Faculty of Science and Engineering, Waseda University, Shinjuku, Tokyo, Japan
Consolidated Research Institute for Advanced Science and Medical Care (ASMeW), Waseda University, Shinjuku, Tokyo, Japan
Address correspondence and reprint requests to Naoya Sawamura, Faculty of Science and Engineering, Waseda University, #03C309, TWIns, 2-2 Wakamatsu, Shinjuku, Tokyo 162-8480, Japan. E-mail: firstname.lastname@example.org
Amyloid β protein (Aβ) plays a central role in the pathogenesis of Alzheimer's disease (AD). Point mutations within the Aβ sequence associated with familial AD (FAD) are clustered around the central hydrophobic core of Aβ. Several types of mutations within the Aβ sequence have been identified, and the ‘Arctic’ mutation (E22G) has a purely cognitive phenotype typical of AD. Previous studies have shown that the primary result of the ‘Arctic’ mutation is increased formation of Aβ protofibrils. However, the molecular mechanism underlying this effect remains unknown. Aβ42 binds to a neuronal nicotinic acetylcholine receptor subunit, neuronal acetylcholine receptor subunit alpha-7 (CHRNA7), with high affinity and, thus, may be involved in the pathogenesis of AD. Therefore, to clarify the molecular mechanism of Arctic mutation-mediated FAD, we focused on CHRNA7 as a target molecule of Arctic Aβ. We performed an in vitro binding assay using purified CHRNA7 and synthetic Arctic Aβ40, and demonstrated that Arctic Aβ40 specifically bound to CHRNA7. The aggregation of Arctic Aβ40 was enhanced with the addition of CHRNA7. Furthermore, the function of CHRNA7 was detected by measuring Ca2+ flux and phospho-p44/42 MAPK (ERK1/2) activation. Our results indicated that Arctic Aβ40 aggregation was enhanced by the addition of CHRNA7, which destabilized the function of CHRNA7 via inhibition of Ca2+ responses and activation of ERK1/2. These findings indicate that Arctic Aβ mutation may be involved in the mechanism underlying FAD. This mechanism may involve binding and aggregation, leading to the inhibition of CHRNA7 functions.
Amyloid β protein (Aβ) plays a central role in the pathogenesis of Alzheimer's disease (AD). The ‘Arctic’ mutation within the Aβ sequence has a purely cognitive phenotype typical of AD. Here, we show that Arctic Aβ40 aggregation was enhanced by the addition of neuronal acetylcholine receptor subunit alpha-7 (CHRNA7), which destabilized CHRNA7 functions via inhibition of the Ca2+ response and activation of ERK1/2. These findings may reflect the mechanism underlying familial AD caused by the Arctic Aβ mutation.
The traditional amyloid hypothesis suggests that the cytotoxicity of mature aggregated amyloid fibrils are a toxic form of the protein responsible for disrupting cellular Ca2+ homeostasis, thereby inducing apoptosis (Yankner et al. 1990). Amyloid β proteins (Aβ) are considered to play a significant role in the pathogenesis of Alzheimer's disease (AD) because of their ability to aggregate into β-sheets, which form amyloid plaques (Burdick et al. 1992).
Several autosomal dominant mutations causing early-onset familial AD (FAD) have been identified in the amyloid precursor protein (APP) gene, suggesting that this protein affects either the metabolism of Aβ or its properties of aggregation (Selkoe 1999; Ronnback et al. 2012). Clinical features of FAD are indistinguishable from those of sporadic cases; however, disease onset occurs at a much younger age (Kamino et al. 1992). Point mutations in the Aβ sequence associated with FAD are clustered around the central hydrophobic core of Aβ. The ‘Dutch’ mutation of the Aβ sequence, Aβ(E22Q), gives rise to a highly distinct phenotype of severe amyloid angiopathy, leading to recurrent cerebral hemorrhage (Van Broeckhoven et al. 1990). The ‘Iowa’ mutation, Aβ(D23N), is also associated with severe amyloid angiopathy (Grabowski et al. 2001). The Aβ(E22G) mutation, which causes AD in Swedish families and was first reported in 2001 by Nilsberth et al. (2001), is named Arctic mutation. This mutation has a purely cognitive phenotype, typical of AD, despite the presence of marked amyloid angiopathy (Basun et al. 2008). Carriers of the Arctic mutation have decreased amounts of plasma Aβ(1-42) and Aβ(1-40). Furthermore, Aβ(1-40)E22G forms protofibrils much faster and more abundantly than the wild-type Aβ, although the rate of fibrillization remains the same (Nilsberth et al. 2001; Itkin et al. 2011). Therefore, increased Aβ protofibril formation might be a primary result of Arctic mutation.
Previous studies demonstrated that Aβ42 binds with high affinity to a neuronal nicotinic acetylcholine receptor subunit, neuronal acetylcholine receptor subunit alpha-7 (CHRNA7) (Wang et al. 2000a,b). Interestingly, a decline in senile plaques was detected in over-expressed Aβ transgene mice by blocking CHRNA7 (Dziewczapolski et al. 2009). Taken together, these results have led to the hypothesis that the CHRNA7 subunit plays a role in AD. Moreover, Aβ interacts with CHRNA7 to impair receptor function (Liu et al. 2001; Pettit et al. 2001). CHRNA7 has a high relative permeability to Ca2+, regulating numerous Ca2+-dependent events in the nervous system (Bertrand et al. 1993; Seguela et al. 1993; Liu et al. 2001) and the activation of CHRNA7 can mediate long-term potentiation (LTP) at glutamatergic synapses (Mansvelder and McGehee 2000).
Because the molecular mechanism of Arctic mutation-mediated FAD remains unknown, we investigated Arctic Aβ and CHRNA7. We hypothesized that Arctic Aβ aggregates on CHRNA7, thereby modulating its function and contributing to the pathogenesis of FAD. We observed that Arctic Aβ specifically bound to CHRNA7 and that Arctic Aβ aggregation was enhanced by the addition of CHRNA7. To test the function of CHRNA7, Ca2+ responses to nicotine induction and phospho-p44/42 MAPK (ERK1/2) activation were measured. This study demonstrated that CHRNA7 functions were influenced by binding and aggregation, indicating its involvement in the mechanism of Arctic mutation-mediated FAD.
The following antibodies were used: anti-CHRNA7 polyclonal antibody (Santa Cruz Biotechnology, Dallas, TX, USA); Aβ monoclonal antibody 6E10 (Covance, Berkeley, CA, USA); (ERK1/2) Rabbit monoclonal antibody (Cell Signaling Technology, Beverly, MA, USA); (ERK1/2) (Thr202/Tyr204) Rabbit monoclonal antibody (Cell Signaling Technology); and Anti-HA-tag polyclonal antibody (Medical & Biological laboratories, Nagoya, Japan).
Preparation of Aβ
The synthesized wild-type Aβ proteins (Human 1–40) and the mutant Aβ proteins [Arctic(E22G), Dutch(E22Q), and Iowa (D23N)] were used for in vitro binding assays, and were purchased from Anaspec (Fremont, CA, USA). In further experiments, Arctic Aβ40 was purchased from Bachem (Torrance, CA, USA). All amyloid peptides were dissolved in 0.1% ammonia to a concentration of 1 mM, then diluted to 100 μM in phosphate-buffered saline (PBS).
In vitro binding assay
Glutathione S-transferase (GST)-CHRNA7 was co-incubated with synthetic Aβ at 4°C for 2 h in 1 mL of binding buffer. For immunoprecipitation, anti-CHRNA7 antibodies were coupled to protein G Mag Sepharose beads (5 μL) (GE Healthcare, Waukesha, WI, USA) via a 1-h-incubation at 4°C. The samples were then incubated with magnetic beads for 1 h at 4°C, washed three times with PBS, eluted with sample buffer solution (Wako, Osaka, Japan), and quantified via western blotting.
A standard protocol (Sawamura et al. 2001, 2005) was used but with minor modifications. Proteins were separated using SuperSep gels (Wako) and transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA, USA). Nonspecific binding was then blocked with 5% skim milk (Wako) containing 0.1% Tween 20 (Bio-Rad, Hercules, CA, USA) in PBS. Membranes were then incubated with primary antibodies overnight at 4°C. To detect both monoclonal and polyclonal antibodies, appropriate peroxidase-conjugated secondary antibodies were used in conjunction with Novex ECL (Life Technologies, Grand Island, NY, USA). Images were recorded with the LAS-3000 imager (Fujifilm, Tokyo, Japan), and ImageJ (NIH, Bethesda, MD, USA) was used as automatic image analysis software.
Thioflavin T assay
The degree of Aβ aggregation was determined using the fluorescent dye, Thioflavin T (ThT) (LeVine 1993). An incubated sample (10 μL) was taken every 3 h, treated with 100 μM ThT, and adjusted to 50 mM with glycine NaOH buffer (pH 8.5). Absorbance was measured at excitation and emission wavelengths of 446 nm and 482 nm, respectively. The relative degree of Aβ aggregation was assessed as fluorescence intensity, measured by a RF-5300PC spectrofluorophotometer (Shimadzu, Kyoto, Japan). The intensity of each sample was normalized to buffer-only-sample at each time point. Values were normalized to each 0 h-incubation-sample and then the mean was calculated.
Transmission electron microscopy
Transmission electron microscopy (TEM) was performed according to a previously published method (Ono et al. 2008). Each sample (10 μL) was spotted on to a collodion-coated copper grid (NisshinEM, Tokyo, Japan) and incubated for 20 min. The droplet was then displaced with an equal volume of 2% (v/v) glutaraldehyde in water and incubated for an additional 5 min. This solution was wicked off, and the grid was air-dried. Samples were examined using the transmission electron microscope H-9500 (Hitachi, Tokyo, Japan).
Construction of CHRNA7 expression vectors for transfection
Mammalian Gene Collection Human CHRNA7 sequence-verified cDNA was purchased from Open Biosystems (Huntsville, AL, USA). CHRNA7 cDNAs were amplified from this cDNA, and SalI/NotI fragments containing CHRNA7 were isolated and ligated into the mammalian expression vector, pRK5-HA, then digested with the same endonucleases. The resulting recombinant plasmid, pRK5-HA-CHRNA7 was used for further studies. The entire nucleotide sequence was confirmed by DNA sequencing. A vector containing the CHRNA7 cDNA was transfected into CHO-K1 cells using Polyfect Transfection reagent (Qiagen, Hilden, Germany) according to the manufacturer's instructions. CHO-K1 cells were maintained in Dulbecco's modified Eagle's medium (Wako) supplemented with 10% fetal bovine serum at 37°C in 5% CO2.
Measurement of Ca2+ flux
Ca2+ flux in CHO-K1 cells was measured using Fluo-4 acetoxymethyl ester (Fluo-4/AM; Dojindo, Kumamoto, Japan). Ca2+ flux was induced by 400 nM Aβ or 100 μM nicotine after Aβ incubation. Fluorescence excitation and emission was read at 455 nm and 525 nm, respectively. A change of fluorescence intensity was determined by Powerscan HT and analyzed by GENE 5 software (Biotek, Tokyo, Japan). All measurements of Ca2+ flux experiments were repeated five times and the maximum Ca2+ response peak amplitude at a single time point was plotted on the bar graph.
Data in all experiments were expressed as means ± SD from three to five independent experiments. Data were analyzed using the Student's t-test, where differences between samples was considered statistically significant when p < 0.05.
Arctic Aβ specifically binds to CHRNA7
We first performed an in vitro binding experiment using several types of synthetic Aβ with GST-CHRNA7. We tested the interaction of Dutch, Iowa, and Arctic Aβ40 with CHRNA7. These mutations have the potential to affect all factors (i.e., the production, degradation, and aggregation) that regulate Aβ monomer levels. Only Arctic Aβ40 specifically bound to CHRNA7 (Fig. 1a). Furthermore, quantification of Aβ immunoreactivity by western blotting showed Arctic Aβ40 bound to GST-CHRNA7 with a significantly higher affinity than wild-type Aβ40 (**p < 0.01; Fig. 1a). Because Aβ42 might have a stronger association with AD, we also tested the interaction of wild-type Aβ42 and Arctic Aβ42 with CHRNA7. Although Arctic Aβ42 bound to GST-CHRNA7 more strongly than wild-type Aβ42, a band corresponding to the Aβ42 dimer (~8 kDa) was also detected (Figure S1). It was difficult to clarify whether this interaction between aggregated Arctic Aβ42 and GST-CHRNA7 occurred before or after Arctic Aβ42 accumulated, because aggregates were also visible in the input sample (Figure S1, arrowhead). Physiological phenomena of cells might be affected by the self-accumulation of Arctic Aβ, either bypassing or including CHRNA7. Therefore, to exclude the effect of self-aggregation of Arctic Aβ on cells and to focus on the effect of the mutant Aβ on CHRNA7, we utilized Arctic Aβ 1-40 peptides, which did not show self-aggregation in our experimental conditions (Fig. 2). An in vitro binding experiment using Arctic Aβ40 with GST-CHRNA7 or GST protein showed that Arctic Aβ bound specifically to CHRNA7 (Fig. 1b). This result led us to speculate on a connection between Arctic Aβ and CHRNA7, and that CHRNA7 might play a role in FAD caused by the Arctic mutation.
Enhanced Arctic Aβ aggregation by CHRNA7
The Arctic mutation was reported to induce the rapid formation of protofibrils, thus leading to an alternative pathogenic mechanism of FAD (Nilsberth et al. 2001). Fibril formation by the Arctic mutation was observed both in vitro (Murakami et al. 2003) and in vivo (Cheng et al. 2007) conditions. In the present study, we investigated the effect of CHRNA7 on the aggregation form of Arctic Aβ via the ThT assay and TEM. By ThT, the intensity of each sample was normalized to buffer-only-sample at each time point. These values were normalized to each 0 h-incubation-sample and then the mean was calculated. The ThT assay showed that Arctic Aβ started to aggregate at 6 h when co-incubated with CHRNA7, compared with Arctic Aβ alone, which did not aggregate at this time point (Fig. 2a). Furthermore, aggregation of Arctic Aβ was enhanced at 9 h when co-incubated with CHRNA7, compared with Arctic Aβ alone (Fig. 2a). Aggregation was not observed for Aβ40 when co-incubated with CHRNA7 (Fig. 2a). This phenomenon was also seen via TEM. The accumulation of Arctic Aβ was clearly observed when co-incubated with CHRNA7 for 24 h compared with Arctic Aβ alone (Fig. 2b). Therefore, the aggregation of Arctic Aβ was enhanced by the addition of CHRNA7 (Fig. 2).
Synthetic arctic Aβ does not modify CHRNA7 functions
CHRNA7 has an extremely high relative permeability to Ca2+, and regulates numerous Ca2+-dependent events (Bertrand et al. 1993; Seguela et al. 1993; Liu et al. 2001). Therefore, we investigated the Ca2+ response and Ca2+-downstream signaling activation of ERK1/2 to evaluate whether Arctic Aβ affects the function of CHRNA7. The overexpression of CHRNA7 in CHO-K1 cells was confirmed by western immunoblot analysis. CHRNA7 bands were detected by anti-HA-tag and anti-CHRNA7 antibodies (Fig. 3a). To determine whether Arctic Aβ induces Ca2+ flux, synthetic Aβ was used. CHO-K1 cells produced a rapid and sharp increase in Ca2+ in response to Aβ42, confirming a previous report (Dineley et al. 2002). This increase was not seen when Arctic Aβ was added (Fig. 3b). Although Arctic Aβ40 bound to CHRNA7 similar to Aβ42, the Ca2+ flux was not the same (Fig. 3b). Aβ activates the MAP kinase cascade via CHRNA7 (Dineley et al. 2001) and therefore, we examined the activation of ERK1/2 downstream of CHRNA7 as a further test of a Ca2+-dependent event. The addition of synthetic Arctic mutant Aβ did not activate ERK1/2 (Fig. 3c).
Incubation with Arctic Aβ inhibits CHRNA7 functions
Next, we examined the effect of co-incubation with Aβ on the function of CHRNA7. Nicotine activated the Ca2+ response after 24-h-incubation. Ca2+ responses were not induced by nicotine in mock-transfected CHO-K1 cells. Ca2+ was rapidly increased by nicotine in CHO-K1 cells overexpressing CHRNA7. However, this response was significantly (p <0.05) reduced in cells incubated with Arctic Aβ40. Thus, Arctic Aβ tended to attenuate Ca2+ responses in addition to enhancing aggregation when co-incubated with CHRNA7 (Fig. 4b). Furthermore, Ca2+-downstream activation of ERK1/2 was investigated as a further test of a Ca2+-dependent event. Addition of the Arctic mutant Aβ inhibited the activation of ERK1/2 (Fig. 4b). Therefore, these results indicated that Arctic Aβ was aggregated in the presence of CHRNA7, and suppressed the function of CHRNA7 via the inhibition of Ca2+ flux and activation of ERK1/2.
This study has shown for the first time that Arctic Aβ specifically binds to CHRNA7 with a high affinity, unlike other typical early-onset FAD mutant forms of Aβ. Therefore, this receptor may be a target membrane receptor for Arctic Aβ. Aggregation of Arctic Aβ was enhanced when co-incubated with CHRNA7, suggesting this receptor was required for further aggregation of the mutant form of Aβ. Furthermore, Arctic Aβ aggregated on CHRNA7 and inhibited its function by reducing Ca2+ responses and nicotine-induced activation of ERK1/2.
Arctic mutation-mediated FAD has been studied since the discovery of this association by Nilsberth et al. (2001). Most previous studies focused on the special aggregation pattern of Arctic Aβ, which preferentially forms protofibril assemblies rather than fibrils. Some of these studies were performed under different conditions (Johansson et al. 2006) or by structural analysis (Norlin et al. 2012). These studies suggested that Arctic Aβ protofibrils are quickly generated (Paivio et al. 2004; Johansson et al. 2006), similar to the conclusion formed by Nilsberth et al. (Nilsberth et al. 2001). Although the Arctic mutation was reported to enhance the formation of protofibrils (Nilsberth et al. 2001), it was not confirmed. Different species of Arctic mutant Aβ aggregates (i.e., oligomers, fibrils) were investigated, and the oligomerization pattern of the Arctic mutation was shown to be different from wild-type Aβ with a tendency to form larger oligomers (Gessel et al. 2012). Fibril structures were also studied (Norlin et al. 2012). However, these studies did not report on the molecular signaling pathway affected by membrane receptors. CHRNA7 binds to wild-type Aβ42 (Wang et al. 2000a,b) and does not interact with wild-type Aβ40. The present study showed that Arctic Aβ40 binds to CHRNA7 with a high affinity, indicating CHRNA7 has a critical role in Arctic mutation-mediated FAD by interacting with Arctic Aβ. This suggests the mutated structure of monomer Arctic Aβ40 has an increased affinity for CHRNA7. Therefore, we suggest the inhibition of CHRNA7 functions were caused by a combination of mutant effects and the higher affinity to CHRNA7. This finding led us to clarify mechanisms involved in Arctic FAD apart from its aggregation pattern.
The aggregation pattern of Arctic Aβ40 is different from wild-type Aβ (Gessel et al. 2012) because Arctic Aβ40 can generate larger protofibrils faster during incubation (Paivio et al. 2004; Norlin et al. 2012). In this study, the self-aggregation of Arctic Aβ40 was not observed using our experimental conditions (Fig. 2). Arctic Aβ40 showed enhanced aggregation when co-incubated with CHRNA7. Additionally, the environment is crucial for Arctic Aβ aggregation (Yamamoto et al. 2004), which can be induced in the presence of lipids (Sureshbabu et al. 2010; Pifer et al. 2011). Mixing Arctic Aβ with wild-type Aβ also enhances Arctic aggregation (Lashuel et al. 2003). Therefore, environmental factors might support Arctic Aβ aggregation. Several studies demonstrated that CHRNA7 interacts with Aβ (Wang et al. 2009), and furthermore, exacerbates the pathological features in AD mouse models (Soderman et al. 2008; Dziewczapolski et al. 2009). Therefore, our study suggests that CHRNA7 might enhance the aggregation of Arctic Aβ.
In the present study, although the direct addition of Arctic Aβ40 did not significantly affect the influx of Ca2+, a diminished Ca2+ response was observed after 24-h-incubation. Thus, the inhibition of CHRNA7 function might be due to the aggregation of Arctic Aβ40 when both were co-incubated for 24 h, resulting in a seed role for CHRNA7. Arctic Aβ40 was reported to inhibit hippocampal LTP in vivo (Klyubin et al. 2004). Moreover, the activation of CHRNA7 at glutamatergic synapses promotes LTP (Mansvelder and McGehee 2000). Therefore, CHRNA7 might function as a critical intermediate, and its loss of function might be a primary cause of LTP inhibition by Arctic Aβ.
Our results suggested that enhanced Arctic Aβ aggregation occurred with CHRNA7 as the seed, and inhibited receptor functions. This effect might be associated with LTP inhibition and neurotoxicity. The observation of cognitive deficits in an Arctic APP transgenic mouse model (Ronnback et al. 2011) suggested that memory and cognitive functions are affected by Arctic mutation. The levels of Aβ protofibrils correlate with learning performance in Arctic APP mice (Lord et al. 2009), suggesting a link between aggregation and memory. Therefore, we hypothesize that CHRNA7-mediated changes of Arctic Aβ aggregation may be crucial to our understanding of the molecular mechanisms underlying Arctic FAD. Furthermore, we propose that CHRNA7 may be a critical therapeutic target for treatment of FAD caused by Arctic mutation.
Acknowledgments and conflict of interest disclosure
This work was supported by KAKENHI 21700415. This study was also supported by the High-Tech Research Center (TWIns), the Consolidated Research Institute of Advanced Science and Medical Care (ASMeW), the Global COE ‘Practical Chemical Wisdom’ projects, the Leading Graduate Program in Science and Engineering, Waseda University, and COI STREAM (Center of Innovation Science and Technology based Radical Innovation and Entrepreneurship Program), from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan.
All experiments were conducted in compliance with the ARRIVE guidelines. The authors have no conflict of interest to declare.