The modulation of nicotinic acetylcholine receptors on the neuronal network oscillations in rat hippocampal CA3 area

γ oscillations are associated with higher brain functions such as memory, perception and consciousness. Disruption of γ oscillations occur in various neuro-psychological disorders such as schizophrenia. Nicotinic acetylcholine receptors (nAChR) are highly expressed in the hippocampus, however, little is known about the role on hippocampal persistent γ oscillation. This study examined the effects of nicotine and selective nAChR agonists and antagonists on kainate-induced persistent γ oscillation in rat hippocampal slices. Nicotine enhanced γ oscillation at concentrations of 0.1–10 μM, but reduced it at a higher concentration of 100 μM. The enhancement on γ oscillation can be best mimicked by co-application of α4β2- and α7- nAChR agonist and reduced by a combination of nAChR antagonists, DhβE and MLA. However, these nAChR antagonists failed to block the suppressing role of nicotine on γ. Furthermore, we found that the NMDA receptor antagonist D-AP5 completely blocked the effect of nicotine. These results demonstrate that nicotine modulates γ oscillations via α7 and α4β2 nAChR as well as NMDA activation, suggesting that nAChR activation may have a therapeutic role for the clinical disorder such as schizophrenia, which is known to have impaired γ oscillation and hypo-NMDA receptor function.

F ast network oscillations in the c frequency band  Hz; c oscillation) are associated with brain function such as attention, working memory and sensory information processing [1][2][3][4] . The parvalbumin (PV)-expressing interneurons provide strong inhibitory input to pyramidal neurons and play a critical role in the synchronization of neuronal firing within the network, a basic mechanism for the generation of c oscillations 5 .
Recently, nicotinic acetylcholine receptor (nAChR) agonist, nicotine, has been reported to induce theta activity in the hippocampus 13 and augments stimulation-induced hippocampal theta oscillation via activation of alpha7 acetylcholine receptors 6 . Relatively little is known about the modulation of nAChR on fast network oscillations such as c oscillation. Although nicotine is not able to induce c oscillation, it appears to enhance auditory evoked c oscillations 14 , but the mechanism of nicotinic modulation of c oscillations remains largely unknown.
a7 and a4b2 nAChRs are two subunits of nAChRs commonly expressed in the brain. a7 nAChRs are located on glutamatergic and GABAergic terminals and modulate the release of glutamate and GABA [15][16][17] . a4b2 nAChRs are expressed in GABAergic interneurons and modulate GABA release 16,18,19 . It has been recently reported that a4b2 nAChRs expressed in glutamatergic terminals regulate glutamate release in prefrontal cortex 20 . It is expected that nicotine may activate these receptors and modulate c oscillations 14,21 .
The patients with the neuro-psychological disorders such as schizophrenia are associated with disruption of c oscillations 22,23 , reflecting the dysfunction in sensory information processing and cognitive control in these patients 24,25 . Patients with schizophrenia may be associated with NMDAR hypofunction, as blockade of MDA receptor mimics schizophrenic-like symptoms in both humans and animal model of the disease 26,27 , and induces aberrant c oscillations [28][29][30] . Interestingly, nicotine enhances NMDA-mediated current 31 , ameliorates NMDA receptor antagonist-induced deficits in contextual fear conditioning through a4b2 nAChR in the hippocampus 32 and enhances NMDA cognitive circuits via a7 nAChR activation in dorsolateral prefrontal cortex 33 . These studies indicate that nicotine enhances NMDA receptor function through activation of specific nAChR subunits. Whether NMDA receptor is involved in the modulation of nicotine on c oscillations is unknown, although the pharmacologically-induced persistent c oscillations do not require NMDA receptor activation 34,35 . Therefore, this study aimed to investigate the roles of nAChR activation on c oscillations, clarify the nAChR subunit-specific involvement and determine whether NMDA receptor is involved. We chose the commonly-used model of c oscillations, which can be stable for hours, necessity for the investigation of the roles of various nAChR antagonists and agonists on c. We demonstrated that low concentrations of nicotine enhanced kainate-induced persistent c oscillation through a4b2 and a7 nAChRs as well as NMDA receptor activation and that higher concentration of nicotine reduced c through an NMDA receptor-dependent effect. This study suggests that tonic activation of nAChR modulates hippocampal network oscillations with a positive and negative consequence depending on the concentration of nicotine, thus manipulation of the strength of nAChR activation will be crucial for the improving cognitive function in pathological conditions such as schizophrenia, which is known to have impaired c and NMDA receptor hypofunction.

Methods
Animals. All experimental protocols were approved by the Animal Experimentation Ethics Committees of Xinxiang Medical University and Leeds University, and all efforts were made to minimize animal suffering and reduce the number of animals used. All experiments were performed in accordance with the guidelines of the Animal Care and Use Committee of Xinxiang Medical University and Leeds University. Electrophysiological studies were performed on hippocampal slices prepared from Wistar rats (male, 4-5 week-old). For electrophysiology, the animals were anaesthetised by intraperitoneal injection of Sagatal (sodium pentobarbitone, 100 mg kg 21 , Rhône Mérieux Ltd, Harlow, UK). When all pedal reflexes were abolished, the animals were perfused intracardially with chilled (5uC), oxygenated artificial cerebrospinal fluid (ACSF) in which the sodium chloride had been replaced by iso-osmotic sucrose. This ACSF (305 mosmol l 21 ) contained (in mM): 225 sucrose, 3 KCl, 1.25 NaH 2 PO 4 , 24 NaHCO 3 , 6 MgSO 4 , 0.5 CaCl 2 and 10 glucose. For extracellular field recording, the hippocampal horizontal slices (450 mm) of rat brain were cut at 4-5uC in the sucrose-ACSF, using a Leica VT1000S vibratome (Leica Microsystems UK, Milton Keynes, UK).
Electrophysiological recording, data acquisition and analysis. For extracellular field recordings, the two hippocampal slices were transferred to an interface recording chamber. The slices were maintained at a temperature of 32uC and at the interface between ACSF and warm humidified carbogen gas (95% O 2 -5% CO 2 ). The ACSF contained (in mM): 126 NaCl, 3 KCl, 1.25 NaH 2 PO 4 , 24 NaHCO 3 , 2 MgSO 4 , 2 CaCl 2 and 10 Glucose. The slices were allowed to equilibrate in this medium for 1 h prior to recording. Both channels of an Axoprobe 1A amplifier (Axon Instruments, Union City, CA, USA) were employed for extracellular field recordings, which were made using glass microelectrodes containing ACSF (resistance 2-5 MV). Data were bandpass filtered online between 0.5 Hz and 2 kHz using the Axoprobe amplifier and a Neurolog system NL106 AC/DC amplifier (Digitimer Ltd, Welwyn Garden City, UK). The data were digitized at a sample rate of 5-10 kHz using a CED 1401 plus ADC board (Digitimer Ltd). Electrical interference from the mains supply was eliminated from extracellular recordings online with the use of 50 Hz noise eliminators (HumBug; Digitimer Ltd).
Data were analyzed off-line using software from Spike 2 (CED, Cambridge, UK). Power spectra were generated to provide a quantitative measure of the frequency components in a stretch of recording, where power, a quantitative measure of the oscillation strength, was plotted against the respective frequency. Power spectra were constructed for 30-60 s epochs of extracellular field recordings using a fast Fourier transform algorithm provided by Spike2. The parameters used for measuring the oscillatory activity in the slice were peak frequency (Hz) and area power ( mV 2 ). In the current study, area power was equivalent to the computed area under the power spectrum between the frequencies of 20 and 60 Hz.
All statistical tests were performed using SigmaStat software (SPSS Inc., California, USA). Results are expressed as mean 6 standard error of mean, unless indicated otherwise. Statistical significance for comparison between two groups or among three groups was determined using tests described in the text or in the figure legends, as appropriate. Measures were considered statistically significant, if P , 0.05.

Results
Nicotine increased c frequency oscillations. Kainate (KA, 200 nM) induced persistent c oscillation (20-60 Hz) in rat hippocampal CA3 area. c oscillation usually takes approximately 1 to 2 hours to achieve steady-state and would last for at least three hours (Fig. 1A1, B1, C1), which is in agreement with previous studies [35][36][37] . c oscillations can be blocked by the AMPA/kainate receptor antagonist, NBQX (20 mM), or the GABA A receptor antagonist, bicuculline (20 mM) (n 5 5, data not shown), confirming that these oscillations are mediated by excitatory and inhibitory neurotransmission.
When c oscillations reached a steady state, various concentrations of nicotine (0.1-100 mM) were administered with ACSF. At 0.25 mM, nicotine caused a 23 6 7% increase in the c power (*p , 0.05, compared with control, one-way repeated measures ANOVA, n The increase in c power was associated with a slight decrease in peak frequency after applications of nicotine. On average, the peak frequency was decreased 2.6 6 0.4 Hz (*p , 0.05, n 5 9, one way RM ANOVA, Fig. 1E), 2.7 6 0.4 Hz (**p , 0.01, n 5 13) and 2.0 6 0.5 Hz (*p , 0.05, n 5 10) for applications of 0.25 mM, 1 mM and 10 mM nicotine, respectively. However, 100 mM nicotine had no significant effect on the peak frequency (p . 0.05, n 5 10).
The roles of selective nAChR agonists on c power. To determine which nAChR subunits play a role on c enhancement of nicotine, we further tested the effects of the selective a7 nAChR agonist PNU282987 or the a4b2 nAChR agonist RJR2403 alone or in the combination on c oscillations. Application of PNU282987 (1 mM) or RJR2403 (1 mM) alone enhanced c oscillation as shown in Fig. 2A1-C1, A2-C2 by representative experiments. The combination of two agonists dramatically enhanced c power ( Fig. 2A3-C3). On average, the percent increase in c-power was 28 6 9%, 25 6 6%, and 61 6 13% for PNU282987 (n 5 10), RJR2403 (n 5 9) and PNU282987 1 RJR2403 (n 5 8), respectively. Compared with control, these changes are all of statistical significance (*p , 0.01, one way RM ANOVA, Fig. 2D).
Roles of selective nAChR antagonists on nicotine's role. To determine the involvement of specific nAChR subunits on nicotine's role on c oscillation, the hippocampal slices were pretreated with the selective a4b2 nAChR antagonist DhbE, the selective a7 nAChR antagonist MLA or a combination of both antagonists to see whether these antagonists can preclude nicotine's effects on c. The hippocampal slices were pretreated with DhbE (0.   After the steady state of c oscillations was reached in the presence of these nAChR antagonists, nicotine (1 mM) was applied. Our results showed that MLA (Fig. 3A1-C1) or DhbE (Fig. 3A2-C2) partially reduced nicotinic enhancement on c power, but a combination of both antagonists blocked the nicotinic effect ( Fig. 3A3-C3). On average, nicotine caused 40 6 11% (*p , 0.05, one way RM ANOVA, n 5 6), 33 6 10% (*p , 0.05, n 5 6) and 1 6 3% (p . 0.05, n 5 7) increase in c power for the pretreatment of MLA, DhbE   and MLA 1 DhbE, respectively (Fig. 3D). Two way RM ANOVA also revealed that there was a significant interaction between nAChR antagonists and nicotine for the pretreatment of MLA 1 DhbE (*p , 0.01) and DhbE (*p , 0.05) but not for MLA (p . 0.05). These results indicate that MLA 1 DhbE pretreatment effectively blocks nicotine-induced increase in c power.
In terms of peak frequency, nAChR antagonist alone partially reduced the effect of nicotine on peak frequency of the oscillations; a combination of both antagonists blocked the decrease of peak frequency induced by nicotine. On average,nicotine caused 1.0 6 0.3 Hz (*p , 0.05, one-way RM ANOVA, n 5 6), 0.7 6 0.2 Hz (*p , 0.05, n 5 6) and 0.1 6 0.3 Hz (p . 0.05, n 5 7) decrease in the peak frequency for the pretreatment of MLA, DhbE or MLA 1 DhbE, respectively (Fig. 3E). Two-way RM ANOVA also revealed that there was a significant interaction between nAChR antagonists and nicotine for the pretreatment of MLA 1 DhbE (***p , 0.001), MLA (*p , 0.05) and DhbE (**p , 0.01), indicating that these antagonists either alone or in a combination blocked the nicotineinduced changes in peak frequency.
In a different set of experiments (n 5 10), we also investigated the effects of these antagonists on nicotine's role in the conditions of these antagonists being applied when c power reached a steady state. Similar to the pretreatment of these antagonists, only a combination of both a7 nAChR and a4b2 nAChR antagonists can block nicotine role (data not shown).
Selective nAChR antagonists blocked nicotine-mediated enhancing role but not suppression effect on c oscillations. We then tested whether the combined antagonists affect the role of nicotine at higher concentrations. In the presence of MLA 1 DhbE, 10 mM nicotine caused 11.7 6 2.2% decrease on c power (*p , 0.05, compared with control, n 5 8, Fig. 4A1, B1, C). These results suggest that nAChR antagonists blocked the nicotine-mediated enhancing role on c and exposed a small, inhibitory effect of 10 mM nicotine on c oscillation. Furthermore, we tested the effects of co-application of MLA and DhbE on the role of 100 mM nicotine on c. Our results showed that these antagonists did not affect the c power per se, but enhanced nicotine-mediated suppression of c (Fig. 4 A2, B2). In the presence of DhbE 1 MLA, 100 mM nicotine caused a 70 6 5% decrease on c power (***p , 0.001, n 5 10, Fig. 4C). Compared with c power in the presence of 100 mM nicotine alone (the dashed line shown in Fig. 4C), such a change was of statistical significance (*p , 0.01, two way RM ANOVA). These results indicate that blockage of nAChR enhanced nicotine-mediated suppression on c power.
In the presence of DhbE 1 MLA, further application of 10 mM or 100 mM nicotine (in different set of experiments) did not alter peak frequency. On average, 10 mM and 100 mM nicotine caused 1 6 1 Hz (n 5 8) and 0.4 6 1 Hz (n 5 10) reduction of peak frequency, respectively (p . 0.05, compared with the control).
The co-application of DhbE and MLA both at low micromolar range failed to block the effect of 100 mM nicotine, the concentration of both nAChR antagonists was increased to 10 mM and their effects on the role of nicotine on c were further tested. Co-application of DhbE and MLA both at 10 mM failed to block nicotine-mediated suppression of c power (Fig. 4A3, B3, C, n 5 5), they rather enhanced nicotine-mediated suppression of c. On average, in the presence of DhbE 1 MLA, 100 mM nicotine caused 74 6 9% decrease on c power (*p , 0.05, compared with control). Compared with application of 100 mM nicotine alone, this change was of a statistical significance (*p , 0.01, two-way RM ANOVA).

NMDA receptor involvement in the nicotine's role on c oscillations.
Previous studies indicate that nAChR activation enhanced NMDA receptor function in the hippocampus 31 and dorsolateral prefrontal cortex 33 . We have thus tested whether NMDA receptor activation contributes to the roles of nicotine on c. When c oscillations reached a steady state, NMDA receptor antagonist, D-AP5 (10 mM) was perfused for 40 min and no significant change on c powers was observed, further application of nicotine (1 mM) caused no obvious changes on c power (Fig. 5A1-C1). On average, the percent changes of c powers were 100%, 98.8 6 5.2% and 90.4 6 7.6% for the control (KA alone), D-AP5 and D-AP51nicotine, respectively. There was no statistically significant difference in c powers between control and D-AP5 or D-AP51nicotine (n 5 17, p . 0.05, one way RM ANOVA).
Above results indicate that D-AP5 prevented nicotine-mediated enhancement of c. We further tested whether D-AP5 was able to block the role of nicotine at higher concentrations on c oscillation. 10 mM D-AP5 itself had no significant effect on c oscillation, but completely blocked the enhancing role of 10 mM nicotine on c power (n 5 12, p . 0.05, one way RM ANOVA, Fig. 5A2, B2, D). Interestingly, 10 mM D-AP5 also blocked the suppression role of 100 mM nicotine on c power (n 5 6, p . 0.05, one way RM ANOVA, Fig. 5A3, B3, D).
Moreover, we tested the effects of a low concentration of D-AP5 (1 mM) on various concentrations of nicotine's role on c. Our results showed that at such a low concentration, D-AP5 was able to block the enhancing role of nicotine (1-10 mM) (n 5 8, Fig. 5E) and the suppression effect of nicotine (100 mM) on c oscillations (n 5 8, Fig. 5E). These results indicate that both the enhancing and suppressing effects of nicotine on c oscillations involves NMDA receptor activation.

Discussion
In this study, we demonstrated that nicotine at low concentrations enhanced c oscillations in CA3 area of hippocampal slice preparation. The enhancing effect of nicotine was blocked by pre-treatment of a combination of a7 and a4b2 nAChR antagonists and by NMDA receptor antagonist. However,at a high concentration, nicotine reversely reduced c oscillations, which can not be blocked by a4b2 and a7 nAChR antagonists but can be prevented by NMDA receptor antagonist. Our results indicate that nAChR activation modulates fast network oscillation involving in both nAChRs and NMDA receptors.
Nicotine induces theta oscillations in the CA3 area of the hippocampus via activations of local circuits of both GABAergic and glutamatergic neurons 13,38 and is associated with membrane potential oscillations in theta frequency of GABAergic interneurons 39 . The modulation role of nicotine on c oscillations may therefore involve in similar network mechanism as its role on theta.
In this study, the selective a7 or a4b2 nAChR agonist alone causes a relative small increment in c oscillations, the combination of both agonists induce a large increase in c oscillations (61%), which is close to the maximum effect of nicotine at 1 mM, suggesting that activation of two nAChRs are required to mimic nicotine' effect. These results are further supported by our observation that combined a4b2 and a7 nAChR antagonists, rather than either alone blocked the enhancing role of nicotine on c. Our results indicate that both a7 and a4b2 nAChR activations contribute to nicotine-mediated enhancement on c oscillation. These results are different from the previous reports that only a single nAChR subunit is involved in the role of nicotine on network oscillations. In tetanic stimulation evoked transient c, a7 but not a4b2 nAChR is involved in nicotinic modulation of electrically evoked c 40 ; whereas a4b2 but not a7 nAChR is involved in Our results are also different from the observation that nicotine at even 200 nM attenuats the carbachol-induced c oscillations in the deep layers of rat prefrontal cortex (PFC) 42 . The local network difference between hippocampal CA3 area and prefrontal cortex may not be a factor to explain the different effect of nicotine on c oscillations. A recent study by Acracri et al. (2010) has showed that nicotine decreases inhibitory postsynaptic potentials (IPSPs) rather than increases it when ionotropic glutamate receptors are blocked in the neurons of prefrontal cortex 19 . This study suggests that the role of nicotine on c may be related to the status of ionotropic glutamate  receptors or the level of glutamatergic tone and that a reduced tone of glutamatergic input may reverse the role of nicotine. In our study, KA-induced c may have a higher level of glutamatergic tone than carbachol-induced c, which may explain the different response of nicotine between two studies. This hypothesis, however, needs to be further tested.
Nicotine has been reported to regulate GABA release from interneurons such as perisomatic targeting parvalbumin-expressing cells via activation of nAChR located at presynaptic sites 43 , which may contribute to nicotine's enhancing role on c oscillations.
NMDA receptor appears to be critically involved in both c-enhancing and c-suppressing effects of nicotine at low and high concentration, respectively. The involvement of NMDA receptor in nicotinic modulation of c oscillations was supported by previous study that showed the activation of NMDA receptors on interneurons increased the frequency of cholinergically-induced c oscillations in the mouse hippocampal CA3 region 44 . In this study, the NMDA receptor antagonists, D-AP5, had no obvious effect on KA-induced c,which was in line with previous studies 34,45 . However, this result is different from the observation that acute application of ketamine, another NMDA receptor antagonist, increased KA-induced c oscillations (but reduced the peak frequency) 29 , suggesting that different NMDA receptor antagonists may have differential roles in the modulation of c oscillations.
Acute application of D-AP5 completely blocked the enhancing role of nicotine on c, which was in line with the contributions of NMDA receptors to the nicotinic cholinergic excitation of CA1 interneurons in the rat hippocampus 46 and the modulation of a7 nAChR on presynaptic NMDA receptor expression and structural plasticity of glutamatergic presynaptic boutons 47 as well as the increment of c oscillation in the hippocampal CA3 region by the activation of interneuronal NMDA receptors 44 .
The high concentration of nicotine reversely reduced c oscillations, which can not be blocked by a4b2 and a7 nAChR antagonists but can be prevented by NMDA receptor antagonist. Our results are different from the study that showed nicotine at 100 mM enhanced tetanicstimulation evoked transient c 40 , the difference is likely explained by the different c model used. Tetanic-stimulation evoked transient c is only lasting a few seconds and the stimulation is far away from physiological condition. The compete blockage of down-regulation of nicotine on c suggest that the role of nicotine at the 100 mM is a physiological response rather than non-specific action for such a concentration of nicotine. High concentration of nicotine may impose a rapid and strong NMDA receptor activation, causing a large calcium influx which negatively regulates c oscillations. The reverse relationship between intracellular calcium and c oscillations was demonstrated in previous studies 48,49 . It seems that at the high concentrations (10-100 mM), the activation of nAChRs and NMDA receptor play an opposite role on c oscillations, as nAChR antagonists either exposed or worsen the effects of the down-regulation of nicotine at higher concentrations. Interestingly, it seems all concentrations of nicotine used in this study are able to activate NMDA receptors, as NMDA receptor antagonist at even a low concentration can block the different response of nicotine at various concentrations tested. Nevertheless, this study demonstrates the dose-dependent modulation of nicotine on c oscillations and suggests that nAChR agonists may have a therapeutic effect in neuro-psychological disorders 24 .
Clinical significance. The modulation of nicotine at different concentrations on c oscillations and NMDA receptor function suggests that nAChR activation may be useful for the therapeutic application in schizophrenia, as the abnormal c synchrony was demonstrated in the human [50][51][52] and in the animal models 29,30,53 .