ABSTRACT
Purpose: The high fatality-to-case ratio of ovarian cancer is directly related to platinum resistance. Exportin-1 (XPO1) is a nuclear exporter that mediates nuclear export of multiple tumor suppressors. We investigated possible clinicopathologic correlations of XPO1 expression levels and evaluated the efficacy of XPO1 inhibition as a therapeutic strategy in platinum-sensitive and -resistant ovarian cancer.Experimental Design: XPO1 expression levels were analyzed to define clinicopathologic correlates using both TCGA/GEO datasets and tissue microarrays (TMA). The effect of XPO1 inhibition, using the small-molecule inhibitors KPT-185 and KPT-330 (selinexor) alone or in combination with a platinum agent on cell viability, apoptosis, and the transcriptome was tested in immortalized and patient-derived ovarian cancer cell lines (PDCL) and platinum-resistant mice (PDX). Seven patients with late-stage, recurrent, and heavily pretreated ovarian cancer were treated with an oral XPO1 inhibitor.Results: XPO1 RNA overexpression and protein nuclear localization were correlated with decreased survival and platinum resistance in ovarian cancer. Targeted XPO1 inhibition decreased cell viability and synergistically restored platinum sensitivity in both immortalized ovarian cancer cells and PDCL. The XPO1 inhibitor-mediated apoptosis occurred through both p53-dependent and p53-independent signaling pathways. Selinexor treatment, alone and in combination with platinum, markedly decreased tumor growth and prolonged survival in platinum-resistant PDX and mice. In selinexor-treated patients, tumor growth was halted in 3 of 5 patients, including one with a partial response, and was safely tolerated by all.Conclusions: Taken together, these results provide evidence that XPO1 inhibition represents a new therapeutic strategy for overcoming platinum resistance in women with ovarian cancer. Clin Cancer Res; 23(6); 1552-63. ©2016 AACR.
Subject(s)
Cell Proliferation/drug effects , Drug Resistance, Neoplasm/genetics , Karyopherins/genetics , Ovarian Neoplasms/drug therapy , Receptors, Cytoplasmic and Nuclear/genetics , Acrylates/administration & dosage , Active Transport, Cell Nucleus/genetics , Animals , Apoptosis/drug effects , Cell Line, Tumor , Cell Survival/drug effects , Female , Humans , Hydrazines/administration & dosage , Karyopherins/antagonists & inhibitors , Mice , Ovarian Neoplasms/genetics , Ovarian Neoplasms/pathology , Platinum/administration & dosage , Platinum/adverse effects , Receptors, Cytoplasmic and Nuclear/antagonists & inhibitors , Triazoles/administration & dosage , Xenograft Model Antitumor Assays , Exportin 1 ProteinABSTRACT
The formation and stability of protein-protein interfaces are of obvious biological importance. While a large body of literature exists describing the effect of osmolytes on protein folding, very few studies address the effect of osmolytes on protein association and binding. The plant lectin concanavalin A (ConA), which undergoes a reversible tetramer-to-dimer equilibrium as a function of pH, was used as a model system to investigate the influence of nine osmolytes on protein self-association. The stabilizing or destabilizing impacts of the osmolytes were evaluated from pH titrations combined with circular dichroism spectroscopy. Relative to the dimer, trimethylamine N-oxide, betaine, proline, sarcosine, sorbitol, sucrose, and trehalose all stabilized the ConA tetramer to varying extents. Glycerol had a negligible effect, and urea destabilized the tetramer. From multiple titrations in different osmolyte concentrations, an m-value (a thermodynamic parameter describing the change in the association free energy per molar of osmolyte) was determined for each osmolyte. Experimental m-values were compared with those calculated using two theoretical models. The Tanford transfer model, with transfer free energies determined by Bolen and co-workers, failed to accurately predict the m-values in most cases. A model developed by Record and co-workers, currently applicable only to urea, betaine, and proline, more accurately predicted our experimental m-values, but significant discrepancies remained. Further theoretical work is needed to develop a thermodynamic model to predict the effect of osmolytes on protein-protein interfaces, and further experimental work is needed to determine if there is a general stabilization by osmolytes of such interfaces.
