III neurons

Brain Research 915 (2001) 94–100 www.elsevier.com / locate / bres

Short communication

Synchronized population oscillation of excitatory synaptic potentials dependent of calcium-induced calcium release in rat neocortex layer II / III neurons H. Yoshimura a,b,c , T. Sugai c , N. Onoda c , N. Segami b , N. Kato a , * a

Department of Integrative Brain Science, Kyoto University Graduate School of Medicine, Kyoto 606 -8501, Japan b Department of Dentistry and Stomatology, Kanazawa Medical University, Uchinada 920 -0293, Japan c Department of Physiology, Kanazawa Medical University, Uchinada 920 -0293, Japan Accepted 13 July 2001

Abstract We examined the roles played by calcium-induced calcium release from ryanodine-sensitive calcium stores in induction of neocortical membrane potential oscillation by using caffeine, an agonist of ryanodine receptors. Intracellular recordings were made from neurons in layer II / III of rat visual cortex slices in a caffeine-containing medium. White matter stimulation initially evoked monophasic synaptic potentials. As low-frequency stimulation continued for over 10 min, an oscillating synaptic potential gradually became evoked, in which a paroxysmal depolarization shift was followed by a 8–10-Hz train of several depolarizing wavelets. This oscillating potential was not induced in a medium containing no caffeine with 2 or 0.5 mM [Mg 21 ] o . Under blockade of N-methyl-D-aspartate receptors, induction of this oscillating potential failed even with caffeine application. Experiments with the calcium store depletor, thapsigargin, revealed that this oscillating potential is induced in a manner dependent on intracellular calcium release. Dual intracellular recordings revealed that the oscillation was synchronized in pairs of layer II / III neurons. The oscillating potential was detectable by field potential recordings also, suggesting that the present oscillation seems to reflect a network property.  2001 Elsevier Science B.V. All rights reserved. Theme: Excitable membranes and synaptic transmission Topic: Postsynaptic mechanisms Keywords: Visual cortex; Caffeine; Rat; Intracellular calcium; Membrane potential oscillation

Synchronous membrane potential oscillation in populations of neocortical neurons is suggested to underlie cognitive functions of the brain [9,18]. An essential basis of such oscillation in the neocortex is thought to be rhythmic bursting of action potentials [6]. Recent electrophysiological studies have pointed to critical roles played by intracellular calcium increases and calcium-activated channels in generation of rhythmic burst firings in central neurons including mesencepharic, thalamic and neocortical neurons [2,5,8,10,11,16,22]. As a source of calcium necessary for activating such calcium-activated channels, involvement of low threshold voltage-dependent calcium channels has been well documented especially in thalamic *Corresponding author. Tel.: 181-75-753-4661; fax: 181-75-7534486. E-mail address: [email protected] (N. Kato).

[10] and mesencephalic neurons [22]. In supraoptic nucleus neurons [14] and neocortical neurons [11], calciuminduced calcium release (CICR) has also been proposed to play a critical role in burst firings. Hence, CICR may be essentially involved not only in elicitation of burst firings but also in the induction of synchronous population oscillation. The present report describes a synaptic induction of synchronous population oscillation in visual cortex slices, which turned out to require CICR. It is also shown that NMDA receptor activation as well as CICR is needed for the induction of this oscillation. Visual cortex slices were prepared as described previously [12]. Briefly, Wistar rats at postnatal days 35–60 were decapitated under ether anesthesia and the brains were quickly removed and soaked in a cold medium (2–48C) consisting of (in mM) NaCl (124), KCl (3.3), NaH 2 PO 4 (1.25), MgSO 4 (0.5 or 2), CaCl 2 (2), NaHCO 3

0006-8993 / 01 / $ – see front matter  2001 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 01 )02832-3

H. Yoshimura et al. / Brain Research 915 (2001) 94 – 100

(26) and D-glucose (10), saturated with O 2 –CO 2 (95:5%). Slices (350 mm thick) were cut from the visual cortex with a slicer (Dosaka EM, Kyoto, Japan), and were left at room temperature for at least 1 h before starting the recording session. A submerged-type chamber placed on the stage of an upright microscope (Axioskop, ZEISS, Oberkochem, Germany) was perfused with medium (308C) at 5 ml / min. A bipolar tungsten electrode was inserted into the white matter for stimulation (duration5200 ms; intensity53.5– 6.0 V). According to the purposes of the experiments, the following drugs (purchased from Nakalai Ltd, Kyoto, Japan) were added to medium: caffeine, 4.0–6.0 mM; D-2-amino-5-phosphonovaleric acid (AP5), 10 mM; thapsigargin, 0.1–80 mM. Intracellular recordings were made from layer II / III neurons with a glass micropipette filled with 3 M potassium acetate. Micropipettes for field potential recordings were filled with 3 M NaCl and inserted into layer II / III. Synaptic responses were evoked at 0.03–0.3 Hz, recorded with bridge-equipped amplifier (Axoclamp-2A or 2B, Axon Instruments, Union City, CA, USA), digitized at 0.32–2.0 kHz (TL-1 or Digidata 1200, Axon Instruments) and stored in a PC for off-line analysis. In some experiments, recordings were made from two neurons simultaneously. In no pairs of the recorded neurons were there mutual synaptic connections, as revealed by attempting to elicit action potentials with current injection in one cell and to record synaptic responses from the other. To determine the frequency of membrane

95

potential oscillation, we used the fast Fourier transformation module included in the Origin software (Microcal, Northampton, MA, USA). All the experiments were carried out in accordance with the guidelines of the Japanese Physiological Society. The student’s t-test was used for statistics. Data were expressed by average6s.e.m., unless otherwise noted. The average membrane potential and input resistance of all the cells recorded intracellularly was 78.460.6 mV and 29.560.6 MV (N5108). In medium containing 2.0 mM [Mg 21 ] o [Fig. 1; 2.0 mM, Caffeine (1), top trace], subthreshold stimulation to white matter evoked a solitary postsynaptic potential (PSP) in 57 of 57 neurons examined. Once caffeine was applied into the medium [Fig. 1; 2.0 mM, Caffeine (1), below the open arrow], the amplitude of PSPs started to gradually increase, and paroxysmal depolarizing shifts (PDSs) accompanied by burst spike discharges began to occur. With the stimulation further continued for more than 10 min, the membrane potential oscillation was induced, albeit in only 7 of the 57 neurons examined (12.3%). This low probability of induction might be due to a slightly higher Mg 21 concentration (2 mM) than the physiological concentration (1.2 mM). We therefore reduced it to 0.5 mM in another set of seven neurons, to which the same protocol of stimulation was applied. In all the seven neurons examined, a solitary PSP seen before caffeine application [Fig. 1, 0.5 mM, caffeine (1), top trace] was changed into PDS (second from the top) and the membrane potential

Fig. 1. Induction of oscillatory synaptic potentials in a caffeine-containing medium. From top to bottom, representative traces are arranged in chronological order to show the gradual appearance of oscillatory potentials in caffeine-containing medium [Caffeine (1)], and the absence of oscillation or paroxysmal depolarization shifts (PDSs) in a caffeine-lacking medium [Caffeine (2)]. On the left, the time after caffeine application is indicated. [Mg 21 ] o was 2.0 mM (left) or 0.5 mM (right). The top traces in Caffeine (1) groups were recorded in normal medium before bath-application of caffeine (indicated by open arrows), and the second to fifth traces were obtained after caffeine application.

96

H. Yoshimura et al. / Brain Research 915 (2001) 94 – 100

started to show an oscillation (third trace) in the same time course as in the experiments with 2.0 mM [Mg 21 ] o . The oscillation gradually became prominent with time (50 min later, 60 min later). With 0.5 mM [Mg 21 ] o , the duration of a single episode of oscillation was much longer than in the experiments with 2.0 mM [Mg 21 ] o , and the wavelets conforming the oscillation were larger in amplitude and number. To examine whether the membrane potential oscillation thus induced is synchronized in a population of neurons within layer II / III of the visual cortex, dual intracellular recordings were made from pairs of neurons. In a caffeinecontaining medium with 0.5 mM [Mg 21 ] o , membrane potential oscillation was induced in 7 / 7 pairs examined. For all these seven pairs, the oscillation was synchronized (Fig. 2A). With all these seven pairs, fast-Fourier transformation analysis revealed that the frequency range of the synchronous oscillation was mainly 8–10 Hz, falling within the range (Fig. 2B). Once the recorded neurons started to exhibit the oscillation in response to synaptic stimulation, we attempted to change the strength of stimulation. There seems to be a threshold stimulation strength for generating an oscillation episode (Fig. 2C). With a subthreshold stimulation (2.3 V; Fig. 2C), not a single wavelet was elicited. With the intensity of stimulation increased (2.5 V; Fig. 2C), a full episode of oscillation suddenly occurred, consisting of 7–8 wavelets in both neurons of the pair. Neither neuron showed a membrane potential drift without the other neuron of a pair exhibiting a synchronized change in membrane potential. Furthermore, occurrence of the oscillation was independent of the action potential generation in all the seven pairs of neurons examined, arguing against the possibility that one neuron may directly drive the partner neuron of the pair to synchronize the oscillation. These dual recording experiments may agree with the view that the present oscillation stems from a network property rather than a mechanism intrinsic to a single neuron, which was also suggested by field potential recordings. Population oscillation induced by the same method was also detected in field potential recordings and was shown to consist of wavelets of similar number and frequency (Fig. 2D). In contrast to these findings in the caffeine-containing medium, the same synaptic stimulation in the caffeinelacking medium failed to induce the oscillation either at 2.0 mM [Mg 21 ] o [0 / 5 neurons; Fig. 1; 2.0 mM, caffeine (2)] or at 0.5 mM [Mg 21 ] o [0 / 5 neurons; Fig. 1; 0.5 mM, caffeine (2)]. These findings point to the possibility that the induction of the present oscillation requires CICR from intracellular calcium stores, which is known to be facilitated by caffeine [1,21]. To test a possible involvement of CICR in the induction of population oscillation, we examined the effect of the calcium store depletor, thapsigargin, in the caffeine-containing medium with 0.5 mM [Mg 21 ] o . The effects of

thapsigargin were examined at various concentrations by using field potential recordings (Fig. 3A). With 0.1 mM thapsigargin, we observed synchronized population oscillation similar to that detected by intracellular recordings. With 20 mM thapsigargin in the medium, however, the oscillation was only partially observed and the number of the wavelet in the single oscillatory episode was smaller than in the medium containing no thapsigargin. The wavelet number decreased with increasing concentrations of thapsigargin (Fig. 3B). This dose–response relationship is likely to support the conclusion that the induction of the present membrane potential oscillation depends on CICR. The dependency of the oscillation on [Mg 21 ] o in the present experiment might be due to the involvement of N-methyl-D-aspartate (NMDA) receptors, which is subject to voltage-dependent blockade by [Mg 21 ] o . Alternatively, low [Mg 21 ] o might have enhanced transmitter release, thereby promoting large oscillating synaptic potentials. We, therefore, tested the effect of the NMDA receptor antagonist AP5. This drug completely blocked the induction of oscillation in all the 5 / 5 cells tested in the caffeinecontaining medium (Fig. 3C, from top to bottom in chronological order). The generation of PDSs was also completely blocked. Thus, activation of NMDA receptors is a necessary condition for the induction of the present oscillation, as reported for the induction of membrane potential oscillation in the hippocampus at low Mg 21 [19]. Once induced, however, the expression of the oscillation seems to be largely independent of NMDA receptors, since the amplitude of each wavelet in an oscillatory episode became smaller as the membrane potential was getting depolarized by current injections (Fig. 4). Membrane voltage dependence of this type excludes a major involvement of NMDA receptors in the expression of the oscillation. This finding also indicates that the frequency of the present population oscillation is independent of the membrane potential. The present experiments in the visual cortex have shown that CICR is required for a synaptic induction of synchronized population oscillation of a-frequency range in neocortical neurons. Caffeine was used to reduce the threshold for triggering CICR [21], and we would suggest that this CICR may be triggered by calcium influx coming through synaptically activated NMDA receptors. Once the oscillation was established, not a solitary synaptic potential but a large oscillatory potential consisting of 7–8 wavelets was evoked by a single stimulation. This finding may also indicate that caffeine application drastically enhances neuronal excitability, in agreement with the epileptogenic effect of caffeine (e.g., Ref. [3]). Evidence has recently been accumulated that intracellular calcium increases could often be achieved by cooperation of more than one calcium source: cooperation between NMDA receptors and VGCCs [17], NMDA receptors and CICR [4], VGCCs and CICR [13,20], as well as VGCCs and IICR [15,23]. The functional significance of such

H. Yoshimura et al. / Brain Research 915 (2001) 94 – 100

97

Fig. 2. Synchrony of the population oscillation demonstrated by dual intracellular recordings and field potential recordings. (A) Synaptic responses simultaneously recorded from two different neurons (Cell 1 and Cell 2) with [Mg 21 ] o at 0.5 mM in a representative pair of neurons. The responses were elicited by stimulation of a constant intensity (4.0 V). After bath-application of caffeine (indicated by the arrow), stimulation was continued at 0.03–0.3 Hz, and then a synchronized oscillation of membrane potential gradually emerged in both neurons. Traces were arranged chronologically from the left-hand side to the right. Below the specimen recordings, the time after caffeine application is indicated. (B) The frequency of this synchronized oscillation was shown to be mainly 8–10 Hz by Fourier analysis. (C) In the same pair of neurons, neither a solitary synaptic potential nor an episode of oscillation was elicited by a weaker stimulus intensity (2.3 V). With the intensity increased from 2.3 to 2.5 V, a full episode of oscillation appeared in an all-or-none fashion. For each of the three recording pairs (4.0, 2.5 and 2.3 V), the top and bottom traces represent recordings from Cell 1 and Cell 2, respectively. (D) Induction of the population oscillation demonstrated with field potential recordings. In the specimen recordings, the negativity is downward. Similarly to the intracellular recordings shown in Figs. 1 and 2, the oscillation gradually developed, and it took about 1 h for the oscillation to be fully induced. The time after caffeine application is indicated below recording traces.

98

H. Yoshimura et al. / Brain Research 915 (2001) 94 – 100

Fig. 3. Effects of the calcium store depleter, thapsigargin, and the NMDA receptor antagonist, AP5, on the induction of the oscillation. (A) Field potential recordings in a caffeine-containing medium with 0.5 mM [Mg 21 ] o revealed that bath-application of the calcium store depleter, thapsigargin (Thap), interfered the induction of oscillation in a dose-dependent manner. The number of the wavelet in one oscillating episode depended on thapsigargin concentration. (B) Dose–response relationship of the effect of thapsigargin applied into the medium. The wavelet number was plotted against the concentration of thapsigargin (logarithmic scale). Each open circle represents a recording from one separate neuron. Vertical bars with two circles on both sides (sus) indicate that the two circles completely overlap and occupy the position of the bar in the plot. (C) Blockade of NMDA receptors by AP5 completely abolished induction of the oscillation. Potentials were recorded intracellularly. Caffeine (Caff) was applied after the top recordings were obtained (open arrows). AP5 was applied at the same time as caffeine and washed out at 60 min after application. Scale bars520 mV and 200 ms.

cooperative calcium increases due to interaction between different calcium sources has not been clarified. Recently, Guertin and Hounsgaard [7] have reported a synergistic

action of NMDA receptors and voltage-dependent calcium channels for generating membrane potential oscillation in motoneurons. In the present report, we have proposed a

H. Yoshimura et al. / Brain Research 915 (2001) 94 – 100

99

and Ministry of Health and Welfare of Japan (H.Y., N.S. and N.K.).

References

Fig. 4. Effects of changing membrane potential on the amplitude and frequency of the oscillatory potential. Depolarizing (10.4, 10.2 nA) or hyperpolarizing (20.2, 20.4 nA) currents were injected through the intracellular electrode. The amplitude of the wavelets decreased with depolarization, while the frequency of the oscillation did not change.

possible cooperative action of NMDA receptors and ryanodine-sensitive calcium stores that seems to be critical for generating synchronized population oscillation in neocortical neurons. We are aware that there still is a possibility that NMDA receptors and ryanodine receptors are activated in separate neurons located in different parts within the neuron network responsible for generating the present population oscillations, since the present population oscillation appears to stem from a network property rather than a mechanism intrinsic to a single neuron. If these two receptors are activated in the same neurons, the possibility arises that calcium entering through NMDA receptors may trigger CICR in the presence of caffeine, thereby forming a cooperative link between NMDA receptor-mediated calcium entry and ryanodine-sensitive calcium stores. Among Ca-activated conductances, the apamin-sensitive class of Ca-activated K channels are shown to play a critical role in rhythmic firing in midbrain dopaminergic neurons [22], and involvement of Ca-activated non-specific cation channels in burst firing has been suggested in the neocortex [11]. In the present experiment, however, it was left unanswered what class of calcium-dependent channels contribute to generating the present synchronous oscillation in the visual cortex. Along with calcium-activated mechanisms, sodium-dependent processes, such as slowlyinactivating sodium conductance, might also be involved as reported for the membrane oscillation in mesencephalic neurons [22].

Acknowledgements This work was supported in part by grants from the Ministry of Education, Science, Sports and Culture (N.K.)

[1] M.J. Berridge, Neuronal calcium signaling, Neuron 21 (1998) 13– 26. [2] M. Caeser, D.A. Brown, B.H. Gahwiler, T. Knopfel, Characterization of a calcium-dependent current generating a slow after-depolarization of CA3 pyramidal cells in rat hippocampal slice cultures, Eur. J. Neurosci. 5 (1993) 560–569. [3] V. Dzhala, L. Desfreres, Z. Melyan, Y. Ben-Ari, R. Khazipov, Epileptogenic action of caffeine during anoxia in the neonatal rat hippocampus, Ann. Neurol. 46 (1999) 95–102. [4] N. Emptage, T.V.P. Bliss, A. Fine, Single synaptic events evoke NMDA receptor-mediated release of calcium from internal stores in hippocampal dendritic spines, Neuron 22 (1999) 115–124. [5] D.D. Fraser, B.A. MacVicar, Cholinergic-dependent plateau potential in hippocampal CA1 pyramidal neurons, J. Neurosci. 16 (1996) 4113–4128. [6] C.M. Gray, D.A. McCormick, Chattering cells: superficial pyramidal neurons contributing to the generation of synchronous oscillations in the visual cortex, Science 274 (1996) 109–113. [7] P.A. Guertin, J. Hounsgaard, NMDA-induced intrinsic voltage oscillations depend on L-type calcium channels in spinal motoneurons of adult turtles, J. Neurophysiol. 80 (1998) 3380– 3382. [8] S. Haj-Dahmane, R. Andrade, Calcium-activated cation nonselective current contributes to the fast after-depolarization in rat prefrontal cortex neurons, J. Neurophysiol. 78 (1997) 1983–1989. [9] S. Herculano-Houzel, M.H. Munk, S. Neuenschwander, W. Singer, Precisely synchronized oscillatory firing patterns require electroencephalo-graphic activation, J. Neurosci. 19 (1999) 3992–4010. [10] H. Jahnsen H, R. Llinas, Ionic basis for the electro-responsiveness and oscillatory properties of guinea-pig thalamic neurones in vitro, J Physiol. (London) 349 (1984) 227–247. [11] Y. Kang, T. Okada, H. Ohmori, A phenytoin-sensitive cationic current participates in generating the after-depolarization and burst after-discharge in rat neocortical pyramidal cells, Eur. J. Neurosci. 10 (1998) 1363–1375. [12] N. Kato, H. Yoshimura, Reduced Mg 21 block of N-methyl-D-aspartate receptor-mediated synaptic potentials in developing visual cortex, Proc. Natl. Acad. Sci. USA 90 (1993) 7114–7118. [13] N. Kato, T. Tanaka, K. Yamamoto, Y. Isomura, Distinct temporal profiles of activity-dependent calcium increase in pyramidal neurons of the rat visual cortex, J. Physiol. (London) 519 (1999) 467–479. [14] Z. Li, G.I. Hatton, Ca 21 release from internal store: role in generating depolarizing after-potentials in rat supraoptic neurons, J. Physiol. (London) 498 (1997) 339–350. [15] T. Nakamura, J.-G. Barbara, K. Nakamura, W.N. Ross, Synergistic release of Ca 21 from IP3-sensitive stores evoked by synaptic activation of mGluRs paired with backpropagating action potentials, Neuron 24 (1999) 727–737. [16] T. Okada, Y. Kang, H. Ohmori, Li 1 and muscarine cooperatively enhance the cationic tail current in rat cortical pyramidal cells, Eur. J. Neurosci. 11 (1999) 2397–2402. [17] R. Schneggenburger, Z. Zhou, A. Konnerth, E. Neher, Fractional contribution of calcium to the cation current through glutamate receptor channels, Neuron 11 (1993) 133–143. [18] W. Singer, Synchronization of cortical activity and its putative role in information processing and learning, Annu. Rev. Physiol. 55 (1993) 349–374. [19] R.D. Traub, J.G.R. Jefferys, M.A. Whittington, Enhanced NMDA conductance can account for epileptiform activity induced by low

100

H. Yoshimura et al. / Brain Research 915 (2001) 94 – 100

Mg 21 in the rat hippocampal slice, J. Physiol. (London) 478 (1994) 379–393. [20] Y.M. Usachev, S.A. Thayer, All-or-non Ca 21 release from intracellular stores triggered by Ca 21 influx through voltage-gated Ca 21 channels in rat sensory neurons, J. Neurosci. 17 (1997) 7404– 7414. [21] A. Verkratsky, A. Shmigol, Calcium-induced calcium release in neurones, Cell Calcium 19 (1996) 1–14.

[22] C.J. Wilson, C.J. Callaway, Coupled oscillator model of the dopaminergic neuron of the substantia nigra, J. Neurophysiol. 83 (2000) 3084–3100. [23] K. Yamamoto, K. Hashimoto, Y. Isomura, S. Shimohama, N. Kato, An IP3-assisted form of Ca 21 -induced Ca 21 release in neocortical neurons, NeuroReport 11 (2000) 535–539.