(A) Immunoblots for pEphA/B on lysates from PC-3 and HEK 293-A2 (293-A2) cell lines treated for 30 minutes with indicated doses of doxazosin (DZ). (B) Immunoblots for pEphA/B on lysates from MDA-231-A2 (231-A2) and 293-A2 cell lines treated with 50 µM DZ in 0.2% DMSO for indicated times. (C) Doxazosin selectively activates EphA2 and EphA4 receptors. Immunoblots for pEphA/B on lysates from HEK 293 cell lines expressing given Eph receptors following treatment with indicated doses of doxazosin (DZ) for 60 minutes. Treatment with 1 µg/ml ephrin-A1-Fc ligand (EA1-Fc) for 10 minutes (EphA2) and 30 minutes (Vector, EphA1, EphA4), as well as 30 minute treatment with ephrin-A5-Fc (EA5-Fc) (EphA3) and ephrin-B1-Fc (EB1-Fc) (EphB3) served as positive controls. Blotting for total Eph kinases served as loading controls.

(A) Immunoblot of lysates from A172-A2 cells treated with indicated doses of doxazosin (DZ) in 0.2% DMSO for 90 minutes. Treatment with 1 µg/ml ephrin-A1-Fc ligand for 10 minutes served as a positive control. Lysates were immunoprecipitated with ephrin-A1-Fc (IP: EA1-Fc) as given in Methods, and probed for pEphA/B and total EphA2. Total cell lysates were probed for phosphorylated forms of ERK1/2 and Akt, as well as total ERK1/2 and Akt. (B) Immunoblot of total cell lysates from LK133A-Vec and LK133A-A2 cells treated with indicated doses of doxazosin (DZ) in 0.2% DMSO for 90 minutes. Treatment with 1 µg/ml ephrin-A1-Fc for 10 minutes served as a positive control. Lysates were probed for phosphorylated forms of EphA/B, Akt, and ERK1/2, as well as total EphA2, Akt, and ERK1/2.

MDA-MB-231 (A), A172-A2 (B), and PC3-DAB2IP KD (C) cells were subject to haptotactic cell migration toward fibronectin as described previously (see Methods). Doxazosin at indicated concentrations was presented at the lower chamber of the Transwells. Cells were allowed to migrate toward fibronectin for 4 hours. Data represent average numbers of migrating cells from 6 randomly selected fields. DMSO was used as vehicle control.

(A) Schematic illustration of the predicted effects of small molecule agonists in inducing ligand-dependent signaling. (B) Crystal structure of the EphA2 ligand binding domain (LBD) in complex with ephrin-A1. Highlighted are the hydrophobic pocket and arginine 103 of the EphA2-LBD that interact with the G–H loop of ephrin-A1 and glutamate 119 of ephrin-A1, respectively. EphA2-LBD was rotated ∼10° counter-clockwise to better reveal the binding pocket. (C) Small molecule screening identifies doxazosin (Compound 11) as a novel EphA2 agonist. MDA-231-A2 cells were treated with Compounds 1–11 (50 µM in 0.2% DMSO) for 30 minutes and cell lysates were subject to immunoblot for phosphorylated EphA/B kinases (pEphA/B) and total EphA2. (D) Chemical structure of doxazosin (DZ). (E) Dose-response of EphA2 activation by DZ. MDA-231-A2 cells were treated with the indicated doses of DZ for 30 minutes and lysates were immunoprecipitated with an EphA2-specific antibody and blotted as in (C). Treatment with 1 µg/ml ephrin-A1-Fc (EA1-Fc) for 10 minutes served as a positive control. Note decreasing amount of EphA2 following ephrin-A1 and doxazosin treatment. (F) Immunoblots for pEphA/B on lysates from MDA-231-A2 cells pretreated with 1 µM phenoxybenzamine and then treated for 1 hour with indicated doses of DZ. Treatment with 1 µg/ml ephrin-A1-Fc (EA1-Fc) served as a positive control. Treatment with 0.2% DMSO for either 1 hour (left), or 5 hours (right) served as vehicle controls. (G) Representative plot from Surface Plasmon Resonance (SPR) analysis of DZ binding to the recombinant ligand binding domain of EphA2. Curves from bottom to top represent concentrations of 1.56, 3.13, 6.25, 12.5, 25, 50 µM. Determined K_D value is shown within plot. (H) Molecular modeling of surface area diagram indicating amino acids of EphA2 potentially involved in direct interaction with doxazosin. The four amino acids of the ephrin-A1 loop are shown in red. Images were created using UCSF Chimera.

(A) ¹H-¹⁵N NMR HSQC spectra of the EphA4 LBD in the absence (blue) and in the presence (red) of doxazosin (DZ) at a molar ratio of 1∶5 (EphA4∶DZ). Several residues located over the convex surface of the EphA4 ephrin-binding channel are labeled. (B) Residue-specific chemical shift index (CSI) of the EphA4 LBD in the presence of doxazosin at a molar ratio of 1∶5 (EphA4∶DZ). Significantly-shifted residues shared with C1 are colored in bright brown, while the residues significantly shifted only by doxazosin binding are in red. (C) The docking model of the EphA4-doxazosin complex in ribbon. Binding regions identical to those for the C1-binding were colored in brown, while those unique for the doxazosin binding in red. G, K, M and E are used to donate β-strands of the convex surface of EphA4/ephrin-binding channel. (D) EphA4 residues having direct contacts with doxazosin. Residues on D–E and J–K loops are in brown, those on the convex surface in violet, and Arg106 in cyan. Green dashed lines indicate hydrogen bonds between doxazosin and EphA4 residues. (E)–(F) The same docking model with the electrostatic potential of the EphA4 LBD displayed. (G) EphA4 LBD in free and doxazosin-bound states display different squared generalized order parameter S² (Figure S6A). Blue: S² difference ≤−0.01; red: S² difference ≥0.01; brown: no significant change or S² values not determined. (H) Conformational exchanges of EphA4 in free (left panel) and doxazosin-bound states (right panel). Residues with R_ex>5 (Figure S6B) are displayed in balls and colored in red. (I) A docking model of the EphA2-doxazosin complex. Contact residues in D–E and J–K loops are labeled in brown, on the convex surface in cyan, and Arg103 in violet. The violet dash is used to indicate the hydrogen bonds between doxazosin and EphA2 residues.

(A) Immunofluorescence staining of U373-A2 cells for EphA2 receptor (red) after treatment for 60 minutes with 50 µM DZ in 0.2% DMSO. Treatment with 1 µg/ml ephrin-A1-Fc and DMSO served as positive and negative controls, respectively. DAPI nuclear staining is shown in blue. (B) Immunofluorescence staining of MDA-MB-231 cells for EphA2 receptor (red) after treatment for 120 minutes with 50 µM DZ in 0.2% DMSO. Controls are as given above. Scale bars, 25 µm. (C) Images from cell rounding analysis of PC-3 cells stimulated with 50 µM doxazosin for 30 or 60 min. Stimulation with 2 µg/ml ephrin-A1-Fc for 10 min or 0.2% DMSO served as positive and negative controls, respectively. Cells were seeded on 6-well plates and stimulated after 24 hours.

(A) Fluorescent images of prostate tumors and lung metastases resulting from GFP-tagged PC3-DAB2IP KD cells after 10 days of treatment with either vehicle, or 50 mg/kg doxazosin. Tumors in the prostate gland were imaged in a GFP light box, while lung metastatic foci were visualized under an inverted fluorescence microscope. (B) Graph comparing total number of metastatic lung foci in individual vehicle-treated (n = 7) and doxazosin-treated (n = 8) mice. (C) Quantitative analyses of total number of metastatic lung foci from different size categories. Categories were based on foci diameter measured in number of cells (small = 1–3 cells, medium = 4–6 cells, large = 7–10 cells). (D) Comparison of total metastatic burden (number of foci×foci diameter) in mice treated with vehicle control vs. those treated with doxazosin. (E) Graph comparing bodyweights of vehicle- and doxazosin-treated mice. Bars represent mean bodyweights. Error bars represent the SEM. Experiment was repeated three times with similar results. (F) Kaplan-Meier Plot showing prolonged survival in mice treated with doxazosin (n = 8) compared with those treated with vehicle control (n = 9). Mice were injected with the PC3-DAB2IP KD cells and treated as in (A) and closely monitored for survival. Those that became moribund were sacrificed. Similar results were obtained from three independent experiments.