The Journal of Neurophysiology Vol. 79 No. 4 April 1998, pp. 2111-2118
Copyright ©1998 by the American Physiological Society

Electrophysiological and Pharmacological Characterization of the Direct Perforant Path Input to Hippocampal Area CA3

Julia Berzhanskaya, Nathaniel N. Urban, and German Barrionuevo

Department of Neuroscience and Center for the Neural Basis of Cognition, University of Pittsburgh, Pittsburgh, Pennsylvania 15260

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Berzhanskaya, Julia, Nathaniel N. Urban, and German Barrionuevo. Electrophysiological and pharmacological characterization of the direct perforant path input to hippocampal area CA3. J. Neurophysiol. 79: 2111-2118, 1998. Monosynaptic perforant path responses evoked by subicular stimulation were recorded from CA3 pyramidal cells of rat hippocampal slices. These monosynaptic responses were isolated by using low intensities of stimulation and by placing a cut through the mossy fibers. Perforant path-evoked responses consisted of both excitatory and inhibitory components. Excitatory postsynaptic currents (EPSCs) were mediated by both -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acidreceptors (AMPAR) and N-methyl-D-aspartate receptors (NMDAR).Inhibitory postsynaptic currents consisted of -aminobutyric acid-A (GABA_A-) and -B (GABA_B)-receptor-mediated components. At membrane potentials more positive than -60 mV and at physiological [Ca²⁺]/[Mg²⁺] ratios, >30% of perforant path evoked EPSC was mediated by NMDARs. This value varied as a function of the membrane voltage and external [Mg²⁺]. Two types of responses were observed after low-intensity stimulation of the perforant path. The first type of response showed paired-pulse facilitation and was reduced by 2-amino-4-phosphonobutyric acid (AP4). The second type of response showed paired-pulse depression and was reduced by baclofen. Electrophysiological and pharmacological characteristics of these two types of responses are similar to the properties of lateral and medial perforant path-evoked EPSPs in the dentate gyrus.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

In the classical "trisynaptic circuit" model of the hippocampus, pyramidal cells of the CA3 region of the hippocampus receive their primary input from the mossy fibers (MF), the axons of the dentate granule cells. However, anatomic data show that pyramidal cells of CA3 also receive a direct, monosynaptic input from the entorhinal cortex (EC), via the fibers of the perforant path (PP) (Steward 1976). Because an almost equal number of synapses are formed by PP fibers in CA3 and in the dentate gyrus (DG) (Amaral et al. 1990), the PP projection to CA3 could be as strong as the PP projection to the DG. Findings from electrophysiological experiments in vitro (Doller and Weight 1982) and in anesthetized animals in vivo (Yeckel and Berger 1990, 1995) also suggest that the monosynaptic input from the EC to areas CA1 and CA3 is sufficiently strong enough to excite pyramidal cells in these areas to the level of action potential generation. In addition, McNaughton et al. (1989) have demonstrated that place cell firing in areas CA1 and CA3 of behaving rats persists after destruction of the DG. These data suggest that the direct PP input to areas CA1 and CA3 plays a significant role in hippocampal function. Several computational models of the hippocampal formation already have incorporated the direct PP projection to areas CA3 (O'Reilly and McClelland 1994) and CA1 (Hasselmo and Schnell 1994) as a functional part of cortico-hippocampal connections.

Despite the recognition of the functional importance of the direct input from EC to the hippocampus proper, the pharmacological and electrophysiological properties of the monosynaptic PP input to area CA1 and, especially, to area CA3 have not yet been characterized fully. Previous studies have shown that PP-evoked excitatory postsynaptic currents (EPSPs) are accompanied by inhibitory postsynaptic currents (IPSPs) both in the DG (Buckmaster and Schwartzkroin 1995; Staley and Mody 1992) and area CA1 (Colbert and Levy 1992; Empson and Heinemann 1995a). These IPSPs are thought to be disynaptic (Empson and Heinemann 1995b); and anatomic data suggest that hippocampal interneurons receiving direct PP input mediate this disynaptic inhibition (Buhl et al. 1994; Kiss et al. 1996). However, PP input to the DG was found to be mainly excitatory, whereas in CA1, PP stimulation results in both excitation and strong inhibition (Empson and Heinemann 1995a). It has been demonstrated that inhibition can change dendritic excitability (Kim et al. 1994) and reduce EPSPs by linear or nonlinear summation (Staley and Mody 1992). The role of inhibition in the response of CA3 pyramidal cells to the PP stimulation is still unknown.

PP-evoked responses in the DG and area CA1 (Colbert and Levy 1992; Lambert and Jones 1989) were found to be mediated in part by N-methyl-D-aspartate (NMDA) receptors (NMDAR). The NMDAR-mediated component represents a significant part of the response (~30%) in the DG (Colino and Malenka 1993; Keller et al. 1991). Although the NMDAR-mediated component of the PP-evoked response in area CA1 has been studied (Colbert and Levy 1992; Empson and Heinemann 1995b), it has not been quantified. The presence of an NMDAR-mediated component of PP-evoked EPSPs in area CA3 has not been demonstrated directly. However, the ability to induce NMDAR-dependent long-term potentiation of the PP input to CA3 (Berger and Yeckel 1991) indicates that some portion of PP-evoked responses in this area must be NMDAR mediated at least under conditions of tetanic stimulation.

The heterogeneity of the PP input to the hippocampus adds another dimension to the characterization of this input. Specifically, the PP input to the hippocampus consists of medial and lateral PP fibers (MPP and LPP, respectively) (Amaral 1993; McNaughton 1980; Steward 1976). Thus stimulation of PP fibers may result in combined LPP/MPP responses. In the DG, LPP and MPP form synapses in adjacent portions of stratum moleculare. LPP synapses are located more distally on the dendritic tree, whereas MPP synapses are located more proximally (Steward 1976; Witter 1993). A similar distribution of LPP and MPP terminals exists in area CA3. The LPP- and the MPP-evoked responses in the DG can be differentiated by their kinetics, paired-pulse profile and different sensitivity to neuromodulators (Bramham et al. 1988, 1991; Colino and Malenka 1993; Koerner and Cotman 1991; Lanthorn and Cotman 1981; McNaughton 1980). Similar differences between LPP and MPP inputs to area CA3 have been observed in vivo (Breindl et al. 1994).

Most previous studies of the PP input to CA3 were done in in vivo preparations, in which contamination of the monosynaptic PP input to CA3 by polysynaptic responses is difficult to avoid. Hence, the first goal of this study was to identify conditions under which the isolated monosynaptic PP input to area CA3 could be studied using the advantages of the hippocampal slice preparation. The second goal was to determine which receptors mediate excitatory and inhibitory components of isolated PP-evoked response in area CA3. The third goal was to determine the proportion of the NMDAR-mediated component of PP-evoked EPSCs. Finally, we aimed to investigate differences between LPP- and MPP-evoked responses in area CA3.

	METHODS

Abstract Introduction Methods Results Discussion References

Slice preparation and recordings

Transverse hippocampal slices (450 µm thick) were obtained as described previously (Henze et al. 1995; Urban and Barrionuevo 1996). In all slices, a cut was made through the hilus of the DG up to the suprapyramidal blade of dentate to prevent contamination of the monosynaptic PP response by MF. Recordings were made from slices submerged in carbogenated artificial cerebrospinal fluid [concentrations were (in mM) 125.0 NaCl, 2.0 KCl, 1.2 NaH₂PO₄, 26.0 NaHCO₃, 10.0 dextrose, 3.0 MgCl₂, and 3.0 CaCl₂] at temperature 31-33°C and rate of perfusion 2 ml/min.

A bipolar stimulation electrode (62 µm diam nichrome wire) was positioned in the s. lacunosum moleculare on the border of the subiculum and area CA1 (Fig. 1C), where myelinated PP fibers are clearly visible under the dissecting microscope. In studies of excitatory and inhibitory components, the position of stimulation electrode was close to the hippocampal fissure (350-100 µm) and did not vary throughout the experiment. In experiments in which we tried to isolate LPP responses from MPP responses by changing stimulation site, the stimulation electrode was moved within a range of 100 µm perpendicular to the PP fiber tract. Field potentials for the laminar and the current source density analysis (CSD) study were recorded using glass microelectrodes (1-3 M) filled with 0.5 M NaCl solution at multiple positions within area CA3.

View larger version (36K): [in this window] [in a new window]   FIG. 1. Isolation of the perforant path projection to CA3. A: scheme of a hippocampal slice preparation showing stimulation (STIM) and recording (REC) sites and placement of cut. Recordings were made from CA3 pyramidal cells located in CA3b and CA3c subfields. Extracellular electrode in the CSD study was moved in the direction perpendicular to the cell bodies layer (green line). B: Current source density profile within CA3 indicates selective activation of perforant path synapses. Current source density analysis was calculated from field potentials recorded in CA3 after stimulation of the perforant path as described in METHODS. Contour plots show that the main current sink is located in stratum lacunosum moleculare with very little activation elsewhere on the dendritic tree of CA3 pyramidal cells.

When collecting data for CSD, the first recording site was always in the s. lacunosum moleculare of CA3. While recording at this location, the stimulation intensity was adjusted to be 60% of the value required to elicit a population spike. The depth (z axis) of the recording electrode was adjusted to maximize the peak amplitude of the recorded field potential. Ten consecutive responses were recorded at this location before the electrode was moved 50 µm toward the s. pyramidale (x axis). At this new location, the electrode was placed into the tissue at the same depth as in the previous position. Every 200 µm (i.e., every fourth position) the depth of the electrode was varied by ±50 µm in the z axis, and 10 responses were recorded; if responses recorded in either of these positions were larger than those observed in the original z position, the experiment was terminated and the data were not used. This procedure was continued until the recording electrode was moved 150 mm beyond the cell body layer, at which point it was returned to the original placement in the s. lacunosum moleculare and field potentials were recorded again at this position. If the average peak amplitude of these field potentials differed by >20% of the amplitude originally recorded at this position, then the data were not used. The electrode then was moved to positions away from the cell body layer, until it reached the hippocampal fissure. These field potential data then were transformed into CSD data using the algorithm D2 described by Freeman and Nicholson (1975). Tissue conductivity was assumed to be constant and equal to 1, and therefore, CSDs are expressed in arbitrary units proportional to actual current densities.

Whole cell recordings (WCR) (Blanton et al. 1989) were made via Axopatch-1D patch-clamp amplifier (Axon Instruments). WCR pipettes (3-7 M) were prepared from borosilicate glass and filled with one of three WCR solutions (see further). Series resistances (<20 M) were not compensated and were estimated as described by Langdon et al. (1995). Only traces with stable series resistance were analyzed.

Experiments were begun 30-40 min after breaking into the cell to allow for complete cell dialysis. The stimulation intensity was adjusted to elicit postsynaptic responses the amplitude of which was 25-30% of maximum amplitude. This intensity was less than or equal to the stimulation intensity used in CSD experiments. Paired pulse stimulation (ISI = 60-80 ms) was used in the beginning of all experiments. In studies of MPP/LPP responses, ISI intervals varied from 30 to 200 ms to show the persistence of paired-pulse facilitation (PPF) or paired-pulse depression (PPD). Electrophysiological responses were filtered at 3-5 kHz, digitized, and stored on a computer for off-line analysis. Either responses to single-pulse stimulation or the first response to paired-pulse stimulation were used for I-V plots and NMDAR component estimation. All traces shown are averages of 6-10 individual waveforms.

Solutions and drug application

WCR pipettes were filled with one of three solutions: CsF + PTX [which contained (in mM) 120 cesium fluoride, 20 CsCl, 1 bis-(o-aminophenoxy)-N,N,N',N'-tetraacetic acid (BAPTA), 10 N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES),and 0.15-0.2 picrotoxin]; Cs-gluconate [which contained (in mM)115 cesium gluconate, 20 CsCl, 10 HEPES, 10 ethylene glycol-bis(-aminoethyl ether)-N,N,N',N'-tetraacetic acid (EGTA), 10 sodium phosphocreatine, 4 MgATP, and 0.3 GTP plus 50 U/ml creatine phosphate]; or K-gluconate [which contained (in mM) 115 potassium gluconate, 20 KCl, 10 HEPES, 10 EGTA, 10 sodium phosphocreatine, 4 MgATP, and 0.3 GTP plus 50 U/ml creatine phosphate]. The pH and osmolarity of the solutions were 7.2-7.4 and 270 mosM, respectively.

All drugs were delivered by bath application. The following drugs were used: the AMPA receptor (AMPAR) antagonist, 6-cyano-7-nitroxaline-2,3-dione (CNQX; 10 µM), the NMDAR antagonist, D-2-amino-5-phosphovalerate (D-APV) (25 µM), the GABA_A antagonist bicuculline methiodide (10 µM), and theGABA_B antagonist CGP35348 (500 µM). The GABA_B agonist baclofen was applied in concentrations of 2-5 µM, the mGluR agonist (D,L)-2-amino-4-phosphonobutiric acid (AP4) in concentrations of 25-50 µM to affect selectively either MPP- or LPP-evoked responses (Koerner and Cotman 1991; Lantorn and Cotman 1981). All drugs were purchased from Sigma (St. Louis, MO) with exception of D-APV (Research Biochemicals, Natick, MA) and CGP (generous gift of Giba Geigy, Basel, SZ).

	RESULTS

Abstract Introduction Methods Results Discussion References

Isolation of monosynaptic PP-evoked response

Electric stimulation in the subiculum can result in multiple polysynaptic responses due to activation of intrinsic hippocampal circuits. Specifically, the monosynaptic PP response in area CA3 can be contaminated by disynaptic responses due to activation of MFs, disynaptic responses due to activation of collaterals of CA3 pyramidal cells, and direct recruitment of Schaffer collaterals in area CA1 resulting in either antidromic activation of CA3 pyramidal cells or monosynaptic responses due to activation of collateral branches in this area. To eliminate responses from mossy fibers, we placed a cut between the DG and area CA3 up to the suprapyramidal blade of dentate (Fig. 1A). In addition, we placed the stimulation electrodes far from area CA3 (Fig. 1A), as close as possible to the hippocampal fissure, and used low stimulation intensities that resulted in responses with amplitude <30% of maximum amplitude. These measures were taken to reduce the probability of antidromic stimulation of CA3 pyramidal cells, and of activation of CA3 collaterals. Absence of collateral and antidromic activation was verified by the lack of an evoked population spike in s.pyramidale as measured extracellularly (data not shown) and by the distribution of current sinks and sources, as determined by CSD analysis (Fig. 1B).

Data for CSD analysis that met the criteria described in METHODS were collected from five slices. The strength of stimulation was <60% of that required to elicit a population spike. All CSDs showed a pattern of sources and sinks that is similar to the example in Fig. 1B. PP stimulation resulted in a large, short latency current sink, of <10 ms duration, in an area between 50 and 150 m from the hippocampal fissure. This large current sink was accompanied by a current source in the s. radiatum and was followed a few milliseconds later by a small current source in the CA3 cell body layer. No evidence was seen for current sinks of significant amplitude in either the s. radiatum or the cell body layer (for comparison, a small current negativity at 15-20 ms and at 500 µm from the hippocampal fissure represents <3% of the s. lacunosum moleculare current sink). This indicates that stimulation in s. lacunosum moleculare of the subiculum resulted in specific activation of PP synapses.

To assess the validity of the assumption that all current sinks and sources were located in recording area, the sum of current sinks and sources was calculated for all time points during two time periods: 3-8 ms before PP stimulation and 12-60 ms after PP stimulation. This sum averaged0.45 ± 33 units/time point (mean ± SD) before stimulation and 0.58 ± 27 units/time point (SD) after stimulation. Maximal sum of sinks and sources after stimulation was 39 units/time point, which is within 1.2 SD from the sum value before stimulation.

We concluded that by using low intensities of stimulation, positioning the stimulation electrode far from area CA3 and making a cut as described above, we were able to isolate the monosynaptic PP-evoked responses for the following pharmacological analysis.

Isolation of EPSC and IPSC evoked by PP stimulation

Stimulation of the PP resulted in a composite response in CA3 pyramidal cells that consisted of both inward and outward currents (Fig. 2C) when recorded in the whole cell mode at 80 mV. To isolate EPSCs and different components of IPSCs, we varied the contents of the recording pipette. Both fluoride (F) and picrotoxin block GABA_A Cl channels when applied internally (Bormann 1988; Inomata et al. 1988; Kay 1992), and Cs blocks GABA_B K⁺ channels (Gahwiler and Brown 1985). We therefore used CsF + picrotoxin WCR solution to isolate EPSCs (Figs. 2A and 4). Alternatively, we used Cs-gluconate WCR solution to suppress the GABA_B-mediated component of IPSC and to leave the GABA_A-mediated component intact (Fig. 3, A and B). Finally, we used K-gluconate WCR solution, which supports glutamatergic and both GABAergic conductances (Steilzer et al. 1988), to study the intact PP-evoked response (Figs. 2C and 3C).

View larger version (26K): [in this window] [in a new window]   FIG. 2. I-V relations for perforant path-evoked responses. A: excitatory postsynaptic currents (EPSCs) measured at various holding potentials (CsF+picrotoxin in the whole cell recordings pipette). B: plot of EPSC peak amplitude (at 15 ms from the stimulation artifact) vs. holding potential. C: perforant path-evoked responses recorded with a pipette containing K-gluconate had 3 components: early (peak at 6 ms, reversal at 20 mV), intermediate (peak at 20 ms, reversal at 43 mV), and late (peak at 130 ms, reversal at 100 mV). D: I-V plot for the late and the intermediate components.

View larger version (14K): [in this window] [in a new window]   FIG. 4. Pharmacological isolation of excitatory components of the perforant path-evoked EPSC. A: control recordings ([Mg2+] = 3 mM, holding potential, 60 mV) and N-methyl-D-aspartate receptor (NMDAR)-mediated component of EPSC isolated in the presence of 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; 10 mM) (holding potential, 60 mV, [Mg2+] = 1.25 mM, stimulation intensity increased by 50%). B: -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPAR)-mediated component of EPSC isolated by blocking NMDAR-mediated component with 2-amino-5-phosphovalerate (APV; 25 mM). NMDAR-mediated component calculated by subtracting AMPAR-mediated component from the control response (control-APV) [holding potential, 80 mV, ([Mg2+] = 3 mM)]. C: Proportion of EPSC mediated by NMDARs, calculated as the ratio of the peak amplitude of the NMDAR-mediated EPSC to the peak amplitude of the control EPSC in experiments with APV application as a function of holding potential ([Mg2+] = 3 mM, mean ± SE, n = 6 for each holding potential except for 100 mV where n = 3). In all experiments, the pipette solution contained CsF + picrotoxin.

View larger version (9K): [in this window] [in a new window]   FIG. 3. Perforant path-evoked inhibitory responses. Mixed EPSC with -aminobutyric acid-A GABAA-mediated inhibitory postsynaptic current (IPSC; control), and isolated EPSC after suppression of IPSC with bicuculline at 10 mV (A) and 50 mV (B) (pipette solution contained Cs-gluconate). C: perforant path-evoked response recorded in current-clamp using pipette containing K-gluconate. Long-lasting inhibitory component was suppressed by addition of CGP 35348 (500 mM).

The reversal potential of isolated EPSCs in a sample cell was 8.4 mV (Fig. 2B; mean reversal across experiments was 3.1 ± 6.2 mV (SD, n = 11). These EPSCs were suppressed partially by addition of the NMDAR antagonist D-APV (25 mM) to the bath solution (Fig. 4B) and were suppressed completely in the presence of both D-APV and CNQX. These results indicate that PP-evoked EPSCs are glutamatergic and mediated by both AMPARs and NMDARs.

With Cs-gluconate in the pipette the PP-evoked response reversed at about 40 mV. This reversal potential indicates that the response consists of both an EPSC and a GABA_A-mediated IPSC. We isolated pure GABA_A-mediated IPSCs by recording at 10 mV (close to the EPSC reversal potential; Fig. 3A). These IPSCs were suppressed completely by the addition of bicuculline (10 µM). However, EPSCs and IPSCs did not add linearly. Thus at the reversal potential of the GABA_A-mediated IPSCs (50 mV), application of bicuculline resulted in an increase of the amplitude of the depolarizing response (Fig. 3B). This indicates that in control conditions EPSCs are reduced by GABA_A-ergic shunting (50% reduction at 50-60 mV).

Finally, all three components of the PP-evoked response, excitatory, fast and slow inhibitory, were observed with K-gluconate in the WCR pipette (Fig. 2, C and D). The intermediate, GABA_A-ergic component, as in the previous set of experiments, reversed at about 40 mV. The long-lasting component of the PP-evoked response reversed at 100 mV and was suppressed in the presence of the GABA_B antagonist, CGP35348 (500 µM; Fig. 3C). On the basis of these data, we conclude that the long-lasting inhibitory component is mediated by GABA_B receptor-dependent conductances.

Proportion of the NMDAR-mediated component of PP-evoked EPSC

EPSCs isolated with CsF + picrotoxin WCR solution were analyzed to provide with a quantitative estimate of the proportion of the evoked EPSC that is mediated by NMDARs. With the stimulation paradigm described above, only responses showing PPF were recorded and are analyzed in this section. NMDAR-mediated EPSC (NMDAR-EPSC) component was isolated by applying CNQX (10 µM; Fig. 4A). In the absence of AMPAR-EPSCs, NMDAR component was strongly reduced, therefore the intensity of stimulation was increased to obtain a measurable response. In other experiments, NMDAR-mediated EPSCs were obtained by subtraction of EPSCs recorded in the presence of D-APV (25 µM) from control EPSCs (control-APV, Fig. 4B). The ratio of the peak amplitude of NMDAR-EPSC to the peak amplitude of control EPSCs was used to determine the contribution of NMDAR-EPSCs to the total EPSC at different membrane potentials, and different extracellular [Mg²⁺]. The integral of the EPSC over time (charge) is another quantative measure for NMDAR-mediated EPSC. We calculated both the ratio of the amplitude and ratio of the charge of NMDA to control responses for five cells and found no significant differences, therefore here we report amplitude ratios only.

The amplitude of the NMDAR-mediated response varied as a function of holding potential (Fig. 4C). The NMDAR-mediated EPSC was 40 ± 8% (SE) of the total response at 40 mV and did not differ significantly from this value in a range of potentials between 60 and 20 mV and for both [Mg²⁺] used (1.25 and 3 mM). At more negative potentials, the proportion of the NMDAR-mediated EPSC decreased (Fig. 4C).

Variability of the paired-pulse profiles of PP-evoked EPSCs

The PP input to area CA3 consists of both LPP and MPP fibers. LPP-evoked responses in the DG show PPF of EPSPs and sensitivity to the mGluR agonist AP4 (Koerner and Cotman 1981), whereas MPP-evoked responses show PPD and are suppressed by the GABA_B agonist baclofen (Lanthorn and Cotman 1981). This suppression of synaptic transmission is thought to be caused by activation of presynaptic receptors (Harris and Cotman 1985) located on the PP synaptic terminals. PP inputs to both DG and area CA3 are thought to arise form the same cell population in the EC (Tamamaki and Nojo 1993; Witter et al. 1989) and thus may show the same presynaptic properties. Data from in vivo studies suggest that MPP- and LPP-evoked responses in area CA3 may show the same paired-pulse profiles as those recorded in the DG (Briendl et al. 1994) (Fig. 2). We therefore hypothesized that MPP- and LPP-evoked responses in area CA3 in the slice preparation can be distinguished on the basis of their paired-pulse profiles.

With low stimulation intensities (which resulted in responses of <15% of maximal amplitude), in a number of experiments, we observed PP-evoked EPSCs that exhibited either PPD or PPF (PPD or PPF responses, respectively; Fig. 5, A and B). We hypothesized that MPP fibers were activated selectively in experiments where PPD was evoked by low-intensity stimulation, whereas predominant stimulation of LPP fibers was accompanied by PPF. To test this hypothesis, we applied drugs that are known to affect selectively either MPP- or LPP-evoked responses in the DG. Consistent with observations in the DG, baclofen suppressed PPD responses by 66% (measured as ratio of the amplitude of the first EPSC in the presence of drug to the amplitude of control EPSC), whereas PPF responses were reduced by only 20% (Fig. 5, A and C). On the other hand, AP4 suppressed PPF responses by 49% and PPD responses by 6% (Fig. 5, B and C). On the basis of these data, we conclude that PP-evoked EPSCs that exhibit PPD mainly represent MPP-evoked EPSCs, whereas those that exhibit PPF mainly represent LPP-evoked EPSCs.

View larger version (20K): [in this window] [in a new window]   FIG. 5. Two types of perforant path responses. A: responses of the 1st type demonstrated paired-pulse depression in control conditions (PPD responses). Addition of baclofen suppressed PPD responses by 65%. B: responses of the 2nd type demonstrated paired-pulse facilitation (PPF responses). Addition of AP4 suppressed responses by 50%. C: summary of baclofen and AP4 action on PPD and PPF responses. Suppression of the response was determined from a ratio of the amplitude of the 1st EPSC in the pair recorded in a presence of drug (either AP4 or baclofen) to the EPSC amplitude recorded under control conditions (mean ± SE, for PPF/AP4 experiments, n = 8; for other experiments, n = 4). In all experiments, the pipette solution contained CsF + picrotoxin.

Increasing the stimulation intensity had a similar effect on PPF and PPD responses. Specifically, it decreased PPF in responses that initially showed PPF and decreased PPD in the responses that initially showed PPD. This suggests that positioning of stimulating electrode did not result in selective activation of either pathway but instead resulted in mixed responses. In agreement with this conclusion, selectivity of drug action also was reduced when higher stimulation intensity was used. We concluded that stimulation intensity (30% of maximal) used in the pharmacological characterization of EPSCs and IPSCs reported above resulted in mixed LPP/MPP response in area CA3.

In the DG, it is possible to stimulate selectively LPP or MPP by placing the stimulating electrode in either the outer or the middle part of s. moleculare. We tested whether repositioning of the stimulation electrode further from the hippocampal fissure in the subiculum could result in stimulation of MPP instead of LPP. In our experiments, such change of position resulted in the shift from PPF to PPD of the whole cell measured EPSCs only in a few cases. Changes in stimulation protocol were not effective in isolating MPP- and LPP-evoked responses.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

This study is the first characterization of some of the pharmacological properties of PP-evoked monosynaptic EPSCs and disynaptic IPSCs recorded from CA3 pyramidal cells. With the hippocampal slice preparation, we isolated monosynaptic PP-evoked EPSCs and demonstrated that they are mediated by both AMPARs and NMDARs. We have shown that stimulation of the PP also results in disynaptic IPSCs, which have both GABA_A and GABA_B receptor-mediated components. Finally, in a number of experiments, the stimulation of the PP resulted in two electrophysiologically and pharmacologically distinct types of responses, which we identified as originating from LPP and MPP.

Our estimates of the proportion of the NMDAR-mediated EPSC are close to those reported by Keller et al. (1991) and Colino and Malenka (1993) on PP-evoked EPSCs and EPSPs components in the DG. In contrast to these findings, however, in our experiments, NMDA-mediated component began to decrease only at holding potentials more negative than 60 mV. There are three possible explanations for the apparent difference in voltage dependence for the NMDAR-mediated proportion of EPSC. First, as suggested by Spruston et al. (1993), the quality of the dynamic voltage clamp may not be sufficient to prevent voltage escape from distal dendrites during synaptic events. Thus at the peak of the EPSC, the dendritic membrane potential is more positive than the holding potential at the soma and NMDA channels can be relieved from the Mg²⁺ block at holding potentials more negative than 40 mV. Consistent with this explanation NMDAR component of EPSC at 60 mV was large when AMPAR-EPSCs were activated simultaneously (Fig. 4B), but it was reduced when AMPAR-EPSCs were blocked. Second, the large ratio of NMDAR- to AMPAR-mediated responses relative to that previously reported in the DG at negative membrane potentials (Colino and Malenka 1993; Keller et al. 1991) may be due to differences in electronic length between pyramidal and granule cells. Because of dendritic cable filtering, fast EPSCs (AMPAR-mediated) arriving distally with respect to the soma (as in the case of pyramidal cells) are more attenuated than those arriving proximally (as in the case of granule cells). The difference in attenuation of the slow (NMDAR-mediated) component between pyramidal and granule cells is relatively smaller; therefore the ratio of NMDAR- to AMPAR-mediated components in pyramidal cells can be larger than that in granule cells. Finally, this difference also may represent an actual difference in the proportion of NMDA and AMPA receptors activated at PP-CA3 (compared with PP-DG) synapses.

Activation of voltage-dependent conductances, which exist in the dendrites of pyramidal cells (Magee and Johnston 1995; Stuart and Sakmann 1995) could confound the effects of drug application and distort the I-V function for PP-evoked responses (Bernander et al. 1994; Taylor et al. 1995). To minimize the activation of voltage-dependent conductances, we used low-intensity PP stimulation resulting in EPSCs the amplitude of which was <30% of maximum. EPSCs evoked with this stimulation corresponded to EPSPs <3 mV (measured in current clamp). EPSPs of such small amplitude have been shown to involve active dendritic conductances to only a small degree (Stuart and Sakmann 1995). In addition, CsF + picrotoxin WCR solution, which was used in our EPSC studies inactivates some Ca²⁺ voltage-dependent conductances and blocks K⁺ voltage-dependent conductances. Linearity of the I-V plot for isolated AMPA-EPSCs (not shown) also indicates that voltage-dependent conductances were not activated in these experiments.

Stimulation of the PP resulted in GABA_A and GABA_B receptor-mediated IPSCs in area CA3. Our data are in agreement with other reports on fast and slow IPSP components in PP-evoked responses recorded from granule cells in DG and pyramidal cells in areas CA3 (Buckmaster and Schwartzkroin 1995) and CA1 (Empson and Heinemann 1995b). We also observed nonlinear summation of EPSCs and GABA_A-mediated IPSCs at negative membrane potentials, which suggests that GABA_A inhibition shunts excitation evoked by PP stimulation. However, even with GABA_A inhibition intact, EPSC are reduced only by 50% at membrane potentials near rest.

The PP is heterogeneous fiber tract; both LPP and MPP fibers synapse in s. lacunosum moleculare of area CA3. In the DG, it is possible to distinguish between LPP- and MPP-evoked responses by placement of the stimulating electrode (in the either outer or middle s. moleculare, respectively), kinetics (MPP responses have a faster rising phase), or paired-pulse stimulation (PPD vs. PPF for MPP or LPP, respectively). The same laminar distribution of MPP and LPP exists in CA3 area. Because the PP fibers to CA3 and DG originate from the same cell population in the entorhinal cortex, we expect that the PP-CA3 synapses will have presynaptic properties (including paired pulse facilitation/depression) similar to those observed for the PP synapses in the DG. In the present study, the type of the paired-pulse response recorded from CA3 pyramidal cells did not vary with the distance of the stimulation electrode from the hippocampal fissure. This is in agreement with anatomic studies (Steward 1976), which have demonstrated that in area CA1 and the subiculum, fibers of both the MPP and LPP are present throughout all of the s. lacunosum moleculare without clear separation. However, despite the anatomic overlap of these two pathways, using very low intensities of stimulation we could observe responses that show either PPFor PPD.

We also attempted to exclude the postsynaptic factors that can affect paired-pulse ratio of PP-evoked responses. The observed differences in the paired-pulse ratio can not be accounted for by recruitment of inhibitory responses, because they were blocked by the whole cell solution. Further in LPP/MPP separation experiments, stimulation intensity was very low, which reduced the probability of antidromic or collateral contamination of EPSCs. Finally, we have compared these two types of responses with the LPP- and MPP-evoked responses observed in DG with respect to their pharmacological sensitivity. Results of differential action of baclofen and AP4 on PPD and PPF responses (Fig. 5) are in agreement with the data on selective action of these compounds on MPP and LPP synapses in the DG. Therefore, we conclude that PPD responses, which we elicited with this low intensity of stimulation, were mainly (but not exclusively) due to activation of MPP fibers, whereas PPF responses were mainly due to activation of LPP fibers. Consistent with this interpretation, the degree of PPD (for PPD responses) or PPF (for PPF responses) decreased when stimulation strength was increased to the level used in other experiments described here. This indicates that our regular stimulation protocol resulted in activation of both pathways.

In this study, we observed that PP-evoked response recorded from CA3 pyramidal cells consists of four components: AMPAR- and NMDAR-mediated excitatory and GABA_A- and GABA_B-mediated inhibitory ones. The NMDAR-mediated component of PP EPSP may play significant role in activity-dependent synaptic plasticity and thus will be important in the further characterization of this pathway. As we demonstrated, PP-evoked inhibition can modulate EPSPs in CA3 pyramidal cells. More detailed studies are necessary to determine whether this effect depends on strength of PP stimulation and how it may affect synaptic plasticity in this pathway.

	ACKNOWLEDGEMENTS

  This work was supported by National Institute of Neurological Disorders and Stroke Grant NS-24288 and by a Howard Hughes Medical Institute predoctoral fellowship to N. N. Urban.

	FOOTNOTES

  Address for reprint requests: German Barrionuevo, Dept. of Neuroscience, 446 Crawford Hall, University of Pittsburgh, Pittsburgh, PA 15260.

  Received 17 July 1997; accepted in final form 3 December 1997.

	REFERENCES

Abstract Introduction Methods Results Discussion References

• AMARAL, D. G. Emerging principles of intrinsic hippocampal organization. Curr. Opin. Neurobiol. 3: 225-229, 1993.[Medline]
• AMARAL, D. G., ISHIZUKA, N., CLAIBORNE, B. Neurons, numbers and the hippocampal network. Prog. Brain Res. 83: 1-11, 1990.[Medline]
• BERGER, T. W., YECKEL, M. F. Long-term potentiation of entorhinal afferents to the hippocampus: enhanced propagation of activity through the trisynaptic pathway. In: Long-Term Potentiation: A Debate of Current Issues. Cambridge, MA: MIT Press, 1991, p 327-356.
• BERNANDER, O., KOCH, C., DOUGLAS, R. J. Amplification and linearization of distal synaptic input to cortical pyramidal cells. J. Neurophysiol. 72: 2743-2753, 1994.[Abstract/Free Full Text]
• BLANTON, M. G., LOTURCO, J. J., KRIEGSTEIN, A. R. Whole-cell recording from neurons in slices of reptilian and mammalian cerebral cortex. J. Neurosci. Methods 30: 203-210, 1989.[Medline]
• BORMANN, J. Electrophysiology of GABA_A and GABA_B receptor subtypes. Trends Neurosci. 11: 112-116, 1988.[Medline]
• BRAMHAM, C. R., ERRINGTON, M. L., BLISS, T.V.P. Naloxone blocks the induction of long-term potentiation in the lateral but not the medial perforant pathway in the anaesthetized rat. Brain Res. 449: 352-356, 1988.[Medline]
• BRAMHAM, C. R., MILGRAM, N. W., SREBRO, B. Activation of AP5-sensitive NMDA receptors is not required to induce LTP of synaptic transmission in the lateral perforant path. Eur. J. Neurosci. 3: 1300-1308, 1991.[Medline]
• BREINDL, A., DERRICK, B. E., RODRIGUEZ, S. B., MARTINEZ, J. L. JR Opioid receptor-dependent long-term potentiation at the lateral perforant path-CA3 synapse in rat hippocampus. Brain Res. Bull. 33: 17-24, 1994.[Medline]
• BUHL, E. H., HAN, Z. S., LORINCZI, Z., STEZHKA, V. V., KARNUP, S. V., SOMOGYI, P. Physiological properties of anatomically identified axo-axonic cells in the rat hippocampus. J. Neurophysiol. 71: 1289-1307, 1994.[Abstract/Free Full Text]
• BUCKMASTER, P. S., SCHWARTZKROIN, P. A. Interneurons and inhibition in the dentate gyrus of the rat in vivo. J. Neurosci. 15: 774-789, 1995.[Abstract]
• COLBERT, C. M., LEVY, W. B. Electrophysiological and pharmacological chracterization of perforant path synapses in CA1: mediation by glutamate receptors. J. Neurophysiol. 68: 1-8, 1992.[Abstract/Free Full Text]
• COLINO, A., MALENKA, R. C. Mechanisms underlying induction of long-term potentiation in rat medial and lateral perforant paths in vitro. J. Neurophysiol. 69: 1150-1159, 1993.[Abstract/Free Full Text]
• DOLLER, H. J., WEIGHT, F. F. Perforant pathway activation of hippocampal CA1 stratum pyramidale neurons: electrophysiological evidence for a direct pathway. Brain Res. 237: 1-13, 1982.[Medline]
• EMPSON, R. M., HEINEMANN, U. Perforant path connections to area CA1 are predominantly inhibitory in the rat hipocampal-entorhinal cortex combined slice preparation. Hippocampus 5: 104-107, 1995a.[Medline]
• EMPSON, R. M., HEINEMANN, U. The perforant path projection to hippocampal area CA1 in the rat hippocampal-entorhinal cortex combined slice. J. Physiol. (Lond.) 484: 707-720, 1995b.[Abstract/Free Full Text]
• FREEMAN, J. A., NICHOLSON, C. Experimental optimization of current source density technique for anuran cerebellum. J. Neurophysiol. 38: 369-382, 1975.[Abstract/Free Full Text]
• GAHWILER, B. H., BROWN, D. A. GABA_B-receptor-activated current in voltage-clamped CA3 pyramidal cells in hippocampal cultures. Proc. Natl. Acad. Sci. USA 82: 1558-1562, 1985.[Abstract/Free Full Text]
• HARRIS, E. W., COTMAN, C. W. Effects of synaptic antagonists on perforant path paired-pulse plasticity: differentiation of pre- and postsynaptic antagonism. Brain Res. 334: 348-353, 1985.[Medline]
• HASSELMO, M. E., SCHNELL, E. Laminar selectivity of the cholinergic suppression of synaptic transmission in rat hippocampal region CA1: computational modeling and brain slice physiology. J. Neurosci. 14: 3898-3914, 1994.[Abstract]
• HENZE, D. A., CAMERON, W. E., BARRIONUEVO, G. Dendritic morphology and its effects on the amplitude and rise-time of synaptic signals in hippocampal CA3 pyramidal cells. J. Comp. Neurol. 369: 331-344, 1995.
• INOMATA, N., TOKUTOMI, T., OYAMA, Y., AKAIKE, N. Intracellular picrotoxin blocks phenobarbital-gated Cl conductance. Neuroci. Res. 6: 72-75, 1988.
• KAY, A. An intracellular medium formulary. J. Neurosci. Methods 44: 91-100, 1992.[Medline]
• KELLER, B., KONNERTH, A., YAARI, Y. Patch clamp analysis of excitatory synaptic currents in granule cells of rat hippocampus. J. Physiol. (Lond.) 435: 275-293, 1991.[Abstract/Free Full Text]
• KIM, H. G., BEIERLEIN, M., CONNORS, B. W. Inhibitory control of excitable dendrites in neocortex. J. Neurophysiol. 76: 1810-1814, 1994.
• KISS, J., BUZAKI, G., MORROW, J. S., GLANTZ, S. B., LERANTH, C. Entorhinal innervation of paravalbumin containing neurons (Basket and Chandelier cells) in the rat Ammon's Horn. Hippocampus 6: 239-246, 1996.[Medline]
• KOERNER, J. F., COTMAN, C. W. Micromolar L-2-amino-4-phosphobutiric acid selectively inhibits peforant path synapses from lateral entorhinal cortex. Brain Res. 216: 192-198, 1991.
• LAMBERT, J.D.C., JONES, R.S.G. Activation of N-methyl-D-aspartate receptors contributes to the EPSP at perforant path synapses in the rat dentate gyrus in vitro. Neurosci. Lett. 97: 323-328, 1989.[Medline]
• LANGDON, R. B., JOHNSON, J. W., BARRIONUEVO, G. Posttetanic potentiation and presynaptically induced LTP at the mossy fiber synapse in rat hippocampus. J. Neurobiol. 26: 370-385, 1995.[Medline]
• LANTORN, T. H., COTMAN, C. W. Baclofen selectively inhibits excitatory synaptic transmission in the hippocampus. Brain Res. 225: 171-178, 1981.[Medline]
• MAGEE, J. C., JOHNSTON, D. Characterization of single voltage-gated Na⁺ and Ca²⁺ channels in dendrites of CA1 cells. J. Physiol. (Lond.) 487: 67-90, 1995.[Abstract/Free Full Text]
• MCNAUGHTON, B. L. Evidence for two physiologically distinct perforant pathways to the fascia dentata. Brain Res. 199: 1-19, 1980.[Medline]
• MCNAUGHTON, B. L., BARNES, C. A., MELTZER, J., SUTHERLAND, R. J. Hippocampal granule cells are necessary for normal spatial learning, but not for spatially-selective pyramidal cell discharge. Exp. Brain Res. 76: 485-496, 1989.[Medline]
• O'REILLY, R. C., MCCLELLAND, J. L. Hippocampal conjunctive encoding, storage, and recall: avoiding a trade-off. Hippocampus 4: 661-682, 1994.[Medline]
• SPRUSTON, N., JAFFE, D. B., WILLIAMS, S. H., JOHNSTON, D. Voltage and space-clamp errors associated with the measurement of electrotonically remote synaptic events. J. Neurophysiol. 70: 781-802, 1993.[Abstract/Free Full Text]
• STALEY, K. J., MODY, I. Shunting of excitatory input to dentate gyrus granule cells by a depolarizing GABA_A-receptor-mediated postsynaptic conductance. J. Neurophysiol. 68: 197-212, 1992.[Abstract/Free Full Text]
• STELZER, A., KAY, A. R., WONG, R.K.S. GABA_A-receptor function in hippocampal cells is maintained by phosphorylation factors. Science 241: 339-341, 1988.[Abstract/Free Full Text]
• STUART, G., SAKMANN, B. Amplification of EPSPs by axosomatic sodium channels in neocortical pyramidal neurons. Neuron 15: 1065-1076, 1995.[Medline]
• STEWARD, O. Topographic organization of the projections from the enthorinal area to the hippocampal formation of the rat. J. Comp. Neurol. 167: 285-314, 1976.[Medline]
• TAMAMAKI, N. AND NOJYO, Y. Projection of the entorhinal layer II neurons in the rat as revealed by intracellular pressure-injection of neurobiotin, Hippocampus 3: 471-480, 1993.
• TAYLOR, G. C, COLES, J. A., EILBECK, J. C. Conditions under which Na⁺ channels can boost conduction of small graded potentials. J. Theor. Biol. 172: 379-386, 1995.[Medline]
• URBAN, N. N., BARRIONUEVO, G. Induction of Hebbian and non-Hebbian mossy fiber LTP by distinct patterns of high-frequency stimulation. J. Neurosci. 16: 4293-4299, 1996.[Abstract/Free Full Text]
• WITTER, M. P., GROENEWEGEN, H. J., LOPES DA SILVA, F. H., LOHMAN, A.H.M. Functional organization of the extrinsic and intrinsic circuitry of the parahippocampal region. Prog. Neurobiol. 33: 161-253, 1989.[Medline]
• WITTER, M. P. Organization of the entorhinal-hippocampal system: a review of current anatomical data. Hippocampus 3: 33-44, 1993.
• YECKEL, M. F., BERGER, T. W. Feedforward excitation of the hippocampus by afferents form the entorhinal cortex: redefinition of the role of the trisynaptic pathway. Procs. Natl. Acad. Sci. USA 87: 5832-5836, 1990.[Abstract/Free Full Text]
• YECKEL, M. F., BERGER, T. W. Monosynaptic excitation of hippocampal CA1 pyramidal cells by affarents from the entorhinal cortex. Hippocampus 5: 108-114, 1995.[Medline]

0022-3077/98 $5.00 Copyright ©1998 The American Physiological Society