A small molecule drug promoting miRNA processing induces alternative splicing of MdmX transcript and rescues p53 activity in human cancer cells overexpressing MdmX protein.

Georgios Valianatos,Michaela Vyhnakova,Jitka Mlcochova,Ondrej Slaby,Amandine Verlande,Stjepan Uldrijan,Katerina Growkova,Barbora Valcikova,Lenka Radova,Monika Stetkova

doi:10.1371/journal.pone.0185801

Abstract

MdmX overexpression contributes to the development of cancer by inhibiting tumor suppressor p53. A switch in the alternative splicing of MdmX transcript, leading to the inclusion of exon 6, has been identified as the primary mechanism responsible for increased MdmX protein levels in human cancers, including melanoma. However, there are no approved drugs, which could translate these new findings into clinical applications. We analyzed the anti-melanoma activity of enoxacin, a fluoroquinolone antibiotic inhibiting the growth of some human cancers in vitro and in vivo by promoting miRNA maturation. We found that enoxacin inhibited the growth and viability of human melanoma cell lines much stronger than a structurally related fluoroquinolone ofloxacin, which only weakly modulates miRNA processing. A microarray analysis identified a set of miRNAs significantly dysregulated in enoxacin-treated A375 melanoma cells. They had the potential to target multiple signaling pathways required for cancer cell growth, among them the RNA splicing. Recent studies showed that interfering with cellular splicing machinery can result in MdmX downregulation in cancer cells. We, therefore, hypothesized that enoxacin could, by modulating miRNAs targeting splicing machinery, activate p53 in melanoma cells overexpressing MdmX. We found that enoxacin and ciprofloxacin, a related fluoroquinolone capable of promoting microRNA processing, but not ofloxacin, strongly activated wild type p53-dependent transcription in A375 melanoma without causing significant DNA damage. On the molecular level, the drugs promoted MdmX exon 6 skipping, leading to a dose-dependent downregulation of MdmX. Not only in melanoma, but also in MCF7 breast carcinoma and A2780 ovarian carcinoma cells overexpressing MdmX.Together, our results suggest that some clinically approved fluoroquinolones could potentially be repurposed as activators of p53 tumor suppressor in cancers overexpressing MdmX oncoprotein and that p53 activation might contribute to the previously reported activity of enoxacin towards human cancer cells.

Highlights

Several key tumor suppressor pathways are inactivated during the development of most, if not all, human cancers
The practical use of fluoroquinolones in cancer therapy might be limited by the fact that doses required for enhancing miRNA processing in mammalian cells are higher than those given to patients to treat bacterial infections
We Enoxacin-induced p53 activation in cancer cells show that a switch in MdmX mRNA alternative splicing and down-regulation of MdmX protein levels, leading to non-genotoxic activation of p53-dependent transcription, could partly contribute to the antiproliferative action of fluoroquinolone antibiotics in melanoma cells expressing high levels of the MdmX oncogene

Summary

Introduction

Several key tumor suppressor pathways are inactivated during the development of most, if not all, human cancers. A prime example of such frequently inactivated tumor suppressor is p53, an important regulator of cellular responses to stress stimuli, such as hypoxia, DNA damage, oncogene activation, telomere shortening or metabolic stress [1]. Its function is commonly lost in cancers by mutations in the p53 gene or by overexpression of cellular inhibitory proteins Mdm and MdmX ( known as Mdm or HdmX) that cooperate to inhibit p53-mediated transcription by binding to its transactivation domain and by targeting p53 for proteasomal degradation [2]. There are other cancers with a known or suspected role of MdmX overexpression in tumor development or progression, such as, for example, retinoblastoma [5,6], breast carcinoma [7,8], or chronic lymphocytic leukemia (CLL) [9]

Methods

Results

Conclusion