Stephanie Blaich

DMBT1 functions as pattern-recognition molecule for poly-sulfated and poly-phosphorylated ligands.

Deleted in malignant brain tumors 1 (DMBT1) is a secreted glycoprotein displaying a broad bacterial‐binding spectrum. Recent functional and genetic studies linked DMBT1 to the suppression of LPS‐induced TLR4‐mediated NF‐κB activation and to the pathogenesis of Crohns disease. Here, we aimed at unraveling the molecular basis of its function in mucosal protection and of its broad pathogen‐binding specificity. We report that DMBT1 directly interacts with dextran sulfate sodium (DSS) and carrageenan, a structurally similar sulfated polysaccharide, which is used as a texturizer and thickener in human dietary products. However, binding of DMBT1 does not reduce the cytotoxic effects of these agents to intestinal epithelial cells in vitro. DSS and carrageenan compete for DMBT1‐mediated bacterial aggregation via interaction with its bacterial‐recognition motif. Competition and ELISA studies identify poly‐sulfated and poly‐phosphorylated structures as ligands for this recognition motif, such as heparansulfate, LPS, and lipoteichoic acid. Dose–response studies in Dmbt1−/− and Dmbt1+/+ mice utilizing the DSS‐induced colitis model demonstrate a differential response only to low but not to high DSS doses. We propose that DMBT1 functions as pattern‐recognition molecule for poly‐sulfated and poly‐phosphorylated ligands providing a molecular basis for its broad bacterial‐binding specificity and its inhibitory effects on LPS‐induced TLR4‐mediated NF‐κB activation.

Carcinogen inducibility in vivo and down-regulation of DMBT1 during breast carcinogenesis

Deleted in malignant brain tumors 1 (DMBT1) has been proposed as a candidate tumor suppressor for brain and epithelial cancer. Initial studies suggested loss of expression rather than mutation as the predominant mode of DMBT1 inactivation. However, in situ studies in lung cancer demonstrated highly sophisticated changes of DMBT1 expression and localization, pointing to a chronological order of events. Here we report on the investigation of DMBT1 in breast cancer in order to test whether these principles might also be attributable to other tumor types. Comprehensive mutational analyses did not uncover unambiguous inactivating DMBT1 mutations in breast cancer. Expression analyses in the human and mouse mammary glands pointed to the necessity of DMBT1 induction. While age‐dependent and hormonal effects could be ruled out, 9 of 10 mice showed induction of Dmbt1 expression after administration of the carcinogen 7,12‐dimethybenz(α)anthracene prior to the onset of tumorigenesis or other histopathological changes. DMBT1 displayed significant up‐regulation in human tumor–flanking tissues compared to in normal breast tissues (P < 0.05). However, the breast tumor cells displayed a switch from lumenal secretion to secretion to the extracellular matrix and a significant down‐regulation compared to that in matched normal flanking tissues (P < 0.01). We concluded that loss of expression also is the predominant mode of DMBT1 inactivation in breast cancer. The dynamic behavior of DMBT1 in lung carcinoma is fully reflected in breast cancer, which suggests that this behavior might be common to tumor types arising from monolayered epithelia.

Abstract A51: Development and initial characterization of isogenic cell lines stably overexpressing prostate cancer candidate genes

Microarray analyses of global changes in gene expression patterns have recovered a large number of genes, which are deregulated in prostate cancer, when compared to normal prostate tissue. In order to successfully translate these data into clinical intervention strategies, it is crucial to determine genes that truly contribute to tumor development (which may be referred to as ‘driver genes’) and to separate them from the majority of secondary anomalies in expression levels (‘passenger genes’). To obtain the missing functional information, we construct isogenic cell line libraries derived from commonly used prostate (RWPE‐1) and prostate cancer (LNCaP, PC‐3) cell lines and subject them to a range of functional assays. Using a system for rapid generation of stably‐transfected isogenic cell lines, that was developed in our group, we create a prostate cancer library in a two‐step procedure. At first, acceptor clones are generated from cancer and normal prostate cell lines by stable integration of a plasmid carrying a recombinase target sequence. Incorporation of a second expression plasmid with the gene of interest is then mediated by site‐specific recombination. This procedure allows a facilitated transgene integration into a predefined, transcriptionally active locus. It also minimizes the influence of genetic background, as clones generated within such a library differ among each other only by the presence of the gene of interest in the same chromosomal location. Therefore, in contrast to traditional knock‐in strategies, our system ensures that readily interpretable phenotypes are obtained and thereby provides a highly standardized resource for functional gene analysis. The technology is well suited for medium throughput and was designed to link data from high‐throughput genomic analyses with functional gene analysis in animal models. LNCaP and PC‐3 acceptor cell lines have been generated and characterized. Recombination of PC‐3 with several well‐described, cancer‐relevant genes under the control of an inducible promoter indicated the feasibility of the method. The next step will include initial screening of the library of roughly 100 recombinants for changes in the cell viability and proliferation. The most promising effectors from the primary screen will be followed with an expanded range of functional assays, addressing their impact on cell cycle, apoptosis and cell migration. This can then serve as a starting point for an individual characterization of the novel prostate cancer genes. Citation Information: Cancer Res 2009;69(23 Suppl):A51.