Adult male (10 months old and 20 months old) C57 mice were obtained from Hubei provincial center for disease control and prevention (Wuhan, China). The animals were acclimated for 5-7 days under standard conditions and kept for appropriate age. Ten 10 months old and Ten 20 months old C57 mice were killed, and the brains were quickly removed. Three of them were used for chip assay, and the remaining seven were used for quantitative real time PCR (qRT- PCR) verification in each group. Fresh hippocampi were then dissected from the brain on a chilled glass plate on ice according to the procedures described in reference [17] . The hippocampi were stored at 70˚C until the day of assay. All animal experiments in this study were approved by Medical Ethics Committee of Yangtze University and carried out in accordance with the guiding principles for the care and use of laboratory animals published by the U.S. National Institutes of Health (NIH Publication No.85-23, revised 1996) and the ARRIVE guidelines.

miRNA microarray

Microarray analysis was performed by Kangcheng Bio-tech Inc. (Shanghai, China). Briefly, total RNA was harvested using Trizol (Invitrogen, Carlsbad, CA, USA) and miRNeasy mini kit (Qiagen, Valencia, CA, USA) according to manufacturer’s instructions. After RNA quantity measurement with the NanoDrop 1000, the samples were labeled using the miRCURYTM Hy3TM/Hy5^TM Power labeling kit and hybridized on the miRCURYTM LNA Array (v.16.0, Exiqon, Skelstedet, Vedbaek, Denmark). After hybridization, scanning was performed with the Axon GenePix 4000B microarray scanner (Molecular Devices, Downingtown, PA, USA). The raw intensity of the image was read with GenePix pro V6.0 (Molecular Devices) and the intensity of green signal was calculated after background subtraction. Four replicated spots of each probe on the same slide were averaged. Median Normalization Method was used to obtain “Normalized Data”: Normalized Data = (Foreground − Background)/median, where the median was the 50% quantile of miRNA intensity, which was larger than 50 in all samples after background correction.

Total RNA was extracted using Trizol reagent (Invitrogen). cDNA was synthesized using the PrimeScript RT Reagent Kit (GeneCopoeia, MD, USA). A total of 6 differentially expressed miRNAs were selected based on their function and involvement in pathways and processes important to C57 mouse ageing. Detection of the mature form of miRNAs was performed using Quantitect SYBR Green PCR Kit (GeneCopoeia) and qRT-PCR Primer Sets (GeneCopoeia) with the U6 small nuclear RNA as an internal control.

Up and down-regulated miRNAs of greater than 1.5 fold change from the 20- month-old and 10-month-old were selected from miRNAs expression profiling data (Intersection of three sets of chip results).

Three types of miRNA target prediction software, i.e., Microcosm (http://www.ebi.ac.uk/enright-srv/microcosm/htdocs/targets/v5/), Miranda (http://www.microrna.org/microrna/home.do), Mirdb (http://mirdb.org/miRDB/) were used to predict the target genes of the selected miRNAs. The intersection of these three datasets was used as the prediction results of the target genes of the selected miRNAs.

The hierarchical clustering method [18] was used to classify different group patterns. Principal component analysis (PCA) [19] was used to produce a two-dimensional graph of the distances between different groups.

Gene Ontology (GO) classification systems was used to assign putative function to each clone by way of biological process, molecular function and cellular components. The Database for Annotation, Visualization and Integrated Discovery (DAVID) v6.7b [20] was used to determine pathways and processes of major biological significance and importance through the Functional Annotation Cluster (FAC) tool based on the GO annotation function.

DAVID FAC analysis was conducted on two independent normalized gene lists containing the target genes (Intersection of 5 or more miRNAs) of 1.5-fold up-regulated normalized miRNAs and 1.5-fold down-regulated normalized miRNAs (5 or more). High stringency ease score parameters were selected, to indicate confident enrichment scores of functional significance and importance of the given pathways and processes investigated. The Gene Ontology (GO) system in DAVID was utilized to identify enriched biological themes in both gene lists.

Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway tools were used to visually map cluster of the target genes involved in common pathways and processes for both pathway-specific and molecular overview purposes. KEGG pathway tools were utilized through DAVID online tools. Since visual mapping was the primary objective, all target genes (Intersection of 5 or more miRNAs) of normalized miRNAs that were differentially expressed by 1.5-fold were considered for the KEGG pathway analysis. Heat map analyses were also conducted through DAVID to produce a matrix of enriched GO terms with common target genes (Intersection of 5 or more miRNAs) of normalized miRNAs that were 1.5-fold or more up and down-regulated. The green and black shading on the heat map matrix indicates a positive and negative correlation between the enriched GO term and the given target gene, respectively.

The 6^th generation of miRNA array (Exiqon) contains about 1280 capture probes, covering mouse microRNAs annotated in miRBase 16.0. C57 mice hippocampi of 10 month old and 20 month old were used for miRNAs microarray analysis. Up and down-regulated miRNAs with change more than 1.5 fold from both 20-month-old and 10-month-old (Table 1, Table 2) were chosen for further study.

In total, two replicate microarray hybridizations were performed, and 26 differentially expressed miRNAs (>1.5 fold) were identified through analysis. Of these miRNAs, 16 were up-regulated and 10 down-regulated during hippocampus ageing. The intersections of the predicted target genes from 5 or more selected miRNAs’ were analyzed using the Functional Annotation Cluster (FAC) tool contained in the Database for Annotation, Visualization and Integrated Discovery (DAVID) [20] . DAVID FAC analysis of the 132 predicted target genes of up-regulated miRNAs (>1.5-fold) produced a total of 31 enriched functional clusters under high stringency conditions (Enrichment Score > 1.5). The enrichment score gives an indication of the biological significance of the gene groups being analyzed, and the top 10 considered in our study are cell fraction, synaptic function, regulation of secretion, apoptosis, response to hormone stimulus, protein localization, neuron projection, synaptic transmission, regulation of kinase activity and regulation of apoptosis (Figure 1). DAVID FAC analysis of the 268 predicted target genes of down-regulated miRNAs (>1.5-fold) produced a total of 44 enriched functional clusters under high stringency conditions (Enrichment Score > 1.5). The top 10 were shown in Figure 2. Some significant biological processes were same with the target genes of up-regulated miRNAs, maybe just be regulated and regulation in the opposite direction. Considering the negative regulation is the main characteristic of miRNA function, target genes of up-regulated miRNAs (>1.5-fold) were chosen for further analysis.

Two microarray sets may not be enough to convince differential expression in adult and aged hippocampi, so qRT-PCR was used to validate microarray results. A total of 6 differentially expressed miRNAs were selected based on their function and involvement in pathways and processes important to C57 mice ageing. The qRT-PCR results are correlated with the microarray expression data (Figure 3, Table 1, Table 2).

Synaptic function activity showed a very high enrichment score in the FAC analysis of target genes of up-regulated miRNAs in 20 month old C57 mouse

(10 m: 10 months old; 20 m: 20 months old).

(Figure 1). DAVID heat map analyses identified 16 genes that functionally clustered into common GO terms related to synaptic transmission, synaptic plasticity, behavior, learning and memory, cognition and cell fraction (Figure 4(A)). A number of genes that were identified in the heat map, such as N-methyl-D-aspartate

(10 m: 10 months old; 20 m: 20 months old).

subtype of glutamate receptors (NMDA-R), brain derived neurotrophic factor (BDNF), dopamine receptor D1A, early growth response 1, solute carrier family 24, and tachykinin receptor 1, were found highly related to learning or memory, cognition, behavior, regulation of transmission of nerve impulse, synaptic transmission, regulation of neuronal synaptic plasticity, regulation of neurological system process and membrane fraction. Especially, N-methyl-D-aspartate subtype of glutamate receptors (NMDA-R) is known to play a major role in the induction of synaptic plasticity as well as in the acquisition of memory traces [21] . Impaired long-term potentiation (LTP) expression at hippocampal synapses of the aging brain has recently been linked to weaker NMDA-R activation [22] . And the neurotrophin BDNF has been shown to modulate the development and function of synapses in the nervous system [23] . BDNF is important in modulating dentate gyrus neurogenesis [24] and in synaptogenesis [25] .

These results suggest differentially expressed miRNAs were involved in the regulation of hippocampus synaptic function by regulating the expression of their target genes.

Cell fraction showed the highest enrichment score in the FAC analysis with the target genes of up-regulated miRNAs (>1.5-fold) during the ageing (Figure 1). DAVID two dimensional figure analysis shows that 42 genes are correlated with cell fraction, membrane fraction, etc. of GO classifications (Figure 4(B)). Dopamine receptor D1A, metabotropic and ionotropic glutamate receptors (GluRs), tumour necrosis factor receptor (TNFR) etc. appeared in these genes. Regulation of secretion is another important function cluster (Figure 1), and 22 genes that functionally clustered into common GO terms pertinent to regulation

of secretion, regulation of transport and regulation of position (Figure 4(C)). Among them, dopamine receptor D1A as an important part of the dopamine system, together with the catecholamine neurotransmitter dopamine, participate in motor function, learning and memory function, brain aging, and many other physiological activities. GluRs play vital role in learning, memory, neuronal plasticity, brain development and aging, etc., when bound with the important neurotransmitter glutamate.

FAC analysis identified apoptosis as an important biological process during hippocampus ageing (Figure 1). DAVID heat map analyses identified B-cell translocation gene 2 (Bcl2), transforming growth factor (TGF), transforming growth factor-β receptor 1 (TGF-β R1), presenlin 1, SMAD family member 3 (SMAD3), BDNF, caspase 3, B-cell translocation gene 2 etc. that functionally clustered into common GO terms related to apoptosis and other associated processes such as regulation of apoptosis, anti-apoptosis, regulation of cell death, positive and negative regulation of programmed cell death (Figure 4(D)).

Apoptosis suits itself into antagonistic pleiotropy theory and as it has deleterious effects in aging neurons while it is needed during development where it removes neurons that don’t integrate into the growing neuronal network [26] . However, after the development of the neuronal network into the mature nervous system, the fight of neurons for survival and healthy existence with plasticity becomes anti-apoptotic [27] . In the mitochondrial pathway of apoptosis, intracellular signals result in releasing cytochrome c from mitochondria, which binds to the adaptor protein APAF-1 (apoptotic protease-activating factor-1), which further leads to the activation of Caspase 9 and subsequently Caspase 3 [28] .

Bcl2 as a pro-survival gene is a member of the anti-apoptotic group, which binds to and inhibits the action of multi-domain pro-apoptotic proteins like Bax [29] . Moreover, the up-regulation of Bcl2 expression leads to down-regulation of Bax, and interferes with the release of cytochrome c from the mitochondrial intermembrane space into the cytosol, where it associates with Caspase 9 and Apaf1 to form the apoptosome complex, thus effecting apoptosis [30] .

Therefore the regulation of miRNAs that target key genes such as Bcl2 expression may be interpreted as potential therapeutic candidates for preventing apoptotic death in neuronal cells, thus contributing to cognitive robustness.

DAVID FAC analysis indicated that response to hormone stimulus and protein localization is active in both adult and aged hippocampus (Figure 1, Figure 2). FAC analysis identified responses to hormone stimulus, endogenous stimulus, estrogen stimulus, estradiol stimulus, organic substance and alkaloid as significant biological processes during hippocampus ageing, dopamine receptor D1A, ras homolog gene family, insulin-like growth factor 2, TGF-β R1, regulator of calcineurin 1, tachykinin receptor 1, adenylate cyclase 6, insulin induced gene 1 etc. were identified in the DAVID heat map (Figure 4(E)). DAVID heat map analysis identified 27 target genes of up-regulated miRNAs (Figure 4(F)) and 29 target genes of down-regulated miRNAs (not shown) involved in protein localization. These include dopamine receptor D1A, NMDA-R, mitogen-activated protein kinase 9 (MAPK9) etc.

Several in vitro studies demonstrate potent neurotrophic actions of IGF-1 and IGF-2 in both neurons and glia including stimulation of DNA and RNA synthesis, induction of neurite outgrowth, regulation of neurotransmitter release, synaptogenesis and protection against neurotoxic insults [31] [32] . During hippocampus ageing, organizations secretion, such as hormone will be changed, and miRNAs expression will be changed too, they would involve in tissue autocrine and response to secretion through the regulation of their targets.

DAVID heat map analyses also identified a cluster of predicted target genes found to be involved in neuron projection, synaptic transmission and regulation of kinase activity (not shown). Neuron projection and synaptic transmission are very important physiological activities in CNS; therefore these functions also show significant changes associated with ageing, of course, the related important cytokines will be regulated by miRNAs.

The important GO terms of FAC analysis were chosen for KEGG pathway analysis. Among them, neurotrophin signaling pathway, which widely exists in brain tissue, plays an important physiological role in all kinds of neural cells growth, differentiation, maintenance and regeneration (Figure 5(A)) (http://www.genome.jp/kegg-bin/show_pathway?mmu04722+12064). Long term potentiation (LTP) related signal pathways, widely existing in brain hippocampus, cortex, limbic system, cerebellum, midbrain, thalamus and other brain regions, are closely related to synaptic plasticity, learning and memory process (Figure 5(B)) (http://www.genome.jp/kegg-bin/show_pathway?mmu04720+14811). Cell apoptosis signaling pathway, which is closely related to the hippocampal growth and aging (Figure 5(C)) (http://www.genome.jp/kegg-bin/show_pathway?mmu04210+12043); TGF-β signaling pathway, which can regulate multiple cellular functions, including cell growth, adhesion, cell transfer, differentiation and apoptosis, and plays an important role in regulating immune inflammation, wound healing, immune steady and tolerance (Figure 5(D)) (http://www.genome.jp/kegg-bin/show_pathway?mmu04350+21808). Dopamine synaptic pathway, which participates in the control of the movement in CNS, plays an important role in regulating brain development, ageing, learning and memory function (Figure 5(E))

(http://www.genome.jp/kegg-bin/show_pathway?mmu04728+13488). In the KEGG pathway maps, red colour was used to mark the predicted target genes of

miRNAs and differentially expressed miRNAs (Figure 5, Table 3). If further studies confirm these differentially expressed miRNAs (Table 3) could combine with mRNA 3’UTR sequence of predicted target genes and inhibited the expression of target genes, that suggests these miRNAs are involved in regulation of the above mentioned important signaling pathways by inhibiting the expression of the important cytokines, and attend in regulated hippocampal growth, ageing and maintenance, and then regulate behavior, emotion, learning and memory function.