Tumor suppressor maspin as a rheostat in HDAC regulation to achieve the fine-tuning of epithelial homeostasis

Abstract

Maspin, a class II tumor suppressor, is often downregulated during tumor progression and its depletion from the nucleus is associated with poor prognosis. Recently, we reported that reintroduction of maspin is sufficient for redifferentiation of prostate cancer cells to epithelial phenotype, a reversal of epithelial-to-mesenchymal transition. We have linked this effect of maspin with its ability to directly inhibit HDAC1, thereby influencing the acetylation state of transcription factors and other proteins. Maspin overexpression leads to changes in the expression level of a large number of proteins and these changes are often microenvironment specific. In this review, we summarize the epigenetic effects of maspin and provide comprehensive bioinformatic analysis of microarray-derived gene expression changes caused by maspin in different microenvironments. The analysis was performed on multiple levels, including identification of statistically enriched gene ontology groups, detection of overreprepresented transcription factors binding sites in promoters of differentially expressed genes, followed by searching for key nodes of regulatory networks controlling these transcription factors. The results are consistent with our hypothesis that maspin serves as an endogenous regulator of HDAC activity and suggest that the effect of maspin is primarily mediated by TGFβ, β-catenin/E-cadherin pathways, and network key nodes such as Abl kinase, p62, IL1, and caspases 6 and 8.

FIGURE 1

Gene ontology (GO) categories affected by maspin expression. The categories are colored according to enrichment p-value (see scale).

FIGURE 2

Transcription factors with overrepresented binding sites in promoters of genes differentially regulated by maspin. TFs based on separate analysis of each of the three experimental systems are in black, those based on lists of genes uniquely differentially expressed in specific conditions are in blue, and those based on lists of commonly regulated genes are marked in red. The TFs are listed in descending statistical significance order while only those with p-values less then e⁻⁵ and enrichment ratio at least 1.3 are shown.

FIGURE 3

Network key nodes regulated by maspin according to promoter analysis. Network cluster analysis was performed using the ExPlain tool with overrepresented TFs. Activation is shown with green squares, inhibition with red, and all other interactions or modifications with yellow. TFs are shown as triangles, signaling kinases as ovals, and receptors as parallelograms.