Characterization of a new, secreted key regulator of epithelial plasticity and tumor progression

Epithelial cell tumors (carcinomas) are the largest group of human malignancies, causing >80% of all cancer deaths through their ability to metastasize. While certain aspects of carcinogenesis are well-studied mechanistically, local invasion and metastasis are still ill understood. We studied mammary carcinogenesis in a murine in vitro/in vivo epithelial cell model, i.e. polarized mammary epithelial cells (EpH4) that form organotypic tubular structures in 3D collagen gel cultures. We discovered that epithelial-mesenchymal transition (EMT, involving trans-differentiation to a mesenchymal cell phenotype) is a central process in local invasion and metastasis. In EpH4 cells, EMT and metastatic ability are co-induced by transforming growth factor (TGF) beta and oncogenic Ras, signaling via the ERK/MAPK and PI3K pathways. In this model, genome-wide gene expression profiling of polysome-bound mRNA identified genes and signal transduction pathways important for epithelial plasticity and metastasis, revealing distinct functional roles to signals transduced by TGF beta, MAPK, PI3K and PDGF. In this project we will perform a detailed functional analysis of one of these EMT/metastasis-specific genes, a novel secreted protein structurally related to cytokines (FAM3C/ILEI). First studies suggest that this protein is necessary for metastasis, and may cause tumors and metastases largely on its own, probably only requiring cooperation with anti-apoptotic genes. Firstly we will address the role of ILEI in EMT, tumor cell proliferation and metastasis, using ILEI-over-expressing EpH4-derived cells (focusing on those unable to form tumors or metastases). These will then be analyzed in 3D collagen gels and nude mice. Secondly, we will employ stable knock-down of ILEI by RNA interference (RNAi) to test whether over-expression of endogenous ILEI is required for EMT and metastasis, initially using metastastic, EpH4-derived cells after EMT and then testing other murine and human cancer models. Thirdly, we will address ILEI`s mechanism of action, analyzing possible phenotypic effects of (i) recombinant ILEI and (ii) function-blocking antibodies to ILEI in EpH4-derived cells. We will then analyse (iii) signaling pathways induced by over-expressed or exogenous ILEI and (iv) whether these are important for ILEI function, using pharmacological inhibitors. With Boehringer Ingelheim and academic cooperations, we will analyze the expression of ILEI in human tumors and normal tissues and generate mouse models of ILEI function. The long-term aim of this project is to explore the potential use of function-blocking monoclonal antibodies against ILEI as therapeutics to interfere with late-stage tumor progression.

Dissemination of tumor cells to distant organs (metastasis) is the cause of death in >90% of all epi-thelial cancers. During development of carcinomas towards malignancy and metastasis, epithelia lose most aspects of their ordered polar. organization, referred to as epithelial polarity. In this project, we identified and molecularly characterized several key proteins that cause loss of epithelial organization (EMT). The first key player identified was ILEI (interleukin-like EMT inducer), a secreted protein with a double function. An immature, intracellular form of ILEI caused epithelial disorganization (EMT) in co-operation with an oncogenic protein (Ras), while secreted and `mature` ILEI might function in the (in-creasingly important) interaction of tumor cells with their surrounding stroma (e.g. creating an environ-ment of chronic inflammation). In this context, ILEI was both necessary and sufficient for EMT and me-tastasis in several, in part newly developed tissue culture models that also caused tumor- and meta-stasis formation in experimental animals, including mouse breast carcinoma cells able to colonize the lung for metastasis from the primary mammary tumor. ILEI was localized to an intracellular organelle essential for proper epithelial polarity in normal cells and benign tumors,. During EMT/metastasis, ILEI-expressing vesicles were redistributed to the entire cytoplasm, allowing use of ILEI subcellular localization as a major predictor of metastasis occurrence and bad prognosis in human breast cancer. To substantiate this rather complex way of ILEI action, we studied two other proteins regulating loss of epithelial organization, i.e. CREG (cellular repressor of E1A-regulated genes) and Annexin A1 (AnxA1). Both proteins are secreted and bind to known receptors, but also have distinct intracellular functions. In cells from genetically modified mice lacking Annexin A1, inactivation of a growth factor receptor by vesicular transport was abolished. CREG is routed by its receptor to protein-degrading or-ganelles (lysosomes), affecting genes regulating proliferation inside the cell and regulating cell diffe-rentiation through receptor binding. Overexpression of CREG and experimentally induced loss of AnxA1 by RNA-interference (RNAi) induced EMT and metastasis in several cell models. Induced loss of CREG via RNAi and re-expression of human AnxA1 reversed EMT and abolished metastasis. Expression of human AnxA1 in human metastatic mammary carcinoma cells - which lack AnxA1, show an EMT phenotype and are resistant to low molecular weight anti-cancer drugs (i.e. signaling inhibitors) - restored an almost normal epithelial organization in these cells and rendered them non-metastatic. In AnxA1- knockdown cells, a signal-transducing protein suspected to contribute to cancer - signal transducer and activator of transcription 3 (STAT3) - was essential for EMT. Future work will focus on actual mechanisms how ILEI, CREG and AnxA1 deregulate molecular machines essential for epithelial organization, using an EMT-inducing protein with a known function in such machines - scribble, first identified as a suppressor of tumor cell spread in Drosophila - as a po-sitive contol. For this we engineered the gold standard cellular model for molecular analysis of epithe-lial polarity (MDCK) to undergo EMT inducible by added ligands. In MDCK cells, more molecular tools to analyse epithelial polarity machines exist than in any other model. This will be complemented by work using ILEI-deficient mice and mice expressing a drug- inducible ILEI transgene, that have just been generated by collaborators. We expect, that our work will identify a number of novel drug targets, allowing intervention with cancer by low-MW chemicals or therapeutic monoclonal antibodies - to be done in collaboration with pharmaceutical industry, i.e. Boehringer Ingelheim Austria and - Germany.