The role of receptor tyrosine kinase EphB4 in the tumorigenesis of endometrial cancer

Abstract

Erythropoietin-producing hepatocellular receptor B4 (EphB4), a receptor tyrosine kinasewithin the Eph family, has been increasingly associated with development of humanmalignancies. However, the role of EphB4 depicted in various cancers seems divergentwhich is believed to be highly dependent on the tissue type and cell context. Previousstudies in different tumour cohorts have suggested a positive correlation between EphB4expression and malignant potential in human endometrial cancer; yet its biologicalsignificance in endometrial tumorigenesis has not been well characterised. The aim of thisstudy was to interrogate the role of EphB4 in the regulation of endometrial cancer cellsurvival and migration in vitro, and further elucidate the underlying mechanisms thatcontribute to tumour progression.First, we screened the expression of EPHB4 and EFNB2 transcripts across a diversecollection of human endometrial specimens by quantitative real-time RT-PCR. Inagreement with previous reports, our results showed upregulation of EPHB4 and EFNB2expression in human endometrial adenocarcinomas in comparison to little expression innormal human endometrium. Then, we examined the expression of EPHB4 and EFNB2 inseveral human endometrial epithelial cell lines. Real-time RT-PCR and immunoblottingresults demonstrated that EPHB4 and EFNB2 were commonly coexpressed in the fourendometrial cell lines we analysed, with the strongest EphB4 protein expression detectedin Ishikawa and HEC-1B endometrial cancer cell lines. Additionally, we observeddifferential expression of EPHB4 and EFNB2 in three-dimensional (3D) multicellularspheroids (MCSs) versus traditional two-dimensional (2D) cell monolayers.To further characterise the biological function of EphB4 in endometrial cancer cellprogression, we knocked down EphB4 protein expression in Ishikawa and HEC-1B celllines by transduction with two independent lentiviral shRNAs targeting different exons ofEPHB4, and further assessed the effects of EphB4 downregulation on cell survival andmotility. We found that EphB4 knockdown in Ishikawa and HEC-1B cells resulted inattenuated cell growth in both 2D monolayers and 3D MCSs, concurrent with impairedcolony formation and cell migration. Conversely, overexpression of full-length EPHB4 inHEC-1B cells facilitated a more aggressive phenotype characterised with augmented cellgrowth, clonogenicity and migration. In addition, preliminary evidence in our studysuggests that some focal adhesion signalling components, such as FAK, paxillin and Src, are possibly involved downstream of EphB4 in the regulation of endometrial cancer cellsurvival and motility, although we observed distinctive modulation pattern in 2D versus 3Dcell models as well as in cell lines with differential Ephs-ephrins expression profile.Intriguingly, despite the substantial protein levels of EphB4 in the examined endometrialcancer cell lines, minimal tyrosine phosphorylation was detected. This data suggests thatendogenous EphB4 receptors in these cell lines are not constitutively activated by thecoexpressed ephrin-B2 ligands. Our results however further demonstrated that EphB4receptor kinase in Ishikawa cells remained ligand-responsive since the tyrosinephosphorylation level of EphB4 markedly increased following stimulation with a chimericephrin-B2/Fc ligand. Activation of EphB4 kinase activity by the ligand stimulation led todecreased cell viability in both 2D monolayers and 3D MCSs, suggesting that ephrin-B2ligand-induced EphB4 kinase signalling is anti-proliferative. Moreover, treatment with apotent EphB4 kinase inhibitor NVP-BHG712 alone showed different effects on Ishikawacell growth in 2D versus 3D cultures, with a small growth promoting effect in 2D but anotable proliferative effect in 3D. Furthermore, pre-treated the cells with NVP-BHG712 for1 hour and then followed with chimeric ephrin-B2/Fc ligand stimulation was observed topartially block the growth inhibitory effect of ephrin-B2-induced EphB4 kinase activity in 3Dcultures but not seen in 2D cultures. Hence, we hypothesize that NVP-BHG712 mightimpact the signalling activity of other Eph receptor kinases in addition to the EphB4receptor contributing to the overall cellular response. Indeed, we have identified thatmultiple Eph receptor transcripts including EPHA3, EPHA4, EPHB1, EPHB2 and EPHB3are coexpressed with EPHB4 and EFNB2 in the three endometrial cancer cell lines butwith differential expression profiles. Taken together, these data suggest that aberrantexpression of the EphB4 receptor and ephrin-B2 ligand is likely present in endometrialcancer cells, such that overexpression of EphB4 leading to disruption of anti-oncogenicephrinB2/EphB4 signalling.In summary, it is suggested that overexpression of the EphB4 receptor lacking kinaseactivation with the ephrin-B2 ligand is likely responsible for the malignant potential inendometrial cancer cells. The ligand-independent EphB4 signalling has beendemonstrated to play an oncogenic role in favouring cancer cell survival and migration. Onthe contrary, ligand-dependent EphB4 kinase activity elicits a tumour-suppressingresponse in preventing endometrial cancer cell growth. The dual role of EphB4 incontrolling the growth and migration of endometrial cancer cells is believed to act through multiple mechanisms, such as modulation of focal adhesion complexes and crosstalk withother Eph family members to form different signalling complexes, although furtherinvestigations are warranted. Importantly, this study has revealed the complexities of theEphB4 biology in endometrial cancer cells and highlighted that studying thecomprehensive EphB4 biology in 3D tumour cell environment can unravel novel findingsthat are not replicated in their 2D counterparts. Above all, our results also indicate thatrestoring the anti-oncogenic EphB4 kinase signalling and/or blocking the oncogenickinase-independent activity of EphB4 receptor may provide a possible therapeutic strategyfor endometrial cancer.