RESUMO
Interferons (IFN) are multi-functional proteins that induce a large number of genes which mediate many biological processes including host defense, cell growth control, signaling, and metabolism. Bioinformatics analysis of the 3'-untranslated regions of IFN-stimulated genes (ISGs) showed that the AU-rich elements (ARE) exist in approximately 10% of the mRNA induced by IFN. The human epithelial cell lines, WISH and 293, and the human B cell lines, Daudi and RPMI 1788, were assessed for their response to type-I IFN. Due to their differential response to the anti-viral and anti-proliferative action of IFN-alpha, they were used as cellular models for genome wide ARE-gene expression. The anti-viral and anti-proliferative actions of IFN-alpha were substantially more potent against WISH and Daudi cells than 293 and RPMI 1788 cells, respectively. These results correlated with the Stat1-driven gene expression as assessed by monitoring the expression of Stat1-mediated IFN-inducible 6-16 mRNA. Interferons were able to induce a significant proportion of common and distinct ARE-genes, but the patterns of expression were different and dependent on the type of the cell, type of IFN, and status of the cellular sensitivity to IFN. Clustering algorithms generated two informative expressed gene clusters that were selectively associated with cellular sensitivity and resistance to the anti-viral and anti-proliferative action of IFN. Use of rationally designed microarray experiments in IFN biology yielded informative clusters that may provide candidate genes for diagnostic or for evaluation of therapeutic possibilities.