Modeling human tumor progression - identifying an extracellular matrix interaction network involved in epithelial carcinogenesis

Investigating human tumor progression in patient samples is complicated by etiologic heterogeneity, genetic instability, and an overabundance of precursor lesions that fail to progress. Experimental models of human tumorigenesis offer greater control, but most rely upon the subcutaneous injection of transformed cells and therefore ignore native tissue barriers to transformation, appropriate tumor-stroma interactions during progression, and the somatic modulation of oncogenes and tumor suppressors. In this work, we generated an inducible tissue model of human neoplasia driven by conditionally active Ras and characterized the sequence of gene expression programs engaged in epithelial tumor tissue and adjacent stroma at specific time points during carcinogenesis. Tumor-intrinsic gene expression was further refined by leveraging the increased signaling specificity of downstream oncogene effectors and the resultant core tumor signature was subjected to network modeling to prioritize potential functional targets. Network topology predicted that tumor development depends upon engagement of specific ECM-interacting network hubs. Blockade of one such hub, the beta1 integrin subunit, preferentially disrupted network gene expression and attenuated tumorigenesis in vivo. Thus, integrating temporal gene expression analysis and network modeling of inducible human neoplasia provides an approach to prioritize and then characterize genes functioning in cancer progression.