It is known that GABA, a major inhibitory transmitter in the CNS, acts as an excitatory (or depolarizing) transmitter transiently after intense GABAA receptor activation in adult brains. The depolarizing effect is considered to be dependent on two GABAA receptor-permeable anions, chloride (Cl–) and bicarbonate (HCO3–). However, little is known about their spatial and temporal profiles during the GABAergic depolarization in postsynaptic neurons. In the present study, we show that the amplitude of synaptically induced depolarizing response was correlated with intracellular Cl– accumulation in the soma of mature hippocampal CA1 pyramidal cells, by using whole cell patch-clamp recording and Cl– imaging technique with a Cl– indicator 6-methoxy-N-ethylquinolinium iodide (MEQ). The synaptically activated Cl– accumulation was mediated dominantly through GABAA receptors. Basket cells, a subclass of fast-spiking interneurons innervating the somatic portion of the pyramidal cells, actually fired at high frequency during the Cl– accumulation accompanying the depolarizing responses. These results suggest synaptically activated GABAA-mediated Cl– accumulation may play a critical role in generation of an excitatory GABAergic response in the mature pyramidal cells receiving intense synaptic inputs. This may be the first demonstration of microscopic visualization of intracellular Cl– accumulation during synaptic activation.
Although GABA is generally known as a major inhibitory transmitter in the CNS, GABAergic transmission is dynamically changeable from hyperpolarizing to depolarizing in certain neuronal environments. In mature hippocampal pyramidal cells, high-frequency synaptic stimulation (tetanization) results in a slow posttetanic depolarizing response (Alger and Nicoll 1979), which is often followed by a long-lasting, seizure-like epileptic afterdischarge (Bragin et al. 1997; Stasheff et al. 1989). Several electrophysiological studies revealed that transient depolarizing GABAergic transmission plays a key role in the generation of slow posttetanic depolarization (Alger and Nicoll 1979) and also in the subsequent generation of seizure-like afterdischarge (Fujiwara-Tsukamoto et al. 2003; Velazquez and Carlen 1999). For the slow posttetanic depolarization, some theoretical models for its ionic basis have been proposed so far: GABAA reversal potential depending on preserved bicarbonate (HCO3–) and collapsed chloride (Cl–) gradients (Staley and Proctor 1999; Staley et al. 1995; but see Grover et al. 1993), activation of distinct GABAA receptors with lower Cl–/higher HCO3– permeability and slower kinetics (Perkins and Wong 1996), and nonsynaptic depolarizing effect of extracellular potassium (K+) accumulation in a HCO3– -dependent manner (Kaila et al. 1997; Smirnov et al. 1999). In all the models, two GABAA-permeable anions (Cl– and HCO3–) are thought to act as most critical factors to bring about the slow posttetanic depolarization, but it has been controversial whether intracellular Cl– accumulation is necessary for the depolarizing response.
Several fluorescent Cl– indicators have been developed to visualize Cl– dynamics in living cells: e.g., 6-methoxy-N-ethylquinolinium iodide (MEQ), N-(ethoxycarbonylmethyl)-6-methoxyquinolinium bromide (MQAE), and 6-methoxy-N-(3-sulfopropyl)quinolinium (SPQ). Fluorescence of these 6-methoxyquinolinium derivatives is quenched by their collision to diffused Cl– ions, so they would have little effect to buffer the ion's physiological functions unlike Ca2+ indicators such as fura-2. In addition, the Cl– indicators are relatively insensitive to HCO3– or pH change and quite stable in loaded cells. In recent pharmacological experiments, microscopic imaging techniques with these Cl– indicators successfully visualized intracellular Cl– responses to GABAA receptor agonists or Cl– transporter blockers (using MEQ: Fukuda et al. 1998; Inglefield and Schwartz-Bloom 1997, 1998a; Schwartz and Yu 1995; Yamada et al. 2001; MQAE: Hara et al. 1992; Marandi et al. 2002; SPQ: Frech et al. 1999). By the Cl– imaging technique with the indicator MEQ, we here examined the correlation between the slow posttetanic depolarization and the synaptically activated Cl– accumulation in single hippocampal neurons.
Hippocampal slices (400 μm thick) were prepared from halothane-anesthetized Wistar rats (P20–22) with a microslicer (VT-1000S; Leica, Nussloch, Germany), and the CA1 regions were routinely isolated from the CA3 and the subiculum regions. After the recovery for ≥1 h, each slice was transferred to a submerged-type recording chamber continuously perfused with a normal artificial cerebrospinal fluid (ACSF, 30°C), which consisted of (in mM) 124 NaCl, 2.5 KCl, 1.2 KH2PO4, 26 NaHCO3, 1.2 MgSO4, 2.5 CaCl2 and 25 d-glucose and was saturated with 95% O2-5% CO2 gas. To induce the slow posttetanic depolarization, tetanic stimulation (100 Hz for 0.5 s; intensity: 100–400 μA, duration: 20–400 μs) was delivered to the stratum radiatum by a monopolar glass stimulating electrode (0.5–1 MΩ, filled with 2.5 M NaCl) (Fujiwara-Tsukamoto et al. 2003).
Whole cell patch-clamp recordings were obtained from the CA1 pyramidal cells near the surface of hippocampal slices under visual guidance using a cooled charge-coupled device (CCD) camera (ORCA-ER C4742–95; Hamamatsu Photonics, Hamamatsu, Japan) fitted to an infrared-differential-interference-contrast (IR-DIC) up-right microscope (ECLIPSE E600FN; Nikon, Tokyo) with a ×40 water-immersion objective lens (Nikon Fluor ×40/0.80W DIC M; Nikon). In current-clamp mode, membrane potentials were recorded [resting membrane potential (r.m.p.), –67.5 ± 5.9 (SD) mV] with a patch-clamp amplifier (Axopatch 200B; Axon Instruments, Union City, CA), through glass patch electrodes filled with an internal solution containing (in mM) 140 K-methanesulfonate, 2 KCl, 1 MgCl2, 10 HEPES-NaOH, 0.2 EGTA, 0.2 Mg(ATP)2, 0.04 GTP, and 0.1 MEQ (pH 7.4; 5–10 MΩ). Recorded signals were low-pass-filtered at 1–2 kHz, digitized at 1 kHz with an A/D interface (Digidata 1322A; Axon Instruments), and stored in a Windows PC using the Axon Clampex 8.2. In some experiments, membrane potentials were recorded from basket cells (r.m.p., –59.2 ± 8.9 mV) according to previously described methods (Fujiwara-Tsukamoto et al. 2003).
The Cl– indicator MEQ was excited by single-wavelength illumination from a Xe-arc lamp through ND and band-pass (340–380 nm) filters, and fluorescence images on the basis of emission lights filtered at 435–485 nm were captured with the CCD camera (Fukuda et al. 1998). The illuminated area of the slices was restricted to the somatic portion of the recorded neurons to minimize possible phototoxicity of UV illumination. The captured images were stored in another Windows PC running the image acquisition system AquaCosmos 2.0 (Hamamatsu Photonics), which controlled exposure time (25–100 ms) and sampling rate (10–20 Hz) of the CCD camera. Three regions of interest (ROIs), 4 × 4-pixel squares (5.6 × 5.6 μm), were placed in the somatic region of the fluorescence image, with one ROI on the apical side and two ROIs on the basal side (see Fig. 1A). For the analysis of fluorescence changes in each ROI, background fluorescence was subtracted routinely and photobleaching of the indicator was corrected linearly according to the slope for 2 s immediately before the synaptic stimulation. The index –ΔF/F, as defined in our calcium imaging study (Isomura and Kato 1999), was used to estimate relative change of intracellular Cl– concentration; F is the averaged fluorescent intensity obtained for 2 s before the stimulation, and ΔF is the increase from the F to fluorescent intensity excited at a given time. Therefore quenching of the MEQ fluorescence corresponding to Cl– increase will be expressed as positive value in this index. In the present study, we did not determine the absolute Cl– concentration in recorded neurons.
Bicuculline methiodide and picrotoxin were purchased from Sigma, St. Louis, MO; MEQ was from Molecular Probes, Eugene, OR; other reagents were from Nacalai Tesque, Kyoto, Japan, or from Wako Pure Chemical, Osaka, Japan. All data are expressed as the means ± SD unless otherwise mentioned, and Student's t-test or ANOVA was applied for statistical comparisons. All experiments were carried out in accordance with the guidelines for care and use of animals approved by Hamamatsu University, Tokyo Metropolitan Institute for Neuroscience, and the Physiological Society of Japan.
It was previously shown that strong tetanic stimulation induces slow posttetanic depolarization lasting for 1–5 s, which is dependent on GABAA receptor activation, in the hippocampal CA1 pyramidal cells (Fujiwara-Tsukamoto et al. 2003). In the present study, we attempted to examine intracellular Cl– accumulation during the depolarizing response by microscopic Cl– imaging technique with the Cl– indicator MEQ in combination with whole cell patch-clamp recording. We focused especially on the somatic Cl– dynamics (Fig. 1A, right) because GABAergic synapses are condensed to the somatic and perisomatic portions of pyramidal cells (Papp et al. 2001) and because local GABA application to the soma elicits large depolarizing response after the tetanization (Fujiwara-Tsukamoto et al. 2003). A tetanic stimulation with weak intensity induced a small depolarizing response and only a little change in the MEQ fluorescence (Fig. 1A). In contrast, strong tetanic stimulation induced the prominent posttetanic depolarization followed by the seizure-like afterdischarge. The measurement of MEQ fluorescence exhibited large –ΔF/F increase corresponding to intracellular Cl– accumulation in all three somatic ROIs (Fig. 1B). The fluorescence responses were evoked by the strong tetanus irrespective of the polarity of stimulating electrode; however, they were not observed during strong direct depolarizations by intracellular current injection (∼0.5 nA, data not shown). We failed to observe any fluorescence change associated with single hyperpolarizing inhibitory postsynaptic potential (IPSP) probably because Cl– driving force at resting membrane potential is too small to detect by the present imaging technique (Fukuda et al. 1998).
The peak latencies of the slow posttetanic depolarizations were quite comparable to those of the MEQ fluorescence changes (usually within 2 s). Therefore we analyzed the correlation between the membrane depolarization changes at 2 s after the stimulation and the peak fluorescence changes in individual trials with stimulation intensities varied (Fig. 1C; mean membrane depolarization, 20.6 ± 14.6 mV; peak –ΔF/F, magnitude 5.94 ± 2.45%, latency 1.47 ± 0.35 s; n = 70 trials in 27 cells). The peak –ΔF/F values were positively correlated with the amplitude of posttetanic depolarization (r = 0.69; n = 70; P < 0.001). Moreover, the linear regression suggests that weak synaptic activation eliciting no depolarization can evoke slight Cl– accumulation (3.52% –ΔF/F at 0 mV depolarization), which implies that the Cl– accumulation may underlie the generation of posttetanic depolarization.
Next, it was pharmacologically examined whether the somatic Cl– accumulation is mediated through GABAA receptors during the synaptic activation. Bath application of the GABAA antagonist 25 μM bicuculline, which blocks the posttetanic depolarization (Fujiwara-Tsukamoto et al. 2003), diminished the change in MEQ fluorescence largely but not completely (Fig. 2, A, C, and D; ACSF: 107.1 ± 22.2%, n = 24, P > 0.1; bicuculline: 30.1 ± 8.0%, n = 14; P < 0.001). Similar results were obtained by applying 50 μM picrotoxin, another GABAA antagonist (data not shown). Furthermore, perfusion with Ca2+-free ACSF, which abolished every synaptic transmission in the slices, completely depressed the posttetanic depolarization as well as the fluorescence change (Fig. 2, B, C, and D; 0Ca/5Mg: 13.4 ± 8.3%, n = 6; P < 0.01). As bicuculline reduced changes in MEQ fluorescence by >70%, the major component of somatic Cl– accumulation may be mediated through GABAA receptors. However, the other Cl– influx pathways may also exist because the reduction in the fluorescence change by the bicuculline application was significantly smaller than that by the perfusion with Ca2+-free ACSF (Fig. 2, C and D, P < 0.001).
We also examined that presynaptic interneurons innervating the somatic or peri-somatic portions of the pyramidal cells actually fire action potentials, hence activating postsynaptic GABAA receptors during the posttetanic depolarization. Of known hippocampal interneurons, the basket cell is a typical subtype of fast-spiking GABAergic interneurons, which terminates just at the somata of the pyramidal cells (Freund and Buzsáki 1996). As shown in Fig. 3, we confirmed that morphologically identified basket cells indeed fired at a very high frequency (∼50–100 Hz) during the posttetanic depolarization (n = 8), possibly allowing massive Cl– influx into the postsynaptic pyramidal cells through GABAA receptors.
Our present results show the correlation between the slow posttetanic depolarization and the somatic Cl– accumulation both of which were dependent on GABAA receptor activation in the mature hippocampal pyramidal cells. The peak of the MEQ fluorescence response analyzed here was observed reliably within 2 s after the stimulation. However, the fluorescence change later than the peak (>5 s) might be nonspecifically influenced owing to a delayed increase in cell volume. In fact, Takagi et al. (2002) reported a GABAA receptor-mediated cell swelling observed as changes in IR transmittance in hippocampal CA1 region, which peaked ∼10s after high-frequency stimulation (100 Hz). It is possible that Cl– influx per se may cause the delayed cell swelling during prominent excitation of pyramidal cells (Inglefield and Schwartz-Bloom 1998a,b), which might cause a dilution of intracellular MEQ (Fukuda et al. 2001). On the other hand, the fluorescence of Cl– indicators is intensified, rather than diminished, during hyposmotic or isosmotic cell swelling induced in constantly Cl–-free condition (Srinivas and Bonanno 1997). Taken together, although the later component of the Cl– increase observed here could be actually under- or overestimated to some extent, the early component should indicate changes in intracellular Cl– concentration. Furthermore, a volume change alone could not explain the fluorescence changes exceeding 5% such as observed here. Therefore intracellular Cl– accumulation does occur at least in the somatic portion of pyramidal cells during the posttetanic depolarization.
Consistent with the inhibitory effect of GABAA receptor antagonists on the posttetanic depolarization, these antagonists largely blocked the synaptically activated somatic Cl– accumulation. Usually the somatic Cl– concentration is much lower than the dendritic Cl– concentration (Hara et al. 1992), and the somatic GABAA response is hyperpolarizing while the dendritic GABAA response may be depolarizing in the resting state of pyramidal cells (Gulledge and Stuart 2003). But somatic GABAA response is dynamically converted from hyperpolarizing into depolarizing during the posttetanic depolarization and seizure-like afterdischarge (Fujiwara-Tsukamoto et al. 2003), and such GABAergic depolarization seems to be driven mainly by somatically or peri-somatically innervating interneurons including the basket cells (Y. Fujiwara-Tsukamoto, Y. Isomura, K. Kaneda, M. Takada, personal communications). GABAergic depolarization does not necessarily mean “excitatory” (i.e., enhancing discharge); actually, GABA transmission could be “inhibitory” during the slow posttetanic depolarization owing to its shunting effect (see spike amplitudes in Fig. 1B), though it is indeed excitatory during the seizure-like afterdischarge (Fujiwara-Tsukamoto et al. 2003). Thus the somatic Cl– accumulation mediated through GABAA receptors may participate in the generation of epileptic neuronal activity.
The remaining component of the synaptically activated Cl– accumulation would be accomplished by other Cl– influx/efflux pathways such as cation-Cl– transporters, Cl–/HCO3– exchangers, ATP-driven Cl– pumps, and voltage-sensitive Cl–-channels. In particular, two cation-Cl– cotransporters, KCC2 and NKCC1, might actively contribute to this residual component of the Cl– accumulation (Schomberg et al. 2003). It is likely that K+ accumulation in extracellular spaces increases NKCC1-driven Cl– influx and decreases or reverses KCC2-driven Cl–-efflux (Jarolimek et al. 1999; Ueno et al. 2002; Yamada et al. 2001) during the posttetanic neuronal excitations. Such dynamic changes in the balance between Cl– intrusion and extrusion will greatly influence the normal functions of GABA as an inhibitory transmitter in adult brains. For example, it was recently reported that abnormal excitatory GABAergic transmission facilitates interictal epileptiform activity in hippocampal slices from temporal lobe epilepsy patients (Cohen et al. 2002). In the kindling models, NKCC1 is upregulated (Okabe et al. 2002) while KCC2 is downregulated (Rivera et al. 2002). Thus intracellular Cl– accumulation regulated by GABAA receptors and/or Cl– transporters should play crucial roles in abnormal GABAergic excitation in the epileptic conditions.
In conclusion, using Cl– imaging and whole cell patch-clamp techniques, we revealed that the slow posttetanic depolarization is positively correlated with the somatic Cl– accumulation in hippocampal pyramidal cells. The somatic Cl– accumulation was mediated mainly through postsynaptic GABAA receptors, which might be activated by fast-spiking interneurons such as basket cells.
The present study was supported by the Ministry of Education, Culture, Sports, Science and Technology of Japan, Japan Science and Technology Corporation, and the Japan Society for the Promotion of Science.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵* Y. Isomura and M. Sugimoto contributed equally to this work.
- Copyright © 2003 by the American Physiological Society