Transcriptional profiling of Epstein-Barr virus (EBV) genes and host cellular genes in nasal NK/T-cell lymphoma and chronic active EBV infection

Abstract

Nasal NK/T-cell lymphoma is an aggressive subtype of non-Hodgkin lymphoma (NHL) that is closely associated with Epstein-Barr virus (EBV). The clonal expansion of EBV-infected NK or T cells is also seen in patients with chronic active EBV (CAEBV) infection, suggesting that two diseases might share a partially similar mechanism by which EBV affects host cellular gene expression. To understand the pathogenesis of EBV-associated NK/T-cell lymphoproliferative disorders (LPD) and design new therapies, we employed a novel EBV DNA microarray to compare patterns of EBV expression in six cell lines established from EBV-associated NK/T-cell LPD. We found that expression of BZLF1, which encodes the immediate-early gene product Zta, was expressed in SNK/T cells and the expression levels were preferentially high in cell lines from CAEBV infection. We also analyzed the gene expression patterns of host cellular genes using a human oligonucleotide DNA microarray. We identified a subset of pathogenically and clinically relevant host cellular genes, including TNFRSF10D, CDK2, HSPCA, IL12A as a common molecular biological properties of EBV-associated NK/T-cell LPD and a subset of genes, such as PDCD4 as a putative contributor for disease progression. This study describes a novel approach from the aspects of viral and host gene expression, which could identify novel therapeutic targets in EBV-associated NK/T-cell LPD.

Figure 1

Quantification of EBV viral genome in SNK/T cells. The outlined bar indicates copy numbers of EBV in SNK/T cells. The hatched bar indicates the copy numbers of EBV in untreated supernatant. The black bar indicates the copy numbers of encapsidated EBV in DNase-treated supernatant that was only seen in B95-8 cells. In SNK/T cells, viral load is observed in both cells and untreated supernatant, whereas EBV DNA was undetectable after DNase treatment. Primers and probes are as follows: the forward primer; 5′-GGAACCTGGTCAT CCTTGC, the reverse primer; 5′-ACGTGCATGGACCGGTTAAT, and probe 5′-FAM- CG CAGGCACTCGTA CTGCTCGCT-TAMRA.

Figure 2

Transcriptional profiling of EBV in SNK/T cells and EBV associated diseases. The cluster analysis (K-mean) was done using Genomic Profiler® software. This colour map shows that genes whose expression is greater than EBV negative control is shown as red (see legend). The molecular signature of SNK/T cells is highly variable, but shares partly common features (see the text). Notable genes are indicated by arrowheads, and asterisks indicate lytic genes.

Figure 3

(A) Quantitative relationship between EBV gene expression by microarray and a real-time RT-PCR. Circles: SNK/T cells, Triangle: IM-a cells, and Solid dot: Akata. In general, the normalised log ratio of median was basically associated with the results obtained from real-time RT-PCR. When normalised log ratios were extremely high (i.e., LCL-a) or undetectable (i.e., IM-a), there was a significant correlation between the EBV gene expression levels by microarray analysis and those by a real-time RT-PCR. (B) Quantification of EBV gene expression by real-time RT-PCR. BARF1 is consistently expressed in all the SNK/T cells. The expression of BHRF1 is also seen in all the SNK/T cells, but the expression levels are approximately 2 log lower than those in LCL. (C) Western blot analysis of Zta in SNK/T cells. WaY was used as the negative control, and B95-8 was used as the positive control. All the SNK/T cells were weakly positive for Zta.

Figure 4

(A) The common molecular signature of SNK/T cells. ANOVA analysis was performed by GeneSifter® using microarray analysis data deposited in GEO (GSE2414). Genes whose expression levels are 4-fold greater (left) or lower (right) in SNK/T cells compared to those in normal lymphocytes were navigated by GeneSifter®. A list of gene names is also shown in Table 2. (B) Genes preferentially expressed in lymphoma cell lines. Quantitative real-time RT PCR of FGF14, PDCD4, and PCNA in SNK/T cells. The expression levels are higher in cell lines established from NK/T lymphoma consistent with the results obtained from microarray analysis.