Targeting EphA4 abrogates intrinsic resistance to chemotherapy in well-differentiated tumors

Purpose Alkylating reagent chemotherapy for human cancers is not curative, and relapse occurs due to the continued presence of tumor cells, referred to as minimal residual disease (MRD). Methods The survival of MRD cells after chemotherapy, a phenomenon referred to as intrinsic resistance, depends on reactive oxygen species (ROS). Well-differentiated regions of the tumor are intrinsically resistant to chemotherapy. Results Here, we report that ROS produced by cisplatin exposure induce the tyrosine phosphorylation of the receptor tyrosine kinase erythropoietin-producing human hepatocellular receptor A4 (EphA4). EphA4 protein is highly expressed in the well-differentiated tumor-derived cervical cancer cell line Caski, but not in poorly differentiated tumor-derived cervical cancer cell lines such as HeLa or SiHa. Pharmacological inhibition of EphA4 increased cisplatin-induced cell death in Caski cells. Moreover, we observed increased expression levels of the senescence marker cyclin-dependent kinase inhibitor 2A (p16) and IL-8 in the absence of EphA4 kinase function after stimulation of Caski cells with hydrogen peroxide or cisplatin exposure. Conclusion Our data demonstrate that EphA4 expression levels determine the threshold of alkylating reagent chemotherapy. Therefore, ROS-induced tyrosine phosphorylation of EphA4 confers intrinsic resistance to alkylating reagent chemotherapy in a well-differentiated tumor, and may represent a unifying Achilles’ heel for chemotherapy resistance of well-differentiated tumors. 3 samples Caski cells were grown with cisplatin (20μM) or cisplatin plus C1(EphA4 inhibitor) for 12h, and purified using the RNeasy kit (Qiagen, Valencia, CA). For the first experiment (experiment 1; cisplatin-induced changes in gene expression in Caski cells with the aim of determining changes in gene expression that are modulated by cisplatin), the data were normalized to the median of untreated control Caski cells. For the second experiment (experiment 2; cisplatin-induced changes in gene expression in Caski cells treated with or without C1), the data were normalized to the median of samples without C1.