EGFR-Targeted Immunotoxin Exerts Antitumor Effects on Esophageal Cancers by Increasing ROS Accumulation and Inducing Apoptosis via Inhibition of the Nrf2-Keap1 Pathway

Abstract

Previously, we developed a novel EGFR-targeted antibody (denoted as Pan), which has superior antitumor activity against EGFR-overexpressed tumors. However, it shows marginal effect on the growth of esophageal cancers. Therefore, the variable region of Pan was fused to a fragment of Pseudomonas exotoxin A (PE38) to create the immunotoxin, denoted as Ptoxin (PT). Results indicated that PT shows more effective antitumor activity as compared with Pan both on EGFR-overexpressed KYSE-450 and KYSE-150 esophageal cancer cells, especially on KYSE-450 cells. Moreover, treatment of PT induces regression of KYSE-450 tumor xenografts in nude mice. Furthermore, we investigated the potential mechanism involved in the enhanced antitumor effects of PT. Data showed that PT was more potent in reducing the phosphorylation of EGFR and ERK1/2. More importantly, we for the first time found that PT was more effective than Pan in inducing ROS accumulation by suppression of the Nrf2-Keap1 antioxidant pathway, and then induced apoptosis in KYSE-450 esophageal cancer cells, which may partly explain the more sensitive response of KYSE-450 to PT treatment. To conclude, our study provides a promising therapeutic approach for immunotoxin-based esophageal cancer treatment.

Figure 1

Characterization of Ptoxin. (a) SDS-PAGE analysis of Ptoxin. Lane 1: Ptoxin; M: protein marker. (b) ELISA analysis of Ptoxin-binding potency for recombinant EGFR antigen. EC50 for Ptoxin is 0.4304 μg/mL (95% confidence interval (CI): 0.3571-0.5189 μg/mL). IgG was used as control. Points: mean of 3 independent determinations; bars: SD.

Figure 2

PT was rapidly internalized when bound to EGFR-overexpressed esophageal cancer cells. (a) Flow cytometric results, indicating cell surface EGFR expression in KYSE-150 cell lines. (b) Flow cytometric results, indicating cell surface EGFR expression in KYSE-450 cell lines. (c) Analysis of internalization rates of PT in KYSE-150 cells when binding to EGFR. (d) Analysis of internalization rates of PT in KYSE-450 cells when binding to EGFR. KYSE-150 or KYSE-450 cells were incubated with saturating level of PT (10 μg/mL) for 30 minutes at 4°C. Unbound antibody conjugates were removed by washing cells. Cells were then incubated at either 4°C or 37°C. At the indicated time points, samples were detected by flow cytometry. Points: mean of 3 independent determinations; bars: SD.

Figure 3

Treatment with PT inhibited the growth of esophageal cancer cells, and EGFR downstream signaling molecules were examined by western blot. (a) CCK-8 assay showing the inhibitory effects of Ptoxin on growth of KYSE-150 cells. IC50 is 11.43 nM (95% CI: 8.578-15.24 nM) for PT. Bars: SD. (b) CCK-8 assay showing the inhibitory effects of PT on growth of KYSE-450 cells. IC50 is 2.195 nM (95% CI: 1.904-2.532 nM) for PT. (c) Expression of key molecules involved in the EGFR-AKT, EGFR-ERK1/2, and Nrf2-Keap1 signaling pathway was tested in KYSE-150 and KYSE-450 cells upon medium alone (−) or PT (10 nM) treatment for 12 h. (d) Quantification of western blot signal intensity analysis is expressed relative to the β-actin loading control by using the ImageJ software. Data show the mean ± SD (3 independent experiments). ^∗ p < 0.05, ^∗∗ p < 0.01, and ^∗∗∗ p < 0.001.

Figure 4

Treatment with PT resulted in ROS accumulation and caused repression of the Nrf2-Keap1 pathway. (a) KYSE-450 cells were treated with medium alone (control), Pan (10 nM), PT (1 nM), or PT (10 nM) for 10 hours, respectively, and flow cytometry was used to analyze the level of ROS accumulation in cells after DCFH-DA was added to stain the cells. (b) Key signaling molecules in response to treatments with Pan or PT treatment in KYSE-450 cells. Exponentially growing cells were treated with medium alone (−) or containing Pan (10 nM) or PT (10 nM) for 12 h before being analyzed. Then, EGFR, p-EGFR, ERK, p-ERK, Nrf2, and Keap1 were analyzed by western blot. (c) Bar graphic representations of the DCFH-DA fluorescence intensity upon different treatments relative to control. ^∗∗ p < 0.01 and ^∗∗∗ p < 0.001. (d) Quantification of western blot signal intensity analysis is expressed relative to the β-actin loading control by using the ImageJ software. Data show the mean ± SD (3 independent experiments). ^∗∗ p < 0.01.

Figure 5

PT treatment resulted in caspase-3-dependent apoptosis in KYSE-450 cells. (a) Induction of apoptosis in KYSE-450 cells with medium alone (control), PT (10 nM), or Pan (10 nM) treatment for 16 h. Apoptosis proportion was measured by flow cytometry. (b) Statistical analysis of the percentage of the apoptotic cells. Data was shown with mean ± SD. ^∗∗∗ p < 0.001. (c) Apoptosis-related protein (PARP, cleaved caspase-3, BcL-XL, or Bcl-2) was examined in KYSE-450 cells when treated with medium alone (−), Pan (10 nM), or PT (10 nM) for 16 h.

Figure 6

In vivo efficacy of PT in the KYSE-450 xenograft tumor model. (a) Mean tumor volumes of mice xenografted with KYSE-450 cells and treated with PT. There were 6 mice per treatment group. PT treatment started as indicated in the graphs (black arrows). Error bars show ±SD (^∗∗∗ p < 0.001). (b) Histological examination was conducted in nude mice post 2 weeks after the last injection with Pan or PT. Representative images (magnification, ×400) of the liver from nude mice injected with indicated agents for two times were obtained by staining with hematoxylin and eosin. Scale bars: 100 μm. (c) Effect of PT on nude mice body weight was determined using KYSE-450 tumor-bearing nude mice. Mice were weighed at regular intervals during the whole period to monitor therapy-related toxicity.