Transcriptomic Analysis Of Depressive Model Brain Tissue Compared With A Database-based Metabolite System

Background:Major depressive disorder(MDD)is a complex psychiatric disorder with high prevalence and suicide rates,which is a major contributor of global burden of disease.Currently,MDD is considered a multifactorial disorder,influenced by a variety of stress,genetic,environmental and psychological factors.Stress is often a major factor in the onset and exacerbation of MDD.The primary source of stress experienced by humans and most mammals is social in nature.The chronic social defeat stress(CSDS)model,in which mice are repeatedly exposed to socially defeating stress,mimics this social source of stress and is a good model for conducting research on the disease mechanisms of MDD.Most previous studies have focused on the role of independent brain region and single gene in disease mechanisms.However,increasing evidences from imaging and autopsy samples of patients with depression indicates that MDD may involve abnormalities in multiple functions in multiple brain regions.Therefore,using transcriptomics to explore the gene expression profiles of multiple brain regions may provide reliable research clues to the complex synergistic regulatory mechanisms of depression.Objective:The purpose of this study is to compare the transcriptional expression profiles of four brain regions in depressive-like mice with chronic social defeat stress through multiple analyses,to explore the patterns of changes in depression-related gene clusters across brain regions and functional dysregulation under social stress,and to discover the molecular mechanisms of synergistic alterations in multiple brain regions.Methods:1.A chronic social defeat stressed mouse model of depression was established.Then the genome-wide based m RNA sequencing analysis was performed for four brain regions including hippocampus(HIP),habenula nucleus(HA),hypothalamus(HYP)and prefrontal cortex(PFC)from control and depression-like mice.2.Rank-rank hypergeometric overlap test(RRHO)and gene set enrichment analysis(GSEA)were performed to compare the overall transcriptional characteristics and altered biological functions among brain regions.3.Linear model was used to identify differentially expressed genes in four brain regions and IPA pathway analysis was applied to explore the biological functions of these differential genes.4.Weighted gene co-expression network analysis was performed to identify gene modules with significantly altered gene co-expression relationships across brain regions in depressive-like mice.The exploring of the biological functions involved in these modules was accomplished by IPA pathway analysis and cell type enrichment analysis.5.Gene interaction network was constructed using the STRING database to identify potential key regulators.Results:1.Behavioral tests results showed that mice in the stress group had significant weight loss,decreased sucrose preference,decreased social interaction and increased immobility time in forced swimming test.These results indicate that stressed mice showed significant depression-like behaviors compared to control mice.2.The hippocampus and the other three brain regions of depression-like mice showed strong overlapping pattern in the overall transcriptional features and synchronous up-regulation in the biological functions related to neuronal migration,axon and synaptic development as well as synchronous down-regulation in biological functions related to mitochondrial energy metabolism.3.A total of 5,237 genes with differential expression in at least one brain region were identified according to the criteria of p<0.05 and fold change>1.3.Among them,six differential genes shared in all brain regions were significantly upregulated,namely Trpc5,Shc3,Sgcd,Fam205 c,Adgrl3and Adcy1.IPA analysis showed that 29 canonical pathways were coenriched in at least two brain regions,and these pathways showed consistent trends of alteration in four brain regions.Among them,CREB signal in neurons was the pathway that was significantly up-regulated in all brain regions.4.Weighted gene co-expression network analysis identified a gene module with a significantly enhanced co-expression pattern in depressionlike mice.IPA pathway analysis revealed that this module is associated with apoptosis-related pathways including NAD signaling pathway,spliceosome cycle and senescence pathway.Cell type analysis showed that the module was significantly enriched in oligodendrocyte type.5.The protein interaction network constructed based on the genes within this enhanced co-expression module identified 17 potential network regulators,including Notch1,Nipbl,Ep300,Bptf,Kmt2 c,Kdm6b,Cnot1,Traf6,Polr2 a,Thoc2,Tia1,Ranbp2,Kat6 a,Kdm2a,Xrn1,Smad4,Kdm2 a,Xrn1,Smad4,and Dicer1.Conclusions:In this study,we used multiple unbiased data analysis methods to identify synergistic changes in depression-related gene clusters in multiple brain regions.We found significant up-regulation of "CREB signaling in neurons" in the hippocampus,habenula nucleus,hypothalamus,and prefrontal cortex of depressive-like mice under chronic social defeat stress,and depression-related gene module showed enhanced gene co-expression pattern,which may be related to neuronal apoptotic functions.Background:Major depressive disorder(MDD)is a psychiatric disorder caused by various etiologies.Chronic stress models are used to simulate the heterogeneous pathogenic processes of depression.However,few studies have compared transcriptional features between stress models and MDD patients.Objective:The purpose of this study is to elucidate the ability of stress models to recapitulate the mechanisms of MDD by comparing the brain transcriptional characteristics between stress models and depressed patients to explore common dysfunctions and provide clues for further studies of the molecular mechanisms of depression.Methods:1.The hippocampal gene expression profiles of the chronic social defeat stress(CSDS)mouse model of depression constructed in the first part of the work were integrated with raw sequencing data downloaded from public databases for the chronic unpredictable mild stress model of mice(CUMS),the learned helplessness model(LH),and the post-mortem hippocampal tissue of MDD patients.The data were standardized and preprocessed,and uniformly annotated by human genome annotations.2.The comparation of overlapping patterns of the overall transcriptional features of the 3 depression models with those of depressed patients was performed.Gene set enrichment analysis(GSEA)was performed to detect the common altered biological functions.3.The comparation of overlapping patterns of the differential gene expression between 3 depression models and depressed patients was performed and IPA pathway analysis was applied to explore common altered pathways.4.Gene network analysis was performed to explore the reproducibility of co-expressed gene networks of depression patients in three depression models.IPA pathway analysis and cell type enrichment analysis were used to identify the biological functional alterations involved in depressionrelated co-expressed gene networks.Results:1.A total of 10242 mouse–humans one-to-one orthologues were detected in four datasets after gene filter and cross-species gene conversion.2.Overall comparison of transcriptional features showed that CUMS and LH models showed strong transcriptional regulatory similarities with MDD patients.The GSEA analysis identified a series changes of biological functions related to cell fate,inflammation and mitochondrial gene expression.Among thenm,three significantly enriched functions shared by all depression models and MDD patients,which include ensheathment of neurons,IL-17 production and protein-containing complex remodeling.3.A total of 2126,1102,1067 and 263 differentially expressed genes were identified in CUMS model,CSDS model,LH model and MDD patients.Each depression model shared 11.4% to 16.3% of differentially expressed genes with MDD patients.IPA pathway analysis identified four pathways that were significantly enriched in MDD patients and all three mouse models,including CREB signaling in Neurons,PTEN signaling,glioblastoma multiforme signaling and liver fibrosis signaling pathway.4.At the gene network level,each stress model shared 20.9% to 41.6%of co-expressed genes with MDD patients.IPA pathway analysis based on these genes found that axon guidance signaling is the most significantly enriched pathway that was shared by all stress models and MDD patients.The cell type analysis showed that these co-expressed genes were significantly enriched in oligodendrocyte type.Conclusions:The results of this part of the study show that hippocampal transcriptional features of stress models partially overlap with those of MDD patients.The canonical pathways of MDD patients,including ensheathment of neurons,PTEN signaling,and axonal guidance signaling,were shared with all stress models.These findings provide further clues to understand the molecular mechanisms of depression.Background:In view of the findings in Part I and Part II,stress-induced differential expression of genes in depressed mouse brain were significantly correlated with metabolic function,suggesting that metabolites may play an important role in the pathogenesis of depression as the most downstream product of gene expression to achieve function.In view of the rapid development of detection technology of metabolites in the last decade,metabolomics provides unbiased and comprehensive analytical strategies for the screening of diagnostic and therapeutic biomarkers of depression and the study of mechanisms.Numerous researches have been conducted on the metabolomic changes in animal models of depression.However,there is no general agreement about which metabolites exhibit consist changes.Objective:The aim of this study was to identify consistently altered metabolites across models or tissues by collecting,integrating,and systematically comparing metabolomics data of animal model of depression globally.The discovery of the underlying biological functions in the pathogenesis of depression driven by data from large-scale metabolomics studies,can provide reliable clues to the molecular mechanisms of depression at the metabolic level.Methods:1.A total of 5 literature databases including Pub Med,Cochran Library,Web of Science,Embase and Psyc INFO,as well as 5 metabolomics databases including HMDB,Omics Discovery Index,Metabolomics Workbench,Metabo Lights and The Metabolome Xchange were searched for all depression-related metabolomics studies worldwide to update our previously constructed Metabolite Network of Depression Database(MENDA).2.The screening of metabolomics studies conducted on animal models of depression were performed according to the inclusion and exclusion criteria based on the updated database.3.Vote counting analyses were performed to identify consistently upregulated or down-regulated metabolites in the brain,blood,and urine of animal models of depression based on all differential metabolites we included.Results:1.The back-end data of the MENDA was expanded from 464 studies with totaling 5675 metabolites to 768 studies with totaling 12,097 metabolites.2.A total of 241 animal metabolomics studies with a total of 3743 differential metabolites were included based on the updated database.3.Regarding to the brain: Serotonin,dopamine,gamma-aminobutyric acid,norepinephrine,N-acetyl-L-aspartic acid,anandamide,and tryptophan were consistently down-regulated,while kynurenine,myo-inositol,hydroxykynurenine,and the kynurenine to tryptophan ratio were consistently up-regulated.4.Regarding to the blood: Tryptophan,leucine,tyrosine,valine,trimethylamine N-oxide,proline,oleamide,pyruvic acid,and serotonin were consistently down-regulated,while N-acetyl glycoprotein,corticosterone,and glutamine were consistently up-regulated.5.Regarding to the urine: Citric acid,oxoglutaric acid,proline,tryptophan,creatine,betaine,L-dopa,palmitic acid,and pimelic acid were consistently down-regulated,and hippuric acid was consistently upregulated.6.We also identified consistently altered metabolites in the secondary analyses of hippocampus,prefrontal cortex,serum,and plasma,such as kynurenine,glutamine,tryptophan and leucine.Conclusions:These findings suggested that metabolomic changes in depression models are characterized by decreased neurotransmitter and increased kynurenine metabolite levels in the brain,decreased amino acid and increased corticosterone levels in blood,and imbalanced energy metabolism and microbial metabolites in urine.This study contributes to existing knowledge of metabolomic changes in depression and revealed that the reproducibility of candidate metabolites was inadequate in previous studies.