Recombinant human endostatin enhances the radioresponse in esophageal squamous cell carcinoma by normalizing tumor vasculature and reducing hypoxia

Abstract

The aim of this study was to investigate the effect of recombinant human endostatin (rh-Endo) in combination with radiation therapy (RT) on esophageal squamous cell carcinoma (ESCC) and explore the potential mechanisms. ECA109-bearing nude mice were administered RT and/or rh-Endo treatment. Tumor volume, survival, hypoxia and vascular parameters were recorded during the treatment schedule and follow-up as measures of treatment response. ESCC cell lines (ECA109 and TE13) and human umbilical vein endothelial cells (HUVECs) were developed to investigate the outcomes and toxicities of rh-Endo and RT in vitro. Hypoxia inducible factor-1α (HIF-1α) and vascular endothelial growth factor (VEGF) were also evaluated. In vivo studies of ECA109-bearing xenografts showed that rh-Endo improved the radioresponse, with normalization of tumor vasculature and a reduction in hypoxia. In vitro studies showed that rh-Endo did not radiosensitize ESCC cell lines but did affect endothelial cells with a time- and dose-dependent manner. Studies of the molecular mechanism indicated that the improved radioresponse might be due to crosstalk between cancer cells and endothelial cells involving HIF and VEGF expression. Our data suggest that rh-Endo may be a potential anti-angiogenic agent in ESCC especially when combined with RT. The improved radioresponse arises from normalization of tumor vasculature and a reduction in hypoxia.

Figure 1. rh-Endo improves the radioresponsiveness of ESCC in vivo.

ECA109 xenograft-bearing male BALB/c mice were divided into four treatment groups (n = 6): control; IR alone; rh-Endo alone; and a combination of rh-Endo and irradiation. 10 days after inoculation of ECA109 cells (1 × 10⁶ cells/mouse), the mice were irradiated with a single fraction of 8-Gy X-rays. (a) Measurements of tumor size. Data represent the mean tumor volume; error bars represent the SD. **P < 0.01 vs. the control group; ^#P < 0.05 vs. the rh-Endo group. (b) Representative images of ECA109 xenograft-bearing mice. (c) Comparison of tumor doubling times between the four groups. *P < 0.05. (d) Kaplan-Meier survival curves for the mice in the four groups.

Figure 2. rh-Endo normalizes ESCC vascularization in vivo.

Tumor tissues from mice in the four treatment groups were removed and stained with anti-CD31, anti-CD105, anti-Lectin and anti-α-SMA antibodies on days 10, 16 and 22 (D10, D16 and D22) after transplantation of ECA109 cells. (a) Representative images of CD31 immunostaining (used as a microvessel marker). (b) Microvessel densities determined from CD31 immunostaining in the various experimental groups. (c) Representative images showing immunostaining for CD105 (a marker of new vessels), Lectin (a marker of functional vessels) and α-SMA (a marker of pericyte coverage). (d) Integral optical densities (IODs) of the various experimental groups (upper, CD105; middle, Lectin; lower, α-SMA) on D22 (the end of the treatment period). *P < 0.05 vs. IR group, **P < 0.01 vs. IR group.

Figure 3. rh-Endo improves ESCC tissue hypoxia in vivo.

Immunofluorescence assessment of tumor tissues were conducted on D22 (the end of the treatment period). (A) Pimonidazole staining showed an improvement of hypoxia in the rh-Endo + IR group. (B) Quantify analysis of CD31 and pimonidazole. *P < 0.05 vs. control, **P < 0.01 vs. control.

Figure 4. rh-Endo does not improve radiosensitivity of EC cell lines in vitro but does affect endothelial cells.

(a–c) rh-Endo did not inhibit the growth of ECA109 and TE13 cells at lower concentrations (200 μg/mL and below) but did inhibit the growth of HUVECs at these lower concentrations. *P < 0.05 vs. 0 μg/mL at the same time point; **P < 0.01 vs. 0 μg/mL at the same time point. (d–f) Flow cytometric analysis revealed that rh-Endo did not induce apoptosis of ECA109 cells and TE13 cells, but did affect the apoptosis of HUVECs. **P < 0.01. (g,h) Clonogenic assays showed that rh-Endo did not affect the radioresponsiveness of ESCC cells in vitro, whereas DDP improved the sensitivity to radiation.

Figure 5. The improved radioresponse with rh-Endo is associated with decreased expression of HIF-1α and VEGF.

(a) Serum VEGF-A levels measured in vivo in the four experimental groups. (b) Supernatant levels of VEGF-A from ECA109 cells, TE13 cells and HUVECs normalized to cell numbers. (c,d) Immunofluorescence assessment of VEGF and HIF-1α in the four experimental groups on D22. (e,f) Quantity analysis of c and d. *P < 0.05 vs. control or 0 μg/mL, **P < 0.01 vs. control or 0 μg/mL.