A Systematic Bioinformatics Workflow With Meta-Analytics Identified Potential Pathogenic Factors of Alzheimer’s Disease

Abstract

Potential pathogenic factors, other than well-known APP, APOE4, and PSEN, can be further identified from transcriptomics studies of differentially expressed genes (DEGs) that are specific for Alzheimer’s disease (AD), but findings are often inconsistent or even contradictory. Evidence corroboration by combining meta-analysis and bioinformatics methods may help to resolve existing inconsistencies and contradictions. This study aimed to demonstrate a systematic workflow for evidence synthesis of transcriptomic studies using both meta-analysis and bioinformatics methods to identify potential pathogenic factors. Transcriptomic data were assessed from GEO and ArrayExpress after systematic searches. The DEGs and their dysregulation states from both DNA microarray and RNA sequencing datasets were analyzed and corroborated by meta-analysis. Statistically significant DEGs were used for enrichment analysis based on KEGG and protein–protein interaction network (PPIN) analysis based on STRING. AD-specific modules were further determined by the DIAMOnD algorithm, which identifies significant connectivity patterns between specific disease-associated proteins and non-specific proteins. Within AD-specific modules, the nodes of highest degrees (>95th percentile) were considered as potential pathogenic factors. After systematic searches of 225 datasets, extensive meta-analyses among 25 datasets (21 DNA microarray datasets and 4 RNA sequencing datasets) identified 9,298 DEGs. The dysregulated genes and pathways in AD were associated with impaired amyloid-β (Aβ) clearance. From the AD-specific module, Fyn, and EGFR were the most statistically significant and biologically relevant. This meta-analytical study suggested that the reduced Aβ clearance in AD pathogenesis was associated with the genes encoding Fyn and EGFR, which were key receptors in Aβ downstream signaling.

Introduction

Alzheimer’s disease is a neurodegenerative disease that is the major cause of dementia worldwide (Alzheimer’s Association, 2009). The AD brain is characterized by the distribution of amyloid plaques and neurofibrillary tangles, which are composed of Aβ and hyperphosphorylated tau proteins, respectively (Wang and Mandelkow, 2016; Wang J. et al., 2017). The generation of Aβ from amyloid precursor protein (APP) is a central theme in the field of AD. APP is a single transmembrane protein with a large extracellular domain that is generated in large quantities and efficiently metabolized. APP can be processed via non-amyloidogenic or amyloidogenic generation of Aβ (Chow et al., 2010). Aβ mainly exists in two forms, Aβ₄₀ and a more hydrophobic Aβ₄₂, which consist of 40 and 42 amino acids, respectively (Marina et al., 2003). The Aβ monomer is intrinsically disordered, and present as a dynamic conformational structure (Chen et al., 2017), which allows the monomer binds to numerous substrate with a high binding affinity. The monomers of Aβ aggregate into oligomers, fibrils, and plaques, when a critical concentration of amyloid is reached due to dysfunctional homeostasis (Knowles et al., 2014). The soluble Aβ oligomer is recognized as the major neurotoxic species, and exists in an equilibrium with fibrils (Yang et al., 2017). And the toxicity of Aβ oligomers is inversely correlated with their oligomer size (Sengupta et al., 2016). Aβ is directly neurotoxic and induces a series of cellular responses; for example, it can induce excitotoxic signaling by increasing glutamate release (Brito-Moreira et al., 2011), and alter Ca²⁺ homeostasis and synaptic functions in neurons, astrocytes, and microglia (Parihar and Brewer, 2010). Aβ located on the mitochondria generates ROS, thus increasing oxidative stress and disrupting oxidative phosphorylation (Caspersen et al., 2005). Aβ also activates protein kinase to phosphorylate tau protein, resulting in the formation of neurofibrillary tangles (Hernández and Avila, 2008). The toxicity of both Aβ and hyperphosphorylated tau proteins are dependent on one another; an absence of tau protein reduces Aβ-induced memory impairment, while increased Aβ levels promote tau pathology (Oddo et al., 2006; Roberson et al., 2007).

The amyloid cascade hypothesis was proposed in 1992, and depicts Aβ as the causative factor in AD development (Hardy and Higgins, 1992). This hypothesis has genetic-based support, because mutations in genes related to Aβ formation (APP and PSEN) and clearance (APOE4) induce early-onset AD (Rovelet-Lecrux et al., 2006; Kline, 2012; St George-Hyslop and Fraser, 2012). However, this hypothesis is challenged by two facts, although there are also counterarguments that have been proposed from other scientific findings. First, amyloid imaging reveals that Aβ deposition also occurs in healthy aging subjects, suggesting that Aβ deposition could be a normal phenomenon of aging (Edison et al., 2007). However, Esparza et al. (2013) suggested that Aβ oligomer levels, which are the neurotoxic form of Aβ deposition, are higher in AD brains compared with healthy subjects. Second, the degree of dementia does not correlate well with amyloid plaque formation in AD (Nelson et al., 2012). In fact, Aβ oligomers may exert their neurotoxic effects much earlier, before the formation of amyloid plaques (Broersen et al., 2010). Also, the drug development for the treatment of AD has been guided by the amyloid cascade hypothesis for the pass two decade, although most anti-Aβ drugs fail in clinical trials and no novel drugs have been brought to market since 2003 (Cummings et al., 2014). For example, Solanezumab (currently in a phase 3 trial) recognizes Aβ at the 13 to 28 amino acid position and lowers amyloid levels, but does not show beneficial effects in subjects with mild AD (Honig et al., 2018). Verubecestat (currently in a phase 3 trial), a β-secretase inhibitor, reduces Aβ levels in a dose-dependent manner (Kennedy et al., 2016). However, a recent clinical trial (Egan et al., 2019) indicated that verubecestat does not have clinical effects in AD subjects. The failure of development of anti-Aβ drugs does not discredit amyloid as the central theme for AD treatment, because the sporadic AD is caused by the changes of genes which potentiate the neurotoxicity of Aβ in the brain over years, rather than solely Aβ overproduction. Indeed, the cerebral Aβ₄₂ kinetics were modified by Aβ accumulation (Potter et al., 2013), resulting in the disease preventative treatment required at least 95% lowering Aβ₄₂ production if the treatment was started after Aβ accumulation (Roberts et al., 2017). This suggests that the development of anti-Aβ drugs should not be solely dependent on inhibition of Aβ₄₂ production. However, the incomplete knowledge of AD pathogenic factors may hindered the investigation of AD (Scott et al., 2014), not just Aβ production, but also the Aβ-relevant receptors and their corresponding downstream signaling cascades.

Alzheimer’s disease is a complex disorder, expressed as a malfunction of defects in multiple genes. The discovery of AD pathogenic factors can be revealed by transcriptome profiling approach [i.e., DNA microarray and RNA-Seq (Sutherland et al., 2011)] to identify DEGs in AD subjects compared with healthy subjects. DEGs are genes that are potentially associated with disease pathology. Table 1 shows the publications using transcriptomic data and bioinformatics tools for identifying AD pathogenic factors. Puthiyedth et al. (2016) identified common DEGs from microarray datasets across six different brain regions. All datasets were integrated using the Colored (α, β)-κ Feature set approach (Puthiyedth et al., 2015), and a few common DEGs were identified across the brain regions, revealing an AD-specific signature. Kawalia et al. (2017) used the BC3Net10 algorithm to infer AD gene regulatory networks. This network was generated by the integration of literature-based knowledge and data-driven analysis. Inferences did not solely depend on the network topology, but also integrated information about DEGs and AD-related genes, after systematic searching from databases. However, the DEGs obtained from different transcriptomic studies of AD are sometimes conflicting because of many factors, such as the use of different experimental designs or statistical methods between studies. The integration of these studies based on meta-analysis therefore resolves inconclusive results and obtains a generalization (Haidich, 2010). The results of meta-analysis are presented as logORs based on the number of dysregulation events in both disease and control samples. The effect sizes are integrated to assess the overall effect size based on either fixed- or random-effect models. Meta-analysis has been applied in genomic analyses to assess the effectiveness of a particular gene in AD. Moradifard et al. (2018) conducted a meta-analysis using the R package RobustRankAggreg to identify statistically significant DEGs in AD across six microarray datasets. The DEGs were further corroborated with three RNA-Seq datasets and input for biological enrichment analysis to reveal the miRNAs that regulate the DEGs. Furthermore, Wang Q. et al. (2017) conducted a meta-analysis using the R package RankProd to explore the molecular mechanisms of AD across seven microarray datasets. They found 37 DEGs commonly shared by six brain regions, and most of these DEGs were downregulated. The significantly enriched pathways of DEGs were associated with mitochondrial oxidative phosphorylation and synaptic vesicle function. However, both studies only focused on a single microarray platform (Affymetrix), and the meta-analysis of DEGs did not include the gene expression profiles from RNA-Seq. For the study by Moradifard et al., the DEGs from RNA-Seq were only used to validate the results of microarray datasets; that is, the DEGs common to both the RNA-Seq and microarray datasets, in the same dysregulated direction. It seems that the results of currently available meta-analyses of DEGS for identifying AD pathogenic genes are limited by the use of microarray platforms and the inability to integrate results from both microarray and RNA-Seq.

This study aimed to identify potential pathogenic factors based on the meta-analysis of transcriptomic data from different brain regions in AD. The meta-analysis was conducted according to PRISMA guidelines (Moher et al., 2009), including those for dataset search, dataset selection, and statistical analysis. The raw transcriptomic data were retrieved from databases after a systematic search and were analyzed using a consistent workflow, from quality control and normalization to DEG determination. A statistical meta-analysis was performed when more than one study reported the same DEG. The DEGs that remained statistically significant after P-value adjustment were collected to perform a biological enrichment analysis and subgroup analysis, and to construct a PPIN. The network information and AD seed genes, which were obtained from the GWAS Catalog (Welter et al., 2014), were collected for the DIseAse Module Detection (DIAMoND) algorithm (Ghiassian et al., 2015) to examine the significance of the connectivity patterns among network components. This algorithm identified an AD-specific module from the network, and the nodes within the module with a high degree were treated as potential pathogenic factors. The overall study design is illustrated in Figure 1.

FIGURE 1

Materials and Methods

Database Search and Dataset Selection

The datasets were collected from GEO¹ (Barrett et al., 2012) and ArrayExpress² (Kolesnikov et al., 2015) from their respective inception dates until January 6, 2019. The search strategy is shown in Supplementary File S1.

Raw DNA microarray and RNA-Seq data reporting the transcriptomes of brain regions from AD and healthy subjects were included. The datasets were excluded if they were (1) not case–control studies; (2) focused on other neurodegenerative diseases; (3) demonstrated treatment effects on AD; or (4) used cell-line or artificial AD models. Sample details were collected if they were provided. If the datasets contained several comparisons, DEGs were determined from each comparison independently.

Data Processing

Each microarray dataset was processed consistently and independently using R packages from Bioconductor. Each dataset was processed through background correction, quantile normalization, and log₂-transformation of the averaged expression value of duplicate probes, using the affy (Gautier et al., 2004), oligo (Carvalho and Irizarry, 2010), or limma (Ritchie et al., 2015) package. For each dataset, the quality of each array was assessed using the arrayQualityMetrics (Kauffmann et al., 2009) package. The array was omitted from the subsequent analysis if it was detected as an outliner in any of the three metrics provided by the package. The gene expression ratios between AD and healthy subjects were determined for each dataset using the limma package. Genes with an adjusted P-value (by FDR) of less than 0.05 were considered to be statistically significant DEGs. The DEGs were annotated using the gene annotation table provided by GEO and ArrayExpress. The probes with missing gene symbols were removed. When multiple probes were annotated with the same gene symbol on the same array, the probe with the highest expression value was kept.

The raw sequence data were converted to FASTQ format using the SRA Toolkit (version 2.9.2) (Leinonen et al., 2011). Quality control on FASTQ was conducted using FastQC (version 0.11.8) (Andrews, 2010): (1) sequences with less than 80% of bases with at least quality of 25 were filtered; and (2) the bases with quality less than 25 were trimmed. The cleaner FASTQ files were aligned to the Homo.sapiens reference genome, GRCh38.94 (downloaded from Ensemble), using TopHat (version 2.1.1) (Kim et al., 2013). The aligned sequences were sorted based on the aligned position to the reference genome using SAMtools (version 1.9) (Li H. et al., 2009). After sorting, the count-based gene expression was obtained using HTSeq (version 0.8.0) (Anders et al., 2015), and the reads with alignment scores less than 10 were omitted. Differential expression was performed using the edgeR (Robinson et al., 2009) package for R software. RNA-Seq datasets were processed on a Linux-based HP ProLiant DL580 Gen8 workstation [Inter^® Xeon^® E7-4820 CPU v2 @ 2.00 GHz; 4 processors with 32 total cores enabled; 128 GB RAM].

Meta-Analysis, Biological Enrichment Analysis, and Subgroup Analysis

The list of DEGs and their dysregulation states (i.e., upregulated or downregulated) from each study were processed using the meta for Chung et al. (2015) package for R software. The meta-analysis was conducted under a random-effects model, and their outcomes were logORs with P-values. For each gene in the i-th study, the effect (θ_i) based on the numbers of dysregulation events in both AD and control samples was first calculated, then the overall effect was computed according to formula , where w_i is the weight and is equal to 1/v_i, where v_i is the sample variance. The genes with logORs above or below 0 were considered upregulated or downregulated, respectively. The P-values were adjusted by FDR, and DEGs with adjusted P-values less than 0.05 were regarded as statistically significant. After meta-analysis, the statistically significant DEGs were used to perform a biological enrichment analysis. The enrichment analysis was conducted by the hypergeometric test using the clusterProfiler (Yu et al., 2012) package for R software, based on the KEGG³ (Kanehisa et al., 2012). The statistically significant DEGs after meta-analysis were collected as input for the enrichment analysis to identify biological pathways among them. The pathways with FDR-adjusted P-values less than 0.05 were considered statistically significant.

The DEGs were split into different subgroups based on which brain regions they were from, to compare expression profiles among brain regions. According to Roth et al. (2006), brain regions can be hierarchically clustered into three major branches based on cytological differences. The first branch includes the cerebellum. The second branch is sub-categorized into three sub-branches; the first sub-branch includes the putamen and nucleus accumbens, the second sub-branch includes the amygdala and hippocampus, and the third sub-branch includes the cerebral neocortex. The third branch includes the thalamus, brain stem, and spinal cord. Biological enrichment analysis was also conducted in each subgroup.

Protein–Protein Interaction Network and Potential Pathogenic Factors

After meta-analysis, statistically significant DEGs were used to construct a PPIN based on the data from STRING (version 11⁴ ; Szklarczyk et al., 2011). Only interactions with the highest confidence (0.9) were kept. The PPIN was visualized using Cytoscape (Shannon et al., 2003). Disease-associated genes are not topologically interacted into a dense network module because disease can be the consequence of perturbation in many functional units. AD seed genes were retrieved from the GWAS Catalog (Welter et al., 2014)⁵ using the key word “AD” and the following selection criteria: a significance p-Value cutoff of no more than 1^∗10^–8. Genes with official gene symbols were kept as AD seed genes. The PPIN information (e.g., edge list in the co-expression network) and AD seed genes were put into the DIAMOnD algorithm (Ghiassian et al., 2015) to evaluate the significance of the connections that each node had with the seed genes in the biological network. In each iteration until the stopping condition was satisfied, the node with the lowest connectivity P-value was treated as the most significantly connected node for output. The nodes within the module that had a high degree (>95th percentile) were treated as potential pathogenic factors.

Risk of Bias

All studies were collected according to the publication information of datasets. If there was more than one publication for one dataset, the publication with the earlier publication year was selected. The risk of bias for the included studies was evaluated according to MIAME (Minimum information about a microarray experiment) (Brazma, 2009) and MINSEQE (Minimum information about a high-throughput nucleotide sequencing experiment⁶, proposed by FGED Society in 2012) guidelines by two authors (SY and HZ), independently. Disagreements between the authors were resolved by discussion with the third author (SWL). The MIAME and MINSEQE guidelines are designated to report the quality of DNA microarray and RNA-Seq studies, respectively. And the quality of transcriptomic parts of included studies was assessed, including raw data collection, result presentation, sample annotation, experimental design, array annotation, and data processing pipeline. Items with low risk were counted +1, suggesting high reproducibility; items with unclear risk were counted 0, suggesting ambiguous reproducibility; and items with high risk were counted –1, suggesting low reproducibility.

Results

Included Datasets

The selection process of the datasets is shown in Figure 2. A total of 225 datasets were initially identified from GEO and ArrayExpress. After removing duplicated datasets and datasets that met the exclusion criteria, 25 datasets (21 for microarray and 4 for RNA-Seq) met the eligibility criteria. The dataset characteristics are shown in Table 2. Among the included datasets, the hippocampus (n = 8) was the most extensively studied brain region, followed by the entorhinal cortex and prefrontal cortex (n = 3).

FIGURE 2

Preprocessing of Datasets and Determination of Differentially Expressed Genes

After the selection process of the datasets, all included datasets contained complete transcriptomic raw data, and were undergone a quality control as a prioritization procedure. In the included microarray datasets, the quality of the normalized dataset was assessed using the R package arrayQualityMetrics. Different numbers of arrays were identified as outliers and omitted in the subsequent analysis (Table 2). In the included RNA-Seq datasets, the reads were mapped to the Homo.sapiens reference genome (GRCh38.94) after sequence quality assessment. The mapping rates were above 95% for the RNA-Seq datasets. From the mapped reads, the genes were counted to determine DEGs. After preprocessing steps, different numbers of DEGs were determined from each comparison, ranging from 0 to over 10,000 (Table 3). There were 21,064 DEGs in all comparisons that compared AD brain samples with healthy brain samples, and 16,810 DEGs (79.76%) were reported in at least two comparisons (Supplementary File S2). In the meta-analysis of DEGs, 9,298 DEGs were found to be statistically significant (Supplementary File S3); 4,960 genes were downregulated and 4,338 genes were upregulated. The most reported downregulated DEGs were DPP6 and FXYD7, which were reported in 16 comparisons; RHOQ was the most reported upregulated DEG among all 15 reported comparisons. A loss of DPP6 or FXYD7 is reported to dysregulate neuronal excitation (Hoos et al., 2013; Cacace et al., 2019), while RHOQ is reported to enhance Aβ oligomerization (Aguilar et al., 2017). Multiple DEGs involved in Aβ clearance were found statistically significant in meta-analysis (Table 4). These DEGs could be categorized into UPS, autophagy, and UPR, implying the downregulated Aβ clearance. The statistically significant DEGs in UPS and autophagy were mostly downregulated, while those in UPR were mostly upregulated, proposing a dysfunctional homeostasis. The impaired Aβ clearance resulting from dysfunctional homeostasis is suggested to induce Aβ accumulation in sporadic AD, in contrast with the overproduction of Aβ in familial AD (Mawuenyega et al., 2010; Potter et al., 2013).

However, several genes involved in Aβ clearance were not found statistically significant in meta-analysis, although they were reported as DEGs in separated comparisons. The heat shock proteins work coordinately with UPS for protein clearance, remodeling the misfolded proteins before the degradation by UPS. Both systems are upregulated to restore dysfunctional homeostasis at the beginning (Kästle and Grune, 2012). However, the genes for most heat shock proteins, such as ATXN1, ATXN3, and ATXN7, were not statistically significant (Supplementary File S3), suggesting that the early event of AD might not be truly reflected. PICALM and SQSTM1 which are involved in autophagosome formation (Moreau et al., 2014; Ntsapi and Loos, 2016), were not reported as DEGs in the meta-analysis. Several autophagic marker genes, MAP1LC3B and ATG12, associated with autophagosomal function (Ma et al., 2010), were also found as a statistically insignificant in the meta-analysis. The expression of TREM2 was upregulated, which was contradictory to the role of TREM2 reported in autophagy (Lucin et al., 2013), but consistent with that reported in Aβ plaque-activated microglia (Yuan et al., 2016; Yin et al., 2017). Meanwhile, those genes involved in Aβ plaque-activated microglia, MS4A4A, PLCG2, and ABI3, were also found significantly upregulated in the meta-analysis.

On the enrichment analysis, AD itself was identified as significant (Figure 3A). In the AD pathway, components related to the mitochondrial respiratory chain were downregulated, while components related to Ca²⁺ channels that transport Ca²⁺ from the ER into the cytoplasm were upregulated (Figure 3B). Several AD-related pathways, including proteasome, oxidative phosphorylation, and retrograde endocannabinoid signaling pathways, were also identified as statistically significant; most components in these pathways were downregulated (Figures 3C–E). Downregulation of the proteasome decreases the clearance of Aβ (Hong et al., 2014), while downregulation of oxidative phosphorylation decreases the efficacy of the mitochondrial respiratory chain, suggesting hypometabolism in the AD brain (Mosconi et al., 2008). The genes (e.g., MFN1, OPA1, and DNM1L) involved in mitochondrial fusion and fission were downregulated (Westermann, 2012), disturbing mitochondrial biogenesis to adopt energetic demands. The aberrant mitochondria were removed by the process, mitophagy. However, the genes responsible for mitophagy, PINK1 and PRKN (Narendra et al., 2008), were downregulated, which further increased the burden of damaged protein clearance. The regulation of genes involved in oxidative phosphorylation in mitochondria was severely impaired in the present study, including genes for NADH dehydrogenase, succinate dehydrogenase, cytochrome c reductase, and cytochrome c oxidase in the electron transport chain. eCBs are signaling molecules used among nearby cells over short distances, and they modulate neuronal transmission. There is a high density of cannabinoid receptors in presynaptic terminals in the hippocampus and limbic cortex, indicating a role for cannabinoids in cognitive information processing (Zou and Kumar, 2018). A disruption of eCB signaling has been reported in AD, which influences neurotransmitter release (Basavarajappa et al., 2017). In the current study, most members of the retrograde endocannabinoid signaling pathway were downregulated, interfering with the role of eCBs in the release of glutamate and GABA in synapses.

FIGURE 3

Subgroup Analysis

Among 77 comparisons, 15 investigated the amygdala and hippocampus of AD subjects, while 28 focused on the cerebral cortex (Table 5). Statistically significant DEGs for each subgroup analysis are shown in Supplementary File S4. There were 2,260 DEGs that were reported as significant in the amygdala and hippocampus, while 6,636 DEGs were reported as significant in the cerebral cortex. Among the 1,390 commonly found DEGs, 359 upregulated and 1,020 downregulated DEGs were consistently found in two subgroups. Eleven DEGs were upregulated in one subgroup but downregulated in another subgroup. The most highly upregulated and downregulated DEGs were S100A6 and STMN4, respectively, in both subgroups. The protein calcyclin (encoded by S100A6) is a calcium-binding protein that is highly expressed in astrocytes surrounding Aβ deposits, and is involved in neuronal death via zinc depletion (Bartkowska et al., 2017). The protein stathmin (encoded by STMN4) binds tubulin to mediate microtubule dynamics, and dysregulation of stathmin-mediated microtubule stability induces memory loss (Uchida et al., 2014).

In the subgroup enrichment analyses, the pathways for proteasome and retrograde endocannabinoid signaling were found in both subgroups (Figures 4, 5). For the subgroup of the amygdala and hippocampus, the synaptic vesicle cycle was identified as statistically significant, and most components in this pathway were downregulated (Figure 4F). Downregulated synaptic functions decrease the release of neurotransmitters from neurons, inducing memory-related symptoms in AD. For the subgroup of the cerebral cortex, several synaptic signaling pathways were identified in AD, including the MAPK signaling pathway, dopaminergic synapses, and the insulin signaling pathway; the components in these pathways were mostly downregulated (Figures 5D–F). The role of the MAPK effector, JNK, is to dissociate Beclin 1 from Bcl-2 to mediate autophagy, and downregulation of JNK reduces autophagy ability (Zhou et al., 2015). The loss of dopaminergic synapses in the ventral tegmental area reduces the release of dopamine toward the cortex and hippocampus, and is reported to be associated with memory loss in AD before the formation of amyloid plaques (Nobili et al., 2017). Although all AD datasets collected in this study did not report suffering from T2DM, the dysfunctional insulin signaling was identified, suggesting an effect of Aβ in insulin signaling pathway. Aβ shares a common sequence with insulin, antagonizing the regulatory effects of insulin (Folch et al., 2018). In the canonical insulin signaling pathway, insulin binds receptors to activate Akt for mediating downstream effects, e.g., translocating glucose receptors to cellular membrane, activating mTOR for autophagy regulation, and inhibiting GSK3β for tau protein phosphorylation. Losing the regulatory effects of insulin may induce dysfunctional autophagy and tau protein hyperphosphorylation, potentiating AD development. The increased level of free insulin due to Aβ blockage reduces Aβ clearance via IDE, because insulin has a higher affinity to IDE than Aβ (Farris et al., 2003). On the other hand, the AD progression reduces the transport of insulin across BBB, decreasing insulin signaling in brain (Stanley et al., 2016).

FIGURE 4

FIGURE 5

In the present study, several datasets focused on the hippocampus and entorhinal cortex because these two regions suffer the most significant neuronal loss in the early stage of AD, while no significant neuronal loss is observed in these regions during normal aging. The transcriptome from these brain regions therefore represents the chronic response to the causes and consequences, or the treatment, of AD.

Quality Assessment of Studies

MIAME and MINSEQE guidelines were used to assess the transcriptomic analysis of studies (Blalock et al., 2004, 2011; Liang et al., 2007; Berchtold et al., 2008; Bronner et al., 2009; Williams et al., 2009; Nunez-Iglesias et al., 2010; Tan et al., 2010; Durrenberger et al., 2012; Antonell et al., 2013; Fischer et al., 2013; Janssen et al., 2013; Miller et al., 2013; Zhang et al., 2013; Hokama et al., 2014; Narayanan et al., 2014; Magistri et al., 2015; Scheckel et al., 2016; Wang et al., 2016; Lachen-Montes et al., 2017; Friedman et al., 2018; Nativio et al., 2018; Stopa et al., 2018) that published the datasets that were used in the present study. The results of quality assessments are shown in Figure 6. Among the 19 studies that published the microarray datasets, 80% did not provide the brain sample size or the weight used for RNA extraction for sample annotation and experimental variables, and only two studies provided sufficient information about experimental design, including quality control of samples. Furthermore, the annotations of array designs, including microarray quality indicators, were not fully reported for most studies (95%). For 40% of the studies, experimental and data processing details, including normalization methods and cut-offs for DEGs, were not fully reported. Among the four studies that published the RNA-Seq datasets, no study provided the brain sample size or the weight used for RNA extraction, or the details of experimental parameters used for RNA-Seq procedures. Furthermore, half of the studies did not provide sufficient information about experimental and data processing protocols, including the cut-offs for DEGs.

FIGURE 6

The relatively low quality of the studies indicates the insufficient information related to transcriptomic analysis provided in the studies, raising the reproducibility issue. Also, the transcriptomic results might be influenced, blurring some pathological events in the postmortem AD brains. The genes encoded heat shock proteins (e.g., ATXN1, ATXN3, and ATXN7) were not found statistically significant in the meta-analysis, while the multiple genes involved in neuroinflammation (e.g., CXCL3, IFNG, IL6, IL13, and CXCL9) (Garwood et al., 2011) were only reported as DEGs in one dataset, and therefore were not collected for meta-analysis. Although both heat shock proteins recruitment and neuroinflammation are associated with AD development, their pathological roles in AD development might not fully reflected in this study.

Construction of a Protein–Protein Interaction Network to Identify Potential Pathogenic Factors

The DEGs from the meta-analysis were used to retrieve the corresponding proteins in interactions according to STRING. The PPIN consisted of 4,781 nodes (2,083 upregulated nodes and 2,698 downregulated nodes) and 51,076 edges (Figure 7A) and there were 105 AD seed genes (Supplementary File S5) retrieved from the GWAS Catalog. The PPIN exhibited scale free topology with the degree distribution following a power law distribution (y = 3779.2x^–1.⁴⁸¹) and the most proteins were closely linked. In the PPIN, polyubiquitin-C (endcoded by UBC) and E3 ubiquitin-protein ligase RBX1 (encoded by RBX1), were the two proteins with the highest degree (372 and 222, respectively). These proteins were downregulated and are parts of UPS for protein degradation. The dysfunctional proteasome is regarded as an early event of AD (Manavalan et al., 2013), e.g., the decreased proteasomal clearance enhances mitochondrial dysfunction (Braun et al., 2015).

FIGURE 7

The PPIN for each protein pair was extracted for the subsequent DIAMOnD algorithm with AD seed genes to obtain a disease module. Not every seed gene was biologically linked to AD; some were involved in AD comorbidity disease. Among the disease modules (Figure 7B), the two nodes with a high degree (>95th percentile) were selected as potential pathogenic factors, and included EGFR and Fyn. The EGFR was the Aβ oligomer receptor, in which Aβ oligomer is regarded as the causative toxic species. The hydrophobic residues of Aβ oligomers are more easily accessible and interacting with cellular proteins, compared with Aβ fibrils (William et al., 2011). Fyn is involved in Aβ signal transduction and tau protein phosphorylation, and is changed in synaptic signaling pathways, including MAPK signaling pathway. These two proteins are both Aβ-relevant receptors, mediating the Aβ oligomers downstream neurotoxicity.

Discussion

We used a meta-analysis approach based on DNA microarray and RNA-Seq datasets from multiple brain regions to reveal potential pathogenic factors of AD. Perturbations in the genes involved in the proteasome, oxidative phosphorylation, and retrograde endocannabinoid signaling pathways were identified. The proteasome is required to degrade the damaged proteins to maintain the cellular functionality. The decreased proteasome activity results in an accumulation of ubiquitinated and damaged proteins (Ciechanover, 2015). The ubiquitination degradation of APP by the proteasome includes several key genes, such as PRKN, NEDD8, PSMC1, UCHL1, and UCHL3 (Bedford et al., 2008; Chen et al., 2012; Zhang et al., 2015; Nomura et al., 2016). Tau-specific ubiquitin–proteasome related genes, including UBE2D2 (Shimura et al., 2004) and STUB1 (Saidi et al., 2015), were also downregulated. The downregulation of these genes might result in impaired Aβ and phosphorylated tau protein clearance (Table 4). The oligomeric Aβ itself disrupts the catalytic activity of proteasome (Tseng et al., 2008), forming a positive feedback loop to worsen the Aβ clearance by proteasome. The disrupted catalytic activity of proteasome may also lose its inhibition of presenilin complex, which is the catalytic complex for γ-secretase, enhancing Aβ production (Chadwick et al., 2012). The accumulated Aβ, located on the mitochondria, deleteriously impairs oxidative phosphorylation, decreasing ATP production in the mitochondria (Chen and Du Yan, 2007; Moreira et al., 2010). The ATP-dependent proteins (i.e., the Ca²⁺ pumps located on the endoplasmic reticulum (ER) and ATP-dependent ion pumps) are unable to maintain their physiological functions, interfering neurotransmitter release and memory formation (Sudhof, 2012). The ATP-dependent molecular chaperones are also failed to work coordinately with UPS to remodel the misfolded proteins (Mattoo and Goloubinoff, 2014).

The reduced activity of proteasome enhances the burden of Aβ clearance on the autophagy and UPR. Autophagy is the process of degrading redundant cellular components by delivering them to lysosomes and forming autophagosomes. Beclin 1 (encoded by BECN1) is an essential component of autophagososomal structure (Jaeger and Wyss-Coray, 2010), but its autophagy function is inhibited by interacting with Bcl-2 (encoded by BCL2) leading to cellular death (Marquez and Xu, 2012). The downregulation of BECN1 and upregulation of BCL2 in the present study indicated autophagic flux failure. Autophagy is inhibited by the upregulation of mTOR and activated by the downregulation of TFEB (Caccamo et al., 2010; Polito et al., 2014). The downregulation of mTOR and upregulation of TFEB in the present study induced the formation of immature autophagosomes, impairing Aβ clearance. In addition, the downregulation of mTOR is associated with AD development (François et al., 2014), memory impairment (Slipczuk et al., 2009), and neuronal apoptosis (Yin et al., 2011). The depletion of autophagy-related genes, including Atg5 and Atg17, impairs the elongation of phagopores to form autophagosomes, resulting in Aβ accumulation (Hara et al., 2006; Nilsson et al., 2013). Accumulated Aβ on the ER membrane induces ER stress, which is relieved by the activation of UPR. The activation of UPR is induced by the accumulation of misfolded proteins (Stutzbach et al., 2013), and has been reported in AD subjects (Scheper and Hoozemans, 2015). Two main UPR proteins, including PERK (encoded by EIF2AK3) and IRE-1 (encoded by ERN1), were upregulated and activated in AD due to misfolded Aβ oligomers (Joshi et al., 2016), although a recent critical review suggested that the involvement of ER stress in AD might be exaggerated by the misuse of APP/PS1-overexpressing AD mouse models (Hashimoto and Saido, 2018). The activation of UPR is a significant approach to prevent the accumulation of misfolded proteins, but may also increase Aβ production (Scheper and Hoozemans, 2015), and result in synaptic loss (Gong et al., 2016). Taken together, disturbed Aβ clearance could be the upstream event for AD pathological development, to induce Aβ accumulation, altering intracellular ionic gradients and resulting in insufficient oxidative phosphorylation.

The two top pathogenic factors determined in the present study, Fyn and EGFR, are key receptors in Aβ downstream signaling. Fyn is a protein tyrosine kinase belonging to the Src family, and it mediates synaptic plasticity in the central nervous system (Kaufman et al., 2015). The soluble Aβ oligomer binds to PrP^C with high affinity (Laurén et al., 2009), and interrupts the PrP^C’s autoinhibitory mechanism (Wu et al., 2017), overactivating Fyn to increase glutamate release (Trepanier et al., 2012). Overactivation of Fyn, however, alters NMDA receptor function and intracellular Ca²⁺ homeostasis, rendering neurons vulnerable to Aβ-induced neurotoxicity. The Aβ-activated Fyn can phosphorylate downstream tau proteins at Tyr18 (Lee, 2004), and tau proteins participate in transporting Fyn to post-synaptic densities around glutamate receptors (Ittner et al., 2010). Fyn is a key protein that links between extracellular Aβ and intracellular tau protein, and unites these two pathologies in AD. The activity of Fyn is gene-dose-dependent (Um et al., 2012); that is, downregulation of FYN reduces Aβ-induced neurotoxicity, while upregulation of FYN exacerbates Aβ-induced neurotoxicity. In previous studies, overexpression of FYN has also been reported to accelerate synaptic loss, making Fyn inhibition a potential therapeutic treatment for AD (Chin et al., 2004; Chin, 2005). Saracatinib (AZD0530) is a Src kinase inhibitor that is being tested clinically for the treatment of AD. Saracatinib blocks Fyn activation to rescue memory deficits, and exhibits well-tolerated effects in mild-to-moderate AD subjects (Nygaard et al., 2015). A recently published clinical trial (VanDyck et al., 2019) revealed that saracatinib does not slow the decline in the CMRgl, and does not exhibit beneficial effects on several cognitive assessments, compared with the placebo group. However, a trend for slowing the reduction of volumetric measure in brain is observed, and a post hoc exploratory analysis deduces a statistically significant decline in CMRgl in the entorhinal cortex. Although the frequent adverse event of saracatinib, e.g., diarrhea, may limit the feasibility of increasing 125 mg daily to a higher dosage, the authors proposed that a higher dosage of saracatinib may be beneficial to slow decline of CMRgl in AD participants, who have greater tolerability.

EGFR is more commonly associated with cancer than with AD. However, the expression of EGFR is statistically correlated with the expression of γ-secretase, suggesting a significant role of EGFR in AD (Zhang et al., 2007). EGFR is widely distributed in the hippocampus and cerebral cortex (Fu et al., 2003), and EGFR signaling exerts neurotrophic functions to regulate synaptic architecture in the central nervous system (Oyagi et al., 2011). The expression of EGFR is reported to be upregulated in the brains of AD animals, and overactivation of EGFR by Aβ oligomers induces memory loss (Wang et al., 2012). The overactivation of EGFR can also be induced by the prolong UPR (Papa and Germain, 2011). The levels of overactivated EGFR can be suppressed by treatment with the EGFR inhibitor gefitnib to rescue memory loss. Activation of EGFR also induces downstream neuroinflammatory cascades in response to Aβ-induced neurotoxicity. Afatinib, an orally available EGFR tyrosine kinase inhibitor, inhibits EGFR activation to alleviate neuroinflammation by reducing caspase 1 activation and interleukin-1β levels in neurodegenerative diseases (Chen et al., 2019). Antagonizing Aβ-induced activation of EGFR may have beneficial effects to slow memory loss and alleviate neuroinflammation. Figure 8 summarizes the mechanism of action of Aβ-mediated Fyn and EGFR in AD, as described above.

FIGURE 8

Besides the genes involved in Aβ downstream neurotoxicity, the genes related to amyloid clearance proteins (e.g., ECE1 and NEU1) are reported as pathogenic factors of AD by other studies. Endothelin 1 (encoded by ECE1) is implicated in Aβ degradation and downregulation of ECE1 enhances Aβ deposition (Pacheco-Quinto and Eckman, 2013). However, upregulation of ECE1 was found in the present study, suggesting that the pathophysiological role of ECE1 was not solely involved in Aβ metabolism, but also associated with the increase Aβ-induced vasoconstriction in AD (Palmer et al., 2013). Sialidase (encoded by NEU1) is a lysosomal enzyme for Aβ clearance through lysosomal exocytosis (Annunziata et al., 2013). Deficit of NEU1 causes Aβ-induced proteasome inhibition, which was consistent with our results. Disruption of Aβ clearance proteins may be the secondary events of AD progression, rather than the primary cause. Increasing evidences show a correlation between T2DM and higher risk of developing AD, through increasing BBB permeability by chronic peripheral inflammation (Holmes et al., 2009). The BBB leakage was supported by the dysregulated genes in this study, including OCLN, TJP1, CDH5, and CTNNB1, which encode the tight junction proteins. The compromised BBB can be exacerbated by chronic hypertension (Santos et al., 2017) and microbial pathogens (Li H. et al., 2018), and alters peripheral immune cell trafficking to the brain and induces neuroinflammation. The neuroinflammation is mediated by p38 (encoded by MAPK14), which is a protein found in MAPK pathway. p38 is also a kinase to phosphorylate tau protein, making p38 as a potential pathogenic factor of AD. However, upregulation of p38 was not observed in this study, since p38 participates in numerous cellular events not only neuroinflammation. This can be supported by that although the inhibition of p38 alleviates neuroinflammation in AD (Yasuda et al., 2012; Ashabi et al., 2013), most p38 inhibitors are failed in clinical trials due to off-target effects. A gene converting glucose into lactate in glycolysis, LDHA, is regarded as a relevant pathogenic factor in AD, although it may not be specific to AD pathogenesis. Overexpression of LDHA is reported to prevent neurons from Aβ neurotoxicity by shifting mitochondrial glucose metabolism to lactate production, and decreasing ROS production (Newington et al., 2012). However, elevated lactate level may be detrimental to neurons (Harris et al., 2016). Although the expression of LDHA was insignificantly downregulated in this study, the opposing neuroprotective and neurotoxic effects of LDHA suggested that neurons might have its own defense mechanism against Aβ in the early onset of AD.

The strength of this study is the integration of transcriptomic data from both DNA microarrays and RNA-Seq using a systematic search and consistent workflow. However, there are some limitations. First, the articles that published the datasets did not provide full details of the experimental procedures, making the transcriptomic results less reliable and more difficult to repeat. In addition, the transcriptomic data retrieved from postmortem brain tissue using different methods might be biased. For example, laser-microdissection mainly targets neuronal soma, but may miss transcripts transporting to the pre- or post-synapses. Lacking of sufficient information about disposal of transcriptomic analysis, resulting in relatively low quality of the selected studies, might raise a reproducibility issue. Second, the confounding factors of postmortem brain tissues include age, sex, and postmortem interval, and may blur the results, e.g., neuroinflammation. Third, most included datasets in this study focused on hippocampus and entorhinal cortex, in which the brain region AD are believed to start. However, AD might start differently from those two brain regions, e.g., the impaired UPS as an early AD event was reported in brainstem (Irmler et al., 2012). Transcriptomic analysis in different regions of postmortem AD brains would provide more novel insights of AD development.

Conclusion

This meta-analytical study suggested that the reduced Aβ clearance in AD pathogenesis was associated with the genes encoding Fyn and EGFR, which were key receptors in Aβ downstream signaling.

References

• 1

  AguilarB. J.ZhuY.LuQ. (2017). Rho GTPases as therapeutic targets in Alzheimer’s disease.Alzheimers. Res. Ther.9:97. 10.1186/s13195-017-0320-4

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 2

  AndersS.PylP. T.HuberW. (2015). HTSeq–a Python framework to work with high-throughput sequencing data.Bioinformatics31166–169. 10.1093/bioinformatics/btu638

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 3

  AndrewsS. (2010). FastQC: A Quality Control Tool for High Throughput Sequence Data. Available online at: http://www.bioinformatics.babraham.ac.uk/projects/fastqc.

  • Google Scholar

• 4

  AnnunziataI.PattersonA.HeltonD.HuH.MoshiachS.GomeroE.et al (2013). Lysosomal NEU1 deficiency affects amyloid precursor protein levels and amyloid-β secretion via deregulated lysosomal exocytosis.Nat. Commun.4:2734. 10.1038/ncomms3734

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 5

  AntonellA.LladóA.AltirribaJ.Botta-OrfilaT.BalasaM.FernándezM.et al (2013). A preliminary study of the whole-genome expression profile of sporadic and monogenic early-onset Alzheimer’s disease.Neurobiol. Aging341772–1778. 10.1016/j.neurobiolaging.2012.12.026

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 6

  AshabiG.AlamdaryS. Z.RaminM.KhodagholiF. (2013). Reduction of hippocampal apoptosis by intracerebroventricular administration of extracellular signal-regulated protein kinase and/or p38 inhibitors in amyloid beta rat model of Alzheimer’s disease: involvement of nuclear-related factor-2 and nuclear factor-kB.Basic Clin. Pharmacol. Toxicol.112145–155. 10.1111/bcpt.12000

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 7

  Alzheimer’s Association, (2009). 2009 Alzheimer’s disease facts and figures.Alzheimers Dement.5234–270. 10.1016/j.jalz.2009.03.001

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 8

  BarrettT.WilhiteS. E.LedouxP.EvangelistaC.KimI. F.TomashevskyM.et al (2012). NCBI GEO: archive for functional genomics data sets—update.Nucleic Acids Res.41D991–D995. 10.1093/nar/gks1193

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 9

  BartkowskaK.SwiatekI.AniszewskaA.JurewiczE.TurlejskiK.FilipekA.et al (2017). Stress-dependent changes in the CacyBP/SIP interacting protein S100A6 in the mouse brain.PLoS One12:e0169760. 10.1371/journal.pone.0169760

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 10

  BasavarajappaB. S.ShivakumarM.JoshiV.SubbannaS. (2017). Endocannabinoid system in neurodegenerative disorders.J. Neurochem.142624–648. 10.1111/jnc.14098

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 11

  BedfordL.HayD.DevoyA.PaineS.PoweD. G.SethR.et al (2008). Depletion of 26S proteasomes in mouse brain neurons causes neurodegeneration and lewy-like inclusions resembling human pale bodies.J. Neurosci.288189–8198. 10.1523/JNEUROSCI.2218-08.2008

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 12

  BerchtoldN. C.CribbsD. H.ColemanP. D.RogersJ.HeadE.KimR.et al (2008). Gene expression changes in the course of normal brain aging are sexually dimorphic.Proc. Natl. Acad. Sci. U.S.A.10515605–15610. 10.1073/pnas.0806883105

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 13

  BlalockE. M.BuechelH. M.PopovicJ.GeddesJ. W.LandfieldP. W. (2011). Microarray analyses of laser-captured hippocampus reveal distinct gray and white matter signatures associated with incipient Alzheimer’s disease.J. Chem. Neuroanat.42118–126. 10.1016/j.jchemneu.2011.06.007

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 14

  BlalockE. M.GeddesJ. W.ChenK. C.PorterN. M.MarkesberyW. R.LandfieldP. W. (2004). Incipient Alzheimer’s disease: microarray correlation analyses reveal major transcriptional and tumor suppressor responses.Proc. Natl. Acad. Sci. U.S.A.1012173–2178. 10.1073/pnas.0308512100

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 15

  BraunR. J.SommerC.LeibigerC.GentierR. J. G.DumitI.PaduchK.et al (2015). Accumulation of basic amino acids at mitochondria dictates the cytotoxicity of aberrant ubiquitin.Cell Rep.101557–1571. 10.1016/j.celrep.2015.02.009

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 16

  BrazmaA. (2009). Minimum information about a microarray experiment (MIAME)–successes, failures, challenges.ScientificWorldJournal9420–423. 10.1100/tsw.2009.57

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 17

  Brito-MoreiraJ.Paula-LimaA. C.BomfimT. R.OliveiraF. B.SepulvedaF. J.De MelloF. G.et al (2011). Aβ oligomers induce glutamate release from hippocampal neurons.Curr. Alzheimer Res.8552–562. 10.2174/156720511796391917

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 18

  BroersenK.RousseauF.SchymkowitzJ. (2010). The culprit behind amyloid beta peptide related neurotoxicity in Alzheimer’s disease: oligomer size or conformation?Alzheimers. Res. Ther.2:12. 10.1186/alzrt36

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 19

  BronnerI. F.BochdanovitsZ.RizzuP.KamphorstW.RavidR.vanSwietenJ. C.et al (2009). Comprehensive mRNA expression profiling distinguishes tauopathies and identifies shared molecular pathways.PLoS One4:e6826. 10.1371/journal.pone.0006826

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 20

  BrowneF.WangH.ZhengH. (2015). A computational framework for the prioritization of disease-gene candidates.BMC Genomics16(Suppl. 9):S2. 10.1186/1471-2164-16-S9-S2

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 21

  CacaceR.HeemanB.VanMosseveldeS.DeRoeckA.HoogmartensJ.DeRijkP.et al (2019). Loss of DPP6 in neurodegenerative dementia: a genetic player in the dysfunction of neuronal excitability.Acta Neuropathol.137901–918. 10.1007/s00401-019-01976-3

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 22

  CaccamoA.MajumderS.RichardsonA.StrongR.OddoS. (2010). Molecular interplay between mammalian target of rapamycin (mTOR), amyloid-β, and tau.J. Biol. Chem.28513107–13120. 10.1074/jbc.M110.100420

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 23

  CarvalhoB. S.IrizarryR. A. (2010). A framework for oligonucleotide microarray preprocessing.Bioinformatics262363–2367. 10.1093/bioinformatics/btq431

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 24

  CaspersenC.WangN.YaoJ.SosunovA.ChenX.LustbaderJ. W.et al (2005). Mitochondrial Aβ: a potential focal point for neuronal metabolic dysfunction in Alzheimer’s disease.FASEB J.192040–2041. 10.1096/fj.05-3735fje

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 25

  ChadwickL.GentleL.StrachanJ.LayfieldR. (2012). Review: Unchained maladie – a reassessment of the role of Ubb +1-capped polyubiquitin chains in Alzheimer’s disease.Neuropathol. Appl. Neurobiol.38118–131. 10.1111/j.1365-2990.2011.01236.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 26

  ChandrasekaranS.BonchevD. (2016). Network topology analysis of post-mortem brain microarrays identifies more Alzheimer’s related genes and micrornas and points to novel routes for fighting with the disease.PLoS One11:e0144052. 10.1371/journal.pone.0144052

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 27

  ChenG. F.XuT. H.YanY.ZhouY. R.JiangY.MelcherK.et al (2017). Amyloid beta: structure, biology and structure-based therapeutic development.Acta Pharmacol. Sin.381205–1235. 10.1038/aps.2017.28

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 28

  ChenJ. X.Du YanS. (2007). Amyloid-β-induced mitochondrial dysfunction.J. Alzheimer’s Dis.12177–184. 10.3233/JAD-2007-12208

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 29

  ChenY.NeveR. L.LiuH. (2012). Neddylation dysfunction in Alzheimer’s disease.J. Cell. Mol. Med.162583–2591. 10.1111/j.1582-4934.2012.01604.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 30

  ChenY.-J.HsuC.-C.ShiaoY.-J.WangH.-T.LoY.-L.LinA. M. Y. (2019). Anti-inflammatory effect of afatinib (an EGFR-TKI) on OGD-induced neuroinflammation.Sci. Rep.9:2516. 10.1038/s41598-019-38676-7

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 31

  ChinJ. (2005). Fyn kinase induces synaptic and cognitive impairments in a transgenic mouse model of Alzheimer’s disease.J. Neurosci.259694–9703. 10.1523/JNEUROSCI.2980-05.2005

  • CrossRef
  • Google Scholar

• 32

  ChinJ.PalopJ. J.YuG. Q.KojimaN.MasliahE.MuckeL. (2004). Fyn kinase modulates synaptotoxicity, but not aberrant sprouting, in human amyloid precursor protein transgenic mice.J. Neurosci.244692–4697. 10.1523/JNEUROSCI.0277-04.2004

  • CrossRef
  • Google Scholar

• 33

  ChowV. W.MattsonM. P.WongP. C.GleichmannM. (2010). An overview of APP processing enzymes and products.Neuromolecular Med.121–12. 10.1007/s12017-009-8104-z

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 34

  ChungS. S.SunK.BolechC. J. (2015). Matrix product ansatz for Fermi fields in one dimension.Phys. Rev. B91:121108. 10.1103/PhysRevB.91.121108

  • CrossRef
  • Google Scholar

• 35

  CiechanoverA. (2015). The unravelling of the ubiquitin system.Nat. Rev. Mol. Cell Biol.16322–324. 10.1038/nrm3982

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 36

  CummingsJ. L.MorstorfT.ZhongK. (2014). Alzheimer’s disease drug-development pipeline: few candidates, frequent failures.Alzheimers. Res. Ther.6:37. 10.1186/alzrt269

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 37

  DurrenbergerP. F.FernandoF. S.MagliozziR.KashefiS. N.BonnertT. P.FerrerI.et al (2012). Selection of novel reference genes for use in the human central nervous system: a brainnet europe study.Acta Neuropathol.124893–903. 10.1007/s00401-012-1027-z

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 38

  EdisonP.ArcherH. A.HinzR.HammersA.PaveseN.TaiY. F.et al (2007). Amyloid, hypometabolism, and cognition in Alzheimer disease: an [11C]PIB and [18F]FDG PET study.Neurology68501–508. 10.1212/01.wnl.0000244749.20056.d4

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 39

  EganM. F.KostJ.VossT.MukaiY.AisenP. S.CummingsJ. L.et al (2019). Randomized trial of verubecestat for prodromal Alzheimer’s disease.N. Engl. J. Med.3801408–1420. 10.1056/NEJMoa1812840

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 40

  EsparzaT. J.ZhaoH.CirritoJ. R.CairnsN. J.BatemanR. J.HoltzmanD. M.et al (2013). Amyloid-beta oligomerization in Alzheimer dementia versus high-pathology controls.Ann. Neurol.73104–119. 10.1002/ana.23748

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 41

  FarrisW.MansourianS.ChangY.LindsleyL.EckmanE. A.FroschM. P.et al (2003). Insulin-degrading enzyme regulates the levels of insulin, amyloid -protein, and the -amyloid precursor protein intracellular domain in vivo.Proc. Natl. Acad. Sci. U.S.A.1004162–4167. 10.1073/pnas.0230450100

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 42

  FischerM. T.WimmerI.HöftbergerR.GerlachS.HaiderL.ZrzavyT.et al (2013). Disease-specific molecular events in cortical multiple sclerosis lesions.Brain1361799–1815. 10.1093/brain/awt110

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 43

  FolchJ.EttchetoM.BusquetsO.Sánchez-LópezE.Castro-TorresR. D.VerdaguerE.et al (2018). The implication of the brain insulin receptor in late onset Alzheimer’s disease dementia.Pharmaceuticals11:11. 10.3390/ph11010011

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 44

  FrançoisA.Rioux BilanA.QuellardN.FernandezB.JanetT.ChassaingD.et al (2014). Longitudinal follow-up of autophagy and inflammation in brain of APPswePS1dE9 transgenic mice.J. Neuroinflammation11:139. 10.1186/s12974-014-0139-x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 45

  FriedmanB. A.SrinivasanK.AyalonG.MeilandtW. J.LinH.HuntleyM. A.et al (2018). Diverse brain myeloid expression profiles reveal distinct microglial activation states and aspects of Alzheimer’s disease not evident in mouse models.Cell Rep.22832–847. 10.1016/j.celrep.2017.12.066

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 46

  FuL.Abu-KhalilA.MorrisonR. S.GeschwindD. H.KornblumI. (2003). Expression patterns of epidermal growth factor receptor and fibroblast growth factor receptor 1 mRNA in fetal human brain.J. Comp. Neurol.462265–273. 10.1002/cne.10727

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 47

  GarwoodC. J.PoolerA. M.AthertonJ.HangerD. P.NobleW. (2011). Astrocytes are important mediators of Aβ-induced neurotoxicity and tau phosphorylation in primary culture.Cell Death. Dis.2:e167. 10.1038/cddis.2011.50

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 48

  GautierL.CopeL.BolstadB. M.IrizarryR. A. (2004). Affy – Analysis of affymetrix genechip data at the probe level.Bioinformatics20307–315. 10.1093/bioinformatics/btg405

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 49

  GhiassianS. D.MencheJ.BarabásiA. L. (2015). A DIseAse MOdule detection (DIAMOnD) algorithm derived from a systematic analysis of connectivity patterns of disease proteins in the human interactome.PLoS Comput. Biol.11:e1004120. 10.1371/journal.pcbi.1004120

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 50

  GongB.RadulovicM.Figueiredo-PereiraM. E.CardozoC. (2016). The ubiquitin-proteasome system: Potential therapeutic targets for Alzheimer’s disease and spinal cord injury.Front. Mol. Neurosci.9:4. 10.3389/fnmol.2016.00004

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 51

  HaidichA. B. (2010). Meta-analysis in medical research.Hippokratia1429–37. 10.5005/jp/books/10519

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 52

  HaraT.NakamuraK.MatsuiM.YamamotoA.NakaharaY.Suzuki-MigishimaR.et al (2006). Suppression of basal autophagy in neural cells causes neurodegenerative disease in mice.Nature441885–889. 10.1038/nature04724

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 53

  HardyJ.HigginsG. (1992). Alzheimer’s disease: the amyloid cascade hypothesis.Science256184–185. 10.1126/science.1566067

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 54

  HarrisR. A.TindaleL.LoneA.SinghO.MacauleyS. L.StanleyM.et al (2016). Aerobic glycolysis in the frontal cortex correlates with memory performance in wild-type mice but not the APP/PS1 mouse model of cerebral amyloidosis.J. Neurosci.361871–1878. 10.1523/JNEUROSCI.3131-15.2016

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 55

  HashimotoS.SaidoT. C. (2018). Critical review: involvement of endoplasmic reticulum stress in the aetiology of Alzheimer’s disease.Open Biol.8:180024. 10.1098/rsob.180024

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 56

  HernándezF.AvilaJ. (2008). The role of glycogen synthase kinase 3 in the early stages of Alzheimers’ disease.FEBS Lett.5823848–3854. 10.1016/j.febslet.2008.10.026

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 57

  HokamaM.OkaS.LeonJ.NinomiyaT.HondaH.SasakiK.et al (2014). Altered expression of diabetes-related genes in Alzheimer’s disease brains: the hisayama study.Cereb. Cortex242476–2488. 10.1093/cercor/bht101

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 58

  HolmesC.CunninghamC.ZotovaE.WoolfordJ.DeanC.KerrS.et al (2009). Systemic inflammation and disease progression in alzheimer disease.Neurology73768–774. 10.1212/WNL.0b013e3181b6bb95

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 59

  HongL.HuangH.-C.JiangZ.-F. (2014). Relationship between amyloid-beta and the ubiquitin–proteasome system in Alzheimer’s disease.Neurol. Res.36276–282. 10.1179/1743132813Y.0000000288

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 60

  HonigL. S.VellasB.WoodwardM.BoadaM.BullockR.BorrieM.et al (2018). Trial of solanezumab for mild dementia due to Alzheimer’s disease.N. Engl. J. Med.378321–330. 10.1056/NEJMoa1705971

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 61

  HoosM. D.RichardsonB. M.FosterM. W.EverhartA.ThompsonJ. W.MoseleyM. A.et al (2013). Longitudinal study of differential protein expression in an Alzheimer’s mouse model lacking inducible nitric oxide synthase.J. Proteome Res.124462–4477. 10.1021/pr4005103

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 62

  IrmlerM.GentierR. J. G.DennissenF. J. A.SchulzH.BolleI.HölterS. M.et al (2012). Long-term proteasomal inhibition in transgenic mice by UBB+1 expression results in dysfunction of central respiration control reminiscent of brainstem neuropathology in Alzheimer patients.Acta Neuropathol.124187–197. 10.1007/s00401-012-1003-7

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 63

  IttnerL. M.KeY. D.DelerueF.BiM.GladbachA.vanEerselJ.et al (2010). Dendritic function of tau mediates amyloid-β toxicity in alzheimer’s disease mouse models.Cell142387–397. 10.1016/j.cell.2010.06.036

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 64

  JaegerP. A.Wyss-CorayT. (2010). Beclin 1 complex in autophagy and Alzheimer disease.Arch. Neurol.671181–1184. 10.1001/archneurol.2010.258

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 65

  JanssenS. F.van derSpekS. J. F.TenbrinkJ. B.EssingA. H. W.GorgelsT. G.van derSpekP. J.et al (2013). Gene expression and functional annotation of the human and mouse choroid plexus epithelium.PLoS One8:e83345. 10.1371/journal.pone.0083345

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 66

  JoshiA. U.KornfeldO. S.Mochly-RosenD. (2016). The entangled ER-mitochondrial axis as a potential therapeutic strategy in neurodegeneration: a tangled duo unchained.Cell Calcium60218–234. 10.1016/j.ceca.2016.04.010

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 67

  KanehisaM.GotoS.SatoY.FurumichiM.TanabeM. (2012). KEGG for integration and interpretation of large-scale molecular data sets.Nucleic Acids Res.40D109–D114. 10.1093/nar/gkr988

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 68

  KästleM.GruneT. (2012). Interactions of the proteasomal system with chaperones: Protein triage and protein quality control.Prog. Mol. Biol. Transl. Sci.109113–160. 10.1016/B978-0-12-397863-9.00004-3

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 69

  KauffmannA.GentlemanR.HuberW. (2009). arrayQualityMetrics—a bioconductor package for quality assessment of microarray data.Bioinformatics25415–416. 10.1093/bioinformatics/btn647

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 70

  KaufmanA. C.SalazarS. V.HaasL. T.YangJ.KostylevM. A.JengA. T.et al (2015). Fyn inhibition rescues established memory and synapse loss in Alzheimer mice.Ann. Neurol.77953–971. 10.1002/ana.24394

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 71

  KawaliaS. B.RaschkaT.NazM.deMatos SimoesR.SengerP.Hofmann-ApitiusM. (2017). Analytical strategy to prioritize Alzheimer’s disease candidate genes in gene regulatory networks using public expression data.J. Alzheimers Dis.591237–1254. 10.3233/JAD-170011

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 72

  KennedyM. E.StamfordA. W.ChenX.CoxK.CummingJ. N.DockendorfM. F.et al (2016). The BACE1 inhibitor verubecestat (MK-8931) reduces CNS -amyloid in animal models and in Alzheimers disease patients.Sci. Transl. Med.8:363ra150. 10.1126/scitranslmed.aad9704

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 73

  KikuchiM.OgishimaS.MiyamotoT.MiyashitaA.KuwanoR.NakayaJ.et al (2013). Identification of unstable network modules reveals disease modules associated with the progression of Alzheimer’s disease.PLoS One8:e76162. 10.1371/journal.pone.0076162

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 74

  KimD.PerteaG.TrapnellC.PimentelH.KelleyR.SalzbergS. L. (2013). TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions.Genome Biol.14:R36. 10.1186/gb-2013-14-4-r36

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 75

  KlineA. (2012). Apolipoprotein E, amyloid-ß clearance and therapeutic opportunities in Alzheimer’s disease.Alzheimers. Res. Ther.4:32. 10.1186/alzrt135

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 76

  KnowlesT. P. J.VendruscoloM.DobsonC. M. (2014). The amyloid state and its association with protein misfolding diseases.Nat. Rev. Mol. Cell Biol.15384–396. 10.1038/nrm3810

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 77

  KolesnikovN.HastingsE.KeaysM.MelnichukO.TangY. A.WilliamsE.et al (2015). ArrayExpress update–simplifying data submissions.Nucleic Acids Res.43D1113–D1116. 10.1093/nar/gku1057

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 78

  Lachen-MontesM.ZelayaM. V.SeguraV.Fernández-IrigoyenJ.SantamaríaE. (2017). Progressive modulation of the human olfactory bulb transcriptome during Alzheimer’s disease evolution: novel insights into the olfactory signaling across proteinopathies.Oncotarget869663–69679. 10.18632/oncotarget.18193

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 79

  LaurénJ.GimbelD. A.NygaardH. B.GilbertJ. W.StrittmatterS. M. (2009). Cellular prion protein mediates impairment of synaptic plasticity by amyloid-β oligomers.Nature4571128–1132. 10.1038/nature07761

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 80

  LeeG. (2004). Phosphorylation of Tau by Fyn: implications for Alzheimer’s Disease.J. Neurosci.242304–2312. 10.1523/JNEUROSCI.4162-03.2004

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 81

  LeinonenR.SugawaraH.ShumwayM. (2011). The sequence read archive.Nucleic Acids Res.39D19–D21. 10.1093/nar/gkq1019

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 82

  LiH.HandsakerB.WysokerA.FennellT.RuanJ.HomerN.et al (2009). The sequence alignment/map format and SAMtools.Bioinformatics252078–2079. 10.1093/bioinformatics/btp352

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 83

  LiH.LiuC. C.ZhengH.HuangT. Y. (2018). Amyloid, tau, pathogen infection and antimicrobial protection in Alzheimer’s disease -conformist, nonconformist, and realistic prospects for AD pathogenesis.Transl. Neurodegener.7:34. 10.1186/s40035-018-0139-3

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 84

  LiX.WangH.LongJ.PanG.HeT.AnichtchikO.et al (2018). Systematic Analysis and Biomarker Study for Alzheimer’s Disease.Sci. Rep.8:17394. 10.1038/s41598-018-35789-3

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 85

  LiangW. S.DunckleyT.BeachT. G.GroverA.MastroeniD.WalkerD. G.et al (2007). Gene expression profiles in anatomically and functionally distinct regions of the normal aged human brain.Physiol. Genomics28311–322. 10.1152/physiolgenomics.00208.2006

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 86

  LucinK. M.O’BrienC. E.BieriG.CzirrE.MosherI.AbbeyR. J.et al (2013). Microglial beclin 1 regulates retromer trafficking and phagocytosis and is impaired in Alzheimer’s disease.Neuron79873–886. 10.1016/j.neuron.2013.06.046

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 87

  MaJ. F.HuangY.ChenS. D.HallidayG. (2010). Immunohistochemical evidence for macroautophagy in neurones and endothelial cells in Alzheimer’s disease.Neuropathol. Appl. Neurobiol.36312–319. 10.1111/j.1365-2990.2010.01067.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 88

  MagistriM.VelmeshevD.MakhmutovaM.FaghihiM. A. (2015). Transcriptomics profiling of Alzheimer’s disease reveal neurovascular defects, altered amyloid-β homeostasis, and deregulated expression of long noncoding RNAs.J. Alzheimers Dis.48647–665. 10.3233/JAD-150398

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 89

  ManavalanA.MishraM.FengL.SzeS. K.AkatsuH.HeeseK. (2013). Brain site-specific proteome changes in aging-related dementia.45:e39. 10.1038/emm.2013.76

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 90

  MarinaG. B.KirkitadzeD.LomakinA.VollersS. S.BenedekG. B.TeplowD. B. (2003). Amyloid β-protein (Aβ) assembly: Aβ40 and Aβ42 oligomerize through distinct pathways.Proc. Natl. Acad. Sci. U.S.A.100330–335. 10.1073/pnas.222681699

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 91

  MarquezR. T.XuL. (2012). Bcl-2:Beclin 1 complex: multiple, mechanisms regulating autophagy/apoptosis toggle switch.Am. J. Cancer Res.2214–221.

  • Pubmed Abstract
  • Google Scholar

• 92

  MattooR. U. H.GoloubinoffP. (2014). Molecular chaperones are nanomachines that catalytically unfold misfolded and alternatively folded proteins.Cell. Mol. Life Sci.713311–3325. 10.1007/s00018-014-1627-y

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 93

  MawuenyegaK. G.SigurdsonW.OvodV.MunsellL.KastenT.MorrisJ. C.et al (2010). Decreased clearance of CNS β-amyloid in Alzheimer’s disease.Science330:1774. 10.1126/science.1197623

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 94

  MengG.ZhongX.MeiH. (2016). A systematic investigation into Aging related genes in brain and their relationship with Alzheimer’s disease.PLoS One11:e0150624. 10.1371/journal.pone.0150624

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 95

  MillerJ. A.WoltjerR. L.GoodenbourJ. M.HorvathS.GeschwindD. H. (2013). Genes and pathways underlying regional and cell type changes in Alzheimer’s disease.Genome Med.5:48. 10.1186/gm452

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 96

  MoherD.LiberatiA.TetzlaffJ.AltmanD. G.Prisma Group. (2009). Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement.PLoS Med.6:e1000097. 10.1371/journal.pmed.1000097

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 97

  MoradifardS.HoseinbeykiM.GanjiS. M.MinuchehrZ. (2018). Analysis of microRNA and gene expression profiles in Alzheimer’s disease: a meta-analysis approach.Sci. Rep.8:4767. 10.1038/s41598-018-20959-0

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 98

  MoreauK.FlemingA.ImarisioS.Lopez RamirezA.MercerJ. L.Jimenez-SanchezM.et al (2014). PICALM modulates autophagy activity and tau accumulation.Nat. Commun.5:4998. 10.1038/ncomms5998

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 99

  MoreiraP. I.CarvalhoC.ZhuX.SmithM. A.PerryG. (2010). Mitochondrial dysfunction is a trigger of Alzheimer’s disease pathophysiology.Biochim. Biophys. Acta Mol. Basis Dis.18022–10. 10.1016/j.bbadis.2009.10.006

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 100

  MosconiL.PupiA.DeLeonM. J. (2008). Brain glucose hypometabolism and oxidative stress in preclinical Alzheimer’s disease.Ann. N. Y. Acad. Sci.1147180–195. 10.1196/annals.1427.007

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 101

  NarayananM.HuynhJ. L.WangK.YangX.YooS.McElweeJ.et al (2014). Common dysregulation network in the human prefrontal cortex underlies two neurodegenerative diseases.Mol. Syst. Biol.10:743. 10.15252/msb.20145304

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 102

  NarendraD.TanakaA.SuenD. F.YouleR. J. (2008). Parkin is recruited selectively to impaired mitochondria and promotes their autophagy.J. Cell Biol.183795–803. 10.1083/jcb.200809125

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 103

  NativioR.DonahueG.BersonA.LanY.Amlie-WolfA.TuzerF.et al (2018). Publisher correction: dysregulation of the epigenetic landscape of normal aging in Alzheimer’s disease.Nat. Neurosci.211018–1018. 10.1038/s41593-018-0124-2

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 104

  NelsonP. T.AlafuzoffI.BigioE. H.BourasC.BraakH.CairnsN. J.et al (2012). Correlation of Alzheimer disease neuropathologic changes with cognitive status: a review of the literature.J. Neuropathol. Exp. Neurol.71362–381. 10.1097/NEN.0b013e31825018f7

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 105

  NewingtonJ. T.RapponT.AlbersS.WongD. Y.RylettR. J.CummingR. C. (2012). Overexpression of pyruvate dehydrogenase kinase 1 and lactate dehydrogenase A in nerve cells confers resistance to amyloid β and other toxins by decreasing mitochondrial respiration and reactive oxygen species production.J. Biol. Chem.28737245–37258. 10.1074/jbc.M112.366195

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 106

  NilssonP.LoganathanK.SekiguchiM.MatsubaY.HuiK.TsubukiS.et al (2013). Aβ secretion and plaque formation depend on autophagy.Cell Rep.561–69. 10.1016/j.celrep.2013.08.042

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 107

  NobiliA.LatagliataE. C.ViscomiM. T.CavallucciV.CutuliD.GiacovazzoG.et al (2017). Dopamine neuronal loss contributes to memory and reward dysfunction in a model of Alzheimer’s disease.Nat. Commun.8:14727. 10.1038/ncomms14727

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 108

  NomuraJ.HosoiT.KanekoM.OzawaK.NishiA.NomuraY. (2016). Neuroprotection by endoplasmic reticulum stress-induced HRD1 and chaperones: possible therapeutic targets for Alzheimer’s and Parkinson’s disease.Med. Sci4:14. 10.3390/medsci4030014

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 109

  NtsapiC.LoosB. (2016). Caloric restriction and the precision-control of autophagy: a strategy for delaying neurodegenerative disease progression.Exp. Gerontol.8397–111. 10.1016/j.exger.2016.07.014

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 110

  Nunez-IglesiasJ.LiuC.-C.MorganT. E.FinchC. E.ZhouX. J. (2010). Joint genome-wide profiling of miRNA and mRNA expression in Alzheimer’s disease cortex reveals altered miRNA regulation.PLoS One5:e8898. 10.1371/journal.pone.0008898

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 111

  NygaardH. B.WagnerA. F.BowenG. S.GoodS. P.MacAvoyM. G.StrittmatterK. A.et al (2015). A phase Ib multiple ascending dose study of the safety, tolerability, and central nervous system availability of AZD0530 (saracatinib) in Alzheimer’s disease.Alzheimers. Res. Ther.7:35. 10.1186/s13195-015-0119-0

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 112

  OddoS.VasilevkoV.CaccamoA.KitazawaM.CribbsD. H.LaFerlaF. M. (2006). Reduction of soluble Aβ and Tau, but not soluble Aβ Alone, ameliorates cognitive decline in transgenic mice with plaques and tangles.J. Biol. Chem.28139413–39423. 10.1074/jbc.M608485200

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 113

  OyagiA.MoriguchiS.NittaA.MurataK.OidaY.TsurumaK.et al (2011). Heparin-binding EGF-like growth factor is required for synaptic plasticity and memory formation.Brain Res.141997–104. 10.1016/j.brainres.2011.09.003

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 114

  Pacheco-QuintoJ.EckmanE. A. (2013). Endothelin-converting enzymes degrade intracellular β-amyloid produced within the endosomal/lysosomal pathway and autophagosomes.J. Biol. Chem.2885606–5615. 10.1074/jbc.M112.422964

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 115

  PalmerJ. C.TaylerH. M.LoveS. (2013). Endothelin-converting enzyme-1 activity, endothelin-1 production, and free radical-dependent vasoconstriction in Alzheimer’s disease.J. Alzheimers Dis.36577–587. 10.3233/JAD-130383

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 116

  PangX.ZhaoY.WangJ.ZhouQ.XuL.KangD.et al (2017). The Bioinformatic analysis of the dysregulated genes and MicroRNAs in entorhinal cortex, hippocampus, and blood for Alzheimer’s disease.Biomed Res. Int.2017:9084507. 10.1155/2017/9084507

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 117

  PapaL.GermainD. (2011). Estrogen receptor mediates a distinct mitochondrial unfolded protein response.J. Cell Sci.124 (Pt 9)1396–1402. 10.1242/jcs.078220

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 118

  PariharM. S.BrewerG. J. (2010). Amyloid-β as a modulator of synaptic plasticity.J. Alzheimers Dis.22741–763. 10.3233/JAD-2010-101020

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 119

  PolitoV. A.LiH.Martini−StoicaH.WangB.YangL.XuY.et al (2014). Selective clearance of aberrant tau proteins and rescue of neurotoxicity by transcription factor EB.EMBO Mol. Med.61142–1160. 10.15252/emmm.201303671

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 120

  PotterR.PattersonB. W.ElbertD. L.OvodV.KastenT.SigurdsonW.et al (2013). Increased in vivo amyloid-b42 production, exchange, and loss in presenilin mutation carriers.Sci. Transl. Med.5189ra77. 10.1126/scitranslmed.3005615

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 121

  PuthiyedthN.RiverosC.BerrettaR.MoscatoP. (2015). A new combinatorial optimization approach for integrated feature selection using different datasets: a prostate cancer transcriptomic study.PLoS One10:e0127702. 10.1371/journal.pone.0127702

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 122

  PuthiyedthN.RiverosC.BerrettaR.MoscatoP. (2016). Identification of differentially expressed genes through integrated study of Alzheimer’s disease affected brain regions.PLoS One11:e0152342. 10.1371/journal.pone.0152342

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 123

  RahmanM. R.IslamT.TuranliB.ZamanT.FaruqueeH. M.RahmanM. M.et al (2019). Network-based approach to identify molecular signatures and therapeutic agents in Alzheimer’s disease.Comput. Biol. Chem.78431–439. 10.1016/j.compbiolchem.2018.12.011

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 124

  RitchieM. E.PhipsonB.WuD.HuY.LawC. W.ShiW.et al (2015). Limma powers differential expression analyses for RNA-sequencing and microarray studies.Nucleic Acids Res.43:e47. 10.1093/nar/gkv007

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 125

  RobersonE. D.Scearce-LevieK.PalopJ. J.YanF.ChengI. H.WuT.et al (2007). Reducing endogenous tau ameliorates amyloid beta-induced deficits in an Alzheimer’s disease mouse model.Science316750–754. 10.1126/science.1141736

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 126

  RobertsB. R.LindM.WagenA. Z.RembachA.FrugierT.LiQ. X.et al (2017). Biochemically-defined pools of amyloid-β in sporadic Alzheimer’s disease: correlation with amyloid PET.Brain1401486–1498. 10.1093/brain/awx057

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 127

  RobinsonM. D.McCarthyD. J.SmythG. K. (2009). edgeR: a bioconductor package for differential expression analysis of digital gene expression data.Bioinformatics26139–140. 10.1093/bioinformatics/btp616

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 128

  RothR. B.HeveziP.LeeJ.WillhiteD.LechnerS. M.FosterA. C.et al (2006). Gene expression analyses reveal molecular relationships among 20 regions of the human CNS.Neurogenetics767–80. 10.1007/s10048-006-0032-6

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 129

  Rovelet-LecruxA.HannequinD.RauxG.MeurN. LeLaquerrièreA.VitalA.et al (2006). APP locus duplication causes autosomal dominant early-onset Alzheimer disease with cerebral amyloid angiopathy.Nat. Genet.3824–26. 10.1038/ng1718

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 130

  SaidiL. J.PolydoroM.KayK. R.SanchezL.MandelkowE. M.HymanB. T.et al (2015). Carboxy terminus heat shock protein 70 interacting protein reduces tau-associated degenerative changes.J. Alzheimers Dis.44937–947. 10.3233/JAD-142094

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 131

  SantosC. Y.SnyderP. J.WuW. C.ZhangM.EcheverriaA.AlberJ. (2017). Pathophysiologic relationship between Alzheimer’s disease, cerebrovascular disease, and cardiovascular risk: a review and synthesis.Alzheimers Dement. (Amst)769–87. 10.1016/j.dadm.2017.01.005

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 132

  ScheckelC.DrapeauE.FriasM. A.ParkC. Y.FakJ.Zucker-ScharffI.et al (2016). Regulatory consequences of neuronal ELAV-like protein binding to coding and non-coding RNAs in human brain.elife5:e10421. 10.7554/eLife.10421

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 133

  ScheperW.HoozemansJ. J. M. (2015). The unfolded protein response in neurodegenerative diseases: a neuropathological perspective.Acta Neuropathol.130315–331. 10.1007/s00401-015-1462-8

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 134

  ScottT. J.O’ConnorA. C.LinkA. N.BeaulieuT. J. (2014). Economic analysis of opportunities to accelerate Alzheimer’s disease research and development.Ann. N. Y. Acad. Sci.131317–34. 10.1111/nyas.12417

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 135

  SenguptaU.NilsonA. N.KayedR. (2016). The role of amyloid-β oligomers in toxicity, propagation, and immunotherapy.EBioMedicine642–49. 10.1016/j.ebiom.2016.03.035

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 136

  ShannonP.MarkielA.OzierO.BaligaN. S.WangJ. T.RamageD.et al (2003). Cytoscape: a software environment for integrated models of biomolecular interaction networks.Genome Res.132498–2504. 10.1101/gr.1239303

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 137

  ShimuraH.SchwartzD.GygiS. P.KosikK. S. (2004). CHIP-Hsc70 complex ubiquitinates phosphorylated tau and enhances cell survival.J. Biol. Chem.2794869–4876. 10.1074/jbc.M305838200

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 138

  SlipczukL.BekinschteinP.KatcheC.CammarotaM.IzquierdoI.MedinaJ. H. (2009). BDNF activates mTOR to regulate GluR1 expression required for memory formation.PLoS One4:e6007. 10.1371/journal.pone.0006007

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 139

  St George-HyslopP.FraserP. E. (2012). Assembly of the presenilin γ-/ε-secretase complex.J. Neurochem.12084–88. 10.1111/j.1471-4159.2011.07505.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 140

  StanleyM.MacauleyS. L.HoltzmanD. M. (2016). Changes in insulin and insulin signaling in Alzheimer’s disease: cause or consequence?J. Exp. Med.

  • Google Scholar

• 141

  StopaE. G.TanisK. Q.MillerM. C.NikonovaE. V.PodtelezhnikovA. A.FinneyE. M.et al (2018). Comparative transcriptomics of choroid plexus in Alzheimer’s disease, frontotemporal dementia and Huntington’s disease: implications for CSF homeostasis.Fluids Barriers CNS15:18. 10.1186/s12987-018-0102-9

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 142

  StutzbachL. D.XieS. X.NajA. C.AlbinR.GilmanS.LeeV. M. Y.et al (2013). The unfolded protein response is activated in disease-affected brain regions in progressive supranuclear palsy and Alzheimer’s disease.Acta Neuropathol. Commun.1:31. 10.1186/2051-5960-1-31

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 143

  SudhofT. C. (2012). Calcium control of neurotransmitter release.Cold Spring Harb. Perspect. Biol.4:a011353. 10.1101/cshperspect.a011353

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 144

  SutherlandG. T.JanitzM.KrilJ. J. (2011). Understanding the pathogenesis of Alzheimer’s disease: will RNA-Seq realize the promise of transcriptomics?J. Neurochem.116937–946. 10.1111/j.1471-4159.2010.07157.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 145

  SzklarczykD.FranceschiniA.KuhnM.SimonovicM.RothA.MinguezP.et al (2011). The STRING database in 2011: functional interaction networks of proteins, globally integrated and scored.Nucleic Acids Res.39D561–D568. 10.1093/nar/gkq973

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 146

  TanM. G.ChuaW. T.EsiriM. M.SmithA. D.VintersH. V.LaiM. K. (2010). Genome wide profiling of altered gene expression in the neocortex of Alzheimer’s disease.J. Neurosci. Res.881157–1169. 10.1002/jnr.22290

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 147

  TrepanierC. H.JacksonM. F.MacDonaldJ. F. (2012). Regulation of NMDA receptors by the tyrosine kinase Fyn.FEBS J.27912–19. 10.1111/j.1742-4658.2011.08391.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 148

  TsengB. P.GreenK. N.ChanJ. L.Blurton-JonesM.LaFerlaF. M. (2008). Aβ inhibits the proteasome and enhances amyloid and tau accumulation.Neurobiol. Aging291607–1618. 10.1016/j.neurobiolaging.2007.04.014

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 149

  UchidaS.MartelG.PavlowskyA.TakizawaS.HeviC.WatanabeY.et al (2014). Learning-induced and stathmin-dependent changes in microtubule stability are critical for memory and disrupted in ageing.Nat. Commun.5:4389. 10.1038/ncomms5389

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 150

  UmJ. W.NygaardH. B.HeissJ. K.KostylevM. A.StagiM.VortmeyerA.et al (2012). Alzheimer amyloid-β oligomer bound to postsynaptic prion protein activates Fyn to impair neurons.Nat. Neurosci.151227–1235. 10.1038/nn.3178

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 151

  VanDyckC. H.NygaardH. B.ChenK.DonohueM. C.RamanR.RissmanR. A.et al (2019). Effect of AZD0530 on cerebral metabolic decline in alzheimer disease: a randomized clinical trial.JAMA Neurol.761219–1229. 10.1001/jamaneurol.2019.2050

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 152

  VargasD. M.DeBastianiM. A.ZimmerE. R.KlamtF. (2018). Alzheimer’s disease master regulators analysis: search for potential molecular targets and drug repositioning candidates.Alzheimers. Res. Ther.10:59. 10.1186/s13195-018-0394-7

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 153

  WangJ.GuB. J.MastersC. L.WangY.-J. (2017). A systemic view of Alzheimer disease — insights from amyloid-β metabolism beyond the brain.Nat. Rev. Neurol.13612–623. 10.1038/nrneurol.2017.111

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 154

  WangL.ChiangH.-C.WuW.LiangB.XieZ.YaoX.et al (2012). Epidermal growth factor receptor is a preferred target for treating Amyloid–induced memory loss.Proc. Natl. Acad. Sci. U.S.A.10916743–16748. 10.1073/pnas.1208011109

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 155

  WangM.RoussosP.McKenzieA.ZhouX.KajiwaraY.BrennandK. J.et al (2016). Integrative network analysis of nineteen brain regions identifies molecular signatures and networks underlying selective regional vulnerability to Alzheimer’s disease.Genome Med.8:104. 10.1186/s13073-016-0355-3

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 156

  WangQ.LiW. X.DaiS. X.GuoY. C.HanF. F.ZhengJ. J.et al (2017). Meta-analysis of Parkinson’s disease and Alzheimer’s disease revealed commonly impaired pathways and dysregulation of NRF2-dependent genes.J. Alzheimers Dis.561525–1539. 10.3233/JAD-161032

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 157

  WangY.MandelkowE. (2016). Tau in physiology and pathology.Nat. Rev. Neurosci.1722–35. 10.1038/nrn.2015.1

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 158

  WelterD.MacArthurJ.MoralesJ.BurdettT.HallP.JunkinsH.et al (2014). The NHGRI GWAS catalog, a curated resource of SNP-trait associations.Nucleic Acids Res.42D1001–D1006. 10.1093/nar/gkt1229

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 159

  WestermannB. (2012). Bioenergetic role of mitochondrial fusion and fission.Biochim. Biophys. Acta Bioenerg.18171833–1838. 10.1016/j.bbabio.2012.02.033

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 160

  WilliamT. L.JohnsonB. R. G.UrbancB.JenkinsA. T. A.ConnellS. D. A.SerpellL. C. (2011). Aβ42 oligomers, but not fibrils, simultaneously bind to and cause damage to ganglioside-containing lipid membranes.Biochem. J.43967–77. 10.1042/BJ20110750

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 161

  WilliamsC.Mehrian ShaiR.WuY.HsuY.-H.SitzerT.SpannB.et al (2009). Transcriptome analysis of synaptoneurosomes identifies neuroplasticity genes overexpressed in incipient Alzheimer’s disease.PLoS One4:e4936. 10.1371/journal.pone.0004936

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 162

  WruckW.SchröterF.AdjayeJ. (2016). Meta-analysis of transcriptome data related to hippocampus biopsies and iPSC-derived neuronal cells from Alzheimer’s disease patients reveals an association with FOXA1 and FOXA2 gene regulatory networks.J. Alzheimers Dis.501065–1082. 10.3233/JAD-150733

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 163

  WuB.McDonaldA. J.MarkhamK.RichC. B.McHughK. P.TatzeltJ.et al (2017). The N-terminus of the prion protein is a toxic effector regulated by the C-terminus.elife6:e23473. 10.7554/eLife.23473

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 164

  XiangS.HuangZ.WangT.HanZ.YuC. Y.NiD.et al (2018). Condition-specific gene co-expression network mining identifies key pathways and regulators in the brain tissue of Alzheimer’s disease patients.BMC Med. Genomics11:115. 10.1186/s12920-018-0431-1

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 165

  XuM.ZhangD. F.LuoR.WuY.ZhouH.KongL. L.et al (2018). A systematic integrated analysis of brain expression profiles reveals YAP1 and other prioritized hub genes as important upstream regulators in Alzheimer’s disease.Alzheimers Dement.14215–229. 10.1016/j.jalz.2017.08.012

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 166

  YanT.DingF.ZhaoY. (2019). Integrated identification of key genes and pathways in Alzheimer’s disease via comprehensive bioinformatical analyses.Hereditas156:25. 10.1186/s41065-019-0101-0

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 167

  YangT.LiS.XuH.WalshD. M.SelkoeD. J. (2017). Large soluble oligomers of amyloid β-protein from alzheimer brain are far less neuroactive than the smaller oligomers to which they dissociate.J. Neurosci.37152–163. 10.1523/JNEUROSCI.1698-16.2016

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 168

  YasudaS.SugiuraH.TanakaH.TakigamiS.YamagataK. (2012). p38 MAP kinase inhibitors as potential therapeutic drugs for neural diseases.Cent. Nerv. Syst. Agents Med. Chem.1145–59. 10.2174/187152411794961040

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 169

  YinG.LiL. Y.QuM.LuoH. B.WangJ. Z.ZhouX. W. (2011). Upregulation of AKT attenuates amyloid-β-induced cell apoptosis.J. Alzheimers Dis.25337–345. 10.3233/JAD-2011-110104

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 170

  YinZ.RajD.SaiepourN.VanDamD.BrouwerN.HoltmanI. R.et al (2017). Immune hyperreactivity of Aβ plaque-associated microglia in Alzheimer’s disease.Neurobiol. Aging55115–122. 10.1016/j.neurobiolaging.2017.03.021

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 171

  YuG.WangL.-G.HanY.HeQ.-Y. (2012). clusterProfiler: an R package for comparing biological themes among gene clusters.OMICS16284–287. 10.1089/omi.2011.0118

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 172

  YuanP.CondelloC.KeeneC. D.WangY.BirdT. D.PaulS. M.et al (2016). TREM2 haplodeficiency in mice and humans impairs the microglia barrier function leading to decreased amyloid compaction and severe axonal dystrophy.Neuron90724–739. 10.1016/j.neuron.2016.05.003

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 173

  ZhangB.GaiteriC.BodeaL. G.WangZ.McElweeJ.PodtelezhnikovA. A.et al (2013). Integrated systems approach identifies genetic nodes and networks in late-onset Alzheimer’s disease.Cell153707–720. 10.1016/j.cell.2013.03.030

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 174

  ZhangM.CaiF.ZhangS.ZhangS.SongW. (2015). Overexpression of ubiquitin carboxyl-terminal hydrolase L1 (UCHL1) delays Alzheimer’s progression in vivo.Sci. Rep.4:7298. 10.1038/srep07298

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 175

  ZhangY. W.WangR.LiuQ.ZhangH.LiaoF. F.XuH. (2007). Presenilin/γ-secretase-dependent processing of β-amyloid precursor protein regulates EGF receptor expression.Proc. Natl. Acad. Sci. U.S.A.10410613–10618. 10.1073/pnas.0703903104

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 176

  ZhouY. Y.LiY.JiangW. Q.ZhouL. F. (2015). MAPK/JNK signalling: a potential autophagy regulation pathway.Biosci. Rep.35:e00199. 10.1042/BSR20140141

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 177

  ZouS.KumarU. (2018). Cannabinoid receptors and the endocannabinoid system: signaling and function in the central nervous system.Int. J. Mol. Sci.19:833. 10.3390/ijms19030833

  • Pubmed Abstract
  • CrossRef
  • Google Scholar