Detection of γ-aminobutyric acid-induced glutamate release in acute mouse hippocampal slices with a patch sensor

ANALYTICAL BIOCHEMISTRY Analytical Biochemistry 353 (2006) 83–92 www.elsevier.com/locate/yabio

Detection of c-aminobutyric acid-induced glutamate release in acute mouse hippocampal slices with a patch sensor Mitsuyoshi Shimane, Kaori Miyagawa, Masao Sugawara

*

Department of Chemistry, College of Humanities and Sciences, Nihon University, Sakurajousui, Setagaya, Tokyo 156-8550, Japan Received 26 December 2005 Available online 4 April 2006

Abstract c-Aminobutyric acid (GABA)-stimulated release of L-glutamate from various neuronal regions of acute mouse hippocampal slices was detected with a patch sensor that responds to L-glutamate at the sub-micromolar level. The response of the patch sensor to L-glutamate was evaluated in terms of an integrated current. The integrated current increased with the concentration of L-glutamate ranging from 0.50 to 5.0 lM. By using the patch sensor, GABA-induced L-glutamate release from acute mouse hippocampal slices was detected. The effect of antagonists for GABAA and GABAB receptors on the L-glutamate release was also investigated. The GABA (25 lM) stimulation induced the release of L-glutamate via GABAA receptor in the CA1 region, but GABA did not induce L-glutamate release in the CA3 region. However, in the presence of the GABAB receptor antagonist (3-aminopropyl)(diethoxymethyl)phosphinic acid (CGP35348), release of L-glutamate in the CA3 region was evoked by GABA stimulation. The glutamate release was completely suppressed when both GABAA and GABAB receptor were inhibited. The current results show that the glutamate release in the CA3 region occurs via a GABAA pathway when GABAB receptors are inhibited.  2006 Elsevier Inc. All rights reserved. Keywords: GABA-induced L-glutamate release; Patch sensor; Mouse hippocampal slice; GABA receptor

and c-aminobutyric acid (GABA)1 are the major excitatory and inhibitory neurotransmitters that play a key role in the central nervous system (CNS) of mammalian brains [1–4]. An inhibitory action of GABA has been recognized in a mature CNS [5]; however, new views have been presented that, at early stages of development, GABA functions predominantly as an excitatory neurotransmitter [6–10]. The mechanism by which synaptic transmission is modulated by the interaction between L-Glutamate

*

Corresponding author. Fax: +81 3 5317 9433. E-mail address: [email protected] (M. Sugawara). 1 Abbreviations used: GABA, c-aminobutyric acid; CNS, central nervous system; DG, dentate gyrus; GluR, glutamate receptor; DNQX, 6,7-dinitroquinoxaline-2,3(1H,4H)-dione; APV, D-(–)-2-amino-5-phosphonovaleric acid; CGP-35348, (3-aminopropyl)(diethoxymethyl)phosphinic acid; Hepes, 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid; CA, cornu ammonis; NMDA, N-methyl-D-aspartate; AMPA, a-amino3-hydroxy-5-methyl-4-isoxazolepropionic acid. 0003-2697/$ - see front matter  2006 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2006.03.028

GABAergic and glutamatergic systems has been addressed in a number of studies [5,11–15]. Relatively few studies have been performed on detection of L-glutamate release from the neuronal regions of hippocampus, which may be modulated by GABAergic activity [16,17]. For the in vivo study of glutamate, microdialysis sampling methods [17–20] frequently are used, although several authors have questioned the neuronal origin of glutamate in dialysates [20]. By combining the microdialysis technique with a built-in enzyme electrode, an electric stimulation-induced L-glutamate release in the dentate gyrus (DG) region of rat hippocampus was monitored in vivo [21]. On the other hand, for acute and cultured brain slices where microdialysis sampling is not adaptable, sampling of extracellular fluid with glass capillaries in combination with an online electrode [22] and a built-in electrode [23,24] has been proposed for quantification of stimulant-induced release of L-glutamate. An alternative method has also been proposed by Kasai and coworkers [25], who detected

84

Detection of GABA-induced glutamate release / M. Shimane et al. / Anal. Biochem. 353 (2006) 83–92

GABA-stimulated glutamate release in cultured rat hippocampal slices with an electrode array. The spatial and time resolution of L-glutamate detection can be improved with a patch sensor prepared by the socalled patch clamp technique [26]. By excising an outsideout patch of biomembranes containing glutamate receptor (GluR) ion channels, a microsensor of 1–2 lm can be constructed. With this microsensor, L-glutamate is detected without sampling of extracellular fluid containing L-glutamate. In addition, the patch sensor is expected to be highly sensitive to L-glutamate and to have a response time of approximately milliseconds [27–30]. Such features of patch sensors are important for monitoring glutamate release induced by chemical stimulation because the stimulation usually is achieved by injection or perfusion of a stimulant solution to a brain slice submerged in a bath solution, and hence the sensor needs to detect a very low concentration of L-glutamate locally spilled from the slice. The patch sensors for L-glutamate reported by several authors [31–33] could detect micromolar concentrations of L-glutamate released in hippocampal slices by electric stimulation. In the current study, we describe detection of GABAstimulated release of L-glutamate from various neuronal regions of acute mouse hippocampal slices by using patch sensors that respond to L-glutamate at the level of sub- to several micromolar concentration. It is demonstrated that GABA stimulation induces the release of L-glutamate via GABAA receptor in the CA1 region but that L-glutamate release in the CA3 region is not evoked due to the inhibitory action of GABAB receptors. Materials and methods Reagents

MgSO4, 26 mM NaHCO3, 1.25 mM NaH2PO4, and 15 mM D-glucose and was saturated with a 95% O2/5% CO2 gas mixture. Hippocampal regions were taken out and transferred into a superfusing chamber positioned under an upright microscope (BX50WI, Olympus, Tokyo, Japan). Superfusate (bath solution) contained 130 mM NaCl, 2.8 mM KCl, 1.0 mM CaCl2, 1.0 mM MgCl2, 10 mM Hepes/NaOH (pH 7.2), 15 mM D-glucose, and 10 lM glycine. In a bath solution, hippocampal slices appeared to be alive longer than 2 h; however, we exchanged the slice used for measurements with a fresh one within 2 h. Preparation of a patch sensor and current recordings The experimental setup used for current recordings with a patch sensor is shown schematically in Fig. 1. An Ag–AgCl electrode served as a reference electrode. Patch pipets were pulled from borosilicate glass capillaries (1.5 mm o.d. and 0.86 mm i.d., Harvard Apparatus, Kent, UK) using a three-pull technique with a Sutter micropipet puller (model P-97, Sutter Instrument, Novato, CA, USA). Pipets having resistance of 15–30 MX in a pipet (inner) solution were used for current recordings. The composition of a pipet (inner) solution was the same as that of a bath solution except that glycine was omitted. After a tight seal (1–10 GX) outside-out patch excised from neurons in the stratum pyramidale region of cornu ammonis 3 (CA3) was formed, current recordings were performed at an applied potential of 60 mV. The height of the patch pipet from the slice surface was determined by measuring the scale of a micromanipulator in the following manner. First, a patch pipet was pressed on the cell membrane to establish a gigaseal. A microscope was focused on the slice surface, and a whole cell patch was formed by application of gentle

GABA and L-glutamate were obtained from Wako Chemicals (Osaka, Japan). 6,7-Dinitroquinoxaline-2, 3(1H,4H)-dione (DNQX), D-(–)-2-amino-5-phosphonovaleric acid (APV), and (3-aminopropyl)(diethoxymethyl)phosphinic acid (CGP-35348) were obtained from Sigma (St. Louis, MO, USA). (+)-Bicuculline was obtained from MP Biomedicals (Irvine, CA, USA). 4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid (Hepes) was obtained from Dojindo Laboratories (Kumamoto, Japan). All other chemicals used were of analytical reagent grade. Milli-Q water (Millipore reagent water system, Bedford, MA, USA) was used throughout the experiments. Slice preparation Adult male ddY mice (7 weeks old) were decapitated under ether anesthesia. Coronal slices (thickness 200 or 300 lm) were cut using a Dosaka DTK-1000 microslicer (Kyoto, Japan) in an ice bath. The slices were incubated for 1 h in a recovery solution at 30–32 C and held at room temperature until use. The recovery solution contained 124 mM NaCl, 3.0 mM KCl, 2.0 mM CaCl2, 1.3 mM

Fig. 1. Schematic illustration of the experimental setup used for current recordings with a patch sensor. The excised membrane contained both NMDA and AMPA receptors, but NMDA receptors existed predominantly.

Detection of GABA-induced glutamate release / M. Shimane et al. / Anal. Biochem. 353 (2006) 83–92

suction. Then the pipet was moved up until the slice surface to form an excised (outside-out) patch, so that the tip of the pipet could be seen clearly. By measuring the scale of a micromanipulator to know the distance from the slice surface, the pipet was further moved up approximately 100 lm and kept at the position. If any channel activities were observed at 60 mV, the patch membrane was discarded and the procedure for preparing an excised patch membrane was repeated. When no channel activities were evoked, a 5.0-ll portion of a 2.5mM GABA solution was injected with a micropipet into a bath solution (final concentration, 25 lM) in the presence of a hippocampal slice. Current recordings usually were started 1 min after the injection of GABA using an Axopatch 200B patch clamp amplifier with a built-in 1.0-kHz filter (Axon Instruments, Burlingame, CA, USA) and stored online using a Physio PC computer (Physio-Tech, Tokyo, Japan) in which pCLAMP software (version 6.0.4 or 8.0, Axon Instruments) was installed. Data acquisition for the currents was continued for 20 s with a sampling interval of 100 ls. For preparing a calibration curve for L-glutamate in the presence of a hippocampal slice, the bath solution was perfused by a fresh bath solution containing a known concentration of L-glutamate. The same procedure was repeated for each concentration with the identical or newly prepared membranes. Data analysis The current responses were analyzed in terms of an integrated current (coulomb) with pCLAMP software (Fetchan, version 6.0.6). The observed channel current at an applied potential (60 mV) was integrated during an entire recording time of 20 s. Results and discussion Response characteristics of a patch sensor for L-glutamate A patch sensor was prepared by excising cell membranes from the CA3 region of a hippocampal slice. The sensor was then positioned 100 lm above the DG region of the slice. The typical current traces before and after the perfusion of a 0.50 lM L-glutamate solution are shown in Fig. 2. In the absence of L-glutamate, the patch membrane remained silent until 16 min (Fig. 3A), showing that nonstimulated (background) release of L-glutamate from the DG region of the acute slice is negligible, that is, below the detection limit. On the other hand, the response of the sensor to L-glutamate was observed as induction of a channel current accompanied by a shift in the chord current. The response of the patch sensor induced by each concentration (0.50 and 5.0 lM) of L-glutamate remained constant until 5 min (Figs. 3B and C). The response of a patch sensor to L-glutamate Q(r) was defined as follows:

85

Fig. 2. Typical current traces before and after perfusion of 0.50 lM L-glutamate: (A) before perfusion of L-glutamate; (B) 1 min after perfusion (2 min) of L-glutamate. The bath solution (500 ll) contained 130 mM NaCl, 2.8 mM KCl, 1.0 mM CaCl2, 1.0 mM MgCl2, 10 mM Hepes/NaOH (pH 7.2), 15 mM D-glucose, and 10 lM glycine. The sensor was prepared from the CA3 region and positioned approximately 100 lm above the DG region. An applied potential was 60 mV.

QðrÞ ¼ QðgÞ  QðoÞ;

ð1Þ

where Q(g) is an integrated current (coulomb) in a unit time (s) obtained for a given concentration of L-glutamate and Q(o) is that in the absence of L-glutamate. Although the background current without L-glutamate showed no significant time dependence until 16 min, the recorded currents in response to each L-glutamate concentration exhibited multiple conductance levels, as is typical for receptor ion channels. Therefore, the integration of currents with respect to time was more preferable as the sensor signal. The excised membranes that evoke channel currents in response to L-glutamate appeared to contain both N-methyl-D-aspartate (NMDA) and a-amino-3-hydroxy-5-methyl4-isoxazolepropionic acid (AMPA) receptors. The contribution of each receptor to the L-glutamate-induced current was examined by recording currents in the presence of glutamate receptor antagonists, that is, DNQX (5.0 lM) for AMPA receptors and APV (30 lM) for NMDA receptors. As shown in Fig. 4, the response to L-glutamate (1.0 lM) was significantly suppressed by the presence of each antagonist, but the suppression by APV was much larger (82 ± 23%, n = 3) than that by DNQX (28 ± 25%, n = 3). These results suggest that both NMDA and AMPA receptors exist in the patch membrane excised from the CA3 region of hippocampal slices. However, at 1.0 lM Lglutamate, NMDA receptors predominantly contributed to evoking the response because NMDA receptors have higher affinity to L-glutamate than do AMPA receptors [34] and hence are activated at a lower concentration. When 50 lM acetylcholine was injected instead of L-glutamate, the patch sensor evoked a channel current, indicating the presence of acetylcholine receptors in the

86

Detection of GABA-induced glutamate release / M. Shimane et al. / Anal. Biochem. 353 (2006) 83–92

investigated with L-glutamate in the concentration range from 0.50 to 5.0 lM. The response was evaluated at 1 min after 2 min perfusion of a bath solution containing each concentration of L-glutamate. As shown in Fig. 5, the response increased in a sigmoid manner with an increase in the concentration of L-glutamate from 0.50 to 5.0 lM. The lower concentration limit is superior to that with the patch sensor reported by Maeda and coworkers [31,32]. The reason for the higher sensitivity of our sensor is that the output signal was an integrated one of channel currents, and at low concentration of L-glutamate NMDA receptors, rather than AMPA receptors, contributed to the induction of the sensor response (vide supra). At 50 lM L-glutamate, the sensor exhibited a response, 26 · 1012 C/s, which is roughly twice as large as the response at 5.0 lM, seemingly due to the contribution of AMPA receptors (EC50 = 500 lM [2]). Because NMDA receptors existed predominantly in the patch membrane (vide supra), the gradual saturation of the response up to 5.0 lM was regarded as due to NMDA receptors. Regarding that the response reached a maximal one at 5.0 lM L-glutamate, the half-maximal response (EC50) value was 1.0 lM, which was in agreement with the reported value (1.0 lM) for NMDA receptors [2]. Detection of GABA-induced L-glutamate release

Fig. 3. Time courses of the responses of patch sensors to L-glutamate. The sensors were prepared from the CA3 region of hippocampal slices and positioned approximately 100 lm above the DG region. Concentrations of L-glutamate were 0 lM (n = 2) (A), 0.50 lM (n = 2) (B), and 5.0 lM (n = 3) (C). For panels A and B, the average of two current recordings was plotted.

excised membrane. However, at 1.0 lM acetylcholine, which is much larger than the reported concentration of acetylcholine in brain (6 nM) [35], no channel activities were evoked. This suggests that acetylcholine, even if released in the acute slice, does not affect the L-glutamate detection. Concentration dependence For preparing an in situ calibration curve for L-glutamate, patch sensors were set approximately 100 lm above the DG region of hippocampal slices, where nonstimulated release of L-glutamate was negligibly small. The concentration dependence of the patch sensor was

The potentiality of a patch sensor for the detection of GABA-induced release of L-glutamate from a target neuronal region was tested. The hippocampal slice was stimulated by injecting a 5.0-ll portion of a 2.5-mM GABA solution into a bath solution (500 ll). The final GABA concentration (25 lM) was much larger than the EC50 value (10 lM) for GABAA receptors [36]. In the case that the patch sensor was set in a bath solution apart from the slice, induction of a channel current was not observed (Fig. 6A) when the slice was stimulated by GABA. However, when the same sensor was moved to approximately 100 lm above stratum pyramidale in the CA1 region, channel currents appeared immediately (Fig. 6B). These results indicate that the GABA-evoked channel current is due to L-glutamate released from the CA1 region of the hippocampal slice. The effect of height of a patch sensor from the surface of a hippocampal slice on the response time and magnitude was investigated by setting the patch sensor 100 lm above the slice or in contact with the surface of pyramidale cells of the CA1 region. The response of the sensor was evaluated as time required for induction of a channel current when a 5.0-ll portion of a 2.5-mM GABA solution was injected into a bath solution (500 ll). Because the injection of a GABA solution was performed at a position apart from the patch sensor to avoid physical disturbance of the sensor, a time lag was needed for GABA to diffuse toward the target neuronal region. Although variation was noticed among

Detection of GABA-induced glutamate release / M. Shimane et al. / Anal. Biochem. 353 (2006) 83–92

87

Fig. 4. Effect of glutamate receptor inhibitors on the responses of a patch sensor to 1.0 lM L-glutamate. (A) Current traces obtained without inhibitors (1), in the presence of 5.0 lM DNQX (2), and in the presence of 30 lM APV (3). (B) Inhibitor dependence of L-glutamate currents. The sensors were prepared from the CA3 region of hippocampal slices and positioned in bath solutions apart from the slices. The responses were measured 1 min after injection of a 5.0-ll portion of 100 lM L-glutamate. In recording currents without inhibitors (case 1 in panel A), L-glutamate was applied by perfusion (2 min). An applied potential was 60 mV. Solution conditions were the same as in Fig. 2.

Fig. 5. Concentration dependence of patch sensors at an applied potential of 60 mV. (A) A current trace recorded before L-glutamate perfusion. (B) Current traces recorded 1 min after perfusion (2 min) of each L-glutamate solution. (C) An in situ calibration curve. The mean values of integrated currents from three to six membranes were plotted against concentration of L-glutamate. Error bars indicate means ± SEM. The sensors were prepared from the CA3 region of hippocampal slices and positioned approximately 100 lm above the DG region. Solution conditions were the same as in Fig. 2.

five current recordings, the average time required for GABA to induce an L-glutamate current was 22 ± 10 s (n = 5) for the surface set sensors (Fig. 7A) and 44 ± 15 s (n = 5) for the sensors set approximately 100 lm above the slice surface (Fig. 7B). The response Q(r) = (7.8 ± 2.1) · 1012 C/s of the former sensor was larger than the response Q(r) = (5.7 ± 2.4) · 1012 C/s

of the latter sensor. Thus, an increase in the distance between the sensor and slice surface resulted in a delay of response and a drop of sensitivity. These results suggest that the patch sensor set approximately 100 lm above the target neuronal region detected L-glutamate overflowed from the neuronal region into a bath solution close to the surface of the patch sensor.

88

Detection of GABA-induced glutamate release / M. Shimane et al. / Anal. Biochem. 353 (2006) 83–92

Fig. 6. Responses of a patch sensor to L-glutamate released in an acute mouse hippocampal slice stimulated by GABA. (A) The sensor was positioned in a bath solution apart from a hippocampal slice. The current at an applied potential of 60 mV was recorded 5 min after injection of GABA (final concentration, 25 lM). (B) The same sensor was moved to approximately 100 lm above the stratum pyramidale (str. pyr.) of the CA1 region, and the response was recorded after 1 min (7.5 min after the GABA stimulation). The arrows indicate the tip of the sensor.

Fig. 7. Examples of current traces recorded at 60 mV after injection of a 5.0-ll portion of a 2.5-mM GABA solution. (A) The sensor was in contact with the cell surface of the CA1 region. (B) The sensor was placed approximately 100 lm above the CA1 region. Solution conditions were the same as in Fig. 2.

It is noted that L-glutamate released from the slice needed to diffuse toward the surface of the sensor set approximately 100 lm above the slice surface. Because L-glutamate, locally released at a neuronal site, diffuses away if distance of the sensor from the slice surface was increased, the sensor needed to be kept at the distance near to the surface, in the current case approximately 100 lm

above the slice, to compare the magnitudes of the responses between different neuronal regions and also between different sets of patch sensors. In the case that the patch sensor was set in contact with the slice surface, the magnitude of the response was larger, as observed above, reflecting the shorter diffusion distance of L-glutamate. However, the surface set sensors were less robust and often

Detection of GABA-induced glutamate release / M. Shimane et al. / Anal. Biochem. 353 (2006) 83–92

exhibited large background currents caused by a leaky membrane. GABA-induced release of L-glutamate from different neuronal regions The patch sensor was applied to detection of GABA-stimulated release of L-glutamate from different neuronal regions (CA1, CA3, and DG) of acute hippocampal slices. The time courses of L-glutamate release from each neuronal region are shown in Fig. 8. The release of L-glutamate in response to the GABA stimulation was neuronal region dependent. In the CA1 and DG

CA1 (str. pyr.) 15

5

2

4

6

4

6

Time (min) CA3 15

10

5

0

regions, release of L-glutamate was clearly seen, whereas in the CA3 region, L-glutamate release was not observed. Without GABA stimulation, release of L-glutamate was not observed for all of the neuronal regions. On the basis of an in situ calibration curve (Fig. 5), the concentration of L-glutamate was evaluated as given in Table 1. The local concentration of L-glutamate 1 min after GABA stimulation was at the level of sub- to several micromolar concentration. The L-glutamate release increased in the order of CA1 (stratum pyramidale) DG  CA3 (below detection limit), but after 5 min the L-glutamate release in the DG region was gradually enhanced, resulting in the order of CA1  DG  CA3 (below detection limit). In the CA1 region, GABA-induced glutamate release at stratum pyramidale was larger than that at stratum radiatum and at stratum moleculare. Effect of GABA receptor antagonists on L-glutamate release

10

0

89

2

To know whether the glutamate release in the CA1 region is related to a GABAA receptor pathway, GABA stimulation was performed in the presence of bicuculline (25 lM), a GABAA receptor antagonist. GABAA receptors are bicuculline-sensitive Cl channels; in contrast, GABAB receptors are insensitive to bicuculline [37,38]. In the presence of bicuculline, the bath application of 25 lM GABA exhibited a marked suppression (89 ± 3%) of L-glutamate release in the CA1 region, as shown in Fig. 9. The results indicate that the major part of L-glutamate release from the CA1 region was via a GABAA receptor pathway. This mechanism is in accordance with the view that the high density of GABAA binding is observed in the CA1 and DG regions of the hippocampus, whereas GABAB binding sites are distributed in the CA3 region [39,40]. Also, our results obtained above support the results of Kasai and coworkers [25], who showed that GABA enhances glutamate release via a GABAA receptor pathway in the CA1 and DG regions of rat hippocampus slices.

Time (min) DG

15

Table 1 Concentration of L-glutamate released from different neuronal regions of acute mouse hippocampal slices under stimulation of 25 lM GABA

10

5

0

2

4

6

Time (min) Fig. 8. Time courses of GABA-induced L-glutamate release from different neuronal regions in acute mouse hippocampal slices: CA1 (stratum pyramidale [str. pyr.]) (top), CA3 (middle), and DG (bottom). d, Stimulated by 25 lM GABA (n = 3 or 7); s, control (without GABA, n = 2). Error bars indicate means ± SEM. The sensors were positioned approximately 100 lm above each target region of hippocampal slices.

Neuronal region

Concentration of glutamatea (lM)

Timeb (min)

Number of measurements

CA1 (stratum pyramidale) CA1 (stratum radiatum) CA1 (stratum moleculare) CA3

1.5 ± 0.04 1.6 ± 0.3

1 5 1 5 1 5 1 5 1 5

3 3 3 3 3 3 3 3 7 7

DG

Note.
90

Detection of GABA-induced glutamate release / M. Shimane et al. / Anal. Biochem. 353 (2006) 83–92

Fig. 9. Effect of inhibitors on GABA-stimulated L-glutamate release in the stratum pyramidale of the CA1 region. (A) Current traces after injection of 25 lM GABA obtained in the absence of bicuculline (1), in the presence of 25 lM bicuculline and 1.0% dimethyl sulfoxide (DMSO) (2), and in the presence of 1.0% DMSO without bicuculline (control) (3). (B) Inhibitor dependence of GABA-induced L-glutamate currents. The responses were measured 1 min after the injection of GABA. The sensors were positioned approximately 100 lm above the stratum pyramidale of the CA1 region. Error bars indicate means ± SEM.

Fig. 10. Effect of inhibitors on GABA-stimulated L-glutamate release in the CA3 region. (A) Current traces after injection of 25 lM GABA obtained in the absence of inhibitors (1), in the presence of 100 lM CGP-35348 (2), and in the presence of 100 lM CGP-35348 and 25 lM bicuculline (3). (B) Inhibitor dependence of GABA-induced L-glutamate currents. The responses were measured 1 min after the injection of GABA. The sensors were positioned approximately 100 lm above the stratum pyramidale of the CA3 region. Error bars indicate means ± SEM.

Detection of GABA-induced glutamate release / M. Shimane et al. / Anal. Biochem. 353 (2006) 83–92

Similarly, release of L-glutamate in the CA3 region was measured in the presence of the GABAB receptor antagonist CGP-35348. In the presence of 100 lM CGP-35348, the GABA stimulation induced release of L-glutamate (Fig. 10), in contrast to no release in the absence of inhibitors (vide supra). However, in the presence of both bicuculline (25 lM) and CGP-35348 (100 lM), release of L-glutamate was not evoked. These results show that the inhibition of GABAB receptors in the CA3 region enhances release of L-glutamate via a GABAA receptor pathway in the CA3 region. GABA has been shown to suppress depolarization (0.1 M KCl)-evoked glutamate release in the CA2 region of rat hippocampus, which is mediated by GABAB receptors [17]. In accordance with the results, our data show that L-glutamate release from the CA3 region is enhanced when GABAB receptors were inhibited. It appears that GABAA and GABAB receptors compete with each other for L-glutamate release.

[6]

[7] [8] [9]

[10]

[11] [12]

[13] [14]

Conclusions [15]

The patch sensors prepared by excising the cell membrane from the CA3 region of mouse hippocampal slices was shown to be useful for detection of L-glutamate at the level of sub- to several micromolar concentration. By using the patch sensor, L-glutamate release from various neuronal regions of acute mouse hippocampal slices under stimulation of GABA could be monitored. The results obtained revealed that release of L-glutamate in the CA1 region is induced via GABAA receptors but that GABA does not induce L-glutamate release in the CA3 region due to the inhibitory action of GABAB receptors. The patch sensor can be used for detecting L-glutamate release in acute brain slices submerged in various stimulant solutions. Acknowledgments

[16]

[17]

[18]

[19]

[20] [21]

This work was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, and Culture (Japan) and by a Research Grant for General Individual Research from Nihon University. References [1] R. Luja´n, R. Shigemoto, G. Lo´pez-Bendito, Glutamate and GABA receptor signalling in the developing brain, Neuroscience 130 (2005) 567–580. [2] R. Dingledine, K. Borges, D. Bowie, S.F. Traynelis, The glutamate receptor ion channels, Pharmacol. Rev. 51 (1999) 7–61. [3] H. Mori, M. Mishina, Structure and function of the NMDA receptor channel, Neuropharmacology 34 (1995) 1219–1237. [4] E. Rabow, S.J. Russek, D.H. Faarb, From ion currents to genomic analysis: recent advances in GABAA receptor research, Synapse 21 (1995) 189–274. [5] R.O. Beleboni, R.O.G. Carolino, A.B. Pizzo, L. Castellan-Baldan, J. Coutinho-Netto, W.F. dos Santos, N.C. Coimbra, Pharmacological and biochemical aspects of GABAergic neurotransmission: patho-

[22]

[23]

[24]

[25]

[26]

91

logical and neuropsychobiological relationships, Cell. Mol. Neurobiol. 24 (2004) 707–728. J.J. LoTurco, D.F. Owens, M.J.S. Heath, M.B.E. Davis, A.R. Kriegstein, GABA and glutamate depolarize cortical progenitor cells and inhibit DNA synthesis, Neuron 15 (1995) 1287–1298. A. Stelzer, R.K. Wong, GABAA responses in hippocampal neurons are potentiated by glutamate, Nature 337 (1989) 170–173. Y. Ben-Ari, Excitatory actions of GABA during development: the nature of the nurture, Nat. Rev. Neurosci. 3 (2002) 728–739. Y. Ben-Ari, E. Cherubini, R. Corradetti, J.L. Gaiarsa, Giant synaptic potentials in immature rat CA3 hippocampal neurons, J. Physiol. 416 (1989) 302–325. T. Taira, K. Lamsa, K. Kaila, Posttetanic excitation mediated by GABAA receptors in rat CA1 pyramidal neurons, J. Neurophysiol. 77 (1997) 2213–2218. A.V. Sem’yanov, Diffusional extrasynaptic neurotransmission via glutamate and GABA, Neurosci. Behav. Physiol. 35 (2005) 253–266. T. Plant, C. Schirra, D. Garaschuk, J. Rossier, A. Konnerth, Molecular determinants of NMDA receptor function in GABAergic neurons of rat forebrain, J. Physiol. 499 (1997) 47–63. J. Bormann, The ‘‘ABC’’ of GABA receptors, Trends Pharmacol. Sci. 21 (2000) 16–19. P.V. Belan, P.G. Kostyuk, Glutamate-receptor-induced modulation of GABAergic synaptic transmission in the hippocampus, Pflu¨gers Arch. 444 (2002) 26–37. G. Bonanno, A. Pittaluga, E. Fedele, G. Fontana, M. Raiteri, Glutamic acid and c-aminobutyric acid modulate each other’s release through heterocarriers sited on the axon terminals of rat brain, J. Neurochem. 61 (1993) 222–230. S. Tanaka, A. Tsuchida, Y. Kiuchi, K. Oguchi, S. Numazawa, T. Yoshida, GABAergic modulation of hippocampal glutamatergic neurons: an in vivo microdialysis study, Eur. J. Pharmacol. 465 (2003) 61–67. M. Docherty, H.F. Bradford, J-Y. Wu, Co-release of glutamate and aspartate from cholinergic and GABAergic synaptosomes, Nature 330 (1987) 64–66. S. Zhang, Y. Takeda, S. Hagioka, K. Takata, H. Aoe, H. Nakatsuka, M. Yokoyama, K. Morita, Measurement of GABA and glutamate in vivo levels with high sensitivity and frequency, Brain Res. Protocols 14 (2005) 61–66. K.L. Dew, E.A. Pehek, B.T. Rasley, Y.L. Ma, T.K. Green, Sampling glutamate and GABA with microdialysis: suggestions on how to get the dialysis membrane closer to the synapse, J. Neurosci. Methods 140 (2004) 127–131. W. Timmerman, B.H.C. Westerink, Brain microdialysis of GABA and glutamate: what does it signify? Synapse 27 (1997) 242–261. M.L. Errington, P.T. Galley, T.V.P. Bliss, Long-term potentiation in the dentate gyrus of the anesthetized rat is accompanied by an increase in extracellular glutamate: real-time measurements using a novel dialysis electrode, Phil. Trans. R. Soc. Lond. B 358 (2003) 675–678. K. Torimitsu, O. Niwa, Real-time detection of GABA-induced synaptic glutamate release in cultured rat cortex, Neuroreport 8 (1997) 1353–1358. K. Nakajima, T. Yamagiwa, A. Hirano, M. Sugawara, A glass capillary microelectrode based on capillarity and its application to the detection of L-glutamate release from mouse brain slices, Anal. Sci. 19 (2003) 55–60. N. Nakamura, K. Negishi, A. Hirano, M. Sugawara, Real time monitoring of L-glutamate release from mouse brain slices under ischemia with a glass capillary based enzyme electrode, Anal. Bioanal. Chem. 383 (2005) 660–667. N. Kasai, Y. Jimbo, O. Niwa, T. Matsue, K. Torimitsu, Real-time multisite observation of glutamate release in rat hippocampal slices, Neurosci. Lett. 304 (2001) 112–116. H. Saitoh, Y. Namatame, A. Hirano, M. Sugawara, An excised patch membrane sensor for arachidonic acid released in mouse hippocampal slices under stimulation of L-glutamate, Anal. Biochem. 329 (2004) 163–172.

92

Detection of GABA-induced glutamate release / M. Shimane et al. / Anal. Biochem. 353 (2006) 83–92

[27] T. Yamakura, K. Shimoji, Subunit- and site-specific pharmacology of the NMDA receptor channel, Prog. Neurobiol. 59 (1999) 279–298. [28] D. Bleakman, Kainate receptor pharmacology and physiology, Cell. Mol. Life Sci. 56 (1999) 558–566. [29] J.D. Clements, A. Feltz, Y. Sahara, G.L. Westbrook, Activation kinetics of AMPA receptor channels reveal the number of functional agonist binding sites, J. Neurosci. 18 (1998) 119–127. [30] J.D. Clements, G.L. Westbrook, Activation kinetics reveal the number of glutamate and glycine binding sites on the N-methyl-Daspartate receptor, Neuron 7 (1991) 605–613. [31] T. Maeda, Y. Shimoshige, K. Mizukami, S. Shimohama, S. Kaneko, A. Akaike, M. Satoh, Patch sensor detection of glutamate release evoked by a single electrical shock, Neuron 15 (1995) 253–257. [32] T. Maeda, S. Kaneko, A. Akaike, M. Satoh, Direct evidence for increase in excitatory amino acids release during mossy fiber LTP in rat hippocampal slices as revealed by the patch sensor methods, Neurosci. Lett. 224 (1997) 103–106. [33] J.J. O’Connor, J. Wu, M.J. Rowan, R. Anwyl, Potentiation of N-methyl-D-aspartate receptor-mediated currents detected using the excised patch technique in the hippocampal dentate gyrus, Neuroscience 69 (1995) 363–369.

[34] D.K. Patneau, M.L. Mayer, Structure–activity relationships for amino acid transmitter candidates acting at N-methyl-D-aspartate and quisqualate receptors, J. Neurosci. 10 (1990) 2385–2399. [35] T. Huang, L. Yang, J. Gitzen, P.T. Kissinger, M. Vreeke, A. Heller, Detection of basal acetylcholine in rat brain microdialysate, J. Chromatogr. B 670 (1995) 323–327. [36] J. Kardos, D.J. Cash, Transmembrane 36Cl flux measurements and desensitization of the c-aminobutyric acidA receptor, J. Neurochem. 55 (1990) 1095–1099. [37] R.W. Olsen, GABA–benzodiazepine–barbiturate receptor interactions, J. Neurochem. 37 (1981) 1–13. [38] R.A. Nicoll, R.C. Malenka, J.A. Kauer, Functional comparison of neurotransmitter receptor subtypes in mammalian central nervous system, Physiol. Rev. 70 (1990) 513–565. [39] W. Wisden, D.J. Laurie, H. Monyer, P.H. Seeburg, The distribution of 13 GABAA receptor subunit mRNAs in the rat brain: I. telencephalon, diencephalon, mesencephalon, J. Neurosci. 12 (1992) 1040–1062. [40] C. Knott, J.J. Maguire, R. Moratalla, N.G. Bowery, Regional effects of pertussis toxin in vivo and in vitro on GABAB receptor binding in rat brain, Neuroscience 52 (1993) 73–81.