α₇-nAChR agonist enhances neural plasticity in the hippocampus via a GABAergic circuit

This article describes a potential circuit-level mechanism of action by which α₇-nAChR agonists increase synaptic plasticity and therefore may improve cognition. α₇-nAChR agonists have shown promise in treating cognitive impairment in Alzheimer disease and schizophrenia, and these data support continued research on these compounds.

Abstract

Agonists of the α₇-nicotinic acetylcholine receptor (α₇-nAChR) have entered clinical trials as procognitive agents for treating schizophrenia and Alzheimer's disease. The most advanced compounds are orthosteric agonists, which occupy the ligand binding site. At the molecular level, agonist activation of α₇-nAChR is reasonably well understood. However, the consequences of activating α₇-nAChRs on neural circuits underlying cognition remain elusive. Here we report that an α₇-nAChR agonist (FRM-17848) enhances long-term potentiation (LTP) in rat septo-hippocampal slices far below the cellular EC₅₀ but at a concentration that coincides with multiple functional outcome measures as we reported in Stoiljkovic M, Leventhal L, Chen A, Chen T, Driscoll R, Flood D, Hodgdon H, Hurst R, Nagy D, Piser T, Tang C, Townsend M, Tu Z, Bertrand D, Koenig G, Hajós M. Biochem Pharmacol 97: 576–589, 2015. In this same concentration range, we observed a significant increase in spontaneous γ-aminobutyric acid (GABA) inhibitory postsynaptic currents and a moderate suppression of excitability in whole cell recordings from rat CA1 pyramidal neurons. This modulation of GABAergic activity is necessary for the LTP-enhancing effects of FRM-17848, since inhibiting GABA_A α₅-subunit-containing receptors fully reversed the effects of the α₇-nAChR agonist. These data suggest that α₇-nAChR agonists may increase synaptic plasticity in hippocampal slices, at least in part, through a circuit-level enhancement of a specific subtype of GABAergic receptor.

NEW & NOTEWORTHY

This article describes a potential circuit-level mechanism of action by which α₇-nAChR agonists increase synaptic plasticity and therefore may improve cognition. α₇-nAChR agonists have shown promise in treating cognitive impairment in Alzheimer disease and schizophrenia, and these data support continued research on these compounds.

agonists of the α₇-nicotinic acetylcholine receptor (α₇-nAChR) are procognitive in many preclinical animal models (Medeiros et al. 2014; Stoiljkovic et al. 2015). Several agonists and positive allosteric modulators (PAMs) have advanced into clinical trials for treatment of cognitive impairment in disorders such as schizophrenia and Alzheimer's disease (AD) (Wallace and Porter 2011). Levels of α₇-nAChR mRNA and protein are reduced in the hippocampus, cerebral cortex, and thalamus of schizophrenia patients (Deutsch et al. 2005; Guillozet-Bongaarts et al. 2014), which may contribute to cognitive impairment that is not adequately treated by current therapies that treat positive and negative symptoms (Wallace and Porter 2011). A genetic linkage to the α₇-nAChR gene locus, which includes CHRNA7 and CHRFAM7A, has been identified in families with a history of schizophrenia, suggesting that defects in α₇-nAChR function may contribute to the etiology of the disease (Gault et al. 2003; Sinkus et al. 2015).

In AD, the loss of cholinergic neurons in the basal forebrain is an early hallmark of the disease and results in a deficit in acetylcholine (ACh) signaling throughout the cerebral cortex and hippocampus (Bartus et al. 1982; Whitehouse 1998). Acetylcholinesterase inhibitors (AChEIs), which slow the clearance of ACh during neurotransmission, remain the standard of care for patients with mild to moderate AD (Anand et al. 2014). While a direct link between α₇-nAChR dysfunction or dysregulation and AD is less well established (Neri et al. 2012), selectively restoring α₇-nAChR function with agonists may offer a novel, highly targeted, and well-tolerated approach to improving cognition in multiple central nervous system (CNS) indications including AD and schizophrenia (Toyohara and Hashimoto 2010; Wallace and Porter 2011).

The α₇-nAChR is a homopentameric ligand-gated ion channel (Drisdel and Green 2000). Upon binding its endogenous ligand ACh, the α₇-nAChR becomes selectively permeable to calcium and rapidly desensitizes (Albuquerque et al. 2009; Dani and Bertrand 2007). The relationships between α₇-nAChR activation and desired therapeutic effects on cognitive function are not fully understood. Using Xenopus laevis oocytes, agonists activate α₇-nAChR with cellular EC₅₀ values of >100 nM (Prickaerts et al. 2012), although these typically obtained EC₅₀ values may be artificially elevated 10-fold by the recording conditions employed (Papke and Thinschmidt 1998). Yet, lower concentrations of an α₇-nAChR agonist (at or below detectable α₇-nAChR activation and the binding K_i) can increase the response to the natural ligand ACh (Papke et al. 2011; Prickaerts et al. 2012; Quik et al. 1997; Stoiljkovic et al. 2015). This phenomenon, referred to as “priming,” was observed and modeled at neuromuscular nAChRs (Mukhtasimova et al. 2009). Cognition is improved in rodents with free-drug concentrations of α₇-nAChR agonist below the cellular EC₅₀ values (but similar to the priming concentrations) (Bitner et al. 2010; Prickaerts et al. 2012; Stoiljkovic et al. 2015; Werkheiser et al. 2011).

In the hippocampus, α₇-nAChRs are predominantly expressed on γ-aminobutyric acid (GABAergic) interneurons and to some extent on glutamatergic neurons and astrocytes (Fabian-Fine et al. 2001; Frazier et al. 1998; Halff et al. 2014; Sharma and Vijayaraghavan 2001). A selective α₇-nAChR agonist (PNU-282987) increased synaptic GABAergic activity in hippocampal slices (Hajos et al. 2005) by enhancing the excitability of interneurons (Hurst et al. 2005). Similar results were seen with another α₇-nAChR agonist S24795 (Lagostena et al. 2008), and nicotine (Alkondon et al. 1997). While it is clear that α₇-nAChRs are expressed on GABAergic interneurons and α₇-nAChR agonists have a pronounced effect on GABAergic synaptic transmission in the hippocampus, the functional consequences on cognition remain to be established.

To assess how α₇-nAChR agonists promote cognitive function at a cellular and circuit level, we have examined the pharmacology of FRM-17848, [(R)-7-cyano-N-quinuclidin-3-yl]benzo[b]thiophene-2-carboxamide, using electrophysiological recording techniques in rat brain septo-hippocampal slice preparations. In this preparation, measures of synaptic strength and plasticity such as long-term potentiation (LTP) are widely considered a model for the cellular basis of learning and memory (Izquierdo 1994). Previous studies have demonstrated that α₇-nAChR agonists or PAMs can enhance LTP in the hippocampus of rodents (Kroker et al. 2011; Ondrejcak et al. 2012). This enhancement is absent in CHRNA7 knockout mice (Biton et al. 2007; Lagostena et al. 2008) and α₇-nAChR antagonists such as methylcaconitine (MLA) fail to enhance LTP, indicating that agonists are activating rather than desensitizing α₇-nAChRs (Griguoli et al. 2013).

We recently reported that the procognitive effects of an α₇-nAChR agonist could be accounted for in terms of the concentration-response function of α₇-nAChR receptor pharmacology (Stoiljkovic et al. 2015). The bell-shaped efficacy dose response was shown to align with the free drug levels across assays (oocytes, brain slice LTP, in vivo theta-power, and rodent cognition). In this study, we extend those findings by examining how an α₇-nAChR agonist enhances LTP in the hippocampal brain circuit at select concentrations. We demonstrate that the α₇-nAChR-mediated enhancement of synaptic plasticity (LTP) depends on increased GABAergic neurotransmission that is mediated by GABA_A α₅-receptors (GABA_A α₅R). These data indicate that priming and procognitive concentrations of α₇-nAChR agonists promote synaptic plasticity, at least in part, through a circuit-level enhancement of the activity of a specific subtype of GABAergic receptor.

METHODS

Reagents.

FRM-17848 (FRM-2 in Tang et al. 2014) was synthesized by SAI Life Sciences (Hyderabad, India) and was prepared as 31.6-μM stock solutions in dimethyl sulfoxide (DMSO). The GABA_A α₅R inhibitor (hydroxypropylthio-derivative of MRK-536 and named FRM-35440 herein) (Atack 2011) and the GABA_A α₅R PAM, SH-053-2′F-R-CH3 (Drexler et al. 2013), were synthesized by WuXi Apptech (MaShan, China). Bicuculline, 2,3-dioxo-6-nitro-1,2,3,4-tetrahydrobenzo[f]quinoxaline-7-sulfonamide (NBQX), dl-2-amino-5-phosphonopentanoic acid sodium salt (d-AP5), and CGP-55845 were purchased from Abcam (Bristol, UK). MLA and donepezil were purchased from Tocris Biosciences (Bristol, UK) and dissolved in DMSO. All compounds were stored at −20°C and diluted to the required concentrations in artificial cerebrospinal fluid (aCSF; composition in mM: 127.0 NaCl, 1.6 KCl, 1.24 KH₂PO₄, 1.3 MgSO₄, 2.4 CaCl₂, 26.0 NaHCO₃, and 10 d-glucose) immediately before use. All other reagents were obtained from Sigma-Aldrich.

Animals.

Male Sprague-Dawley rats (5–8 wk of age) were obtained from Charles River UK (Kent, UK) and were maintained on a 12:12-h light-dark cycle with free access to food and water. All experiments were approved by the Cerebrasol Insitutional Animal Care and Use Committee (IACUC) and were carried out in compliance with the UK Animals (Scientific Procedures) Act, 1986, including the recent revision incorporating the European Directive 2010/63/EU on the protection of animals used for scientific purposes.

Female Xenopus laevis frogs were obtained from Xenopus Express (Haute-Loire, France) and used for in vitro oocyte electrophysiology. All experiments were approved by the local IACUC and were carried out in compliance with the European Union regulations (Directive 2010/63/EU).

Characterization of FRM-17848 at α₇-nAChRs.

Binding assays at rat α₇-nAChR receptors were performed at Perkin-Elmer Discovery Services (formerly Caliper Life Sciences/Novascreen, Hanover, MD) according to their standard protocols, which follow published methods (Marks et al. 1986; Meyer et al. 1998). Briefly, rat brains were rapidly removed, homogenized in buffer, and prepared for incubation with the radioactive ligand [¹²⁵I]-α-bungarotoxin. The reaction was terminated by diluting with buffer, followed immediately by filtration through glass fiber filters soaked in buffer containing polyethylenimine. Binding of the radioactive ligand was measured using a scintillation counter. Nonspecific binding was determined with unlabeled ligand. Each condition was measured in duplicate. The reference compound was MLA. K_i values were calculated by the equation of Cheng and Prusoff (1973).

Electrophysiological experiments were carried out with human α₇-nAChR receptors expressed in Xenopus laevis oocytes. Oocytes were prepared, injected with cDNA encoding for the α₇-nAChR, and recorded using standard procedures (Hogg et al. 2008; Stoiljkovic et al. 2015). Oocytes were grown in the presence of penicillin, and all recordings were made in antibiotic-free OR2 medium containing the following (in mmol/l): 88.5 NaCl, 2.5 KCl, 5 HEPES, 1.8 CaCl₂·2H₂O, and 1 MgCl₂·6H₂O, pH 7.4 at 18°C. Four oocytes were used in generating each set of data from at least two preparations.

Field potential recordings for LTP.

Detailed methods of the septo-hippocampal slice preparation and field potential recording are described in Stoiljkovic et al. (2015). For all electrophysiology experiments, the experimenters were blinded to the identity of the compounds.

Whole cell voltage/current clamp.

Whole cell patch-clamp recordings were performed from pyramidal neurons of the CA1 region of the hippocampus at room temperature (17–21°C) using the “blind” version of the patch-clamp technique and Axopatch 1D or Multiclamp 700B amplifiers (Molecular Devices). Patch pipettes were pulled from thin-walled borosilicate glass (GF150TF-10; Harvard Apparatus) with resistances of 3–8 MΩ when filled with intracellular solution of the following composition (in mM): 140 potassium gluconate, 10 KCl, 1 EGTA-Na, 10 HEPES, 2 Na₂ATP, and 0.3 GTP (Sigma-Aldrich). All recorded neurons were morphologically confirmed as pyramidal neurons post-experiment via passive diffusion of Lucifer yellow (1 mg/ml) from the pipette recording solution into the recorded cell and fixation of the hippocampal slice tissue overnight in 4% paraformaldehyde followed by 0.1 M phosphate buffer (pH 6.8) for 24–48 h before visualization under a fluorescent microscope (Zeiss) (data not shown).

A 10-min stable baseline was established for each pyramidal neuron. Inhibitory postsynaptic currents (IPSCs) and inhibitory postsynaptic potentials (IPSPs) were isolated by the addition of 10 μM NBQX and 10 μM AP-5 to block glutamatergic receptors. FRM-17848 (with or without FRM-35440) or donepezil was applied for 10 min followed by compound washout for 20 min. In control experiments, 50 nM MLA alone or MLA + FRM-17848 was perfused over slices after isolation of IPSCs/IPSPs. A final 5-min application of 10 or 20 μM bicuculline (as indicated), a GABA_A receptor antagonist, was used to demonstrate that the remaining recorded currents were GABAergic. Compounds were administered to the slices by bath perfusion from 50-ml syringes arranged in series with the main perfusion line from the aCSF reservoir via three-way valves. Maximum final DMSO concentration was 0.3%.

Profiling FRM-35440 selectivity.

The activity of FRM-35440, a GABA_A α₅R inhibitor, was tested at ChanTest/Charles River (Cleveland, OH) using HEK-293 cells expressing GABA_A α₁-, α₂-, α₃-, α₄-, or α₅-subunits with β₃/γ₂-subunits using an IonWorks Barracuda system (Molecular Devices). The intracellular solution contained the following (in mM): 50 CsCl, 90 CsF, 5 MgCl₂, 1 EGTA, and 10 HEPES, pH adjusted to 7.2 with CsOH. In preparation for a recording session, the intracellular solution was loaded into the intracellular compartment of a planar patch-clamp electrode. The extracellular solution contained the following (in mM): 137 NaCl, 4.0 KCl, 1.8 CaCl₂, 1 MgCl₂, 10 HEPES, and 10 d-glucose, pH adjusted to 7.4 with NaOH. Two recordings (scans) were performed. First the test article was added to measure agonist effects. Five minutes later, a second scan was recorded during the application of GABA (30 μM ∼ EC₉₀) to detect antagonism. Percent inhibition was calculated as (1 − Current_FRM-35440/Current_{GABA EC90}) × 100%.

Analysis.

Data were captured online and stored on a computer running pCLAMP data acquisition software for later offline analysis. Analysis of all data was carried out using Clampfit (Molecular Devices) and Microsoft Excel software. IPSP/IPSC frequencies were calculated from 3-min continuous recordings. LTP data are expressed as the mean ± SE. Statistical analysis was performed using ANOVA followed by Dunnett's multiple comparisons test. For studies with sequential treatments, a repeated-measures ANOVA was applied to the data followed by a Bonferroni multiple comparisons test (PRISM 6, GraphPad Software, San Diego, CA) with P < 0.05 taken to indicate statistical significance. All other data are expressed as the mean ± SD.

RESULTS

FRM-17848 enhances LTP in a bell-shaped concentration-response curve.

In Stoiljkovic et al. (2015), it was established that the α₇-nAChR agonist FRM-17874 enhanced hippocampal LTP at concentrations below the K_i for displacement of α-bungarotoxin. A similar compound FRM-17848 (K_i = 11 nM for rat α₇-nAChR, EC₅₀ = 455 nM) was tested over a wider range of concentrations for effects on LTP in rat septo-hippocampal slices. As shown in Fig. 1, A and C, 3.16 nM FRM-17848 induced a significant enhancement of LTP, measured during the 50- to 60-min interval following theta-burst stimulation (TBS), compared with vehicle-treated slices. The activity of this single concentration was confirmed in an additional study (see Fig. 6D), and no other tested concentration significantly enhanced LTP compared with control. This narrow efficacious window is consistent the results in Stoiljkovic et al. (2015) that showed a similar molecule (FRM-17874) to be active only at 3.2 and 5 nM (Tang et al. 2014).

Fig. 1.

The α₇-nicotinic acetylcholine receptor (α₇-nAChR) agonist FRM-17848 enhances long-term potentiation (LTP) in a narrow concentration range. Extracellular recordings were made of the field excitatory postsynaptic potential (fEPSP) amplitude in the CA1 region of rat septo-hippocampal slices. Theta-burst stimulation (TBS) of the Schaffer collateral afferents evoked stable LTP in all treatment groups. The average fEPSP amplitude is shown from slices treated with vehicle (control), 3.16 nM FRM-17848 (A), and 500 nM donepezil (B). The period of compound application is indicated by the black bar. C: each point in the scatterplot depicts the fEPSP amplitude averaged from 50–60 min post-TBS in brain slices treated with the indicated compounds. At 50–60 min post-TBS, both 3.16 nM FRM-17848 and 500 nM donepezil significantly increased LTP over vehicle (vehicle = 129 ± 4% SE, n = 16; 3.16 nM FRM-17848 = 148 ± 4% SE, n = 11; and 500 nM donepezil = 145 ± 6% SE, n = 7; ANOVA, Dunnett's multiple comparison test compared with vehicle, *P < 0.05). D: α₇-nAChR antagonist MLA blocked the enhancement of LTP by 3.16 nM FRM-17848. Coapplication of 50 nM MLA partially blocked the enhancement of LTP by 3.16 nM FRM-17848, while 100 nM MLA fully blocked the effect. E: the scatterplot depicts the average fEPSP amplitude at 50–60 min as a percent of baseline (vehicle = 129 ± 4% SE, n = 16; 50 nM MLA = 132 ± 5% SE, n = 8; 3.16 nM FRM-17848 = 148 ± 4% SE, n = 11; FRM-17848 + 50 nM MLA = 137 ± 3% SE, n = 8; FRM-17848 + 100 nM MLA = 131 ± 7% SE, n = 6; ANOVA, Dunnett's multiple comparison test for vehicle vs. 3.16 nM FRM-17848, *P < 0.05, and 3.16 nM FRM-17848 vs. 100 nM MLA + 3.16 nM FRM-17848, #P < 0.05).

Fig. 3.

Spontaneous inhibitory postsynaptic current (IPSC) frequency and amplitude is increased by 3.16 nM FRM-17848. A: the input resistance for an individual CA1 pyramidal neuron was measured before and after application of 3.16 nM FRM-17848. Compound application had no significant effect on input resistance [change in input resistance (ΔIR) = +5.4 ± 14.9 MΩ, repeated-measures ANOVA, Bonferroni's multiple comparison test, P = 0.14, n = 18]. B: the membrane potential became significantly more hyperpolarized upon application of 3.16 nM FRM-17848 [change in membrane potential (ΔV_m) = −1.4 ± 2.0 mV, repeated-measures ANOVA, Bonferroni's multiple comparison test, *P < 0.01, n = 38]. This hyperpolarization was not observed with 5.6 nM FRM-17848 (ΔV_m = +0.4 ± 3.9 mV, n = 8, data not shown) or when 50 nM MLA was coapplied with 3.16 nM FRM-17848 (ΔV_m = +0.3 ± 4.2 mV, n = 4, data not shown). C: spontaneous IPSC activity was recorded by inhibiting glutamatergic currents with 10 μM AP-5 and 10 μM NBQX. Three-minute time blocks are shown from a pyramidal neuron excerpted from a continuous recording. Addition of 3.16 nM FRM-17848 to the perfusion resulted in an increase in IPSC frequency which was prevented by 50 nM MLA. The measured currents were GABAergic, as shown by the absence of IPSCs after addition of 10 μM bicuculline. V_h = −40 mV. D: the scatterplot depicts the IPSC frequency averaged over 3 min from 8 pyramidal neurons as they were subjected to sequential treatment with compounds. The addition of the α₇-nAChR agonist FRM-17848 (3.16 nM) resulted in a significant increase in IPSC frequency (baseline = 0.48 Hz ± 0.13, n = 8; 3.16 nM FRM-17848 = 0.70 Hz ± 0.15; wash = 0.41 Hz ± 0.06; repeated-measures ANOVA, Bonferroni's multiple comparison test, *P < 0.05 vs. baseline). Coapplication of the α₇-nAChR antagonist MLA fully inhibited the effect of 3.16 nM FRM-17848 on IPSC frequency (50 nM MLA = 0.47 Hz ± 0.08; 50 nM MLA + 3.16 nM FRM-17848 = 0.46 Hz ± 0.09, n = 3; repeated-measures ANOVA, Bonferroni's multiple comparison test, P = 0.97). Inhibitory postsynaptic potentials (IPSPs) frequency was similarly affected by FRM-17848 (baseline = 0.17 Hz ± 0.04; 3.16 nM FRM-17848 = 0.27 Hz ± 0.06; wash = 0.19 Hz ± 0.05, n = 15, data not shown). E: as with frequency, IPSC amplitude was increased in individual neurons upon wash-in of 3.16 nM FRM-17848 (baseline = 43.9 pA ± 26.8, n = 8; 3.16 nM FRM-17848 = 95.6 pA ± 67.4; wash = 53.3 pA ± 33.9; repeated-measures ANOVA, Bonferroni's multiple comparison test, *P < 0.05 vs. baseline; 50 nM MLA = 36.7 pA ± 26.0; 50 nM MLA + 3.16 nM FRM-17848 = 37.6 pA ± 24.6, n = 3; repeated-measures ANOVA, Bonferroni's multiple comparison test P = 0.71). F: a sample recording of IPSCs from an individual neuron reveal that application of a higher concentration of FRM-17848 (5.6 nM) had no effect on IPSC frequency (V_h = −40 mV). G: the scatterplot depicts the IPSC frequency (averaged over 3 min) of 8 cells upon addition and washout of 5.6 nM FRM-17848 (baseline = 0.31 Hz ± 0.14; 5.6 nM FRM-17848 = 0.29 Hz ± 0.14; wash = 0.33 Hz ± 0.13; n = 8; repeated-measures ANOVA, Bonferroni's multiple comparison test, P = 0.19).

Fig. 6.

A GABA_A α₅R inhibitor (FRM-35440) prevents the LTP-enhancing effect of the α₇ nAChR agonist FRM-17848. A: FRM-35440 acted as a selective and potent inhibitor of GABA_A α5Rs. FRM-35440 (hydroxypropylthio- derivative of MRK-536) was profiled on HEK293 cells expressing GABA_A α₁-, α₂-, α₃-, α₄-, or α₅-subunits with β₃/γ₂-subunits using an IonWorks Barracuda system. Peak currents were recorded in quadruplicate during agonist and antagonist scans. No intrinsic agonist activity of FRM-35440 was detected at 0.1–300 nM in the absence of GABA. The subsequent application of 30 μM GABA revealed that FRM-35440 concentration-dependently inhibited the evoked GABA_A α₅R current (IC₅₀ = 158 nM) under these conditions. The activity of FRM-35440 may be more potent in brain slices than in recombinant cells (Farzampour et al. 2015). B: FRM-35440 was inactive on the other α₁-α₄ GABA_A R subtypes up to 3 μM. C: field potential recordings were used to measure LTP in the CA1 region of hippocampal slices. At a concentration of 50 nM, the GABA_A α₅R inhibitor FRM-35440 alone had no effect on TBS-induced LTP (control = 134 ± 4% SE, n = 7, 50 nM FRM-35440 = 133 ± 6% SE, n = 4). D: coapplication of FRM-35440 prevented the enhancement of LTP induced by 3.16 nM FRM-17848 in a concentration-dependent manner. E: each circle reports the LTP (average fEPSP amplitude from 50–60 min post-TBS) recorded from an individual slice. FRM-35440 inhibited the LTP-enhancing effects of FRM-17848 (control = 134 ± 4% SE, n = 7, 3.16 nM FRM-17848 = 151 ± 6% SE, n = 6; 3.16 nM FRM-17848 + 5 nM FRM-35440 = 143 ± 3% SE, n = 8; 3.16 nM FRM-17848 + 50 nM FRM-35440 = 132 ± 4% SE, n = 8; 50 nM FRM-35440 = 133 ± 6%, n = 4; ANOVA, Dunnett's multiple comparison test, *P < 0.05 for 3.16 nM FRM-17848 vs. vehicle, #P < 0.05 for 3.16 nM FRM-17848 vs. 3.16 nM FRM-17848 + 50 nM FRM-35440).

The AChE inhibitor donepezil was used as a positive control for enhancement of LTP (Fig. 1, B and C) (Kapai et al. 2012). At 500 nM, donepezil showed a significant enhancement of LTP that was comparable to 3.16 nM FRM-17848. Neither compound had an effect on baseline (untetanized) field excitatory postsynaptic potential (fEPSP) amplitude (data not shown). The 3.16-nM concentration of FRM-17848 is comparable to the free drug levels of similar α₇-nAChR agonists such as FRM-17848 and EVP-6124, which have proven benefits on cognition (Keefe et al. 2015; Prickaerts et al. 2012; Stoiljkovic et al. 2015).

To confirm that the enhancement of LTP by FRM-17848 was mediated by α₇-nAChR, experiments were repeated with the selective α₇-nAChR antagonist MLA. At a concentration of 100 nM MLA, the effect of 3.16 nM FRM-17848 was fully inhibited (Fig. 1, D and E). This result confirmed that the enhancement of LTP by 3.16 nM FRM-17848 was through activation of α₇-nAChRs rather than desensitization.

3.16 nM FRM-17848 primes α₇-nAChRs.

To examine the effects of 3.16 nM FRM-17848 on α₇-receptor pharmacology, whole cell recordings were made from Xenopus oocytes expressing human α₇-nAChRs. The addition of 3.16 nM FRM-17848 to the perfusion enhanced the α₇-nAChR response to the subsequent application of 40 μM ACh (Fig. 2, A and B). This phenomenon described as priming in Stoiljkovic et al. (2015) establishes a concentration-response function that is proven to be highly predictive of α₇-nAChR agonist efficacy in LTP and cognition (Prickaerts et al. 2012). Thus LTP in rat hippocampal slices is enhanced at a concentration of FRM-17848 that primes α₇-nAChRs.

Fig. 2.

3.16 nM FRM-17848 can prime α₇-nAChRs. Electrophysiological recordings were made from Xenopus oocytes overexpressing human α₇-nAChRs. A: sample trace from a single oocyte, shows the response to 40 μM ACh before and during the coapplication of the α₇-nAChR agonist (3.16 nM FRM-17848). Cells were pulsed with 40 μM ACh for 5 s at 2- or 10-min intervals. B: the histogram shows the average response from 4 oocytes. A stable baseline was established with 4 pulses of 40 μM ACh (A). Addition of 3.16 nM FRM-17848 to the bath enhanced the ACh-evoked currents recorded at 2-min intervals (3 currents, B) and then 10-min intervals (2 currents, C and D). Finally, FRM-17848 was removed from the bath and an additional 3 ACh-evoked currents were recorded at 2-min intervals (E). The addition of FRM-17848 increased the response to ACh, enhancing the peak ACh-evoked current [40 μM ACh (A) = 1.0 ± 0.06; 40 μM ACh + 3.16 nM FRM-17848 (B) = 1.63 ± 0.22; 40 μM ACh + 3.16 nM FRM-17848 + 10 min (C) = 1.86 ± 0.32; 40 μM ACh + 3.16 nM FRM-17848 + 20 min (D) = 2.16 ± 0.30; 40 μM ACh washout (E) = 1.74 ± 0.19; repeated-measures ANOVA, Bonferroni's multiple comparison test, all treatments were significantly different from 40 μM ACh (A; P < 0.01)]. The priming effect persisted for a short interval after removal of FRM-17848 from the bath.

FRM-17848 hyperpolarizes hippocampal CA1 pyramidal neurons.

To understand the effects of FRM-17848 on cellular physiology, whole cell recordings were made from CA1 pyramidal cells in septo-hippocampal slices. For individual neurons, wash-in of 3.16 nM FRM-17848 did not significantly change the input resistance from baseline (Fig. 3A). In contrast, application 3.16 nM FRM-17848 induced a small, but highly significant hyperpolarization of the membrane potential in most cells (Fig. 3B). This result was initially unexpected, because in normal physiological saline, α₇-nAChRs have a reversal potential of ∼0 mV. Thus activation of α₇-nAChRs on hippocampal CA1 pyramidal neurons would be expected to be depolarizing rather than hyperpolarizing.

FRM-17848 increases GABAergic IPSCs.

Based on previous work showing that α₇-nAChRs are predominantly expressed on GABAergic interneurons in the hippocampus and our new finding that α₇-nAChR activation induced a hyperpolarization of pyramidal neurons, we recorded IPSCs from pyramidal neurons. Glutamatergic currents were inhibited with NBQX and AP-5, leaving bicuculline-sensitive IPSCs. Addition of 3.16 nM FRM-17848 induced an increase in IPSC frequency within minutes (Fig. 3C). This effect was reversed by a 20 min washout of FRM-17848 and was inhibited by coapplication of 50 nM MLA (Fig. 3, C and D). IPSC amplitude was similarly affected in individual neurons upon wash-in of 3.16 FRM-17848 (Fig. 3E), as were IPSPs (data not shown). Thus a concentration of FRM-17848 that was found to prime the α₇-nAChR and enhance LTP also increased IPSC frequency and amplitude in hippocampal slices.

The results from the LTP experiments identified a narrow concentration-response range of FRM-17848 with 3.16 nM being the only active concentration. Since 5.6 nM FRM-17848 was inactive in LTP, this concentration was tested for its effects on spontaneous IPSCs. As shown in Fig. 3, F and G, 5.6 nM FRM-17848 had no effect on the frequency (or amplitude, data not shown) of spontaneous IPSCs. Therefore, 3.16 nM FRM-17848 was found to enhance both LTP and increase spontaneous IPSC frequency, while 5.6 nM FRM-17848 affected neither.

Fig. 5.

Pyramidal neurons are less excitable upon application of 3.16 nM FRM-17848. Whole cell current-clamp recordings were made from CA1 pyramidal neurons. Based on the salt concentrations in the electrode and artificial cerebrospinal fluid (aCSF), the chloride reversal potential was predicted to be −63 mV. In a sample recording, depolarizing square wave current injection in pyramidal neurons induced a series of accommodating action potentials. Perfusion with 3.16 nM FRM-17848 hyperpolarized the resting membrane potential (V_m) and reduced action potential firing, which was restored after a 10 min washout period (baseline = 5.7 ± 3.6 Hz; 3.16 nM FRM-17848 = 3.9 ± 3.3 Hz; wash = 4.9 ± 3.5 Hz; n = 11; repeated-measures ANOVA, Bonferroni's multiple comparison test baseline vs. 3.16 nM FRM-17848, *P < 0.05). Addition of 50 nM MLA increased action potential frequency which was unaffected by coapplication of 3.16 nM FRM-17848.

Donepezil increases GABAergic IPSCs.

Observing that the same concentration of FRM-17848 increased LTP and IPSC frequency, we considered whether the correlation would also apply to donepezil. To test this hypothesis, IPSC frequency was measured in response to ascending concentrations of donepezil (50–500 nM, Fig. 4A). There was a trend to increased IPSC frequency beginning at 50 nM donepezil. However, only 500 nM donepezil, the same concentration that significantly enhanced LTP, showed a statistically significant increase in IPSC frequency (Fig. 4B). These data demonstrate that two distinct cognition-enhancing cholinergic mechanisms (AChE inhibitor and α₇-nAChR agonist) each modulate IPSC frequency at the same concentrations that enhance LTP.

Fig. 4.

Donepezil increases IPSC frequency at concentrations that enhanced LTP. Pyramidal neurons were voltage clamped and IPSCs were isolated with 10 μM NBQX and 10 μM AP-5. A: sequential recordings from an individual pyramidal neuron show that step increases in donepezil concentration cause an increase in spontaneous IPSC frequency. Washout of donepezil restored the IPSC frequency to baseline and bicuculline eliminated the currents confirming their GABAergic origin (V_h = −50 mV). B: the scatterplot depicts the IPSC frequency averaged over 3 min for 9 cells as each was treated with ascending concentrations of donepezil. Only 500 nM donepezil produced a statistically significant increase in IPSC frequency compared with baseline (baseline = 0.21 Hz ± 0.11; 50 nM donepezil = 0.32 Hz ± 0.21; 100 nM donepezil = 0.34 Hz ± 0.24; 250 nM donepezil = 0.33 Hz ± 0.19; 500 nM donepezil = 0.43 Hz ± 0.29; wash = 0.21 Hz ± 0.13; n = 9; repeated-measures ANOVA, Bonferroni's multiple comparison test, *P < 0.05 vs. baseline). Two of the recorded cells showed no change in IPSC frequency in response to any concentration of donepezil.

Increased GABA IPSCs modestly hyperpolarize pyramidal neurons.

To determine the consequences of increased IPSC activity on the hippocampal circuitry, current-clamp recordings were used to investigate the effects of FRM-17848 on the firing properties of pyramidal neurons. A current-step protocol was applied to induce action potential firing (Fig. 5). Wash-in of 3.16 nM FRM-17848 reduced the firing activity triggered by a depolarizing current injection, and the effect could be reversed upon washout of compound. The α₇-nAChR antagonist MLA had the opposite effect, eliciting an increase in spike activity in response to a depolarizing current step. Importantly, 50 nM MLA + 3.16 nM FRM-17848 exhibited the same effect as MLA alone. GABAergic inhibitory neurons are known to exert a tonic hyperpolarizing effect on pyramidal cells (Caraiscos et al. 2004). These results support the conclusion that by increasing IPSC activity, 3.16 nM FRM-17848 induces a modest hyperpolarization of the resting membrane potential, thereby making pyramidal neurons less excitable.

GABA_A α₅R mediation of α₇-nAChR agonist effects.

We then considered whether the effects of an α₇-nAChR agonist on GABAergic synaptic activity may be directly linked to the enhancement of LTP. This hypothesis appears counterintuitive, since FRM-17848 induced a hyperpolarization of pyramidal neurons and reduced excitability. It is well known that benzodiazepines, which are nonselective PAMs of GABA_A receptors, inhibit LTP and are sedatives in humans (del Cerro et al. 1992). Therefore, as a procognitive agent that increases GABAergic activity, α₇-nAChR agonists could not simply mimic the activity of benzodiazepines and nonspecifically increase GABAergic activity. Previous work has demonstrated that selective modulators of GABA_A α₅Rs can be procognitive (Drexler et al. 2013; Johnstone et al. 2011; Koh et al. 2013). Therefore we postulated that α₇-nAChR agonists enhance LTP by activating a subtype of GABAergic interneuron or receptor in the neural circuit.

To test this, we identified a hydroxypropylthio- variant of MRK-536 (herein called FRM-35440) that was reported to act as a highly selective and potent partial agonist at GABA_A α₅Rs (K_i = 2 nM). As a partial agonist, FRM-35440 suppresses the full activation of the receptor by GABA (Atack 2011). Measuring Cl⁻ currents in HEK cells overexpressing GABA_A α_1–5, FRM-35440 functions as a selective antagonist of the GABA_A α₅R under these conditions (Fig. 6, A and B). FRM-35440 showed no binding activity at α₇-nAChRs (IC₅₀ > 10 μM, data not shown). When tested in LTP experiments, 50 nM FRM-35440 alone had no effect on LTP (Fig. 6C). Thus the activity of FRM-35440 at the GABA_A α₅R did not disrupt the basic mechanism of TBS-induced potentiation. As in Fig. 1, 3.16 nM FRM-17848 was again found to enhance LTP. Strikingly, the coapplication of FRM-17848 with 50 nM FRM-35440 fully inhibited the increase in LTP by 3.16 nM FRM-17848, while a lower concentration of FRM-35440 (5 nM) partially inhibited (Fig. 6, D and E).

Based on this result, we predicted that GABA_A α₅R inhibition by FRM-35440 should also disrupt the effect of 3.16 nM FRM-17848 on IPSC frequency. As shown in Fig. 7, A and B, treatment with 50 nM FRM-35440 modestly suppressed the frequency of spontaneous IPSCs recorded in CA1 pyramidal neurons. Maintenance of the majority of IPSCs is consistent with the prevalence of other GABA_AR subtypes, such as GABA_A α₁R in this circuit, that remain active in the presence of the GABA_A α₅R subtype-specific FRM-35440. Subsequent application of 3.16 nM FRM-17848, had no effect on IPSC frequency. Washout of both compounds restored IPSC frequency to baseline. FRM-35440 also fully inhibited the increase of IPSC amplitude by 3.16 nM FRM-17848 (Fig. 7C) without affecting total IPSC amplitude. These data indicate that activating α₇-nAChRs at priming concentrations of FRM-17848 causes an enhancement of LTP and IPSC frequency and that this effect can be reversed by coapplication of a GABA_A α₅R inhibitor.

Fig. 7.

The GABA_A α₅R inhibitor (FRM-35440) prevents the increase in IPSC frequency caused by the α₇-nAChR agonist FRM-17848. A: a sample trace from an individual pyramidal neuron, shows that preapplication of FRM-35440 blocked the increase in IPSC frequency induced by the subsequent addition of FRM-17848. B: quantification of the IPSC frequency in 5 pyramidal neurons demonstrates that the GABA_A α₅R inhibitor FRM-35440 modestly reduced IPSC frequency, and prevents the increase in IPSC frequency that occurs with 3.16 nM FRM-17848 alone. Washout of both compounds restored IPSC frequency to baseline levels (baseline = 0.25 ± 0.07 Hz; 50 nM FRM-35440 = 0.20 ± 0.07 Hz; 50 nM FRM-35440 + 3.16 nM FRM-17848 = 0.21 ± 0.08 Hz; washout = 0.25 ± 0.09 Hz, n = 5; repeated-measures ANOVA, Bonferroni's multiple comparisons test, *P < 0.05 vs. baseline). C: a similar result was observed with IPSC amplitude (baseline = 30.3 ± 12.6 pA; 50 nM FRM-35440 + 3.16 nM FRM-17848 = 30.9 ± 12.0 pA; washout = 31.2 ± 12.0 pA, n = 5; repeated-measures ANOVA, Bonferroni's multiple comparisons test, P > 0.05).

Since FRM-35440 suppressed the effects of the α₇-nAChR agonist FRM-17848, we tested whether a selective GABA_A α₅R PAM (SH-053-2'F-R-CH3) would be sufficient to enhance LTP. A concentration of 250 nM was chosen based on the reported selectivity of SH-053-2'F-R-CH3 at GABA_A α₅Rs (Savíc et al. 2010). As shown in Fig. 8A, a 250-nM concentration of SH-053-2'F-R-CH3 significantly enhanced LTP measured during the 50- to 60-min interval post-TBS. This effect was concentration dependent (Fig. 8B). These data demonstrate that potentiating the activity of the GABA_A α₅Rs in this hippocampal circuit is sufficient to promote synaptic plasticity.

Fig. 8.

SH-053-2'F-R-CH3, a selective GABA_A α₅R PAM, enhanced LTP in septo-hippocampal brain slices. A: field potential recordings were made in CA1 of rat septo-hippocampus slices. Addition of 250 nM SH-053-2'F-R-CH3 enhanced LTP similar to FRM-17848. B: the scatterplot depicts the fEPSP amplitude averaged from the 50–60 min post-TBS interval from individual slices. SH-053-2'F-R-CH3 exhibits a concentration-dependent enhancement of LTP (vehicle = 123 ± 4% SE, n = 7; 50 nM SH-053-2'F-R-CH3 = 121 ± 4% SE, n = 6; 100 nM SH-053-2'F-R-CH3 = 129 ± 5% SE, n = 6; 250 nM SH-053-2'F-R-CH3 = 135 ± 4% SE, n = 7; Dunnett's multiple comparisons test compared with vehicle, *P < 0.05).

DISCUSSION

Accumulating evidence suggests that selective activation of α₇-nAChRs may be an effective therapy for enhancing cognitive function in patients with schizophrenia and AD. Previous reports have demonstrated that α₇-nAChR agonists can enhance LTP in rodent brain slices with a bell-shaped concentration-response relationship and a narrow effective concentration range (Kroker et al. 2011; Lagostena et al. 2008). The effective concentration range for LTP was found to be similar to the binding K_i at α₇-nAChRs.

In this report we have studied the effects of an α₇-nAChR agonist on the basic hippocampal neuronal physiology and neuroplasticity to elucidate a possible mechanism of action underlying the procognitive effects of α₇-nAChR agonists. Using concentrations of an α₇-nAChR agonist (FRM-17848) (FRM-2 in Tang et al. 2014), well below the agonist EC₅₀ (455 nM) and approximately a half-log below the binding K_i (11 nM), but consistent with ACh priming concentration (3.16 nM), we now show enhanced LTP in rat brain slices in a narrow concentration range. The percent increase in LTP after FRM-17848 treatment was comparable to that for donepezil tested at concentrations that are likely to exceed the unbound concentration achievable in patients (5–10 nM) (Gomolin et al. 2011; Seltzer 2005; Yang 2013a). In addition to being an agonist at the α₇-nAChR, FRM-17848 is also a potent 5-HT₃ receptor antagonist (binding K_i = 35 nM and cellular IC₅₀ in oocytes = 9 nM). Since the enhancement of LTP with FRM-17848 was completely blocked by MLA, a highly specific α₇-nAChR antagonist, it is reasonable to postulate that activation of α₇-nAChRs is mediating the observed effects (Palma et al. 1996). Although other classes of 5-HT₃ antagonists can enhance hippocampal LTP, they also reduce GABAergic activity in hippocampal interneurons (Dale et al. 2014; Reznic and Staubli 1997). On the contrary, 3.16 nM FRM-17848 increased GABAergic activity, suggesting that at this concentration FRM-17848 activity at α₇-nAChRs predominated over activity at 5-HT₃ receptors. In addition the enhancement of LTP was lost when the concentration of FRM-17848 was increased from 3.16 nM to 5.6 nM, which is inconsistent with a 5-HT₃ antagonism hypothesis.

The observation that FRM-17848 causes a hyperpolarization, rather than depolarization, of CA1 pyramidal neurons led us to investigate the effects of FRM-17848 on spontaneous IPSCs. IPSC frequency increased in response to FRM-17848, as might be expected based on the predominant expression of α₇-nAChRs on GABAergic interneurons (Fabian-Fine et al. 2001; Frazier et al. 1998; Freedman et al. 1993; Jones 1997; Kawai et al. 2002). These results are consistent with previous findings that showed that α₇-nAChR agonists increase the synaptic transmission of GABAergic interneurons at or near the binding K_i (PNU-282987 K_i = 26 nM and FRM-17874 K_i = 4.6 nM for rat α₇-nAChRs) (Hajos et al. 2005; Stoiljkovic et al. 2015). Similarly, an α₇-nAChR PAM, PNU-120596, increased IPSC frequency near the EC₅₀ for potentiating ACh-evoked responses (Hurst et al. 2005).

One potential explanation for the finding that increases in both LTP and IPSC frequency occurred at concentrations near the binding K_i values of α₇-nAChR agonists is the phenomenon described as priming (Papke et al. 2011; Quik et al. 1997), which proposes that binding of a single molecule of an agonist such as FRM-17848 at an orthosteric binding site primes the receptor to open in response to a second binding event of one molecule of an endogenous ligand such as ACh. The correlation among priming concentrations, enhanced LTP, and increased IPSC frequency is significant because they align with the concentrations for improving performance in cognition and memory-related tasks in animal models (Stoiljkovic et al. 2015).

Our experiments demonstrate that FRM-17848 exhibits a narrow, bell-shaped concentration-response function. While this narrow efficacy concentration range for FRM-17848 may be prohibitive for the development of this compound into a commercial drug, understanding the biological parameters that set the upper and lower boundaries may be useful in developing other molecules that widen the concentration-response function. In the canonical model of nAChR activation, at least two molecules of ACh are required for activation of the receptor (Changeux 1992). Without rapid clearance of the ligand, α₇-nAChRs quickly desensitize (Dani and Bertrand 2007). Similarly, at higher concentrations of FRM-17848, two molecules could bind, activate, and desensitize the α₇-nAChR in the absence of ACh. We have reported that a similar compound, FRM-17874, enhances LTP at 3.16 and 5 nM but not at 10 nM (Stoiljkovic et al. 2015). Thus our model would predict a narrow concentration-response function between priming a sufficient number of receptors to observe efficacy and desensitizing them. Designing α₇-nAChR agonists that show reduced desensitization or exhibit long a half-life (T½) may be desirable to maintain drug levels in an efficacious concentration range.

An alternative hypothesis could explain the bell-shaped concentration-response function using a GABAergic circuit mechanism rather than a receptor-level mechanism. By this reasoning, low concentrations of FRM-17848 would induce a modest increase in GABAergic activity. By moderately hyperpolarizing pyramidal neurons, spontaneous background activity in the hippocampus would be dampened, thereby enhancing the contrast of burst signal-to-noise. However, at higher FRM-17848 concentrations, a further increase in GABAergic activity would hyperpolarize pyramidal neurons to the point of disrupting neurotransmission. A similar phenomenon has been shown in the frontal cortex, where α₇-nAChRs are localized on glutamatergic neurons, escalating concentrations of an α₇-nAChR agonist (PHA543613) continue to increase glutamate release, yet the procognitive effect diminished (Yang 2013b), suggesting that higher concentrations of α₇-nAChR agonists disrupt neuronal network functioning before α₇-nAChRs become desensitized.

However, our data do not support the latter hypothesis for this hippocampal circuit. First, we show that 500 nM donepezil exhibited a comparable effect on LTP as FRM-17848, yet 500 nM donepezil doubled the IPSC frequency (spontaneous IPSC frequency with 500 nM donepezil = 204% of baseline) in contrast to FRM-17848's more modest increase (spontaneous IPSC frequency with 3.16 nM FRM-17848 = 146% of baseline). Therefore, further increasing the IPSC frequency beyond that achieved with 3.16 nM FRM-17848 is not consistent with suppression of plasticity in this circuit. Second, we showed that 5.6 nM FRM-17848 was ineffective at enhancing LTP and that IPSC frequency was lower (not higher) than with 3.16 nM FRM-17848. The receptor-level model can account for this if 3.16 nM FRM-17848 is optimal for priming the α₇-nAChR in the presence of ACh, while 5.6 nM begins to desensitize it. These data are consistent with previous studies using other α₇-nAChR agonist/PAM compounds, which also showed narrow active concentration ranges (Hajos et al. 2005; Hurst et al. 2005; Kroker et al. 2011; Lagostena et al. 2008). Together, these studies suggest that receptor desensitization rather than overactivation of the GABAergic circuit may set the upper boundary of the concentration response curve in the hippocampus.

Our data indicate that the modulation of GABA is directly linked to the enhanced LTP, because the GABA_A α₅R inhibitor blocked the LTP-enhancing effects of an α₇-nAChR agonist without affecting baseline LTP. These results provide direct pharmacological evidence that activation of α₇-nAChRs enhances the activity of a specific subtype of GABAergic synapse, which in turn modulates synaptic plasticity in the hippocampus. Further confirmation could be established by testing α₇-nAChR agonists in brain slices derived from GABA_A α₅R knoc kout mice. There remains some controversy regarding the synaptic vs. extrasynaptic localization of GABA_A α₅Rs in the hippocampus. Our results with FRM-35440 in Fig. 7B are consistent with the existence of a synaptic pool GABA_A α₅Rs as has been shown directly by electron microscopy (Serwanski et al. 2006). The data reported here provide a testable framework for a “procognitive circuit” that may contribute to the understanding of cognitive enhancement throughout the brain.

While this report has focused on the effects of FRM-17848 on the GABAergic circuit and its impact on synaptic plasticity, there may be effects on glutamatergic currents as well (Gu et al. 2012). In Fig. 2 we show that FRM-17848 and donepezil have no effect on baseline fEPSPs, which is consistent with a previous report (Lagostena et al. 2008). In a separate report (manuscript in preparation) we looked at the effects of FRM-17848 on evoked glutamatergic EPSCs and found no enhancement. Our data therefore support the conclusion that the primary effect of FRM-17848 is on GABAergic rather than glutamatergic neurotransmission in this hippocampal circuit.

At first glance, it is unexpected that an increase in GABAergic activity would be associated with enhanced LTP. Nonselective GABA_A receptor PAMs such as benzodiazepines are known to suppress LTP (del Cerro et al. 1992) and are associated with amnesia in humans (Brown and Dundee 1968; Clarke et al. 1970). Moreover, the increase in IPSC activity with 3.16 nM FRM-17848 clearly coincided with a modest hyperpolarization in pyramidal neurons. We observed that the addition of bicuculline caused a depolarization in pyramidal neurons, indicating that GABA contributes significantly to the hyperpolarized resting membrane potential (data not shown). We have demonstrated that the hyperpolarization of pyramidal neurons caused by the increase in GABAergic activity with FRM-17848 leads to reduced firing of spontaneous action potentials in response to a depolarizing current step. Mechanistically, there are many potential circuit-based explanations such as feed forward inhibition (Larson and Lynch 1986) or rebound after-hyperpolarization (Aizenman et al. 1998; Sah and Bekkers 1996) that may account for this apparent paradox. In a recent set of experiments it was shown in cortical circuits that while a nonselective GABA_A receptor PAM such as diazepam decreased the discharge rate of neocortical neurons during up states (active firing), a GABA_A α₅R PAM such as SH-053-2'F-R-CH3 had the opposite effect by increasing bursting activity (Drexler et al. 2013). Some additional recent work by Koh et al. (2013) has demonstrated that GABA_A α₅R PAMs are also procognitive in rats. We favor the possibility that α₇-nAChR agonists may mimic this type of activation of a subset of GABAergic synapses in the circuit, thereby enhancing bursting activity of glutamatergic neurons. This would have the anticipated effect of improving signal-to-noise in the hippocampal circuit, which may be relevant to cognition (Lisman 1997). While additional studies are needed to test this hypothesis, the data reported here support a convergence of the cholinergic (both α₇-nAChR activation and AChE inhibition) and GABAergic systems on regulating neuronal plasticity in the hippocampus.

In conclusion, we report that an α₇-nAChR agonist enhanced LTP in the hippocampus of rat brain slices in a narrow concentration range consistent with a concentration that primes the receptor. The addition of FRM-17848 stimulated an increase in GABAergic activity among a subset of GABAergic synapses. The coapplication of a GABA_A α₅R inhibitor blocked the effects of FRM-17848 on LTP and IPSC frequency. We conclude that the α₇-nAChR agonists promote the activity of a subset of GABAergic synapses as part of a cognition circuit in the hippocampus. These studies help to elucidate the mechanism of action by which α₇-nAChR agonists produce their procognitive activity and demonstrate their potential utility in treating cognitive impairments in schizophrenia and AD.

GRANTS

This work was funded by FORUM Pharmaceuticals.

DISCLOSURES

M. Townsend, D. G. Flood, L. Leventhal, H. Patzke, and G. Koenig were employees of FORUM Pharmaceuticals, a privately owned corporation. A. Whyment, J.-S. Walczak, R. Jeggo, and M. van den Top are employees of Cerebrasol, which was contracted to perform electrophysiology experiments.

AUTHOR CONTRIBUTIONS

M.T., L.L., and G.K. conception and design of research; M.T., A.W., and R.J. analyzed data; M.T., D.G.F., H.P., and G.K. interpreted results of experiments; M.T. prepared figures; M.T. and D.G.F. drafted manuscript; M.T., D.G.F., H.P., and G.K. edited and revised manuscript; M.T., A.W., J.-S.W., R.J., M.v.d.T., D.G.F., L.L., H.P., and G.K. approved final version of manuscript; A.W., J.-S.W., and M.v.d.T. performed experiments.