Faculty Bibliography

OBJECTIVES: To evaluate the role of LPA in regulating E-cadherin cell surface expression, adhesion, and invasion in epithelial ovarian carcinoma (EOC) cells. METHODS: E-cadherin mRNA expression in OVCA429 and IOSE-29 cells was evaluated by real-time RT-PCR. Immunofluorescence and Western blot analysis were performed to determine cell surface expression and shedding of E-cadherin 80-kDa soluble fragment by LPA. Kinetics of LPA-induced uPA activity was followed with a colorimetric enzymatic assay. Invasion assays were performed in a modified Boyden chamber where cells were allowed to migrate to the bottom compartment through a porous filter coated with collagen. Additionally we measured the 80-kDa form from the ascites of women with stage III/IV EOC. RESULTS: LPA induces E-cadherin shedding of a soluble 80-kDa fragment. We found that this process is mediated by the uPA proteolytic cascade. High levels of soluble E-cadherin were found in the ascites from women with advanced stage EOC. LPA and a soluble recombinant E-cadherin-Fc chimera promotes invasion of OVCA429 cells. CONCLUSIONS: LPA induces shedding of an 80-kDa E-cadherin-soluble fragment in an uPA-dependent manner and promotes in vitro invasion. High levels of soluble E-cadherin in malignant ascites may also affect ovarian metastasis

OBJECTIVES: Within the tumor microenvironment the invasiveness of epithelial ovarian carcinoma (EOC) cells is stimulated by biologically active lipids such as lysophosphatidic acid (LPA). We tested the in vitro effect of another bioactive lysophospholipid, sphingosine-1-phosphate (S1P), on the invasiveness of EOC cells. METHODS: Dov13 EOC cells were tested for invasion through matrigel-coated chambers and for gelatinase activity using a fluorogenic assay. cDNA was analyzed through real-time PCR. Cell surface proteins, isolated through biotinylation and affinity purification, were analyzed by Western blots. RESULTS: Invasion of Dov13 cells was enhanced by low (0.5 microM) and inhibited by high (20 microM) concentrations of S1P, which correlated with increased and reduced gelatinase activity in conditioned media. Low and high S1P dose also differently affected the presentation of surface S1P receptors; low S1P dose increased S1P1 and decreased S1P2, while high S1P increased S1P3. LPA and S1P differently altered transcript levels of their respective and reciprocal receptors; receptors that were upregulated by one lysophospholipid (S1P2,3 and LPA1 by LPA, LPA3,4 and S1P1,4,5 by S1P) were downregulated or unchanged by the other. CONCLUSIONS: The dual effect of high and low S1P concentration on invasion was probably caused by the diverse changes to the presentation of surface S1P receptors. The opposite effect of S1P and LPA on expression of each receptor suggests a homeostatic transcriptional mechanism that abrogates the effects of LPA and S1P on EOC cells. Altogether this study demonstrates a complex role of S1P in EOC cell invasion, a process highly balanced and regulated by LPA and S1P within the tumor microenvironment