Purinergic P2X, P2Y and adenosine receptors differentially modulate hippocampal gamma oscillations

Purinergic P2X, P2Y and adenosine receptors differentially modulate hippocampal gamma oscillations

Neuropharmacology 62 (2012) 914e924 Contents lists available at SciVerse ScienceDirect Neuropharmacology journal homepage: www.elsevier.com/locate/n...

1MB Sizes 0 Downloads 72 Views

Neuropharmacology 62 (2012) 914e924

Contents lists available at SciVerse ScienceDirect

Neuropharmacology journal homepage: www.elsevier.com/locate/neuropharm

Purinergic P2X, P2Y and adenosine receptors differentially modulate hippocampal gamma oscillations Steffen B. Schulz a, b,1, Zin-Juan Klaft a,1, Anton R. Rösler a, Uwe Heinemann a, b, Zoltan Gerevich a, * a b

Institute of Neurophysiology, Charité-Universitätsmedizin Berlin, Oudenarder Str. 16, D-13347 Berlin, Germany NeuroCure Research Centre, Charité-Universitätsmedizin Berlin, Charitéplatz 1, D-10117 Berlin, Germany

a r t i c l e i n f o

a b s t r a c t

Article history: Received 14 June 2011 Received in revised form 20 September 2011 Accepted 22 September 2011

The present study was designed to investigate the role of extracellular ATP and its receptors on neuronal network activity. Gamma oscillations (30e50 Hz) were induced in the CA3 region of acute rat hippocampal slices by either acetylcholine (ACh) or kainic acid (KA). ATP reduced the power of KA-induced gamma oscillations exclusively by activation of adenosine receptors after its degradation to adenosine. In contrast, ATP suppressed ACh-induced oscillations through both adenosine and ATP receptors. Activation of adenosine receptors accounts for about 55%, activation of P2 receptors for w45% of suppression. Monitoring the ATP degradation by ATP biosensors revealed that bath-applied ATP reaches w300 times lower concentrations within the slice. P2 receptors were also activated by endogenous ATP since inhibition of ATPhydrolyzing enzymes had an inhibitory effect on ACh-induced gamma oscillations. More specific antagonists revealed that ionotropic P2X2 and/or P2X4 receptors reduced the power of ACh-induced gamma oscillations whereas metabotropic P2Y1 receptor increased it. Intracellular recordings from CA3 pyramidal cells suggest that adenosine receptors reduce the spiking rate and the synchrony of action potentials during gamma oscillations whereas P2 receptors only modulate the firing rate of the cells. In conclusion, our results suggest that endogenously released ATP differentially modulates the power of ACh- or KA-induced gamma oscillations in the CA3 region of the hippocampus by interacting with P2X, P2Y and adenosine receptors. This article is part of a Special Issue entitled ‘Post-Traumatic Stress Disorder’. Ó 2011 Elsevier Ltd. All rights reserved.

Keywords: Adenosine receptors Acetylcholine Gamma oscillations P2X receptors P2Y receptors ATP

1. Introduction Synchronized fast neural network oscillations in the gamma frequency band (w30e90 Hz) can be found in a variety of brain regions and are associated with various cognitive functions, including sensory processing, selective attention, learning and memory (Bartos et al., 2007; Buzsáki and Draguhn, 2004). Conversely, impairment of gamma oscillations might underlie cognitive dysfunction in diseases such as schizophrenia and Alzheimer’s disease (Nakazawa et al., 2011; Palop and Mucke, 2010). Gamma oscillations can be evoked in the hippocampus in vitro for example by bath application of muscarinic acetylcholine receptor

Abbreviations: ACh, acetylcholine; AP, action potential; ATP, Adenosine 50 -triphosphate; FP, field potential; KA, kainic acid; KAR, kainate receptor; mAChR, muscarinic acetylcholine receptor; Physo, physostigmine; PPADS, pyridoxal phosphate-6-azo(benzene-2,4-disulfonic acid; str. pyr., stratum pyramidale; TNP-ATP, 20 ,30 -O-(2,4,6-trinitrophenyl)adenosine-50 -triphosphate. * Corresponding author. Tel.: þ49 30 450528155; fax: þ49 30 450528962. E-mail address: [email protected] (Z. Gerevich). 1 These authors contributed equally to this work. 0028-3908/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropharm.2011.09.024

(mAChR) agonists (mimicking cholinergic input from the septum, Fisahn et al., 1998) or by glutamate via activation of metabotropic glutamate or kainate receptors (KARs) (Fisahn et al., 2004; Hájos et al., 2000). These models have different properties and may differ in their reliance on excitation and inhibition (Bartos et al., 2007). Adenosine 50 -triphosphate (ATP) is an extracellular signaling molecule released e.g. by neurons and astrocytes (Abbracchio et al., 2009). Once released it acts on ionotropic P2X and metabotropic P2Y receptors and, after metabolism to adenosine, on metabotropic adenosine receptors. Functional P2Y1 receptors have been found on interneurons in the CA3 stratum oriens close to the pyramidal cell layer, but not on pyramidal cells. These interneurons are suggested to be O-LM or basket cells, and calretinin-positive interneurons in the CA1 stratum radiatum and lacunosum-moleculare. In response to P2Y1 activation, interneurons fire more action potentials and release more GABA onto their postsynaptic targets (Bowser and Khakh, 2004; Kawamura et al., 2004). The expression of functional P2X receptors in the hippocampus seems to be restricted to axons and axon terminals of pyramidal cells of both CA1 and CA3 pyramidal neurons (Khakh et al., 2003; Rodrigues et al., 2005).

S.B. Schulz et al. / Neuropharmacology 62 (2012) 914e924

A1 and A2A receptors are the predominant adenosine receptors in the hippocampus. A1 receptors are localized mainly presynaptically on axons and axon terminals and inhibit the release of glutamate (Sasaki et al., 2011; Schubert et al., 1986) but are also present postsynaptically on pyramidal cells and on CA1 interneurons where they reduce neuronal excitability (Greene and Haas, 1991; Li and Henry, 2000). In contrast, A2A receptors have lower affinity for adenosine and are expressed at much lower levels in the hippocampus, mainly postsynaptically on pyramidal cells and enhance excitability and synaptic plasticity (Rebola et al., 2008). The less densely distributed presynaptic A2A receptors have been found to stimulate glutamate (Cunha et al., 1994) and inhibit GABA release (Cunha and Ribeiro, 2000) from hippocampal nerve terminals. The net effect of ATP on the hippocampal network activity may be determined by the local release of ATP, its catabolism by ectonucleotidases, the cellular and subcellular expression of purinergic ATP and adenosine receptors and the affinity of the receptors for the different ATP metabolites (Cunha et al., 1998). In the present study we investigated in the CA3 subfield of the hippocampus whether ATP by activation of purinergic receptors modulates gamma oscillations induced either by ACh or KA. 2. Materials and methods 2.1. Slice preparation Hippocampal slices were prepared from Wistar rats of either sex at an age of 5e7 weeks (150e200 g). Animal procedures were conducted in accordance with the guidelines of the European Communities Council and the institutional guidelines approved by the Berlin Animal Ethics Committee (Landesamt für Gesundheit und Soziales Berlin, T0096/02). Animals were anesthetized with isoflurane and then decapitated. Their brains were rapidly removed and washed with ice-cold ACSF containing (in mM): NaCl, 129; KCl, 3; NaH2PO4, 1.25; NaHCO3, 21; CaCl2, 1.6; MgSO4, 1.8; D-glucose, 10, saturated with carbogen (95% O2/5% CO2). For recordings with the ATP sensors, 2 mM glycerol was added. The brain was cut into 400 mm thick horizontal hippocampal slices with a vibratome (DSK microslicer DTK-1000, Dosaka, Japan). Slices were immediately transferred to an interface-type recording chamber perfused with warm and carbogenated ACSF (36  C, flow rate 1.6e1.7 ml/min, pH 7.4). Slices were left for recovery for at least 1 h before commencing with the experiments. 2.2. Extracellular recordings Extracellular field potentials (FPs) were recorded from stratum pyramidale (str. pyr.) of area CA3b with glass pipettes filled with ACSF (resistance < 3 MU) and placed 80e120 mm below the cut surface of the slice. Recordings were amplified by a custom-made amplifier, low-pass filtered at 1 kHz and sampled at 5 kHz by a CED 1401 interface (Cambridge Electronic Design, Cambridge, UK). Gamma oscillations were induced by bath perfusion of either 10 mM acetylcholine (ACh) and 2 mM physostigmine (Physo) or 100 nM kainic acid (KA). ACh/Physo and KA were perfused for 90 and 50 min, respectively, to allow stabilization of gamma oscillations before drug application. Note that gamma oscillations were evoked in an interface-type chamber known for slower equilibration of slices with a given drug than in submerged chambers (Hájos et al., 2009). 2.3. Intracellular recordings Intracellular recordings were made from CA3b pyramidal cells with sharp glass microelectrodes filled with 2 M Kþ-acetate (resistance 60e100 MU). Intracellular signals were amplified by a SEC-05 LX amplifier (npi electronics, Tamm, Germany), low-pass filtered at 2 kHz and sampled at 10 kHz using the CED 1401 interface. Recordings were done in bridge mode. Cells were penetrated during the induction of gamma oscillations. The measurements were started after the stabilization of gamma oscillations but at least 20 min after penetration. Only cells were accepted which showed stable overshooting action potentials (APs) over the full period of the experiment. The mean membrane potentials of pyramidal cells were 50.2  1.0 (n ¼ 9) and 50.4  1.6 mV (n ¼ 14) after the stabilization of KA- and ACh-induced gamma oscillations, respectively. 2.4. Measurement of changes in the extracellular ATP concentration Changes in the extracellular ATP concentration were detected using microelectrode electrochemical biosensors (Sarissa Biomedical, Coventry, UK) which were used in parallel with FP recordings in CA3b. The function of the sensor is described in

915

detail elsewhere (Llaudet et al., 2005). In short, the sensor consists of a platinum (Pt) wire coated with the enzymes glycerol kinase and glycerol-3-phosphatase oxidase. These enzymes degrade ATP by producing H2O2 which gets subsequently oxidized on the Pt wire, finally resulting in two free electrons per ATP molecule. For the oxidation, a potential of 500 mV had to be supplied permanently to the Pt wire. Both the ATP and a reference sensor, which lacks the specific enzyme layer to detect ATP (subsequently called null sensor), were connected to a potentiostat (Duostat ME200þ, Sarissa Biomedical, Coventry, UK) where the signals were low-pass filtered at 0.1 Hz before they were sampled by the CED 1401 interface at 1 kHz. To estimate the ATP concentration on basis of the measured current, a one-point-calibration with 10 mM ATP was done before and after each experiment (Frenguelli et al., 2007). The ATP sensor and the null sensor were inserted over the full length (500 mm, angle of the sensor to the vertical >37 ) in stratum pyramidale at least 45 min before the beginning of the experiments to exclude detection of ATP signals potentially produced by tissue damage. 2.5. Drugs All drugs were purchased from SigmaeAldrich (ACh 10 mM; ATP 0.1 mM e 3 mM; CGS-15943 10 mM; PPADS (pyridoxal phosphate-6-azo(benzene-2,4disulfonic acid)) 30 mM; physostigmine 2 mM) or Tocris (A 740003 10 mM; ARL 67156 50 mM; KA 100 nM; MRS 2179 30 mM; 2meS-ADP 30 mM; MRS 2211 30 mM; MRS 2578 5 mM; NF 110 1 mM; NF 279 1 mM; NF 340 10 mM; NF 449 10 mM; RO-3 10 mM; TNP-ATP (20 ,30 -O-(2,4,6-trinitrophenyl)adenosine-50 -triphosphate) 10 mM) and dissolved in ACSF except A 740003, CGS-15943, MRS 2578, and RO-3 which were first dissolved in DMSO and then further diluted in ACSF with a final [DMSO] of 0.2& (v/v). 2.6. Data analysis and statistics For analysis of oscillations in FP, power spectra were calculated for 2-min periods, and peak power, peak frequency and spectral gamma area of each power spectrum were determined by using a custom-made script for the Spike2 software (Cambridge Electronic Design, Cambridge, UK). Because of the large variability in absolute power of the oscillations, the oscillation power was normalized to a 10-min period prior to drug application or the corresponding time in control. In control experiments, slices received only ACh/Physo or KA during the whole recording. Phase histograms of APs from intracellular recordings and the corresponding FP waveform averages were calculated by the Spike2 software over time windows covering 1000 APs each. Thereby 0 represents the trough of the FP gamma cycles. Occurrence of fast components at the negative peak of gamma oscillations (most probably spikes in pyramidal cells adjacent to the electrode tip) made a low-pass filtering (100 Hz) of the data necessary (Fisahn et al., 1998). The filtering of the recordings caused a w25 shift in phase of APs of pyramidal cells, since without lowpass filtering of our registrations the phase of the APs was found to be w0 (as published by e.g. Hájos et al., 2004; Gulyás et al., 2010). This applies to all tested drug conditions. Nevertheless we used the filtered data, because for the analysis of the accuracy of spike timing it was more important to obtain spike-corrected field troughs rather than a precise phase preservation. The resulting phase shift is a systematic error which basically does not bias the calculated changes in phase accuracy or preferred phase induced by purinergic ligands. The mean vector for each cell was calculated and the resulting mean phases F and vector lengths r were used to calculate the mean vectors and the circular standard deviations for the cell populations of different drug conditions (Batschelet, 1981; Mardia and Jupp, 1972). Time-frequency-analysis of FPs was done using the Morlet wavelet transform (Farge, 1992). The so called mother wavelet is constructed in such a way that it has zero mean and is localized in both time and frequency space. The family of the Morlet pffiffiffi R þN wavelets ja,b(t) was then generated by ja;b ðtÞ ¼ ð1= aÞ N j0 ðt  b=aÞdt, where J0 is the nondimensional angular frequency set to 6 to satisfy the admissibility condition, b is the translation variable and a the scale variable. The normalization pffiffiffi 1= a assures equal energy for all scaled wavelets. The continuous wavelet transR þN formation Wða; bÞ ¼ N xðtÞja;b *ðtÞdt is the convolution of a signal xðtÞ with a Morlet wavelet ja;b ðtÞ. The Morlet power spectrum, defined as jWða; bÞj2 , has the disadvantage of the slant phenomenon where equal amplitude Fourier components with different frequencies show different Morlet power. Therefore we used a modified equal amplitude wavelet power spectrum (mMPS) (Shyu and Sun, 2002) which corresponds more closely to power spectra obtained with the discrete Fourier pffiffiffiffi transform: mMPS ¼ ð2=a pÞ jWða; bÞj2 . To calculate the power sum, the mMPS were scale summed, averaged for each 5 s and normalized to the control period of 5 min before drug administration. All Morlet wavelet calculations were done by usage of a custom-made script for MATLAB software (MathWorks, Natick, MA). Data were represented as mean  s.e.m unless otherwise denoted. Statistical comparisons between the drug-induced changes and the time-matched changes in control experiments were made using one-way ANOVA, unpaired or, where appropriate, paired Student’s t-test. Each drug effect was compared to its matched control group. For circular data, Rayleigh-test, Moore’s test, WatsoneWilliams test, and, where appropriate, Hotelling test for paired samples were performed (Batschelet, 1981; Zar, 2010). Significance level was set at p < 0.05.

916

S.B. Schulz et al. / Neuropharmacology 62 (2012) 914e924

3. Results 3.1. Kainate-induced gamma oscillations are affected by adenosine receptors but not by purinergic type 2 (P2) receptors Bath application of 100 nM kainate (KA) for 50 min reliably induced persistent gamma oscillations in the str. pyr. of area CA3 with an average frequency of 46.7  2.2 Hz (n ¼ 13) and a gamma peak power of 1913  1308 mV2. Bath application of ATP (300 mM) inhibited the peak power of gamma oscillations to 57.2  7.1% (n ¼ 9; p < 0.05 compared to control; Fig. 1B, E), whereas it did not change the peak frequency (99.0  3.6% of control, p > 0.05). To investigate if P2 receptors are involved in this effect of ATP, we applied PPADS (30 mM; Wirkner et al., 2004), a broad spectrum P2 receptor antagonist which blocks most P2Y and P2X receptors, prior to ATP and found that it did not antagonize the inhibitory effect of ATP (peak gamma power: 52.6  16.0% of control, n ¼ 6; p > 0.05 compared to 300 mM ATP alone; Fig. 1D, E). However, ATP is known to be metabolized by ectonucleotidases to adenosine which was recently found to modulate KA-induced gamma oscillations in the hippocampus by A1 and A2A receptors (Pietersen et al., 2009). Therefore, we next blocked A1 and A2A adenosine receptors by CGS-15943 (10 mM; Heidemann et al., 2005). While the blockade of

the adenosine receptors itself increased the oscillations (peak power: 155.2  16.7% of control, n ¼ 5, p < 0.05) which is in line with previous findings of Pietersen et al. (2009), it subsequently abolished the effect of ATP (peak gamma power: 103.5  4.0% of control period before ATP application, n ¼ 5; p < 0.05 compared to 300 mM ATP alone; Fig. 1C, E). These data indicate that ATP by the activation of P2 receptors does not modulate KA-induced gamma oscillations in the hippocampus but does so after degradation to adenosine by the activation of adenosine receptors, as described by Pietersen et al. (2009). 3.2. Determination of ATP levels within the slices during bath application of ATP To investigate which concentration of ATP is reached within the slice during its application, we next bath-applied ATP at different concentrations and simultaneously measured the concentrations reaching a biosensor positioned in the str. pyr. as described in Section 2.4. We registered much lower concentrations than those which were applied (Fig. 2). During application of 300 mM ATP, a concentration producing w50% inhibition of the gamma power, only 1.06  0.56 mM ATP (geometric mean  s.e.m.) was measured in the slice (n ¼ 11; Fig. 2B).

Fig. 1. Kainate-(KA)-induced gamma oscillations in the CA3 area of the hippocampus are inhibited by adenosine but not by ATP receptors. (A) Gamma oscillations induced by bath application of KA (100 nM), recorded extracellularly in the CA3 pyramidal cell layer from a rat hippocampal slice. Top: field recording before, during and after the application of ATP (300 mM). Middle: wavelet spectrum showing the oscillation frequency over time (warmer colors indicate higher power, with color scaled linearly between zero and maximum value). Bottom: change of the total power over all gamma frequencies shown in the wavelet spectrum (30e60 Hz) normalized to the 5-min period before ATP application. (BeD) Left: field recordings before and during the application of 300 mM ATP under certain conditions. Right: corresponding power spectra of oscillations induced by KA (plus purine receptor antagonist where applicable, black) and following the application of ATP (300 mM, gray or colored, resp.). (B) Effect of ATP on KA-induced gamma oscillations. (C) Effect of ATP on KA-induced oscillations in the presence of the adenosine A1/A2A receptor antagonist CGS-15943 (10 mM) (D) Effect of ATP on KA-induced gamma oscillations in the presence of the broad spectrum P2 receptor antagonist PPADS (30 mM). (E) Quantification of the experiments shown in BeD. Bars represent the change of the peak power of KA-induced gamma oscillations normalized to the baseline directly before the wash-in of ATP (n ¼ 5e13). *p < 0.05 compared to time-matched controls. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

S.B. Schulz et al. / Neuropharmacology 62 (2012) 914e924

917

was 1.93  0.16 mM (bath-applied: 733  108 mM) (n ¼ 5e8) with a Hill slope of 2.32  0.38. In comparison, the IC50 value and the Hill slope for ATP alone were found to be 0.96  0.51 mM (bath-applied: 312  107 mM) and 1.06  0.46 (n ¼ 4), respectively (Fig. 3D). These data suggest that net ATP concentrations at the low micromolar range are able to considerably affect gamma oscillations. 3.4. Acetylcholine-induced gamma oscillations are affected by endogenously released ATP

Fig. 2. Bath-applied ATP reaches a much lower concentration in the hippocampal slice than its actual concentration in the bath medium. (A) Recordings of ATP biosensor, null biosensor and the calculated net ATP signal in CA3 stratum pyramidale after bath application of different concentrations of ATP. (B) Concentration-response curve for measured ATP concentration in the slice after application of ATP (geometric mean  s.e.m.).

3.3. Acetylcholine-induced gamma oscillations are inhibited by activation of P2 and adenosine receptors To investigate whether ATP has an impact on cholinergically induced gamma oscillations, we applied 10 mM ACh and 2 mM physostigmine for 90 min in order to evoke persistent gamma oscillations with a peak power of 2074  972 mV2 (n ¼ 8) and a peak frequency of 38.6  2.9 Hz. Application of ATP (300 mM) reduced the peak power to 48.1  8.4% (n ¼ 4; p < 0.05 compared to control; Fig. 3A, E), whereas the peak frequency was not changed (96.7  5.7% of control; p > 0.05; Fig. 3A). In contrast to KA-induced gamma oscillations, CGS-15943 alone did not alter this oscillation (peak power: 96.5  9.4% of control, frequency: 104.6  1.5% of control, n ¼ 8, p > 0.05 each) but significantly inhibited the effect of ATP (peak gamma power: 81.7  5.6% of control, n ¼ 8; p < 0.05 compared to both ATP and control; Fig. 3C, E). Addition of PPADS to the CGS-15943-containing bath solution, completely abolished the inhibitory effect of ATP (106.8  5.3% of control, n ¼ 8; p > 0.05 compared to control; Fig. 3E) indicating that in contrast to KA-induced gamma oscillations, ACh-induced gamma oscillations are modulated by both adenosine and P2 receptors. The reduction of the ACh-induced gamma power by ATP was found to be concentration-dependent. By plotting the measured concentration during bath application of ATP against its inhibitory effect, the half maximum effect (IC50) in the presence of CGS-15943

To test if endogenously released ATP and the following activation of ATP receptors is also able to affect gamma oscillations we applied the selective ecto-ATPase inhibitor ARL-67156 (50 mM) to increase extracellular ATP levels in the slice (Bowser and Khakh, 2004). The peak power of the ACh-induced oscillations was reduced to 56.1  16.9% (n ¼ 3, p < 0.05 compared to control; Fig. 4), suggesting that ATP released from the slice itself is able to attenuate gamma oscillations. The peak frequency did not change significantly after application of ARL-67156 (101.9  2.7% of control; p > 0.05). Next we applied ATP at a low concentration (30 mM) in the presence of ARL-67156 to investigate whether it is more effective in the presence of the ATPase inhibitor. In addition to the inhibitory effect of ARL-67156, ATP further reduced the gamma peak power to 38.6  10.5% of control period prior to ATP application (n ¼ 6) and much stronger than without ARL-67156 (to 87.4  2.6% of control, n ¼ 4, p < 0.05; Figs. 3 and 4D). These data indicate that bathapplied ATP is rapidly degraded in the slice by ATPases reaching lower concentrations and weaker effects on purinergic receptors. Since endogenous ATP and the subsequent activation of P2 receptors are able to inhibit ACh-induced gamma oscillations, we aimed to examine if ATP is released during gamma oscillations and if the activated P2 receptors control their power. We applied the broad spectrum P2 receptor antagonist PPADS (30 mM) during ACh-induced gamma oscillations and found that the peak power dramatically increased to 233.8  48.2% after a 60-min wash in period (n ¼ 6, p < 0.05 compared to control; Fig. 5A and C). As it was the case with other manipulations, the peak frequency of the oscillations was not affected by PPADS (96.2  1.9% of control; p > 0.05). We also found that inhibition of P2 receptors by PPADS attenuated the decay of gamma oscillations after the washout of ACh/Physo (Fig. 5Ac). The peak power of ACh-induced gamma oscillations declined 30 min after the washout of ACh/Physo to 17.9  3.6% (n ¼ 11). When PPADS was present in the bath, the decay of the gamma oscillations was slower, the power was still 41.9  7.2% after 30 min (n ¼ 5, p < 0.05), suggesting that P2 receptors may temporarily limit oscillation periods. In comparison, KA-induced gamma oscillations were not affected by PPADS (peak power: 106.6  17.6% of control, n ¼ 8, p > 0.05 compared to control; peak frequency: 90.9  2.3% of control, p > 0.05; Fig. 5B and C), confirming again that P2 receptors are not involved in the modulation of kainate induced gamma oscillations and that PPADS did not affect ACh-induced gamma oscillations via a non-ATP receptor mediated mechanism. In summary, these data suggest that ATP is present in the extracellular space when ACh is used to induce gamma oscillations and limits the power of gamma oscillations. 3.5. P2Y and P2X receptors differentially modulate ACh-induced gamma oscillations In the next series of experiments we aimed to examine which P2 receptor subtypes are responsible for the effect of ATP on ACh-induced gamma oscillations. In the hippocampus, functional P2Y1 and P2X receptors are expressed on interneurons and

918

S.B. Schulz et al. / Neuropharmacology 62 (2012) 914e924

Fig. 3. Acetylcholine-(ACh)-induced gamma oscillations in the CA3 area of the hippocampus are inhibited by both adenosine and ATP receptors. (A) Gamma oscillations induced by bath application of 10 mM ACh and 2 mM physostigmine, recorded extracellularly in the CA3 pyramidal cell layer from a rat hippocampal slice. For further explanation see the legend of Fig. 1. (B) Top: field recording before, during and after the application of ATP (300 mM) in the presence of CGS-15943 (10 mM) to block A1 and A2A adenosine receptors. The wavelet spectra underneath show the oscillation frequency over time for 1-s periods before (a) and during (b) the application of ATP. Warmer colors indicate higher power, with color scaled linearly between zero and the overall maximum value of both spectra. (C) Comparison of the oscillation properties induced by ACh/Physo in the presence of CGS-15943 (10 mM, black line) and following application of 300 mM ATP (gray line) using power spectra of representative time windows. (D) Concentration-response curve for inhibition of ACh-induced gamma oscillations by ATP in the absence (n ¼ 4, open circles) and presence of CGS-15943 (10 mM; n ¼ 5e8, gray circles). The x-axis represents local ATP concentrations in str. pyr. measured after the application of ATP with the sensors as shown in Fig. 2 (geometric means  s.e.m.). (E) Bars represent the change of the peak power of ACh-induced gamma oscillations after application of ATP (300 mM) in the absence (open, n ¼ 8) and presence of the antagonists (gray, n ¼ 4e8) in comparison to time-matched controls (black, n ¼ 8). *p < 0.05 compared to time-matched controls. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

principal cells, respectively (Bowser and Khakh, 2004; Kawamura et al., 2004; Khakh et al., 2003). Therefore, we first applied MRS 2179 (30 mM), a specific antagonist for P2Y1 receptors (Gerevich et al., 2004; Guzman et al., 2010). As seen on Fig. 6A, 30 mM MRS 2179 reduced the peak power of oscillations to 75.5  14.4% (n ¼ 5, p < 0.05 compared to control). In contrast, MRS 2179 did not influence the peak frequency of the oscillations (106.8  2.0% of control; p > 0.05). To explore whether additional P2Y receptors are also able to modulate gamma oscillations we applied specific antagonists for other P2Y subtypes (Fig. 6C). Neither MRS 2578 (5 mM; Mamedova et al., 2004; peak power: 91.6  11.0% of control, n ¼ 9), NF 340 (10 mM; Meis et al., 2010; 97.7  10.8%, n ¼ 8) nor MRS 2211 (30 mM, Ortega et al., 2011; 84.7  6.1%, n ¼ 8), specific antagonists for the P2Y6, P2Y11 and P2Y13 receptors, respectively, affected the power of gamma oscillations (p > 0.05 each, compared to control). Next we investigated whether P2X receptors also have modulatory effects on gamma oscillations and applied TNP-ATP (10 mM), a selective antagonist for P2X1, P2X2, P2X3 and P2X4 receptors (Lorca et al., 2011). We found that TNP-ATP enhanced the peak power of gamma oscillations to 236.7  27.8% (n ¼ 7, p < 0.05

compared to control; Fig. 6B) and reduced the peak frequency of the oscillations to 90.0  3.5% (p < 0.05). In contrast to this, neither the P2X1 selective antagonists NF 279 (1 mM; Rettinger et al., 2000; 91.4  12.0% of control, n ¼ 4) and NF 449 (10 mM; Hausmann et al., 2006; 104.8  3.6%, n ¼ 3), the P2X3 and P2X2/3 selective NF 110 (1 mM; Hausmann et al., 2006; 134.0  26.4, n ¼ 8), the P2X3 selective RO-3 (10 mM; Gever et al., 2006; 83.6  20.1, n ¼ 6) nor the P2X7 selective A 740003 (10 mM; Honore et al., 2006; 105.8  34.7, n ¼ 6) changed significantly the power of gamma oscillations (p > 0.05, compared to control; Fig. 6D). Together, these experiments suggest that P2Y1 receptors enhance ACh-induced gamma oscillations whereas P2X2 or P2X4 receptors inhibit it. 3.6. P2 and adenosine receptors differentially alter spiking rate and spike timing of CA3 pyramidal cells Changes in FP power can be due to alterations in the firing rate of neurons or in the synchrony of action potentials between cells. In order to get some insight into the underlying mechanisms, we recorded intracellularly from CA3b pyramidal cells during gamma oscillations induced either by KA or ACh/Physo. All recorded cells

S.B. Schulz et al. / Neuropharmacology 62 (2012) 914e924

919

Fig. 4. Endogenously released ATP inhibits gamma oscillations induced by ACh in the CA3 area of the hippocampus. (A) Field recordings after induction of gamma oscillations by ACh before (top) and during (bottom) the application of the selective ecto-ATPase inhibitor ARL-67156 (50 mM) to increase extracellular ATP levels. (B) Power spectrum of ACh-induced gamma oscillations from the experiment shown in A before (black) and after (gray) the application of ARL-67156. (C) Bars represent the change of the peak power of ACh-induced gamma oscillations after application of ARL-67156 (gray, n ¼ 3) compared to the time-matched controls (black, n ¼ 8). (D) Subsequent application of ATP (30 mM) further reduces the peak power of gamma oscillations (gray bar, normalized to control period prior to ATP application, n ¼ 6) with higher effectiveness than ATP alone (black, n ¼ 4). *p < 0.05.

Fig. 5. P2 receptors modulate ACh- but not KA-induced gamma oscillations. (A) ACh-induced gamma oscillations are enhanced by the application of PPADS. (Aa) Field recordings before (top) and during the application of PPADS (30 mM; bottom). (Ab) Power spectrum of ACh-induced gamma oscillations from the experiment shown in Aa before (black) and after (gray) the application of PPADS. (Ac) Inhibition of P2 receptors by PPADS attenuates the decay of gamma oscillations after the washout of ACh. Normalized peak power of gamma oscillations over time after washout of ACh in the presence (black circles, n ¼ 5) and absence (open circles, n ¼ 11) of PPADS. (B) KA-induced gamma oscillations are not affected by the application of the broad-spectrum P2 receptor antagonist PPADS. (Ba) Field recordings before (top) and during the application of PPADS (30 mM; bottom). (Bb) Power spectrum of KA-induced gamma oscillations from the experiment shown in Ba before (black) and after (gray) the application of PPADS. (C) Bars represent the change of the peak power of ACh-induced (black, n ¼ 8) and of KA-induced (gray, n ¼ 8) gamma oscillations after application of PPADS (30 mM).

920

S.B. Schulz et al. / Neuropharmacology 62 (2012) 914e924

Fig. 6. P2Y1 receptors potentiate and P2X1e4 receptors inhibit ACh-induced gamma oscillations in the CA3 area of rat hippocampal slices. (A) Selective inhibition of P2Y1 receptors by MRS 2179 inhibits the power of ACh-induced gamma oscillations. (Aa) Field recordings before (top) and during the application of MRS 2179 (30 mM; bottom). (Ab) Properties of the oscillations induced by ACh and physostigmine (black) and following application of MRS 2179 (gray) compared by using power spectra of representative time windows. (B) Inhibition of the P2X1-4 receptors by TNP-ATP enhances the power of ACh-induced gamma oscillations. (Ba) Field recordings before (top) and during the application of TNP-ATP (10 mM; bottom). (Bb) Properties of the ACh-induced oscillations before (black) and following application of TNP-ATP (gray) compared by using power spectra of representative time windows. (C, D) Bars show normalized peak power of ACh-induced gamma oscillations after application of P2Y (C, gray) and P2X (D, black) receptor antagonists (n ¼ 3e9) selective for the indicated receptor subunits.*p < 0.05 compared to time-matched controls.

(n ¼ 23) in all tested drug conditions as well as the cells as populations showed firing behavior phase-locked to the FP of the gamma oscillation (Rayleigh test, p < 0.001 for each cell and condition and Moore’s test, p < 0.005 for each population and condition, respectively, Figs. 7A and 8C). Analysis of circular data (as is the case here due to the FP oscillation) provides two key parameters: i) the mean phase F, calculated by averaging the phases of all individual APs of a cell, states the phase within an oscillation cycle when the average AP occurs; and ii) the mean vector length r, which indicates how accurate the neuron fires within the cycle (in a case of r ¼ 1, the cell would fire all APs at the very same phase with maximal accuracy; if all APs were equally distributed over the oscillation cycle, vector length would be r ¼ 0). Given that the same number of APs for each cell is used, F and r of a cell population can be obtained by second order statistics. These parameters are compared here between different pharmacological conditions by appropriate statistical tests (Batschelet, 1981; Zar, 2010). Phase analysis of APs related to the gamma cycle (trough ¼ 0 ) during KA-induced gamma oscillations revealed a mean phase F

of 22.8 and a vector length r ¼ 0.863 (n ¼ 9) after low-pass filtering (see Section 2.6). Since P2 receptor activation didn’t show any effect on KA-induced gamma oscillations, to activate adenosine receptors we applied ATP as the physiological source of adenosine. Application of 300 mM ATP significantly altered spike timing (F ¼ 21.0 , r ¼ 0.774, Hotelling test for paired samples, p < 0.05, Fig. 7A) by reducing the accuracy of spiking (r; paired t-test, p < 0.01, Fig. 7A, C) whereas the preferred phase (F) was not affected (p > 0.05). In addition, the spike rate significantly decreased in response to ATP (ATP: 5.58  0.77 s1, control: 8.31  0.88 s1, p < 0.05), an effect that was completely reversed after ATP was washed out (7.70  0.82 s1, p > 0.05 compared to control, Fig. 7B). We found weak correlations between the reduction in FP power by ATP and the reduction in the spike rate and accuracy of spiking among cells, respectively (r ¼ 0.626, p < 0.07 and r ¼ 0.670, p < 0.048, resp.; Fig. 7D, E). Thus, these data suggest that ATP inhibits KA-induced gamma oscillations via activation of adenosine receptors after its degradation to adenosine by means of reducing spiking rate and synchrony of action potentials of CA3 pyramidal neurons.

S.B. Schulz et al. / Neuropharmacology 62 (2012) 914e924

921

Fig. 7. ATP reduces the spiking rate and alters spike timing of CA3 pyramidal cells during KA-induced gamma oscillation. (A) From top to bottom: intracellular recording from a CA3 pyramidal cell and the corresponding field recording of CA3 str. pyr. after induction of gamma oscillations by KA; waveform average of the corresponding oscillatory FP cycles and phase histograms of action potentials (APs) of the shown cell (bin size: 10 ); mean vectors of each recorded cell (gray arrows, n ¼ 9) and the mean vector  circular standard deviation of the whole population (black, bold arrow). For each cell, 1000 APs were analysed before (left column) and during ATP application (300 mM, right column). 0 represents the troughs of the gamma cycles after low-pass filtering. Arrow lengths represent mean vector lengths r (circle radius ¼ 1). (B) Normalized peak power of KA-induced gamma oscillations (upper trace) and spike rate of the corresponding recordings from CA3 pyramidal cells (lower trace) over time before, during and after the application of 300 mM ATP. Bars summarize the effect of ATP on the spike rate of 10 min periods before (black), during (gray) and after (white) ATP application. (C) Changes in mean vector length r of each recorded cell after ATP application (300 mM) (D) Scatter plot showing the reduction in mean vector length r and FP peak power during ATP application normalized to the phase before application for each cell-FP recording pair. (E) Scatter plot showing spike rate and FP peak power during ATP application relative to the phase before application for each cell-FP recording pair. *p < 0.05 compared to control period.

Fig. 8. Blockade of P2 receptors increased the spiking rate but did not affect spike timing of CA3 pyramidal cells during ACh-induced gamma oscillation. (A) Intracellular recording from a CA3 pyramidal cell and the corresponding field recording in CA3 str. pyr. (B) Bars summarize the effects of PPADS (30 mM) and CGS-15943 (10 mM) on the firing rate. (C) Mean vectors of each recorded cell (gray arrows) and the mean vector  circular standard deviation of the whole population (black, bold arrow). 1000 APs are analyzed in the absence (control) and presence of drugs. 0 represents the troughs of the gamma cycles after low-pass filtering. Arrow lengths represent mean vector lengths r (circle radius ¼ 1). *p < 0.05 compared to control period.

922

S.B. Schulz et al. / Neuropharmacology 62 (2012) 914e924

During ACh-induced gamma oscillations, CA3 pyramidal cells fired at a mean phase of F ¼ 26.1 (r ¼ 0.741, n ¼ 7) after low-pass filtering (see Section 2.6), which was not altered after blockade of P2 and adenosine receptors by PPADS and CGS-15943, respectively (PPADS: F ¼ 24.9, r ¼ 0.680, CGS-15943: F ¼ 15.9 , r ¼ 0.823; n ¼ 7; Hotelling test for paired samples and WatsoneWilliams test, resp., p > 0.05, Fig. 8C). However, while the blockade of adenosine receptors by CGS-15943 did not alter the firing rate of the neurons (5.32  1.06 s1; p > 0.05, compared to control: 6.04  1.71 s1), PPADS significantly increased the firing rate compared to the preceding control phase (PPADS: 7.85  2.09 s1, p < 0.05, Fig. 8B). These data suggest that activation of ATP receptors by endogenous ATP inhibits ACh-induced gamma oscillations by reducing the firing rate in CA3 pyramidal neurons and not by changing the synchrony of action potentials. 4. Discussion The main finding of our data is that extracellular ATP modulates the power of both KA- and ACh-induced hippocampal gamma oscillations. While ATP reduces the power of KA-induced gamma oscillations exclusively by activation of adenosine receptors after its degradation, we show here for the first time that ACh-induced gamma oscillations are inhibited by both adenosine and ATP sensitive P2 receptors. Under our experimental conditions, the inhibitory effect of ATP comprises an adenosine receptor component of w55% and a P2 receptor component of w45% (Fig. 3D, E). 4.1. Effect of P2Y and P2X receptors on gamma oscillations Analyzing the P2 receptor component of the inhibition in more details we found two opposing effects: activation of P2Y1 receptors by ATP enhanced gamma oscillations whereas P2X receptors reduced them. Among the P2Y receptor antagonists only MRS 2179, a selective inhibitor of P2Y1 receptors, was able to modulate gamma oscillations. These receptors are functionally expressed on CA3 interneurons but not on pyramidal cells (Bowser and Khakh, 2004; Kawamura et al., 2004). Their stimulation may increase the precisely timed GABA release from perisomatic-targeting interneurons which may lead to increased synchronization of pyramidal cell firing and subsequently to enhanced oscillation power (Mann and Paulsen, 2007). Among the P2X receptors, P2X2, P2X4 and P2X6 receptors are known to be expressed in the hippocampus (Kanjhan et al., 1999; Rubio and Soto, 2001). We found that only TNP-ATP but not other applied P2X receptor antagonists affected gamma oscillations. TNP-ATP blocks P2X1, P2X2, P2X3 and at a less extent P2X4 receptors (Jarvis and Khakh, 2009). Since antagonists of P2X1 and P2X3 receptors did not modulate gamma oscillations, it can be suggested that P2X2 and/or P2X4 receptors are the predominant P2X subunits involved in the suppression of gamma oscillations. Further studies e.g. with KO mice are needed to prove the exact P2X receptor subtype due to the lack of more specific antagonists. P2X channels seem to be restricted to the axons of CA3 pyramidal cells (Khakh et al., 2003; Khakh, 2009; Rodrigues et al., 2005). Ionotropic receptors expressed on axon terminals (Schicker et al., 2008) or along the axons of CA3 pyramidal cells (Sasaki et al., 2011) can modify network activity by altering synaptic transmission or AP conduction. P2X receptors seem to have similar effects (Khakh et al., 2003; Rodrigues et al., 2005). Since generation of ectopic spikes in CA3 pyramidal cell axons may be essential for the development of gamma oscillations (Traub et al., 2000, 2004), their inhibition by P2X receptors can explain the inhibitory effect of P2X receptors on gamma oscillations. Support for this hypothesis are our findings showing that a blockade of P2X channels increases the

firing rate of pyramidal cells, possibly due to higher phasic excitatory input via the recurrent collaterals in the CA3 region. 4.2. Effect of adenosine receptors on gamma oscillations It has been previously shown that adenosine inhibits KA-induced and spontaneous gamma oscillations, particularly via the activation of A1 receptors (Pietersen et al., 2009). Here we found that both KA- and ACh-induced gamma oscillations were inhibited by adenosine receptors. Hippocampal adenosine A1 receptors are known to reduce transmitter release and excitability of pyramidal cells and interneurons (Li and Henry, 2000; Schubert et al., 1986). Moreover, axonal A1 receptors on CA3 pyramidal cells were described to decrease AP conduction and synaptic efficacy (Sasaki et al., 2011). These effects may account for the inhibition of gamma oscillations by adenosine and are in accordance with our results showing that adenosine receptor activation during KA-induced gamma oscillations reduced both the spiking rate of CA3 pyramidal cells and the synchrony of action potentials in these neurons. Selective activation of A2A receptors enhances gamma oscillations (Pietersen et al., 2009). However, the expression of these receptors (Fredholm et al., 2005) and their affinity for adenosine are lower (Ciruela et al., 2006) making A1 receptors to be the functionally dominating adenosine receptor subtype in the hippocampus. 4.3. Extracellular ATP levels We found that ATP concentrations reach w300 times lower levels in the slice after bath application. This has been observed also by others (Frenguelli et al., 2007) and may be explained by the rapid degradation of ATP. The equilibration time course of ATP and the development of its effects were found to be much faster than that observed for other drugs in interface chambers (Decker et al., 2009; Wójtowicz et al., 2009). An explanation might be that ATP induces waves of astrocyte activation mediated by P2Y receptors and subsequent ATP release from astrocytes resulting in a rapid equilibration of the slice but at much lower ATP concentrations (Bowser and Khakh, 2004; Guthrie et al., 1999). Our results indicate that the strength of cholinergically induced gamma oscillations is also controlled by ambient extracellular ATP since both blocking the degradation of ATP to adenosine and application of purinergic antagonists alone were able to modulate gamma oscillations. Thus, endogenous ATP at concentrations in the range of a few hundred nanomolar to few micromolar can finely tilt the balance between excitation and inhibition in the hippocampal network via the activation of P2Y, P2X and adenosine receptors. Since the net impact of exogenously applied ATP is an inhibition of the gamma oscillations, the inhibitory effect of P2X and adenosine receptors seems to dominate this effect. 4.4. Modulation of KA-induced and cholinergic gamma oscillations by neuromodulators A number of neurotransmitters and neuropeptides have been shown to modulate gamma oscillations in the hippocampus such as dopamine (Weiss et al., 2003; Wójtowicz et al., 2009), 5-HT (Krause and Jia, 2005; Wójtowicz et al., 2009), noradrenaline (Hajós et al., 2003; Wójtowicz et al., 2009), cannabinoids (Hájos et al., 2000), C-type natriuretic peptide (Decker et al., 2009), histamine (Andersson et al., 2010) and the aforementioned adenosine (Pietersen et al., 2009). Among them, monoamines such as dopamine and 5-HT seem to inhibit both ACh- and KA-induced gamma oscillations (Weiss et al., 2003; Wójtowicz et al., 2009), whereas histamine was found to decrease only KA- but not carbachol-induced gamma

S.B. Schulz et al. / Neuropharmacology 62 (2012) 914e924

oscillations in the hippocampus (Andersson et al., 2010). The selective modulatory effect of P2 receptors for ACh-induced gamma oscillations presented in our study further confirms that cholinergic and KA-evoked gamma oscillations involve different neuronal pathways, can be differentially modulated and may have distinct physiological functions. Besides gamma oscillations, activation of KARs induces also epileptiform bursts and kainate application has long been used as an animal model for epileptogenesis (Ben-Ari and Cossart, 2000; Nadler, 1981). Our findings that ATP and its receptors do not modulate KA-induced network activity are in line with reports about the missing or very limited effects of P2 receptors on epileptiform activity (Lopatár et al., 2011; Ross et al., 1998). 4.5. Functional relevance We preferred using an interface-type recording chamber to submerged conditions because gamma oscillations require a high oxygen supply (Huchzermeyer et al., 2008) guaranteed only in an interface-type chamber. Hypoxia causes gamma oscillations to collapse within minutes (Fano et al., 2007; Huchzermeyer et al., 2008) associated with elevated extracellular ATP levels (Dale and Frenguelli, 2009; Frenguelli et al., 2007). It can be suggested that the hypoxia-induced collapse of gamma oscillations is caused by the increased ATP levels and the subsequent inhibition of gamma oscillations by activation of P2X and adenosine receptors. This would be an energy conserving mechanism since oxygen consumption seems to be highest during gamma network activity (Huchzermeyer et al., 2008). ATP is, among glutamate and D-serine, one of the most important gliotransmitters (Haas et al., 2006; Halassa et al., 2009) known to be released from astrocytes enabling these cells to modulate neuronal processes such as synaptic transmission (Pascual et al., 2005). Our findings suggest that astrocytes, by releasing ATP and the subsequent activation of purinergic receptors, may also be able to control neuronal gamma oscillations. The power and the synchrony of gamma oscillations have been found to be reduced in schizophrenia and the degree of this reduction seems to correlate with the severity of negative symptoms (Lee et al., 2003). Thus, drugs which selectively activate fast spiking interneurons could provide an effective approach in schizophrenia to increase the synchronization of pyramidal cell firing at gamma frequencies and consequently improving the symptoms of the disease (Gandal et al., 2011). Our results suggest that the selective activation of P2Y1 receptors or antagonism of P2X receptors may provide a possible approach to reverse the loss of gamma power and the subsequent cognitive impairments in schizophrenia. Acknowledgements This work was supported by the Deutsche Forschungsgemeinschaft (He 1173/17-1) and by the Hertie foundation. S.B.S. was a stipendiary of NeuroCure (DFG Exc 257) and Z.-J.K. was supported by the Graduiertenkolleg 1123 (Learning & Memory). The authors thank Dr. Arpad Mike for helpful discussion. References Abbracchio, M.P., Burnstock, G., Verkhratsky, A., Zimmermann, H., 2009. Purinergic signalling in the nervous system: an overview. Trends Neurosci. 32, 19e29. Andersson, R., Lindskog, M., Fisahn, A., 2010. Histamine H3 receptor activation decreases kainate-induced hippocampal gamma oscillations in vitro by action potential desynchronization in pyramidal neurons. J. Physiol. 588, 1241e1249. Bartos, M., Vida, I., Jonas, P., 2007. Synaptic mechanisms of synchronized gamma oscillations in inhibitory interneuron networks. Nat. Rev. Neurosci. 8, 45e56. Batschelet, E., 1981. Circular Statistics in Biology. Academic Press, London.

923

Ben-Ari, Y., Cossart, R., 2000. Kainate, a double agent that generates seizures: two decades of progress. Trends Neurosci. 23, 580e587. Bowser, D.N., Khakh, B.S., 2004. ATP excites interneurons and astrocytes to increase synaptic inhibition in neuronal networks. J. Neurosci. 24, 8606e8620. Buzsáki, G., Draguhn, A., 2004. Neuronal oscillations in cortical networks. Science 304, 1926e1929. Ciruela, F., Casadó, V., Rodrigues, R.J., Luján, R., Burgueño, J., Canals, M., Borycz, J., Rebola, N., Goldberg, S.R., Mallol, J., Cortés, A., Canela, E.I., López-Giménez, J.F., Milligan, G., Lluis, C., Cunha, R.A., Ferré, S., Franco, R., 2006. Presynaptic control of striatal glutamatergic neurotransmission by adenosine A1eA2A receptor heteromers. J. Neurosci. 26, 2080e2087. Cunha, R.A., Milusheva, E., Vizi, E.S., Ribeiro Sebastiao, J.A., Sebastiao, A.M., 1994. Excitatory and inhibitory effects of A1 and A2A adenosine receptor activation on the electrically evoked [3H]acetylcholine release from different areas of the rat hippocampus. J. Neurochem. 63, 207e214. Cunha, R.A., Sebastião, A.M., Ribeiro, J.A., 1998. Inhibition by ATP of hippocampal synaptic transmission requires localized extracellular catabolism by ectonucleotidases into adenosine and channeling to adenosine A1 receptors. J. Neurosci. 18, 1987e1995. Cunha, R.A., Ribeiro, J.A., 2000. Purinergic modulation of [(3)H]GABA release from rat hippocampal nerve terminals. Neuropharmacology 39, 1156e1167. Dale, N., Frenguelli, B.G., 2009. Release of adenosine and ATP during ischemia and epilepsy. Curr. Neuropharmacol. 7, 160e179. Decker, J.M., Wójtowicz, A.M., Ul Haq, R., Braunewell, K.H., Heinemann, U., Behrens, C.J., 2009. C-type natriuretic peptide decreases hippocampal network oscillations in adult rats in vitro. Neuroscience 164, 1764e1775. Fano, S., Behrens, C.J., Heinemann, U., 2007. Hypoxia suppresses kainate-induced gamma-oscillations in rat hippocampal slices. Neuroreport 18, 1827e1831. Farge, M., 1992. Wavelet transforms and their application to turbulence. Annu. Rev. Fluid Mech. 24, 395e457. Fisahn, A., Pike, F.G., Buhl, E.H., Paulsen, O., 1998. Cholinergic induction of network oscillations at 40 Hz in the hippocampus in vitro. Nature 394, 186e189. Fisahn, A., Contractor, A., Traub, R.D., Buhl, E.H., Heinemann, S.F., McBain, C.J., 2004. Distinct roles for the kainate receptor subunits GluR5 and GluR6 in kainateinduced hippocampal gamma oscillations. J. Neurosci. 24, 9658e9668. Fredholm, B.B., Chen, J.F., Cunha, R.A., Svenningsson, P., Vaugeois, J.M., 2005. Adenosine and brain function. Int. Rev. Neurobiol. 63, 191e270. Frenguelli, B.G., Wigmore, G., Llaudet, E., Dale, N., 2007. Temporal and mechanistic dissociation of ATP and adenosine release during ischaemia in the mammalian hippocampus. J. Neurochem. 101, 1400e1413. Gandal, M.J., Edgar, J.C., Klook, K., Siegel, S.J., 2011. Gamma synchrony: Towards a translational biomarker for the treatment-resistant symptoms of schizophrenia. Neuropharmacology, doi:10.1016/j.neuropharm.2011.02.007. Gerevich, Z., Borvendeg, S.J., Schröder, W., Franke, H., Wirkner, K., Nörenberg, W., Fürst, S., Gillen, C., Illes, P., 2004. Inhibition of N-type voltage-activated calcium channels in rat dorsal root ganglion neurons by P2Y receptors is a possible mechanism of ADP-induced analgesia. J. Neurosci. 24, 797e807. Gever, J.R., Cockayne, D.A., Dillon, M.P., Burnstock, G., Ford, A.P., 2006. Pharmacology of P2X channels. Pflugers Arch. 452, 513e537. Greene, R.W., Haas, H.L., 1991. The electrophysiology of adenosine in the mammalian central nervous system. Prog. Neurobiol. 36, 329e341. Gulyás, A.I., Szabó, G.G., Ulbert, I., Holderith, N., Monyer, H., Erdélyi, F., Szabó, G., Freund, T.F., Hájos, N., 2010. Parvalbumin-containing fast-spiking basket cells generate the field potential oscillations induced by cholinergic receptor activation in the hippocampus. J. Neurosci. 30, 15134e15145. Guthrie, P.B., Knappenberger, J., Segal, M., Bennett, M.V., Charles, A.C., Kater, S.B., 1999. ATP released from astrocytes mediates glial calcium waves. J. Neurosci. 19, 520e528. Guzman, S.J., Schmidt, H., Franke, H., Krügel, U., Eilers, J., Illes, P., Gerevich, Z., 2010. P2Y1 receptors inhibit long-term depression in the prefrontal cortex. Neuropharmacology 59, 406e415. Haas, B., Schipke, C.G., Peters, O., Söhl, G., Willecke, K., Kettenmann, H., 2006. Activity-dependent ATP-waves in the mouse neocortex are independent from astrocytic calcium waves. Cereb. Cortex 16, 237e246. Hajós, M., Hoffmann, W.E., Robinson, D.D., Yu, J.H., Hajós-Korcsok, E., 2003. Norepinephrine but not serotonin reuptake inhibitors enhance theta and gamma activity of the septo-hippocampal system. Neuropsychopharmacology 28, 857e864. Hájos, N., Katona, I., Naiem, S.S., MacKie, K., Ledent, C., Mody, I., Freund, T.F., 2000. Cannabinoids inhibit hippocampal GABAergic transmission and network oscillations. Eur. J. Neurosci. 12, 3239e3249. Hájos, N., Pálhalmi, J., Mann, E.O., Németh, B., Paulsen, O., Freund, T.F., 2004. Spike timing of distinct types of GABAergic interneuron during hippocampal gamma oscillations in vitro. J. Neurosci. 24, 9127e9137. Hájos, N., Ellender, T.J., Zemankovics, R., Mann, E.O., Exley, R., Cragg, S.J., Freund, T.F., Paulsen, O., 2009. Maintaining network activity in submerged hippocampal slices: importance of oxygen supply. Eur. J. Neurosci. 29, 319e327. Halassa, M.M., Fellin, T., Haydon, P.G., 2009. Tripartite synapses: roles for astrocytic purines in the control of synaptic physiology and behavior. Neuropharmacology 57, 343e346. Hausmann, R., Rettinger, J., Gerevich, Z., Meis, S., Kassack, M.U., Illes, P., Lambrecht, G., Schmalzing, G., 2006. The suramin analog 4,40 ,40 ,400 -(carbonylbis(imino-5,1,3-benzenetriylbis (carbonylimino)))tetra-kis-benzenesulfonic acid (NF110) potently blocks P2X3 receptors: subtype selectivity is determined by location of sulfonic acid groups. Mol. Pharmacol. 69, 2058e2067.

924

S.B. Schulz et al. / Neuropharmacology 62 (2012) 914e924

Heidemann, A.C., Schipke, C.G., Kettenmann, H., 2005. Extracellular application of nicotinic acid adenine dinucleotide phosphate induces Ca2þ signaling in astrocytes in situ. J. Biol. Chem. 280, 35630e35640. Honore, P., Donnelly-Roberts, D., Namovic, M.T., Hsieh, G., Zhu, C.Z., Mikusa, J.P., Hernandez, G., Zhong, C., Gauvin, D.M., Chandran, P., Harris, R., Medrano, A.P., Carroll, W., Marsh, K., Sullivan, J.P., Faltynek, C.R., Jarvis, M.F., 2006. A-740003 [N-(1-{[(cyanoimino)(5-quinolinylamino) methyl]amino}-2,2-dimethylpropyl)2-(3,4-dimethoxyphenyl)acetamide], a novel and selective P2X7 receptor antagonist, dose-dependently reduces neuropathic pain in the rat. J. Pharmacol. Exp. Ther. 319, 1376e1385. Huchzermeyer, C., Albus, K., Gabriel, H.J., Otáhal, J., Taubenberger, N., Heinemann, U., Kovács, R., Kann, O., 2008. Gamma oscillations and spontaneous network activity in the hippocampus are highly sensitive to decreases in pO2 and concomitant changes in mitochondrial redox state. J. Neurosci. 28, 1153e1162. Jarvis, M.F., Khakh, B.S., 2009. ATP-gated P2X cation-channels. Neuropharmacology 56, 208e215. Kanjhan, R., Housley, G.D., Burton, L.D., Christie, D.L., Kippenberger, A., Thorne, P.R., Luo, L., Ryan, A.F., 1999. Distribution of the P2X2 receptor subunit of the ATP-gated ion channels in the rat central nervous system. J. Comp. Neurol. 407, 11e32. Kawamura, M., Gachet, C., Inoue, K., Kato, F., 2004. Direct excitation of inhibitory interneurons by extracellular ATP mediated by P2Y1 receptors in the hippocampal slice. J. Neurosci. 24, 10835e10845. Khakh, B.S., Gittermann, D., Cockayne, D.A., Jones, A., 2003. ATP modulation of excitatory synapses onto interneurons. J. Neurosci. 23, 7426e7437. Khakh, B.S., 2009. ATP-gated P2X receptors on excitatory nerve terminals onto interneurons initiate a form of asynchronous glutamate release. Neuropharmacology 56, 216e222. Krause, M., Jia, Y., 2005. Serotonergic modulation of carbachol-induced rhythmic activity in hippocampal slices. Neuropharmacology 48, 381e390. Lee, K.H., Williams, L.M., Breakspear, M., Gordon, E., 2003. Synchronous gamma activity: a review and contribution to an integrative neuroscience model of schizophrenia. Brain Res. Brain Res. Rev. 41, 57e78. Li, H., Henry, J.L., 2000. Adenosine action on interneurons and synaptic transmission onto interneurons in rat hippocampus in vitro. Eur. J. Pharmacol. 407, 237e244. Llaudet, E., Hatz, S., Droniou, M., Dale, N., 2005. Microelectrode biosensor for realtime measurement of ATP in biological tissue. Anal. Chem. 77, 3267e3273. Lopatár, J., Dale, N., Frenguelli, B.G., 2011. Minor contribution of ATP P2 receptors to electrically-evoked electrographic seizure activity in hippocampal slices: evidence from purine biosensors and P2 receptor agonists and antagonists. Neuropharmacology 61, 25e34. Lorca, R.A., Rozas, C., Loyola, S., Moreira-Ramos, S., Zeise, M.L., Kirkwood, A., Huidobro-Toro, J.P., Morales, B., 2011. Zinc enhances long-term potentiation through P2X receptor modulation in the hippocampal CA1 region. Eur. J. Neurosci. 33, 1175e1185. Mamedova, L.K., Joshi, B.V., Gao, Z.G., von Kügelgen, I., Jacobson, K.A., 2004. Diisothiocyanate derivatives as potent, insurmountable antagonists of P2Y6 nucleotide receptors. Biochem. Pharmacol. 67, 1763e1770. Mann, E.O., Paulsen, O., 2007. Role of GABAergic inhibition in hippocampal network oscillations. Trends Neurosci. 30, 343e349. Mardia, K.V., Jupp, P.E., 1972. Statistics of Directional Data. Academic Press, London. Meis, S., Hamacher, A., Hongwiset, D., Marzian, C., Wiese, M., Eckstein, N., Royer, H.D., Communi, D., Boeynaems, J.M., Hausmann, R., Schmalzing, G., Kassack, M.U., 2010. NF546 [4,40 -(carbonylbis(imino-3,1-phenylene-carbonylimino-3,1-(4-methyl-phenylene)-carbonylimino))-bis(1,3-xylene-alpha, alpha’-diphosphonic acid) tetrasodium salt] is a non-nucleotide P2Y11 agonist and stimulates release of interleukin-8 from human monocyte-derived dendritic cells. J. Pharmacol. Exp. Ther. 332, 238e247.

Nadler, J.V., 1981. Kainic acid as a tool for the study of temporal lobe epilepsy. Life Sci. 29, 2031e2042. Nakazawa, K., Zsiros, V., Jiang, Z., Nakao, K., Kolata, S., Zhang, S., Belforte, J.E., 2011. GABAergic interneuron origin of schizophrenia pathophysiology. Neuropharmacology, doi:10.1016/j.neuropharm.2011.01.022. Ortega, F., Pérez-Sen, R., Delicado, E.G., Miras-Portugal, M.T., 2011. ERK1/2 activation is involved in the neuroprotective action of P2Y(13) and P2X7 receptors against glutamate excitotoxicity in cerebellar granule neurons. Neuropharmacology 61, 1210e1221. Palop, J.J., Mucke, L., 2010. Amyloid-beta-induced neuronal dysfunction in Alzheimer’s disease: from synapses toward neural networks. Nat. Neurosci. 13, 812e818. Pascual, O., Casper, K.B., Kubera, C., Zhang, J., Revilla-Sanchez, R., Sul, J.Y., Takano, H., Moss, S.J., McCarthy, K., Haydon, P.G., 2005. Astrocytic purinergic signaling coordinates synaptic networks. Science 310, 113e116. Pietersen, A.N., Lancaster, D.M., Patel, N., Hamilton, J.B., Vreugdenhil, M., 2009. Modulation of gamma oscillations by endogenous adenosine through A1 and A2A receptors in the mouse hippocampus. Neuropharmacology 56, 481e492. Rebola, N., Lujan, R., Cunha, R.A., Mulle, C., 2008. Adenosine A2A receptors are essential for long-term potentiation of NMDA-EPSCs at hippocampal mossy fiber synapses. Neuron 57, 121e134. Rettinger, J., Schmalzing, G., Damer, S., Müller, G., Nickel, P., Lambrecht, G., 2000. The suramin analogue NF279 is a novel and potent antagonist selective for the P2X(1) receptor. Neuropharmacology 39, 2044e2053. Rodrigues, R.J., Almeida, T., Richardson, P.J., Oliveira, C.R., Cunha, R.A., 2005. Dual presynaptic control by ATP of glutamate release via facilitatory P2X1, P2X2/3, and P2X3 and inhibitory P2Y1, P2Y2, and/or P2Y4 receptors in the rat hippocampus. J. Neurosci. 25, 6286e6295. Ross, F.M., Brodie, M.J., Stone, T.W., 1998. Modulation by adenine nucleotides of epileptiform activity in the CA3 region of rat hippocampal slices. Br. J. Pharmacol. 123, 71e80. Rubio, M.E., Soto, F., 2001. Distinct Localization of P2X receptors at excitatory postsynaptic specializations. J. Neurosci. 21, 641e653. Sasaki, T., Matsuki, N., Ikegaya, Y., 2011. Action-potential modulation during axonal conduction. Science 331, 599e601. Schicker, K.W., Dorostkar, M.M., Boehm, S., 2008. Modulation of transmitter release via presynaptic ligand-gated ion channels. Curr. Mol. Pharmacol. 1, 106e129. Schubert, P., Heinemann, U., Kolb, R., 1986. Differential effect of adenosine on preand postsynaptic calcium fluxes. Brain Res. 376, 382e386. Shyu, H.C., Sun, Y.S., 2002. Construction of a morlet wavelet power spectrum. Multidim. Syst. Sing. P. 13, 101e111. Traub, R.D., Bibbig, A., Fisahn, A., LeBeau, F.E., Whittington, M.A., Buhl, E.H., 2000. A model of gamma-frequency network oscillations induced in the rat CA3 region by carbachol in vitro. Eur. J. Neurosci. 12, 4093e4106. Traub, R.D., Bibbig, A., LeBeau, F.E., Buhl, E.H., Whittington, M.A., 2004. Cellular mechanisms of neuronal population oscillations in the hippocampus in vitro. Annu. Rev. Neurosci. 27, 247e278. Weiss, T., Veh, R.W., Heinemann, U., 2003. Dopamine depresses cholinergic oscillatory network activity in rat hippocampus. Eur. J. Neurosci. 18, 2573e2580. Wirkner, K., Schweigel, J., Gerevich, Z., Franke, H., Allgaier, C., Barsoumian, E.L., Draheim, H., Illes, P., 2004. Adenine nucleotides inhibit recombinant N-type calcium channels via G protein-coupled mechanisms in HEK 293 cells; involvement of the P2Y13 receptor-type. Br. J. Pharmacol. 141, 141e151. Wójtowicz, A.M., van den Boom, L., Chakrabarty, A., Maggio, N., Haq, R.U., Behrens, C.J., Heinemann, U., 2009. Monoamines block kainate- and carbacholinduced gamma-oscillations but augment stimulus-induced gamma-oscillations in rat hippocampus in vitro. Hippocampus 19, 273e288. Zar, J.H., 2010. Biostatistical Analysis, fifth ed. Prentice-Hall, Upper Saddle River, NJ.