Transcriptome analysis of human dorsal striatum implicates attenuated canonical WNT signaling in neuroinflammation and in age-related impairment of striatal neurogenesis and synaptic plasticity

Abstract

Background:

Motor and cognitive decline as part of the normal aging process is linked to alterations in synaptic plasticity and reduction of adult neurogenesis in the dorsal striatum. Neuroinflammation, particularly in the form of microglial activation, is suggested to contribute to these age-associated changes.

Objective and Methods:

To explore the molecular basis of alterations in striatal function during aging we analyzed RNA-Seq data for 117 postmortem human dorsal caudate samples and 97 putamen samples acquired through GTEx.

Results:

Increased expression of neuroinflammatory transcripts including TREM2, MHC II molecules HLA-DMB, HLA-DQA2, HLA-DPA1, HLA-DPB1, HLA-DMA and HLA-DRA, complement genes C1QA, C1QB, CIQC and C3AR1, and MHCI molecules HLA-B and HLA-F was identified. We also identified down-regulation of transcripts involved in neurogenesis, synaptogenesis, and synaptic pruning, including DCX, CX3CL1, and CD200, and the canonical WNTs WNT7A, WNT7B, and WNT8A. The canonical WNT signaling pathway has previously been shown to mediate adult neurogenesis and synapse formation and growth. Recent findings also highlight the link between WNT/β-catenin signaling and inflammation pathways.

Conclusions:

These findings suggest that age-dependent attenuation of canonical WNT signaling plays a pivotal role in regulating striatal plasticity during aging. Dysregulation of WNT/β-catenin signaling via astrocyte–microglial interactions is suggested to be a novel mechanism that drives the decline of striatal neurogenesis and altered synaptic connectivity and plasticity, leading to a subsequent decrease in motor and cognitive performance with age. These findings may aid in the development of therapies targeting WNT/β-catenin signaling to combat cognitive and motor impairments associated with aging.

Keywords

RNA sequencing caudate putamen aging WNT striatal neurogenesis synaptic plasticity neuroinflammation

1 Introduction

Normal aging is associated with a progressive decline in motor and cognitive functions. The dorsal striatum, which includes the caudate nucleus and putamen, receives dopaminergic input from the substantia nigra and glutamatergic input from the cortex and thalamus (Do et al., 2012). The dorsal striatum is involved in controlling motor movement and action-contingent learning (Delgado, Miller, Inati, & Phelps, 2005; Delgado, Nystrom, Fissell, Noll, & Fiez, 2000; Haber, 2016; Haruno et al., 2004; Kreitzer & Malenka, 2008; Tricomi, Delgado, & Fiez, 2004; C. Wu, Garamszegi, Xie, & Mash, 2017). Altered striatal plasticity has been implicated in the age-related deterioration in motor performance indicators such as balance deficits and slowed movement (Cham, Perera, Studenski, & Bohnen, 2007; Seidler et al., 2010; van Dyck et al., 2008), and cognitive impairment such as deficits in working memory, episodic memory, and executive function (Erixon-Lindroth et al., 2005; Klostermann, Braskie, Landau, O’Neil, & Jagust, 2012). The molecular mechanisms involved in the process of aging, however, remain poorly understood.

Many studies have demonstrated structural brain changes in normal aging (Fjell & Walhovd, 2010). MRI studies showed longitudinal shrinkage of the caudate nucleus and putamen in healthy adults (Gunning-Dixon, Head, McQuain, Acker, & Raz, 1998; Raz et al., 2003). A study in aged cats (Levine et al., 1986) found decreased spine density and dendritic length in medium-sized spiny neurons of the caudate nucleus. Recently, using ¹⁴C birth-dating, adult neurogenesis has been verified in the adult human subventricular zone (SVZ) of the lateral ventricle wall and the subgranular zone (SGZ) of the hippocampal dentate gyrus (Ernst et al., 2014; Spalding et al., 2013). In contrast to rodents, SVZ-derived neuroblasts in humans migrated only to the caudate nucleus and putamen, but not to the olfactory bulb (Ernst et al., 2014). Since adult-generated striatal neurons are preferentially destroyed in the neurodegenerative Huntington’s disease (HD) and adult neurogenesis in the human striatum has been shown to decline during aging (Ernst et al., 2014), age-associated reductions in adult striatal neurogenesis, to the extent that those adult-generated neurons are inadequate for normal striatal function, may contribute to neurodegenerative disorders.

Rodent experiments have suggested that inflammation plays a role in age-associated changes in adult neurogenesis and neuroplasticity, and these often relate to aberrant microglial activation (Gebara, Sultan, Kocher-Braissant, & Toni, 2013; L’Episcopo et al., 2013; Zhang et al., 2013). Notably, up-regulation of genes reflecting microglial activation has been reported in the aging human brain (Cribbs et al., 2012). Microglia, the resident immune cells of the brain, survey and shape neural circuits through phagocytosis of unnecessary synaptic elements and newborn neurons in the developing and adult brain (Brown & Neher, 2014; Paolicelli & Gross, 2011; Sierra et al., 2010; Wake, Moorhouse, Miyamoto, & Nabekura, 2013). On the other hand, inflammatory activation of microglia has been implicated in the pathogenesis of various neurological disorders (Frick, Williams, & Pittenger, 2013; Mondelli, Vernon, Turkheimer, Dazzan, & Pariante, 2017; Mosher & Wyss-Coray, 2014; Wyss-Coray, 2016).

In aged mice, activated microglia may contribute to the decline of SVZ neurogenesis through the up-regulation of microglial pro-inflammatory mediators (L’Episcopo et al., 2013). Recent studies indicate that astrocytes respond to microglia-derived inflammatory molecules by increasing expression of secreted WNT ligands which act to restrain microglial activation by attenuating pro-inflammatory cytokine overexpression (L’Episcopo et al., 2014; L’Episcopo et al., 2011; L’Episcopo et al., 2012; Marchetti & Pluchino, 2013). Microglia overactivation in aged mice has, however, been shown to impair the neurogenic capacity of neural progenitor cells (NPCs) in the SVZ through aberrant astrocyte–microglia crosstalk and subsequent down-regulation of WNT/β-catenin signaling in both astrocytes and microglia (L’Episcopo et al., 2012; L’Episcopo et al., 2013). Nevertheless, the role of WNT-glial interactions in the aged human brain is not known.

To identify age-dependent changes in gene expression networks we performed differential expression and pathway enrichment analyses on 117 postmortem human caudate and 97 putamen samples. These samples were from the Genotype-Tissue Expression (GTEx) database, a NIH founded project that provides extensive gene expression and clinical data from a large cohort of postmortem tissues unbiased for any diseases (Consortium, 2013, 2015; Mele et al., 2015).

2 Materials and methods

2.1 GTEx database and pre-processing of RNA-Sequencing dataset

The GTEx database (v6) collected 1632 brain samples from 13 brain regions. Detailed information on sample collection, RNA sequencing, and data processing pipelines can be found in the GTEx Consortium paper (Mele et al., 2015). Specifically, only optimal (SMTORMVE not equal to “FLAGGED”) TrueSeq.v1 (SMGEBTCHT = “TrueSeq.v1”) RNA-Seq samples in the caudate (N = 117) and putamen (N = 97) were selected for this study (Table S1). Gene expression datasets encompassing raw read counts and RPKM (Reads Per Kilobase of transcript per Million mapped reads) were generated using RNA-SeQC. In addition, GTEx dataset with GENECODE v19 as a reference transcriptome resulted in 56,318 annotated genes.

We evaluated similarities in gene expression between the caudate and putamen samples for each individual by performing hierarchical clustering and inter-array correlation (IAC) (Oldham et al., 2008). Read counts were transformed using the voom function in Limma before IAC analysis (Law, Chen, Shi, & Smyth, 2014). Relationships among RNA-Seq samples were measured using average linkage hierarchical clustering with a distance metric of 1-IAC.

2.2 Identification of significant age-associated genes

Genes differentially expressed based on age were identified using a procedure similar to that described previously (C. Wu, Bendriem, Garamszegi, Song, & Lee, 2017; C. Wu, Chen, Shu, & Lee, 2017; C. Wu, Xu, Tsai, Freed, & Lee, 2017). We performed differential gene expression analysis with DESeq2 (v1.12.3), and considered genes to be expressed if they had more than five reads in at least 25%of total samples (Love, Huber, & Anders, 2014). Next, we used the “svaseq” function from the sva R package to identify hidden batch effects and cofounding factors (Leek, 2014). In addition to gender, 15 surrogate variables were added to the formula in DESeq2. In either caudate or putamen samples, the residuals for normalized read counts, after gender and surrogate variables correction, were tested against continuous age using the following negative binomial (NB) generalized linear regression model (GLM): $K_{ij} \sim NB (μ_{ij}, α_{j})$ (1) $μ_{ij} = s_{i} q_{ij}$ (2)

${log}_{2} (q_{ij}) = β_{0 j} + β_{1 j} AG E_{i} + β_{2 j} GENDE R_{i} + \sum_{k = 15}^{N} β_{kj} S V_{ki} + ɛ_{ij}$ (3)

where Kij is the read counts for gene j in sample i, fitted with a negative binomial distribution. α_j is a gene-specific dispersion parameter. μ_ij represents fitted mean, containing a sample-specific size factor s_i and a covariate-dependent part q_ij (Love et al., 2014). In Equation 3, β₀ is the regression intercept for gene j, ɛ_ij is the error term. β_1j, β_2j, and β_kj (k = 1, . . . , 15) denote the regression coefficients of age, gender and k_th surrogate variables for gene j respectively.

The Wald test P-values were adjusted for multiple testing using the Benjamini-Hochberg (BH) algorithm. Linear association with the continuous “AGE” variable was used as the primary strategy for detecting age-associated genes. To further validate our results, we arbitrarily divided RNA-Seq samples into three age groups: young (AGE≤55), middle-aged (55 < AGE < 65), and old (AGE≥65). The same formula shown in Equation 3 above was applied with the exception of the primary variable being changed to categorical age groups (“AGEGROUP”) with three-factor levels:

${log}_{2} (q_{ij}) = β_{0 j} + β_{1 j} AGEGROU P_{i} + β_{2 j} GENDE R_{i} + \sum_{k = 15}^{N} β_{kj} S V_{ki} + ɛ_{ij}$ (4)

The list of differentially expressed genes that differed between old- and young-age groups was compared to that of the age-associated genes derived from linear association with continuous age. The regression coefficients β_1j indicate the log₂ fold changes for gene j from the contrast between the age groups, whereas log₂ fold changes for continuous age means change per unit (year) (Love et al., 2014). The threshold for significance of differential expression of genes between old- and young-age groups was at a false discovery rate (FDR) level of 0.2. We defined significant age-associated genes at FDR < 0.2 and uncorrected P-value < 0.01 from linear association with continuous age as the primary strategy.

2.3 Detecting age-associated KEGG pathways and gene ontology terms

2.3.1 Gene set enrichment analysis

Gene set enrichment analysis (GSEA) was performed by using clusterProfiler (v3.0.2) in R (Yu, Wang, Han, & He, 2012), which implements the algorithm developed by the Broad Institute (Subramanian et al., 2005). Specifically, we constructed a pre-ranked gene list of all expressed genes ordered by log2 fold changes for continuous age (beta regression coefficients) from DESeq2 package. Enrichment score (ES) and the significance level of ES (nominal P-value) were calculated by 1000 phenotype-based permutation test. Pre-defined pathways from Kyoto Encyclopedia of Genes and Genomes (KEGG PATHWAY database) with minimal gene size of 10 and maximal of 500 were used in GSEA. q-values were calculated for FDR control. Significant pathways with q-value less than 0.05 were reported.

2.4 Overrepresentation test

We performed overrepresentation tests on 1832 age-associated genes by using meta-analysis. The function “enrichKEGG” or “enrichGO” of the cluster Profiler package was used to test whether up- or down-regulated age-associated genes are significantly overrepresented in given pathways from KEGG or gene ontology (GO) categories (Biological Process, BP). Hypergeometric distribution was used to evaluate the P-value. q-values were reported for FDR control. For pathway-based analyses, we defined significant age-associated biological functions at the level of q-value less than 0.05. Selected significant age-associated KEGG pathways were visualized through “pathview” R package (W. Luo & Brouwer, 2013).

2.5 Expression correlation analysis of synaptic and immune genes

To evaluate the gene-gene correlations within synaptic and immune gene sets, Pearson correlation matrix based on gene expression in RPKM was constructed and displayed by corrplot package in R. The function “rcorr” from Hmisc package was used to compute the significance levels for Pearson correlations. Any correlation that did not pass the significance level (P-value < 0.01) was left blank.

2.6 Replication by using human putamen dataset from UKBEC and meta-analysis

The UK Human Brain Expression Consortium (UKBEC) dataset is the largest control brain gene expression dataset generated from microarray analysis. This dataset contains 129 putamen samples from Caucasians aged 16–102 years (Table S1). A detailed description of tissue collection and dissection, RNA isolation, and processing of brain samples was described in Trabzuni’s study (Trabzuni et al., 2011). Briefly, the expression profiles of postmodern brain tissues were performed on the Affymetrix GeneChip® Human Exon 1.0 ST Arrays (Affymetrix, High Wycombe, UK). The gene-level expression dataset (GSE60862) was already normalized using Robust Multichip Analysis (RMA) with GC background correction. We extracted 129 putamen samples with 26493 gene signatures. For each gene signature, we performed linear regression of age against gender and surrogate variable corrected expression values by using limma R package (Leek & Storey, 2007; Ritchie et al., 2015): $y_{ij} = β_{0 j} + β_{1 j} AG E_{i} + β_{2 j} GENDE R_{i} + β_{3 j} S V_{i} + ɛ_{ij}$ (5)

Where y_ij is RMA-normalized gene expression value for gene j in sample i. The “logFC” for continuous age variable produced by limma package is the beta regression coefficient (β_1j). Multiple testing P-values were adjusted using the BH method.

To identify common age-associated genes in the dorsal striatum, we performed a meta-analysis by using P-value combination approaches (Fisher and inverse normal) to increase detection power (Rau, Marot, & Jaffrezic, 2014). All significant genes (FDR < 0.05) without differential expression conflicts were reported.

2.7 Statistical analysis and dataset access

All statistical computing was performed in the R (v3.3, https://www.r-project.org/). Distribution of clinical characteristics and Pearson/Spearman correlation were performed with JMP v12.2.0 (SAS).

RNA-Seq dataset can be downloaded from dbGaP, study accession no. phs000424.v6.p1. Details about the GTEx resource can be found in the GTEx portal (http://www.gtexportal.org). Microarray dataset from UKBEC can be downloaded from GEO, accession no. GSE60862.

3 Results

3.1 Identifying significant age-associated genes in the human caudate and putamen

The inter-array correlation (IAC) analysis showed that RNA-Seq samples from either the caudate (mean IAC = 0.93) or putamen (mean IAC = 0.927) were highly correlated (Fig. S1A-D). This confirms an overall consistency amongst samples in each of the two brain regions.

Genes that were differentially expressed between young- (AGE≤55) and old-age (AGE≥65) groups with an FDR level of 0.2 were compared with age-associated genes derived from linear association with continuous age (Fig. 1B). A large overlap of the age-associated genes between these two strategies was observed (Fig. 1B), indicating consistency between the results of the two analyses. Since group comparison ignores variances within each age group, we opted to use the linear association with continuous age strategy for the subsequent analyses. Volcano plots (Fig. 1C, D) display significant age-associated gene expression changes in the caudate and putamen. At the level of FDR < 0.2 and P-value < 0.01, we identified 2494 age-associated genes in the caudate and 1004 age-associated genes in the putamen. Lists of these genes are presented in Tables S2 and S3. A potential explanation for the larger list of age-associated genes in the caudate samples compared to those of the putamen is the small sample size and lower IAC of the putamen dataset.

Fig. 1

Transcriptome analysis reveals a large number of significant age-associated genes in caudate and putamen. (A) Histograms showing age distribution of subjects in caudate and putamen. (B) Shared significant age-associated genes between two different data analysis strategies. Red circles represent age-associated genes derived from linear association with continuous age (FDR < 0.2 and uncorrected P-value < 0.01), and green circles show the significant differentially expressed (DE) genes in old age group contrasted with young age group (FDR < 0.2). (C, D) Volcano plots show the -log10 (P-value) and beta regression coefficient of age for all expressed genes in the caudate (C) and putamen (D). Age-associated genes at the FDR level of 0.05, 0.1, and 0.2 are indicated by red, green, and yellow, respectively.

3.2 Consistent age-related gene expression changes between caudate and putamen samples

The dorsal striatum houses the caudate and putamen which have similar organization of their neuronal circuitry (Do, Kim, Bakes, Lee, & Kaang, 2012). The question then arises whether age-related gene expression changes in these two regions are coordinated. We have shown that caudate and putamen share 472 common significant age-associated genes (Fig. 2A), suggesting a substantial overlap between the caudate and putamen. Moreover, beta regression coefficients for age were highly correlated (Pearson r = 0.56) between the caudate and putamen (Fig. 2B). A negative regression coefficient of age for gene j signifies that for every unit (year) increase, we expect a beta unit decrease in gene expression (log2 normalized read counts; Equation 3), holding all other variables constant. Amongst a total of 472 shared genes, all directions of age-related changes (beta coefficient) were identical between the caudate and putamen, wherein 255 were down-regulated and 217 were up-regulated (Table S4). These results suggest consistent age-related gene expression changes between these two brain regions.

Fig. 2

Correlation of age-related gene expression changes between caudate and putamen. (A) Venn diagram showing 472 shared significant age-associated genes between caudate and putamen. (B) Each point shows the beta regression coefficient of age in both brain regions. Both the caudate and putamen showing a high correlation in age-related gene expression changes (P-value < 0.0001 and Pearson’s r = 0.56).

3.3 Gene Set Enrichment Analysis (GSEA) reveals impaired synaptic function and neurogenesis, immune activation, and increased inflammation with chronological age

Gene Set Enrichment Analysis (GSEA) revealed a total of 65 and 21 significant pathways with a q-value less than 0.05 in the caudate and putamen, respectively (Tables S5 and S6). Notably, 86%(18 out of 21) of enriched pathways in the putamen overlapped with those in the caudate (Fig. 3A). All directions of these age-associated pathway changes (Normalized Enrichment Score) were identical. This substantial overlap between the caudate and putamen indicates that although expression changes are weaker in the putamen, they follow a trend similar to those identified in the caudate.

Fig. 3

Gene set enrichment analysis (GSEA) reveals a high similarity between the caudate and putamen in age-associated pathways. (A) Venn diagram showing the overlap of the significant age-associated KEGG pathways (FDR q-value < 0.05) between the caudate and putamen. (B) Top 10 significant age-associated KEGG pathways in the caudate. (C) Common significant age-associated pathways between the caudate and putamen.

Amongst eighteen pathways found in both the caudate and putamen, the pathways which were up-regulated throughout aging pertained to functions of the immune system and inflammation, including asthma, allograft rejection, autoimmune thyroid disease, staphylococcus aureus infection, viral myocarditis, complement and coagulation cascades, graft-versus-host disease, and antigen processing and presentation (Fig. 3C). The pathways which were down-regulated throughout the aging process were involved in proper synapse maintenance and function, including glutamatergic synapse, morphine addiction, retrograde endocannabinoid signaling, GABAergic synapse, and cholinergic synapse (Fig. 3C). These results are consistent with previous gene expression studies in the aging human brain (Mohan, Mather, Thalamuthu, Baune, & Sachdev, 2016; Sibille, 2013).

Notably, the RAP1 signaling pathway was down-regulated throughout aging in both caudate and putamen (Fig. 3C). RAP1 is essential in maintaining genome stability by protecting telomeric DNA ends from being shortened or altered via non-homologous end joining or homologous recombination (Bae & Baumann, 2007; Martinez et al., 2010; Sarthy, Bae, Scrafford, & Baumann, 2009; Sfeir, Kabir, van Overbeek, Celli, & de Lange, 2010). Telomeres are naturally shortened with age and telomere length has been previously linked to life expectancy in humans (Heidinger et al., 2012; Shammas, 2011). RAP1 down-regulation throughout aging; therefore, suggests that RAP1 signaling may partially drive observed age-related changes in the brain.

Recent evidence using histological and ¹⁴C dating approaches indicated that the turnover rates of neuronal populations in the striatum of the adult human brain decline with age (Ernst et al., 2014). In agreement with these findings, we found that expression of DCX, a marker of newly born neurons in the adult brain, was decreased in both caudate and putamen (Table S4), suggesting that striatal neurogenesis decreases in the aging human brain. WNT signaling has been identified as an essential regulatory pathway in adult neurogenesis and synaptic plasticity (Inestrosa & Arenas, 2010; Lie et al., 2005). Importantly, abnormal crosstalk between the WNT signaling and inflammatory response has been shown to impact SVZ plasticity and cause nigrostriatal injury (L’Episcopo et al., 2012; L’Episcopo et al., 2013). The WNT signaling pathway was found to be down-regulated throughout aging and is one of the top enriched pathways in the caudate (Fig. 3B, Table S5). GSEA in the putamen also showed a down-regulated WNT signaling pathway (P-value = 0.02, q-value = 0.11), but it was not included in our final list due to stringent cutoff criteria (Table S6). Moreover, the PI3K-Akt signaling pathway, which has been shown to positively modulate WNT signaling (Fukumoto et al., 2001; G. Lee et al., 2010), was down-regulated in the putamen (Table S6), further supporting the claim that altered WNT signaling participates in the aging process in the striatum.

We next used the “pathview” R package to visualize molecular networks that cause progressive loss of functional integrity of the striatum during the aging process (W. Luo & Brouwer, 2013). As shown in Fig. 4, expression of WNT signaling genes revealed a concordant decrease (negative regression coefficient) in both caudate and putamen. Moreover, concordant repression of age-associated GABAergic and glutamatergic synapse pathway genes was observed in both caudate and putamen (Fig. 5, Fig. S2). The putamen exhibited a slightly lower concordance decrease in expression of age-associated dopaminergic synapse pathway genes than those of the caudate (Fig. S3). Nevertheless, a concordant age-associated decrease in expression of genes required for long-term potentiation in both caudate and putamen (Fig. S4) is consistent with loss of synaptic function during aging.

Fig. 4

Gene expression changes (beta regression coefficients) in the WNT signaling pathway visualized using Pathview. Expression of WNT signaling pathway genes showing concordant negative associations with age in the caudate (A) and putamen (B). Up-regulated genes (positive regression coefficient for age) appear in red and down-regulated genes (negative regression coefficient for age) appear in green.

Fig. 5

Gene expression changes (beta regression coefficients) in the GABAergic synapse pathway visualized using Pathview. Expression of GABAergic synapse pathway genes showing concordant negative associations with age in the caudate (A) and putamen (B). Up-regulated genes (positive regression coefficient for age) appear in red, and down-regulated genes (negative regression coefficient for age) appear in green.

Expression of antigen processing and presentation genes, including both class I and class II MHC molecules (MHCI and MHCII, respectively), showed a concordant increase in both caudate and putamen (Fig. 6), suggesting activation of immune responses. Correlation analysis of expression of age-related synaptic and immune genes showed a highly coordinated network within functionally similar gene groups in both caudate (Fig. S5) and putamen (Fig. S6), implicating an association between impaired synaptic function and neuroinflammation in aging.

Fig. 6

Gene expression changes (beta regression coefficients) in the antigen processing and presentation pathway visualized using Pathview. Expression of antigen processing and presentation pathway genes showing concordant positive associations with age in the caudate (A) and putamen (B). Up-regulated genes (positive regression coefficient for age) appear in red, and down-regulated genes (negative regression coefficient for age) appear in green.

3.4 Validation of GTEx results using UKBEC putamen dataset and meta-analysis

To validate the results, we compared the RNA-seq data to the UKBEC microarray-based gene expression dataset for the putamen. The UKBEC dataset contains 129 putamen samples from Caucasian individuals aged 16–102 (Fig. 7A) (Trabzuni et al., 2011). We performed a similar linear regression approach based on the concepts previously described to analyze gene expression changes associated with age (Equation 5). Our results consistently showed that a similarly large number of age-associated genes (2462) were identified at the level of FDR < 0.2 and uncorrected P-value < 0.01 (Fig. 7B). All shared expressed genes in three datasets of UKBEC-putamen, GTEx-putamen, and GTEx-caudate (n = 15382) were merged by gene symbols. Beta regression coefficients of age for all 15382 genes were significantly correlated between these three datasets (p < 0.0001, Fig. S7). Moreover, the Venn diagrams (Fig. 7C, 7D) show an overlap of the significant age-associated genes and pathways amongst these three datasets. Notably, both WNT and PI3K-Akt signaling pathways were down-regulated in the putamen using the UKBEC dataset (Table S7), suggesting consistent age-related changes in WNT signaling in the human dorsal striatum across the three different datasets.

Fig. 7

Validation using microarray-based UKBEC putamen gene expression dataset. (A) Age distribution of 129 donors from UKBEC. (B) Volcano plot shows the -log10 (P-value) and beta regression coefficient of age for all expressed genes in the putamen from UKBEC. Age-associated genes at the FDR level of 0.05, 0.1, and 0.2 are indicated by red, green, and yellow, respectively. (C) Three-way Venn diagram showing shared significant age-associated genes among three datasets (FDR < 0.2 and uncorrected P-value < 0.01). (D) Three-way Venn diagram showing shared significant age-related KEGG pathways among three datasets using GSEA (q-value < 0.05). (E) Five-way Venn diagram comparing the list of age-associated genes at a 5%BH threshold using three datasets and two P-value combination methods. (F) Each point shows the beta regression coefficient of age for all significant age-associated genes from the meta-analysis (n = 1832). A correlation probability matrix showing that three datasets are highly correlated. Histograms showing distributions of beta coefficients. Red lines represent the linear fit and 95%density contours.

We next performed a meta-analysis of age-related gene expression profiles on three datasets of UKBEC-putamen, GTEx-putamen, and GTEx-caudate using P-value combination approaches (Fisher and inverse normal) to identify shared molecular changes in the aging human dorsal striatum (Fig. 7E, S8). Significant age-associated genes at a 5%FDR threshold by meta-analysis were identified (Table S8). Among these 1832 genes, 1390 were negatively associated with age and 442 were positively associated with age. Figure 7F showed that beta regression coefficients of age for all significant age-associated genes from the meta-analysis (n = 1832) were significantly correlated between these three datasets (p < 0.0001). Notably, there was an age-related decrease in expression of the marker for newborn neurons DCX amongst the three datasets, further suggesting a reduction in newly-generated neurons (Table S8).

Overrepresentation tests for up- and down-regulated genes using GSEA revealed similar transcriptional signatures to those from the GTEx dataset. As shown in Table S9, down-regulated aging genes were significantly enriched in the PI3K-Akt signaling pathway and pathways acting on synaptic communication in the brain, whereas up-regulated genes were mainly enriched in immune and inflammatory pathways. Overrepresentation tests based on GO biological process database provided additional insights into biological functions of these significant age-associated genes from the meta-analysis. Among the down-regulated genes, the most enriched categories were cognition, learning or memory, WNT signaling, neurogenesis, neuronal migration, axonal and synaptic plasticity, and synaptic transmission. Among the up-regulated genes, the most enriched categories were antigen processing and presentation, and T-cell activation and aggregation (Table S10). These results showed consistent aging-related gene expression profiles amongst three datasets.

4 Discussion

We identified a comprehensive list of aging-related genes and pathways in two human dorsal striatal regions, caudate and putamen. Genes pertaining to neurogenesis, synaptic plasticity, and striatal functions were significantly down-regulated, whereas genes within immune response and inflammation pathways were significantly up-regulated. Notably WNT signaling, which functions at the neuroimmune interface and plays pivotal roles in regulating neural progenitors, post-mitotic neurons, and microglia in the adult human brain, was down-regulated in association with aging. Other prior studies have examined gene expression patterns in aged human brain, but none have focused specifically on the dorsal striatum (Dillman et al., 2017; Donertas et al., 2017). The present study identifies a potential link between WNT signaling and age-mediated changes in neurogenesis, synaptic function, and immune cell activation in the dorsal striatum.

4.1 Age-mediated down-regulation of neurogenesis

Adult neurogenesis is the continuous generation and integration of neurons into existing neural circuitry. These adult-born neurons are structurally plastic and functional. Adult hippocampal neurogenesis has been implicated in cognitive processes such as hippocampal-dependent learning, memory, and mood regulation (Anacker & Hen, 2017; Goncalves, Schafer, & Gage, 2016). Age-dependent impairment of adult neurogenesis in the dentate gyrus (DG) has been linked to hippocampal dysfunction, and counteracting this aging-induced pathophysiology can reduce age-related cognitive decline (Casadesus et al., 2004; Drapeau et al., 2003; Driscoll et al., 2006; Mayo et al., 2003; Sun, Evans, Hsieh, Panici, & Bartke, 2005). The full spectrum of functions of striatal neurogenesis remains unknown; yet, the lower turnover rate of adult-generated striatal neurons in HD patients suggests that the interruption of striatal neurons turnover might be a cause of the motor and cognitive deficits that these patients exhibit (Ernst et al., 2014). It is therefore reasonable to assume that newly generated striatal neurons have a functional role in the striatum. Our results indicate an age-dependent reduction in the neuronal marker DCX transcript in both caudate and putamen (Table S4) and significant enrichment of down-regulated genes in neurogenesis-related pathways (Table S10), both of which are consistent with reports of a decline in striatal neuronal cell turnover during human aging (Ernst et al., 2014). The above observations suggest that reduced striatal neurogenesis in the aged brain relates to a gradual decay in the functionality of the dorsal striatum.

4.2 Age-induced changes in inflammatory and synaptic pathways

Correlation analysis of expression of age-mediated synaptic and immune response genes implicates the association of synaptic deficits and immune activation with aging in both caudate and putamen (Fig. S5, S6). The neuroinflammatory response following microglial activation is thought to contribute to age-related brain plasticity impairments (Dickson et al., 1992; X. G. Luo, Ding, & Chen, 2010). Microglia can be neuroprotective in the developing brain but have also been shown to become neurotoxic and destructive in the aged brain when activated by adverse stimuli, including neurodegeneration (X. G. Luo et al., 2010; Perry & Holmes, 2014; Streit, Miller, Lopes, & Njie, 2008). Microglia-derived proinflammatory mediators, including TNF-α and IL-1β, have been shown to impact adult neurogenesis (Iosif et al., 2006; L’Episcopo et al., 2013; Zunszain et al., 2012); however, transcript expression for these proinflammatory cytokines did not change with age in the dorsal striatum in the current study.

On the other hand, CX3CL1, a chemokine ligand expressed in neurons that plays a vital role in neuroprotection when bound to microglia-specific CX3CR1 (Cardona et al., 2006; Paolicelli & Gross, 2011), was down-regulated with age (Table S8). CX3CR1-knockout mice exhibit decreased adult hippocampal neurogenesis, impaired LTP in the hippocampus, and impaired learning and memory (Bachstetter et al., 2011; Rogers et al., 2011), and exogenous CX3CR1 administration was able to reverse these phenotypes (Bachstetter et al., 2011)(Bachstetter et al., 2011). Decreased levels of CX3CL1 in the aging hippocampus have also been reported (Bachstetter et al., 2011; Vukovic, Colditz, Blackmore, Ruitenberg, & Bartlett, 2012). Recent studies also showed that the neuron-microglia interface via CX3CL1/R1 signaling plays a role in synaptic pruning, a process crucial in eliminating malfunctioning synapses in order to establish efficient or functional neuronal circuits (Paolicelli & Gross, 2011). CX3CR1-knockout mice exhibited reduced microglia numbers and subsequent synaptic pruning deficits, including problems with synaptic transmission and brain connectivity (Paolicelli & Gross, 2011; Zhan et al., 2014). Decreased expression of CX3CL1 with age in the dorsal striatum (Table S8) therefore suggests that neuron-microglia interaction mediated by CX3CL1/CX3CR1 signaling may be impaired during the aging process, possibly contributing to deficits in neurogenesis and synaptic plasticity.

Similar to CX3CL1, the membrane glycoprotein CD200 expressed by neurons inhibits microglia activation when bound to its receptor CD200R (Hoek et al., 2000). CD200 was down-regulated in the aging caudate (Table S2), further pointing to a potential link between increased microglial activation and aging. In addition, the microglia-specific cell surface receptor TREM2 (Colonna & Wang, 2016) was up-regulated with age in the caudate (Table 2), which can either be explained by excessive microglial proliferation or increased TREM2 expression by individual microglia (Frank et al., 2008; Raha et al., 2017). Interestingly, a rare mutant variant of TREM2 was identified as a risk factor for Alzheimer’s disease (Guerreiro et al., 2013; Jonsson et al., 2013), and TREM2 transcription was increased in the frontal cortex and hippocampus of various Alzheimer’s disease mouse models (Matarin et al., 2015; Raha et al., 2017). TREM2 is now well-known as an inductor of microglial activation and enhancer of phagocytic abilities, while also suppressing induction of proinflammatory cytokines (Neumann & Takahashi, 2007), suggesting an age-related increase of TREM2 in the caudate may influence the capacity of microglia to react to aging-related processes.

Microglia are the chief MHCII-expressing antigen-presenting cells (APCs) in the brain (X. G. Luo et al., 2010; Schetters, Gomez-Nicola, Garcia-Vallejo, & Van Kooyk, 2017). Expression of MHCII in the brain can be rapidly up-regulated on the surface of microglia and is often used as a marker of their activation (Wyss-Coray & Mucke, 2002). Previous studies have shown that, in non-human primates, microglial expression of MHCII increases with age in parallel with increases of microglial phagocytic activity, suggesting that microglial will respond more efficiently to stimulation (Peters, Josephson, & Vincent, 1991; Sheffield & Berman, 1998; Sloane, Hollander, Moss, Rosene, & Abraham, 1999). Moreover, post-mortem brain tissue transcriptomics revealed significant up-regulation of MHCII genes in the healthy aging human brain and across neurodegenerative diseases (Bossers et al., 2010; Durrenberger et al., 2015; Katsel, Tan, & Haroutunian, 2009; Parachikova et al., 2007). Our results showed significant up-regulation of MHCII antigen processing and presentation machinery in GO enrichment results in dorsal striatum from meta-analysis, including HLA-DMB, HLA-DQA2, HLA-DPA1, HLA-DPB1, HLA-DMA, and HLA-DRA (Table S10). It is known that MHCII enables antigen presentation to CD4 + T cells; however, little is known about interactions between microglia and CD4 + T cells in the aged brain. Nevertheless, age-dependent induction of MHCII transcripts in the dorsal striatum indicates functional changes in microglia, which may have detrimental effects in brain aging.

Age-dependent shortening of telomeres has been previously reported in the brain (Flanary & Streit, 2003). Given that microglia proliferate substantially throughout development and adulthood, as opposed to post-mitotic neurons, which no longer replicate once mature, telomere shortening in microglia likely significantly contributes to the overall reduction of brain telomere length. Studies have previously identified significant telomere shortening and reduction of telomerase activity in microglia in aged rats (Flanary, Sammons, Nguyen, Walker, & Streit, 2007; Flanary & Streit, 2004). Telomere shorting in humans has been correlated with cognitive decline and dementia (Eitan, Hutchison, & Mattson, 2014; Grodstein et al., 2008; Honig, Kang, Schupf, Lee, & Mayeux, 2012; Honig, Schupf, Lee, Tang, & Mayeux, 2006; Valdes et al., 2010). Age-associated down-regulation of the telomere binding protein RAP1 signaling pathway in the dorsal striatum identified in the present study (Fig. 3C; Table S7, S9) suggests that telomere dysfunction in microglia could influence age-related changes in brain function.

Recent studies suggest that immune-system related molecules, including complement and MHC class I (MHCI) proteins, which play a role in synapse formation and pruning in the young brain, may erroneously eliminate functional synapses in late adulthood, contributing to neurodegenerative diseases (Lazarczyk et al., 2016; Y. Wu, Dissing-Olesen, MacVicar, & Stevens, 2015). C3, an essential player in both classical and alternative complement cascades, localizes to immature synapses and recruits complement receptor 3 (CR3)-expressing microglia for synapse elimination (Cribbs et al., 2012; Stephan, Barres, & Stevens, 2012; Stevens et al., 2007). Microglia-mediated synapse elimination is decreased in C3 and CR3-deficient mice (Schafer et al., 2012), suggesting that complement proteins located at immature synapses result in the targeted removal of these synapses by microglia. Interestingly, CR3 expression levels are increased with age in humans (Cribbs et al., 2012) and play a role in the erroneous elimination of functional synapses, thereby contributing to age-associated cognitive decline (Shi et al., 2015). C1q, another player in the complement cascade (Stephan et al., 2012; Stevens et al., 2007), is also overexpressed in aged human and mouse brains but does not appear to mediate synapse elimination (Stephan et al., 2013). Our results revealed significant up-regulation of C1QA, C1QB, CIQC, and C3AR1 in the caudate with age (Table S2), and meta-analysis showed an age-dependent increase of C1QB and C3AR1 expression in the dorsal striatum (Table S8)

Like complement cascade proteins, MHCI proteins have been implicated in synapse maintenance and synaptic plasticity in the developing CNS throughout non-pathological aging (Goddard, Butts, & Shatz, 2007; Huh et al., 2000; Lazarczyk et al., 2016; H.Lee et al., 2014). MHCI protein expression was significantly increased in the cognitively normal aging human brain (Katsel et al., 2009), indicating that increase in MHCI protein levels might be important in regulating and preserving cognitive function. The age-associated up-regulation of antigen processing and presentation pathway in both caudate and putamen (Fig. 3C), together with up-regulation of antigen processing and presentation machinery in GO enrichment results in dorsal striatum from meta-analysis, including MHCI molecules HLA-B and HLA-F (Table S10), suggest that enhanced MHCI gene expression in the dorsal striatum might contribute to synaptic formation and plasticity during normal aging. However, whether up-regulation of MHCI genes in healthy aging subjects is functionally associated with antigen presentation or is involved in interactions between neural and glial cells needs further investigation.

In contrast to the complement and coagulation cascades pathways that were up-regulated, glutamatergic, GABAergic, and cholinergic synapse proteins were down-regulated in both caudate and putamen in aging striatum (Fig. 3C). Age-associated down-regulation of dopaminergic synaptic pathways and long-term potentiation and long-term depression (LTP/LTD) pathways were identified as two of the most significantly enriched pathways in the caudate (Fig. 3B).

4.3 Canonical WNT signaling in age-mediated neuroinflammation and plasticity

The WNT pathway recently has been linked with the regulation of multiple processes in the adult brain, including neurogenesis, axonal growth, synapse formation and plasticity, and cognitive health (Inestrosa & Arenas, 2010; Maiese, Li, Chong, & Shang, 2008; Oliva, Vargas, & Inestrosa, 2013). In addition, WNT molecules act as anti-inflammatory factors, commonly secreted by astrocytes to suppress pro-inflammatory cytokines released by neighboring microglial and prevent their activation (L’Episcopo et al., 2014; Marchetti & Pluchino, 2013). The extracellular WNT molecules are compartmentalized in the cell into three separate major pathways: the canonical WNT/β-catenin pathway and the non-canonical WNT/planar cell polarity (PCP) or WNT/Ca2⁺ pathways (Inestrosa & Varela-Nallar, 2014). Our results showed significant down-regulation of WNT7A, WNT7B, and WNT8A (Table S10), which are activators of the canonical WNT/β-catenin pathway. Moreover, the PI3K-Akt signaling pathway, a positive regulator of the canonical WNT pathway, was down-regulated with age (Table S6, S7, S9) (Fukumoto et al., 2001), suggesting age-mediated down-regulation of canonical WNT signaling. Many studies have shown that canonical WNT/β-catenin pathway participates in adult neurogenesis, axon remodeling, and synaptic plasticity (Lie et al., 2005; Oliva et al., 2013). In addition, β-catenin down-regulation was shown to be involved in inflammation-induced neurogenic impairment (L’Episcopo et al., 2013). Therefore, age-associated inhibition of canonical WNT/β-catenin signaling may both exacerbate a deterioration in the neuroprotective milieu and promote inflammatory cascades.

The blood-brain barrier (BBB) is a semi-permeable membrane made up of a specialized endothelial lining whose integrity is sustained by microglia, astrocytes, pericytes, and neurons. In a healthy state, the BBB restricts access of immune cells into the CNS and maintains homeostasis. During inflammation, BBB integrity is compromised and immune cells can enter the CNS parenchyma. An impaired BBB is a hallmark of the aging brain (Erdo, Denes, & de Lange, 2017). Recent studies using MRI in healthy subjects with no or mild cognitive impairment found that BBB permeability and overall loss of integrity in the hippocampus increased linearly with age, and was highest in subjects with mild cognitive impairment. This suggests that loss of BBB integrity is an early event in the aging human hippocampus that may contribute to cognitive impairment (Montagne et al., 2015). Notably, the death receptor 6 (TNFRSF21) and TROY (TNFRSF19) were shown to act downstream of WNT/β-catenin signaling in BBB endothelial cells to regulate CNS angiogenesis and barrier genesis (Tam et al., 2012). Dysregulation of TNFRSF21/TNFRSF19 signaling was shown to disrupt the BBB (Sonar & Lal, 2015). Our results identified down-regulation of both these TNFRSF members with age (Table S8). Thus down-regulation of the WNT/β-catenin signaling pathway may influence TNFRSF21 and TNFRSF19 expression and contribute to structural and functional impairment in the BBB during aging.

5 Conclusion

Down-regulation of transcripts involved in neurogenesis, synaptogenesis, synapse pruning, and canonical WNTs, with contrasting up-regulation of neuroinflammatory transcripts was identified in the aging caudate and putamen. These data suggest that age-dependent attenuation of canonical WNT signaling in the dorsal striatum may play a pivotal role at the neuron-glial interface to regulate striatal plasticity during aging. WNT/β-catenin signaling dysregulation via astrocyte–microglial interactions during aging sheds light on novel mechanisms driving the decline of striatal neurogenesis and altered synaptic connectivity and plasticity, ultimately leading to impaired motor and cognitive abilities. These findings suggest promising therapeutic opportunities for targeting WNT/β-catenin signaling in order to enhance endogenous restoration or to promote preservation of brain function during normal aging.

Author contributions

C.W. conducted data analysis, C.W., R.M.B., W.J.F., and C-T.L. wrote the manuscript and contributed to the interpretation of data. C.W. and C-T.L. are the guarantors of this work and take responsibility for the accuracy of the data analysis.

Footnotes

Acknowledgments

The authors thank the GTEx program and the GTEx Consortium. This study would not be possible without the wealth of information contained within this enormous database.

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

References

Anacker,

, & Hen,

(2017). Adult hippocampal neurogenesis and cognitive flexibility - linking memory and mood. Nature Reviews: Neuroscience, 18(6), 335–346. doi: 10.1038/nrn.2017.45

Bachstetter,

A.D.

, Morganti,

J.M.

, Jernberg,

, Schlunk,

, Mitchell,

S.H.

, Brewster,

K.W.

,... Gemma,

(2011). Fractalkine and CX 3 CR1 regulate hippocampal neurogenesis in adult and aged rats. Neurobiology of Aging, 32(11), 2030–2044. doi: 10.1016/j.neurobiolaging.2009.11.022

Bae,

N.S.

, & Baumann,

(2007). A RAP1/TRF2 complex inhibits nonhomologous end-joining at human telomeric DNA ends. Molecular Cell, 26(3), 323–334. doi: 10.1016/j.molcel.2007.03.023

Bossers,

, Wirz,

K.T.

, Meerhoff,

G.F.

, Essing,

A.H.

, van Dongen,

J.W.

, Houba,

,... Swaab,

D.F.

(2010). Concerted changes in transcripts in the prefrontal cortex precede neuropathology in Alzheimer’s disease. Brain, 133(Pt 12), 3699–3723. doi: 10.1093/brain/awq258

Brown,

G.C.

, & Neher,

J.J.

(2014). Microglial phagocytosis of live neurons. Nature Reviews: Neuroscience, 15(4), 209–216. doi: 10.1038/nrn3710

Cardona,

A.E.

, Pioro,

E.P.

, Sasse,

M.E.

, Kostenko,

, Cardona,

S.M.

, Dijkstra,

I.M.

,... Ransohoff,

R.M.

(2006). Control of microglial neurotoxicity by the fractalkine receptor. Nature Neuroscience, 9(7), 917–924. doi: 10.1038/nn1715

Casadesus,

, Shukitt-Hale,

, Stellwagen,

H.M.

, Zhu,

, Lee,

H.G.

, Smith,

M.A.

, & Joseph,

J.A.

(2004). Modulation of hippocampal plasticity and cognitive behavior by short-term blueberry supplementation in aged rats. Nutritional Neuroscience, 7(5-6), 309–316. doi: 10.1080/10284150400020482

Cham,

, Perera,

, Studenski,

S.A.

, & Bohnen,

N.I.

(2007). Striatal dopamine denervation and sensory integration for balance in middle-aged and older adults. Gait and Posture, 26(4), 516–525. doi: 10.1016/j.gaitpost.2006.11.204

Colonna,

, & Wang,

(2016). TREM2 variants: new keys to decipher Alzheimer disease pathogenesis. Nature Reviews: Neuroscience, 17(4), 201–207. doi: 10.1038/nrn.2016.7

10.

Consortium, G.T. (2013). The Genotype-Tissue Expression (GTEx) project. Nature Genetics, 45(6), 580–585. doi: 10.1038/ng.2653

11.

Consortium, G.T. (2015). Human genomics. The Genotype-Tissue Expression (GTEx) pilot analysis: multitissue gene regulation in humans. Science, 348(6235), 648–660. doi: 10.1126/science.1262110

12.

Cribbs,

D.H.

, Berchtold,

N.C.

, Perreau,

, Coleman,

P.D.

, Rogers,

, Tenner,

A.J.

, & Cotman,

C.W.

(2012). Extensive innate immune gene activation accompanies brain aging, increasing vulnerability to cognitive decline and neurodegeneration: a microarray study. Journal of Neuroinflammation, 9, 179. doi: 10.1186/1742-2094-9-179

13.

Delgado,

M.R.

, Miller,

M.M.

, Inati,

, & Phelps,

E.A.

(2005). An fMRI study of reward-related probability learning. Neuroimage, 24(3), 862–873. doi: 10.1016/j.neuroimage.2004.10.002

14.

Delgado,

M.R.

, Nystrom,

L.E.

, Fissell,

, Noll,

D.C.

, & Fiez,

J.A.

(2000). Tracking the hemodynamic responses to reward and punishment in the striatum. Journal of Neurophysiology, 84(6), 3072–3077. doi: 10.1152/jn.2000.84.6.3072

15.

Dickson,

D.W.

, Crystal,

H.A.

, Mattiace,

L.A.

, Masur,

D.M.

, Blau,

A.D.

, Davies,

,... Aronson,

M.K.

(1992). Identification of normal and pathological aging in prospectively studied nondemented elderly humans. Neurobiology of Aging, 13(1), 179–189. doi: 10.1016/0197-4580(92)90027-u

16.

Dillman,

A.A.

, Majounie,

, Ding,

, Gibbs,

J.R.

, Hernandez,

, Arepalli,

,... Cookson,

M.R.

(2017). Transcriptomic profiling of the human brain reveals that altered synaptic gene expression is associated with chronological aging. Scientific Reports, 7(1), 16890. doi: 10.1038/s41598-017-17322-0

17.

Do,

, Kim,

J.I.

, Bakes,

, Lee,

, & Kaang,

B.K.

(2012). Functional roles of neurotransmitters and neuromodulators in the dorsal striatum. Learning and Memory, 20(1), 21–28. doi: 10.1101/lm.025015.111

18.

Donertas,

H.M.

, Izgi,

, Kamacioglu,

, He,

, Khaitovich,

, & Somel,

(2017). Gene expression reversal toward pre-adult levels in the aging human brain and age-related loss of cellular identity. Scientific Reports, 7(1), 5894. doi: 10.1038/s41598-017-05927-4

19.

Drapeau,

, Mayo,

, Aurousseau,

, Le Moal,

, Piazza,

P.V.

, & Abrous,

D.N.

(2003). Spatial memory performances of aged rats in the water maze predict levels of hippocampal neurogenesis. Proceedings of the National Academy of Sciences of the United States of America, 100(24), 14385–14390. doi: 10.1073/pnas.2334169100

20.

Driscoll,

, Howard,

S.R.

, Stone,

J.C.

, Monfils,

M.H.

, Tomanek,

, Brooks,

W.M.

, & Sutherland,

R.J.

(2006). The aging hippocampus: a multi-level analysis in the rat. Neuroscience, 139(4), 1173–1185. doi: 10.1016/j.neuroscience.2006.01.040

21.

Durrenberger,

P.F.

, Fernando,

F.S.

, Kashefi,

S.N.

, Bonnert,

T.P.

, Seilhean,

, Nait-Oumesmar,

,... Reynolds,

(2015). Common mechanisms in neurodegeneration and neuroinflammation: a BrainNet Europe gene expression microarray study. J Neural Transm (Vienna), 122(7), 1055–1068. doi: 10.1007/s00702-014-1293-0

22.

Eitan,

, Hutchison,

E.R.

, & Mattson,

M.P.

(2014). Telomere shortening in neurological disorders: an abundance of unanswered questions. Trends in Neurosciences, 37(5), 256–263. doi: 10.1016/j.tins.2014.02.010

23.

Erdo,

, Denes,

, & de Lange,

(2017). Age-associated physiological and pathological changes at the blood-brain barrier: A review. Journal of Cerebral Blood Flow and Metabolism, 37(1), 4–24. doi: 10.1177/0271678X16679420

24.

Erixon-Lindroth,

, Farde,

, Wahlin,

T.B.

, Sovago,

, Halldin,

, & Backman,

(2005). The role of the striatal dopamine transporter in cognitive aging. Psychiatry Research, 138(1), 1–12. doi: 10.1016/j.pscychresns.2004.09.005

25.

Ernst,

, Alkass,

, Bernard,

, Salehpour,

, Perl,

, Tisdale,

,... Frisen,

(2014). Neurogenesis in the striatum of the adult human brain. Cell, 156(5), 1072–1083. doi: 10.1016/j.cell.2014.01.044

26.

Fjell,

A.M.

, & Walhovd,

K.B.

(2010). Structural brain changes in aging: courses, causes and cognitive consequences. Reviews in the Neurosciences, 21(3), 187–221. doi: 10.1515/revneuro.2010.21.3.187

27.

Flanary,

B.E.

, Sammons,

N.W.

, Nguyen,

, Walker,

, & Streit,

W.J.

(2007). Evidence that aging and amyloid promote microglial cell senescence. Rejuvenation Res, 10(1), 61–74. doi: 10.1089/rej.2006.9096

28.

Flanary,

B.E

, & Streit,

W.J.

(2003). Telomeres shorten with age in rat cerebellum and cortex in vivo. Journal of Anti-Aging Medicine, 6(4), 299–308. doi: 10.1089/109454503323028894

29.

Flanary,

B.E.

, & Streit,

W.J.

(2004). Progressive telomere shortening occurs in cultured rat microglia, but not astrocytes. Glia, 45(1), 75–88. doi: 10.1002/glia.10301

30.

Frank,

, Burbach,

G.J.

, Bonin,

, Walter,

, Streit,

, Bechmann,

, & Deller,

(2008). TREM2 is upregulated in amyloid plaque-associated microglia in aged APP23 transgenic mice. Glia, 56(13) 1438–1447. doi: 10.1002/glia.20710

31.

Frick,

L.R.

, Williams,

, & Pittenger,

(2013). Microglial dysregulation in psychiatric disease. Clinical & Developmental Immunology, 2013, 608654. doi: 10.1155/2013/608654

32.

Fukumoto,

, Hsieh,

C.M.

, Maemura,

, Layne,

M.D.

, Yet,

S.F.

, Lee,

K.H.

,... Lee,

M.E.

(2001). Akt participation in the Wnt signaling pathway through Dishevelled. Journal of Biological Chemistry, 276(20), 17479–17483. doi: 10.1074/jbc.C000880200

33.

Gebara,

, Sultan,

, Kocher-Braissant,

, & Toni,

(2013). Adult hippocampal neurogenesis inversely correlates with microglia in conditions of voluntary running and aging. Frontiers in Neuroscience, 7, 145. doi: 10.3389/fnins.2013.00145

34.

Goddard,

C.A.

, Butts,

D.A.

, & Shatz,

C.J.

(2007). Regulation of CNS synapses by neuronal MHC class I. Proceedings of the National Academy of Sciences of the United States of America, 104(16), 6828–6833. doi: 10.1073/pnas.0702023104

35.

Goncalves,

J.T.

, Schafer,

S.T.

, & Gage,

F.H.

(2016). Adult Neurogenesis in the Hippocampus: From Stem Cells to Behavior. Cell, 167(4), 897–914. doi: 10.1016/j.cell.2016.10.021

36.

Grodstein,

, van Oijen,

, Irizarry,

M.C.

, Rosas,

H.D.

, Hyman,

B.T.

, Growdon,

J.H.

, & De Vivo,

(2008). Shorter telomeres may mark early risk of dementia: preliminary analysis of 62 participants from the nurses’ health study. PloS One, 3(2), e1590. doi: 10.1371/journal.pone.0001590

37.

Guerreiro,

, Wojtas,

, Bras,

, Carrasquillo,

, Rogaeva,

, Majounie,

,...Alzheimer Genetic Analysis, G. (2013). TREM variants in Alzheimer’s disease. New England Journal of Medicine, 368(2), 117–127. doi: 10.1056/NEJMoa1211851

38.

Gunning-Dixon,

F.M.

, Head,

, McQuain,

, Acker,

J.D.

, & Raz,

(1998). Differential aging of the human striatum: a prospective MR imaging study. AJNR Am J Neuroradiol, 19(8), 1501–1507.

39.

Haber,

S.N.

(2016). Corticostriatal circuitry. Dialogues in Clinical Neuroscience, 18(1), 7–21.

40.

Haruno,

, Kuroda,

, Doya,

, Toyama,

, Kimura,

, Samejima,

,... Kawato,

(2004). A neural correlate of reward-based behavioral learning in caudate nucleus: a functional magnetic resonance imaging study of a stochastic decision task. Journal of Neuroscience, 24(7), 1660–1665. doi: 10.1523/JNEUROSCI.3417-03.2004

41.

Heidinger,

B.J.

, Blount,

J.D.

, Boner,

, Griffiths,

, Metcalfe,

N.B.

, & Monaghan,

(2012). Telomere length in early life predicts lifespan. Proceedings of the National Academy of Sciences of the United States of America, 109(5), 1743–1748. doi: 10.1073/pnas.1113306109

42.

Hoek,

R.M.

, Ruuls,

S.R.

, Murphy,

C.A.

, Wright,

G.J.

, Goddard,

, Zurawski,

S.M.

,... Sedgwick,

J.D.

(2000). Down-regulation of the macrophage lineage through interaction with OX2 (CD200). Science, 290(5497), 1768–1771. doi: 10.1126/science.290.5497.1768

43.

Honig,

L.S.

, Kang,

M.S.

, Schupf,

, Lee,

J.H.

, & Mayeux,

(2012). Association of shorter leukocyte telomere repeat length with dementia and mortality. Archives of Neurology, 69(10), 1332–1339. doi: 10.1001/archneurol.2012.1541

44.

Honig,

L.S.

, Schupf,

, Lee,

J.H.

, Tang,

M.X.

, & Mayeux,

(2006). Shorter telomeres are associated with mortality in those with APOE epsilon and dementia. Annals of Neurology, 60(2), 181–187. doi: 10.1002/ana.20894

45.

Huh,

G.S.

, Boulanger,

L.M.

, Du,

, Riquelme,

P.A.

, Brotz,

T.M.

, & Shatz,

C.J.

(2000). Functional requirement for class I MHC in CNS development and plasticity. Science, 290(5499), 2155–2159. doi: 10.1126/science.290.5499.2155

46.

Inestrosa,

N.C.

, & Arenas,

(2010). Emerging roles of Wnts in the adult nervous system. Nature Reviews: Neuroscience, 11(2), 77–86. doi: 10.1038/nrn2755

47.

Inestrosa,

N.C.

, & Varela-Nallar,

(2014). Wnt signaling in the nervous system and in Alzheimer’s disease. Journal of Molecular Cell Biology, 6(1), 64–74. doi: 10.1093/jmcb/mjt051

48.

Iosif,

R.E.

, Ekdahl,

C.T.

, Ahlenius,

, Pronk,

C.J.

, Bonde,

, Kokaia,

,... Lindvall,

(2006). Tumor necrosis factor receptor is a negative regulator of progenitor proliferation in adult hippocampal neurogenesis. Journal of Neuroscience, 26(38), 9703–9712. doi: 10.1523/JNEUROSCI.2723-06.2006

49.

Jonsson,

, Stefansson,

, Steinberg,

, Jonsdottir,

, Jonsson,

P.V.

, Snaedal,

,... Stefansson,

(2013). Variant of TREM2 associated with the risk of Alzheimer’s disease. New England Journal of Medicine, 368(2), 107–116. doi: 10.1056/NEJMoa1211103

50.

Katsel,

, Tan,

, & Haroutunian,

(2009). Gain in brain immunity in the oldest-old differentiates cognitively normal from demented individuals. PloS One, 4(10), e7642. doi: 10.1371/journal.pone.0007642

51.

Klostermann,

E.C.

, Braskie,

M.N.

, Landau,

S.M.

, O’Neil,

J.P.

, & Jagust,

W.J.

(2012). Dopamine and frontostriatal networks in cognitive aging. e. Neurobiology of Aging, 33(3), 615–624. doi: 10.1016/j.neurobiolaging.2011.03.002

52.

Kreitzer,

A.C.

, & Malenka,

R.C.

(2008). Striatal plasticity and basal ganglia circuit function. Neuron, 60(4), 543–554. doi: 10.1016/j.neuron.2008.11.005

53.

L’Episcopo,

, Tirolo,

, Caniglia,

, Testa,

, Morale,

M.C.

, Serapide,

M.F.

,... Marchetti,

(2014). Targeting Wnt signaling at the neuroimmune interface for dopaminergic neuroprotection/repair in Parkinson’s disease. Journal of Molecular Cell Biology, 6(1), 13–26. doi: 10.1093/jmcb/mjt053

54.

L’Episcopo,

, Tirolo,

, Testa,

, Caniglia,

, Morale,

M.C.

, Cossetti,

,... Marchetti,

(2011). Reactive astrocytes and Wnt/beta-catenin signaling link nigrostriatal injury to repair in 1-methyl–phenyl-1,2,3,6-tetrahydropyridine model of Parkinson’s disease. Neurobiology of Disease, 41(2), 508–527. doi: 10.1016/j.nbd.2010.10.023

55.

L’Episcopo,

, Tirolo,

, Testa,

, Caniglia,

, Morale,

M.C.

, Deleidi,

,... Marchetti,

(2012). Plasticity of subventricular zone neuroprogenitors in MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine) mouse model of Parkinson’s disease involves cross talk between inflammatory and Wnt/beta-catenin signaling pathways: functional consequences for neuroprotection and repair. Journal of Neuroscience, 32(6), 2062–2085. doi: 10.1523/JNEUROSCI.5259-11.2012

56.

L’Episcopo,

, Tirolo,

, Testa,

, Caniglia,

, Morale,

M.C.

, Impagnatiello,

,... Marchetti,

(2013). Aging-induced Nrf2-ARE pathway disruption in the subventricular zone drives neurogenic impairment in parkinsonian mice via PI3K-Wnt/beta-catenin dysregulation. Journal of Neuroscience, 33(4), 1462–1485. doi: 10.1523/JNEUROSCI.3206-12.2013

57.

Law,

C.W.

, Chen,

, Shi,

, & Smyth,

G.K.

(2014). voom: Precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biology, 15(2), R29. doi: 10.1186/gb-2014-15-2-r29

58.

Lazarczyk,

M.J.

, Kemmler,

J.E.

, Eyford,

B.A.

, Short,

J.A.

, Varghese,

, Sowa,

,... Dickstein,

D.L.

(2016). Major Histocompatibility Complex class I proteins are critical for maintaining neuronal structural complexity in the aging brain. Scientific Reports, 6, 26199. doi: 10.1038/srep26199

59.

Lee,

, Goretsky,

, Managlia,

, Dirisina,

, Singh,

A.P.

, Brown,

J.B.

,... Barrett,

T.A.

(2010). Phosphoinositide -kinase signaling mediates beta-catenin activation in intestinal epithelial stem and progenitor cells in colitis. Gastroenterology, 139(3), 869-881, e861–869. doi: 10.1053/j.gastro.2010.05.037

60.

Lee,

, Brott,

B.K.

, Kirkby,

L.A.

, Adelson,

J.D.

, Cheng,

, Feller,

M.B.

,... Shatz,

C.J.

(2014). Synapse elimination and learning rules co-regulated by MHC class I H2-Db. Nature, 509(7499), 195–200. doi: 10.1038/nature13154

61.

Leek,

J.T.

(2014). svaseq: removing batch effects and other unwanted noise from sequencing data. Nucleic Acids Research, 42(21). doi: 10.1093/nar/gku864

62.

Leek,

J.T.

, & Storey,

J.D.

(2007). Capturing heterogeneity in gene expression studies by surrogate variable analysis. Plos Genetics, 3(9), 1724–1735. doi: 10.1371/journal.pgen.0030161

63.

Levine,

M.S.

, Adinolfi,

A.M.

, Fisher,

R.S.

, Hull,

C.D.

, Buchwald,

N.A.

, & McAllister,

J.P.

(1986). Quantitative morphology of medium-sized caudate spiny neurons in aged cats. Neurobiology of Aging, 7(4), 277–286. doi: 10.1016/0197-4580(86)90008-4

64.

Lie,

D.C.

, Colamarino,

S.A.

, Song,

H.J.

, Desire,

, Mira,

, Consiglio,

,... Gage,

F.H.

(2005). Wnt signalling regulates adult hippocampal neurogenesis. Nature, 437(7063), 1370–1375. doi: 10.1038/nature04108

65.

Love,

M.I.

, Huber,

, & Anders,

(2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq. Genome Biology, 15(12), 550. doi: 10.1186/s13059-014-0550-8

66.

Luo,

, & Brouwer,

(2013). Pathview: an R/Bioconductor package for pathway-based data integration and visualization. Bioinformatics, 29(14), 1830–1831. doi: 10.1093/bioinformatics/btt285

67.

Luo,

X.G.

, Ding,

J.Q.

, & Chen,

S.D.

(2010). Microglia in the aging brain: relevance to neurodegeneration. Molecular Neurodegeneration, 5, 12. doi: 10.1186/1750-1326-5-12

68.

Maiese,

, Li,

, Chong,

Z.Z.

, & Shang,

Y.C.

(2008). The Wnt signaling pathway: aging gracefully as a protectionist? Pharmacology and Therapeutics, 118(1), 58–81. doi: 10.1016/j.pharmthera.2008.01.004

69.

Marchetti,

, & Pluchino,

(2013). Wnt your brain be inflamed? Yes, it Wnt!. Trends in Molecular Medicine, 19(3), 144–156. doi: 10.1016/j.molmed.2012.12.001

70.

Martinez,

, Thanasoula,

, Carlos,

A.R.

, Gomez-Lopez,

, Tejera,

A.M.

, Schoeftner,

,... Blasco,

M.A.

(2010). Mammalian Rap controls telomere function and gene expression through binding to telomeric and extratelomeric sites. Nature Cell Biology, 12(8), 768–780. doi: 10.1038/ncb2081

71.

Matarin,

, Salih,

D.A.

, Yasvoina,

, Cummings,

D.M.

, Guelfi,

, Liu,

,... Edwards,

F.A.

(2015). A genome-wide gene-expression analysis and database in transgenic mice during development of amyloid or tau pathology. Cell Reports, 10(4), 633–644. doi: 10.1016/j.celrep.2014.12.041

72.

Mayo,

, George,

, Darbra,

, Bouyer,

J.J.

, Vallee,

, Darnaudery,

,... Abrous,

(2003). Individual differences in cognitive aging: implication of pregnenolone sulfate. Progress in Neurobiology, 71(1), 43–48. doi: 10.1016/j.pneurobio.2003.09.006

73.

Mele,

, Ferreira,

P.G.

, Reverter,

, DeLuca,

D.S.

, Monlong,

, Sammeth,

,... Guigo,

(2015). Human genomics. The human transcriptome across tissues and individuals. Science, 348(6235), 660–665. doi: 10.1126/science.aaa0355

74.

Mohan,

, Mather,

K.A.

, Thalamuthu,

, Baune,

B.T.

, & Sachdev,

P.S.

(2016). Gene expression in the aging human brain: an overview. Curr Opin Psychiatry, 29(2), 159–167. doi: 10.1097/YCO.0000000000000238

75.

Mondelli,

, Vernon,

A.C.

, Turkheimer,

, Dazzan,

, & Pariante,

C.M.

(2017). Brain microglia in psychiatric disorders. Lancet Psychiatry, 4(7), 563–572. doi: 10.1016/S2215-0366(17)30101-3

76.

Montagne,

, Barnes,

S.R.

, Sweeney,

M.D.

, Halliday,

M.R.

, Sagare,

A.P.

, Zhao,

,... Zlokovic,

B.V.

(2015). Blood-brain barrier breakdown in the aging human hippocampus. Neuron, 85(2), 296–302. doi: 10.1016/j.neuron.2014.12.032

77.

Mosher,

K.I.

, & Wyss-Coray,

(2014). Microglial dysfunction in brain aging and Alzheimer’s disease. Biochemical Pharmacology, 88(4), 594–604. doi: 10.1016/j.bcp.2014.01.008

78.

Neumann,

, & Takahashi,

(2007). Essential role of the microglial triggering receptor expressed on myeloid cells2 (TREM2) for central nervous tissue immune homeostasis. Journal of Neuroimmunology, 184(1-2), 92–99. doi: 10.1016/j.jneuroim.2006.11.032

79.

Oldham,

M.C.

, Konopka,

, Iwamoto,

, Langfelder,

, Kato,

, Horvath,

, & Geschwind,

D.H.

(2008). Functional organization of the transcriptome in human brain. Nature Neuroscience, 11(11), 1271–1282. doi: 10.1038/nn.2207

80.

Oliva,

C.A.

, Vargas,

J.Y.

, & Inestrosa,

N.C.

(2013). Wnts in adult brain: from synaptic plasticity to cognitive deficiencies. Frontiers in Cellular Neuroscience, 7, 224. doi: 10.3389/fncel.2013.00224

81.

Paolicelli,

R.C.

, & Gross,

C.T.

(2011). Microglia in development: linking brain wiring to brain environment. Neuron Glia Biology, 7(1), 77–83. doi: 10.1017/S1740925X12000105

82.

Parachikova,

, Agadjanyan,

M.G.

, Cribbs,

D.H.

, Blurton-Jones,

, Perreau,

, Rogers,

,... Cotman,

C.W.

(2007). Inflammatory changes parallel the early stages of Alzheimer disease. Neurobiology of Aging, 28(12), 1821–1833. doi: 10.1016/j.neurobiolaging.2006.08.014

83.

Perry,

V.H.

, & Holmes,

(2014). Microglial priming in neurodegenerative disease. Nature Reviews: Neurology, 10(4), 217–224. doi: 10.1038/nrneurol.2014.38

84.

Peters,

, Josephson,

, & Vincent,

S.L.

(1991). Effects of aging on the neuroglial cells and pericytes within area 17 of the rhesus monkey cerebral cortex. Anatomical Record, 229(3), 384–398. doi: 10.1002/ar.1092290311

85.

Raha,

A.A.

, Henderson,

J.W.

, Stott,

S.R.

, Vuono,

, Foscarin,

, Friedland,

R.P.

,... Raha-Chowdhury,

(2017). Neuroprotective Effect of TREM-2 in Aging and Alzheimer’s Disease Model. Journal of Alzheimer’s Disease, 55(1), 199–217. doi: 10.3233/JAD-160663

86.

Rau,

, Marot,

, & Jaffrezic,

(2014). Differential meta-analysis of RNA-seq data from multiple studies. BMC Bioinformatics, 15, 91. doi: 10.1186/1471-2105-15-91

87.

Raz,

, Rodrigue,

K.M.

, Kennedy,

K.M.

, Head,

, Gunning-Dixon,

, & Acker,

J.D.

(2003). Differential aging of the human striatum: longitudinal evidence. AJNR Am J Neuroradiol, 24(9), 1849–1856.

88.

Ritchie,

M.E.

, Phipson,

, Wu,

, Hu,

, Law,

C.W.

, Shi,

, & Smyth,

G.K.

(2015). limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Research, 43(7), e47. doi: 10.1093/nar/gkv007

89.

Rogers,

J.T.

, Morganti,

J.M.

, Bachstetter,

A.D.

, Hudson,

C.E.

, Peters,

M.M

, Grimmig,

B.A.

,... Gemma,

(2011). CX3CR1 deficiency leads to impairment of hippocampal cognitive function and synaptic plasticity. Journal of Neuroscience, 31(45), 16241–16250. doi: 10.1523/JNEUROSCI.3667-11.2011

90.

Sarthy,

, Bae,

N.S.

, Scrafford,

, & Baumann,

(2009). Human RAP1 inhibits non-homologous end joining at telomeres. EMBO Journal, 28(21), 3390–3399. doi: 10.1038/emboj.2009.275

91.

Schafer,

D.P.

, Lehrman,

E.K.

, Kautzman,

A.G.

, Koyama,

, Mardinly,

A.R.

, Yamasaki,

,... Stevens,

(2012). Microglia sculpt postnatal neural circuits in an activity and complement-dependent manner. Neuron, 74(4), 691–705. doi: 10.1016/j.neuron.2012.03.026

92.

Schetters,

S.T. T.

, Gomez-Nicola,

, Garcia-Vallejo,

J.J.

, & Van Kooyk,

(2017). Neuroinflammation: Microglia and T Cells Get Ready to Tango. Frontiers in Immunology, 8, 1905. doi: 10.3389/fimmu.2017.01905

93.

Seidler,

R.D.

, Bernard,

J.A.

, Burutolu,

T.B.

, Fling,

B.W.

, Gordon,

M.T.

, Gwin,

J.T.

,... Lipps,

D.B.

(2010). Motor control and aging: links to age-related brain structural, functional, and biochemical effects. Neuroscience and Biobehavioral Reviews, 34(5), 721–733. doi: 10.1016/j.neubiorev.2009.10.005

94.

Sfeir,

, Kabir,

, van Overbeek,

, Celli,

G.B.

, & de Lange,

(2010). Loss of Rap induces telomere recombination in the absence of NHEJ or a DNA damage signal. Science, 327(5973), 1657–1661. doi: 10.1126/science.1185100

95.

Shammas,

M.A.

(2011). Telomeres, lifestyle, cancer, and aging. Current Opinion in Clinical Nutrition and Metabolic Care, 14(1), 28–34. doi: 10.1097/MCO.0b013e32834121b1

96.

Sheffield,

L.G.

, & Berman,

N.E.

(1998). Microglial expression of MHC class II increases in normal aging of nonhuman primates. Neurobiology of Aging, 19(1), 47–55. doi: 10.1016/s0197-4580(97)00168-1

97.

Shi,

, Colodner,

K.J.

, Matousek,

S.B

, Merry,

, Hong,

, Kenison,

J.E.

,... Lemere,

C.A.

(2015). Complement C3-Deficient Mice Fail to Display Age-Related Hippocampal Decline. Journal of Neuroscience, 35(38), 13029–13042. doi: 10.1523/JNEUROSCI.1698-15.2015

98.

Sibille,

(2013). Molecular aging of the brain, neuroplasticity, and vulnerability to depression and other brain-related disorders. Dialogues in Clinical Neuroscience, 15(1), 53–65.

99.

Sierra,

, Encinas,

J.M.

, Deudero,

J.J.

, Chancey,

J.H.

, Enikolopov,

, Overstreet-Wadiche,

L.S.

,... Maletic-Savatic,

(2010). Microglia shape adult hippocampal neurogenesis through apoptosis-coupled phagocytosis. Cell Stem Cell, 7(4), 483–495. doi: 10.1016/j.stem.2010.08.014

100.

Sloane,

J.A.

, Hollander,

, Moss,

M.B.

, Rosene,

D.L.

, & Abraham,

C.R.

(1999). Increased microglial activation and protein nitration in white matter of the aging monkey. Neurobiology of Aging, 20(4), 395–405. doi: 10.1016/s0197-4580(99)00066-4

101.

Sonar,

, & Lal,

(2015). Role of Tumor Necrosis Factor Superfamily in Neuroinflammation and Autoimmunity. Frontiers in Immunology, 6, 364. doi: 10.3389/fimmu.2015.00364

102.

Spalding,

K.L.

, Bergmann,

, Alkass,

, Bernard,

, Salehpour,

, Huttner,

H.B.

,... Frisen,

(2013). Dynamics of hippocampal neurogenesis in adult humans. Cell, 153(6), 1219–1227. doi: 10.1016/j.cell.2013.05.002

103.

Stephan,

A.H.

, Barres,

B.A.

, & Stevens,

(2012). The complement system: an unexpected role in synaptic pruning during development and disease. Annual Review of Neuroscience, 35, 369–389. doi: 10.1146/annurev-neuro-061010-113810

104.

Stephan,

A.H.

, Madison,

D.V.

, Mateos,

J.M.

, Fraser,

D.A.

, Lovelett,

E.A.

, Coutellier,

,... Barres,

B.A.

(2013). A dramatic increase of Cq protein in the CNS during normal aging. Journal of Neuroscience, 33(33), 13460–13474. doi: 10.1523/JNEUROSCI.1333-13.2013

105.

Stevens,

, Allen,

N.J.

, Vazquez,

L.E.

, Howell,

G.R.

, Christopherson,

K.S.

, Nouri,

,... Barres,

B.A.

(2007). The classical complement cascade mediates CNS synapse elimination. Cell, 131(6), 1164–1178. doi: 10.1016/j.cell.2007.10.036

106.

Streit,

W.J.

, Miller,

K.R.

, Lopes,

K.O.

, & Njie,

(2008). Microglial degeneration in the aging brain–bad news for neurons? Frontiers in Bioscience, 13, 3423–3438. doi: 10.2741/2937

107.

Subramanian,

, Tamayo,

, Mootha,

V.K.

, Mukherjee,

, Ebert,

B.L.

, Gillette,

M.A.

,... Mesirov,

J.P.

(2005). Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proceedings of the National Academy of Sciences of the United States of America, 102(43), 15545–15550. doi: 10.1073/pnas.0506580102

108.

Sun,

L.Y.

, Evans,

M.S.

, Hsieh,

, Panici,

, & Bartke,

(2005). Increased neurogenesis in dentate gyrus of long-lived Ames dwarf mice. Endocrinology, 146(3), 1138–1144. doi: 10.1210/en.2004-1115

109.

Tam,

S.J.

, Richmond,

D.L.

, Kaminker,

J.S.

, Modrusan,

, Martin-McNulty,

, Cao,

T.C.

,... Watts,

R.J.

(2012). Death receptors DR6 and TROY regulate brain vascular development. Developmental Cell, 22(2), 403–417. doi: 10.1016/j.devcel.2011.11.018

110.

Trabzuni,

, Ryten,

, Walker,

, Smith,

, Imran,

, Ramasamy,

,... Hardy,

(2011). Quality control parameters on a large dataset of regionally dissected human control brains for whole genome expression studies. Journal of Neurochemistry, 119(2), 275–282. doi: 10.1111/j.1471-4159.2011.07432.x.

111.

Tricomi,

E.M.

, Delgado,

M.R.

, & Fiez,

J.A.

(2004). Modulation of caudate activity by action contingency. Neuron, 41(2), 281–292. doi: 10.1016/s0896-6273(03)00848-1

112.

Valdes,

A.M.

, Deary,

I.J.

, Gardner,

, Kimura,

, Lu,

, Spector,

T.D.

,... Cherkas,

L.F.

(2010). Leukocyte telomere length is associated with cognitive performance in healthy women. Neurobiology of Aging, 31(6), 986–992. doi: 10.1016/j.neurobiolaging.2008.07.012

113.

van Dyck,

C.H.

, Avery,

R.A.

, MacAvoy,

M.G.

, Marek,

K.L.

, Quinlan,

D.M.

, Baldwin,

R.M.

,... Arnsten,

A.F.

(2008). Striatal dopamine transporters correlate with simple reaction time in elderly subjects. Neurobiology of Aging, 29(8), 1237–1246. doi: 10.1016/j.neurobiolaging.2007.02.012

114.

Vukovic,

, Colditz,

M.J.

, Blackmore,

D.G.

, Ruitenberg,

M.J.

, & Bartlett,

P.F.

(2012). Microglia modulate hippocampal neural precursor activity in response to exercise and aging. Journal of Neuroscience, 32(19), 6435–6443. doi: 10.1523/JNEUROSCI.5925-11.2012

115.

Wake,

, Moorhouse,

A.J.

, Miyamoto,

, & Nabekura,

(2013). Microglia: actively surveying and shaping neuronal circuit structure and function. Trends in Neurosciences, 36(4), 209–217. doi: 10.1016/j.tins.2012.11.007

116.

Wu,

, Bendriem,

R.M.

, Garamszegi,

S.P.

, Song,

, & Lee,

C.T.

(2017). RNA sequencing in post-mortem human brains of neuropsychiatric disorders. Psychiatry and Clinical Neurosciences, 71(10), 663–672. doi: 10.1111/pcn.12550

117.

Wu,

, Chen,

, Shu,

, & Lee,

C.T.

(2017). Whole-genome expression analyses of type 2 diabetes in human skin reveal altered immune function and burden of infection. Oncotarget, 8(21), 34601–34609. doi: 10.18632/oncotarget.16118

118.

Wu,

, Garamszegi,

S.P.

, Xie,

, & Mash,

D.C.

(2017). Altered Dopamine Synaptic Markers in Postmortem Brain of Obese Subjects. Frontiers in Human Neuroscience, 11, 386. doi: 10.3389/fnhum.2017.00386

119.

Wu,

, Xu,

, Tsai,

S.A.

, Freed,

W.J.

, & Lee,

C.T.

(2017). Transcriptional profiles of type 2 diabetes in human skeletal muscle reveal insulin resistance, metabolic defects, apoptosis, and molecular signatures of immune activation in response to infections. Biochemical and Biophysical Research Communications, 482(2), 282–288. doi: 10.1016/j.bbrc.2016.11.055

120.

Wu,

, Dissing-Olesen,

, MacVicar,

B.A.

, & Stevens,

(2015). Microglia: Dynamic Mediators of Synapse Development and Plasticity. Trends in Immunology, 36(10), 605–613. doi: 10.1016/j.it.2015.08.008

121.

Wyss-Coray,

(2016). Ageing, neurodegeneration and brain rejuvenation. Nature, 539(7628), 180–186. doi: 10.1038/nature20411

122.

Wyss-Coray,

, & Mucke,

(2002). Inflammation in neurodegenerative disease–a double-edged sword. Neuron, 35(3), 419–432. doi: 10.1016/s0896-6273(02)00794-8

123.

Yu,

, Wang,

L.G.

, Han,

, & He,

Q.Y.

(2012). clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS: A Journal of Integrative Biology, 16(5), 284–287. doi: 10.1089/omi.2011.0118

124.

Zhan,

, Paolicelli,

R.C.

, Sforazzini,

, Weinhard,

, Bolasco,

, Pagani,

,... Gross,

C.T.

(2014). Deficient neuron-microglia signaling results in impaired functional brain connectivity and social behavior. Nature Neuroscience, 17(3), 400–406. doi: 10.1038/nn.3641

125.

Zhang,

, Li,

, Purkayastha,

, Tang,

, Zhang,

, Yin,

,... Cai,

(2013). Hypothalamic programming of systemic ageing involving IKK-beta, NF-kappaB and GnRH. Nature, 497(7448), 211–216. doi: 10.1038/nature12143

126.

Zunszain,

P.A.

, Anacker,

, Cattaneo,

, Choudhury,

, Musaelyan,

, Myint,

A.M.

,... Pariante,

C.M.

(2012). Interleukin-beta: a new regulator of the kynurenine pathway affecting human hippocampal neurogenesis. Neuropsychopharmacology, 37(4), 939–949. doi: 10.1038/npp.2011.277