br Results br Discussion In this
Discussion In this study, we investigated the role of EP4 on invasion and invadopodia in breast cancer cells. Our results demonstrate that EP4 activation promotes invadopodia-driven ECM degradation, in turn facilitating future intravasation and metastasis of breast cancer ion channel (Gligorijevic et al., 2012, Gligorijevic et al., 2014). Our data demonstrates that EP4 stimulation elevates the invasive capability of breast cancer cells in ways similar to EGFR stimulation. Interestingly, both EGF and CAY10598-induced stimulation of EGFR and EP4 exhibit biphasic dose responses. It was recently suggested that two phosphorylation sites at Src are involved in this dose-dependent switch in EGF signaling, Src-Y416, which activates Src at low EGF concentrations, and the Src-527, which inactivates it at high EGF concentrations (Zhang et al., 2012). The fact that CAY10598 also exhibits the biphasic response may suggest that EP4 mediates the effect on the invasive capability via Src. The presence of EP4 and EGFR signaling cross-talk is supported by the fact that inhibition of EGFR reverses the positive effect of EP4 stimulation on matrix degradation, and the other way around. A number of papers proposed the EGFR pathway as effector for EP4-mediated signaling pathways in cancer (Du et al., 2015, Parida et al., 2016, Tveteraas et al., 2012, Yokoyama et al., 2013, Zhang et al., 2014). Recently, it has been shown that the selective EP4 inhibitor GW627368X restricts EGFR transactivation (Parida et al., 2016). We here demonstrate that EP4 is capable of activating invadopodia-driven ECM degradation in vitro and might also be involved in ECM remodeling in vivo. This activation is dependent on the transactivation of the EGFR. The proposed signaling pathway, based on our findings and on existing literature (Buchanan et al., 2006, Kedziora et al., 2016), mediating the effect of EP4 on invadopodia maturation is depicted in Fig. 5. Furthermore, we have shown that PGE2, or specific EP4 stimulation with CAY10598, can promote invadopodia in the absence of EGF. A substitutive effect of PGE2 on the EGFR signaling has been reported by Brocard et al. (2015). It is important to note that MDA-MB-231 cells are reported to express also EP1 and EP3 (Paquette et al., 2011). The activation of all present EP receptors might result in a complex interplay and intracellular processing of the signals. The dissection of the contribution of each receptor in transducing PGE2-dependent invasion and invadopodia signaling is a complex endeavor, but necessary to better understand the role of EP4 in particular. However, our preliminary results with EP2 specific agonist Butaprost showed it does not induce invasion and therefore EP2 is most likely not involved in promoting invasion (data not shown). Moreover, there was no significant difference in the stimulation of invadopodia matrix degradation as well as mature invadopodia between cells stimulated with the EP4-specific agonist CAY10598 and PGE2, which can activate all four EP receptors subtypes. This indicates that PGE2 mainly affects degradation and invadopodia through activation of EP4. In summary, we have shown that EP4 stimulation increases the invasive capability of breast cancer cells as well as the degradative capability of invadopodia. Recently invadopodia became a very interesting therapeutic target for aggressive cancer because their inhibition is effective to limit invasion, extravasation and subsequent metastasis (Stoletov and Lewis, 2015). Since EP4 has been proposed as potential anti-cancer treatments (Kundu et al., 2014, Majumder et al., 2014) (Kundu et al., 2014, Majumder et al., 2014), further investigation of EP4 targeting to restrict invadopodia formation and cancer cell invasion is warranted.
Materials and methods
Acknowledgements We are grateful to Antoine Khalil and the Microscopic Imaging Center of the Radboud Institute for Molecular Life Sciences, Nijmegen for instrumental help, sharing protocols and resources. We also thank Yujin Chung from the Center for Computational Genetics and Genomics at Temple University for her help in statistical analysis. This work was supported by the Radboud Honours Academy to FT, EU grant NANOVISTA [grant number 2882630] to AC and United States National Institute of Health [NIH 5K99CA172360] to BG.