Recessive genetic alterations in tumor suppressor genes frequently activate expression of wild type alleles of transforming oncogenes that function downstream in a biochemical pathway. In normal epithelium, but not in highly dysplastic epithelium, expression of these oncogenes is limited to specific compartments. GKLF is a major transforming activity in carcinomas that is normally expressed in the differentiating cell layers of stratified squamous epithelium. In dysplastic epithelium, expression is greatly increased and is present in all cell layers, suggesting that suppression in the basal cell layer may be lost early during tumor progression. GKLF expression is consistently upregulated during progression of oral squamous cell carcinoma, and is detected by mRNA in situ hybridization in two-thirds of breast cancers. Enforced expression of GKLF in the basal cell layer of transgenic mouse skin induces features of epithelium dysplasia, including loss of apical-basal polarization and hyperkeratosis. In RK3E epithelial cells, GKLF specifically inhibits expression of integrin receptors for the basement membrane components collagen and laminin. Expression of the vitronectin/fibronectin receptor component aV was not inhibited by GKLF. Consistent with these results, GKLF-transformed cells exhibited a markedly reduced rate of attachment to the collagen or laminin, but retained or increased attachment to the extracellular matrix components vitronectin and fibronectin. These results identify GKLF as a candidate determinant of the dysplastic phenotype through regulation of integrin function. By activating expression of GKLF, tumor cells may acquire a property of normal differentiating epithelial cells, the ability to release the basement membrane and attach to other components of the extracellular matrix. In the first aim of this proposal, Dr. Ruppert will characterize GKLF transgenic mouse skin for molecular alterations found in human neoplastic lesions, including loss of integrin expression. In the second aim, he will also test mice for predisposition to tumor formation, and use an inducible system to test the effects of downregulating GKLF in dysplastic epithelium and tumors. In the third aim he will use RK3E cells to further characterize the mechanism of regulation of integrin expression by GKLF, and to determine the frequency and timing of GKLF activation in breast cancers and oral tumors.
