Actions of β-Amyloid Protein on Human Neurons Are Expressed through the Amylin Receptor

Electrophysiological Recordings from HFNs

HFNs, grown on coverslips, were bathed with oxygenated artificial cerebrospinal fluid that contained (in mmol/L) 140 NaCl, 2.5 KCl, 1.5 CaCl₂, 1.2 MgCl₂, 10 HEPES, and 33 d-glucose (pH 7.4). Whole cell patch-clamp recordings were performed at room temperature (20°C–22°C) using an Axopatch-200B amplifier in combination with a 1200A interface (Axon Instruments, Foster City, CA). Patch electrodes (thin wall with filament, 1.5-mm diameter; World Precision Instruments, Sarasota, FL) were pulled (PP-83 electrode puller; Narishige Scientific Instrument Lab, Tokyo, Japan) to yield resistances of 3–6 MΩ. The internal patch pipette solution contained (in mM) 140 K-methylsulfate, 10 EGTA, 5 MgCl₂, 1 CaCl₂, 10 HEPES, 2.2 Na₂-ATP, and 0.3 Na-GTP (pH 7.2). All whole-cell recordings were made in current-clamp and voltage-clamp mode, and bridge balance and capacitance compensation were used. After whole-cell configuration was established with voltage-clamp mode (holding potential = −80 mV), we waited at least 5 minutes for the cell to stabilize, then started either voltage-clamp studies or switched to current-clamp recording mode. The current and membrane voltages were recorded using a low-pass filter at 5 kHz and were digitized at 10 kHz. All data were acquired and analyzed using pClamp8 software v8.0 (Axon Instruments). Whole-cell currents were activated by a voltage-ramp protocol, where the cells were held at −80 mV and subjected to voltage ramps from −110 to +30 mV at the rate of 20 mV/s. A 1-second-long hyperpolarizing command to −110 mV was applied to remove inactivation of K⁺ channels so that maximum current could be activated during the subsequent slow voltage ramp to +30 mV. Ca²⁺-dependent K⁺ conductances were isolated using the BK channel-specific blocker iberiotoxin. Drugs were either bath-applied or delivered via a focal applicator.

Primary Cultures of HFNs

Neuronal cultures were prepared from 12- to 15-gestational week fetuses with approval of the Human Ethics Research Board at the University of Alberta. The meninges and blood vessels were removed, the brain tissue was washed in minimum essential medium and mechanically dissociated by repeated trituration through a 20-gauge needle. Cells were centrifuged at 1500 g for 10 minutes and resuspended in minimum essential medium with 10% heat-inactivated fetal bovine serum, 0.2% N2 supplement, and 1% antibiotic solution (104 U of penicillin G per milliliter, 10 mg streptomycin per milliliter, and 25 mg amphotericin B per milliliter in 0.9% NaCl).²⁷ Subsequently, cultures were treated with arabinofuranosylcytosine (25 μmol/L) for 2 weeks and were plated (at a density of 5 × 10⁻⁵ per well) on 96-well plates. HFN cultures were grown in a 5% CO₂ humidified incubator at 37°C. Sample wells were immunostained for the neuronal marker microtubule-associated protein 2, and only cultures in which >70% of the cells stained positive for the marker were used for experiments. All reagents were obtained from Invitrogen (Burlington, ON, Canada) and antibodies from Sigma-Aldrich (Oakville, ON, Canada).

Treatments and Cell Death Assays

Soluble oligomeric Aβ_1-42 or the inverse sequence peptide Aβ_42-1, human amylin, and AC253 were prepared according to published protocols.^28,29 All peptides were purchased from American Peptide (Sunnyvale, CA). To determine the dose-dependent toxicity of Aβ peptides and human amylin. HFNs were treated with different concentrations of the peptides (0.5–50 μmol/L) (see Supplemental Figure S1 at http://ajp.amjpathol.org). In each experiment and in subsequent ones described below, two rows of eight wells each (of a 96-well plate) received the same treatment, and each experiment was repeated a minimum of four times. To evaluate the neuroprotective effects of the amylin receptor antagonist AC253 against Aβ toxicity, cultured HFNs were preexposed to AC253 (10 μmol/L) for 24 hours and then to soluble oligomeric 20 μmol/L Aβ_1-42 or human amylin (2 μmol/L). Cells in adjacent rows of wells received applications of either 20 μmol/L Aβ_1-42, 2 μmol/L human amylin, or 10 μmol/L or AC253. Controls included applications of the inverse sequence Aβ peptide Aβ_42-1 or no drug. After 48 hours, the control and treated cultures were examined for neuronal survival using a 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay (Sigma-Aldrich).^26,30

For experiments examining the effects of caspase inhibition on Aβ toxicity, the cell cultures were pretreated for 2 hours with individual caspase inhibitors (20 μmol/L; kit from RD Systems, Minneapolis, MN), followed by exposure to 20 μmol/L Aβ_1-42 for 48 hours. At the end of this period, the cultures were processed for MTT assay. Each experiment was conducted four times.

Immunoblotting

Western blotting was performed as described previously.^26,30 Briefly, samples of control and treated groups of cultured cells (with AC253, Aβ_1-42, or AC253 pretreatment followed by Aβ_1-42) with equal amounts of protein were separated by 12% polyacrylamide gel electrophoresis and the resolved proteins were transferred onto nitrocellulose membranes and probed with anti-caspase 3, 6, and 9 antibodies and also with antibodies to pro- and antiapoptotic intermediaries (cytochrome c, SMAC, PUMA, Bax, and Bcl-2; New England Biolabs, Ipswich, MA). Blots were also probed with anti-β actin (Abcam, Cambridge, MA) as loading control. To assess whether AC253 also blocked Aβ-induced caspase-independent cell death, we performed Western blot and immunohistochemistry to detect cleavage and translocation of apoptosis inducing factor (AIF) from the activated mitochondria using AIF antibody (Santa Cruz Biotechnology, Santa Cruz, CA) and MitoTracker Red dye (Molecular Devices, Sunnyvale, CA). Each experiment was repeated four times. For all quantitative analysis of Western blots, Image J software v1.44j (National Institutes of Health, Bethesda, MD) was used. Relative intensities were calculated by comparing the intensity of protein bands of interest to control β-actin loading bands.

siRNA

siRNA corresponding to the human CTR (NM_001742) and RAMP3 (NM_005856) genes, purchased from Open BioSystems (Huntsville, AL), were inserted into a vector (pGIPz) that produced green fluorescent protein. HFNs were plated at a density of 2 million cells/ml in a 6-well plate and cultured for 24 hours before transfection. On the day of transfection, culture media was replaced with OptiMEM (Invitrogen). Four micrograms of plasmid siRNA mixed with Lipofectamine 2000 (Invitrogen) was incubated for 30 minutes and then added to the cultures. Cultures were gently mixed and placed in an incubator at 37°C with 5% CO₂ for 10 hours, after which normal culture media was replaced. Subsequently, transfected and non-transfected HFN cultures in 96-well plates were exposed to soluble oligomeric Aβ_1-42 (20 μmol/L) or the inverse sequence peptide, Aβ_42-1 (20 μmol/L), and cell survival was measured using the MTT assay.

Enzyme-Linked Immunosorbent Assay for Measurement of Aβ Levels in Brain Regions of TgCRND8 APP and Control Mice

Brains were quickly removed from 1-, 4-, and 6-month-old TgCRND8 APP mice (human APP695 transgene array incorporating Swedish K670M/N671L and Indiana V717F mutations superimposed on a C57BL6/C3H genetic background) and from control nontransgenic C57BL6/C3H littermate mice. The cortex, hippocampus, diagonal band of Broca, brainstem, and cerebellum were dissected and weighed before the preparation of normalized lysates. Proteins from each dissected region were isolated according to the protocol provided with the Aβ_1-42 enzyme-linked immunosorbent assay kit (BioSource, Burlington, ON, Canada). In brief, in this protocol a monoclonal antibody specific to the NH-₂ terminus of the Aβ_1-42 is coated to the wells of microtiter strips. Proteins are incubated in the well. A secondary antibody specific to the COOH-terminus of the Aβ_1-42 sequence is then incubated in the wells. Bound rabbit antibody is detected with a horseradish peroxidase-labeled anti-rabbit antibody. Excess anti-rabbit antibody is removed by washing, and then a substrate solution is added and is converted, producing a color. The color is measured at 450 nm and the intensity of color is directly proportional to the amount of Aβ_1-42 in the tissue.

Expression of Amylin Receptor in Brain Regions of TgCRND8 APP-Overexpressing Mice

Under halothane anesthesia, TgCRND8 APP mice (1, 4, and 6 months of age, n = 5 for each age group) and age-matched control mice (n = 5 for each age group) were decapitated and brains were quickly removed. Brains were hemisected and one half of the brain was used for dissecting the cortex, hippocampus, diagonal band of Broca, cerebellum, and brainstem. Tissue was weighed and the placed in cold RIPA (radio-immunoprecipitation assay) buffer with protease inhibitors and proteins were isolated and measured using a Bio-Rad protein assay kit (Mississauga, ON, Canada). Proteins were diluted with SDS sample buffer (New England Biolabs) and were loaded at 30 μg per lane on a 12% agarose gel. Proteins were transferred to nitrocellulose blocked with nonfat milk. Nitrocellulose membranes were incubated with antibodies overnight at 4°C on a shaker. Blots were incubated in the following antibodies, one at a time, and were stripped between each application: CTR (1:1000 rabbit; kindly supplied by Professor P. Sexton, Monash University, VIC, Australia), RAMP3 (1:1000 goat; Santa Cruz Biotechnology), 6E10 for Aβ (1:5000 mouse; 6E10; Signet Laboratories, Dedham, MA), and β-actin (1:1000 mouse; Sigma-Aldrich). Blots were subsequently incubated with corresponding secondary antibodies labeled with horseradish peroxidase and were washed with Immobilon Western chemiluminescent substrate (Millipore, Billerica, MA). Blots were exposed to X-ray film (Kodak BioMax MR) for 10 seconds to 5 minutes. X-ray blots were then scanned and analyzed with Image J software. The other half of the brains from TgCRND8 and control mice was fixed for immunohistochemical analysis of brain sections from cortex and hippocampus to determine Aβ plaque deposition, and the same sections were also double-stained using the CTR antibody.

Statistical Analysis

Data are presented as means ± SEM. Unless otherwise indicated, group data were compared using one-way analysis of variance followed by Newman-Keuls post hoc test with the P-value set at 0.05.

Electrophysiological Effects of Aβ Are Blocked by Amylin Receptor Antagonist

In HFNs, applications of oligomeric Aβ_1-42 caused a time-independent and concentration-dependent reduction in whole-cell outward currents (WCCs) in the voltage range −30 to +30 mV (see Supplemental Figure S2 at http://ajp.amjpathol.org). WCCs were not affected by applications of the inverse sequence Aβ_42-1 peptide (data not shown). AC253 is a selective peptidergic amylin receptor antagonist that, applied in the concentration range 10⁻⁵ to 10⁻⁶ mol/L, has been shown to block the actions of human amylin.^17,31 We investigated whether the effects of Aβ could also be blocked using this amylin receptor antagonist. Application of 10 nmol/L AC253 blocked Aβ_1-42-induced reduction in WCCs (20 nmol/L, n = 8, P < 0.01) (Figure 1, A and B). Peak WCC (at +30 mV) in the presence of Aβ_1-42 was significantly reduced, compared with the control level of current (27.7%, n = 8, P < 0.01), but in the presence of AC253, the Aβ_1-42-evoked response was abolished (Figure 1C). When the high conductance Ca²⁺-activated K⁺ channels Ic (BK channels) were inhibited with iberiotoxin (25 nmol/L), Aβ_1-42-induced reduction in WCCs was significantly blocked (Figure 1D), indicating that Aβ effects are expressed via this ionic conductance.

Under current-clamp conditions, the number of spikes generated by injection of a current pulse in HFNs was significantly increased in the presence of Aβ_1-42 (10 nmol/L), compared with control conditions (Figure 2A). Application of AC253 (10 nmol/L) blocked the increase in excitability induced by Aβ_1-42. The average number of spikes elicited by current injection was 1.2 ± 0.3 Hz under control conditions, which significantly increased to 16.6 ± 3.5 Hz in the presence of Aβ_1-42 (Figure 2B) (n = 5, P < 0.01). In the presence of AC253, application of Aβ_1-42 did not result in a significant increase in firing (n = 5, P > 0.05), compared with control. Under resting conditions, application of Aβ_1-42 (10 nmol/L) to HFNs resulted in membrane depolarization and an increase in action potential firing, which were blocked in the presence of AC253 (Figure 2C).

Aβ Neurotoxicity Is Expressed through the Amylin Receptor

To determine whether Aβ toxicity is expressed via amylin receptors, we used primary cultures of HFNs to initially establish that application of oligomeric Aβ_1-42 (see Supplemental Figure S3 at http://ajp.amjpathol.org and also the Materials and Methods section) or human amylin evokes concentration-dependent cell death in such neurons (see Supplemental Figure S1 at http://ajp.amjpathol.org). Pretreatment of HFNs for 24 hours with the amylin receptor antagonist AC253 (10 μmol/L) resulted in a significant improvement in neuronal survival of HFNs exposed to either Aβ_1-42 (20 μmol/L) or human amylin (2 μmol/L) (Figure 3, A and B). Concentrations of extracellularly applied oligomeric Aβ_1-42 used were based on our data (see Supplemental Figure S1 at http://ajp.amjpathol.org) and data reported in the literature for studies on human neurons.^32,33,34

To corroborate our pharmacological observations concerning neuroprotective effects of AC253, we also examined whether down-regulation of the amylin receptor using an siRNA approach blunts Aβ_1-42-induced neurotoxicity in HFNs. To suppress amylin receptor in HFNs, we generated siRNA corresponding to the two components of the amylin receptor, the human CTR (NM_001742) and RAMP3 (NM_005856) genes, and inserted these into a vector (pGIPz) that produces green fluorescent protein (Figure 4A). Controls were packaged with scrambled siRNA. We first confirmed amylin receptor suppression by the siRNAs in HFNs using separately validated antibodies (Figure 4B). HFN cultures were exposed to Aβ_1-42 at 4 days after transfection and neuronal viability was assessed. Knockdown of the amylin receptor using siRNA rendered HFNs more resistant to Aβ_1-42-induced cell death (Figure 4C).

Amylin Receptor Blockade Attenuates Aβ-Induced Caspase-Dependent and -Independent Apoptotic Cell Death

In animal in vitro models, Aβ may induce apoptosis by activation of a number of cell-death signaling pathways involving caspases, a family of cysteine proteases.^35,36 Alternatively, Aβ-induced cell death may be caspase-independent, occurring through mitochondrial release, cleavage, and translocation of AIF to the nucleus with consequent DNA fragmentation and nuclear condensation.³⁷ We initially identified caspases that are relevant in our HFN model of Aβ-induced apoptosis. HFNs pretreated with the pan-caspase inhibitor Z-VAD-FMK or with inhibitors of specific caspases (3, 6, 8, and 9) demonstrated significant resistance to oligomeric Aβ_1-42 neurotoxicity (see Supplemental Figure S4 at http://ajp.amjpathol.org). Subsequently, we determined whether antagonism of the amylin receptor is able to block the Aβ-induced activation of specific caspase pathways. HFN cultures were initially treated with oligomeric Aβ_1-42 and time-dependent activation of caspases 3, 6, and 9 was detected using antibodies to the cleaved forms of these caspases (Figure 5A). Pretreatment of HFNs with the amylin receptor antagonist AC253 for 24 hours effectively blocked the cleavage of caspases 3, 6, and 9 induced by oligomeric Aβ_1-42 (Figure 5B). Quantitative analysis of pro-caspase 3 confirmed its reduction (due to conversion to cleaved caspase 3 product) after exposure of HFNs to Aβ_1-42, but not Aβ_42-1 (see Supplemental Figure S5A at http://ajp.amjpathol.org). In the presence of AC253, the Aβ_1-42-induced decrease in the pro-caspase 3 substrate was significantly attenuated, indicating that less of the activated (cleaved) caspase 3 was generated to cause apoptotic cell death.

A similar quantitative analysis revealed that the Aβ_1-42-evoked increase in cleaved caspase 9 could be blocked by AC253 pretreatment (see Supplemental Figure S5B at http://ajp.amjpathol.org). Proapoptotic intermediaries in the caspase 9 pathway, including cytochrome c, Bax, SMAC, and PUMA, are also increased after exposure to Aβ_1-42. Pretreatment with AC253 significantly attenuated or blocked not only the cleavage of caspases 9 but also activation of cytochrome c, BAX, SMAC, and PUMA (Figure 6). At the same time, Aβ_1-42-evoked reduction in antiapoptotic mediators Bcl-2, Akt, and PI3-kinase was attenuated by AC253 (Figure 6).

To determine whether blockade of the amylin receptor also attenuated apoptotic cell death via a caspase-independent pathway, we examined the effects of AC253 pretreatment on Aβ_1-42-induced cleavage and translocation of AIF from the mitochondria to the nucleus. Aβ_1-42 (20 μmol/L), but not Aβ_42-1 (20 μmol/L), caused a time-dependent increase in AIF from its 67-kDa form to the cleaved 57-kDa moiety (Figure 7A), which was significantly reduced in HFNs pretreated with AC253 (Figure 7, B and C). Exposure of HFNs to Aβ_1-42 under the same conditions also resulted in activation of activation of caspase 9 through a caspase-dependent pathway (Figure 5). To confirm cellular localization of AIF within the mitochondria and its translocation to the nucleus, we used immunofluorescence methods in SK-N-SH neuroblastoma cells, which (by virtue of their larger size, relative to HFNs) provide better visualization of the cytoplasm and the spatial relationship of the mitochondria to the nucleus. Aβ_1-42-treated SK-N-SH cells demonstrated a significant decrease in retention of AIF within the mitochondria (identified using MitoTracker Red dye), but a corresponding increase in AIF within the nuclei (identified using the nuclear stain 4′,6-diamidino-2-phenylindole), compared with control cultures (Figure 7D). This nuclear translocation of AIF was not apparent in SK-N-SH cells pretreated with AC253 and subsequently exposed to Aβ_1-42. In such cells, AIF was retained within the cytoplasm and colocalized with the mitochondrial-specific marker.

Up-Regulation of Amylin Receptor in Brains of Transgenic Mice That Overexpress Aβ

We sought to determine the relationship between expression of two components of the amylin receptor, RAMP3 and CTR, and the amount of Aβ present in the brains of APP-overexpressing transgenic mice (TgCRND8 APP swe/APP+/−V717F), which serve as a surrogate in vivo model of AD.³⁸ Western blot analyses of brains from TgCRND8 mice at 1, 4, and 6 months of age revealed an age-dependent up-regulation of CTR and RAMP3 (Figure 8). Brains of age-matched control mice did not reveal an increase in either amylin receptor components or Aβ. The age-dependent increase in Aβ in TgCRND8 mice relative to control age-matched mice was confirmed by enzyme-linked immunosorbent assay including measurement of soluble Aβ species (see Supplemental Figure S6 at http://ajp.amjpathol.org). Furthermore, the increases in CTR and RAMP3 levels in TgCRND8 were specific to brain regions (the cortex, hippocampus, and basal forebrain) that demonstrate a high amyloid burden, compared with other regions (the brainstem and cerebellum) that are relatively spared of amyloid deposition (Figure 8, see also Supplemental Figures S6 and S7 at http://ajp.amjpathol.org). Immunohistochemical localization CTR was observed in close vicinity of Aβ plaque deposition within the cortex, hippocampus, and basal forebrain of TgCRND8 mice (Figure 9).