We have previously shown that the EphA2 receptor tyrosine kinase is overexpressed in glioblastoma multiforme (GBM) and represents a novel, attractive therapeutic target for the treatment of brain tumors. Here, we have developed an EphA2-targeted agent, ephrinA1-PE38QQR, a novel cytotoxin composed of ephrinA1, a ligand for EphA2, and PE38QQR, a mutated form of Pseudomonas aeruginosa exotoxin A. EphrinA1-PE38QQR showed potent and dose-dependent killing of GBM cells overexpressing the EphA2 receptor in cell viability and clonogenic survival assays, with an average IC(50) of approximately 10(-11) mol/L. The conjugate was also highly effective in killing breast and prostate cancer cells overexpressing EphA2. The cytotoxic effect of ephrinA1-PE38QQR was specific, as it was neutralized by an excess of EphA2 ligands. Moreover, normal human endothelial cells and breast cancer cells that do not overexpress EphA2, as well as GBM cells that have down-regulated EphA2, were not susceptible to the cytotoxin. EphrinA1-PE38QQR-mediated cytotoxicity induced caspase-dependent apoptosis, which was, however, not responsible for cell death in response to the conjugate. In addition, the conjugate elicited no changes in the activity of survival pathways such as phosphoinositide 3-kinase, measured by AKT phosphorylation. This is the first attempt to create a cytotoxic therapy using any of the ephrin ligands of either class (A or B) conjugated to a bacterial toxin. EphrinA1-PE38QQR is very potent and specific, produces cell death that is caspase independent, and forms the basis for the further development of clinically applicable EphA2-targeted cytotoxins.