| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Departamento de Fisiología, Biofísica y Neurociencias, Centro de Investigación y de Estudios Avanzados del IPN, 07000; and 2Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México, Ciudad Universitaria, Mexico D.F. 04510
Submitted 30 October 2002; accepted in final form 6 February 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
-amino butyric acid (GABA), and the vesicular GABA transporter mRNA. Their expression is regulated in an activity-dependent manner and coincides with the appearance of GABAergic transmission from the mossy fibers (MF) to pyramidal cells in area CA3. These data support the hypothesis that MF are able to release glutamate and GABA. Following the principle that a given neuron releases the same neurotransmitter(s) onto all its targets, we here demonstrate the emergence, after a generalized convulsive seizure, of MF GABAergic signaling sensitive to activation mGluR-III onto pyramidal cells and interneurons of CA3. Despite this, excitation overrides inhibition in interneurons, preventing disinhibition. Furthermore, on blockade of GABA and glutamate ionotropic receptors, an M1-cholinergic depolarizing signal is also revealed in both targets, which postsynaptically modulates the glutamatergic and GABAergic fast neurotransmission. The emergence of these nonglutamatergic signals depends on protein synthesis. In contrast to cholinergic responses evoked by associational/commissural fibers activation, cholinergic transmission evoked by DG stimulation is only observed after seizures and is strongly depressed by the activation of mGluR-II, whereas both are depressed by M2-AChR activation. With immunohistological experiments, we show that this cholinergic pathway runs parallel to the MF. Thus seizures compromise a delicate balance of excitation and inhibition, on which a complex interaction of different neurotransmitters emerges to counteract excitation at pre- and postsynaptic sites. Particularly, MF GABAergic inhibition emerges to exert an overall inhibitory action on CA3. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
); therefore after being excited, they inhibit pyramidal cells and other interneurons (Miles et al. 1996; Urban and Barrionuevo 1998). Thus excitatory and inhibitory synaptic responses in CA3 to MF stimulation disappear when glutamatergic transmission is blocked, indicating that GABAergic responses are disynaptically mediated (Buzsáki 1984; Crawford and Connor 1973; Dichter and Spencer 1969; Yamamoto 1972).
However, we recently demonstrated that after seizures (Gutiérrez 2000; Gutiérrez and Heinemann 2001) or after synaptic or direct kindling-like stimulation of the DG in vitro without inducing epileptiform activity (Gutiérrez 2002), MF activation evokes monosynaptic GABAergic potentials in pyramidal cells. This and immunohistochemical (Ramírez and Gutiérrez 2001; Schwarzer and Sperk 1995; Sloviter et al. 1996) and neurochemical evidence (Gómez-Lira et al. 2002; Taupin et al. 1994a,b) suggest that granule cells can synthesize and release GABA in an activity-dependent manner. Furthermore, we also demonstrated that the granule cells and their MF contain and express the vesicular GABA transporter (VGAT) mRNA in an activity-dependent manner (Lamas et al. 2001).
It is a general principle that all the terminals of a given neuron release the same neurotransmitter(s). However, because the MF are functionally compartmentalized (Maccaferri et al. 1998; Tóth et al. 2000), it is reasonable to consider the possibility that release of glutamate and GABA from MF could also be segregated according to the type of target onto which they impinge. Therefore to support the hypothesis that MF release glutamate and GABA, it was important to test whether the simultaneous glutamatergic and GABAergic transmission is expressed onto both MF targets in CA3: the pyramidal cells and the interneurons.
Interestingly, in a previous study (Gutiérrez and Heinemann 2001), we described that after the blockade of the aforementioned MF-evoked GABAergic response a putative cholinergic response could be uncovered. However, cholinergic transmission to area CA3 has been shown to come exclusively from the associational/commissural fibers (A/C), and its modulation in the MF projection is exerted through indirect synaptic interactions, whereby inhibitory interneurons are excited and presynaptically modulate MF glutamate release (Vogt and Regehr 2001). In view of this, we decided to characterize the seizure-induced DG-evoked cholinergic transmission onto CA3 and compare it to that evoked by the stimulation of the stratum radiatum, where the A/C pathway is known to pass (Vogt and Regehr 2001). This would also clarify how these responses interact in CA3 neurons.
Here we demonstrate that after a generalized convulsive seizure, DG activation evokes monosynaptic GABAergic and M1-cholinergic responses both in pyramidal cells and interneurons of CA3, through the activation of different pathways, each with a distinctive pharmacological profile that permits their selective presynaptic modulation. Their emergence depends on protein synthesis.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
For our electrophysiological experiments, we used control and pentylenetetrazol (PTZ)-injected Wistar rats (230–250 g). The intra-peritoneal injection of PTZ (Sigma; 70 mg/kg) invariably induced a generalized convulsive cloniconic seizure that lasted 80–120 s, after which myoclonias and wet-dog shakes were apparent. The rats were decapitated under deep ether anesthesia. All experimental procedures were approved by the Committee on Ethical Animal Research of our Institution and of the Ministry of Health. Brains were rapidly dissected and combined entorhinal cortex-hippocampus slices (400 µm) were cut with a vibro slicer (Campden Instruments, England) submerged in oxygenated artificial cerebrospinal fluid (ACSF) at 4°C. The PTZ-treated group was decapitated 1 h after the injection. For some experiments, slices from control and PTZ-treated rats were incubated in artificial cerebrospinal fluid (ACSF) in the presence of cycloheximide (60 µM) or anisomycin (100 µM) (Jones et al. 1992; Otani et al. 1992) for 
2 h before they were transferred to the recording chamber. All slices, including those preincubated with the protein synthesis blockers, were transferred to a moisturized air-liquid interface recording chamber where they were constantly perfused with oxygenated normal ACSF at 35°C for 1 h before the recordings started. The ACSF contained (in mM) 124 NaCl, 3 KCl, 1.25 NaH2PO4, 2 MgSO4, 2 CaCl2, 26 NaHCO3, and 10 glucose; pH 7.35. The drugs used were diluted at the desired concentration in the ACSF, namely: the NMDA type glutamate receptor antagonist, (DL)-2-amino-5-phosphonovaleric acid (APV; 30 µM; Tocris), and the non-NMDA type glutamate receptor antagonist 6-nitro-7-sulfamoylbenzo(f)quinolaxine-2,3-dione (NBQX; 10 µM; Tocris); the GABAA receptor antagonist bicuculline methiodide (20 µM; Sigma); the GABAB receptor antagonist CGP55845A (1 µM; a gift from Novartis); the M1-cholinergic antagonists: pirenzepine (10 µM; RBI), atropine (10 µM, Sigma), and muscarinic toxin (MT7; 100 nM; Peptide Institute); the M2-cholinergic agonist oxotremorine (1 µM; Sigma); the group II mGluR agonist (2S,2'R,3'R)-2-(2',3'-dicarboxycyclopropyl)glycine (DCG-IV; 1 µM; Tocris); the group III mGluR agonist L(+)-2-amino-4-phosphonobutyric acid (L-AP4; 10 µM; Tocris); and the protein synthesis inhibitors, cycloheximide (60 µM; Sigma) and anisomycin (100 µM; Sigma).
An hour after incubation in normal ACSF, intracellular recordings were done with an AxoClamp 2B amplifier (Axon Instruments) and borosilicate microelectrodes (70–90 M
; Brown-Flaming Puller) filled with 2M potassium acetate and 2% biocytin (Sigma). Electrophysiologically identified pyramidal cells were recorded from regions CA3a and -b, and interneurons were all recorded from the marginal zone between the stratum lucidum and s. pyramidale of region CA3b (Fig. 1). Anatomical identification of the recorded cells was later carried out by the intracellular injection of biocytin. The signals were acquired and analyzed off-line with the program pClamp8 (Axon Instruments). The duration of the action potential, and the duration and amplitude of the afterhyperpolarization were measured as previously described (Scharfman 1995). The onset latency of the synaptic responses was measured from the beginning of the stimulation artifact to the first inflection point of the first derivative of the synaptic potential with the cursor measure function of Clampfit. Membrane input resistance, rise time, and time to peak of the synaptic responses were measured with the built-in functions of Clampfit. The results are expressed as means ± SE.
|
Electrical stimulation was provided with bipolar glass-insulated platinum wire (50 µm) electrodes over the molecular layer of the DG and over the s. radiatum of the CA3 region (Vogt and Regehr 2001) (Fig. 1A). Pulses of 0.1 ms were delivered at a current intensity that evoked 75% of the excitatory postsynaptic potential (EPSP) amplitude needed to reach threshold for evoking action potentials. In the CA3 region, a glass pipette (1–2 M) filled with ACSF was positioned 200–300 µm apart from the recorded cell to stimulate local interneurons with monopolar pulses of 0.1 ms.
Immunohistological experiments
MF INTRACELLULAR LABELING. Slices to which a crystal of dextran, fluorescein (MW 3000; Molecular Probes) was inserted in the granular cell layer of the DG with a patch pipette were perfused with ACSF for 5 h and immersed in phosphate buffer (0.1 M; pH 7.4) with paraformaldehyde (4%) for 12 h and then transferred to a buffer with sucrose (30%) for 24 h. They were frozen and cut at 14 µm in a cryostate (Leica, CM 1850) and recovered in gelatin-coated slices. Some slices were counterstained with Evans blue (Sigma). These were covered with Vectashield (Vector) and observed and photographed under an epifluorescence microscope with appropriate filters for FITC and Texas Red (Nikon Optiphot-2).
For detection of cholin-acetyl-transferase (ChAT) immunoreactivity, four slices from 2 PTZ-treated rats and from two control rats were processed after MF dextran, fluorescein staining. The slices were fixed in phosphate buffer (0.1 M; pH 7.4) with paraformaldehyde (2%) and picric acid (0.18%) for 12 h and transferred to a buffer with sucrose (30%) for 24 h. The slices were frozen, cut, and mounted as described in the preceding text. Sections were preincubated with BSA (5%) for 2 h and incubated with goat polyclonal anti-ChAT (human) antibody (1:200; Chemicon) in PBS, Triton X-100 (0.3%) and BSA (5%) overnight in a humid atmosphere at room temperature. For control purposes, some sections were simultaneously processed in the absence of the primary antibody. After rinsing, they were incubated with biotinilated anti-goat IgG (1:100; Vector) 1 h, followed of avidine-Texas Red (1.200, Vector) 1 h. Finally sections were coverslipped with Vectashield fluorescent mounting medium (Vector) and examined with an epifluorescent microscope. Selected sections were analyzed with a scanning confocal microscope (MRC 1024 Bio-Rad) equipped with an argon/kripton laser. Images of the preparations stained with dextran, fluorescein were acquired with an excitation wavelength of 488 nm and ChAT Texas-red fluorescence with 568 nm. Acquisitions for each wavelength were done separately over the same fields and the combined renditions were digitally processed with the computer program Confocal Assistant (Tood Clark Brelje). The red signal corresponds to ChAT Texas red fluorescence and the green signal to dextran, fluorescein fluorescence. Colocalization of both signals would render a yellow labeling.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
Synaptic responses of pyramidal cells and interneurons to MF activation in control preparations
DG stimulation evoked in pyramidal cells and interneurons EPSP/inhibitory postsynaptic potential (IPSP) sequences, which were completely blocked by the perfusion of non-N-methyl-D-aspartate (NMDA) and NMDA receptors' antagonists (NBQX and APV; Fig. 1, A and B). The values of the amplitude, onset latency, time to peak and rise time of the monosynaptic control EPSP in pyramidal cells and interneurons are summarized in Table 1.
MF activation provokes GABAergic synaptic responses in pyramidal cells and interneurons after seizures
On DG stimulation, the pyramidal cells of the PTZ group responded with EPSP/IPSP sequences (Figs. 2, A1, and C1, and 3A1). In interneurons, however, only an EPSP was evident (Figs. 2, A2 and C2, 3A2). The values of the amplitude, onset latency, time to peak, and rise time of the monosynaptic EPSP in pyramidal cells and interneurons are summarized in Table 1. As compared with the control preparations, the amplitude, rise time, and time to peak of the EPSP evoked in both types of cells were statistically different in the PTZ group (Student's t-test; P < 0.01; the stimulation intensity was normalized as described in METHODS.
|
|
During the blockade of glutamatergic transmission, MF activation evoked fast IPSPs in pyramidal cells (Figs. 2, B1 and C1, and 3A1) and in interneurons (Figs. 2, B2 and C2, and 3A2), whose onset latencies were, for pyramidal cells: 4.3 ± 0.3 ms (n = 25) and for interneurons, 4.8 ± 0.06 ms (n = 10). The latency of the isolated IPSP for each cell type is not different to that of the corresponding control EPSP as determined by a one-way ANOVA test and Scheffé post-ANOVA contrast P < 0.01 (pyramidal cells, F = 0.07; interneurons, F = 6.46) and the variability of the latency of the IPSP (and that of the EPSP) for each neuron type is consistent with a monosynaptic contact (pyramidal cells, F = 2.5; interneurons, F = 1.4; Figs. 3A, 1 and 2, bar graphs).
The values of the amplitude, onset latency, time to peak, and rise time of the MF IPSP in pyramidal cells and interneurons are summarized in Table 1. The reversal potential of the IPSP obtained in pyramidal cells (-67 ± 0.7 mV; n = 25) and in interneurons (-69.4 ± 0.6 mV; n = 10; Fig. 2C, 1 and 2) was consistent with GABAergic responses, as also evidenced by its blockade with bicuculline (Fig. 3A, 1 and 2).
DG stimulation evokes cholinergic responses in CA3 after seizures
In contrast to control slices, in PTZ-treated preparations the stimulation of the DG in the presence of glutamate receptors' antagonists and bicuculline provoked a slow depolarizing potential (EPSPm), both in pyramidal cells (Fig. 3A1) and interneurons (Fig. 3A2). The mean amplitude, rise time, duration, and onset latency for this EPSPm both in pyramidal cells and interneurons are summarized in Table 1. On higher stimulation intensity, this EPSPm was enhanced but did not evoke action potentials in neither of the CA3 targets (not shown).
We then compared the EPSPm evoked by stimulation of the DG to that evoked by stimulation of the A/C fibers (Vogt and Regehr 2001). The characteristics of the A/C-evoked EPSPm are depicted in Table 1. Figure 3B shows the experimental arrangement and the responses to DG and A/C stimulation. Perfusion of glutamate receptors' antagonists block all synaptic responses to DG stimulation in control animals, but A/C stimulation evoked an EPSPm that could be pharmacologically isolated by perfusion of glutamatergic and GABAA,B antagonists (Fig. 3B2). On the contrary, in the PTZ-group, DG stimulation evoked an EPSPm in the presence of glutamatergic and GABAergic antagonists. The selective M1 cholinergic blockers: atropine (10 µM; n = 40), pirenzepine (10 µM; n = 50), and MT7 (100 nM; n = 6), completely blocked the EPSPm evoked by both sites of stimulation, confirming unequivocally that they were mediated by M1 cholinergic receptors (Fig. 3C). Current pulses applied during the EPSPm revealed an increment of the input resistance of the pyramidal cells by 39 ± 7% (n = 5; Fig. 3D).
Alternatively, to determine if the blockade of cholinergic receptors prior to the perfusion of the GABAA antagonist could alter the MF-evoked IPSP, we perfused cholinergic antagonists in the presence of glutamate receptors' antagonists. In these conditions, the MF IPSP evoked in pyramidal cells and interneurons had the characteristics depicted in Table 1, which show that the amplitude of the MF IPSP increases due to the blockade of the antagonizing depolarizing action of the cholinergic signal (Fig. 4, A and B) and the reversal potential of the IPSP shifts to -72.4 ± 0.6 (n = 9), coinciding with the reversal potential of the IPSP evoked by direct stimulation of interneurons (-72 ± 0.8 mV; n = 10), which does not recruit cholinergic fibers.
|
DG-evoked synaptic signals have a distinctive presynaptic modulation
Because activation of mGluR reduces neurotransmitter release from MF, we tested the effect of the mGluR-II agonist, DCG-IV, and of the mGluR-III agonist, L-AP4 on the MF-evoked responses. We tested the effect of both agonists in a group of cells and found that the inhibitory effect of DCG-IV on the MF IPSP was milder (26%) than the effect of L-AP4, which depressed the MF IPSP by 84% (n = 11; Fig. 4, D and E), coinciding with previously reported data (Gutiérrez 2000, 2002; Walker et al. 2001). This effect is selective for the MF-evoked IPSP because the inhibitory potentials evoked by direct stimulation of intrinsic interneurons of area CA3 were not affected (n = 6; Fig. 4D). Moreover, we found that L-AP4 perfused during the maximal effect of DCG-IV further inhibited the MF IPSP (Fig. 4F). Thereafter, the effect of L-AP4 was tested in the GABAergic synaptic responses in the first and last cells recorded from each slice (we usually recorded 3–4 cells per slice). L-AP4 inhibited the MF-IPSP to the same extent in 44 of 44 cells tested.
Because L-AP4 has the ability to depress the MF-evoked IPSP, therefore in the presence of glutamate receptors' antagonists, it should depress the IPSP and would uncover the underlying EPSPm, provided that the latter is not sensitive to L-AP4. As shown in Fig. 4C, L-AP4 strongly inhibits the IPSP without altering the EPSPm (n = 4). On the other hand, perfusion of L-AP4 had no effect on the isolated mEPSP (not shown).
To study the presynaptic modulation of the different cholinergic responses of pyramidal cells evoked by DG and A/C stimulation, EPSPm were evoked successively in the same pyramidal cells by alternating the site of stimulation (as depicted in Fig. 3B) in the presence of glutamatergic and GABAergic antagonists (n = 93). Also, the GABAB-dependent IPSP was blocked with CGP55845A (n = 12). Under these conditions, we found that DCG-IV virtually blocked the DG-but not the A/C-evoked cholinergic responses, which remained unaltered (n = 5; Fig. 5, A and B). The responses to a paired pulse potentiation protocol showed that the effects are consistent with a presynaptic action of DCG-IV (Fig. 5B). Thus the activation of mGluR-II, but not of mGluR-III, inhibits the DG-but not the A/C-driven EPSPm. Due to this, one can think that MF could also be mediating the cholinergic potential. However, because, unlike extrinsic cholinergic fibers, MF do not express M2 receptors, we tested the effect of the M2 agonist, oxotremorine, on the EPSPm evoked by stimulation of both pathways. Oxotremorine completely and reversibly depressed both EPSPm (n = 5; Fig. 5, C and D), showing that the cholinergic transmission from the DG to CA3 is not of MF origin and, contrary to the A/C cholinergic responses, is modulated by mGluR-II. For comparison purposes, we assessed the effect of oxotremorine on the fast glutamatergic DG-evoked EPSP of control preparations and confirmed its lack of effect (n = 5; Fig. 5D).
|
Emergence of nonglutamatergic signaling from the DG to CA3 after seizures depends on protein synthesis
Because the induction of the MF IPSP in vitro depends on protein synthesis (Gutiérrez 2002), we tested the effect of cycloheximide and anisomycin, protein synthesis blockers, on the seizure-induced nonglutamatergic signaling from the DG to CA3. After preincubation of the slices from seizing-rats in the protein synthesis blockers (see METHODS, synaptic potentials were evoked by DG and direct interneuron stimulation in pyramidal cells (9 for each inhibitor) and interneurons (n = 3; anisomycin). The perfusion of glutamate receptors' antagonists blocked all synaptic responses evoked by DG but not by direct stimulation of local interneurons in all the pyramidal cells and interneurons tested (Fig. 6). For comparison purposes, we carried out the same experiments in slices from control nontreated rats and from slices prepared from PTZ-treated rats 3 h after the onset of the seizure. In the former, EPSP/IPSP sequences could normally be evoked by DG stimulation. The perfusion of glutamate receptors' antagonists blocked all synaptic potentials evoked by DG stimulation, whereas an IPSP was evoked by direct interneuron stimulation, as previously shown in control nontreated slices (n = 5 for each inhibitor). On the other hand, in slices prepared 3 h after the seizure and then incubated in the blockers, nonglutamatergic transmission from DG to CA3 could still be evoked (not shown).
|
ChAT-positive fibers pass through the DG and along MF in CA3
After anatomically identifying the MF with topical application of dextran, fluorescein into the DG (Fig. 7, A and B), we carried out immunofluorescence staining to ChAT (Fig. 7, C–E). Control experiments, in which slices were processed in the absence of primary antibody, did not present immunoreactivity (not shown). We were able to discard the colocalization of ChAT and dextran, fluorescein in the granule cells or their MF both in control and PTZ-treated preparations. The perisomatic region of pyramidal cells and granule cells was densely immunoreactive to ChAT, where ChAT positive terminals were evident (Fig. 7C, 1 and 2). Also, fine fibers could be localized in stratum lucidum (Fig. 7C2). In addition, we found few cholinergic interneurons in the molecular layer of the DG (Fig. 7D1), whose orientation and terminals were restricted to the DG. Fibers positive to ChAT could be traced passing through the molecular (Fig. 7D1) and granular cell layers (Fig. 7D2). We found no apparent differences in ChAT staining between the control (Fig. 7, C1 and D1) and PTZ (Fig. 7, C2 and D2) groups.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
Seizures induce simultaneous glutamatergic and GABAergic transmission from MF to its targets in area CA3
Supporting the hypothesis that MF are able to release GABA, we show that the different targets of MF in area CA3 respond with simultaneous glutamatergic and GABAergic potentials. In control preparations, glutamate receptors' antagonists block all synaptic responses to DG stimulation both in pyramidal cells and interneurons, indicating that GABAergic potentials are disynaptically mediated. However, slices of the PTZ-treated group presented MF-evoked GABAergic responses in a condition in which all excitatory components are blocked, discarding polysynaptic contamination (Sik et al. 1994; Weisskopf and Nicoll 1995). However, there is evidence of the presence of GABAergic inputs to CA3, other than intrinsic interneurons, and of interneurons within area CA3c that can reach the hilar region (Sik et al. 1997). Although activation of these inputs cannot be entirely ruled out, the possibility that the GABAergic responses seen in our experiments are of interneuronal origin seems unlikely for three main reasons. 1) Stimulation of the hilus (Gutiérrez and Heinemann 2001) and of the granule and molecular layers of the DG (Gutiérrez 2000, 2002; Gutiérrez and Heinemann 2001) of slices of healthy nontreated animals does not produce GABAergic responses on pyramidal cells and interneurons of areas CA3a,b; however, after seizures or LTP-like stimulation, these responses appear. 2) The projections of the interneurons with CA3 targets are really CA3 interneurons, restricted to CA3c. This type of neuron, described by Sik et al. (1997) is in CA3 s. oriens and only few dendrites project to the hilus proper. 3) In slices from nontreated animals, after completely blocking synaptic responses by the perfusion of glutamate receptors' antagonists, LTP-like stimulation of the DG induces the emergence of GABAergic responses. These are sensitive to mGluR activation and to protein synthesis blockade (Gutiérrez 2002), whereas GABAergic responses evoked by direct interneuron stimulation are not. This evidence strongly supports the notion of an activity-dependent mechanism and discards the possibility of stimulating interneurons projecting from the molecular layer of the DG to as far as CA3a,b.
Accordingly, the control EPSP and the pharmacologically isolated MF-evoked IPSP had the same latency, as previously reported (Gutiérrez 2000, 2002; Gutiérrez and Heinemann 2001; Walker et al. 2001). The activation of mGluR reduces neurotransmitter output from MF (Manzoni et al. 1995; Min et al. 1998; Salin et al. 1996) but has no effects on GABAergic transmission from interneurons in the DG (Tong et al. 1996) and in CA3 (this work; Walker et al. 2001). Thus that the latency of the IPSP isolated by the perfusion of glutamate receptors' antagonists is not affected in the presence of mGluR agonists, or of baclofen, a GABAB agonist known to inhibit neurotransmitter release from MF and interneurons (Brown and Johnston 1983; Gutiérrez 2002), and even when perfusing them in a low-Ca2+ medium (Gutiérrez 2000) provide strong evidence that our recordings represent monosynaptic IPSPs of MF origin. Further support to the hypothesis that MF can release GABA is provided from works showing that GAD (Ramírez and Gutiérrez 2001; Schwarzer and Sperk 1995; Sloviter et al. 1996), GABA (Gómez-Lira et al. 2002), and VGAT mRNA (Lamas et al. 2001) are expressed in the granule cells and their MF in an activity-dependent fashion. This is further confirmed by our experiments showing that protein synthesis is required to trigger the MF GABA-releasing process (see also Gutiérrez 2002). However, the cellular mechanisms by which the blockade of protein synthesis disrupts the expression of nonglutamatergic signaling cannot be inferred from our present results.
Walker et al. (2001) have found this MF inhibitory transmission to be normally present in young guinea pig slices. Although marked differences are known to exist in regard to MF physiology in the two species, we have recently determined (R. Gutiérrez, H. Romo-Parra, J. Maqueda, M. Ramírez,
M. A. Morales and M. Lamas, unpublished results) that MF monosynaptic GABAergic signals can be normally detected in juvenile rats. Therefore the release of GABA from MF in the rat has possibly been previously observed (Ben-Ari et al. 1989) but not identified as such. In the light of our previous (Gutiérrez 2002) and present results, MF-GABAergic transmission can be recognized by its sensitivity to mGluR-III activation. It has been considered that MF glutamatergic neurotransmission in the rat is sensitive to the mGluR-II agonist, DCG-IV, but not to the mGluR-III agonist, L-AP4. The contrary is true for MF neurotransmission in the guinea pig (Bradley et al. 1996; Lanthorn et al. 1984; Shigemoto et al. 1997; but see Ohishi et al. 1993, 1995). That MF GABAergic signaling is strongly inhibited by the activation of the mGluR-III permits to suggest a link of mGluR-III to the GABA-releasing machinery and of mGluR-II to the glutamate releasing machinery, which has been shown to produce a downregulation of the exocytotic machinery (Kamiya and Ozawa 1999). Therefore not only an inhibition of Ca2+ entrance explains inhibition of neurotransmitter release in MF as has been suggested (Vogt and Regehr 2001).
The granule cells and MF of the rat DG express groups II/III mGluR mRNA (Ohishi et al. 1993, 1995) and mGluR2,4,7 (Bradley et al. 1996; Lie et al. 2000; Shigemoto et al. 1997). Electron microscopy has revealed immunolabeling for the mGluR-III predominantly in presynaptic active zones of asymmetrical and symmetrical synapses, whereas mGluR-II immunolabeling was found in preterminal rather than terminal portions of axons (Shigemoto et al. 1996,1997). Our data suggest a presynaptic segregation of mGluR receptors according to the class of neurotransmitter to be released. Postsynaptically, GABAA receptors cluster apposed to glutamatergic terminals in cultured pyramidal hippocampal cells suggesting that there is an element common to GABA and glutamate synapses (Rao et al. 2000).
Seizures induce mGluR-sensitive cholinergic transmission from the DG to pyramidal cells and interneurons in area CA3
Cholinergic responses of pyramidal cells and interneurons of the hippocampus usually appear on high-frequency or -intensity stimulation of their cholinergic afferents (Brunner and Misgeld 1994; Cole and Nicoll 1984; Morton and Davies 1997; Morton et al. 2001; Müller and Misgeld 1986), whereas single-pulse stimulation over the DG in the presence of glutamate receptors' antagonists does not elicit synaptic responses in control preparations. However, after seizures, single-pulse stimulation over DG evokes M1-AChR-dependent responses, suggesting that these fibers are probably silent in normal conditions, and they become transiently active after seizures (Gutiérrez and Heinemann 2001). We now establish that this is a process dependent on protein synthesis.
It has been described that the cholinergic input from the septum exerts its modulation in the MF pathway through the activation of interneurons, which in turn affect pyramidal cells (Martin and Alger 1999; Williams and Johnston 1990; Vogt and Regehr 2001). Our data show that M1-cholinergic signaling of DG origin directly activates both pyramidal cells and interneurons within CA3. This, contrary to the A/C evoked-cholinergic responses, is sensitive to activation of mGluR-II, whereas M2-AChR activation depresses ACh release from both pathways. It has been described that activation of mGluR-II depresses MF but not A/C synapses (Manzoni et al. 1995) and muscarine directly inhibits the A/C synapse without directly affecting the MF synapse (Vogt and Regehr 2001), which does not have M2 receptors (Hájos et al. 1998), as we also confirm.
The cholinergic activation of different subsets of neurons and/or interneurons and, as we show, the activation of different cholinergic pathways with differential presynaptic modulation, regulate the overall synaptic integration in the hippocampus (Hasselmo and Fehlau 2001). Thus in PTZ-treated preparations, the DG-cholinergic signal postsynaptically modulates the convergent glutamatergic and GABAergic neurotransmission. In this way, this adds to other modulatory mechanisms in a pathway where spill over is a means by which collateral multi-target modulation is exerted together with pre- and postsynaptic interactions in a parallel fashion (Hájos et al. 1998; Min et al. 1998; Rusakov et al. 1999; Semyanov and Kullmann 2000; Vogt and Regehr 2001).
Anatomical findings
Our immunohistological data discard the presence of ChAT in the MF and disclose a pathway that probably runs in parallel to this projection. As previously shown (Frotscher and Léránth 1985), we found ChAT-positive fibers to form a network around granule cells in the DG and pyramidal cells and interneurons in CA3. The terminals observed in the CA3 region are likely to be collaterals of those innervating the DG. In accordance with previous reports (Blaker et al. 1988; Frotscher et al. 2000), we observe the presence of few cholinergic interneurons, whose arborizations terminate locally within the s. moleculare of the DG. Additionally, Frotscher et al. (2000) showed that neither GAD65 nor GAD67 mRNA colocalizes with ChAT in cholinergic interneurons.
Functional implications
Our present data showing that MF glutamatergic and GABAergic neurotransmission can be observed both in pyramidal cells and interneurons and that MF GABAergic signaling has a distinctive presynaptic modulation suggest that distinct MF fiber collaterals are involved in MF GABAergic neurotransmission (Gutiérrez 2002). Its emergence has net inhibitory effects on pyramidal cells because the activation of MF at high frequencies provokes the more pronounced IPSPs in pyramidal cells of PTZ-treated rats to summate, hyperpolarizing the cells (Gutiérrez and Heinemann 2001). Although here we prove that MF GABAergic transmission is also expressed onto interneurons, which are the main targets of the MF (Acsády et al. 1998), we show that MF glutamatergic excitation surpasses the inhibitory transmission. Thus the interneurons are more readily excited than pyramidal cells, which continue to receive a strong inhibitory control from the interneurons within CA3, besides the MF GABAergic signal.
Beyond the attractive hypothesis of glutamate/GABA corelease from the MF, other important functional consequences arise. Our model (Fig. 8) proposes that selective activation of mGluR-II would inhibit excitatory transmission (glutamatergic from the MF, and cholinergic from the associated parallel fibers) but not MF inhibitory transmission, which in turn, can still exert presynaptic GABAB-R activation and postsynaptically act on pyramidal cells and interneurons. Contrariwise, activation of mGluR-III would selectively inhibit inhibitory transmission without consequences in excitatory transmission, which could still further activate presynaptic mGluR. Finally, short- and long-term changes in the cholinergic system due to epileptic and stress-related phenomena are known to exist (Ferencz et al. 2001; Friedman et al. 1996; Kaufer et al. 1998; Lupica and Berman 1988; Mingo et al. 1997, 1998; Serra et al. 1997), whether these are related phenomena needs to be disclosed.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
This work was supported in part by Consejo Nacional de Ciencia y Tecnología, México, Grants 29309-N and 36178-N to R. Gutiérrez.
|
FOOTNOTES |
|---|
* H. Romo-Parra and C. Vivar contributed equally to this work. 
Address for reprint requests: R. Gutiérrez, Departamento de Fisiología, Biofísica y Neurociencias, Centro de Investigación y Estudios Avanzados del IPN, Apartado Postal 14-740, México D.F. 07000 (E-mail: grafael{at}fisio.cinvestav.mx).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Ben-Ari Y, Cherubini E, Corradetti R, and Gaiarsa JL. Giant synaptic potentials in immature rat CA3 hippocampal neurons. J Physiol 416: 303–325, 1989.
Blaker SN, Armstrong DM, and Gage FH. Cholinergic neurons within the rat hippocampus: response to fimbria-fornix transection. J Comp Neurol 272: 127–138, 1988.[ISI][Medline]
Bradley SR, Levey AI, Hersch SM, and Conn PJ. Immunocytochemical localization of group III metabotropic glutamate receptors in the hippocampus with subtype-specific antibodies. J Neurosci 16: 2044–2056, 1996.
Brown TH and Johnston D. Voltage-clamp analysis of mossy fiber synaptic input to hippocampal neurons. J Neurophysiol 50: 487–507, 1983.
Brunner H and Misgeld U. Muscarinic amplification of fast excitation in hilar neurons and inhibition in granule cells in the guinea pig hippocampus. J Physiol 480: 513–526, 1994.[ISI]
Buzsáki G. Feed-forward inhibition in the hippocampal formation. Prog Neurobiol 22: 131–153, 1984.[ISI][Medline]
Cole AE and Nicoll RA. Characterization of a slow cholinergic postsynaptic potential recorded in vitro from rat hippocampal pyramidal cells. J Physiol 352: 173–188, 1984.
Crawford IL and Connor JD. Localization and release of glutamic acid in relation to the hippocampal mossy fiber pathway. Nature 244: 442–443, 1973.[Medline]
Dichter M and Spencer WA. Penicillin-induced interictal discharges from the cat hippocampus. II. Mechanisms underlying origin and restriction. J Neurophysiol 32: 663–687, 1969.
Ferencz I, Leanza G, Nanobashvili A, Kokaia Z, Kokaia M, and Lindvall O. Septal cholinergic neurons suppress seizure development in hippocampal kindling in rats: comparison with noradrenergic neurons. Neuroscience 102: 819–832, 2001.[ISI][Medline]
Friedman A, Kaufer D, Shemer J, Hendler I, Soreq H, and Tur-Kaspa I. Pyridostigmine brain penetration under stress enhances neuronal excitability and induces early immediate transcriptional response. Nat Med 2: 1382–1385, 1996.[ISI][Medline]
Frotscher M and Léránth C. Cholinergic innervation of the rat hippocampus as revealed by choline acetyltransferase immunocytochemistry: a combined light and electron microscopic study. J Comp Neurol 239: 237–246, 1985.[ISI][Medline]
Frotscher M, Vida I, and Bender R. Evidence for the existence of non-GABAergic, cholinergic interneurons in the rodent hippocampus. Neuroscience 96: 27–31, 2000.[ISI][Medline]
Gómez-Lira G, Trillo E, Ramírez M, Asai M, Sitges M, and Gutiérrez R. Expression of GABAergic transmission in the mossy fiber synapse coincides with the seizure-induced expression of GABA in mossy fiber synaptosomes. Exp Neurol 177: 276–283, 2002.[ISI][Medline]
Gutiérrez R. Seizures induce simultaneous GABAergic and glutamatergic neurotransmission in the dentate gyrus-CA3 system. J Neurophysiol 84: 3088–3090, 2000.
Gutiérrez R. Activity-dependent expression of simultaneous glutamatergic and GABAergic neurotransmission from the mossy fibers in vitro. J Neurophysiol 87: 2562–2570, 2002.
Gutiérrez R and Heinemann U. Kindling induces transient fast inhibition in the dentate gyrus–CA3 projection. Eur J Neurosci 13: 1371–1379, 2001.[ISI][Medline]
Hájos N, Papp ECS, Acsády L, Levey A,I and Freund TF. Distinct interneuron types express M2 muscarinic receptor immunoreactivity on their dendrites or axon terminals in the hippocampus. Neuroscience 82: 355–376, 1998.[ISI][Medline]
Hasselmo ME and Fehlau BP. Differences in time course of ACh and GABA modulation of excitatory synaptic potentials in slices of rat hippocampus. J Neurophysiol 86: 1792–1802, 2001.
Jones LS, Grooms SY, Lapadula DM, and Lewis DV. Protein synthesis inhibition blocks maintenance but not induction of epileptogenesis in hippocampal slice. Brain Res 599: 338–344, 1992.[ISI][Medline]
Kamiya H and Ozawa S. Dual mechanism for presynaptic modulation by axonal metabotropic glutamate receptor at the mouse mossy fiber-CA3 synapse. J Physiol 518: 497–506, 1999.
Kaufer D, Friedman A, Seidman S, and Soreq H. Acute stress facilitates long-lasting changes in cholinergic gene expression. Nature 28: 373–377, 1998.
Lamas M, Gómez-Lira G, and Gutiérrez R. Vesicular GABA transporter mRNA expression in the dentate gyrus and in mossy fiber synaptosomes. Mol Brain Res 93: 209–214, 2001.[Medline]
Lanthorn TH, Ganong AH, and Cotman CW. 2-Amino-4-phosphonobutyrate selectively blocks mossy fiber-CA3 responses in guinea pig but not rat hippocampal neurons. Neuron 4: 741–749, 1984.
Lie AA, Becker A, Behle K, Beck H, Malitschek B, Conn PJ, Kuhn R, Nitsch R, Plaschke M, Schramm J, Elger CE, Wiestler OD, and Blumcke I. Upregulation of the metabotropic glutamate receptor mGluR4 in hippocampal neurons with reduced seizure vulnerability. Ann Neurol 47: 26–35, 2000.[ISI][Medline]
Lupica CR and Berman RF. Atropine slows olfactory bulb kindling while diminished cholinergic innervation does not. Brain Res Bull 20: 203–209, 1988.[ISI][Medline]
Maccaferri G, Tóth K, and McBain CJ. Target-specific expression of presynaptic mossy fiber plasticity. Science 279: 1368–1370, 1998.
Manzoni O, Castillo PE, and Nicoll RA. Pharmacology of metabotropic glutamate receptors at the mossy fiber synapses of the guinea pig hippocampus. Neuropharmacology 34: 965–971, 1995.[ISI][Medline]
Martin LA and Alger BE. Muscarinic facilitation of the occurrence of depolarization-induced suppression of inhibition in rat hippocampus. Neuroscience 92: 61–71, 1999.[ISI][Medline]
Miles R,Tóth K, Gulyáas AI, Hájos N, and Freund TF. Differences between somatic and dendritic inhibition in the hippocampus. Neuron 16: 815–823, 1996.[ISI][Medline]
Min MY, Rusakov DA, and Kullmann DM. Activation of AMPA, kainate, and metabotropic receptors at hippocampal mossy fiber synapses: role of glutamate diffusion. Neuron 21: 561–570, 1998.[ISI][Medline]
Mingo NS, Cottrell GA, Mendonca A, Gombos Z, Eubanks JH, and Burnham WM. Amygdala-kindled and electroconvulsive seizures alter hippocampal expression of the m1 and m3 muscarinic cholinergic receptor genes. Brain Res 810: 9–15, 1998.[ISI][Medline]
Mingo NS, Cottrell GA, Zhang L, Wallace MC, Burnham WM, and Eubanks JH. Kainic acid-induced generalized seizures alter the regional hippocampal expression of the rat m1 and m3 muscarinic acetylcholine receptor genes. Epilepsy Res 29: 71–79, 1997.[ISI][Medline]
Morton RA and Davies CH. Regulation of muscarinic acetylcholine receptor-mediated synaptic responses by adenosine receptors in the rat hippocampus. J Physiol 502: 75–90, 1997.[ISI][Medline]
Morton RA, Manuel NA, Bulters DO, Cobb SR, and Davies CH. Regulation of muscarinic acetylcholine receptor-mediated synaptic responses by GABAB receptors in the rat hippocampus. J Physiol 535: 757–766, 2001.
Müller W and Misgeld U. Slow cholinergic excitation of guinea pig hippocampal neurons is mediated by two muscarinic receptor subtypes. Neurosci Lett 18: 101–112, 1986.
Ohishi H, Akazawa C, Shigemoto R, Nakanishi S, and Mizuno N. Distributions of the mRNAs for L-2-amino-4-phosphonobutyrate-sensitive metabotropic glutamate receptors, mGluR4 and mGluR7, in the rat brain. J Comp Neurol 360: 555–570, 1995.[ISI][Medline]
Ohishi H, Shigemoto R, Nakanishi S, and Mizuno N. Distribution of the mRNA for a metabotropic glutamate receptor (mGluR3) in the rat brain: an in situ hybridization study. J Comp Neurol 335: 252–66, 1993.[ISI][Medline]
Otani S, Roisin-Lallemand MP, and Ben-Ari Y. Enhancement of extracellular protein concentrations during long-term potentiation in the rat hippocampal slice. Neuroscience 47: 265–272, 1992.[ISI][Medline]
Rao A, Cha EM, and Craig AM. Mismatched appositions of presynaptic and postsynaptic components in isolated hippocampal neurons. J Neurosci 20: 8344–8353, 2000.
Ramírez M and Gutiérrez R. Activity-dependent expression of GAD67 in the granule cells of the rat hippocampus. Brain Res 917: 139–146, 2001.[ISI][Medline]
Rusakov DA, Kullmann DM, and Stewart MG. Hippocampal synapses: do they talk to their neighbors? Trends Neurosci 22: 382–388, 1999.[ISI][Medline]
Salin PA, Scanziani M, Malenka RC, and Nicoll RA. Distinct short-term plasticity at two excitatory synapses in the hippocampus. Proc Natl Acad Sci USA 93: 13304–13309, 1996.
Scharfman H. Electrophysiological diversity of pyramidal-shaped neurons at the granule cell/hilus border of the rat dentate gyrus recorded in vitro. Hippocampus 5: 287–305, 1995.[ISI][Medline]
Schwarzer C and Sperk G. Hippocampal granule cells express glutamic acid decarboxilase 67 after limbic seizure in the rat. Neuroscience 69: 705–709, 1995.[ISI][Medline]
Semyanov A and Kullmann DM. Modulation of GABAergic signaling among interneurons by metabotropic glutamate receptors. Neuron 25: 663–672, 2000.[ISI][Medline]
Serra M, Dazzi L, Cagetti E, Chessa MF, Pisu MG, Sanna A, and Biggio G. Effect of pentylenetetrazole-induced kindling on acetylcholine release in the hippocampus of freely moving rats. J Neurochem 68: 313–318, 1997.[ISI][Medline]
Shigemoto R, Kinoshita A, Wada E, Nomura S, Ohishi H, Takada M, Flor PJ, Neki A, Abe T, Nakanishi S, and Mizuno N. Differential presynaptic localization of metabotropic glutamate receptor subtypes in the rat hippocampus. J Neurosci 17: 7503–7522, 1997.
, the regions depicted in the next panels. B: MF in the stratum lucidum of CA3 stained with dextran, fluorescein at 2 magnifications. C1: in this preparation of a control rat, 
show MF and their thorny spines (in green) surrounded by a plexus of ChAT-positive terminals and fibers (in red) in s. lucidum of CA3. No colocalization of the dextran, fluorescein and ChAT-positive signals was observed. C2: some (ChAT-positive) cholinergic fibers could be followed (