Modulation of AMPA receptor mediated current by nicotinic acetylcholine receptor in layer I neurons of rat prefrontal cortex

Layer I neurons in the prefrontal cortex (PFC) exhibit extensive synaptic connections with deep layer neurons, implying their important role in the neural circuit. Study demonstrates that activation of nicotinic acetylcholine receptors (nAChRs) increases excitatory neurotransmission in this layer. Here we found that nicotine selectively increased the amplitude of AMPA receptor (AMPAR)-mediated current and AMPA/NMDA ratio, while without effect on NMDA receptor-mediated current. The augmentation of AMPAR current by nicotine was inhibited by a selective α7-nAChR antagonist methyllycaconitine (MLA) and intracellular calcium chelator BAPTA. In addition, nicotinic effect on mEPSC or paired-pulse ratio was also prevented by MLA. Moreover, an enhanced inward rectification of AMPAR current by nicotine suggested a functional role of calcium permeable and GluA1 containing AMPAR. Consistently, nicotine enhancement of AMPAR current was inhibited by a selective calcium-permeable AMPAR inhibitor IEM-1460. Finally, the intracellular inclusion of synthetic peptide designed to block GluA1 subunit of AMPAR at CAMKII, PKC or PKA phosphorylation site, as well as corresponding kinase inhibitor, blocked nicotinic augmentation of AMPA/NMDA ratio. These results have revealed that nicotine increases AMPAR current by modulating the phosphorylation state of GluA1 which is dependent on α7-nAChR and intracellular calcium.

nAChRs can regulate synaptic plasticity in layer I by increasing the spontaneous EPSC 14 . This may be achieved by enhancing AMPA receptor (AMPAR) mediated current which has been demonstrated in dopaminergic neurons 15 . In contrast to AMPAR, the NMDA receptor (NMDAR) is inactivated by nicotine in layer V pyramidal neurons in PFC 16 , suggesting that these glutamatergic receptors might be differentially regulated by nAChR in a region specific manner. However, how nAChRs might regulate the function of AMPAR and/or NMDAR in layer I remains unclear.
In this study, we assessed nicotine effect on AMPAR-and NMDAR-mediated currents in layer I neurons, and identified the potential subunits of nAChR and AMPAR that might be involved in this regulation. Results showed that activation of nAChRs led to the selective enhancement of AMPAR mediated current, which was dependent on α 7-nAChR, intracellular calcium and the phosphorylation state of AMPAR subunit GluA1.

Materials and Methods
Slice preparation. All protocols were approved by the Commission of Chongqing Medical University for ethics of experiments on animals and were conducted in accordance with international standards. Male Sprague-Dawley (SD) rats (2weeks) were obtained from the Experimental Animal Center of Chongqing Medical University. Brain slices were prepared as previously reported [17][18][19] . Briefly, SD rats were anesthetized with 10% chloralic hydras (300 mg/kg). PFC slices (350 μ m) were prepared with a Leica (Germany) VP1200S Vibratome and then incubated in artificial CSF (ACSF, in mM: 119 NaCl, 26NaHCO 3 , 2.5 KCl, 1MgCl 2 ,1.25NaH 2 PO 4 ,2CaCl 2 and 25 glucose, pH 7.4, 310 mOsm) at room temperature (25 °C) bubbled with 5% CO 2 and 95% O 2 for at least 1 hr before recording.
Patch clamp recordings. Whole-cell recording was performed as described previously [20][21] . Glass microelectrodes (Sutter, USA) were shaped by a pipette puller (P-97, Sutter, USA) with a resistance of 3-5 MΩ when filled with internal solution. The liquid junction potential was around − 10 mV, which was corrected before sealing. A Multi-clamp 700B amplifier (Axon, USA) was used for the recordings. Signals were sampled at 10 kHz and filtered at 2 kHz. A stable baseline was obtained for at least 5 min before experiments and data were discarded when the access resistance (15)(16)(17)(18)(19)(20) was changed by 20% at the end of recording.
The evoked EPSC currents were generated with a 40μ s pulse (0.1 Hz) from a stimulation isolation unit controlled by an AMPI generator (Master-8, USA). A bipolar stimulation electrode (FHC) was positioned ~50μ m rostral to the recording electrode in the same layer 16 . The internal solution contained (in mM):130Cs-methanesulfonate, 10HEPES, 10CsCl, 4NaCl, 1MgCl 2 , 1EGTA, 5NMG, 5MgATP, and 0.5Na 2 GTP and 12 phosphocreatine, pH 7.2, 275-290 mOsm. Bicuculine(10 μ M) was added to the bath solution to block GABA A receptors. The evoked currents at − 70 mV were identified as the AMPAR-mediated currents. Neurons were then voltage clamped at +40 mV, and the amplitude of the evoked EPSC 50 ms post-stimulus was considered as the NMDAR-mediated currents [22][23] . AMPA/NMDA ratio was calculated by AMPAR-mediated current (AMPA current) relative to NMDAR-mediated current (NMDA current). For the pure NMDA current recording, AMPAR antagonist CNQX (20 μ M) and bicuculine(10 μ M) were added in the ACSF when cells were voltage-clamped at +40 mV. For paired-pulse ratio recording, a paired-pulse protocol of two stimuli at an inter pulse interval of 50 ms was applied while the cells were voltage-clamped at − 70 mV. Paired-pulse ratio (PPR) was defined as the second peak amplitude (P2) divided by the first peak amplitude (P1).
For the measurement of inward rectification of AMPA current, spermine (100 μ M) was added in the internal solution, while bicuculine (10 μ M) and APV (50 μ M) were added in ACSF. AMPARs are heterotetramer composed of GluA1-A4 (also termed GluR1-R4). The GluA2 lacking (containing GluA1) receptors are permeable to Ca 2+ that can be blocked by polyamines such as spermine at positive potentials, thus resulting in a characteristic inwardly rectifying current-voltage (I-V) relationship [24][25][26] . Neurons were held at − 60, − 40, − 20, 0, +20 or +40 mV, respectively. The resultant inward rectification (IR) was calculated by dividing the absolute amplitude of average EPSC at − 60 mV by that at +40 mV 15,26 . Reagents. The following reagents (final concentration) were included in this study. Data analysis. Ten consecutive traces of evoked EPSC current were averaged to reach a final value.
All values were expressed as mean ± SEM. Paired Student's t-test was used for comparing nicotine effect in the same cells before and after application. Two way ANOVA and post-hoc testing was used to compare different groups. Mini Analysis program (Synaptosoft, Leonia, NJ) was used to analyze mEPSC amplitude and frequency. Individual synaptic events with fast onset and exponential decay kinetics were captured with threshold detectors in Mini Analysis software.

Results
Nicotine selectively increased AMPA current. We have previously reported that nicotine increases the amplitude and frequency of sEPSC in layer I neurons of PFC 14 . To identify the potential role of AMPAR and NMDAR, we first recorded AMPA and NMDA currents in the presence of nicotine. The amplitude of AMPA current, NMDA current and AMPA/NMDA ratio before nicotine treatment were 158.27 ± 11.79 pA, 247.92 ± 24.08 pA and 0.67 ± 0.05, respectively. Then, nicotine (5 μ M) was bath applied in ACSF for 10 minutes 14,27-28 . The resultant amplitude of AMPA current, NMDA current and AMPA/NMDA ratio were 220.02 ± 17.08 pA, 255.41 ± 24.72 pA and 0.89 ± 0.04, respectively. As shown in Fig. 1A,C, nicotine significantly increased the amplitude of AMPAR current (P < 0.001) and AMPA/ NMDA ratio (P < 0.001), but it did not cause significant change of NMDA current in PFC layer I neurons (P = 0.08, paired Student's t-test, n = 10). To clearly identify the effect of nicotine on pure NMDA component, evoked EPSC at + 40 mV was measured in the presence of AMPA antagonist CNQX (20 μ M). As shown in Fig. 1B, nicotine did not cause significant changes of pure NMDA current at + 40 mV, which was 136.38 ± 13.04 pA before and 132.61 ± 17.003 pA after nicotine application, respectively (P = 0.566, paired Student's t-test; n = 8). These results suggest that nicotine significantly increases APMA current, leaving the mixed NMDA or chemically isolated NMDA current unchanged.
To further determine the detailed effect of nicotine on evoked EPSC at different holding potentials, the time course experiments were performed. As shown in Fig. 1D, after nicotine application, the evoked EPSC gradually increased and reached a steady state in 15-25 min when the holding potential was at − 70, − 40, − 20 and +20 mV (160.20 ± 4.85%, n = 7; 141.89 ± 5.07%, n = 5; 138.01 ± 7.52%, n = 5; and 115.46 ± 3.53%, n = 5; respectively; P < 0.001 in each group, paired Student's t-test, data were collected at 20 min after nicotine). However, nicotine treatment did not change the evoked EPSC relative to basal level at + 40 or +60 mV (99.36 ± 1.37%, P = 0.26; 102.88 ± 2.31%, P = 0.30; respectively; paired Student's t-test, n = 5 in each group). These results indicated that although nicotine proportionally increased AMPA current at negative holding potentials, the nicotine effect was smaller at + 20 mV than that at − 20 mV and was lost at + 40 or +60 mV, suggesting that nicotine causes an inward rectification of AMPA current.
We also measured AMPA current in response to different concentrations of nicotine. As shown in Enhancement of AMPA current by nicotine was dependent on α7-nAChR. Layer I neurons express both α 7-and non α 7-nicotinic receptors 13 . To determine which nAChR subtype might be responsible for nicotine-induced increase of AMPA current, slices were pre-incubated with a selective α 4β 2-nAChR antagonist dihydro-β -erythroidine (DHβ E, 1 μ M) or a selective α 7-nAChR antagonist methyllycaconitine (MLA,10 nM) for one hour before recording. Figure 2A,B showed that DHβ E failed to abolish nicotine-induced enhancement of AMPA/NMDA ratio (0.87 ± 0.08 before and 1.08 ± 0.10 after nicotine, p < 0.001, paired Student's t-test, n = 7). MLA pre-incubation prevented nicotine effect on AMPA/NMDA ratios, which were 0.69 ± 0.09 before and 0.66 ± 0.09 after nicotine administration, respectively (p = 0.16, n = 6). These results suggest that nicotine increases AMPA/NMDA ratio via α 7-nAChR mediated mechanism in our experimental conditions. Activation nAChRs could promote Ca 2+ influx 29 . Thus we tested if nicotine effect on AMPA current was Ca 2+ -dependent. As shown in Fig. 2C, pipette inclusion of calcium chelator BAPTA (10 mM) blocked nicotine enhancement of AMPA/NMDA ratio (0.55 ± 0.07 before and 0.56 ± 0.07 after nicotine, P = 0.87, paired Student's t-test, n = 5). Previous study has indicated that systemic nicotine-induced enhancement of AMPA/NMDA ratio is dependent on NMDA receptor in dopamine neurons in the ventral tegmental area (VTA) 15 . Thus we also assessed nicotine effect in the presence of NMDAR inhibitor APV (50 μ M). As shown in Fig. 2D, the amplitude of AMPA current was 138.32 ± 14.75 pA before and 135.08 ± 12.68 pA after APV treatment, respectively (P = 0.37, paired Student's t-test, n = 6). However, subsequent application of nicotine in the presence of APV could still induce an augmentation of AMPA current (197.09 ± 25.28 pA; P = 0.009, APV +nicotine vs. APV alone, paired Student's t-test, n = 6). These results suggest that the enhancement of AMPA current is dependent on intracellular Ca 2+ but not NMDAR.
To further validate that the IR was mediated by CP-AMPAR 33-34 , we assessed nicotine effect on AMPA/NMDA ratio in the presence of a selective CP-AMPAR inhibitor IEM-1460 35 . As shown in Fig. 5D, nicotine caused an increase of AMPA/NMDA ratio in control condition (0.63 ± 0.06 before and 0.91 ± 0.07 after nicotine, P = 0.001, two way ANOVA followed by Bonferroni's LSD post hoc test, n = 5). Subsequent application of IEM-1460 alone (50 μ M) for 10 min caused a significant reduction of AMPA/ NMDA ratio (0.47 ± 0.04, P < 0.001, IEM-1460 vs. nicotine; p = 0.005, IEM-1460 vs. control, two way ANOVA followed by Bonferroni's LSD post hoc test, n = 5). However, the effect of nicotine was absent in the presence of IEM-1460 (0.42 ± 0.05, P = 0.47, IEM-1460+ nicotine vs. IEM-1460 alone, two way ANOVA followed by Bonferroni's LSD post hoc test, n = 5). The results suggest that nicotine enhancement of AMPAR or AMPA/ NMDA ratio is via an increase in postsynaptic CP-AMPAR.
ERK and calcineurin (CN) are two important molecules in nAChR signaling 37 . Thus we also tested the effect of their inhibitors on nicotine regulation of AMPA current. However, neither ERK inhibitor U0126 (10 μ M) nor CN inhibitor FK506 (10 μ M) was able to block nicotine enhancement of AMPA/NMDA ratios (Fig. 6C), which were 0.68 ± 0.05 before and 0.83 ± 0.06 after nicotine in U0126 (P < 0.001, paired Student's t-test n = 8); and 0.82 ± 0.15 before and 1.01 ± 0.16 after nicotine in FK506 (P = 0.003, paired Student's t-test, n = 6). These results indicated that ERK and CN were not involved in nicotinic regulation of AMPA current.

Discussion
The current study has shown that nicotine selectively increases the AMPA current in PFC layer I neurons. This effect is blocked by α 7-nAChR antagonist and is dependent on intracellular calcium. The characteristic inward rectification of CP-AMPAR suggests an involvement of GluA1 in nicotine augmentation of AMPAR current, which is validated by the use of synthetic peptides that are designed to block phosphorylation sites of GluA1.
The present study provides evidence that nAChR causes an increase in AMPAR current but not NMDAR current in layer I neurons. The unresponsiveness of NMDAR current in layer I is in contrast to what has been found in pyramidal neurons, in which nAChR activation causes a sustained reduction of NMDAR current involving ER calcium store and ERK activation 16 . Since almost all neurons in layer I are GABAergic 8 , our results suggest that different mechanisms might underlie NMDAR regulation between inhibitory and excitatory neurons. Indeed, ERK or calcineurin inhibitor does not block nicotinic regulation of AMPA current (Fig. 6C). With respect to AMPAR, a consistent response seems to be existed in both layer I and layer V pyramidal neurons (Fig. 5). Consistently, a recent study has shown that nicotine facilitates the expression of GluA1 containing AMPAR in hippocampal neurons 38 .
In layer I neurons, the mRNAs of α 4β 2-and α 7-nAChR are predominantly expressed 13 . And the selective antagonist of α 4β 2-or α 7-nAChR blocks nicotine enhancement of neuronal excitation in layer I neurons [13][14] , suggesting that both receptor subtypes mediate nicotinic regulation of neuronal activity in layer I. In our study however, the enhancement of AMPAR by nicotine is blocked by MLA but not DHβ E, suggesting that α 7-nAChR is crucial (Fig. 2B). It is reasonable that activation of α 7-nAChR might promote Ca 2+ influx and trigger Ca 2+ -dependent cellular processes including synaptic plasticity, neurotransmitter release, cell migration, and survival [39][40] . Consistently, intracellular inclusion of calcium chelator blocks nicotinic effect on AMPA current. However, this effect cannot be inhibited by NMDAR inhibitor (Fig. 2). Previous study has shown that NMDAR antagonist prevents systemic nicotine induced increase of AMPA/NMDA ratio in dopaminergic ventral tegmental area (VTA) 15 . In their study, increased AMPA/NMDA ratio in VTA neurons occurs in 1 hour later and lasts for 3 days, but not in 10 minutes after a single systemic administration of nicotine. And the NMDAR antagonist MK-801 is also injected intraperitoneally, which is different from our experimental condition. The exact mechanisms underlying these differences are currently unknown.
In our study, nicotine effect on mEPSC frequency and PPR is blocked by MLA but not Dhβ E, supporting a presynaptic role of α 7-nAChR. However, in the case of AMPAR regulation, the following might also suggest a postsynaptic function of α 7-nAChR. First, ultrastructural study reveals that at least in the hippocampus, α 7-nAChR is also located at dendritic spines in both GABAergic and glutamatergic neurons 30 . Second, the intracellular inclusion of calcium chelator, as well as GluA1 peptides, prevents nicotinic regulation of AMPA current, indicating postsynaptic events. This is further supported by the characteristic inward rectification of AMPA current, which reflects postsynaptic APMAR trafficking 15,26,31 . Finally, a recent study provides a direct evidence that postsynaptic α 7-nAChR contributes to GluA1 containing AMPAR accumulation on dendritic spines in the cultured hippocampal neurons 38 . Although it is a general consensus that mEPSC frequency reflects presynaptic release probability [41][42] , postsynaptic AMPAR might also play a role. Béïque and colleagues have shown that in PSD-95 (post-synaptic density 95) transfected organotypic brain slices, a larger AMPA/NMDA ratio is accompanied by increased AMPAR-mediated EPSC frequency but not amplitude, whereas the PPR is without change, suggesting no alterations of presynaptic release probability 43 . This could be explained by increased number of synapse rather than increased number or function of APMAR. The same research group demonstrates that PSD-95 KO mice exhibit greater occurrence of silent synapse 44 . It might be possible in our experiment that some of the neurons in layer I are silent under normal condition, which are then activated by nicotine after synaptic insertion of GluA1 containing AMPARs, resulting in increased frequency but not amplitude of mEPSC.
It is well documented that GluA1 contains phosphorylation sites of CAMKII, PKC and PKA, which regulate its synaptic insertion, while GluA2 trafficking is linked to NSF and adaptor protein AP2, among others [31][32] . Consistently, these kinase inhibitors as well as inhibitory peptides that block nicotine effect on AMPA/NMDA ratios further support a role of GluA1 (Fig. 6). Therefore, our study suggests that nicotine controls AMPAR current by regulating GluA1 phosphorylation state. Although PKC site is also seen GluA2, this phosphorylation leads to GluA2 internalization rather than synaptic insertion 45 .
Nicotinic effect on mEPSC (the action potential is inhibited) excludes the postsynaptic involvement of L-type voltage-dependent calcium channels; and nicotinic effect on AMPA current is not prevented by NMDAR inhibitor, which excludes other major calcium entry mediators that might play important role in this regulation. Therefore, we propose that nAChR activation through α 7-nAChR results in increased intracellular calcium, which in turn might activate CAMKII and PKC; and the resultant increase in phosphorylation of GluA1 leads to enhanced synaptic insertion and AMPAR current in layer I neurons. PKA phosphorylation of GluA1 seems to help "prime" GluA1 containing AMPARs for LTP, thus might maintain their retention at synaptic sites 46 . However, despite that PKA contributes to α 7-nAChR enhancement of glutamatergic neurotransmission 47 , direct evidence that links nAChR and PKA activation is sparse. On the other hand, inhibitors of calcineurin and ERK, two important nAChR downstream signaling molecules, fail to block nicotine regulation of AMPAR current (Fig. 6), suggesting that calcineurin and ERK do not play a role in this case. Rather, ERK 16 and calcineurin 48 might be involved in NMDAR regulation.
Clinical significance. AMPAR trafficking is involved in synaptic remodeling, such as long-term potentiation and depression. Growing evidence suggests that AMPAR-dependent synaptic plasticity is closely related to hippocampus-and amygdale-dependent learning and memory 49 . It is known that addictive drugs can cause behavioral alterations through AMPAR trafficking in the mesolimbic reward system including PFC and the ventral tegmetal area 5,49 . And the role of nicotine in cognition by acting on PFC networks has been well-documented [3][4] . Therefore, nicotinic regulation of AMPAR trafficking in layer I in our study might provide an important mechanism underlying nicotine related behaviors.