Abstract

The challenge of microarray analysis is to unveil the biological mechanisms behind the chip data. Due to the sometimes counteracting influences of de novo transcription, RNA processing and degradation, the discovery of any particular mechanism is difficult. Therefore, a combination of data- and knowledge-driven analysis appears to be the best way to attack the problem. We analyzed human astrocytes stably expressing the HIV-1 nef gene by microarray analyses to elucidate the effects of constitutive HIV-1 Nef expression on the transcriptome of astrocytes. Statistical evaluation of microarray results revealed small clusters of genes specifically up-regulated by native Nef protein in contrast to astrocytes expressing a non-myristoylated Nef variant. At least three significantly overrepresented gene ontology groups (small GTPase signaling, regulation of apoptosis and lipid metabolism) were detected. The JAK/STAT pathway was clearly associated with those genes. This finding agreed well with a literature-based approach, where a network was derived by combined literature and promoter sequence analysis. Promoter organization suggested potentially coordinated transcriptional regulation of some of these genes. Both results were in line with previously reported phenotypic changes.