PRL-3 promotes epithelial-mesenchymal transition and confers resistance to apoptosis

Abstract

PRL-3 is a multi-tasking phosphatase involved in cancer metastasis. It promotes various cancer-related properties, such as motility, invasiveness and tumorigenicity. Two important features of cancer progression are Epithelial-Mesenchymal Transition (EMT) and evasion of programmed cell death. EMT is a process that is important for embryonic development and oncogenesis. This process causes epithelial cells to adopt a migratory mesenchymal phenotype. Evasion of apoptosis allows transformed cells to survive in the circulation to reach distant secondary sites. Here we attempt to understand the role of PRL-3 in these two processes. In this study, we found that cells expressing PRL-3 exhibited reduced focal adhesion proteins paxillin, phosphorylated paxillin-Tyr31 and vinculin. Additionally, there was reduction in RhoA-GTP, Racl-GTP and filamentous-actin (F-actin). DLD-1 human colorectal cancer cells stably expressing EGFP-PRL-3 showed activation of Akt by PRL-3 and inactivation of glycogen synthase kinase-3?. In these cells, PRL-3 also down-regulated epithelial markers E-cadherin, gamma-catenin (plakoglobin) and integrin ?3 while up-regulating mesenchymal markers fibronectin and snail. These changes could all be abrogated by the phophoinositide 3-kinase (PI3K) inhibitor LY294002. Thus PRL-3 would act upstream of PI3K to initiate EMT during cancer metastasis. PRL-3 also down-regulates phosphatase and tensin homologue deleted on chromosome 10 (PTEN), which is a key antagonist of PI3K, reinforcing PI3K/Akt activation. These changes point to PRL-3 promoting EMT by signalling through PI3K and down-regulating PTEN. Activation of Akt is also known to have effects on cell cycle regulation and cell survival. We found that MCF7 cells and DLD-1 cells stably expressing EGFP-PRL-3 were more resistant to genotoxic agents methyl methanesulfonate (MMS) and doxorubicin, as well as oxidative stress induced by H2O2. We screened protein lysates from these cells using a commercial antibody array to find cell cycle related proteins whose phosphorylation status could be modified by PRL-3, and for proteins that could interact with PRL-3. We identified the Growth Arrest and DNA Damage protein GADD34 as a candidate that fulfils both conditions consistently across different cell lines. To understand how PRL-3 could mediate GADD34 phosphorylation, we looked for changes in Lyn, a Src family kinase known to phosphorylated GADD34 and negatively regulate its pro-apoptotic response to genotoxic apoptosis. Surprisingly, we found that PRL-3 down-regulated Lyn protein as well as mRNA. We take this finding to indicate that PRL-3 might promote phosphorylation of GADD34 by an alternative pathway that is independent of Lyn. These events require PRL-3 phosphatase activity, as the catalytically inactive mutant PRL-3 (C104S) could not effect these changes. We propose that PRL-3 confers resistance to apoptosis through phosphorylation and inhibition of GADD34. These findings are part of the growing body of evidence that PRL-3 is a multi-tasking phosphatase involved in cancer metastasis through a variety of processes. PRL-3 is thus a promising molecular target for cancer therapy.