Dickkopf 3 (Dkk3) Improves Amyloid-β Pathology,Cognitive Dysfunction,and Cerebral Glucose Metabolism in a Transgenic Mouse Model of Alzheimer’s Disease

Abstract

Dysfunctional Wnt signaling is associated with Alzheimer’s disease (AD), and activation of the Wnt signaling pathway inhibits AD development. Dickkopf 3 (Dkk3) is a modulator of the Wnt signaling pathway and is physiologically expressed in the brain. The role of Dkk3 in the pathogenesis of AD has not been evaluated. In the present study, we determined that Dkk3 expression was significantly decreased in brain tissue from AD patients and the AD transgenic mouse model APPswe/PS1dE9 (AD mice). Transgenic mice with brain tissue-specific Dkk3 expression were generated or crossed with AD mice to study the effects of Dkk3 on AD. In AD mice, transgenic expression of Dkk3 improved abnormalities in learning, memory, and locomotor activity, reduced the accumulation of amyloid-β, and ameliorated glucose uptake deficits. Furthermore, we determined that Dkk3 downregulated GSK-3β, a central negative regulator in canonical Wnt signaling, and upregulated PKCβ1, a factor implicated in noncanonical Wnt signaling. This indicates that increased activation of GSK-3β and the inhibition of PKCβ1 in AD patients may be responsible for the dysfunctional Wnt signaling in AD. In summary, our data suggest that Dkk3 is an agonist of Wnt signaling, and the ability of transgenic expression of Dkk3 to compensate for the decrease in Dkk3 expression in AD mice, reverse dysfunctional Wnt signaling, and partially inhibit the pathological development of AD suggests that Dkk3 could serve as a therapeutic target for the treatment of AD.

Keywords

Alzheimer’s disease cognition Dkk3 transgenic mice Wnt signaling pathway

INTRODUCTION

Alzheimer’s disease (AD) is a neurodegenerative disorder clinically characterized by progressive cognitive dysfunction [1 –3] as a consequence of structural and functional brain abnormalities, including the deposition of amyloid-β (Aβ) plaques, intracellular tangles of hyper-phosphorylated tau protein, and diminished cerebral glucose metabolism [4]. Despite intensive study of AD over the past decade, the mechanism underlying AD pathogenesis remains elusive, and no effective treatments are available for this disorder so far [2].

Wnt ligands and/or antagonists interact with their receptors in the plasma membrane and subsequently activate intracellular signaling pathways during development. Wnt signaling is involved in programming and regulating cellular proliferation, differentiation, translocation, polarization, and fate decisions during both embryonic development and tissue homoeostasis in adulthood [5]. Dysfunctional Wnt signaling is implicated in the development of AD, and the activation of the Wnt signaling pathway slows down the progression of AD [6 –8].

Dickkopf3 (Dkk3), a known regulator of Wnt signaling pathways, is frequently downregulated in solid cancer types and hematological malignancies due to aberrant promoter methylation, microRNA deregulation, or repressive histone modifications [9, 10]. Dkk3 overexpression suppresses tumor cell growth [11 –14], and compelling evidence from numerous in vitro and in vivo studies implicates Dkk3 as a central tumor suppressor. Much of our knowledge of the effects of Dkk3 on Wnt signaling has been obtained in cancer studies, which have shown that Dkk3 activates or inhibits the Wnt signaling pathway in different cell types or tissues [15].

Dkk3 is expressed in pyramidal neurons of the hippocampus and cortex [16, 17]. However, the role of Dkk3 and its interaction with Wnt signaling pathways in the pathogenesis of AD have not been elucidated. In the present study, we determined that Dkk3 expression was significantly decreased in brain tissue from both AD patients and APPswe/PS1dE9 transgenic AD model mice (AD mice) [18 –22]. Transgenic expression of Dkk3 to compensate for the decrease in endogenous expression in the brains of AD mice results in the improvement of AD phenotypes by rescuing the inhibition of Wnt signaling caused by reduced expression of Dkk3.

MATERIALS AND METHODS

Human brain tissue samples

Fresh frozen postmortem AD and control human brain samples were acquired from the Human Brain Bank of the Chinese Academy of Medical Sciences & Peking Union Medical College with the approval of the Institutional Review Board of the Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences (NO.009-2014). Tissue from the anterior prefrontal cortex was obtained from the brain samples by dissection and fresh frozen with a postmortem delay of <24 h [23, 24]. As shown in Table 1, the average age at time of death of the 4 non-AD controls and 4 AD individuals was similar (controls, 87.5 y; AD cases, 87.2 y).

Table 1

Controls and Alzheimer’s disease (AD) patients

Group	ID	Gender	Age (years)
Control	1	female	95
	2	male	85
	3	female	84
	4	female	86
AD patients	1	female	82
	2	male	89
	3	female	98
	4	female	80

Animals

The APPswe/PS1dE9 double-transgenic mouse model of AD (referred to as “AD”) was described in our previous study [18 –22]. The cDNA encoding mouse Dkk3 (GenBank Accession No. NM_015814) was cloned into an expression plasmid under the PDGF promoter (Supplementary Figure 1), and the brain-specific Dkk3 transgenic mice (referred to as “Dkk3”) were generated by microinjection. Genotyping was performed by PCR using primers 5^′-TCCTTCCCCGACGGTCACTT-3^′ and 5^′-CTGTCTCGGGTGCATAGCATCTG-3^′. The protein level of the target gene in the brain was analyzed by western blotting using antibodies against Dkk3 (ab186409, Abcam). F3-generation Dkk3 transgenic mice were crossed to APPswe/PS1dE9 transgenic mice to generate Dkk3×APPswe/PS1dE9 transgenic mice (referred to as “Dkk3×AD”). Non-transgenic littermates (referred to as “NTG”) and AD mice were used as controls.

All the mice used in this study were maintained on a C57BL/6J genetic background and bred in an AAALAC-accredited facility. The use of animals in this study was approved by the Animal Care and Use Committee of the Institute of Laboratory Animal Science of Peking Union Medical College (ILAS-GC-2015-002).

Behavioral tests

Behavioral tests of the NTG mice (n = 23), AD mice (n = 20), Dkk3 mice (n = 21), and Dkk3×AD mice (n = 17) were performed as previously described, with some modifications [18–22 , 26]. The spontaneous behavioral changes of the mice in a novel environment were measured using the open-field test. Ethovision XT monitoring and analysis software (Noldus Company, Netherlands) was used for scoring. Each mouse was allowed to freely explore the open field (50×50 cm), which was divided into a peripheral zone and the central zone, for 5 min. The time that each mouse spent in each zone and the number of rearing was recorded. Thigmotaxis (the tendency to remain close to walls) was assessed by duration in central zone and peripheral zone.

The Morris water maze (MWM) test was administered using a circular tank (100 cm in diameter) filled with water (22±1°C) that was made turbid with powdered milk. An escape platform (10 cm in diameter) was submerged 0.5 cm below the water’s surface in the center of one of four quadrants. Each mouse was subjected to two trials per day of 1 min each on 5 consecutive days. The time taken to reach the platform (escape latency) was recorded and the average of the two trials was determined. In the subsequent probe trial, each mouse was allowed to swim freely in the tank without the platform for 1 min, and the time and frequency spent in each quadrant, including the quadrant that formerly contained the platform, were recorded. Each mouse was allowed to recover in a plastic holding cage on an electric heater after the swim. The room temperature and light levels were maintained constant. Monitoring was performed using the Noldus Ethovision XT system.

PET/CT imaging

Glucose uptake was examined using PET/CT scans. Briefly, following 6 h of fasting, 6-month-old mice (n = 4, each group) were anaesthetized using 1.5% isoflurane with 31% O₂ inhalation through a nose cone (flow rate: 2.5 L/min), prior to injection of the tracer. Each mouse was imaged on a small animal scanner (MicroPET/CT, Inveon, Siemens). Prior to the dynamic small animal procedure, 18.5 MBq (500 μCi) ¹⁸F-FDG tracer (FDG, provided by Cancer Hospital, Chinese Academy of Medical Sciences) was injected intraperitoneally as a bolus (approximately 200 μL), and the animals were kept at room temperature for 45 min. Each mouse was then exposed to a 10-min PET scan, followed by a 10-min CT scan for attenuation corrections of the small-animal PET images. The images were reconstructed using the filtered back-projection algorithm with CT-based photon-attenuation correction [27]. The reconstructed images were examined on a 3D display in axial, coronal, and sagittal views. These images were used to manually define 3 regions of interest (ROIs) consisting of the frontal cortex, the temporal cortex, and the hippocampus. The injected dose (ID) per gram of tissue was calculated automatically using Inveon Research Workplace software (IRW). The field of view was 11.28×12.66 cm.

Histochemical staining

Following PET/CT scanning, brain tissue was removed from the mice (n = 3–4 for each group) and were immediately coronally sectioned (8-μm-thick sections) on a freezing microtome (LEICA CMl850, Nussloch, Germany). The cryosections were air-dried and fixed in 4 % (w/v) paraformaldehyde for 20 min as described by Yao et al. [28].

For thioflavin-S staining, the sections were incubated in 0.25% (w/v) potassium permanganate and 1% (w/v) oxalic acid until becoming white, followed by immersion in water and staining for 5 min with 1% (w/v) thioflavin-S (T1892, Sigma-Aldrich) solution in 50% (v/v) ethanol. Finally, the sections were washed, dehydrated, mounted on coverslips, and visualized using a stereo fluorescence microscope (Leica MZ16F). Hemispheres fields including cortex and hippocampus were imaged and analyzed by using Image J software (1.43u, NIH, USA). Plaques were quantified by an operator blind to genotype.

Immunofluorescence procedures were performed using conventional methods. The sections were incubated with primary antibody (anti-Aβ, 1:300, SIG-39320, Covance; anti-Dkk3, 1:50, 10365-1-AP, Proeintech; anti-GFAP,1:200, ab4674, Abcam; anti-Glut1, 1:200, ab652, Abcam) overnight at 4°C. After rinsing with 0.01 M PBS, immunoreactivity was visualized by incubation with Alexa Fluor^® 488-conjugated goat anti-mouse IgG (H + L) secondary antibody (1:200, A28175, Thermo Scientific) or Alexa Fluor^® 555 conjugated goat anti-rabbit IgG (H + L) secondary antibody (1:200, A27039, Thermo Scientific). After rinsing thoroughly with 0.01 M PBS, the coverslips were mounted on slides using fluorescence mounting medium (ZLI-9557, ZSGB-BIO). Images were captured using a Leica TCS SPE laser scanning confocal microscope (Leica Microsystems GmbH, Mannheim, Germany).

Western blotting

After the behavioral tests, the mice (n = 3 for each group) were randomly chosen and sacrificed by cervical dislocation. Brain hemispheres were directly homogenized in RIPA buffer containing 0.1% (w/v) PMSF (36978, Thermo Scientific) and 0.1% (w/v) protease inhibitor cocktail (87785, Thermo Scientific). The lysates were centrifuged at 14,000 rpm for 30 min at 4°C and the protein concentration in the supernatants was determined using a BCA assay. The protocol used for differential detergent extraction was modified from previously reported methods [29]. Briefly, freshly dissected brain tissues from the mice (n = 3 for each group) were homogenized with TBS extraction buffer containing 50 mM Tris (pH 7.4), 150 mM NaCl, 1 mM NaVO₃, 1 mM NaF, a protease inhibitor mixture (87785, Thermo Scientific), a phosphatase inhibitor cocktail (78420, Thermo Scientific), and 1 mM phenylmethylsulfonylfluoride (PMSF, 36978, Thermo Scientific) using a Polytron PT2500 E homogenizer (Kinematica Inc., Bohemia, NY, USA). The TBS-insoluble pellets were then resuspended and homogenized in 2% (w/v) SDS extraction buffer. Finally, the supernatant fractions were collected as TBS-insoluble or 2% (w/v) SDS-soluble fractions for Aβ detection.

Western blotting was performed according to standard procedures. Briefly, proteins were separated by 10% SDS-PAGE or 4–12% NuPAGE^® Novex^® Bis-Tris Gels (for Aβ, NP0321BOX, Thermo Scientific), and then transferred onto nitrocellulose membranes (Immobilon NC, Millipore, Molsheim, France). For Aβ analysis, the membranes were cross-linked with 0.5% (v/v) glutaraldehyde solution before blocking. After blocking in 5% (w/v) nonfat milk, the membranes were incubated with the following antibodies: Aβ (1:1000, SIG-39320, Covance), Dkk3 (1:1000, ab186409, Abcam), PKCβ1 (1:100, ab172919, Abcam), phopho-PKCβ1 Thr642 (1:500, ab75657, Abcam), GSK-3β (1:1000, #9315, CST), phopho-GSK-3β Ser9 (1:1000, #9323, CST), p-PHF-Tau Ser202/Thr205 (AT8) (1:1000, MN1021, Thermo Scientific), and Tau-5 (1:1000, 577801, Calbiochem). Primary antibodies were visualized using anti-rabbit or mouse HRP-conjugated secondary antibodies (Santa Cruz Biotechnology, Inc.) and a chemiluminescent detection system (Western Blotting Luminol Reagent; Santa Cruz Biotechnology, Inc.). Variations in sample loading were accounted for by normalization to GAPDH. Quantitative analysis was performed by densitometry using NIH Image J software.

Statistical analyses

The results are presented as means±SEM. With the exception of data derived from the MWM experiments, statistical analysis was performed by parametric one-way analysis of variance (ANOVA), and t-tests were used for comparisons between two groups. For the MWM tests, escape latency in the hidden platform trial was analyzed using repeated measures two-way ANOVA, and the data obtained from the probe trial was analyzed by one-way ANOVA. Differences were considered statistically significant at p < 0.05.

RESULTS

Dkk3 expression is downregulated in brain tissue of AD patients and AD mice

To gain insight into the putative role of Dkk3 in AD, we measured the expression of Dkk3 in brain tissues from AD patients. Dkk3 expression was significantly decreased by 66.8% in the brain tissues of AD patients (Fig. 1a, n = 4, p < 0.05) compared to control brain tissue (n = 4), as determined by western blotting. We further assessed Dkk3 expression in brain tissues from 4- and 6-month-old AD mice and age-matched NTG mice. Consistent with the results obtained in AD patient tissues, the average expression of Dkk3 was reduced by 30.0% and 74.6% in 4- and 6-month-old AD mice, respectively, compared with age-matched NTG mice (Fig. 1b, n = 3, p < 0.05). Decreased expression and distribution of Dkk3 in coronal brain sections from 6-month-old AD mice were also observed by confocal microscopy (Fig. 1d and Supplementary Figure 2). In these AD mice, early behavioral changes occur at 4 months of age, suggesting that the abnormal decrease in Dkk3 expression could be an early event in the pathological development of AD. Thus, transgenic mice overexpressing Dkk3 were generated (Supplementary Figure 1). Based on the level of Dkk3 expression as determined by western blotting (Fig. 1c), one line (Founder #16) was selected for crossing with the AD mouse line to obtain Dkk3×AD mice. Increased expression and distribution of Dkk3 in brain sections from Dkk3 and Dkk3×AD mice were observed by immunofluorescent staining (Fig. 1d and Supplementary Figure 2). Moreover, double-fluorescence staining indicated that neurons (NeuN+) rather than astrocytes (GFAP+) were the main source of Dkk3 in the brain (Fig. 1e and Supplementary Figure 3). These mice were further investigated to study the putative neuroprotective effects of Dkk3 in AD.

Fig.1

Determination of Dkk3 expression in brain tissues and the generation of Dkk3 transgenic mice. Protein levels of Dkk3 in brain tissues from postmortem AD patients (a) and from AD model mice (b) were detected by western blotting and quantitatively compared with those of age-matched controls. Brain-specific Dkk3 transgenic mice in which Dkk3 is overexpressed 2.5-fold compared to the NTG mice were generated (c). The expression and distribution of Dkk3 in coronal brain sections from 6-month-old NTG, AD, Dkk3, and Dkk3×AD mice were observed by immunofluorescent with an anti-Dkk3 antibody (d), which revealed an abnormal decrease in Dkk3 expression in AD mice and an expected increase in Dkk3 expression in Dkk3 and Dkk3×AD mice. Further, double-fluorescent staining indicated that Dkk3 was predominantly localized to the neurons rather than the astrocytes (GFAP+) (e). No co-localization was observed between Dkk3 (red) and GFAP (green), as indicated by the white arrows; DAPI-stained nuclei are shown in blue. The results shown are representative from three independent experiments. All data are presented as the mean±SEM. For a, *p < 0.05, AD patients’ brains versus control brains; for b, *p < 0.05, AD mice versus NTG mice; for c, *p < 0.05, Dkk3 mice versus NTG mice.

Transgenic expression of Dkk3 attenuates abnormal exploratory activity and cognitive deficits in AD mice

The exploratory activities of the NTG, AD, Dkk3, and Dkk3×AD mice in a novel environment were determined at 4, 6, and 8 months of age using the open-field test. The AD mice displayed increased rearing number, increased total distance covered and time spent in the central zone, and decreased time spent in the peripheral zone (Fig. 2a-c and Supplementary Table 1). However, this abnormal exploratory activity was ameliorated in Dkk3×AD mice, as assessed by the decreases in the number of rearing (32.8%, p < 0.05) and in total distance covered (19.9%, p < 0.05) and by the increase in the amount of time spent in the peripheral zone (7.5%, p > 0.05) compared with the AD mice (Fig. 2a-c and Supplementary Table 1).

Fig.2

Analysis of the behavior and cognitive phenotypes. The open-field test was administered to NTG (n = 23), AD (n = 20), Dkk3 (n = 21) and Dkk3×AD (n = 17) mice and the number of rearing (a), total distance moved (b), and thigmotaxis (c) were measured to evaluate the spontaneous behavioral changes of the mice in a novel environment. The Morris Water Maze test was also performed to NTG (n = 23), AD (n = 20), Dkk3 (n = 21), and Dkk3×AD (n = 17) mice to measure the escape latency (d), the duration in the target zone (e), and number of target crossings (f) to evaluate the spatial learning and memory of the mice. All data are presented as the mean±SEM. *p < 0.05, **p < 0.01, ***p < 0.001, AD mice versus NTG mice; ^# p < 0.05, Dkk3×AD mice versus AD mice.

The spatial learning and memory abilities of NTG, AD, Dkk3, and Dkk3×AD mice were also analyzed using the MWM. In the hidden platform test, the escape latencies of AD mice were significantly longer than those of NTG mice (Fig. 2d and Supplementary Table 2, p < 0.05 for day 2–5, respectively). However, the escape latencies of Dkk3×AD mice were significantly decreased compared with those of AD mice (Fig. 2d and Supplementary Table 2, p < 0.05 for day 4–5, respectively). In the probe test, AD mice showed an obvious decrease in the time spent in the target zone (Fig. 2e and Supplementary Table 3, 41.9%, p < 0.05) and in the number of target crossings (Fig. 2f and Supplementary Table 3, 65.2%, p < 0.001) compared with NTG mice. By contrast, the transgenic expression of Dkk3 significantly improved the preference for the target quadrant of Dkk3×AD mice (Fig. 2e, f and Supplementary Table 3).

Dkk3 overexpression reduces Aβ accumulation in AD mice

The accumulation of Aβ in NTG, AD, Dkk3, and Dkk3×AD mice was detected using thioflavin-S and 6E10 antibody staining as well as western blotting. No visible plaques were observed in NTG and Dkk3 transgenic mice (Fig. 3a). However, a number of plaques were visible in the hippocampus and cortex in AD mice, whereas transgenic expression of Dkk3 significantly reduced the total number and area of plaques (Fig. 3a-c, n = 3, p < 0.05) and intracellular Aβ deposition (Fig. 3d), and single plaque size (shown in Supplementary Fig. 4) in coronal brain sections when compared with AD mice. Furthermore, transgenic expression of Dkk3 also dramatically reduced the levels of monomeric and oligomeric Aβ in the TBS-insoluble fractions of brain tissue from Dkk3×AD mice compared with tissue from AD mice (Fig. 3e, f, n = 3, p < 0.001).

Fig.3

Observation of pathological changes. Aβ plaques were stained by thioflavin-S in brain sections from 6-month-old NTG, AD, Dkk3, and Dkk3×AD mice (a), and the number (b) and area (c) of Aβ plaques in the hippocampus and cortex (n = 3 for each group, 30 slices for each mice) were analyzed. Aβ plaques in the hippocampus and cortex were further confirmed by immunofluorescent staining with anti-Aβ antibody; cortical areas were zoomed in to observe Aβ deposition in neurons (d). Aβ monomers, dimers, and trimers in the TBS-insoluble fractions of brain tissue from NTG, AD, Dkk3, and Dkk3×AD mice were detected by western blotting (e) and quantitatively analyzed using GAPDH for normalization (f). The results shown are representative of three independent experiments. All data are presented as the mean±SEM. *p < 0.05, AD mice versus NTG mice; ^# p < 0.05, Dkk3×AD mice versus AD mice.

Dkk3 overexpression improves abnormal glucose uptake in AD mice

Cerebral glucose metabolism is reduced in AD brains and FDG-PET imaging studies have shown that this reduction is an early feature of disease progression [30]. Glucose uptake levels in the brains of NTG, AD, Dkk3, and Dkk3×AD mice were comparatively analyzed using ¹⁸F-FDG PET scans. The frontal cortex, the temporal cortex, and the hippocampus were defined as ROIs using CT images as an anatomical reference (Fig. 4a). Static PET images of the brain showed that glucose uptake was dramatically inhibited in AD mice and rescued in the Dkk3×AD mice (Fig. 4b). ¹⁸F-FDG uptake in the frontal cortex, the temporal cortex, and the hippocampus was significantly decreased by 28.3%, 31.1% and 30.2%, respectively, in AD mice compared with NTG mice. By contrast, transgenic expression of Dkk3 in the Dkk3×AD mice rescued glucose uptake in these three ROIs to normal levels (Fig. 4c, n = 3, p < 0.01).

Fig.4

Analysis of changes in energy metabolism. Glucose uptake in NTG, AD, Dkk3, and Dkk3×AD mice was compared by ¹⁸F-FDG PET/CT scans. CT images (a) were used as an anatomical reference to define the frontal cortex (arrow), temporal cortex (arrowhead), and hippocampus (asterisk) as regions of interest (ROIs). Representative static PET images of the brain (axial view) consisting of the ROIs were obtained to show the level of glucose uptake in whole brains (b). ¹⁸F-FDG uptake in the ROIs was calculated using Inveon Research Workplace software (c). The expression of Glut1 in coronal brain sections was observed under a confocal scanning microscope by immunofluorescence (d). DG, dentate gyrus. The results shown are representative of three independent experiments. n = 4 mice per group. All data are presented as the mean±SEM. *p < 0.05, AD mice versus NTG mice; ^# p < 0.05, Dkk3×AD mice versus AD mice.

The expression of glucose transporter type-1 (Glut-1) is decreased in AD patients and APP/PS1 mice, leading to further impairment of energy metabolism [31]. Immunofluorescent staining revealed reduced expression of Glut1 consistent with the decreased glucose uptake in AD mice, whereas Glut1 expression was increased in the Dkk3×AD mice due to the transgenic expression of Dkk3 (Fig. 4d).

Dkk3 overexpression inhibits the activity of GSK-3β and increases the phosphorylation of PKCβ1 in AD mice

GSK-3β, a key factor in the canonical Wnt signaling pathway, is constitutively active in most tissues and is most commonly regulated by inhibitory phosphorylation of Ser9, which is negatively correlated with Tau hyperphosphorylation [32]. Our results showed that the abnormal decrease in Dkk3 expression corresponded with decreased levels of phosphorylation of Ser9 of GSK-3β and increased tau phosphorylation in AD mice compared with the NTG mice, whereas the transgenic expression of Dkk3 enhanced the phosphorylation of GSK-3β Ser9 and reduced tau phosphorylation in the Dkk3×AD mice compared with AD mice (Fig. 5a-d, n = 3, p < 0.05). PKCα, PKCɛ, PKCβ1, and CaMKII belong to the noncanonical Wnt signaling pathway. Our results showed that the levels of phosphorylated PKCα, PKCɛ, and CaMKII did not change in correspondence with either the decrease or increase in the level of Dkk3 in AD mice or Dkk3×AD mice, respectively (shown in Supplementary Figure 5). Notably, the phosphorylation of PKC β1 was inhibited by 37.1% in AD mice compared with the NTG mice, consistent with the reduced expression of Dkk3 in the AD mice, but increased by 90.5% in Dkk3×AD mice compared with AD mice due to the transgenic expression of Dkk3 (Fig. 5a, e, p < 0.05). Consistent with these results, the phosphorylation of GSK-3β at Ser9 and phosphorylation of PKCβ1 were also significantly downregulated in brain tissue from AD patients (Fig. 5f-h, p < 0.05).

Fig.5

Determination of the effects of Dkk3 on the Wnt signaling pathway. The abnormal decrease in Dkk3 expression in AD mice was compensated for by transgenic expression of Dkk3 in Dkk3×AD mice. The phosphorylation levels and total levels of Wnt signaling proteins in brain tissue from NTG, AD, Dkk3, and Dkk3×AD mice were detected by western blotting and quantitatively analyzed using GAPDH for normalization, including Dkk3 (a, b), GSK-3β (a, c), PHF-Tau (a, d) and PKCβ1 (a, e). The phosphorylation levels and total levels of GSK-3β (f, g), and PKCβ1 (f, h) in brain tissue from AD patients and controls were also detected by western blotting and quantitatively analyzed using GAPDH for normalization. The results shown are representative of three independent experiments. All data are presented as the mean±SEM. For b-e, *p < 0.05, AD mice versus NTG mice; ^# p < 0.05, Dkk3×AD mice versus AD mice. For g and h, n = 4 per group, *p < 0.05, AD patients’ brains versus control brains.

Fig.6

Schematic diagram showing the protection of Dickkopf 3 (Dkk3) against AD through Wnt signaling pathway. Wnt binds to the classical receptor Frizzled (Fz) and its coreceptors and sequentially activates canonical Wnt and/or non-canonical Wnt signaling pathways. ApoE4 inhibits the activation of the canonical Wnt signaling pathway, while Dkk3 activates both of canonical Wnt and non-canonical Wnt signaling pathways followed by inhibition of GSK-3β and activation of PKC. The GSK-3β inhibition leads to the increase of ChAT activity and compensates the cholinergic hypofunction as well as prevents the Tau hyperphosphorylation. PKC activation increases the activity of ADAM10, which contributes to α-secretase cleavage and prevents deposition of amyloid-β. CM, cell membrane; Dvl1, Disheveled 1; NE, nucleolus.

DISCUSSION

Dickkopf 3 (Dkk3) is a member of the Dickkopf family, which comprises four members (Dkk-1 to Dkk-4) and a unique Dkk-3-related protein (DKKL1, or SGY-1) [33, 34]. Dkk3 is frequently downregulated in solid cancer types, hematological malignancies [11 –14] and liver steatosis [35], and unregulated in dilated cardiomyopathy [36, 37]. Dkk3 has been shown to act as an agonist [37, 38] or antagonist [11 , 39–41] of the Wnt signaling pathway, depending on the cell types or tissues; the role of Dkk3 appears to be highly dependent on context, including cell surface receptors, co-receptors, and co-expressed ligands.

Dkk3 expression was significantly decreased in brain tissues from both AD patients and AD model mice (Fig. 1), suggesting that an abnormal decrease in Dkk3 expression may be involved in AD pathogenesis. Thus, we generated brain-specific Dkk3 transgenic mice and crossed these transgenic mice with AD mice to compensate for the decrease in Dkk3 expression in the brains of AD mice. We then examined the effect of Dkk3 on spatial learning and memory using the MWM test and spontaneous behaviors using the open-field test. The results indicated that transgenic expression of Dkk3 improves abnormalities in learning, memory, and locomotor activity of AD mice (Fig. 2). Furthermore, the transgenic expression of Dkk3 also reduced the accumulation of Aβ and ameliorated disturbances in brain energy metabolism in AD mice (Figs. 3 and 4).

Wnt signaling can be classified into canonical and noncanonical pathways. Canonical Wnt signaling is dependent on the β-catenin pathway, of which GSK-3β is a critical downstream element. The constitutive activation of GSK-3β is inhibited in the presence of the Wnt protein, leading to the accumulation and stabilization of β-catenin in the cytosol and its translocation into the nucleus [42, 43]. Noncanonical Wnt signaling involves at least two β-catenin-independent pathways, the Wnt/JNK pathway, and the calcium pathway [37]. Protein kinase C (PKC) is a key element of the calcium pathway and is activated in response to the release of intracellular Ca²⁺. PKC decreases cyclic guanosine monophosphate (cGMP) by Wnt ligand-receptor interactions via phospholipase-C (PLC) [44]. Altered expression of key Wnt pathway components has been observed in the brains of AD patients [45, 46], as well as the APPswe/PS1dE9 mouse model of AD [47], supporting a relationship between deregulation of Wnt pathways and Tau hyperphosphorylation and cognitive decline in AD. Age-related downregulation of Wnt pathways was recently observed in the hippocampus in 9- and 12-month-old SAMP8 mice [48]. Similarly, we observed decreased levels of phosphorylated GSK-3β Ser9 and PKCβ1 in both AD patients and 6-month-old AD mice, which was in accordance with the abnormal decrease in Dkk3 expression in both AD patients and AD mice. Interestingly, the phosphorylation of GSK-3β and PKC β1 were significantly enhanced when the decreased Dkk3 expression in AD mice was compensated by the transgenic expression of Dkk3 (Fig. 5).

GSK-3β has been suggested as a critical molecular link between senile plaques and neurofibrillary tangles [49]. GSK-3β phosphorylates APP (Thr668), Tau, and presenilin1 [50 –52] and enhances the formation of Aβ plaques and neurofibrillary tangles [53]. Our previous studies have revealed an abnormal increase in GSK-3β activity and its relationship to AD pathogenesis [21, 54]. The activated GSK-3β also inhibits the choline acetyltransferase (ChAT) activity [55, 56], leading to reduction of ACh and the cholinergic hypofunction that contributes to cognitive and behavioral deficits in patients with AD [57]. By contrast, inhibitors of GSK-3β [49] or M1 muscarinic agonists [58] have been reported to rescue cognitive deficits in several murine models of AD. In the present paper, we observed that the decrease in inhibitory phosphorylation of GSK-3β at Ser9 and the abnormal decrease in Dkk3 expression occurred simultaneously in AD mice, along with the increase in Tau hyperphosphorylation. Furthermore, compensation for the abnormal decrease in endogenous Dkk3 by transgenic expression of Dkk3 reversed the increase in GSK-3β activity and Tau phosphorylation in Dkk3×AD mice (Fig. 5).

The role of PKC in memory processes has been well-established in animal models [59, 60] and has been found to be defective in AD [61, 62]. The ADAM10, a member of the a disintegrin and metalloproteinase (ADAM) family, is a constitutive α-secretase and plays a critical role in amyloid-β protein precursor cleavage and inhibition of the Aβ deposition [63]. The activation of PCK increases the activity of ADAM10 [64, 65], and the treatment of the AD mouse models with PCK activators, such as benzolactam and bryostatin, reduces Aβ depostion in the brain and improves premature death and behavioral outcomes [66]. In the present study, we assessed the levels of the phosphorylation of CaMKII and several PKC family members, including PKCα, PKCɛ, and PKCβ1 (shown in Supplementary Figure 5), in all four mouse genotypes, which revealed that the phosphorylation of PKCβ1 was decreased in the brains of the AD mice and increased in the brains of the Dkk3×AD mice. The reduction of Aβ deposition in the Dkk3×AD mice (Fig. 3) might be partly through the activation of PCK induced by Dkk3.

Allele 4 of apolipoprotein E (apoE4) is the most important risk factor of the sporadic form of AD [67, 68]. By contrast with Dkk3, apoE4 causes inhibition of the canonical Wnt signaling pathway in PC-12 cells, when combined with very low-density lipoproteins or in cells over-expressing the low-density lipoprotein receptor-related protein, LRP6 [69].

Taken together, the amelioration of AD by Dkk3 may be through canonical and noncanonical Wnt signaling pathways (Fig. 6).

Similar to PKC, AMPK, a highly conserved serine/threonine protein kinase and metabolic sensor, is also relevant to Wnt signaling [70]. It has been proposed that the activation of AMPK is involved in the beneficial effects of Leptin on AD-related biochemical pathways, including Tau phosphorylation and Aβ production [71 –74]. However, there are contradictory reports of AMPK activation in AD. Increased AMPK activity was observed in the hippocampus in the J20 transgenic mouse model [75] and APP/PS2 mice [76]; by contrast, a slightly decreased ratio of activated p-AMPK (Thr172)/total APMK was observed in 3-month-old APPsw/PS1dE9 mice, whereas AMPK activity was unchanged in 6-month-old APPsw/PS1dE9 mice [77]. Consistent with the previous report in APPsw/PS1dE9 mice, our results indicated that p-AMPKα was not altered in 6-month-old APPsw/PS1dE9 mice (Supplementary Fig. 5), suggesting that the effect of Dkk3 on reduced Tau phosphorylation and Aβ production is independent of the AMPK pathway in our model.

In summary, Dkk3 expression is abnormally decreased in AD patients and an AD mouse model and is associated with the activation of GSK-3β and inhibition of PKCβ1, key elements of the Wnt signaling pathways. Transgenic expression of Dkk3 in the brains of AD mice compensates for the abnormal decrease in endogenous levels, thus reversing the changes in GSK-3β and PKCβ1 and improving the phenotypes of AD mice. These results suggest that Dkk3 is an agonist of Wnt signaling and could partially inhibit the pathological development of AD through the Wnt signaling pathway.

Footnotes

ACKNOWLEDGMENTS

The present work was supported by National Natural Science Foundation of China (No.81571222), Fundamental Research Funds for the Central Universities (2016ZX310039), and CAMS Innovation Fund for Medical Sciences (CIFMS, 2016-I2M-1-004). Tissue provided by: Human Brain Bank, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China. This study was sponsored by the China Human Brain Bank Consortium.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Holmes

, Boche

, Wilkinson

, Yadegarfar

, Hopkins

, Bayer

, Jones

, Bullock

, Love

, Neal

, Zotova

, Nicoll

(2008) Long-term effects of Abeta42 immunisation in Alzheimer’s disease: Follow-up of a randomised, placebo-controlled phase I trial. Lancet 372, 216–223.

Mattson

(2004) Pathways towards and away from Alzheimer’s disease. Nature 430, 631–639.

Stokin

, Lillo

, Falzone

, Brusch

, Rockenstein

, Mount

, Raman

, Davies

, Masliah

, Williams

, Goldstein

(2005) Axonopathy and transport deficits early in the pathogenesis of Alzheimer’s disease. Science 307, 1282–1288.

Chen

, Zhong

(2013) Decoding Alzheimer’s disease from perturbed cerebral glucose metabolism: Implications for diagnostic and therapeutic strategies. Prog Neurobiol 108, 21–43.

Kim

, Kim

, Jho

(2013) Wnt/β-catenin signalling: From plasma membrane to nucleus. Biochem J 450, 9–21.

Inestrosa

, Montecinos-Oliva

, Fuenzalida

(2012) Wnt signaling: Role in Alzheimer disease and schizophrenia. J Neuroimmune Pharmacol 7, 788–807.

Silva-Alvarez

, Arrázola

, Godoy

, Ordenes

, Inestrosa

(2013) Canonical Wnt signaling protects hippocampal neuronsfrom Aβ oligomers: Role of non-canonical Wnt-5a/Ca(2+) inmitochondrial dynamics. Front Cell Neurosci 7, 97.

Vargas

, Fuenzalida

, Inestrosa

(2014) In vivo activation of Wnt signaling pathway enhances cognitive function of adult mice and reverses cognitive deficits in an Alzheimer’s disease model. J Neurosci 34, 2191–2202.

Veeck

, Bektas

, Hartmann

, Kristiansen

, Heindrichs

, Knuchel

, Dahl

(2008) Wnt signalling in human breast cancer: Expression of the putative Wnt inhibitor Dickkopf-3 (DKK3) is frequently suppressed by promoter hypermethylation in mammary tumours. Breast Cancer Res 10, R82.

10.

Valdora

, Banelli

, Stigliani

, Pfister

, Moretti

, Kool

, Remke

, Bai

, Brigati

, Hielscher

, Romani

, Servidei

, Zollo

, Cinalli

, Oberthuer

, Tonini

, Coco

(2013) Epigenetic Silencing of DKK3 in medulloblastoma. Int J Mol Sci 14, 7492–7505.

11.

Hoang

, Kubo

, Healey

, Yang

, Nathan

, Kolb

, Mazza

, Meyers

, Gorlick

(2004) Dickkopf 3 inhibits invasion and motility of Saos-2 osteosarcoma cells by modulating the Wnt-beta-catenin pathway. Cancer Res 64, 2734–2739.

12.

Lodygin

, Epanchintsev

, Menssen

, Diebold

, Hermeking

(2005) Functional epigenomics identifies genes frequently silencedin prostate cancer. Cancer Res 65, 4218–4227.

13.

Kawano

, Kitaoka

, Hamada

, Walker

, Waxman

, Kypta

(2006) Regulation of prostate cell growth and morphogenesis by Dickkopf-3. Oncogene 25, 6528–6537.

14.

Abarzua

, Sakaguchi

, Takaishi

, Nasu

, Kurose

, Ebara

, Miyazaki

, Namba

, Kumon

, Huh

(2005) Adenovirus-mediated overexpression of REIC/Dkk-3 selectively induces apoptosis in human prostate cancer cells through activation of c-Jun-NH2-kinase. Cancer Res 65, 9617–9622.

15.

Zhan

, Rindtorff

, Boutros

(2017) Wnt signaling in cancer. Oncogene 36, 1461–1473.

16.

Diep

, Hoen

, Backman

, Machon

, Krauss

(2004) Characterisation of the Wnt antagonists and their response toconditionally activated Wnt signalling in the developing mouseforebrain. Brain Res Dev Brain Res 153, 261–270.

17.

Meister

, Papatriantafyllou

, Nordström

, Kumar

, Ludwig

, Lui

, Boyd

, Popovic

, Fleming

, Moldenhauer

, Nawroth

, Gröne

, Waldmann

, Oelert

, Arnold

(2015) Dickkopf-3, a tissue-derived modulator of local T-cell responses. Front Immunol 6, 78.

18.

Zhang

, Hao

, Liu

, Zhang

, Lei

, Su

, Miao

, Li

(2011) Soluble Aβ levels correlate with cognitive deficits in the 12-month-old APPswe/PS1dE9 mouse model of Alzheimer’s disease. Behav Brain Res 222, 342–350.

19.

Zong

, Wang

, Dong

, Quan

, Zhu

, Xu

, Huang

, Ma

, Qin

(2011) miR-29c regulates BACE1 protein expression. Brain Res 1395, 108–115.

20.

Wang

, Yang

, Zhang

, Guan

, Zhang

(2013) Naringin enhances CaMKII activity and improves long-term memory in a mouse model of Alzheimer’s disease. Int J Mol Sci 14, 5576–5586.

21.

Wang

, Li

, Gao

, Lu

, Zhang

, Ma

, Ye

, Zhang

(2013) Cardiotrophin-1 (CTF1) ameliorates glucose-uptake defects and improves memory and learning deficits in a transgenic mouse model of Alzheimer’s disease. Pharmacol Biochem Behav 107, 48–57.

22.

Wang

, Gao

, Li

, Shen

, Zhang

, Ma

, Qin

, Zhang

(2012) Long-term naringin consumption reverses a glucose uptake defect and improves cognitive deficits in a mouse model of Alzheimer’s disease. Pharmacol Biochem Behav 102, 13–20.

23.

, Gao

, Zhan

, Xiong

, Qiu

, Qian

, Wang

, Zhang

, Yang

, Wang

, Bao

, Dou

, Tian

, Meng

, Gai

, Huang

, Yan

, Ge

, Ma

(2016) Quantitative protein profiling of hippocampus during human aging. Neurobiol Aging 39, 46–56.

24.

, Xiong

, Tian

, Zhan

, Gao

, Qiu

, Wang

, Ge

, Ma

(2016) Temporal lobe in human aging: A quantitative protein profiling study of samples from Chinese Human Brain Bank. Exp Gerontol 73, 31–41.

25.

Laczó

, Vlcek

, Vyhnálek

, Vajnerová

, Ort

, Holmerová

, Tolar

, Andel

, Bojar

, Hort

(2009) Spatial navigation testing discriminates two types of amnestic mildcognitive impairment. Behav Brain Res 202, 252–259.

26.

Liang

, Hon

, Tyan

, Liao

(1994) Involvement of hippocampal NMDA and AMPA receptors in acquisition, formation and retrieval of spatial memory in the Morris water maze. Chin J Physiol 37, 201–212.

27.

Chow

, Rannou

, Chatziioannou

(2005) Attenuation correction for small animal PET tomographs. Phys Med Biol 50, 1837–1850.

28.

Yao

, Zhang

, Huang

, Zhu

, Liu

, Ma

, Sheng

, Qin

(2013) Regional and cell-type specific distribution of HDAC2 in the adult mouse brain. Brain Struct Funct 218, 563–573.

29.

Choi

, Kim

, Hebisch

, Sliwinski

, Lee

, D’Avanzo

, Chen

, Hooli

, Asselin

, Muffat

, Klee

, Zhang

, Wainger

, Peitz

, Kovacs

, Woolf

, Wagner

, Tanzi

, Kim

(2014) A three-dimensional human neural cell culture model of Alzheimer’s disease. Nature 515, 274–278.

30.

Cohen

, Klunk

(2014) Early detection of Alzheimer’s disease using PiB and FDG PET. Neurobiol Dis 72, 117–122.

31.

Hooijmans

, Graven

, Dederen

, Tanila

, van Groen

, Kiliaan

(2007) Amyloid beta deposition is related to decreased glucose transporter-1 levels and hippocampal atrophy in brains of aged APP/PS1 mice. Brain Res 1181, 93–103.

32.

Jin

, Yin

, Yu

, Cao

, Gong

, Iqbal

, Ding

, Gu

, Liu

(2015) Truncation and activation of GSK-3β by calpain I: Amolecular mechanism links to tau hyperphosphorylation in Alzheimer’s disease. Sci Rep 5, 8187.

33.

Niehrs

(2006) Function and biological roles of the Dickkopf family of Wnt modulators. Oncogene 25, 7469–7481.

34.

Veeck

, Dahl

(2012) Targeting the Wnt pathway in cancer: The emerging role of Dickkopf-3. Biochim Biophys Acta 1825, 18–28.

35.

Xie

, Wang

, Zhang

, Zhao

, Wang

, Guo

, Zhu

, Zhang

, Li

(2016) DKK3 expression in hepatocytes defines susceptibility to liver steatosis and obesity. J Hepatol 65, 113–124.

36.

Bao

, Cai

, Zhang

, Li

, Liu

, Wan

, Hu

, Wan

, Zhang

, Zhu

, Xia

, Li

(2015) Dickkopf-3 protects against cardiac dysfunction and ventricular remodelling following myocardial infarction. Basic Res Cardiol 110, 25.

37.

, Bao

, Dong

, Liu

, Zhang

, Gao

, Ge

, Gao

, Zhang

(2016) Dkk3 prevents familial dilated cardiomyopathy development through Wnt pathway. Lab Invest 96, 239–248.

38.

Zenzmaier

, Sklepos

, Berger

(2008) Increase of Dkk-3 bloodplasma levels in the elderly. Exp Gerontol 43, 867–870.

39.

Lee

, Jo

, Rho

, Park

, Yoo

, Park

, Chae

, Zhang

, Lee

(2009) Dkk3, downregulated in cervical cancer, functions as a negative regulator of beta-catenin. Int J Cancer 124, 287–297.

40.

Wang

, Yin

, Yuan

, Sakamaki

, Okano

, Glazer

(2008) Musashi1 modulates mammary progenitor cell expansion through proliferin-mediated activation of the Wnt and Notch pathways. Mol Cell Biol 28, 3589–3599.

41.

Yue

, Sun

, Dacic

, Landreneau

, Siegfried

, Yu

, Zhang

(2008) Downregulation of Dkk3 activates beta-catenin/TCF-4 signaling in lung cancer. Carcinogenesis 29, 84–92.

42.

Angers

, Moon

(2009) Proximal events in Wnt signal transduction. Nat Rev Mol Cell Biol 10, 468–477.

43.

Port

, Basler

(2010) Wnt trafficking: New insights into Wntmaturation, secretion and spreading. Traffic 11, 1265–1271.

44.

Farias

, Godoy

, Cerpa

, Varela-Nallar

, Inestrosa

(2010) Wnt signaling modulates pre- and postsynaptic maturation:Therapeutic considerations. Dev Dyn 239, 94–101.

45.

Boonen

, van Tijn

, Zivkovic

(2009) Wnt signaling in Alzheimer’s disease: Up or down, that is the question. Ageing Res Rev 8, 71–82.

46.

Riise

, Plath

, Pakkenberg

, Parachikova

(2015) Aberrant Wnt signaling pathway in medial temporal lobe structures of Alzheimer’s disease. J Neural Transm (Vienna) 122, 1303–1318.

47.

Toledo

, Inestrosa

(2010) Activation of Wnt signaling by lithium and rosiglitazone reduced spatial memory impairment and neurodegeneration in brains of an APPswe/PSEN1DeltaE9 mouse model of Alzheimer’s disease.. Mol Psychiatry 15, 272–285, 228.

48.

Bayod

, Felice

, Andrés

, Rosa

, Camins

, Pallas

, Canudas

(2015) Downregulation of canonical Wnt signaling inhippocampus of SAMP8 mice. Neurobiol Aging 36, 720–729.

49.

Llorens-Martin

, Jurado

, Hernández

, Avila

(2014) GSK-3β, a pivotal kinase in Alzheimer disease. FrontMol Neurosci 7, 46.

50.

Avila

, Lucas

, Perez

, Hernandez

(2004) Role of tau protein in both physiological and pathological conditions. Physiol Rev 84, 361–384.

51.

Hanger

, Hughes

, Woodgett

, Brion

, Anderton

(1992) Glycogen synthase kinase-3 induces Alzheimer’s disease-like phosphorylation of tau: Generation of paired helical filament epitopes and neuronal localisation of the kinase. Neurosci Lett 147, 58–62.

52.

Kirschenbaum

, Hsu

, Cordell

, McCarthy

(2001) Substitution of a glycogen synthase kinase-3beta phosphorylation site in presenilin 1 separates presenilin function from beta-catenin signaling. J Biol Chem 276, 7366–7375.

53.

Grimes

, Jope

(2001) The multifaceted roles of glycogen synthase kinase 3beta in cellular signaling. Prog Neurobiol 65, 391–426.

54.

Zhang

, Liu

, Wu

, Tao

, Sui

, Yao

, Xu

, Huang

, Zhu

, Sheng

, Qin

(2014) Tubastatin A/ACY-1215 improves cognition in Alzheimer’s disease transgenic mice. J Alzheimers Dis 41, 1193–1205.

55.

Zhao

, Chu

, Li

, Yang

, Niu

, Qin

, Hao

, Dong

, Guan

, Hu

, Wang

(2013) Glycogen synthase kinase-3 reducesacetylcholine level in striatum via disturbing cellulardistribution of choline acetyltransferase in cholinergicinterneurons in rats. Neuroscience 255, 203–211.

56.

Wang

, Tian

, Liu

, Zhao

, Song

, Liu

, Ren

, Jiang

, Zeng

, Yang

, Wang

(2017) Activation of GSK-3 disrupts cholinergic homoeostasis in nucleus basalis of Meynert and frontal cortex of rats. J Cell Mol Med doi: 10.1111/jcmm.13262 .

57.

Bartus

(2000) On neurodegenerative diseases, models, and treatment strategies: Lessons learned and lessons forgotten a generation following the cholinergic hypothesis. Exp Neurol 163, 495–529.

58.

Fisher

(2007) M1 muscarinic agonists target major hallmarks of Alzheimer’s disease–an update. Curr Alzheimer Res 4, 577–580.

59.

Olds

, Alkon

(1993) Protein kinase C: A nexus in the biochemical events that underlie associative learning. Acta Neurobiol Exp (Wars) 53, 197–207.

60.

Pascale

, Govoni

, Battaini

(1998) Age-related alteration of PKC, a key enzyme in memory processes: Physiological and pathological examples. Mol Neurobiol 16, 49–62.

61.

Cole

, Dobkins

, Hansen

, Terry

, Saitoh

(1988) Decreased levels of protein kinase C in Alzheimer brain. Brain Res 452, 165–174.

62.

Wang

, Pisano

, Friedman

(1994) Attenuated protein kinase C activity and translocation in Alzheimer’s disease brain. Neurobiol Aging 15, 293–298.

63.

Yuan

, Sun

, Tan

, Yu

, Tan

(2017) The role of ADAM10 in Alzheimer’s disease. J Alzheimers Dis 58, 303–322.

64.

Marchesi

, Amadio

, Colombrita

, Govoni

, Ratti

, Pascale

(2016) PKC activation counteracts ADAM10 deficit in HuD-silenced neuroblastoma cells. J Alzheimers Dis 54, 535–547.

65.

Saraceno

, Marcello

, Di Marino

, Borroni

, Claeysen

, Perroy

, Padovani

, Tramontano

, Gardoni

, Di Luca

(2014) SAP97-mediated ADAM10 trafficking from Golgi outposts depends on PKC phosphorylation. Cell Death Dis 5, e1547.

66.

Etcheberrigaray

, Tan

, Dewachter

, Kuiperi

, Van der Auwera

, Wera

, Qiao

, Bank

, Nelson

, Kozikowski

, Van Leuven

, Alkon

(2004) Therapeutic effects of PKC activators in Alzheimer’s disease transgenic mice. Proc Natl Acad Sci U S A 101, 11141–11146.

67.

Donahue

, Johanson

(2008) Apolipoprotein E, amyloid-beta, and blood-brain barrier permeability in Alzheimer disease. J Neuropathol Exp Neurol 67, 261–270.

68.

De Ferrari

, Papassotiropoulos

, Biechele

, Wavrant De-Vrieze

, Avila

, Major

, Myers

, Sáez

, Henriquez

, Zhao

, Wollmer

, Nitsch

, Hock

, Morris

, Hardy

, Moon

(2007) Common genetic variation within the low-density lipoproteinreceptor-related protein 6 and late-onset Alzheimer’s disease. Proc Natl Acad Sci U S A 104, 9434–9439.

69.

Caruso

, Motolese

, Iacovelli

, Caraci

, Copani

, Nicoletti

, Terstappen

, Gaviraghi

, Caricasole

(2006) Inhibition of the canonical Wnt signaling pathway by apolipoprotein E4 in PC12 cells. J Neurochem 98, 364–371.

70.

Zhao

, Yue

, Zhu

, Sreejayan

, Du

(2010) AMP-activated protein kinase (AMPK) cross-talks with canonical Wnt signaling via phosphorylation of beta-catenin at Ser 552. Biochem Biophys Res Commun 395, 146–151.

71.

Greco

, Hamzelou

, Johnston

, Smith

, Ashford

, Tezapsidis

(2011) Leptin boosts cellular metabolism by activating AMPK and the sirtuins to reduce tau phosphorylation and β-amyloid in neurons. Biochem Biophys Res Commun 414, 170–174.

72.

Greco

, Sarkar

, Casadesus

, Zhu

, Smith

, Ashford

, Johnston

, Tezapsidis

(2009) Leptininhibits glycogen synthase kinase-3betato prevent tau phosphorylation in neuronal cells. NeurosciLett 455, 191–194.

73.

Greco

, Sarkar

, Johnston

, Tezapsidis

(2009) Leptinregulates tau phosphorylation and amyloid through AMPK in neuronalcells. Biochem Biophys Res Commun 380, 98–104.

74.

Greco

, Sarkar

, Johnston

, Zhu

, Su

, Casadesus

, Ashford

, Smith

, Tezapsidis

(2008) Leptin reduces Alzheimer’sdisease-related tau phosphorylation in neuronal cells. Biochem Biophys Res Commun 376, 536–541.

75.

Mairet-Coello

, Courchet

, Pieraut

, Courchet

, Maximov

, Polleux

(2013) The CAMKK2-AMPK kinase pathway mediates the synaptotoxic effects of Aβ oligomers through Tau phosphorylation. Neuron 78, 94–108.

76.

Lopez-Lopez

, Dietrich

, Metzger

, Loetscher

, Torres-Aleman

(2007) Disturbed cross talk between insulin-like growth factor I and AMP-activated protein kinase as a possible cause of vascular dysfunction in the amyloid precursor protein/presenilin 2 mouse model of Alzheimer’s disease. J Neurosci 27, 824–831.

77.

Pedrós

, Petrov

, Allgaier

, Sureda

, Barroso

, Beas-Zarate

, Auladell

, Pallàs

, Vázquez-Carrera

, Casadesús

, Folch

, Camins

(2014) Early alterations inenergy metabolism in the hippocampus of APPswe/PS1dE9 mouse modelof Alzheimer’s disease. Biochim Biophys Acta 1842, 1556–1566.