Effective Tumor Targeting by EphA2-Agonist-Biotin-Streptavidin Conjugates

Abstract

We recently reported on a potent synthetic agent, 135H11, that selectively targets the receptor tyrosine kinase, EphA2. While 135H11 possesses a relatively high binding affinity for the ligand-binding domain of EphA2 (Kd~130 nM), receptor activation in the cell required the synthesis of dimeric versions of such agent (namely 135H12). This was expected given that the natural ephrin ligands also need to be dimerized or clustered to elicit agonistic activity in cell. In the present report we investigated whether the agonistic activity of 135H11 could be enhanced by biotin conjugation followed by complex formation with streptavidin. Therefore, we measured the agonistic EphA2 activity of 135H11-biotin (147B5) at various agent/streptavidin ratios, side by side with 135H12, and a scrambled version of 147B5 in pancreatic- and breast-cancer cell lines. The (147B5)_n-streptavidin complexes (when n = 2, 3, 4, but not when n = 1) induced a strong receptor degradation effect in both cell lines compared to 135H12 or the (scrambled-147B5)₄-streptavidin complex as a control, indicating that multimerization of the targeting agent resulted in an increased ability to cause receptor clustering and internalization. Subsequently, we prepared an Alexa-Fluor-streptavidin conjugate to demonstrate that (147B5)₄-AF-streptavidin, but not the scrambled equivalent complex, concentrates in pancreatic and breast cancers in orthotopic nude-mouse models. Hence, we conclude that these novel targeting agents, with proper derivatization with imaging reagents or chemotherapy, can be used as diagnostics, and/or to deliver chemotherapy selectively to EphA2-expressing tumors.

Keywords: agonistic EphA2 peptides; breast cancer; cancer imaging; orthotopic cancer models; pancreatic cancer; streptavidin.

Figure 1

Affinity of EphA2-LBD-targeting agents as detected by ITC (panels A and B) or DELFIA (panel C). (a) ITC curve for agent 147B5 for EphA2-LBD (K_d = 117 nM). (b) ITC curve for the scrambled version of 147B5 against EphA2-LBD, no binding detected. (c) DELFIA dose–response curves for agents 135H11 (purple, IC₅₀ = 130 nM), 135H12 (blue, IC₅₀ = 150 nM), or the scrambled version of 135H11 (green, no inhibition).

Figure 2

Schematic representation of EphA2 receptor dimerization by agonistic agents. The structures of the EphA4 subtype were used for schematic illustration (PDB IDs 4M4P and 4M4R were used for the apo and ephrinA5 bound, respectively) [38]. Dimeric agents ephrinA1-Fc and 135H12 can efficiently induce dimer formation and receptor activation, internalization, and degradation. The present study evaluates the biotin–streptavidin system (far right) to affect receptor internalization and for its application for imaging and targeted cancer chemotherapy.

Figure 3

Effect of 135H11-biot (147B5, Table 1), and its complexes with streptavidin at various ratios, on EphA2 receptor degradation. (a) Western blot of MDA-MB-231 cells treated (1 h) with ephrinA1-Fc, EphA2 agents 135H12 (135H11 synthetic dimer; tested at 50 nM and 200 nM monomer concentration), and 147B5 (50 nM to 200 nM) and its scrambled-biot corresponding compound (200 nM) (Table 1), tested in the absence and presence of streptavidin (50 nM) for 1 h. Anti-EphA2 blot suggests that the positive control ephrinA1-Fc (22 nM monomer) treatment led to significant degradation of the EphA2 receptor, similar to (147B5)_n-streptavidin when n= 3 or 4 (150 nM or 200 nM 147B5 concentration in the presence of 50 nM streptavidin) but not when n = 1. (b) Densitometry analysis based on duplicate measurements of band intensities relative to the bands of DMSO-treated cells. * p ≤ 0.05, ** p ≤ 0.01 as determined by a one-way analysis of variance using Dunnett post-test analysis. Uncropped images for the WB are available as supplementary materials.

Figure 4

Effect of 147B5 and its complexes with streptavidin at various ratios on EphA2 receptor degradation. (a) Western blot study of BxPC3 pancreatic cancer cells treated (1 h) with ephrinA1-Fc, EphA2 agents 135H12 (135H11 synthetic dimer; tested at 50 nM and 200 nM monomer concentration), and 147B5 (50 nM to 200 nM) and its scrambled version (200 nM) (Table 1), tested in the absence and presence of streptavidin (50 nM), as indicated and for 1 h. Anti-EphA2 blot indicates that the positive-control ephrinA1-Fc (22 nM monomer) treatment led to nearly complete degradation of the EphA2 receptor, similar to (147B5)_n-streptavidin when n = 2, 3 or 4 (hence at 100 nM, 150 nM, or 200 nM 147B5 concentration in the presence of 50 nM streptavidin), but not when n = 1. (b) Densitometry analysis based on duplicate measurements of band intensities relative to the bands of DMSO-treated cells. * p ≤ 0.05, as determined by a one-way analysis of variance using Dunnett post-test analysis. Uncropped images for the WB are available as supplementary materials.

Figure 5

(147B5)₄-streptavidin-AF accumulates in breast cancer. (a) Mice harboring orthotopic human breast cancer MDA-MB-231-GFP (white circles) were treated with either (147B5)₄-streptavidin-AF or its control (scrambled-147B5)₄-streptavidin-AF (100 μL, via tail vein, 10 mg/kg) and images were taken at various time points. (b) images for the resected tumors from the experiment in (a).

Figure 6

(147B5)₄-streptavidin-AF accumulates in pancreatic cancer. Mice harboring orthotopic human pancreatic cancer BxPC3-GFP (white arrows) were treated with either (147B5)₄-streptavidin-AF or its control (scrambled-147B5)₄-streptavidin-AF (100 μL, via tail vein, 10 mg/kg) and images were taken at various time points.