INTRODUCTION

The excessive glutamate release during brain anoxia or ischemia triggers the death of neurons, causing mental and physical handicaps (Choi and Rothman, 1990; Nishizawa, 2001). The hippocampal CA1 region is highly vulnerable to ischemia, which induces a selective and delayed degeneration of pyramidal neurons. This phenomenon is called delayed neuronal cell death (DND) (Kirino, 1982; Szatokowski and Attwell, 1994). During ischemia, glutamate is released by a reversal operation of glutamate uptake carriers at CA1 pyramidal neurons (Rossi et al., 2000). This glutamate release induces a rise in intracellular Ca²⁺ concentration ([Ca²⁺]_i) in CA1 pyramidal neurons in an NMDA-receptor dependent manner (Silver and Erecinska, 1990). Following the ischemia, [Ca²⁺]_i almost returns to its basal level within 5 min, whereas the neurons exposed to excessive glutamate die within a few hr (Silver and Erecinska, 1992).

Ischemia also generates a novel form of long-term potentiation (LTP), which is called anoxic long-term potentiation (A-LTP) (Hammond et al., 1994). The synaptic mechanisms underlying A-LTP in the hippocampal CA1 regions have been considered to be due to an enhancement of pre-synaptic glutamate release (Hammond et al., 1994). The enhancement of glutamate release leads to DND coupled with postsynaptic NMDA receptor activation (Hammond et al., 1994). This postsynaptic event is defined, but the pre-synaptic ion channel modulation mechanisms underlying the excessive glutamate release were still unknown. In the present study, we applied a combined technique of electro-physiological and infra-red differential interference contrast (IR-DIC) procedures to hippocampal slices in order to determine a presynaptic ionic basis of A-LTP, which induces necrosis like neuronal cell death.

MATERIALS AND METHODS

Hippocampal slices (300 μm) were prepared from 4-week-old Wistar rats using standard procedures (Takagi et al., 1992a, b; Watanabe et al., 1998). Animal care and experimental procedures were performed according to the guidelines of Shinshu University. Briefly, 300 μm thick hippocampal slices were prepared using a microslicer (DTK-1500, Dosaka-EM, Japan) in a submerged holding chamber containing ice-cold standard extracellular solution (sES) (mM: 124 NaCl, 2.5 KCl, 1.0 NaH₂PO₄, 26 NaHCO₃, 2.0 CaCl₂, 1.0 MgCl₂ and 10 dextrose, bubbled with 95% O₂/5% CO₂), and then stored at room temperature for the experiments (<6 hr). After 60 min of pre-incubation, the slices were mounted in a recording chamber and perfused with an external solution (32°C) at a constant rate of about 3–4 ml/min. The field excitatory postsynaptic potential (fEPSP) was recorded from the stratum radiatum in the CA1 region by a glass microelectrode filled with extracellular solution (tip diameter, about 1 μm; DC resistance, 4–8 MΩ). We delivered the orthodromic stimuli of 100 μs square pulses at 0.03 Hz frequency to the orientation of the tips in the CA3 region (the Schaffer collateral) throughout the experiment using a monopolar glass electrode filled with extracellular solution. We set the stimulus intensity (5–50 V) in order to evoke a value of 40–60% of the maximum fEPSP slope. Electrophysiological recordings were performed through a patch clamp amplifier head stage (EPC-7, HEKA, Germany) interfaced with a personal computer (Aptiva J33, IBM, Japan), and the data were acquired, digitized, stored and analyzed using pCLAMP software (ver. 6.0.2, Axon Instruments, CA, USA). A microscope (Axio-scope FS2, Carl Zeiss, Germany) equipped with a x40 water immersion lens and a CCD camera (C2400-50, Hamamatsu Photonics, Japan) was used to monitor morphological changes in the CA1 pyramidal neurons in hippocampal slices using an IR-DIC procedure.

Morphological images were recorded on a VCR, and these images were digitized and stored on a personal computer (Dell V333c, Dell, Japan) through a video capture cable system (USBCAP, I.O data, Japan). Stored morphological images were contrasted by Photoshop software (ver. 4, Adobe Systems, Japan).

Ischemia for hippocampal slices was applied for 5 min by the replacement of extracellular solution from sES to ischemia extracellular solution (iES) (mM: 124 NaCl, 2.5 KCl, 1.0 NaH₂PO₄, 26 NaHCO₃, 2.0 CaCl₂, 1.0 MgCl₂ and 10 Deoxy-glucose, bubbled with 95% N₂/5% CO₂) at 10 min after the onset of experiment. Here, A-LTP was induced by the replacement from O₂ to N₂ and from dextrose to deoxy-glucose. To establish a real time monitoring system for ischemia-induced DND, we monitored fEPSP slope and paired pulse facilitation (PPF) ratio for about 3 hr after ischemia. Throughout the experiment, morphological changes in the CA1 pyramidal neurons in hippocampal slices were monitored for 1 min at 0, 10, 15, 20, 30, 45, 60, 75, 90, 105, 120, 135, 150, 165, 180, 205 and 210 min after the onset of experiment.

The data were expressed as mean±SE. Statistical significance was examined by Student t-test.

RESULTS

After replacement of sES to iES, the fEPSP slope changed in 3 phases (Fig. 1A). After the replacement, the fEPSP slope suddenly decreased to zero within 3 min (Phase 1). The fEPSP slope gradually recovered within 5-10 min after ischemia, and then significantly increased to 2.04±0.89 (n=3) compared to the control level that was normalized to 1 (p<0.001) (Phase 2). The ischemia-induced synaptic potentiation (ISP) lasted for 30–60 min. Then the fEPSP slope gradually decreased to a lower level (0.81± 0.06, p<0.005) than the control level within 135–145 min after ischemia (Phase 3). Furthermore, the pyramidal neurons became swelling at 75 min after ischemia (Fig. 1B). The plasma membranes of the pyramidal neurons became unclear at 195 min after ischemia (Fig. 1B). In no ischemic group, no morphological changes of the pyramidal neurons appeared even at 195 min after ischemia (Fig. 1B). Therefore, these neuronal damages were caused not by artificial infra-red effects, but mainly by ischemia.

Fig. 1

Ischemia-induced fEPSP slope and morphological changes in pyramidal neurons in the hippocampal CA1 region. (A) EPSP traces recorded from CA1 regions are expressed before ischemia (control), during ischemia, and at 60, 85, 160 and 195 min after ischemia (upper column). The time course of the ischemia-induced fEPSP slope change is shown (lower column). 60, 85, 160 and 195 min after ischemia in upper column correspond to 75, 100, 175 and 210 min in lower column. Data are expressed as mean± SE. The control level was normalized to 1. The period of ischemia is expressed as a solid black bar. (B) Morphological images of hippocampal CA1 pyramidal neurons through IR-DIC are expressed at the control condition, 75 min and 195 min after ischemia (upper column: ischemic group; lower column: no ischemic group) (bar: 10 μm). 75 min and 195 min after ischemia correspond to 90 min and 210 min in lower column of (A). Dashed lines indicate the outline changes of the pyramidal neurons.

The ISP was observed in each of experiments (n=3) under the normal extracellular CaCl₂ concentration (2 mM) condition (Fig. 2A). The PPF ratio also changed in every experiment (n=6). During ischemia, the PPF ratio elevated suddenly (Fig. 2B). Then the PPF ratio gradually and significantly decreased after ischemia, and remained at a significantly lower level (1.31±0.003, n=6) compared to the basal (control) PPF ratio level (1.58±0.57, n=6, p<0.001) (Fig. 2B). We thus concluded that the ISP resulted from presynaptic modulation and that this modulation of the presynaptic releasing site was induced in phases 1 and 2.

Fig. 2

Ischemia-induced changes of fEPSP slope and PPF ratio. (A and B) Time-dependent changes of fEPSP slope and PPF ratio of the normal condition ([Ca²⁺]_o=2 mM) group, (C and D) time-dependent changes of fEPSP slope and PPF ratio of the higher concentration Ca²⁺ ([Ca²⁺]_o= 10 mM) applied group, and (E and F) time-dependent changes of fEPSP slope and PPF ratio of the 2 mM 4-aminopyridine (4-AP) added group are shown. Data are expressed as mean±SE. In each case, the control level of fEPSP was normalized to 1. The period of ischemia is expressed as a solid black bar.

To elucidate the presynaptic modulation mechanisms in phases 1 and 2, possible contributions of K⁺ channels, Ca²⁺ channels and releasing machineries in the presynaptic sites were studied. The ISP was not observed under a higher extracellular CaCl₂ concentration (10 mM) condition (n=4) (Fig. 2C). The decline of the fEPSPs slope during ischemia (Phase 1) was not observed, but the fEPSP slope slightly increased under the higher extracellular CaCl₂ condition. During and after ischemia, the PPF ratio did not change under the higher extracellular CaCl₂ condition. Furthermore, the basal (control) PPF ratio was significantly lower (0.86±0.12, n=4) (Fig. 2D) than the normal condition (1.58±0.57, n=6, p<0.001) (Fig. 2B). In addition, 2 mM 4-aminopyridine (4-AP), an A- and D-type K⁺ channel-selective blocker (Storm, 1990; Takagi et al., 1998; Takagi et al., 2000), inhibited the ISP induction (n=4) (Fig. 2E). After ischemia, the PPF ratio did not change under the 2 mM 4-AP condition (n=4) (Fig. 2F). The basal (control) PPF ratio under the 2 mM 4-AP condition (0.57±0.01, n=4) was significantly lower than the normal condition (1.58±0.57, n=6, p< 0.001).

DISCUSSION

Neurotransmitter release is triggered by membrane depolarization and regulated by sequential Ca²⁺ influx which is dependent on the concentration gradient of Ca²⁺ between the extracellular space and the intracellular space (Llinás et al., 1976). 4-AP sensitive K⁺ channels and Ca²⁺-dependent K⁺ channels have been reported to regulate neurotransmitter release at presynaptic nerve terminals (Weeler et al., 1996; Hu et al., 2001). The present results suggest that the ISP results from 4-AP sensitive presynaptic K⁺ channel modulation (maybe ‘suppression’), but not from the modulation of presynaptic Ca²⁺ channels and Ca²⁺-dependent K⁺ channels, causing a more excitable state in presynaptic nerve terminals than that observed under no ischemic conditions.

In this study, we obtained important new findings. (1) The ISP was not observed under a higher extracellular CaCl₂ concentration (10 mM), where the Ca²⁺ influx through presynaptic Ca²⁺ channels and Ca²⁺-dependent K⁺ channels was enhanced. (2) The 4-AP, an A- and D-type K⁺ channel-selective blocker, also inhibited the ISP induction. These results suggest that the modulation of 4-AP sensitive presynaptic K⁺ channels plays a crucial role in the ISP induction. Neurotransmitter overdose results in an over-excitation of postsynaptic membrane, detected as ISP, and then induces cell damage or cell death. The absence of ischemia-induced fEPSP decline under the higher CaCl₂ concentration does not indicate a protective procedure against neuronal cell damage, but it rather means the pre-synaptic membrane is controlled at a more excitable state than under the normal condition due to higher CaCl₂ pre-treatment before ischemia (Llinás et al., 1976). The suppression of ISP under the higher CaCl₂ concentration proves that the amplitude of fEPSP is controlled at a relatively higher level than under the normal condition in the experiments. Furthermore, the ISP is produced only under the condition where synapses can work more efficiently.

We established the best in vitro real-time monitoring system for acute neuronal cell death for the first time, and we also found that 4-AP sensitive presynaptic K⁺ channel modulation (maybe ‘suppression’) is the crucial mechanism of neuronal cell damage and/or death. Our results will provide useful therapeutic procedures against acute neuronal cell death if the modulation of 4-AP sensitive presynaptic K⁺ channels can be reset to normal, since 4-AP also inhibited the ISP induction. Further studies are needed as follows for the establishment of therapeutic procedures against ischemia-induced neuronal cell damage and/or death. (1) To determine whether or not a 4-AP sensitive presynaptic K⁺ channel opener (Pena and Tapia, 2000; Ortiz et al., 2002) can protect against acute neuronal cell damage. (2) To elucidate the cellular mechanisms (eg, IL-1β/nitric oxide system, MAP kinase system, Estrogen system) (Togashi et al., 2001; Nozaki et al., 2001; Jover et al., 2002) of ischemia-induced modulation of 4-AP sensitive presynaptic K⁺ channels. (3) To elucidate the time and spatial dependent correlation between ISP and DND.