Comprehensive Correlation Analysis Of MiRNA-mRNA-protein In Human Peripheral Blood Mononuclear Cells

Peripheral blood mononuclear cells(PBMC)are cells with mononuclear cells in peripheral blood,including lymphocytes and monocytes.It is often used as an important target cell in the fields of immune diseases(rheumatoid arthritis,systemic lupus erythematosus),infectious diseases and cancer research.Studying the multi-omics regulatory networks and correlation patterns in PBMCs can provide evidence for explaining the occurrence of PBMC-related diseases caused by abnormal gene expression,on the other hand,it can find potential biomarkers or drug target for PBMC-related diseases such as cancer and immune diseases.However,the current multi-omics studies in PBMC are more focused on the genome or the correlation between miRNA and mRNA.Therefore,we designed this study to explore the interrelated network of miRNA,mRNA and protein at the omics level in PBMC.The study can be divided into three sections:the first section we perform the Pearson correlation analysis between mRNA expression level and the corresponding protein expression level at the all-to-all level,individual level,genetic level,and biological function category level.It comprehensively expounds the ability of mRNA expression to predict protein expression;Section II establishes the overall correlation pattern and network between miRNA and protein;Section III establishes the miRNA-mRNA-protein association network through integration analysis.A causal inference test was performed to find the miRNA-mRNA-protein causal chain,with which we can explain miRNA-mediated protein expression.Aim:The purpose of this study is to investigate the interrelationship between miRNA,mRNA,and protein in PBMC at the omics level,and to construct a miRNA-mRNA-protein interconnection network in PBMC,with which to identify the miRNAs that play a key regulatory role in the network.Methods:We extracted PBMC from 28 Chinese female subjects.Protein expression levels were quantified by Label-free protein quantitation method.mRNA expression profiling was performed by using the Affymetrix mRNA 4.0 Array.miRNA expression profiling was performed by using the Affymetrix miRNA 4.0 Array.Genes that are expressed at both mRNA level and protein level in all samples were screened out,and the genes selected were subjected to GO(gene ontology)enrichment analysis.Next,we performed Pearson correlation analysis between the mRNA and protein expression levels at all-to-all,individual,gene,and biological function category level of these 669 genes to investigate the ability to predict protein expression by mRNA expression level.Next,the target genes of miRNAs were screened in the database,and we carried out GO enrichment analysis of the selected genes.Finally,we performed a four-level Pearson correlation analysis of these miRNA-protein pairs.miRNA-mRNA-protein networks were established through an integrative analysis.Then causal inference test was followed to detect miRNA-mediated effects on protein expressions.Results:For all genes studied,we observed a strong positive correlation between mRNA and protein(r=0.26,p<0.0001).On a genomic scale of 28 individual samples,correlation for all the studied genes were significant and positive.The strength of correlation varies among different GO terms.The correlation of genes in terms of vesicle membrane(GO:0012506),the r-squared value reached a maximum of 0.766.While in positive regulation of organelle organization(GO:0010638)it is only 0.008.However,for each gene only 33 studied genes had significant positive correlation across all the samples.All in all,our study found a positive correlation between mRNA and protein expression levels.In the study of miRNA-protein correlation patterns,Pearson correlation analysis between miRNA and protein showed a negative but relatively small global correlation in each subject.Among the 371 constructed miRNA-protein pairs(60 unique miRNAs,and 150 unique proteins),32(8.6%)pairs have significant correlations(Padj<0.05).Some highlighted miRNAs(e.g.,hsa-miR-590-3p,hsa-miR-34c-5p)exerted significant regulation on multiple genes.Simultaneously,some genes(e.g.,HSP90B1)were targeted by multiple miRNAs.The genes associated with miRNAs tend to enrich in some important GO terms:biological processes(e.g.,gene expression,protein binding and RNA binding),and molecular functions(protein binding:GO:0005515;RNA binding:GO:0003723).At last,about 66 trios corresponding to 44 proteins,17 miRNAs and 25 mRNAs fulfill miRNA-mRNA-protein interactions,among which 3 significant causal chains of miRNA-mRNA-protein were identified by causal inference tests(CIT),highlighting the intermediate effects of 2 key miRNAs.For miRNA hsa-miR-4448,it has two correlations between mRNA and protein.This miRNA plays an important role in the process of protein metabolism,cellular and tissue homeostasis.Conclusion:Our study revealed an overall positive correlation between mRNA and protein expression levels in PBMC.However,the correlation strength is not strong and varies greatly among different biological function classifications,which indicates that there is a limitation in using mRNA expression levels to predict protein expression levelsIn addition,we provided a global view of the miRNA-protein expression correlation profile in human PBMCs,which would facilitate in-depth investigation of biological functions of key miRNAs/proteins and better understanding of the pathogenesis underlying PBMC related diseases.Finally,a multi-omics integrated network analysis and causal inference test were used to construct a miRNA-mRNA-protein interconnected network in PBMC.At the same time,three regulatory chains that confirmed causality were also found.This provides evidence to explain the occurrence of PBMC-related diseases caused by abnormal gene expression.On the other hand,it can find potential biomarkers or drug targets for PBMC-related diseases such as cancer and immune diseases.