Insights into gene expression changes impacting B-cell transformation: cross-species microarray analysis of bovine leukemia virus tax-responsive genes in ovine B cells

Abstract

Large-animal models for leukemia have the potential to aid in the understanding of networks that contribute to oncogenesis. Infection of cattle and sheep with bovine leukemia virus (BLV), a complex retrovirus related to human T-cell leukemia virus type 1 (HTLV-1), is associated with the development of B-cell leukemia. Whereas the natural disease in cattle is characterized by a low tumor incidence, experimental infection of sheep leads to overt leukemia in the majority of infected animals, providing a model for studying the pathogenesis associated with BLV and HTLV-1. Tax(BLV), the major oncoprotein, initiates a cascade of events leading toward malignancy, although the basis of transformation is not fully understood. We have taken a cross-species ovine-to-human microarray approach to identify Tax(BLV)-responsive transcriptional changes in two sets of cultured ovine B cells following retroviral vector-mediated delivery of Tax(BLV). Using cDNA-spotted microarrays comprising 10,336 human genes/expressed sequence tags, we identified a cohort of differentially expressed genes, including genes related to apoptosis, DNA transcription, and repair; proto-oncogenes; cell cycle regulators; transcription factors; small Rho GTPases/GTPase-binding proteins; and previously reported Tax(HTLV-1)-responsive genes. Interestingly, genes known to be associated with human neoplasia, especially B-cell malignancies, were extensively represented. Others were novel or unexpected. The results suggest that Tax(BLV) deregulates a broad network of interrelated pathways rather than a single B-lineage-specific regulatory process. Although cross-species approaches do not permit a comprehensive analysis of gene expression patterns, they can provide initial clues for the functional roles of genes that participate in B-cell transformation and pinpoint molecular targets not identified using other methods in animal models.

FIG. 1.

Northern blot confirmation of microarray gene expression patterns. Northern blot analysis of five putative differentially expressed genes confirms their transcriptional up-regulation in Tax_BLV-expressing B cells (+) compared to control B cells (−). A and B correspond to RNA extracted from B-cell clones (Clone2_LTAXSN/Clone2 cells) used in two independent microarray hybridizations, hybridizations A and B (Table 2), respectively; K stands for RNA extracted from PBMC-derived B cells (SBL_TCE/SBL_iE) used in microarray hybridization K (Table 4). RNA extracted from ovine B cells was analyzed using the ³²P-labeled human cDNA probes BRF2, LENG5, METAP2, FOSL2, CART1, and TAX_BLV, described in Materials and Methods. G3PDH was used as a control. The numbers below each pair of signals indicate the relative expression ratios quantitated by densitometry analysis. The numbers in parentheses show the ratios from the corresponding microarray data.

FIG. 2.

Tax_BLV provides a growth advantage to ovine B-cell clones. Clone2 cells were transduced with either pSFTaxCMVGFP or control pSFiECMVGFP retroviral vectors and cultured in the presence of mCD154 and γ_c-common cytokines. The resulting B-cell populations comprising transduced (+) and untransduced (−) cells, referred to as Clone2_iE+/− (a) and Clone2_TCE+/− (b), were examined over time for GFP fluorescence using flow cytometry. Percentages of GFP⁺ cells are shown. Data presented are from one representative example of three independent experiments.

FIG. 3.

Flow cytometry sorting of retroviral vector-mediated transduced SBL cells results in highly purified B-cell populations that express Tax_BLV. (a) SBL were derived from ovine PBMC cultured in the presence of mCD154 and γ_c-common cytokines for 14 days. The resulting sIgM⁺ B cells were transduced with either the pSFTaxCMVGFP retroviral vector encoding both Tax and GFP or the control pSFiECMVGFP retroviral vector for the transfer of GFP alone and further cultured in the presence of mCD154 and γ_c-common cytokines for 7 days before sorting of the GFP⁺ cell populations. The resulting GFP⁺ B-cell populations are referred to as SBL_TCE and SBL_iE, respectively. (b) Both the SBL_TCE and SBL_iE populations were examined for Tax expression by Northern analysis using a TAX probe. G3PDH was used as a control of RNA load.

FIG. 4.

Tax_BLV protects PBMC-derived ovine SBL cells from apoptotic cell death and enhances viable B-cell numbers. SBL cells were cultured in the presence of mCD154 and γ_c-common cytokines IL-2, IL-4, IL-7, and IL-15. Apoptotic cell death in both SBL_TCE and SBL_iE cells was examined by a terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end-labeling assay of normal costimulated cultures collected 72 and 96 h after cell passage (a) or after a 24-h culture period in the absence of γ_c-common cytokines (−cyt) in either the presence or the absence of mCD154 (+CD154/−CD154) (b). Data presented in a and b are means ± standard deviations of values of three independent cultures. Viable SBL cell numbers were determined by trypan blue exclusion at each passage throughout a 40-day culture period (c). Data presented in c are from one representative assay of three independent experiments.

FIG. 5.

Distribution of Tax_BLV-associated differentially expressed genes between two distinct ovine B-cell populations, Clone2_LTAXSN and SBL_TCE. Eighty genes differentially expressed in both B-cell clones and PBMC-derived B cells were visualized as the intersect of two rectangles, each representing the overall number of significantly altered genes (median, >1.5 or <0.66) as a result of Tax_BLV expression in three microarray experiments for each B-cell population. Taking relative expression ratio median values, 211 genes were >1.5 and 52 genes were <0.66 (263) in hybridizations A, B, and C (Clone2_LTAXSN/Clone2), and 292 genes were >1.5 and 149 genes were <0.66 (441) in hybridizations K, L, and M (SBL_TCE/SBL_iE). Sixty-five genes were >1.5 and 15 genes were <0.66 (80) in both A, B, and C and K, L, and M hybridizations (overlap).

FIG. 6.

Similar changes in overall gene transcription patterns as a result of Tax_BLV expression in two different ovine B-cell populations. (a) Comparison of overall gene expression profiles generated from Clone2_LTAXSN/Clone2 and SBL_TCE/SBL_iE populations. Log ratio medians were computed from three corresponding microarray experiments, hybridizations A, B, and C (Clone2_LTAXSN/Clone2 cells) and K, L, and M (SBL_TCE/SBL_iE cells), respectively. b is similar to a except that median values were replaced by means. Dots above the upper oblique lines represent genes overexpressed by >2-fold, and dots below the lower oblique lines represent genes underexpressed by >2-fold.