Abstract
Recently, aging societies have been showing an increasingly strong relationship between Alzheimer’s disease (AD) and chronic cerebral hypoperfusion (HP). In the present study, we created a new mouse model for AD with HP, and investigated its clinical and pathological characteristics. Alzheimer’s disease transgenic mice (APP23) were subjected to bilateral common carotid arteries stenosis with ameroid constrictors for slowly progressive cerebral HP. In contrast to simple APP23 mice, cerebral HP exacerbated motor and cognitive dysfunctions with white matter lesions and meningo-parenchymal amyloid-β (Aβ) burdens. Strong cerebrovascular inflammation and severe amyloid angiopathy with cerebrovascular remodeling were also observed in APP23 + HP mouse brains. An acetylcholinesterase inhibitor galantamine improved such clinical dysfunctions, retrieved above neuropathological characteristics, and enhanced nicotinic acetylcholine receptor (nAChR)-binding activity. The present study demonstrates that chronic cerebral HP enhanced cognitive/motor dysfunctions with parenchymal/cerebrovascular Aβ accumulation and cerebrovascular remodeling. These neuropathological abnormalities were greatly ameliorated by galantamine treatment associated with nAChR-mediated neuroprotection by allosterically potentiating ligand action.
INTRODUCTION
With the growth of old populations, the number of Alzheimer’s disease (AD) patients continuously increases every year around the world [1]. Our recent data showed that AD occurs in 69% of the elderly population more than 75 years old with dementia [2], approximately 90% of whom have cerebral amyloid angiopathy (CAA) [1, 3] and white matter lesions (WML) [4].
Chronic cerebral hypoperfusion (HP) is frequently observed in such aged AD patients [5–7], and brain HP probably triggers and reinforces the degenerative pathology which is present in AD (vascular hypothesis) [8–10], while others have claimed that the deposition of fibrillary and insoluble amyloid-β (Aβ) cause neurodegeneration in AD even without cerebrovascular disease (amyloid hypothesis) [10–13]. However, the relationship between cerebral HP and AD is still controversial.
In our previous study, a short period of ischemic stroke caused progressive AD pathology in aged-hypertensive rats [14]. An AD mouse model that expresses human vasculotropic Swedish mutant amyloid precursor protein (APP23 mice) [15] presents both parenchymal senile plaque (SP) and CAA. To mimic chronic cerebral HP, the bilateral common carotid artery (CCA) stenosis (BCAS) with microcoils is frequently used [16]. However, this animal model possesses a limitation that cerebral blood flow (CBF) drops too rapidly to replicate “chronic” cerebral hypoperfusion. On the other hand, bilateral common carotid arteries (BCCA) stenosis (BCCS) with ameroid constrictors may induce such a “chronic” cerebral HP.
Galantamine has been used for not only AD but also vascular dementia (VaD) [17, 18]. Our previous and others’ clinical studies reported a strong protective effect of galantamine on cognitive function in patients with either AD or VaD complicated by cerebrovascular pathology [19–22], suggesting that galantamine may have both neuroprotective and vasculoprotective benefits related to its dual action as an acetylcholinesterase inhibitor and allosteric modulation of nicotinic acetylcholine receptor (nAChR) [19, 22–25].
In the present study, therefore, we created a new AD mouse model with chronic HP, and investigated motor and cognitive dysfunctions in relation to WML, cortical SP/CAA aggregation, vascular remodeling, nAChR-binding activity, and activation of Iba-1-positive microglial cells as well as the efficacy of galantamine.
MATERIALS AND METHODS
Experimental model and drug treatment
All animal experiments were performed in compliance with a protocol approved by the Animal Committee of the Graduate School of Medicine and Dentistry, Okayama University (Neurovascular protection effect of galantamine in APP23 mice; OKU-2012325) and conducted in accordance with ARRIVE guidelines (http://www.nc3rs.org/ARRIVE) and the Okayama University guidelines on the Care and Use of the Laboratory Animals. As a preliminary in vitro study, ameriod constrictors with an inner diameter (D) of 0.75 mm and an open gap (G) of 0.5 mm (26 mg weight, Research Instruments NW, Lebanon, OR, USA) were immersed in 0.9% sodium chloride (NaCl) or in 0.9% NaCl + 6 g/dL albumin (Alb), both at 37°C. The pictures of the constrictors were captured at 3 h, 1, 3, 7, 14, 28, and 60 days (d) to mimic an in vivo study (Fig. 1A). A preliminary in vivo study was simultaneously performed, and the in vivo constrictor images at 7, 14, 28, and 60 d were also recorded (Fig. 1A).
For the in vivo study, four groups of male mice were used in this study: wild type mice (C57BL/6J, only sham surgery, n = 10), APP23 group (APP23 + sham surgery, n = 17), hypoperfusion group (APP23 + HP, n = 12), and galantamine-treated group (APP23 + HP + Gal, n = 10). As previously reported, APP23 mice, as a classical AD animal model, develop both SP and vascular amyloid deposits after the age of 6 months (M) [15, 26]. Until now, APP23 mice have been used in many studies on CAA [26–29]. In order to conduct cerebral HP surgery, a neck incision was made and ameroid constrictors were applied to BCCA (a total of 52 mg of constrictors) of APP23 mice 4 M old (body weight 20–25 g) (Fig. 1B). Additionally, to test whether ameroid constrictors would narrow the trachea and lead to alterations in arterial blood gases, we analyzed pH, pO2 and pCO2 of arterial blood samples. However, there was no significant difference in these three parameters between the WT group and the WT + HP group (Supplementary Figure 1).
CBF was measured with a laser-Doppler flowmeter (FLO-C1, Omegawave, Tokyo, Japan) before and 1, 3, 7, 14 and 28 d after surgery. A laser-Doppler flowmetry probe was fixed perpendicular to the skull 1 mm posterior and 2.5 mm lateral to the bregma where CBF values were measured five times. The maximum CBF value was recorded. At 15 d after surgery when cervical incision had completely healed, APP23 + HP + Gal group of mice began to receive galantamine (5 mg/kg; 1 mg/ml in ultrapure water, 0.15 ml; Takeda Pharmaceutical Company Limited, Osaka, Japan) once daily by oral gavage until sacrifice.
Behavioral analysis
The Rotarod test evaluated motor coordination and balance by measuring latency seconds (s) until 2, 5, 8, and 11 M-old mice fell off from a rotating rod (MK 670; Muromachi Kikai Co., Tokyo, Japan) according to our previous report [30].
An 8-arm radial maze test evaluated behavioral memory (mainly for working memory) described according to our and other’s reports [31, 32]. For each trial, a mouse was allowed to make arm choices until either all four pellets were eaten or 5 min had elapsed, whichever occurred first. Re-entry into the baited arms previously visited was scored as a working memory error. The radial maze task was performed separately when mice were 3, 6, 9, and 12 M old.
A pilot study of the Morris water maze test resulted in the death of many mice by drowning caused by the heavy ameroid constrictors (52 mg for BCCA), which did not allow reference memory to be analyzed in the present study. There was also the possibility that the rotarod results were influenced by the weight of ameroid constrictors. However, the significant improvement of motor performance was found by galantamine treatment in the present study.
Tissue preparation and histochemistry
When 6 and 12 M old, APP23, APP23 + HP, and APP23 + HP + Gal mice were deeply anesthetized by intraperitoneal injection of pentobarbital (40 mg/kg) and transcardially perfused with 20 ml of ice-cold phosphate-buffered saline (PBS) and then 20 ml of ice-cold 4% paraformaldehyde (PFA) in 0.1 mol/L phosphate buffer. The brains were removed and post-fixed in 4% PFA overnight. Floating coronal sections (50 μm thickness) were produced with a vibrating blade microtome (LEICA VT1000S; Leica, Germany). To evaluate white matter (WM), mice brain sections were stained with luxol fast blue (LFB). BF-168 (Wako Pure Chemicals, Osaka, Japan) served as a probe for SP [33]. For immunohistochemistry, brain sections were immersed in 0.6% periodic acid to block intrinsic peroxidase, and were treated with 5% bovine serum in 50 mM PBS, pH 7.4, containing 0.1% triton to block any non-specific antibody responses, and were incubated at 4°C overnight with primary antibody. We used the following primary antibodies: anti-Aβ40 antibody (BA27; Wako), anti-αSMA monoclonal antibody conjugated to Cy3 (1 : 100; Sigma-Aldrich), Iba-1 (1 : 1000; Wako). Immunoreactions were visualized using fluorescent secondary antibody or horseradish peroxidase-conjugated antibody with the diaminobenzidine reaction. Additionally, some researchers previously reported that the fluorescent spots of BF-168 corresponded well to Aβ immunostaining in the APP23 transgenic mouse brain (13 M of age), which can bind both Aβ40-positive and Aβ42-positive plaques [33]. To further test the specificity of BF-168 in our mouse model, we performed double staining of Aβ40/BF-168 and Aβ42/BF-168. In agreement with the previous report, BF-168 probe did recognize not only Aβ40 but also Aβ42 (Supplementary Figure 2).
Autoradiography analysis
For autoradiography, the brains of 12 M old mice were dissected and freshly frozen in powdered dry ice. Coronal brain sections (20 μm thickness) were then cut on a cryostat at – 20°C. To evaluate nAChRs, brain sections were incubated for 90 min with 75 pM [125I]epibatidine (Amersham, UK) in 50 mM Tris-HCl (pH 7.4) as described in our previous reports [34–36]. To evaluate mAChRs, brain sections were incubated for 180 min in an assay buffer (50 mM Tris-HCl, pH 7.4, 120 mM NaCl, 5 mM KCl, 2.5 mM CaCl, 1 mM MgCl2) together with 1.0 nM [3H]QNB (quinuclidinyl benzilate) (Amersham, UK) as described in our previous reports [34, 36]. The [125I]epibatidine-binding and [3H]QNB-binding sections were exposed for 8 h or 7 d, respectively to imaging plates (Fuji Film, Tokyo, Japan).
Quantitative analysis
All images for immunostaining analysis were taken with a digital microscope camera (BX-51; Olympus Optical Co., Japan), and were analyzed using image processing software (ImageJ; National Institutes of Health, Bethesda, USA). For each measurement, we analyzed separated sections from three levels of the caudate putamen (1.0, 0.5, and 0 mm rostral to the bregma) per brain and three or four randomly selected regions per section (i.e., n = 9–12 measurements per mouse) [29]. For LFB staining, the average pixel intensity of signal in the corpus callosum region, as well as the thickness and area of the corpus callosum were measured. For BF-168-positive SP analysis, data were reported as the percentage area occupied by the BF-168-positive signal in the cortex area [37, 38]. To analyze vascular Aβ (VAβ) accumulation, leptomeningeal or parenchymal VAβ40-positive signal was recognized by contrasting with the αSMA-positive vascular smooth muscle signal. Data were reported as the percentage area occupied by the VAβ40-positive signal in the cortex area [37, 38]. To analyze cerebrovascular remodeling, the thickness and inner diameter of parenchymal αSMA-positive vessels were evaluated. Each vessel was categorized according to arterial changes and the severity of Aβ deposition [26, 39]. Type A was classified as visible Aβ deposits within the vessel wall. Type B showed remarkable deposits infiltrating the whole vessel wall. Type C was type B plus amyloid deposits with a thick and complete amyloid coat extending into the brain parenchyma. CAA was histologically assessed by calculating the percentage of total vessel numbers occupied by each type in the cortex and thalamus. To quantify Iba-1-positive microglial cell activation, the number of Iba-1-positive microglial cells was analyzed in 6 fields (microglial cells/mm2) including the cerebral cortex, corpus callosum, thalamus, hypothalamus, leptomeningeal vessel, and parenchymal vessel areas. Similarly, the pixel intensity of Iba-1 immunostaining was measured in the same fields. In the autoradiography, binding was quantified at striatum and dorsal hippocampal levels with ImageJ. The intensity of the binding signal in five fields (cerebral cortex, striatum, thalamus, hypothalamus and hippocampus) was calculated.
Statistical analysis
All data were expressed as mean ± SD. Statistical analyses were performed using one way analysis of variance followed by a Tukey–Kramer postcomparison test. Differences with a probability value of p < 0.05 were considered to be statisticallysignificant.
RESULTS
Changes in CBF with ameroid constrictors
The external diameter of CCA was 0.60 ± 0.05 mm. The gap (G) of the ameroid constrictor shrank progressively and disappeared at 1 d in vitro (Fig. 1A), and the internal diameter (D) shrank in vitro from 1 day (d) reaching 0.46 ± 0.02 mm at 7 d, 0.36 ± 0.04 mm at 28 d, and 0.32 ± 0.05 mm at 60 d. Similarly, G in vivo shrank to 0.46 ± 0.06 mm at 7 d, 0.38 ± 0.05 mm at 28 d and 0.37 ± 0.07 mm at 60 d (Fig. 1A), leading to a slowly progressive stenosis of BCCA (BCCS) (Fig. 1B). The mean CBF maintained its baseline level (93.1% to 108.6%) after sham surgery until 28 d in APP23 group (Fig. 1C, triangles). However, CBF gradually decreased in both APP23 + HP and APP23 + HP + Gal groups from 1 d after surgery (Fig. 1C), and reached a minimum decline at 7 d to 66.3 ± 19.5% in APP23 + HP group (Fig. 1C, dotted squares, *p < 0.05, **p < 0.01 versus APP23) and 61.2 ± 22.3% in APP23 + HP + Gal group (Fig. 1C, filled squares, *p < 0.05, **p < 0.01 versus APP23). CBF remained at a similar level in the two APP23 + HP groups until 28 d.
Motor and memory deficits after chronic cerebral hypoperfusion
Motor performance and 8-arm radial maze tests showed no difference between wild type and APP23 group at 2 and 3 M before BCCS surgery (Fig. 2A, E). When using the accelerating rotarod apparatus, APP23 + HP group showed a significantly inferior performance than the wild type and APP23 group at 5 and 8 M (§p < 0.05, §§p < 0.01 versus wild type, *p < 0.05 versus APP23; Fig. 2B and C). In contrast, APP23 + HP + Gal group preserved rotarod scores at 5 M (#p < 0.05 versus APP23 + HP;Fig. 2B).
In the working memory task with the 8-arm radial maze, there were no differences in revisiting errors among the four mice groups at 6 M (Fig. 2F). However, APP23 + HP group showed significantly more errors than the wild type and APP23 group at 9 and 12 M (§p < 0.05 versus wild type, *p < 0.05, **p < 0.01 versus APP23; Fig. 2G, H). Galantamine treatment rescued this disturbance in memory (#p < 0.05 versus APP23 + HP; Fig. 2G).
White matter lesion and senile plaque accumulation
Compared with APP23 group (6 M, 1.00 ± 0.06; 12 M, 0.91 ± 0.07), LFB staining showed reduced myelin intensity in the corpus callosum regions of APP23 + HP group at 6 and 12 M (6 M, 0.82 ± 0.04; 12 M, 0.74 ± 0.10; **p < 0.01 versus APP23; Fig. 3A, B). However, galantamine treatment significantly recovered it (6 M, 0.98 ± 0.05; 12 M, 0.90 ± 0.05; #p < 0.05, ##p < 0.01 versus APP23 + HP; Fig. 3A, B). The thickness and area also appear to have decreased in APP23 + HP group at 6 M, but the change was not significant (Fig. 3C, D).
Few BF-168-positive SP were detected in the cerebral cortex (Fig. 4A), but the difference among the three APP23 mice groups at 6 M was not significant (Fig. 4B). However, the areas of these SPs increased considerably in APP23 + HP group (APP23 group, 0.20 ± 0.18%; APP23 + HP group, 0.66 ± 0.60%, *p < 0.05, versus APP23; Fig. 4B), and galantamine treatment significantly reduced SP areas at 12 M (0.04 ± 0.34%, #p < 0.05 versus APP23 + HP; Fig. 4A, B).
Galantamine attenuated vascular Aβ deposit and cerebrovascular remodeling
Aβ accumulation in the leptomeningeal (arrowheads) and parenchymal (arrows) small vessels were increased significantly in APP23 + HP group compared with APP23 group (*p < 0.05, **p < 0.01 versus APP23; Fig. 5A top middle and Fig. 5B). Galantamine treatment attenuated leptomeningeal VAβ accumulation at 12 M (#p < 0.05 versus APP23 + HP; Fig. 5A top right and Fig. 5B).
Types A, B, and C of cerebrovascular remodeling were classified according to arterial changes and the severity of cerebrovascular amyloid deposits (Fig. 5G). APP23 + HP group showed a significant increase in thickness and inner diameter of parenchymal small vessels with amyloid deposits than APP23 group in the cerebral cortex and thalamus (*p < 0.05, **p < 0.01 versus APP23; Fig. 5C–F). Galantamine treatment attenuated such cerebrovascular remodeling at 6 and 12 M especially in the thalamus (#p < 0.05, ##p < 0.01 versus APP23 + HP; Fig. 5A bottom panels, Fig. 5E, F). Type C with strong perivascular Aβ deposits was frequently found in the thalamus of APP23 + HP mice (**p < 0.01 versus APP23; Fig. 5H), but was greatly attenuated at 12 M by galantamine treatment (##p < 0.01 versus APP23 + HP; Fig. 5H).
Nicotinic and muscarinic acetylcholine receptor binding sites
An autoradiographic study showed that [125I] epibatidine-positive nAChR was mainly detected in the thalamus and hypothalamus, but also detected in the cerebral cortex and striatum of APP23 mice (Fig. 6A). Densitometric analyses indicated a strong reduction of [125I]epibatidine binding in the striatum of APP23 + HP mice compared with APP23 mice (Fig. 6A, B). In contrast, APP23 + HP + Gal group showed a significant recovery or even an increase of [125I]epibatidine binding in the cerebral cortex, striatum, thalamus and hypothalamus (**p < 0.01 versus APP23; #p < 0.05, ##p < 0.01 versus APP23 + HP; Fig. 6B).
[3H]QNB-positive muscarinic acetylcholine receptor (mAChR) binding sites were mainly detected in the cerebral cortex, striatum and hippocampus, and a small level in thalamus (Fig. 6). However, there were no differences in the above brain regions among the three APP23 groups (Fig. 6D).
Microglial activation after chronic cerebral hypoperfusion
Expression of Iba-1-positive microglial cells was clearly observed in various brain regions, including the cerebral cortex, corpus callosum, thalamus, hypothalamus, and leptomenigeal/parenchymal small vessels at 12 M (Fig. 7A). Quantitative analysis indicated no difference among the three APP23 mice groups in the indicated areas until 6 M, but APP23 + HP group showed a remarkable increase of Iba-1-positive microglial cell number in the corpus callosum, thalamus and hypothalamus, with a prominent increase at and around the parenchymal small vessels at 12 M (*p < 0.05, **p < 0.01 versus APP23; Fig. 7A, B). The intensity of Iba-1 immunostaining showed the similar tendency (*p < 0.05, **p < 0.01 versus APP23; Fig. 7A, C). Galantamine treatment greatly reduced this strong activation of microglia in the brain parenchyma, perivascular parenchyma, and small vessel walls (#p < 0.05, ##p < 0.01 versus APP23 + HP; Fig. 7A–C).
DISCUSSION
In the present study, we established a novel animal model of AD plus chronic cerebral HP in APP23 mice by slowly progressive BCCA stenosis (BCCS) with ameroid constrictors (Fig. 1). Our present results showed that HP exacerbated motor and cognitive dysfunctions (Fig. 2) in this mouse AD model with a significant decrease in myelin intensity (Fig. 3), cortical SP accumulation and CAA aggregation (Figs. 4 and 5). We also found that HP resulted in a decline of binding activity by nAChRs and significant activation of Iba-1-positive microglial cells with remarkable cerebral microvascular remodeling (Figs. 5–7). On the other hand, galantamine, an acetylcholinesterase inhibitor with allosterically potentiating ligand (APL) action, improved such motor and cognitive dysfunctions (Fig. 2, filled squares) and retrieved the above neuropathologic characteristics (Figs. 3–7).
A slowly progressive stenosis of BCCA (BCCS) with ameroid constrictors successfully induced mild cerebral HP and cumulative WML (Fig. 3). In the case of a previous mouse model of HP by microcoils, CBF suddenly dropped just after surgery, followed by a significant recovery [16, 40]. On the other hand, our mouse model showed a gradual reduction of CBF in 1–3 d with a constant mild reduction 7 d after inserting ameroid constrictors (Fig. 1C), suggesting that this model could mimic more closely cerebral HP of aged AD patients with WML. A limitation of this study is the lack of information of long-term CBF measurement in our cohort of the APP23 mice. In this study, since repetitive CBF measurement led to serious skin and skull injury, CBF was measured until 28 d post-HP surgery when the CBF value stabilized (60% to 70% of pre-HP surgery). In this study, at both 6 and 12 months of age, we found progressive WML and enhanced microglial activation in the corpus callosum of APP23 + HP mice, speculating the sustained oligemia in the chronic phase. In the near future, we will do autoradiographic CBF measurements in the chronic phase of cerebral HP.
In the current study, we used APP23 mice that overexpressed vasculotropic mutant Aβ, and HP strongly enhanced both parenchymal SP accumulation and vascular amyloid deposits (Fig. 5A) probably due to both an increase in the synthesis of Aβ [41] and a reduced driving force for perivascular clearance of Aβ [42]. These microvascular dysfunctions could impede the perivascular drainage pathway of Aβ, causing a new type of vascular remodeling with strong VAβ deposition (Fig. 5G). In particular, thalamic vessels revealed a more severe CAA pathology and remodeling compared with cerebrocortical vessels (Fig. 5A), probably due to a more severe hypoperfusion in the thalamus than in the cerebral cortex in this mouse model [43].
Our present results also strongly suggested pathological vascular remodeling in this combined AD + HP model (Fig. 5). In the case of simple CAA, Aβ deposits were primarily present around the tunica media, leading to thickening of the tunica media and an accumulation of collagen fibers in the media [27], and finally resulting in muscle cell degeneration and dilation of the lumen with Aβ deposits outside the vessels [44–46]. On the other hand, in the case of simple hypertensive vascular remodeling, the intimal tunica is primarily affected, followed by smooth muscle cell proliferation and migration into the intima, leading to thickening of the media tunica and narrowing or dilation of the lumen [47]. In the case of our APP23 + HP mice, a significant number of considerably enlarged vessels were found especially in the thalamus, with a heavy Aβ burden in/around the microvessels up to 330 μm in diameter (Fig. 7A). Our data at 12 M was similar to a previous sample in APP23 mice at 27 M [26], indicating that the present HP dramatically accelerated the essential pathology of this CAA mouse model. Such hypoperfusion-induced abnormalities of APP23 mice seem to be similar to those of AD patients [27, 48]. However, in AD, the cerebrovasculature which is affected first is the small vessel. In this animal model, on the other hand, the carotid artery was manipulated. This disparity would limit application of the results presented here to the pathophysiology of human AD.
Galantamine is a rather weak acetylcholinesterase inhibitor but also acts as APL for nAChRs [49, 50]. As previously reported, acetylcholine, when released from perivascular parasympathetic nerves and diffusing through the smooth muscle cell layer, may reach the endothelial cells and stimulate the synthesis of NO causing a rapid and marked increase in cortical CBF [51, 52]. The CBF of white matter is lower than that of gray matter but changes in parallel [53]. We observed no improvement in cortical CBF (Fig. 1C) in the galantamine-treated group. On the other hand, we found that galantamine treatment rescued the decline in nAChRs-binding in the present APP23 + HP mouse model, especially in the thalamus (Fig. 6A) but did not improve mAChR-binding (Fig. 6B), suggesting that galantamine worked as an APL for nAChR. Iba-1 immunostaining results showed that galantamine significantly suppressed microglial activation in various kinds of brain regions, including the corpus callosum (Fig. 7). Previously, it has been demonstrated that galantamine can suppress inflammatory cytokine via activation of nAChRs [54, 55]. Therefore, we propose that galantamine partially rescued WML probably by alleviating inflammation (Fig. 3).
Previous studies have demonstrated nAChR-mediated neuroprotection of galantamine by reducing Aβ, but did not show data of direct binding [56, 57]. Although mAChR-binding was decreased under oxidative stress [34, 58], stimulation of nAChR by galantamine could show neuroprotective and vasculoprotective benefits (Figs. 5–7). Previous papers have demonstrated effect of galantamine on cognition was antagonized by nAChR antagonist rather than mAChR antagonist, suggesting galantamine shows effects mainly by allosterically modulating the function of nAChR [59, 60]. Therefore, direct stimulation of nAChR through the action of APL is suggested to potentiate cerebral cortical function with thalamic activation [19, 61], which could be related to the long-term cognitive benefits of galantamine in this mouse AD model.
In summary, the present study demonstrated that cerebral HP enhanced cognitive/motor dysfunctions with parenchymal/cerebrovascular Aβ accumulation and cerebrovascular remodeling. Such neuropathologic abnormalities were greatly ameliorated by galantamine treatment associated with nAChR-mediated neuroprotection by APL action.
Footnotes
ACKNOWLEDGMENTS
This work was partly supported by Grant-in-Aid for Scientific Research (B) 25293202, (C) 15K09316 and Challenging Research 15K15527 and Young Research 15K21181, and by Grants-in-Aid from the Research Committees (Mizusawa H, Nakashima K, Nishizawa M, Sasaki H, and Aoki M) from the Ministry of Health, Labour and Welfare of Japan.
