Phosphorylation of eIF4E promotes EMT and metastasis via translational control of SNAIL and MMP-3.

The progression of cancers from primary tumors to invasive and metastatic stages accounts for the overwhelming majority of cancer deaths. Understanding the molecular events which promote metastasis is thus critical in the clinic. Translational control is emerging as an important factor in tumorigenesis. The messenger RNA (mRNA) cap-binding protein eIF4E is an oncoprotein that has an important role in cancer initiation and progression. eIF4E must be phosphorylated to promote tumor development. However, the role of eIF4E phosphorylation in metastasis is not known. Here, we show that mice in which eukaryotic translation initiation factor 4E (eIF4E) cannot be phosphorylated are resistant to lung metastases in a mammary tumor model, and that cells isolated from these mice exhibit impaired invasion. We also demonstrate that transforming growth factor-beta (TGFß) induces eIF4E phosphorylation to promote the translation of Snail and Mmp-3 mRNAs, and the induction of epithelial-to-mesenchymal transition (EMT). Furthermore, we describe a new model wherein EMT induced by TGFß requires translational activation via the non-canonical TGFß signaling branch acting through eIF4E phosphorylation.

Mechanisms underlying resistance to streptozotocin in Mer+ and Mer- human tumor lines.

Streptozotocin (STZ) is a monofunctional nitrosourea employed in the treatment of patients with islet cell tumors. To analyze the role of DNA repair mechanisms in causing resistance to STZ, we evaluated the cytotoxicity by this agent in three human tumor lines that differ with respect to their abilities to repair N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) damaged virus (the Mer phenotype). HT-29, A2182, and BE human tumor lines are high, intermediate and low, respectively, with regard to features that define the Mer phenotype. Our results demonstrated that the order of resistance to STZ is HT-29 greater than A2182 greater than BE. The degree of inhibition of DNA synthesis by STZ was in the following order: BE greater than A2182 greater than HT-29. O6-Alkyltransferase activity was increased markedly in HT-29 cells compared to A2182 cells which, in turn, had significantly increased levels compared to the BE line. Other potential factors such as 3-methyladenine DNA glycosylase activity, the induction by STZ of single-stranded DNA breaks, and the kinetics of repair of these breaks do not clearly underlie differences in cytotoxicity among the three tumor lines. However, increased topoisomerase II activity, as well as enhanced sensitivity to agents that interact with topoisomerase II, was present in A2182 cells compared to BE cells. These findings demonstrate that while O6-alkyltransferase contributes to resistance to STZ in some Mer+ tumor lines, other mechanisms may also contribute to resistance to this agent.

Schedule dependence of buthionine sulfoximine in reversing resistance to cisplatin.

We have found that the potentiation of antiproliferative effects by buthionine sulfoximine (BSO) of cell growth inhibition induced by cisplatin are highly schedule dependent in resistant BE colon carcinoma cells. Maintenance of low GSH levels during the 12-h interval after cisplatin (cis-DDP) treatment is critical. A schedule of BSO exposure that results in low GSH levels for 12 h after cisplatin exposure is associated with a marked increase in DNA interstrand cross-link formation as analyzed by alkaline elution. These findings are consistent with a central role of GSH in interfering with the conversion of cis-DDP DNA monoadducts to DNA interstrand cross-links and may prove relevant to the design of clinical trials of BSO with cisplatin.

Characterization of acquired resistance to cis-diamminedichloroplatinum (II) in BE human colon carcinoma cells.

To study mechanisms underlying resistance to cis-diamminedichloroplatinum (II) (cis-DDP) we have induced resistance to this agent in BE human colon carcinoma cells. A 5-fold increase in the IC50 of resistant compared to sensitive cells was noted as analyzed by the inhibition of cellular growth. Up to a 4-fold reduction in interstrand cross-link formation by cis-DDP in resistant compared to sensitive cells was present as measured by alkaline elution. No significant differences were detectable either in the extent of DNA platination as analyzed by atomic absorption spectroscopy or in the induction of cis-DDP DNA adducts as evaluated by an enzyme-linked immunosorbent assay employing antiserum that detects intrastrand cross-links formed by cis-DDP. Further, no differences in the kinetics of excision of DNA interstrand cross-links, cis-DDP DNA adducts, or total platinum in DNA were present. Levels of glutathione, however, were increased about threefold in resistant compared to sensitive cells. Loss of resistance was associated with increased interstrand cross-link formation and declines in glutathione levels. Our results are consistent with a critical role of glutathione in preventing platinum monoadduct rearrangements resulting in lower levels of interstrand cross-links and resistance to cis-DDP in resistant BE cells.

Interactions of cis-diamminedichloroplatinum(II) with 1-beta-D-arabinofuranosylcytosine in LoVo colon carcinoma cells.

Prior reports demonstrated more than additive cytotoxic effects of cis-diamminedichloroplatinum(II) (CDDP) and 1-beta-D-arabinofuranosylcytosine (ara-C) in LoVo colon carcinoma cells. We have extended these findings by analyzing mechanisms that may underlie the effect of ara-C on CDDP-induced cytotoxicity. In contrast to a previous study, ara-C neither enhances DNA interstrand cross-link formation by CDDP nor affects the excision of platinum from DNA. Features peculiar to ara-C, such as its misincorporation into DNA, probably contribute since more than additive cytotoxic effects do not occur by combinations of CDDP with inhibitors of DNA synthesis that are not incorporated into DNA. Also, while ara-C does not significantly enhance the degree of inhibition of DNA synthesis caused by CDDP, the recovery of DNA synthesis after drug removal is significantly slowed when cells are exposed to both drugs. These findings contrast with those obtained with CDDP and aphidicolin (the latter agent resembles ara-C in competing with dCTP for binding to DNA polymerase alpha but, unlike ara-C, is not incorporated into DNA). Lastly, ara-C is incorporated into LoVo cell DNA undergoing replicative synthesis as well as into DNA undergoing repair synthesis after CDDP-induced induced DNA damage.

Potentiation of ara-C induced cytotoxicity by hydroxyurea in LoVo colon carcinoma cells.

The present study was undertaken to determine whether cytotoxicity by 1-beta-D-arabinofuranosylcytosine (ara-C) in LoVo colon carcinoma cells, which are resistant to high concentrations of ara-C, would be enhanced by concurrent exposure to hydroxyurea (HU). Since mechanisms underlying the effects of HU on ara-C induced cytotoxicity are unclear, we also evaluated the effect of HU on the incorporation of ara-C into DNA, as well as potential consequences of misincorporation. Our results demonstrate that HU synergistically enhances cytotoxicity by ara-C in these cells. This effect was not present when HU was combined with aphidicolin, an agent that resembles ara-C in competing with dCTP for binding to polymerase alpha but that is not incorporated into DNA. Further, cells exposed to HU and ara-C incorporated up to 5-fold as much ara-C into DNA as cells solely treated with ara-C. While the extent of inhibition of DNA synthesis was comparable with cells exposed to HU and aphidicolin as those treated with HU and ara-C, recovery of DNA synthesis was delayed more significantly by the latter combination. These findings suggest that HU synergistically potentiates ara-C induced cytotoxicity by enhancing incorporation of ara-C in LoVo cell DNA.