β-Amyloid Directly Inhibits Human α4β2-Nicotinic Acetylcholine Receptors Heterologously Expressed in Human SH-EP1 Cells*

Amyloid-β (Aβ) accumulation and aggregation are thought to contribute to the pathogenesis of Alzheimer's disease (AD). In AD, there is a selective decrease in the numbers of radioligand binding sites corresponding to the most abundant nicotinic acetylcholine receptor (nAChR) subtype, which contains human α4 and β2 subunits (hα4β2-nAChR). However, the relationships between these phenomena are uncertain, and effects of Aβ on hα4β2-nAChR function have not been investigated in detail. We first confirmed expression of hα4 and hβ2 subunits as messenger RNA in transfected, human SHEP1 cells by reverse transcription-polymerase chain reaction and mRNA fluorescence in situ hybridization analyses. Immunoprecipitation Western analyses confirmed α4 and β2 subunit protein expression and co-assembly. Whole cell current recording demonstrated heterologous expression in SH-EP1-hα4β2 cells of functional hα4β2-nAChRs with characteristic responses to nicotinic agonists or antagonists. Nicotine-induced whole cell currents were suppressed by Aβ1-42 in a dose-dependent manner. Functional inhibition was selective for Aβ1-42 compared with the functionally inactive, control peptide Aβ40-1.Aβ1-42-mediated inhibition of hα4β2-nAChR function was non-competitive, voltage-independent, and use-independent. Pre-loading of cells with guanyl-5′-yl thiophosphate failed to prevent Aβ1-42-induced inhibition, suggesting that down-regulation of hα4β2-nAChR function by Aβ1-42 is not mediated by nAChR internalization. Sensitivity to Aβ1-42 antagonism at 1 nm was evident for hα4β2-nAChRs, but not for heterologously expressed human α7-nAChRs, although both nAChR subtypes were functionally inhibited by 100 nm Aβ1-42, with the magnitude of functional block being higher for 100 nm Aβ1-42 acting on hα7-nAChRs. These findings suggest that hα4β2-nAChRs are sensitive and perhaps pathophysiologically relevant targets for Aβ neurotoxicity in AD.

Alzheimer's disease (AD) 1 is a progressive, neurodegenerative disorder manifest as a severe impairment of learning and memory. Pathophysiological hallmarks of AD include extracellular deposits of ␤-amyloid peptide (A␤) in senile plaques, formation of intraneuronal neurofibrillary tangles, and cholinergic neuron death (1). Although the precise mechanisms of AD pathogenesis are only partially understood, it is now widely accepted that the accumulation and aggregation of A␤  plays a key role in the disease (2). Evidence has indicated an interaction between A␤ and the cholinergic system (3). For example, very low concentrations (pico to nanomolar) of A␤ can directly induce cholinergic hypofunction (4 -6). It has been reported that solubilized A␤ inhibits several steps of acetylcholine synthesis and release (4,7), inhibits cholinergic enzyme activity (6), impairs cholinergic metabolism and neurotransmission (8 -10), and depresses hippocampal synaptic function (11).
Recent evidence suggests possible roles for nicotinic acetylcholine receptors (nAChRs) as central targets for A␤-induced neurotoxicity manifest as cholinergic hypofunction and cognitive impairment. First, in the brain of AD patients, there is a marked decrease in the numbers of nAChR radioligand binding sites (12), particularly those that have the highest affinity for nicotine and appear to be the most abundant nAChR subtype in the brain of vertebrates (13)(14)(15)(16). Immunoassays show that these nAChRs contain ␣4 and ␤2 subunits (17,18). Second, nAChR ␣4 and ␤2 subunit immunoreactivity is localized in AD brain to A␤ plaques and neurofibrillary tangles (19). Third, interactions have been reported of A␤  with radioligand binding sites on another abundant nAChR subtype found in the vertebrate brain containing ␣7 subunits (␣7-nAChR; picomolar affinity) and on presumed ␣4␤2-nAChR subtypes (nanomolar affinity) (20,21). Direct inhibition to some extent of ␣7-nAChR function by A␤ has been demonstrated in whole cell current and excised membrane patch single channel recordings from interneurons in rat hippocampal slices (22), in whole cell studies using rat hippocampal cultured neurons (23), and in whole cell recordings from Xenopus oocytes containing heterologously expressed rat or human ␣7-nAChRs (24,25), although one report describes that picomolar concentrations of A␤ directly activated whole cell current responses of Xenopus oocytes heterologously expressing rat ␣7-nAChRs (26). Fourth, it is well known that nAChRs participate in cognitive functions, including learning and memory (16), which are compromised in animals lacking ␣4 or ␤2 subunits (27,28). Thus, nAChR agonists, antagonists, and modulators have received attention as potentially novel agents for therapeutic use in AD (29 -31).
However, surprisingly little attention has been given to the possible functional effects of A␤ on the function of human (h) ␣4␤2-nAChRs. To fill this gap in knowledge, we employed patch clamp techniques to elucidate the acute effects of A␤ 1-42 , control peptide, or related peptides on the function of h␣4␤2-nAChRs heterologously expressed in the SH-EP1 cell line.

EXPERIMENTAL PROCEDURES
Heterologously Expressed h␣4␤2-nAChRs in SH-EP1 Cells-Human ␣4 and ␤2 subunits (kindly provided by Dr. Ortrud Steinlein, Institute for Human Genetics, Rheinische-Wilhelms-Universitat, Bonn, Germany) were subcloned into pcDNA3.1-zeocin and pcDNA3.1-hygromycin vectors, respectively, and transfected using established techniques (32)(33)(34) into native nAChR-null SH-EP1 cells (35) to create the SH-EP1-h␣4␤2 cell line. For this and all other methods, manipulations were conducted at room temperature (23 Ϯ 1°C) unless otherwise noted. Briefly, 3 million SH-EP1 cells in 0.5 ml of 20 mM HEPES, 87 mM NaCl, 5 mM KCl, 0.7 mM NaHPO 4 , 6 mM dextrose, pH 7.05, in an electroporation cuvette were mixed with 100 g of ␣4 and ␤2 subunit cDNA constructs. Samples were subjected to electroporation (Bio-Rad Gene Pulsar model 1652076) at 960 microfarads and 200 volts. After electroporation, cells were suspended to 5 ml in complete medium (36), and 1-ml aliquots were added to 12-ml aliquots of medium in each of five 100-mm dishes before returning the cells to an incubator at 37°C. 48 h later, positive selection of incubated cells was initiated by supplementing the medium with 0.25 mg/ml zeocin (Invitrogen) and 0.4 mg/ml hygromycin (Calbiochem). Colonies of surviving cells were picked by ring cloning and expanded before being screened for radioligand binding and functional evidence for ␣4␤2-nAChR expression, which led to selection of the clone designated as the SH-EP1-h␣4␤2 cell line. Cells were maintained as low passage in medium with 0.25 mg/ml zeocin and 0.4 mg/ml hygromycin to ensure stable expression of phenotype and passaged once weekly by splitting just-confluent cultures 1/10 to maintain cells in proliferative growth. Studies also were done using another cell line transfected with the human ␣4 subunit cDNA and a chimeric ␤2 subunit cDNA encoding nAChR ␤2 subunit sequences, except for cytoplasmic loop amino acids sequence Lys 336 -Ser 448 , which was replaced by amino acids Arg 347 -Ala 448 from the murine serotonin receptor 5-TH 3A subunit sequence via a process described in detail elsewhere (37).
RNA Preparation and Reverse Transcription (RT)-PCR-Total RNA was isolated from the cells growing at ϳ80% confluency in a 100-mm culture dish using 2 ml of TRIzol TM reagent (Invitrogen). Prior to RT-PCR, RNA preparations were treated with amplification-grade RNase-free DNase (Invitrogen) to remove residual genomic DNA contamination. Typically, 1 g of RNA was incubated with 1 unit of DNase I in a 10 l-reaction at room temperature for 15 min, and the DNase I was subsequently inactivated by addition of 1 l of 25 mM EDTA and incubation at 65°C for 10 min. RT was carried out using 0.8 g of DNA-free total RNA and oligo(dT) [12][13][14][15][16][17][18] primer in a 20-l reaction using the Superscript II TM Preamplification system (Invitrogen). At the end of the RT reaction, reverse transcriptase was deactivated by incubation at 75°C for 10 min, and RNAs were removed by adding 1 unit of RNase H followed by incubation at 37°C for 30 min. A typical PCR was performed using 2 l of cDNA preparation, 1 l of 10 M each of 5Ј and 3Ј gene-specific primers, 1 l of 10 mM dNTP, and 2.5 units of RedTaq TM (Sigma) in a 50-l reaction. The primers used in the amplification stage were designed and synthesized based on published gene sequences (GenBank TM accession numbers: ␣4, L35901; ␤2, X53179; GAPDH, M32599). The primer sequences and their predicted product sizes are: nAChR, ␣4 subunit sense 5Ј-gaatgtcacctccatccgcatc-3Ј, ␣4 antisense 5Ј-ccggca(a/g)ttgtc(c/t)ttgaccac-3Ј (product size 790 bp); nAChR, ␤2 subunit sense 5Ј-cggctcccttccaaacaca-3Ј, ␤2 antisense 5Ј-gcaatgatggcgtggctgctgca-3Ј (product size 754 bp); GAPDH, sense 5Ј-cgtattgggcgcctggtcaccag-3Ј, GAPDH antisense 5Ј-gtccttgcccacagccttggcagc-3Ј (product size 624 bp). Amplification reactions were carried out in a RoboCycler (Stratagene, La Jolla, CA) for 35 amplification cycles at 95°C for 1 min, 55°C for 90 s, and 72°C for 90 s, followed by an additional 4-min extension at 72°C. One-tenth of each RT-PCR product was then resolved on a 1% agarose gel before ethidium staining was used to visualize bands under ultraviolet illumination.
In Situ Hybridization for Cell-specific Localization of ␣4 or ␤2 Subunit Message-Riboprobes corresponding to nucleotide sequences coding for large segments of the ␣4 or ␤2 subunit large cytoplasmic domains were prepared from in vitro transcription and used for in situ hybridization. For riboprobe preparation, 354-(for the ␣4 subunit) or 211-bp (for the ␤2 subunit) partial cDNA fragments were amplified from human nAChR ␣4 and ␤2 subunit cDNA constructs using the following two sets of PCR primers: ␣4 sense 5Ј-gcacccgccctcaccgtccttc-3Ј and ␣4 antisense 5Ј-tcagccgagtgggtgttgcgagag-3Ј; ␤2 sense 5Ј-catcattgcgcccgtcagcg-3Ј and ␤2 antisense 5Ј-gccacagccacacggctccc-3Ј. These ␣4 and ␤2 subunit cytoplasmic domain templates were then gel purified and non-directionally ligated to a T7 phage promoter adapter (Lig'n Scribe, Ambion, Austin, TX). Subsequent PCRs using sense or antisense gene-specific primers in conjunction with Lig'n Scribe adapter primers were then performed to selectively amplify ligation products of in vitro transcription templates in antisense or sense orientation. Sense or antisense partial ␣4/␤2 RNAs were separately transcribed from the PCR-derived template in the presence of biotinylated UTPs using the MAXIscript in vitro transcription system (Ambion), and the quality and quantity of the riboprobes were evaluated using denaturing polyacrylamide gel electrophoresis and spectrophotometric reading, respectively. For cell sample preparation, confluent SH-EP1-h␣4␤2 cells were trypsinized, resuspended in medium, seeded on Lab Tek II CC2 chambered slides at a target density of ϳ50,000 cells/chamber, and allowed to grow at 37°C to achieve proper morphology and optimal confluence (ϳ90%). After ϳ2 days, cells were then briefly rinsed in 1ϫ phosphatebuffered saline (PBS), fixed in 4% paraformaldehyde, rinsed again in 1ϫ PBS for 5 min, and acetylated, delipidated with chloroform, and serially dehydrated with ethanol (50, 75, 85, and 95%) at 23 Ϯ 1°C. Samples were then incubated in pre-hybridization solution containing final concentrations of 250 g/ml tRNA, 25% formamide, 10% dextran sulfate, 2.5ϫ Denhardt's solution, 0.05 mg/ml salmon sperm DNA, 4ϫ SSCP, 4 mM EDTA, pH 8.0, for 2 h at 50°C before being incubated in hybridization solution containing antisense or sense probes (final concentration ranging from 0.2 to 0.7 ng/l) for 20 h at 50°C. After hybridization, samples were rinsed in 2ϫ SSC twice for 5 min each, 1ϫ SSC for 10 min, 0.5ϫ SSC for 10 min at 23 Ϯ 1°C, 0.1ϫ SSC for 20 min at 60°C, and briefly rinsed in double distilled water at 23 Ϯ 1°C. Samples were then dehydrated again in a graded ethanol series and vacuum dried under desiccant. Dry samples were then rehydrated by rinsing in 1ϫ PBS for 5 min before submersion in a 1:2000 dilution (in 1ϫ PBS supplemented with 0.1% bovine serum albumin) of avidin-Alexa fluorophore complexes (Molecular Probes, Eugene, OR) for 30 min. To terminate the reaction, samples were rinsed 3 times in 1ϫ PBS for 15 min each, again at 23 Ϯ 1°C. Slides were then coverslipped and stored at 4°C in the dark until being subjected to fluorescence microscopy (Olympus IX70, Melville, NY) and image analysis using ImagePro Plus version 4.1 (Silver Spring, MD).
Immunoprecipitation-Western Analysis for Confirmation of ␣4 and ␤2 Subunit Assembly-Preparation of non-ionic detergent-solubilized membranes for immunoprecipitation began with medium removal from 100-mm dishes containing plated cells, rinsing of dishes 3 times with NP buffer (100 mM NaCl, 25 mM NaPO 4 , pH 7.4), mechanical harvesting of cells using a rubber policeman, and centrifugation of healthy, confluent cells from the suspension for 5 min at 1,000 ϫ g. The cell pellets were then resuspended and homogenized in ice-cold NP buffer (200 l/confluent plate) supplemented with complete protease inhibitor mixture (1 Mini tablet per 10 ml; Roche Diagnostics), followed by centrifugation at 10,000 ϫ g for 10 min at 4°C. After centrifugation, the supernatant was then discarded and the pellet resuspended in TNP buffer (NP buffer supplemented with 1% Triton X-100; 200 l per ml of starting suspension; supplemented with protease inhibitor mixture) by incubating at room temperature for 30 min. The preparation was then centrifuged at 12,500 ϫ g for 10 min at 4°C. The supernatant fraction containing the solubilized membrane protein was then collected, and separate samples were reacted overnight at 4°C with either H133 (rabbit anti-nAChR ␣4 subunit) or H92 (rabbit anti-nAChR ␤2 subunit) antibodies (both from Santa Cruz Biotechnology, Santa Cruz, CA; 5 g/1 mg of solubilized total membrane protein) in the presence of protein G-agarose beads (50 l; Calbiochem). Immunoprecipitants were collected by centrifugation, and the beads containing immune complexes were washed 3 times in TNP buffer before being resuspended in SDS-PAGE sample buffer containing ␤-mercaptoethanol (20 l per sample; ICN Biomedicals), heated to 95°C for 3 min, and centrifuged briefly to remove the stripped agarose beads. The now-resolubilized protein samples were then collected and subjected to SDS-PAGE separation on 10% acrylamide gels and transferred to nitrocellulose membrane for Western analysis. Membranes were then blocked by PBS buffer containing 5% skim milk powder and incubated with primary monoclonal antibody 299 (rat anti-nAChR ␣4; 1:1000; Sigma) and pAb s-1724 (rabbit anti-nAChR ␤2; 1:50; a gift from Dr. J. Patrick, Baylor College of Medicine, Houston, TX) for 1 h. After three 5-min washes in PBS containing 0.05% Tween 20, secondary horseradish peroxidase-conjugated, goat anti-rat IgG (sc-2006) and donkey anti-rabbit IgG (sc-2313) antibodies (1:1000; Santa Cruz Biotechnology) were added to the reaction medium to allow conjugation with subunit-primary antibody targets. The horseradish peroxidase:second antibody-labeled bands were then visualized by using enhanced substrate Opti-4CN (Bio-Rad). Earlier control studies had demonstrated that there was no cross-reactivity on the nitrocellullose membranes between monoclonal antibody 299 and ␤2 subunits, between pAb s-1724 and ␣4 subunits, between secondary antibodies and nAChR subunits, or between goat anti-rat IgG antibody and any residual rabbit antibodies in immunoprecipitates (although cross-reactivity between donkey anti-rabbit IgG and rabbit antibodies carried over from immunoprecipitations was sometimes observed), ensuring that nAChR ␣4 and ␤2 subunits could be identified on the same Western blot when probed with both subunitspecific antibodies and both species-specific anti-IgG.
Patch Clamp Whole Cell Current Recordings-Conventional whole cell current recording, coupled with techniques for fast application and removal of drugs (U-tube), was applied in this study as previously described (38 -40). Briefly, cells plated on poly-lysine-coated 35-mm culture dishes were placed on the stage of an inverted microscope (Olympus iX7, Lake Success, NY) and continuously superfused with standard external solution (2 ml/min). Glass microelectrodes (3-5 M⍀ resistance between pipette and extracellular solutions) were used to form tight seals (Ͼ2 G⍀) on the cell surface until suction was applied to convert to conventional whole cell recording. Cells were then voltageclamped at a holding potential of Ϫ60 mV, and ion currents in response to application of ligands were measured (Axon Instruments 200B amplifier, Foster City, CA), typically using data filtered at 2 kHz, acquired at 5 kHz, displayed and digitized on-line (Axon Instruments Digidata 1200 series A/D board), and stored on hard media for subsequent off-line analysis. Both pipette and whole cell current capacitance were minimized, and the series resistance was routinely compensated to 80%. Before series resistance compensation, whole cell access resistance less than 20 M⍀ was accepted. Data acquisition and analyses were done using Pclamp8.0 (Axon Instruments), and results were plotted using Origin 5.0 (Microcal, North Hampton, MA). nAChR acute desensitization (the decline in inward current amplitude over the course of agonist application) was analyzed for decay half-time (; ϭ 0.693/k for decay rate constant k), peak current (I p ), and steady-state current (I s ), using fits to the mono-(or double-) exponential expression where I i is the intermediate level of current and k 1 and k 2 are rate constants from the two separate decay processes). Curve fitting usually was done using data between 90 and 10% of the difference between peak and steadystate currents. The experimental data are presented as mean Ϯ S.E., and comparisons of different conditions were analyzed for statistical significance using Student's t tests. All experiments were performed at room temperature (23 Ϯ 1°C). Dose-response profiles were fit to the Hill equation and analyzed using Origin 5.0.
Solutions and Drug Application-The standard external solution contained 120 mM NaCl, 3 mM KCl, 2 mM MgCl 2 , 2 mM CaCl 2 , 25 mM D-glucose, 10 mM HEPES, adjusted to pH 7.4 with Tris base. For conventional whole cell recordings, "Tris ϩ electrodes" were used to prevent receptor functional run-down (37,40), and were filled with solution containing 110 mM Tris phosphate dibasic, 28 mM Tris base, 11 mM EGTA, 2 mM MgCl 2 , 0.1 mM CaCl 2 , 4 mM Na-ATP, pH 7.3 (41). To initiate whole cell current responses, nicotinic agonists were delivered into the bath medium near the cell being recorded via a computercontrolled U-tube system so that solution exchange occurred within 20 ms (based on 10 -90% peak current rise times). Intervals between drug applications (3 min) were adjusted specifically to ensure stability of nAChR responsiveness (without functional rundown). Drugs used in The RT-PCR products were electrophoretically resolved on a 1% agarose gel along with a 100-bp ladder molecular weight marker. GAPDH serves as a positive control, indicating success of RT-PCR analysis, and specificity of ␣4 and ␤2 subunit expression only in transfected cells also is demonstrated. B, in situ hybridization showing expression of nAChR ␣4 and ␤2 subunit transcripts in the SH-EP1-h␣4␤2 cell line. Following the protocol described under "Experimental Procedures," cells were probed either with biotinylated ␣4 or ␤2 sense orientation riboprobes (a and b, respectively) or with biotinylated antisense ␣4 or ␤2 riboprobe (aЈ and bЈ, respectively). The hybridized probes were then labeled with avidin-Alexa 488, and the signal was subjected to epifluorescence microscopic analysis. Images were obtained at ϫ200 (a and aЈ) or ϫ600 (b and bЈ) magnification (calibration bars are 20 and 50 m, respectively). C, immunoprecipitation-Western analysis showing nAChR ␣4 and ␤2 subunit expression and co-assembly in SH-EP1-h␣4␤2 cells. Membrane fractions from untransfected SH-EP1 cells (lanes 2 and 4) or SH-EP1-h␣4␤2 cells (lanes 1 and 5) were detergent solubilized and subjected to immunoprecipitation using anti-␣4 antibody (lanes 1 and 2) or anti-␤2 antibody (lanes 4 and 5) before being subjected to SDS-PAGE, vacuum blotting, and Western analyses for both nAChR ␣4 subunit (ϳ70 kDa band; compare with pre-stained SDS-PAGE mass standards in lane 3 having sizes indicated to the left of the image) and ␤2 subunit (ϳ52-kDa band) proteins. The faint band of ϳ55 kDa in lanes 2 and 4 is not reproducibly observed and likely reflects cross-reaction of donkey anti-rabbit IgG with primary rabbit anti-subunit antibodies sometimes carried over into the sample. D, wild-type, untransfected SH-EP1 cells (blue arrow) or SH-EP1-h␣4␤2 cells heterologously expressing human ␣4 and ␤2 subunits (black arrow) were tested for whole cell current responses to 1 M nicotine applied for 4 s. No response was observed in wild-type cells, but an inward current was rapidly induced in transfected cells.
the present study were (Ϫ) nicotine, dihydro-␤-erythroidine (DH␤E), A␤ 40 -1 (Sigma), and A␤ 1-42 (rPeptide, Athens, GA). A␤ peptides were reconstituted per vendor specifications in distilled water to a concentration of 100 M, stored at Ϫ20°C, and used within 10 days of reconstitution. Thawed peptide solutions and working dilutions in standard extracellular solution were used within 4 h before being discarded.

Heterologously Expressed h␣4␤2-nAChRs in SH-EP1 Cells-
RT-PCR analysis confirmed expression of nAChR ␣4 and ␤2 subunit messages in SH-EP1-h␣4␤2 cells ( Fig. 1A; ␣4␤2). By contrast, whereas there was successful amplification of GAPDH message (positive control, lane c), no significant ␣4or ␤2-specific product was observed in the host cell line (wild-type SH-EP1 cells). Results of in situ hybridization studies consistently found an absence of staining for nAChR ␣4 or ␤2 subunit messages reacting with sense or antisense riboprobes in untransfected SH-EP1 host cells (data not shown) or reacting with sense riboprobes in SH-EP1-h␣4␤2 cells (Fig. 1B, a and b). However, for reactions employing antisense riboprobes, nearly all SH-EP1-h␣4␤2 cells displayed strong perinuclear and cytoplasmic hybridization signals for both ␣4 and ␤2 subunits (Fig.  1, B, aЈ and bЈ), respectively. Furthermore, apparent levels of ␣4 or ␤2 subunit transcripts did not seem to fluctuate as a function of cell confluence or between grouped or solitary cells (data not shown). Immunoprecipitation-Western analyses using solubilized membrane samples from transfected SH-EP1-h␣4␤2 cells clearly indicated that ␣4 and ␤2 subunits are expressed as protein and assembled together (Fig. 1C). Finally, screening using whole cell current recordings confirmed that functional responses to 1 M nicotine are evident in transfected SH-EP1-h␣4␤2 cells, but not in untransfected, wild-type SH-EP1 cells (Fig. 1D), indicating that transfected cells express functional h␣4␤2-nAChRs.
Basic Functional Properties of Transfected h␣4␤2-nAChRs-Whole cell current responses of heterologously expressed h␣4␤2-nAChRs show nicotine and acetylcholine (ACh) dose dependence ( Fig. 2A) that translates into a sigmoid dose-response relationship characterized by a functional EC 50 value for nicotine of 3.2 M (Hill coefficient of 0.81) and for ACh of 9.2 M (Hill coefficient of 0.6; Fig. 2B). Voltage dependence of h␣4␤2-nAChR responses to both nicotine and ACh (Fig. 2B) shows inward rectification at positive membrane potentials (Fig. 2C). Nicotine dose-response curves for h␣4␤2-nAChRmediated currents obtained in the presence of 0.3 or 1 M DH␤E, which is a relatively selective antagonist of ␣4␤2-nAChRs compared with other nAChR subtypes, are shifted to the right without showing an effect on peak whole cell current amplitude at maximal nicotine concentration (Fig. 2D). Compared with the EC 50 value for nicotine alone (3.3 Ϯ 0.6 M, mean Ϯ S.E. for separate EC 50 measures from replicate experiments, n ϭ 10), the EC 50 obtained for nicotine in the presence of 300 nM DH␤E is higher, but not significantly (4.7 Ϯ 1.2 M, Each symbol is the average response from 7 to 8 cells. D, the sigmoidal nicotine concentration-peak current response curves in the absence or presence of 300 nM or 1 M DH␤E show that DH␤E shifted the nicotine concentration-response curve to the right without affecting the maximal response, suggesting a competitive mechanism of antagonism. All symbols were normalized to the 100 M nicotinic response (marked with the arrowhead), and each symbol is the average from 4 to 8 cells.
p Ͼ 0.05, n ϭ 11), but the EC 50 obtained in the presence of nicotine plus 1 M DH␤E is significantly higher (9.6 Ϯ 1.5 M, p Ͻ 0.05, n ϭ 4), but nicotine efficacy is not reduced in the presence of DH␤E. Thus, DH␤E exhibits features of competitive antagonism acting on heterologously expressed h␣4␤2-nAChRs. These basic electrophysiological and pharmacological properties indicate that heterologously expressed h␣4␤2-nAChRs in SH-EP1 cells exhibit characteristic features of the subtype and validate SH-EP1-h␣4␤2 cells as models for studying h␣4␤2-nAChR function.
A␤  Acutely Inhibits h␣4␤2-nAChR-mediated Whole Cell Currents-Initial studies done to examine the acute effects of 1 nM A␤ 1-42 on h␣4␤2-nAChR responses to nicotine or ACh indicated that pretreatment (typically done for 3 min) with A␤ peptide was necessary to elicit a statistically significant, functional inhibition of peak current responses (reduced to 75.3 Ϯ 2.1% of control values for nicotine, n ϭ 18, p Ͻ 0.01; reduced to 77.4 Ϯ 3.1% of control values for ACh, n ϭ 8, p Ͻ 0.01; Fig. 3,  A and B). A␤ 1-42 exposure also caused reduction of steady-state whole cell current responses to nicotine (reduced to 75.5 Ϯ 2.2% of control values, n ϭ 18, p Ͻ 0.01) or ACh (reduced to 71.5 Ϯ 4.2% of control values, n ϭ 8, p Ͻ 0.01; Fig. 3, A and B). Increases in rates and extents of acute desensitization, assessed as time constants for inward current decay from peak to steady-state levels and as the magnitude of the steady-state inward current, were evident for A␤ 1-42 action on h␣4␤2-nAChRs responding to nicotine or ACh (Fig. 3A, traces). Because there were no differences in effects of A␤ 1-42 on ACh-or nicotine-induced h␣4␤2-nAChR function, the remainder of the studies focused on the use of nicotine as the prototypical nicotinic agonist. Interestingly, without pretreatment, 1 nM A␤ 1-42 did not show significant reduction of peak current responses to nicotine (decreased to 94.3 Ϯ 2.9% of control values, n ϭ 17, p Ͼ 0.05), but the peptide reduced steady-state currents to 76 Ϯ 6.0% of control values (n ϭ 17, p Ͻ 0.01; data not illustrated). Studies using the reverse-sequence peptide A␤ 40 -1 failed to show significant inhibition of h␣4␤2-nAChR function (data also not illustrated), suggesting that effects were peptide sequencespecific. An additional set of studies examining the effects of different concentrations of A␤ 1-42 on responses to 1 M nicotine showed dose-dependence of h␣4␤2-nAChR functional block (Fig. 3, C and D). Collectively, these results indicate that A␤ 1-42 acutely inhibits h␣4␤2-nAChR function.
Possible Mechanisms Involved in A␤  Inhibition of h␣4␤2-nAChR Function-Dose-response profiles of nicotine-induced whole cell peak currents in the absence or presence of 1 nM A␤ 1-42 showed similar magnitudes of peptide-induced functional inhibition and no dramatic change in nicotine EC 50 values or Hill coefficients (3.3 Ϯ 0.6 and 0.97 Ϯ 0.07 M, respectively, for nicotine alone, n ϭ 10; 4.1 Ϯ 0.9 and 1.1 Ϯ 0.1 M, respectively, for nicotine plus 1 nM A␤ 1-42 , n ϭ 5, p Ͼ 0.05), suggesting a non-competitive mechanism of inhibition (Fig. 4,  A and B). Other studies indicated that 1 nM A␤  showed similar inhibition of 100 M nicotinic-induced responses at different holding potentials (V H ), suggesting a voltage-independent means of functional inhibition by A␤ 1-42 (Fig. 4, C and D). show clear signs of use-dependent inhibition of h␣4␤2-nAChRs (Fig. 5, A and C). Interestingly, after 15 min of continuous exposure to 1 nM A␤ 1-42 followed by peptide washout for 6 min, peak current responses to nicotine did not recover fully (Fig. 5,  A and C). The absence of use-dependence suggests that noncompetitive inhibition of h␣4␤2-nAChR function by A␤ 1-42 is not mediated through open channel block, although the persistence in functional block after washout of fluid phase A␤ suggests a lingering effect of the peptide (see "Discussion"). To address whether A␤ 1-42 -induced h␣4␤2-nAChR internalization is involved in this incomplete functional recovery after relatively long term exposure (15 min) of h␣4␤2-nAChRs to 1 nM A␤ 1-42 , we pre-loaded recorded cells with GDP␤S (600 M) for 20 min, which was reported to prevent ␣-amino-3-hydroxy-5methyl-4-isoxazolepropionate receptor endocytosis (42), and we repeated the experiment shown in Fig. 5A. GDP␤S treatment neither prevented A␤ 1-42 -mediated inhibition, nor improved h␣4␤2-nAChR functional recovery from washout of A␤ 1-42 (Fig. 5, B and C), suggesting that the inhibition of h␣4␤2-nAChR function by A␤ 1-42 is not mediated via h␣4␤2-nAChR internalization through the process affecting ␣-amino-3-hydroxy-5-methyl-4-isoxazolepropionate receptors.
Acceleration of h␣4␤2-nAChR Acute Desensitization by A␤  Is Not Influenced by ␤2 Subunit Intracellular Domains-The data presented thus far indicate that 1 nM A␤ 1-42 accelerates h␣4␤2-nAChR desensitization. We considered whether this modulation could be via mechanisms involving sequences in the nAChR ␤2 subunit large, cytoplasmic loop that are potentially involved in nAChR phosphorylation (46), possibly engaged in coupling to the cytoskeleton or cytoplasmic modulators, or perhaps involved in some other intracellular mechanism mediating A␤ 1-42 action. To test these possibilities, we created chimeric ␤2 (␤2ch) subunits whose natural cytoplasmic loop sequences were replaced by those from the mouse serotonin receptor 5-TH 3A subunit sequence. ␣4␤2ch-nAChRmediated currents showed significant acceleration of receptor desensitization compared with rates of desensitization for fully wild-type ␣4␤2-nAChRs (Fig. 6, A and B). However, 1 nM A␤ 1-42 further accelerated receptor desensitization and lowered peak and steady-state components of nicotine-induced whole cell currents in either ␣4␤2or ␣4␤2ch-nAChRs (Fig. 6,  C and D). These results suggest that accelerated ␣4␤2-nAChR desensitization upon A␤ 1-42 exposure is not influenced by changes in the nAChR ␤2 subunit cytoplasmic loop sequence.
Selective Inhibition of h␣4␤2-nAChR Function at Pathophysiologically Relevant Concentrations of A␤ 1-42 -Given reports of interactions at high concentrations (100 -3000 nM) of A␤ 1-42 with ␣7-nAChRs (22,23,25,26) and rat ␣4␤2-nAChRs (24), we compared effects of A␤ 1-42 on function of heterologously expressed h␣4␤2-nAChRs to effects on heterologously expressed h␣7-nAChRs expressed in the same cellular host (40). A␤ 1-42 , at a pathophysiologically relevant concentration (1 nM), clearly inhibits nicotinic peak current responses mediated by h␣4␤2-nAChRs (inhibition to 76% of control values; Fig. 7A, left), but not responses mediated by h␣7-nAChRs (Fig.  7B, left). However, at a higher concentration (100 nM), A␤  inhibits both h␣4␤2-nAChR-(to 68% of control values; Fig. 7A, right) and h␣7-nAChR-mediated peak currents (to 41% of control values; Fig. 7B, right), although the magnitude of functional inhibition of ␣7-nAChRs is greater. Interestingly, the  ). B, results of replicate studies (as in A) fit to the logistic equation indicate a downward shift reflecting reduction of maximal current response in the agonist dose-response profile in the presence of A␤ 1-42 , but without a change in apparent EC 50 for agonist, characteristic of non-competitive inhibition. All symbols were normalized to the peak current induced by 100 M nicotine (marked by the arrowhead), and each symbol is the average from 10 cells for nicotine, and 5 cells for nicotine plus 1 nM A␤  . C, representative whole cell current traces in response to 100 M nicotine alone (higher amplitude traces) or in the presence of 1 nM A␤ 1-42 (lower amplitude traces) at different holding potentials as indicated. D, plots of whole cell peak current amplitude as a function of holding potentials. All symbols were normalized to nicotine-induced peak current at a V H of Ϫ80 mV (marked by the arrowhead) and are averages of recordings from 5 to 6 cells. Vertical bars indicate standard errors.
inhibition of h␣4␤2-nAChR-mediated currents by 1 or 100 nM A␤ 1-42 appears to occur through different mechanisms because 1 nM A␤ 1-42 accelerated the rate of current decay from peak to steady-state levels without changing the half-time for achievement of peak currents, whereas 100 nM A␤ 1-42 slowed current rising time without changing decay time (Figs. 3D and 7, A and  C). Nevertheless, these results indicate that heterologously expressed h␣4␤2-nAChRs are more sensitive than h␣7-nAChRs to pathophysiologically relevant concentrations (1 nM) of A␤ 1-42 .

DISCUSSION
Principal Observations-These studies validate SH-EP1-h␣4␤2 cells as a model for investigation of the properties of h␣4␤2-nAChRs. The cells express nAChR ␣4 and ␤2 subunits as mRNAs and receptor subunit proteins that co-assemble to yield nAChRs that show functional responses to nicotine or ACh that are sensitive to competitive inhibition by DH␤E and that are manifest as inwardly rectifying whole cell currents on patch clamp analysis. A␤ 1-42 peptides, but not control A␤ 40 -1 peptides, inhibit function of h␣4␤2-nAChRs at concentrations as low as 1 nM. Effects are manifest as a reduction in whole cell peak and steady-state currents and acceleration of rates of desensitization. Effects show no use-dependence or voltagedependence and persist after bulk removal of peptide in solution. Features of functional inhibition are the same for ␣4␤2-nAChRs and nAChRs composed of ␣4 subunits and a chimeric form of ␤2 subunits containing alternate cytoplasmic loop amino acids. A␤ 1-42 -mediated inhibition of human ␣7-nAChR function also is observed, but at higher concentrations (100 nM).
Possible Mechanisms of A␤ 1-42 -mediated Inhibition of h␣4␤2-nAChR Function-The present results clearly support the hypothesis that A␤ directly modulates nAChR function. Our data shows that A␤ 1-42 , at concentrations from 1 pM to 1 M, failed to directly induce whole cell current responses from cells expressing h␣4␤2-or h␣7-nAChRs or from untransfected cells, meaning that A␤, in our hands, and although signal transduction through other means remains a possibility, does not have properties of a nAChR agonist able to induce whole cell currents. This finding is in agreement with previous reports that A␤ 1-42 failed to elicit naturally expressed ␣7-nAChR-mediated currents in cultured hippocampal neurons (23) or heterologously expressed ␣7-nAChR responses in Xenopus oocytes (25), but it is not consistent with A␤ 1-42 binding to rat ␣7or non-␣7-nAChRs (likely ␣4␤2-nAChRs) at an agonist, co-agonist, or other allosteric, positive modulatory site involved in eliciting or enhancing whole cell currents (26,43). Agonism (or competitive antagonism) also is inconsistent with our repeated findings 2 failing to detect direct competition of picomolar to micromolar A␤ 1-42 with radioligand binding sites on heterologously expressed h␣4␤2-or h␣7-nAChRs. On the contrary, our findings indicate that A␤ 1-42 acts as a non-competitive antagonist of h␣4␤2-nAChRs under our experimental conditions. A non-competitive mechanism of naturally or heterologously expressed ␣7-nAChR functional block by A␤ 1-42 also has been invoked (22,25).
As to the domain(s) involved in non-competitive block, our data indicating that effects of A␤ 1-42 on h␣4␤2-nAChR function are not affected by changes in amino acid sequences corresponding to the ␤2 subunit second, major intracellular domain, suggests an alternate site of action. Other chimera studies of ␣7-nAChRs more directly indicated a role for the NH 2 -terminal extracellular domain of the ␣7 subunit in A␤ effects on that subtype (23). Our data showing no use or voltage dependence of A␤ 1-42 inhibition of h␣4␤2-nAChR function suggests that effects are not mediated via open channel block. In FIG. 6. A␤ 1-42 inhibits both wild-type h␣4␤2-nAChR-and h␣4␤2-chimera-nAChR-mediated currents. A, representative whole cell current traces for responses to 1 M nicotine (a) or 1 mM ACh (b) of fully wild-type h␣4␤2-nAChRs and nAChRs containing wild-type ␣4 and chimeric ␤2 subunits (␣4␤2-chimera). Right panels superimpose the nicotinic responses from the left and middle panels. B, bar graphs compare 1 M nicotine-induced whole cell currents recorded from transfected wild-type h␣4␤2-nAChRs and ␣4␤2-chimera-nAChRs. The normalized wild-type h␣4␤2-nAChR-mediated currents (as 100%) are presented as a horizontal dashed line. I p and I s mean the peak and steady-state currents, respectively. C, representative whole cell current traces for responses to 1 M nicotine alone or in the presence of 1 nM A␤ 1-42 (lower amplitude traces) for wild-type h␣4␤2-nAChRs (a) or for ␣4␤2-chimera-nAChRs (b). D, summary of effects of 1 nM A␤ 1-42 on peak and steady-state current amplitudes and on fast and slow phases of nAChR desensitization in 4 -14 replicate studies. The asterisk means p Ͻ 0.05 and the triple asterisk means p Ͻ 0.001. Vertical bars indicate standard errors. agreement with other studies (23)(24)(25)(26), we also found that pretreatment with A␤ 1-42 is necessary to induce convincing inhibition of h␣4␤2-nAChR-mediated nicotinic responses. The nature of this slow onset of inhibition is still unknown, although it has been hypothesized that A␤ 1-42 may drive nAChRs into a long-lived closed conformation. This hypothesis is supported by the present observation that long exposure (15 min) to A␤  led to persistent loss of h␣4␤2-nAChR function, although any loss of nAChR function does not seem to be mediated via A␤ 1-42 -induced h␣4␤2-nAChR internalization. Our finding that A␤ 1-42 exposure accelerated rates of h␣4␤2-nAChR functional desensitization is consistent with another report that A␤ 1-42 accelerated the rate of current decay for muscle-type nAChRs (25). A possible role of changes in nAChR phosphorylation state in regulation of receptor desensitization has been proposed (46), but our studies suggest that any such role for h␣4␤2-nAChR phosphorylation would be because of changes in ␣4 subunit phosphorylation state.
Pathological Relevance of ␣4␤2-nAChR Dysfunction and AD-nAChRs that bind radiolabeled nicotine with the highest affinity contain ␣4 subunits (␣4*-nAChR; for review, see Refs. [47][48][49][50][51]. Immunoassays have shown that the predominant, naturally expressed form of the ␣4*-nAChR in the vertebrate brain contains ␣4 and ␤2 subunits (␣4␤2-nAChR) (17,18). Evidence indicates that a consistent, significant loss of ␣4*-nAChRs has been observed at autopsy in a number of neocortical areas and hippocampi of patients with AD (for reviews, see Refs. 52 and 53). Cortical ␣4*-nAChR deficits are significantly correlated with cognitive impairment in AD patients (54,55). The major pathological features of AD are A␤ protein deposition and a severe cholinergic deficit. It has been shown that A␤ protein serves as a major constituent of senile plaques, a neuropathologic hallmark of AD and a neurotoxin through in vivo and in vitro studies. A␤ 1-42 accumulation and aggregation are thought to contribute to the pathology of AD. However, links between soluble A␤ 1-42 elevation and cholinergic dysfunction remain unclear. The present studies provide the basis for a new hypothesis that A␤ 1-42 -induced h␣4␤2-nAChR functional depression may contribute to cholinergic signaling deficits and/or neurodegeneration that are hallmarks of AD. This hypothesis is not necessarily exclusive of, for example, other mechanisms that would produce selective death of cholinergic neurons, would result in destabilization of membrane properties by A␤ and/or its precursors (56 -59), or would reflect effects of A␤ on intracellular Ca 2ϩ homeostasis through either production of cation ionophores or activation of ligand-and/or voltage-gated channels (60,61). Our perspective on a primary role for h␣4␤2-nAChRs in low concentration effects of A␤ complements other suggestions that the modulation of nAChR function by A␤ could be pathologically relevant (20 -26, 43).
In conclusion, this study provides evidence for a possible link between abnormal deposits of A␤ 1-42 and h␣4␤2-nAChR dys- function in AD patients. Furthermore, the finding that human ␣4␤2-nAChRs are sensitive targets of pathophysiologically relevant concentrations of A␤ 1-42 suggests novel strategies for AD therapy that would involve amelioration of peptide-induced ␣4␤2-nAChR functional block and/or perhaps preservation of ␣4␤2-nAChR activity.