Contrasting responses of non-small cell lung cancer to antiangiogenic therapies depend on histological subtype

Abstract

The vascular endothelial growth factor (VEGF) pathway is a clinically validated antiangiogenic target for non-small cell lung cancer (NSCLC). However, some contradictory results have been reported on the biological effects of antiangiogenic drugs. In order to evaluate the efficacy of these drugs in NSCLC histological subtypes, we analyzed the anticancer effect of two anti-VEGFR2 therapies (sunitinib and DC101) in chemically induced mouse models and tumorgrafts of lung adenocarcinoma (ADC) and squamous cell carcinoma (SCC). Antiangiogenic treatments induced vascular trimming in both histological subtypes. In ADC tumors, vascular trimming was accompanied by tumor stabilization. In contrast, in SCC tumors, antiangiogenic therapy was associated with disease progression and induction of tumor proliferation. Moreover, in SCC, anti-VEGFR2 therapies increased the expression of stem cell markers such as aldehyde dehydrogenase 1A1, CD133, and CD15, independently of intratumoral hypoxia. In vitro studies with ADC cell lines revealed that antiangiogenic treatments reduced pAKT and pERK signaling and inhibited proliferation, while in SCC-derived cell lines the same treatments increased pAKT and pERK, and induced survival. In conclusion, this study evaluates for the first time the effect of antiangiogenic drugs in lung SCC murine models in vivo and sheds light on the contradictory results of antiangiogenic therapies in NSCLC.

Figure 1

Sunitinib induces a differential tumor response between adenocarcinoma (ADC) and squamous cell carcinoma (SCC). A, B In the urethane-induced ADC mouse model, sunitinib (A) and DC101 (B) treatments stabilized tumor growth; Top: Waterfall plots of tumor response to sunitinib (A) or DC101 (B) treatments of ADC-bearing mice (the bold line at 20% indicates disease progression as defined by response in human cancers criteria). Tumor diameter change was calculated for each mouse taking the initial size of every lesion as reference. Each column represents the average tumor diameter change in each mouse after 5 weeks of treatment. Bottom: The pre-and post-treatment mean size for each group is presented in a line plot. Representative micro-CT images of ADC tumors show the progression of the disease in the control group and the stabilization of tumor growth in the sunitinib group. Lung ADC lesions (arrows) are shown at higher magnification (inset). Data are presented as mean ± standard error. C, D In the N-nitroso-tris-chloroethylurea-induced SCC model disease progression was observed. Top: Waterfall plots of SCC tumor response to sunitinib (C) or DC101 (D) treatments showing the percentage of area change for each mouse. Bottom: The pre-and post-treatment mean tumor area for each group is presented in a line plot. Representative micro-CT images of SCC tumors (dotted line) show the progression of the disease in the sunitinib and vehicle (control) groups. Data are presented as mean ± standard error.

Figure 2

VEGFR2 blockade differentially affects adenocarcinoma (ADC) and squamous cell carcinoma (SCC) tumor cell proliferation and apoptosis. A–D Representative images of immunohistochemical analysis of Ki67 and cleaved caspase 3 in ADC tumor-bearing mice. In the urethane-induced ADC model, sunitinib (A) and DC101 (B) reduced the tumor cell proliferation index. Increased tumor apoptosis as measured by cleaved caspase 3 staining was observed in sunitinib (C) and DC101 (D) groups. Scale bar, 25 m. Data are presented as mean ± standard error. E–H Representative images of immunohistochemical analysis of Ki67 and cleaved caspase 3 in SCC. Sunitinib (E) and DC101 (F) treatments in SCC tumors significantly induced proliferation as compared to vehicle. The percentage of apoptotic cells was increased in the sunitinib group (G), but not in the DC101 group (H). Scale bar, 25 μm. Data are presented as mean ± standard error.

Figure 3

VEGFR2 inhibitors induce contrasting responses in adenocarcinoma (ADC) and squamous cell carcinoma (SCC) tumorgrafts. Immunodeficient mice bearing subcutaneous ADC or SCC tumors (using UN-ADC12 or UN-SCC680 cell lines, respectively) were treated with: sunitinib (40 mg/kg) or vehicle (PBS), or DC101 (40 mg/kg) or isotype antibody (40 mg/kg). Treatment was initiated at day 17 post-injection in the ADC model and at day 25 post-injection in the SCC model. The quantification of Ki67 and cleaved caspase 3 expression after sunitinib and DC101 treatments was performed automatically. Data are presented as mean ± standard error (n = 6 for all groups).

Figure 4

Anti-VEGFR2 therapies induce opposite effects on cell survival and VEGFR2 downstream signaling in vitro. A MTT cell proliferation assay of adenocarcinoma (ADC)-(UN-ADC12 and UN-ADC18) and squamous cell carcinoma (SCC)-(UN-SCC679 and UN-SCC680) tumor-derived cell lines treated with sunitinib. Data are presented as mean ± standard error. B, C Colony formation assay of ADC-and SCC tumor-derived cell lines treated with sunitinib (B) or DC101 (C). Data are presented as mean ± standard error. D, E Representative western blotting images of pAKT and pERK levels showing a differential activation of VEGFR2 downstream signaling in ADC (D) and SCC (E) cell lines. Source data are available for this figure.

Figure 5

Sunitinib induces the expression of stem cell markers in squamous cell carcinoma tumors. A, B ALDH1A1 expression. Representative images of immunohistochemical analysis of ALDH1A1 in vehicle (control) and sunitinib-treated mice (A). Automatic quantification of tumor cell staining was performed. The mRNA expression of ALDH1A1 was evaluated by quantitative PCR (B). Each value was normalized to the reference gene (GUSB) and the expression in paired normal tissue was used for calibration. Scale bar, 25 μm. Data are presented as mean ± standard error. C, D CD133 expression analysis. Representative analysis of CD133 expression by immunohistochemistry (C) and by quantitative PCR (D). Each value was normalized to the reference gene (GUSB) and the expression in paired normal tissue was used for calibration. Scale bar, 25 m. Data are presented as mean ± standard error. E, F CD15 expression analysis by immunohistochemistry (E) and by quantitative PCR (F). Each value was normalized to the reference gene (GUSB) and the expression in paired normal tissue was used for calibration. Scale bar, 25 μm. Data are presented as mean ± standard error.

Figure 6

DC101 treatment induces the expression of stem cell markers in squamous cell carcinoma tumors. The expression of stem cell markers ALDH1A1, CD133, and CD15 was analyzed by immunohistochemistry (A, C, E) and quantitative PCR (qPCR) (B, D, F). Automatic quantification of tumor cell staining was performed. For qPCR, each value was normalized to the reference gene (GUSB) and the expression of paired normal tissue was used as calibrator sample. Data are presented as mean standard error. The expression of stem cell markers ALDH1A1, CD133, and CD15 was analyzed by immunohistochemistry (A, C, E) and quantitative PCR (qPCR) (B, D, F). Automatic quantification of tumor cell staining was performed. For qPCR, each value was normalized to the reference gene (GUSB) and the expression of paired normal tissue was used as calibrator sample. Data are presented as mean ± standard error. A, B ALDH1A1 expression. C, D CD133 expression E, F CD15 expression