In the present study, 2 healthy male volunteers (aged 38 and 40 years) were enrolled at The King Chulalongkorn Memorial Hospital, Faculty of Medicine, Chulalongkorn University (Bangkok, Thailand). The present study was approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University (Bangkok, Thailand) (approval no. 358/46) and both subjects provided written informed consent for participation.

PBMCs were isolated from healthy donor blood according to the manufacturer's protocol using density gradient separation with Isoprep (Robbins Scientific Corporation). PBMCs were adjusted to 1x10⁶ cells/ml with R10 medium; RPMI-1640 supplemented with 100 U/ml penicillin and 100 U/ml streptomycin (all from Gibco; Thermo Fisher Scientific, Inc.) and 10% heat-inactivated FBS (BioWhittaker, Inc.). For the EBV-containing culture supernatant, the marmoset B-lymphoblastoid cell line (B95-8) were maintained in R10 media at an exponentially growing rate in a humidified incubator at 37˚C with 5% CO₂. After 3 days, the supernatant was collected and centrifuged at 300 x g for 10 min at 4˚C to separate the EBV-containing culture supernatant from the cells. The supernatant was filtered through a 0.45 µm filter, aliquoted and stored at -80˚C. Typically, the culture supernatant contains >100-1,000 transforming units per ml (17). EBV-transformed BLCL were generated by incubating 1x10⁶ cells/ml PBMCs with 1 ml EBV supernatant from the B95-8 cell line for 1 h at 37˚C with 5% CO₂. BLCL were maintained in R20 medium; RPMI-1640 supplemented with 100 U/ml penicillin, 100 U/ml streptomycin and 20% heat-inactivated FBS, with 1 mg/ml cyclosporin A. The B95-8 cell line was kindly provided by The National Institute of Health, Thailand. Following B cell transformation, the BLCL were used for RNA extraction or cryopreserved for future use.

Total RNA was extracted from the cells mentioned above using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. RNA purity and integrity qualification were evaluated on a ND-1000 Spectrophotometer (NanoDrop Technologies; Thermo Fisher Scientific, Inc.) and an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc.). RNA samples were used for RNA labeling and hybridization using an Agilent One-Color Microarray-Based Gene Expression Analysis protocol (version 6.5; Agilent Technologies, Inc.). Briefly, 100 ng total RNA from each sample was linearly amplified using the oligo dT-promoter primer included in the kit, and labeled with Cy3-dCTP and Cy5-dCTP at 40˚C for 2 h (Agilent Technologies, Inc.) for BLCL and normal B cells, respectively. Following purification, the concentration and specific activity of the labeled cRNAs (pmol Cy3/µg cRNA) were measured using a NanoDrop ND-1000.

A transcriptome database of each RNA sample was created using Agilent SurePrint G3 Human GE 8x60 K Microarray kit (Agilent Technologies Inc.). Briefly, 600 ng of each labeled cRNA was fragmented by adding 5 µl 10x blocking agent and 1 µl 25x fragmentation buffers, and then heated at 60˚C for 30 min. Finally, 25 µl 2x GE hybridization buffer was added to dilute the labeled cRNA. Hybridization solution (40 µl) was dispensed into the gasket slide and assembled into the Agilent SurePrint G3 Human GE 8X60K, V3 Microarrays. The slides were incubated for 17 h at 65˚C in an Agilent hybridization oven and then washed at room temperature using the Agilent One-Color Microarray-Based Gene Expression Analysis protocol. The hybridized array was immediately scanned with an Agilent Microarray Scanner D (Agilent Technologies, Inc.)

Raw data were extracted using Agilent Feature Extraction Software (version 11.0.1.1; Agilent technologies, Inc.). The raw data for the same genes were then summarized automatically using the Agilent feature extraction protocol to generate a raw data text file, providing expression data for each gene probed on the array. Array probes with Flag A in the samples were filtered out. The selected gProcessedSignal value was log-transformed and normalized using the Quantile method (18). Gene-Enrichment and Functional Annotation analysis for the significant probe list was performed using g: Profiler, a web server for functional enrichment analysis (19) and Kyoto Encyclopedia of Genes and Genomes (KEGG) (20).

R version 2.15.1 software (21) was used for image and statistical analysis. Statistical significance of the expression data was determined using the fold-change and local-pooled-error test for identifying significantly differentially expressed genes in microarray experiments in which the null hypothesis was that no difference existed between 2 groups. Hierarchical cluster analysis was performed using complete linkage and Euclidean distance as a measure of similarity. Pearson's correlation (scatter plot) and multidimensional scaling plot (2-D graphics) were used to analyze the degree of reproducibility and similarity between samples, respectively. Log2 fold-change values >4.0 and <-4.0 were set as the cut off for significantly upregulated and downregulated genes, respectively. False discovery rate adjusted P<0.05 was considered to indicate a statistically significant Gene Ontology (22,23) and KEGG pathway.

Following the infection of PBMCs and the transformation of B cells into BLCL, the BLCL grew as small cell clusters that were visible by light microscopy within a week (Fig. 1A). With continued cell culturing, microscopic clusters became larger and single cells exhibited large nuclei and numerous vacuoles on weeks 2 and 3 (Fig. 1B and C). The highest degree of cell clumping, cell elongation, cell trajectory and formation of rosette shapes was visible macroscopically on week 4 (Fig. 1D). The infection rate was ~100%, at this time point, and the B cell transformation was considered successful (17,24-26); BLCL were considered ready for RNA extraction.

Microarray analysis was used to differentiate gene expression profiles between normal B cells (S1_A and S2_A) and EBV-transformed BLCL (S1_B and S2_B). A multidimensional scaling plot was used to investigate the similarities amongst samples. Of note, the two EBV-transformed BLCL datasets exhibited similar RNA expression profiles that were represented by two points (S1_B and S2_B; Fig. 2B), whereas the two normal B cell datasets exhibited a long distance of RNA expression profiles as represented by two separated points (S1_A and S2_A; Fig. 2B). The scatter plots and correlation analysis showed a close correlation between S1_B and S2_B and S1_A and S2_A (Fig. 2C).

Following data normalization, heat map clustering was plotted to express the distance similarity between normal B cells and EBV-transformed BLCL (Fig. 2A). A log2 fold-change of >4.0 and <4.0 were set as the cut off points for the differential gene expression analysis of up and downregulated genes, respectively. A total of 1,018 and 1,061 upregulated, and 1,169 and 1,306 downregulated genes were identified in samples S1_B and S2_B, respectively. There were 623 overlapping upregulated genes with a >4.0 fold-change and 503 overlapping downregulated genes with a <4.0 fold-change in the two datasets (S1_B and S2_B) (Table SI). The top 30 most up and downregulated genes are shown in Table I.

Gene-enrichment and functional annotation analysis were performed using KEGG pathway and gene ontology analysis in g:Profiler. A total of 623 overlapping upregulated genes were used for complete gene ontology analysis; 178 terms in biological process (BP), 37 terms in cellular component (CC), 24 terms in molecular function (MF) and 15 KEGG pathways were identified. The top 10 BP, CC, MF and KEGG pathways are shown in Fig. 3, and the top 3 pathways of the 30 most upregulated genes and gene lists are shown in Table II. The most significant BPs included immune system process, cell activation, response to stimulus and leukocyte activation. The most significant CCs included cell periphery, plasma membrane and plasma membrane part. The most significant MF pathways included enzyme binding, cytokine binding and molecular transducer activity. The most significant KEGG pathways included osteoclast differentiation, cytokine-cytokine receptor interaction, hematopoietic cell lineage and pathways involved in cancer.

Functional and pathway analysis results of 503 downregulated genes are shown in Fig. 4. There were 250 terms in BP, 112 terms in CC, 28 terms in MF and 6 KEGG pathways identified by g: Profiler. The most significant BPs included cell cycle, mitotic cell cycle process and mitotic cell cycle. The most significant CCs included chromosome, chromosomal region and chromosomal part. The most significant MF pathways included protein binding, ATPase activity and DNA helicase activity. The most significant KEGG pathways included cell cycle, the p53 signaling pathway and oocyte meiosis (Fig. 4). The top 3 pathways of the 30 most downregulated genes with gene lists are displayed in Table III.

EBV infection affected several cellular pathways; according to BP and KEGG pathway analysis, EBV infection induced the expression of cellular genes involved in cancer-related pathways, including the MAPK signaling pathway (ARRB2, CD14, CDC25B, DUSP1, DUSP6, HSPA6, JUND, MAP3K1, MAP3K6, MAPKAPK3, PRKACB, PRKCA, RASGRP2, RASGRP4, TGFBR2 and TNFRSF1A), the WNT signaling pathway (CCND3, CTBP2, FRAT1, FRAT2, LEF1, PLCB1, PLCB2, PRKACB, PRKCA and WNT10B) and the PI3K-Akt signaling pathway (CCND3, CDK4, CDK6, CSF1R, CSF3R, F2R, GNG11, GNG2, IL4R, IL6R, ITGA5, ITGB1, JAK1, LAMB2, PPP2R2B, PRKCA, SGK3 and TLR2). These genes promote proliferation of BLCL and their transformation to immortalized cells. Cellular genes that function and are located in the plasma membrane and cell surface, such as CX3CR1, CD8A, CD3D, CD14, CD2, CDH23, CSF3R, TYROBP, CSF1R, S100A10, NKG7, LILRB3, CD2, GP1BB, NRGN, ITGAM, ITGB and ITGAX were highly expressed, and these genes were associated with the MF pathways; protein binding and cell adhesion. As shown in Fig. 1D, EBV-transformed BLCL exhibited cell clumping and rosette shape formation.

Downregulated genes were primarily involved in cell cycle (CCNB1, CCNB2, CCNE1, CCNE2, CDC20, CDC45, CDK1, CDKN2A, CHEK1, E2F1, ESPL1, MAD2L1, MCM2, MCM4, MDM2, PCNA, PKMYT1, PTTG1, PTTG2 and SFN), p53 signaling pathway (CCNB1, CCNB2, CCNE1, CCNE2, CDK1, CDKN2A, CHEK1, DDB2, FAS, MDM2, RRM2B and SFN) and cellular senescence (CCNB1, CCNB2, CCNE1, CCNE2, CDK1, CDKN2A, CHEK1, E2F1, MAPK11, MDM2, MYBL2 and SLC25A4).