to that of Prp50 mice (P > 0.05; n = 46; t test).
As in Prp 50 mice, both NPY (1 µM) and SST (1 µM) showed a reversible inhibition of NPY-neurons in the hilus (Fig. 9A) of the tau-sapphire mice. NPY and SST decreased the spike frequency by 60.1 ± 23.1% (P < 0.01; n = 4; ANOVA) and 64.0 ± 18.4% (P < 0.05; n = 5; ANOVA), respectively (Fig. 9C). Both peptides also hyperpolarized the membrane potential by 3.7 ± 1.0 mV (P < 0.05; n = 7; ANOVA) and 3.6 ± 0.7 mV (P < 0.01; n = 7; ANOVA), respectively (Fig. 9D). Figure 9A shows a typical response of NPY-sapphire neurons to NPY and SST.
|
Gramicidin perforated patch was used to record the GABA reversal potential in hilar NPY cells from sapphire-tau NPY transgenic mice (17 days old). In these recordings, KMeSO4 pipette solution was used to exclude the possible artifact that might occur from leakage into the recorded cell from a pipette solution containing high Cl–. Similar to results obtained in PrP50 mice, the mean reversal potential of the muscimol-induced current was –42.1 ± 5.4 mV (n = 5, range –59.5 to –26.0 mV), substantially positive to the resting membrane potential (–72.4 ± 2.4 mV). These data provide further corroborating evidence for potential GABA-mediated depolarization in these hilar NPY neurons, and suggest these actions of GABA are not caused by some genetic anomaly found only in the PrP50 mouse.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Electrophysiological characterization of hilar interneurons
Interneurons in the hilus are both morphologically and electrophysiologically diverse, and therefore particular cell types have been difficult to identify in live slices (Amaral, 1978; Buckmaster & Schwartzkroin, 1995; Mott et al. 1997; Lübke et al. 1998). The average input resistance of hilar NPY neurons was 335 M, higher than the 198–252 M in rat dentate-hilus border interneurons (Mott et al. 1997). Compared with the slower pyramidal cells, NPY cells were capable of very rapid firing with modest spike frequency adaptation, and a deep 15 mV monophasic AHP, previously described in some unidentified hilar neurons (Mott et al. 1997).
GABA responses
When we combined the data from gramicidin and cell-attached recording experiments, both leaving intracellular Cl– unperturbed, half the NPY cells were excited by GABA agonists. Consistent with this, blocking GABAA receptors with BIC or PIC inhibited half of the NPY neurons recorded, whereas blocking GABA synaptic activity increased activity in all pyramidal cells. Together, 55% of 42 NPY neurons were excited by exogenous or synaptic GABA. Developing neurons show depolarizing responses to GABA, primarily due to a high intracellular Cl– (Ben-Ari et al. 1994; Gao et al. 1998). In pyramidal cells the depolarizing effects reversed to hyperpolarizing before the end of the second week (Leinekugel et al. 1997), and all recordings in the present study were made at later developmental stages. With cell-attached recordings, GABAA agonists increased spike frequency even under conditions where synaptic effects were reduced with low Ca2+ and high Mg2+. Similarly, gramicidin perforated-patch recording showed an elevated GABA reversal potential, in some cases 25–30 mV positive to the resting membrane potential and also positive to the spike threshold, consistent with high levels of intracellular chloride. The excitatory actions of GABA were selective for NPY neurons, as all control CA3 pyramidal neurons in the same hippocampal slices were consistently inhibited.
The excitatory action of GABA is unlikely to be due to GFP expression, as we previously recorded from hundreds of GFP-expressing hypocretin, melanin Concentrating hormone (MCH), noradrenaline, or GABA neurons, and have not found GABA-mediated excitation in neurons of the same age (Li et al. 2002; van den Pol et al. 2002, 2004; Fu et al. 2004; Acuna-Goycolea et al. 2005). Furthermore, two different transgenic mice on different genetic backgrounds used here showed similar GABA-mediated excitation. The reporter gene in one was GFP, and in the other it was sapphire-tau. It is unlikely that two different constructs would insert into the same site in the same host chromosome. Thus it is unlikely that the responses were due to disruption of some particular gene, or to a specific biochemical response to two different reporter gene products in the two unrelated transgenic mice used.
Excitatory actions mediated through GABAA receptors have been reported in mature hippocampal and cortical neurons (Michelson & Wong, 1991, 1994; Gulledge & Stuart, 2003). Alterations in ion gradients of GABAA receptor-permeable anions, such as the intracellular accumulation of Cl– ions and the redistribution of HCO3– ions (Staley et al. 1995; Perkins & Wong, 1996; Staley & Proctor, 1999), as well as the extracellular accumulation of K+ ions (Wong & Watkins, 1982; Kaila et al. 1997) can be involved in the excitatory activity of GABA. As substitution of Hepes for bicarbonate did not block GABA excitation, and we maintained K+ in the bath at a constant level, the effect was probably not based on bicarbonate or K+. The very positive GABA reversal potential suggests the depolarization is probably based on a mechanism of high intracellular Cl– in these NPY cells. These findings, together with our data that GABA excitation was present even in adult NPY cells, argue that hilar NPY cells do not show the developmental switch from GABA excitation to inhibition found in pyramidal neurons prior to postnatal day 12; after day 12, pyramidal cells are hyperpolarized by GABA (Ben-Ari et al. 1989). It is possible that some of our data showing depolarizing actions of GABA between postnatal days 14 and 21 may be due to a slow rate of maturation and shift from GABA depolarization to GABA hyperpolarization, but as depolarizing actions were also found in more mature mice, and as depolarizing actions of GABA have been identified in other adult hilar interneurons (Michelson & Wong, 1991, 1994), it seems unlikely that all the excitatory actions of GABA on these NPY cells were due only to developmental immaturity.
Some cells showed biphasic responses to GABA agonists when recorded with KMeSO4 pipettes. The initial hyperpolarization may be a direct effect of GABA on the recorded cell which is perfused with KMeSO4, and the secondary depolarization may be due to GABA actions on the dendrites which have not yet been subject to infusion of the pipette buffer and may still maintain high Cl–. A differential response to GABA has been reported in soma and dendrites, with dendrites showing a GABA-mediated depolarization (Alger & Nicoll, 1982; Scharfman & Sarvey, 1987; Hara et al. 1992). The inhibition by BIC and PIC on these neurons may also in part be due to actions on other hilar NPY neurons, also excited by GABA, that are presynaptic to the recorded cell; GABAergic hilar cells have a high degree of synaptic connectivity (Forti & Michelson, 1998). The contribution of local GABA synaptic interaction is consistent with the block of NPY neuron bursting by BIC.
Although about half the NPY cells showed GABA-mediated excitation, other NPY neurons were inhibited by GABA agonists, indicating a heterogeneity of GABA actions. Inhibitory actions could be due to an decrease in intracellular Cl–, resulting in GABA-mediated hyperpolarization. In addition, a shunting inhibition can also occur during a modest depolarization concurrent with an increase in membrane conductance. Similar to our results on identified NPY neurons, some unidentified guinea pig hilar cells showed both inhibitory and excitatory responses to synaptic GABA release (Michelson & Wong, 1991, 1994), paralleling work on interneurons in CA1 that showed heterogenous response profiles to many neurotransmitters (Parra et al. 2000). Our data show that NPY cells may show differences in the structure of the dendritic tree, as well as in the range of the dendritic trees which could be primarily restricted to the hilus, or could penetrate the granule cell and molecular layer. In addition, we found that some hilar NPY neurons send an axonal projection to the septum, but the majority do not show retrograde labelling after septal dye injections; as the entire septum was not labelled in the dye injections, it is probable that the frequency of retrograde-labelled neurons underestimates the actual number of hilar NPY cells projecting to the septum. Together, these data suggest that there may be different classes of hilar NPY cells, similar to different classes of somtatostatin cells recently described in the cortex (Ma et al. 2006).
Hilar NPY/SST neurons receive long-distance GABAergic input from the medial septum. It has been suggested that the GABA-containing septohippocampal pathway could facilitate granule cell activity by inhibiting the hilar interneurons (Yamano & Luiten, 1989; Milner & Veznedaroglu, 1993). However, the data here suggest an alternative possibility, that GABAergic axons from the septum could potentially increase hilar NPY/GABA cell activity, secondarily leading to an enhanced inhibition of principal excitatory hippocampal neurons.
Neuropeptide Y and somatostatin inhibit NPY interneurons
Both NPY and SST consistently inhibited NPY neurons. NPY evoked no hyperpolarization in the presence of TTX or ionotropic glutamate receptor antagonists, suggesting an effect possibly due to depression of glutamate release, consistent with previous reports that NPY inhibits pyramidal cells by reducing excitatory synaptic transmission presynaptically (Colmers et al. 1987, 1991). Our results indicate that SST inhibited NPY neurons in the hilus not only by an indirect mechanism, but also by a direct hyperpolarization of the membrane potential; SST decreased input resistance, suggesting the possibility of activation of K+ channels. SST inhibits pyramidal neurons by depressing presynaptic glutamate release (Boehm & Betz, 1997) and augmenting postsynaptic K+ currents (Moore et al. 1988; Schweitzer et al. 1998).
GABAergic interneurons play an essential role in the control of hippocampal network activity. Hilar GABAergic interneurons are particularly important since they receive excitatory input from entorhinal afferents, and can gate the propagation of epileptiform activity into the hippocampus (Pare et al. 1992). The GABA interneurons that coexpress SST/NPY in the hilus are vulnerable to seizures, and loss of these neurons has been found in animal models of epilepsy (Sperk et al. 1992; Schwarzer et al. 1995; Mitchell et al. 1995, 1997) and human epileptic tissue (de Lanerolle et al. 1989; Robbins et al. 1991; Mathern et al. 1995). The loss of these cells, possibly related to a low threshold for activation by excitatory input (Scharfman, 1991), may reduce the functional inhibition of granule cells in the dentate gyrus and lower seizure threshold (de Lanerolle et al. 1989). Here, the consistent inhibition of NPY neurons by NPY and SST may confer some protective inhibitory effect against hyperexcitability.
GABA-mediated excitation of NPY cells would not only increase GABA release, but, importantly, may also enhance release of the coexpressed inhibitory peptides NPY and SST onto granule cells. This would enhance inhibition by both postsynaptic actions and presynaptic attenuation of excitatory synaptic input. The release of neuropeptides per spike is enhanced during spike bursts (Dreifuss et al. 1971; Dutton & Dyball, 1979); as GABA appears to be critical for NPY neuron bursting, local hilar excitatory actions of GABA may ultimately enhance inhibition of dentate granule cells by increasing release of the three inhibitory transmitters, GABA, SST, and NPY. The excitatory responses of some hilar NPY cells to GABA may provide an important rapid mechanism for synchronization of inhibitory output regulating dentate postsynaptic excitatory circuits.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Acuna-Goycolea C, Tamamaki N, Yanagawa Y, Obata K & van den Pol AN (2005). Mechanisms of neuropeptide Y, peptide YY, and pancreatic polypeptide inhibition of identified green fluorescent protein-expressing GABA neurons in the hypothalamic neuroendocrine arcuate nucleus. J Neurosci 25, 7406–7419.
Akaike N (1996). Gramicidin perforated patch recording and intracellular chloride activity in excitable cells. Prog Biophys Mol Biol 65, 251–264. Review.[CrossRef][Medline]
Alger B & Nicoll R (1982). Pharmacological evidence for two kinds of GABA receptor on rat hippocampal pyramidal cells studied in vitro. J Physiol 328, 125–141.
Amaral DG (1978). A Golgi study of cell types in the hilar region of the hippocampus in the rat. J Comp Neurol 182, 851–914.[CrossRef][Medline]
Baraban SC, Hollopeter G, Erickson JC, Schwartzkroin PA & Palmiter RD (1997). Knock-out mice reveal a critical antiepileptic role for neuropeptide Y. J Neurosci 17, 8927–8936.
