Screening Differentially Expressed Genes In EBV-associated Lymphoma Model

Objective: To detect and analyse differentially expressed genes of the host cellgenome between EBV-associated lymphoma in vivo in Scid mouse and normal humanlymphocytes by cDNA microarray techniques, to screen important candidate genesrelated to the lymphoma induced by EBV, and to investigate the mechanism of theoccurrence of EBV associated lymphoma.Method: Human lymphocytes were separated from fresh peripheral blood ofhealth donors, replicated EBV-associated lymphoma model in Scid mouse in vivo.Genes of two groups of samples were detected by4×44K Agilent whole GenomeMicroarray. Then obviously different genes were analyzed by comparing the host cellgenome bwteen EBV-associated lymphomas and normal lymphocytes respectively attranscriptional level of whole human genome, with fold change≥2as obviousup-regulated genes and fold change≤0.5as obvious down-regulated genes.Result:1. Tumors-induced in Scid mouse were diagnosed as diffuse large B-celllymphoma by pathological examination. Moreover, PCR showed that theEBV-associated lymphoma is human-derived; successful replicated theEBV-associated tumor model in vivo. Using LIMMA, BRB-Random Variance Model,SAM analysis for high-thoroughly microarray data, differential gene expressionprofiles of the host cell genome were obtained between six homologousEBV-associated lymphomas (T) and normal lymphocytes (N). A test (PDR<0.0001)estimated that3928probesets measured noticeably changed expression betweenNorms and Tumors, after removing probes unmapped to HUGO, selected thefold-change2-fole as to express significant difference, selected202differentiallyexpressed genes, including44up-regulated genes and158down-regulated genes,constructed differential gene expression profiles of six cases of homologous EBV-associated lymphoma (T) in vivo and normal lymphocytes (N). By unsupervisedhierarchical clustering and analysis, differential gene expression profiles clearlydistinguished EBV-associated lymphoma from normal lymphocyte. GOcategorization, KEGG metabolic pathways, Biocarta, Reactom regulatory pathwayand DAVID online software were applied to functional annotate and pathway analyzedifferentially expressed genes, combined with the STRING and Cytscape analysis ofthe interaction of differentially expressed genes, to predict the biological function ofthe differentially expressed genes.2. The result of systematically bioinformatic analysis of the differentialexpresssed genes files.Gene Ontology classfication analysis had shown, there were30up-regulatedgenes participated in27BP categories, in which “cell cycle”、“cell cycle phase”、“nuclear division”、“mitosis”、“M phase of mitotic cell cycle”、“cell cycle process”、“mitosis”、“nuclear division”、“organelle fission”、“M phase” and “mitotic cellcycle” had the lowest EASE score, all <2.6E-09. Up-regulated genes CDC6, KIFC1,OIP5, NCAPG, KIF15, BUB1, CDCA2, AURKA, CEP55, PBK involved in a numberof biological processes related to cell cycle.126down-regulated genes participated in116GO BP categories, in which “inflammatory response”,“response to wounding”,“immune response”,“defense response”,“taxis”,“chemotaxis”,“regulation ofsecretion”,“behavior”,“anti-apoptosis” and “protein kinase cascade” had the lowestEASE Score, all <1.6E-04. Down-regulated genes CSF2, CCL2, CXCL5, CXCL2,TLR2, FCN1, LILRA5, IL1RAP, IL1B, THBS1, CFD, PTX3, FCGR3A, IL1A,IL1RN and so on were mainly related to cell wounding, inflammatory response,immune response and other biological processes.There were25up-regulated genes participated in13GO MF categories, whichwere mainly related to nucleotide binding activity, such as “ATP binding”,“adenylribonucleotide binding”,“adenyl nucleotide binding”,“purine nucleoside binding”and “nucleoside binding” had the lowest EASE Score, all <0.006. Up-regulated genesCDC6, KIFC1, KIF15, BUB1, AURKA, PBK, TOP2A, GSG2, RAD51wereninvolved in several molecular functions. There were123up-regulated genes participated in12GO MF categories, which were mainly related to molecular signaland cytokine receptor binding, such as “carbohydrate binding”,“cytokine binding”,“polysaccharide binding”,“protein binding”,“glycosaminoglycan binding”,“cytokine activity”and “interleukin-1receptor binding” had EASE Score <0.001.Down regulated genes SELP, TNFAIP6, CCL2, C6ORF25, TLR2, PTX3, THBS1,NLRP3, IL1RN, IL1B, IL1A were involved in several molecular functions.The pathway analysis had shown, through the database of “KEGG Pathway”,“BioCarta” and “Reactome”, in up-regulated genes, no one participated in “KEGGPathway”,2genes participated in1“BioCarta-pathway”, and5genes participated in1“Reactome-pathway”. In down-regulated genes, there were64genes participated in7“KEGG-pathway”,30genes participated in1“BioCarta-pathway” and33genesparticipated in2“Reactome-pathway”. All these pathways have the EASE Score<0.05.The result of DAVID “Functional Annotation Clustering” analysis had shown,there were15pathway and GO classes enriched for up-regulated genes, which had thehighest Enrichment Score classes described aspects of cell cycle and cell division;there were63pathway and GO classes enriched for down-regulated genes, which hadthe highest Enrichment Score classes described aspects of chemotaxis, intrinsic tomembrane and protein binding activity.By conbinating STRING, Cytoscape, PATHWAY and GO categorization toanalyze202differentially expressed genes, include44up-regulated and158down-regulated genes, as well as fuction analyze and prediction, up-regulated genesCDC6, KIFC1, OIP5, NCAPG, KIF15, BUB1, CDCA2, AURKA, CEP55and PBKwere maily participated in cell cycle, mitosis and other biological process;up-regulated genes TNFRSF13B, TNFRSF17, CXCL9and down-regulated genesCSF2, CCL2, FOS, EGF, IL1A, IL1B and DUSP6were closely related to tumorassociated signal pathways, such as Inflammation, Angiogenesis, MAPK signalpathway, Adherens junction, NOD-like receptor signaling pathway. Down-regulatedgenes CSF2, CCL2, CXCL5, CXCL2, TLR2, FCN1, LILRA5, IL1RAP, IL1B,THBS1, CFD, PTX3, FCGR3A, IL1A and IL1RN were mainly participated in response to wounding, inflammatory response, immune response and other molecularfuctions. The result of combination of these bioinformatics analysis had shown,15differentially expressed genes were in a central location in protein network, EGF,IL1B, PBK, CSF2and TLR2were in front of the ranking of degree and betweenness,therefore we speculate that EGF, IL1B, of PBK, CSF2, and TLR2may be the keymolecules which leading to the accurrence of EBV-associated lymphoma.Integrated signaling pathways, gene biological classification, gene interactionanalysis and literatures that has been reported, suggesting EBV may up regulate PBKto promote cell proliferation, down regulate EGF to paly a role of anti-apoptosis anddown regulate IL1β，CSF2，TLR2to decrease cell immunity, leading to the occurrenceof EBV-associated lymphoma.Conclusion:1. We have systematic constructed differential gene expression profiles of six casesof homologous in vivo EBV-associated lymphoma (T) and normal lymphocytes(N), we have indicated there are obviously difference in genes expressed betweenEBV-associated lymphomas and normal lymphocytes.2. Bioinformatics analysis identified202differentially expressed genes, including44up-regulated genes and158down-regulated genes, indicate the process of theoccurrence of EBV-associated lymphoma refers to multiple genes and pathwayparticipated, and including the interacitve of virus genome and host genome.3. We suppose that EGF, IL1B, CSF2，PBK and TLR2are the target molecule inEBV-associated lymphoma.