Type I and II positive allosteric modulators differentially modulate agonist-induced up-regulation of α7 nicotinic acetylcholine receptors

Authors


Address correspondence and reprint requests to Morten Skøtt Thomsen, Ph.D., Neurobiology Research Unit, University Hospital Copenhagen, Juliane Maries Vej 24, DK-2100 Copenhagen, Denmark. E-mail: morten.s.thomsen@nru.dk

Abstract

Long-term treatment with nicotine or selective α7 nicotinic acetylcholine receptor (nAChR) agonists increases the number of α7 nAChRs and this up-regulation may be involved in the mechanism underlying the sustained procognitive effect of these compounds. Here, we investigate the influence of type I and II α7 nAChR positive allosteric modulators (PAMs) on agonist-induced α7 nAChR up-regulation. We show that the type II PAMs, PNU-120596 (10 μM) or TQS (1 and 10 μM), inhibit up-regulation, as measured by protein levels, induced by the α7 nAChR agonist A-582941 (10 nM or 10 μM), in SH-EP1 cells stably expressing human α7 nAChR, whereas the type I PAMs AVL-3288 or NS1738 do not. Contrarily, neither type I nor II PAMs affect 10 μM nicotine-induced receptor up-regulation, suggesting that nicotine and A-582941 induce up-regulation through different mechanisms. We further show in vivo that 3 mg/kg PNU-120596 inhibits up-regulation of the α7 nAChR induced by 10 mg/kg A-582941, as measured by [125I]-bungarotoxin autoradiography, whereas 1 mg/kg AVL-3288 does not. Given that type II PAMs decrease desensitization of the receptor, whereas type I PAMs do not, these results suggest that receptor desensitization is involved in A-582941-induced up-regulation. Our results are the first to show an in vivo difference between type I and II α7 nAChR PAMs, and demonstrate an agonist-dependent effect of type II PAMs occurring on a much longer time scale than previously appreciated. Furthermore, our data suggest that nicotine and A-582941 induce up-regulation through different mechanisms, and that this confers differential sensitivity to the effects of α7 nAChR PAMs. These results may have implications for the clinical development of α7 nAChR PAMs.

Abbreviations used
BTX

α-bungarotoxin

mPFC

medial prefrontal cortex

nAChR

nicotinic acetylcholine receptor

PAM

positive allosteric modulator

VLO

ventrolateral orbitofrontal cortex

The α7 nicotinic acetylcholine receptor (nAChR) is a pentameric cation-selective ligand-gated ion channel that is widely expressed in the brain (Dani and Bertrand 2007; Albuquerque et al. 2009). Numerous studies have demonstrated that α7 nAChR agonists improve a wide range of cognitive parameters in animal models (Hajos 2009; Thomsen et al. 2010). Clinical studies have also demonstrated cognitive improvement in healthy volunteers and patients with schizophrenia, although somewhat restricted to effects on attention and working memory domains (Kitagawa et al. 2003; Olincy et al. 2006; Freedman et al. 2008). Consequently, α7 nAChR ligands are currently being developed for the treatment of cognitive deficits in diseases such as schizophrenia and Alzheimer’s disease (Wallace and Porter 2011).

Positive allosteric modulators (PAMs) of the α7 nAChR have no intrinsic effect on channel activation, but increase the effectiveness of an agonist. They are normally divided into two types depending on whether they decrease receptor desensitization (type II) or not (type I) (Grønlien et al. 2007). It has been argued that allosteric modulation may be preferable over direct agonism for therapeutic purposes because this merely modulates the effects of endogenous transmitters, rather than activating the receptor per se, thus restricting the drug effects to areas where acetylcholine is being released (Faghih et al. 2008), but the lack of activation by the PAMs alone may diminish their effect in patients with decreased levels of endogenous activation. As there are no published clinical results with α7 nAChR PAMs, it is not clear at the moment, whether α7 nAChR agonists or PAM are preferable for clinical use. But from the limited data on the behavioral effects of α7 nAChR PAMs in animals, they seem to produce similar acute effects as α7 nAChR agonists, including improvements of pre-pulse inhibition and auditory gating as well as short- and long-term memory (Hurst et al., 2005; Ng et al. 2007; Timmermann et al. 2007; Dunlop et al. 2009; Dinklo et al. 2011). There are currently no behavioral data demonstrating in vivo differences between the effects of type I and II α7 nAChR PAMs.

It is well known that nicotine and selective α7 nAChR agonists increase binding of the α7 nAChR antagonist [125I]-bungarotoxin (BTX) in the rodent brain after systemic administration (Marks et al. 1985, 1986a; Rasmussen and Perry 2006; Christensen et al. 2010). Several mechanisms may be involved in agonist-induced up-regulation of nAChRs, such as decreased cell surface turnover, increased receptor trafficking to the surface, increased subunit maturation and assembly, decreased subunit degradation, and conformational changes at the membrane leading to increased ligand affinity (reviewed in (Lester et al. 2009; Govind et al. 2009)). However, the mechanisms underlying agonist-induced up-regulation of the α7 nAChR are not well understood.

Underscoring the potential behavioral effects of nAChR up-regulation, it has been shown that increased [125I]-BTX binding and enhanced novel object recognition in mice occur in parallel 4–48 h after administration of the specific α7 nAChR agonist AZD0328 (Werkheiser et al. 2011), and that up-regulation of [3H]-nicotine binding correlates directly with performance in the Morris water maze task several days after nicotine administration (Abdulla et al. 1996). These data suggest that increased nAChR levels may partly underlie the cognitive improvements seen with nAChR agonists. We have recently demonstrated that α7 nAChR up-regulation does not occur with administration of α7 nAChR PAMs in the rat brain, demonstrating a fundamental in vivo difference between agonists and PAMs (Christensen et al. 2010).

Here, we use an SH-EP1 cell line transfected with the α7 nAChR coupled to yellow fluorescent protein and ex vivo [125I]-BTX binding to show that type II α7 nAChR PAMs inhibit up-regulation of the α7 nAChR by the α7 nAChR agonist A-582941, whereas type I α7 nAChR PAMs do not affect A-582941-induced up-regulation. Neither the type I or II PAMs affect α7 nAChR up-regulation by nicotine. These results demonstrate a fundamental difference between type I and II PAMs that may affect the long-term behavioral effects of these compounds. Furthermore, the results suggest that different agonists may mediate up-regulation of the α7 nAChR through different mechanisms.

Methods

Compounds

2-methyl-5-(6-phenyl-pyridazin-3-yl)-octahydro-pyrrolo[3,4-c]pyrrole (A-582941) (Bitner et al. 2007), 1-(5-chloro-2,4-dimethoxy-phenyl)-3-(5-methyl-isoxazol-3-yl)-urea (PNU-120596) (Hurst et al., 2005), and 1-(5-chloro-2-hydroxy-phenyl)-3-(2-chloro-5-trifluoromethyl-phenyl)-urea (NS1738) (Timmermann et al. 2007) were synthesized at the Department of Medicinal Chemistry at NeuroSearch A/S (Ballerup, Denmark). N-(4-chlorophenyl)-α-[[(4-chloro-phenyl)amino]methylene]-3-methyl-5-isoxazoleacet-amide (AVL-3288, also known as XY4083 and CCMI) (Ng et al. 2007) was a kind gift from Kelvin W. Gee, University of California, Irvine. 4-naphthalen-2-yl)-3a,4,5,9b-tetrahydro-3H-cyclopenta[c]quinoline-8-sulfonamide (TQS) (Grønlien et al. 2007) was synthesized by the Chembridge Corporation (San Diego, CA, USA). Methyllycaconitine citrate (MLA) was purchased from Ascent Scientific (Bristol, UK). (−)-Nicotine hydrogen tartrate salt was purchased from Sigma Aldrich (St. Louis, MO, USA).

Cell Culture

SH-EP1 cells, wild-type or stably and heterologously expressing enhanced yellow-fluorescent protein inserted into the C2 loop of the mouse α7 subunit (Murray et al. 2009) – a kind gift from Dr. Paul Whiteaker at the Barrow Neurological Institute – were maintained at 37°C in a humidified atmosphere with 5% CO2 in tissue culture-treated polystyrene flasks coated with 5 μg/mL poly-l-lysine. The culture medium was high-glucose Dulbecco’s modified Eagle’s medium, supplemented with 10% heat-inactivated horse serum, 5% fetal bovine serum, 8 mM l-glutamine, 1 mM sodium pyruvate, 100 Units/mL penicillin, and 100 μg/mL streptomycin. Hygromycin B was added at a final concentration of 130 μg/ml to the culture medium used for the transfected cell line to prevent proliferation of non-transfected cells. For analysis of fluorescence, confluent cells were trypsinized for ∼3 min at 37°C, and replated at a 1 : 4 ratio in dark-walled cell-culture tissue culture-treated 96-well plates coated with 5 μg/mL poly-l-lysine.

Stimulation of the cells was performed 24 h after plating. The stimulation was performed by addition of compounds to the culture medium followed by gentle mixing. A-582941, MLA, and nicotine were dissolved in culture media. PNU-120596, NS1738, AVL-3288, and TQS were dissolved in dimethyl sulfoxide (DMSO), and then diluted in culture media to a maximal final concentration of 2.5% DMSO.

When coincubating with two compounds, A-582941 or nicotine was added 15 min after the other compound. After another 24 h, the cells were washed three times in pre-warmed Dulbecco’s phosphate-buffered saline (PBS). Micrographs were taken in the middle of each well using a 10 × objective with a FITC filter on an Olympus IX51 (Olympus, Ballerup, Denmark) inverted microscope using Visiopharm Integrator System software version 3.4.2.0 after adjusting black levels in a well only containing Dulbecco’s PBS. Background subtraction using the rolling-ball method with a sliding paraboloid and a radius of 50 pixels, and subsequent mean OD measurements of the micrographs were performed using ImageJ version 1.43μ (http://rsbweb.nih.gov/ij/). The average mean OD for wells containing wild-type cells was subtracted from all values, and data were normalized to that of the vehicle-treated cells in each experiment.

Cell density was measured in separate experiments using Hoechst 33342 (Molecular Probes®, Nærum, Denmark) to stain nuclei. Briefly, cells were incubated for 5 min in 1 μg/mL Hoechst 33342 in culture media, followed by two washes in pre-warmed Dulbecco’s PBS. Micrographs were taken as stated above, but using a DAPI filter (Olympus). The number of cells in the micrographs were counted using the Analyze Particles function in ImageJ version 1.43μ.

Animals and drug treatment

Seventy-five juvenile (34 days old at the time of the experiment) male Wistar rats were purchased from Taconic Europe (Ll. Skensved, Denmark). The animals were acclimatized under standardized conditions with free access to food and water for 7 days after arrival. All experiments were conducted in accordance with the Declaration of Helsinki, the Danish National Guide for Care and Use of Laboratory animals, and the European Communities Council Directive of 24 November 1986 (86/609/EEC). All compounds were injected s.c. at 2 mL/kg.

For the first in vivo experiment, rats were injected with 3 mg/kg PNU-120596 or vehicle (1% DMSO in 0.9% saline). Fifteen minutes later, they were injected with 10 mg/kg A-582941 or vehicle (0.9% saline).

For the second in vivo experiment, rats were injected with 1 mg/kg AVL-3288 or vehicle (1% DMSO in 0.9% saline). Fifteen minutes later, they were injected with 10 mg/kg A-582941 or vehicle (0.9% saline).

For the third in vivo experiment, rats were injected with 3 mg/kg PNU-120596, 1 mg/kg AVL-3288, or vehicle (5% DMSO and 8% Solutol in 0.9% saline). Fifteen minutes later, they were injected with 1 mg/kg nicotine (free base weight) or vehicle (0.9% saline).

For all three experiments, the rats were killed by decapitation 4 h after the last administration, and their brains dissected and divided into two hemispheres, one of which was frozen directly in powdered dry ice and used for autoradiography. The other hemisphere was used outside of the context of this article.

Autoradiography

Brain hemispheres were cut in 12 μm serial coronal sections on a cryostat and directly thaw-mounted onto super frost glass slides. Sections were collected in parallel series with four to six sections per glass slide throughout the prefrontal cortex (six series, 2.8–3.2 mm anterior to Bregma) and the dorsal hippocampal region (six series, 3.6–4.0 mm posterior to Bregma) (Paxinos and Watson 1986). Two adjacent slides from each animal were thawed at 22°C for 30 min, followed by 30-min hydration in 50 mM Tris buffer, pH 7.3 (binding buffer). One slide was then incubated for 2 h in binding buffer containing 0.5 nM [125I]Tyr-54-mono-iodo-α-bungarotoxin (2,200Ci/mmol, Perkin Elmer, Skovlunde, Denmark) to assess total binding. For analysis of non-specific binding, 1 mM (−)-nicotine (Sigma-Aldrich, Brøndby, Denmark) was included in the incubation for the adjacent slide. Slides were then briefly washed in binding buffer, followed by 2 × 30-min washes in ice-cold binding buffer and rinsed briefly in ice-cold distilled water. Finally, the slides were dried for 2–3 h, and exposed to Kodak Biomax MS film (GE Healthcare, Little Chalfont, UK) for ∼18 h. Mean optical densities from autoradiographies were quantified in the regions of interest using a computer image analysis system (Quantity One®, Bio-Rad, CA, USA) by an observer blinded to the treatment of the animals. The individual value for each region was calculated as the average measurement from three individual sections. The average value from a corresponding slide with non-specific binding was subtracted to yield specific binding.

Cytotoxicity assay

The release of cytoplasmic lactate dehydrogenase (LDH) into the culture medium from transfected SH-EP1 cells was analyzed using the CytoTox 96® Non-Radioactive Cytotoxicity Assay (Promega, San Luis Obispo, CA, USA), according to the manufacturer’s instructions. Briefly, the culture medium was removed and spun to remove cellular residue. The remaining cells were washed in PBS and resuspended in fresh culture medium, lysed by incubating for 30 min at −80°C followed by 15 min at 37°C, and spun to remove cellular residue. A volume of 50 μl of each sample was incubated with 50 μl of a Substrate Mix for 30 min in the dark, after which the reaction was stopped by addition of 50 μl 1 M acetic acid. Finally, the color intensity at the 490 nm wavelength was measured using a plate reader. The results are presented as the percentage of LDH released, i.e. LDH from the medium divided by the total amount of LDH from the medium and the cells.

Quantification and statistical analysis

Data were analyzed using one-way anova followed by Dunnett’s or Tukey’s Multiple Comparison Tests. Unpaired t-tests were used to confirm the effect of agonists when co-applied with other compounds in cell cultures. The statistical calculations were performed using GraphPad Prism version 5.03 for Windows (GraphPad Software, San Diego, USA). All data are presented as mean ± standard error of the mean, and a p-value of less than 0.05 was considered statistically significant.

Results

Agonists, but not PAMs or antagonists, increase α7 nAChR protein levels in vitro

The fluorescent signal from enhanced yellow fluorescent protein was used to measure the level of α7 subunit protein expression in transfected SH-EP1 cell cultures (Fig. 1a).

Figure 1.

 α7 nicotinic acetylcholine receptor agonists increase α7 subunit expression in transfected SH-EP1 cells. (a) Representative micrograph taken at 10 × (the inset is a 40 × section) magnification with a FITC filter of an SH-EP1 cell culture expressing enhanced yellow-fluorescent protein coupled to the mouse α7 subunit. (b) Incubation with A-582941 or nicotine for 24 h dose-dependently increased α7 protein expression in SH-EP1 cell cultures. (c) Co-incubation with methyllycaconitine (MLA) dose-dependently inhibited the increase in α7 protein expression induced by A-582941 or nicotine. (d) Incubation with MLA, AVL-3288, NS1738, PNU-120596, or TQS for 24 h did not significantly change α7 protein expression in SH-EP1 cell cultures. n > 8. *< 0.05, **< 0.01, and ***< 0.001 indicate significant difference from vehicle controls (a) or A-582941- or nicotine-treated groups (d) in a one-way anova with Dunnett’s Multiple Comparisons Test. ###< 0.001 indicates significant difference from the vehicle-treated group in an unpaired t-test. Error bars indicate the standard error of the mean.

Incubation for 24 h with the non-selective nAChR agonist nicotine or the selective α7 nAChR agonist A-582941 significantly increased α7 subunit protein expression to a maximum of ∼128% and ∼137%, respectively (Fig. 1b). A-582941 produced a maximal increase at 1 nM, and was thus approximately 1000-fold more potent compared with nicotine, which produced a maximal increase at 1 μM.

To verify the involvement of the α7 nAChR in the effect of nicotine and A-582941, they were co-incubated with the α7 nAChR antagonist MLA. MLA at 0.1 and 1 μM almost completely inhibited the effect of 10 μM A-582941 or nicotine on α7 protein expression (Fig. 1c).

Incubation for 24 h with the α7 nAChR antagonist MLA (0.01–10 μM) or the α7 PAMs AVL-3288 (0.01–100 μM), NS1738 (0.01–100 μM), PNU-120596 (0.01–10 μM), or TQS (0.01–10 μM) did not affect α7 protein expression (Fig. 1d).

In a separate experiment, nuclear staining using Hoechst 33342 showed that neither of the above-described treatments significantly affected the number of cells in the wells (data not shown).

Type I and II α7 nAChR positive allosteric modulators differentially regulate α7 nAChR up-regulation by A-582941 and nicotine in vitro

Co-incubation of 10 nM or 10 μM A-582941 with the type II PAMs PNU-120596 or TQS dose-dependently inhibited the effect of A-582941 on α7 protein expression, with complete inhibition at 10 μM of PAM in both cases (Fig. 2a and b). Contrarily, 0.1–10 μM concentrations of the type I PAMs AVL-3288 or NS1738 did not affect the A-582941-induced increase in α7 protein expression.

Figure 2.

 α7 nicotinic acetylcholine receptor up-regulation by A-582941 and nicotine in transfected SH-EP1 cells is differentially regulated by α7 nicotinic acetylcholine receptor positive allosteric modulators. (a, b) Co-incubation of A-582941 with 10 μM PNU-120596 or 1 or 10 μM TQS for 24 h significantly inhibited the increase in α7 protein expression induced by 10 nM or 10 μM A-582941 in SH-EP1 cell cultures. Co-incubation of A-582941 with AVL-3288 or NS1738 did not affect A-582941-induced up-regulation of α7 protein expression. (c) Co-incubation of 10 μM nicotine with PNU-120596, TQS, AVL-3288, or NS1738 did not affect nicotine-induced up-regulation of α7 protein expression. n > 8. *< 0.05, **< 0.01, and ***< 0.001 indicate significant difference from A-582941- or nicotine-treated groups in a one-way anova with Dunnett’s Multiple Comparisons Test. #< 0.05, ##< 0.01, and ###< 0.001 indicate significant difference from respective vehicle group in an unpaired t-test. Error bars indicate the standard error of the mean.

Co-incubation of 10 μM nicotine with 0.1–10 μM concentrations of PNU-120596, TQS, AVL-3288, or NS1738 did not affect the nicotine-induced increase in α7 protein expression (Fig. 2c).

In a separate experiment, nuclear staining using Hoechst 33342 showed that neither of the above described treatments significantly affected the number of cells in the wells (data not shown).

A type II, but not type I, PAM inhibits A-582941-induced increase in [125I]-BTX binding in the rat frontal cortex and hippocampus

[125I]-BTX autoradiography was performed on brain sections from juvenile rats that had received an acute injection of 3 mg/kg PNU-120596 or its vehicle followed by an injection with 10 mg/kg A-582941 or its vehicle (Fig. 3a). A-582941 increased [125I]-BTX binding in all regions measured, although only significantly so in layer V–VI of the medial prefrontal cortex (mPFC) as well as the CA1, CA2/3, and dentate gyrus regions of the hippocampus, which corresponds with our previously published data (Christensen et al. 2010). Administration of PNU-120596 alone did not significantly alter [125I]-BTX binding, although a trend towards decreased binding was evident in the mPFC and ventrolateral orbitofrontal cortex (VLO), similar to what has previously been shown after repeated administration of PNU-120596 (Christensen et al. 2010). Notably, pre-administration with PNU-120596 lead to an almost complete inhibition of the A-582941-induced increase in [125I]-BTX binding in all regions measured, except the CA2/3 region.

Figure 3.

 PNU-120596, but not AVL-3288, inhibits A-582941-induced α-bungarotoxin (BTX) binding in the frontal cortex and hippocampus. Juvenile rats were administered vehicle or A-582941 (A-58, 10 mg/kg) preceded by an injection with PNU-120596 (PNU, 3 mg/kg) or AVL-3288 (AVL, 1 mg/kg) or vehicle and killed by decapitation. [125I]-BTX binding was analyzed using optical density (OD) in the outer (I–IV) and inner (V–VI) layers of the medial prefrontal cortex (mPFC), ventrolateral orbitofrontal cortex (VLO), as well as the CA1, CA2/3, and dentage gyrus (DG) subregions of the hippocampus. In both experiments, A-582941 significantly increased [125I]-BTX binding in most regions analyzed. (a) In all regions except the CA2/3, PNU-120596 significantly reduced A-582941-induced [125I]-BTX binding. (b) Contrarily, AVL-3288 did not affect A-582941-induced [125I]-BTX binding. In the VLO, however, AVL-3288 significantly increased [125I]-BTX binding. n = 5–8. *< 0.05, **< 0.01, and ***< 0.001 indicate significant difference from vehicle controls, #< 0.05, ##< 0.01, and ###< 0.001 indicate significant difference from the vehicle+A-582941-treated group, and ¤< 0.05, ¤¤< 0.01, and ¤¤¤< 0.001 indicate significant difference from the AVL-3288 + vehicle-treated group, respectively, in a one-way anova with Tukey’s Multiple Comparisons Test performed separately for each region. Error bars indicate the standard error of the mean. Data were normalized to the BTX binding in mPFC layer I–IV in the vehicle group.

In a parallel experiment, juvenile rats received an acute injection of 1 mg/kg AVL-3288 or its vehicle followed by an injection with 10 mg/kg A-582941 or its vehicle (Fig. 3b). Here, A-582941 significantly increased [125I]-BTX binding in the mPFC, VLO, and CA2/3. AVL-3288 only affected [125I]-BTX binding in the VLO, where it produced a small, but significant increase. Pre-administration with AVL-3288 did not significantly affect the A-582941-induced increase in [125I]-BTX binding in any region measured.

Acute nicotine does not increase BTX binding in the frontal cortex or hippocampus in the presence or absence of PNU-120596 or AVL-3288

[125I]-BTX autoradiography was performed on brain sections from juvenile rats that had received an acute injection of 3 mg/kg PNU-120596, 1 mg/kg AVL-3288, or vehicle followed by an injection with 1 mg/kg (free base weight) nicotine or its vehicle (Fig. 4). Nicotine did not significantly affect [125I]-BTX in layer I–IV or V–VI of the mPFC or in the VLO. Administration of PNU-120596 or AVL-3288 alone or in combination with nicotine did not affect [125I]-BTX binding in any of the regions measured.

Figure 4.

 Acute nicotine does not increase α-bungarotoxin (BTX) binding in the frontal cortex or hippocampus in the presence or absence of PNU-120596 or AVL-3288. Juvenile rats were administered vehicle or nicotine (Nico, 1 mg/kg free base) preceded by an injection with PNU-120596 (PNU, 3 mg/kg) or AVL-3288 (AVL, 1 mg/kg) or vehicle and killed by decapitation 4 hours later. [125I]-BTX binding was analyzed using optical density (OD) in the outer (I–IV) and inner (V–VI) layers of the medial prefrontal cortex (mPFC) and ventrolateral orbitofrontal cortex (VLO). Nicotine did not significantly increase [125I]-BTX binding in any region measured. Administration of PNU or AVL alone or in combination with nicotine did not affect [125I]-BTX binding. n = 6. Data were analyzed using a one-way anova with Tukey’s Multiple Comparisons Test performed separately for each region. Error bars indicate the standard error of the mean. Data were normalized to the BTX binding in mPFC layer I–IV in the vehicle group.

Ligands of the α7 nAChR do not induce cytotoxicity

The release of LDH, as a measure of cytotoxicity, was measured from transfected SH-EP1 cells incubated with α7 nAChR ligands. The assay was performed on cells exposed to the highest concentrations of ligands used in the present experiments. None of the α7 nAChR ligands or combinations of ligands had any significant effect on LDH release (Fig. 5). Five percent DMSO was used as a positive control in the assay and yielded a highly significant ∼four-fold increase in released LDH compared with untreated cells.

Figure 5.

 Ligands of the α7 nicotinic acetylcholine receptor do not induce cytotoxicity in SH-EP1 cells. Incubation with A-582941, nicotine, methyllycaconitine (MLA), PNU-120596, TQS, AVL-3288 or NS1738 for 24 h in SH-EP1 cell cultures did not affect the release of cytoplasmic lactate dehydrogenase (LDH), as a measure of cytotoxicity. Neither did co-incubation of A-582941 or nicotine with either of the other compounds. 5% dimethyl sulfoxide (DMSO) was used as a positive control, and markedly increased LDH release. n = 8 (24 for control group). ***< 0.001 indicates significant difference from untreated controls in a one-way anova with Dunnett’s Multiple Comparisons Test. Error bars indicate the standard error of the mean.

Discussion

The main findings in this article are that type II α7 nAChR PAMs inhibit A-582941-induced up-regulation of α7 nAChR levels in vitro and in vivo, whereas type I PAMs do not. Furthermore, nicotine-induced up-regulation of α7 nAChR levels in vitro is not affected by type I or II PAMs, suggesting that A-582941 and nicotine increase α7 nAChR levels through different mechanisms.

Both A-582941 and nicotine produced a dose-dependent increase in α7 subunit protein expression in SH-EP1 cells transfected with the α7 nAChR coupled to enhanced yellow-fluorescent protein. Nicotine-induced up-regulation of BTX binding in cell cultures is a well-known phenomenon (Molinari et al. 1998; Kawai and Berg 2001), and it has previously been shown that nicotine increases BTX binding in SH-EP1 cells transfected with the human α7 subunit by ∼80% (Peng et al. 1999). Our data suggest that at least part of the nicotine-induced increase in BTX binding is because of increased protein expression of the α7 subunit. A-582941 was ∼1000-fold more potent than nicotine and slightly more efficacious. The increased potency of A-582941 over nicotine may be because of a higher affinity and potency of A-582941 for the α7 nAChR. Thus, the Ki for A-582941 binding to rat and human brain tissue is 10.8 and 16.7 nM, respectively (Bitner et al. 2007), compared to a Ki of ∼400–8900 nM for nicotine (Marks et al. 1986b; Gotti et al. 1997; Haustein and Groneberg 2009). And, A-582941 activates rat and human α7 nAChRs expressed in Xenopus oocytes with an EC50 of 2.45 and 4.26 μM, respectively (Bitner et al. 2007), compared with an EC50 of ∼18–91 μM for nicotine (Peng et al. 1994b; Briggs et al. 1995; Haustein and Groneberg 2009).

The α7 nAChR antagonist MLA inhibited the up-regulation of α7 protein levels by A-582941 or nicotine. This confirms that binding of the agonist to the orthosteric site of the α7 nAChR is required for up-regulation induced by either compound. MLA has previously been shown to increase BTX binding in SH-SY5Y and transfected HEK293 cells (Molinari et al. 1998; Ridley et al. 2001), but MLA alone had no effect on α7 subunit protein expression in SH-EP1 cells. Although these effects may vary between cell types, one explanation is that MLA affects nAChR levels, as reflected in BTX binding, downstream of protein synthesis.

Neither the type I PAMs, AVL-3288 and NS1738, nor the type II PAMs, PNU-120596 and TQS, affected α7 subunit protein expression in SH-EP1 cells, at doses well above their EC50 for increasing acetylcholine-evoked currents (0.7, 3.4, 0.21, and 5.5 μM, respectively) (Hurst et al., 2005; Grønlien et al. 2007; Ng et al. 2007; Timmermann et al. 2007). The lack of effect of type I or II PAMs corresponds with the lack of effect on in vivo brain BTX binding, previously reported after repeated administration of PNU-120596 or NS1738 (Christensen et al. 2010). This provides further support for the notion that allosteric modulation of the α7 nAChR does not affect the basal levels of the receptor, contrary to what is seen with α7 nAChR agonists.

Importantly, the type II PAMs PNU-120596 and TQS dose-dependently inhibited the increase in α7 subunit protein expression induced by both 10 nM and 10 μM A-582941 in SH-EP1 cells. The inhibition with PNU-120596 or TQS occurred at concentrations which correspond well with those needed to affect agonist current responses at the α7 nAChR (Grønlien et al. 2007). This provides further support that the effect of A-582941 on α7 protein expression occurs directly at the α7 nAChR. Up-regulation of the α7 subunit may depend on specific conformational states, such as desensitized states induced by the ligand. In line with this, it has been shown that receptor desensitization initiates nicotine-induced up-regulation of α4β2 nAChRs (Fenster et al. 1999). Binding of allosteric modulators affects receptor conformations by favoring certain states, and this may diminish the propensity of receptors to enter desensitized states (Williams et al. 2011). Thus, the type II α7 nAChR PAMs, PNU-120596 and TQS, greatly diminish agonist-induced desensitization of the α7 nAChR, and can even revert desensitized receptors into a conducing state (Hurst et al., 2005; Grønlien et al. 2007). It may therefore be the effect on receptor desensitization that enables PNU120596 and TQS to inhibit A-582941-induced up-regulation of the α7 nAChR, whereas the type I α7 nAChR PAMs, AVL-3288 and NS1738, which do not affect receptor desensitization, did not affect the A-582941-induced increase in α7 protein expression in SH-EP1 cells. However, it is also possible that the different PAM compounds bind to different allosteric sites on the α7 nAChR (Thomsen and Mikkelsen 2012), which may confer different modulation of A-582941-mediated effects. Finally, metabotropic signaling through the α7 nAChR has previously been described (Chernyavsky et al. 2009, 2010), and we cannot exclude the possibility that agonists induce receptor up-regulation by activating intracellular cascades through metabotropic signaling, similar to what is observed with, e.g., the regulation of NMDA receptors through metabotropic glutamate receptors (Benquet et al. 2002).

As an extension of the in vitro data, we demonstrated that 3 mg/kg PNU-120596 inhibited the A-582941-induced increase in BTX binding in the frontal cortex and hippocampus in vivo, whereas 1 mg/kg AVL-3288 did not. These data are the first to demonstrate an in vivo difference between type I and II PAMs, and they suggest that type II PAMs have activity-dependent effects that occur on a much longer time scale (i.e. several hours) than previously reported using electrophysiology. Furthermore, as the endogenous agonist, acetylcholine, might also induce up-regulation of nAChRs, and given the potential involvement of receptor up-regulation in the prolonged effect of nAChR ligands, these data allow us to speculate that repeated treatment with type II versus I PAMs may have very different cognitive effects, particularly in situations with high levels of endogenous or experimental agonists. It should be noted, however, that these data are based on single-point BTX measurements, so although we have previously shown with saturation curves that α7 nAChR agonists increase α7 nAChR numbers without affecting ligand affinity, we cannot exclude that α7 nAChR PAMs reduce BTX binding by reducing ligand affinity. In addition, although we have chosen behaviorally active doses of both PAMs (Hurst et al., 2005; Ng et al. 2007), different in vivo kinetics of the two PAMs might influence the results. Thus, administration of 10 mg/kg PNU-120596 results in an approximate brain level of 15 ng/mL, corresponding to 0.048 μM, which declines with a half life of 7.9 h (McLean et al. 2011). Administration of 0.3 mg/mL AVL-3288 results in brain levels of 1 μM after 10 min, which declines to 0.3 μM after 90 min (Ng et al. 2007). So, although PNU-120596 reaches a lower concentration than AVL-3288, the prolonged presence of the compound in the brain may facilitate its effect on an A-582941-induced increase in BTX binding.

The correspondence between our results from in vitro protein data and in vivo BTX-binding data for A-582941 and PNU-120596 suggests that at least part of the observed increase in BTX binding is because of increased α7 subunit protein expression. However, it has been demonstrated in vitro and in vivo that the majority of α7 nAChRs exist as an intracellular pool (Fabian-Fine et al. 2001; Zhong et al. 2008; Mielke and Mealing 2009; Murray et al. 2009). Therefore, an increase in BTX binding might readily occur without increased protein synthesis, and it is possible that other factors, such as increased transport of α7 nAChRs to the cell surface occur in parallel with increased protein synthesis. Vice versa, increases in α7 protein may not necessarily lead to increased BTX binding. Thus, a single administration of 1 mg/kg nicotine did not affect BTX binding in the frontal cortex in vivo. It is well documented that prolonged exposure to nicotine increases nAChR levels, including the α7 nAChR (Marks et al. 1985, 1986a), whereas to our knowledge, there are no reports on the effects of acute nicotine administration on α7 nAChR levels. This effect of nicotine might therefore require prolonged exposure. This illustrates another difference between selective experimental α7 nAChR agonists and nicotine, regarding the mechanism of α7 nAChR up-regulation. This difference may be related to different affinities for the α7 nAChR, as the experimental agonists that demonstrate α7 nAChR up-regulation after acute administration all have higher affinities for the α7 nAChR than nicotine (Christensen et al. 2010). However, although we have previously shown, A-582941-induced up-regulation of the α7 nicotinic receptor to be evident several days after administration (Christensen et al. 2010), we cannot exclude the possibility that the shorter half-life of nicotine (∼45 min) (Matta et al. 2007) compared to A-582941 (∼2 h) in rats (Tietje et al. 2008) precludes us from observing an effect of nicotine.

PNU-120596 has previously been shown to reduce cell viability in α7 nAChR-expressing SH-SY5Y cells, because of the prolonged opening of the α7 nAChR, characteristic of type II PAMs, leading to Ca2+-mediated toxicity (Ng et al. 2007). Another study, however, found no effect of PNU-120596 on cell viability in PC12 cells or rat primary cortical neurons (Hu et al. 2009). We measured LDH release, as a measure of cytotoxicity, and found no effect of any of the α7 nAChR ligands used in our studies. Therefore, the reduced expression of α7 in SH-EP1 cells co-incubated with A-582941 and PNU-120596 or TQS is not because of cytotoxicity of the compounds.

Contrary to the effects on A-582941-induced up-regulation, neither type I nor II α7 nAChR PAMs affected nicotine-induced α7 up-regulation in vitro. This indicates a mechanistic difference in how A-582941 and nicotine induce α7 up-regulation. It has been suggested that nicotine induces up-regulation of high-affinity nAChRs primarily by acting as a molecular chaperone, promoting assembly and maturation of nAChRs, in the endoplasmic reticulum (Kuryatov et al. 2005; Lester et al. 2009). Because of its unique chemical properties, nicotine readily crosses the cellular membrane and is enriched in intracellular organelles such as the endoplasmic reticulum (Lester et al. 2009). It is therefore possible that nicotine primarily affects intracellular receptors, and that these receptors are not as accessible to A-582941 or the α7 nAChR PAMs used in our studies. However, one study has indicated that the degree of up-regulation of high-affinity nAChRs that occurs with nicotine in transfected HEK cells cannot be accounted for merely by increased maturation of already existing subunits (Vallejo et al. 2005), and recent studies have shown that nicotine-induced increases in α4β2*-binding sites can be explained by increased α4 and β2 protein levels (Moretti et al. 2010; Marks et al. 2011). Conformational changes leading to increased ligand binding, and reduced receptor turnover has also been shown to affect ligand binding after nicotine treatment (Peng et al. 1994a; Kuryatov et al. 2005; Vallejo et al. 2005). Furthermore, it has been shown that repeated nicotine-induced up-regulation of α7 nAChRs, but not high-affinity nAChRs, requires protein synthesis and glycosylation in cortical cultures (Kawai and Berg 2001). Proteasomal activity is also an important regulator of α7 nAChR levels (Christianson and Green 2004), and repeated nicotine-mediated inhibition of proteasomal activity has been shown to mediate α7 nAChR up-regulation (Rezvani et al. 2007). In summary, nicotine may affect α7 nAChR levels through several mechanisms, and it is currently not known whether other nAChR agonists, such as A-582941, exhibit the same properties. However, our data with PAM inhibition suggest that A-582941 is dependent on receptor-specific conformations, such as desensitized states, to increase α7 nAChR levels whereas nicotine may not be.

Our data demonstrate a fundamental difference between the effects of α7 nAChR agonists and type I and II PAMs on receptor up-regulation that is evident both in vitro at the level of protein expression in single cells, and in vivo at the level of assembled α7 nAChRs in several brain regions. As nAChR up-regulation may underlie the ability of nAChR agonists to produce, particularly long-lasting, cognitive effects (Abdulla et al. 1996; Briggs et al. 1997; Buccafusco et al. 2005; Werkheiser et al. 2011), it is pertinent to investigate the potential behavioral consequences of this differential effect on agonist-induced up-regulation of the α7 nAChR. Therefore, further studies are warranted to determine whether type II PAMs may inhibit the long-lasting cognitive effects of α7 nAChR agonists.

It is currently not known to what degree endogenous cholinergic signaling affects nAChR levels or on what time scale. However, if up-regulation does occur in response to increased acetylcholine or choline levels, e.g. during performance, then inhibiting this up-regulation may interfere with long-term cognitive performance. Finally, our results highlight the importance of the arousal state of the animals when performing experiments with PAMs, as they are dependent on endogenous signaling for their effect. Thus, both the acute and long-term effects of PAM administration may be very different in non-performing animals with low levels of endogenous neurotransmitter, and in aroused animals with high levels of endogenous neurotransmitter.

Acknowledgements

MST and JDM conceived of and designed the experiment and wrote the article. MST acquired the data. The authors would like to thank Hans Jørgen Jensen for skilful technical assistance, Dr. Paul Whiteaker at the Barrow Neurological Institute for providing the SH-EP1 cell line, and Dr. Kelvin W. Gee, University of California, Irvine, and Dr. Dan Peters and Dr. Daniel B Timmermann, NeuroSearch A/S, for donating α7 nAChR ligands.

This work was supported by the Danish Medical Research Council, the Lundbeck Foundation, the NOVO Nordisk Foundation, and the Danish Ministry of Science, Technology and Innovation. The authors declare that they have no conflicts of interest.

Ancillary