Hedgehog signalling pathway orchestrates angiogenesis in triple-negative breast cancers

Abstract

Background: Several evidences suggest a marked angiogenic dependency in triple-negative breast cancer (TNBC) tumorigenesis and a potential sensitivity to anti-angiogenic agents. Herein, the putative role of Hedgehog (Hh) pathway in regulating TNBC-dependent angiogenesis was investigated.

Methods: Expression and regulation of the Hh pathway transcription factor glioma-associated oncogene homolog1 protein (GLI1) were studied on the endothelial compartment and on TNBC-initiated angiogenesis. To evaluate the translational relevance of our findings, the combination of paclitaxel with the Smo inhibitor NVP-LDE225 was tested in TNBC xenografted mice.

Results: Tissue microarray analysis on 200 TNBC patients showed GLI1 overexpression paired with vascular endothelial growth factor receptor 2 (VEGFR2) expression. In vitro, Hh pathway promotes TNBC progression in an autocrine manner, regulating the VEGF/VEGFR2 loop on cancer cell surface, and in a paracrine manner, orchestrating tumour vascularisation. These effects were counteracted by Smo pharmacological inhibition. In TNBC xenografted mice, scheduling NVP-LDE225 rather than bevacizumab provided a better sustained inhibition of TNBC cells proliferation and endothelial cells organisation.

Conclusions: This study identifies the Hh pathway as one of the main regulators of tumour angiogenesis in TNBC, thus suggesting Hh inhibition as a potential new anti-angiogenic therapeutic option to be clinically investigated in GLI1 overexpressing TNBC patients.

Figure 1

GLI1 expression correlates with VEGFR2 in TNBC patients. (A) Left, pie chart representing the percentage of samples included in the TMA that show immunoreactive score (IS) equal to 0, 1<IS<3 and IS>3 for VEGFR2 expression. Right, immunoistochemical images representing VEGFR2 negative (IS=0) or -positive tumours with moderate (1<IS<3) and high (IS>3) expression levels, respectively. Endothelial cells positivity represents internal control (× 20 magnification). The red arrows indicate representative highly VEGFR2 positive signal. (B) Left, pie chart representing the percentage of samples included in the TMA that show IS equal to 0, 1<IS<3 and IS>3 for GLI1 expression, respectively. Right, immunoistochemical images representing GLI1-negative (IS=0) or -positive tumours with moderate (1<IS<3) or high (IS>3) expression levels, respectively (× 20 magnification). The red arrows indicate representative highly GLI1 positive signal.

Figure 2

GLI1 is frequently overexpressed in TNBC primary and immortalised cell lines. (A, B) Western blot analysis of protein expression in a panel of A nTNBC or TNBC immortalised cell lines, and (B) nTNBC or TNBC primary cell lines. (C) Capillary tubes formation assay on HUVEC cells grown on Matrigel in: RPMI serum-free medium; RPMI serum-free medium supplemented with VEGF-A (100 ng ml⁻¹); conditioned growth media from nTNBC (MDA-MB-361) or TNBC (MDA-MB-468). Cells were observed under an inverted microscope, and pictures were taken at T0 hours and after 3 h. Scale bars, 100 μm. (D) Percent of cell density of bEND5 endothelial cells co-cultured with nTNBC or TNBC cells for 3 days, as measured by MTT assay. Data represent the mean (±s.d.) of three independent experiments, each performed in triplicate. (E, F) ELISA assay for the determination of VEGF-A concentrations (pg ml⁻¹) in conditioned media from (E) nTNBC or TNBC immortalised cells, or (F) primary cells. Data represent the mean (±s.d.) of three independent experiments, each performed in triplicate. Bars, s.d. Asterisks indicate statistical significance, as determined by the Student t-test (**P<0.01).

Figure 3

Hh pathway regulates the production of pro- and anti-angiogenic secreted factors. (A) Capillary tubes formation assay on HUVEC cells grown on Matrigel in the presence of conditioned growth media of TNBC (MDA-MB-468) cells treated with NVP-LDE225 (2.5 μM) or bevacizumab (1 μM).Cells were observed under an inverted microscope, and pictures were taken at T0 hours and after 3 h. Scale bars, 100 μm. (B) Quantitative analysis of VEGF-A in conditioned media (CM) of MDA-MB-231, SUM-159, SUM-149, HCC70 and MDA-MB-468 cells treated with NVP-LDE225 (2.5 μM) or bevacizumab (1 μM). (C, D) Quantitative analysis of (C) sVEGFR2 and (D) THBS1 in CM of MDA-MB-231, SUM-159, SUM-149, HCC70 and MDA-MB-468 cells treated with NVP-LDE225 (2.5 μM), as performed by using a multiplexed immunoassay. (E, F) Relative expression of (E) sVEGFR2 and (F) THBS1 mRNA in MDA-MB-231, SUM-159, SUM-149, HCC70 and MDA-MB-468 treated with NVP-LDE225 (2.5 μM), as performed by real-time RT–PCR (qRT-PCR) analysis. Data were calculated with mean cycle threshold (CT) values, normalised to endogenous control. Data represent the mean (±s.d.) of three independent experiments, each performed in triplicate. Bars, s.d. Asterisks indicate statistical significance, as determined by the Student t-test (*P<0.05, **P<0.01, ***P<0.001).

Figure 4

GLI1 regulates VEGFR2 expression. (A) Western blot analysis of protein expression in MDA-MB-468 and HUVEC cells treated with NVP-LDE225 (2.5 μM). (B) Relative expression of sVEGFR2 mRNA in MDA-MB-468 and HUVEC cells treated with NVP-LDE225 (2.5 μM), as performed by real-time RT–PCR (qRT-PCR) analysis. Data were calculated with mean cycle threshold (CT) values, normalised to endogenous control. Data represent the mean (±s.d.) of three independent experiments, each performed in triplicate. (C) Western blot analysis of protein expression in MDA-MB-468 and HUVEC cells, 24 and 48 h after transfection with scramble or GLI1 siRNA pool (50 nmol l⁻¹) using DharmaFECT 1 Transfection Reagent in DMEM. (D) Western blot analysis of protein expression in MDA-MB-231 cells, 24 h after transfection with either pCMV6-GFP empty vector, pCMV6-GFP GLI1 or pCMV6-GFP tGLI1 plasmids using lipofectamine 2000 in DMEM. (E) Relative luciferase units in MDA-MB-468 cells transfected with the empty pGL4 plasmid or the pGL4 plasmid containing 500 bp fragment of VEGFR2 promoter, and treated with NVP-LDE225 5 μM for 24 h after transfection. Luciferase activity was determined 48 h after transfection. Results were the average of three independent experiments. (F) Chromatin immunoprecipitation (ChIP) assay in MDA-MB-468 cells by using a GLI1 antibody and primers specific for the VEGFR2 promoter. Results were reported as fold change compared to negative control (no antibody); results were the average of three independent experiments. Bars, s.d. Asterisks indicate statistical significance, as determined by the Student t-test (**P<0.01). (G) Graphical representation of VEGFR2 proximal promoter region, containing the binding site of tGLI1 indicated by oval.

Figure 5

NVP-LDE225 increases the efficacy of paclitaxel in nude mice xenografted with TNBC tumours. (A) Tumour volume of MDA-MB-468 orthotopic xenografts in nude mice, randomised (10 per group) to receive NVP-LDE225 or bevacizumab in combination with paclitaxel, as described in the Methods section. The one-way ANOVA test was used to compare tumour sizes among treatment groups at the median survival time of the control group (35 days). Comparison of tumour sizes, evaluated by the one-way ANOVA test, was statistically significant for the combination NVP-LDE225 and paclitaxel vs control (P⩽0.05). (B) Number of surviving mice orthotopically xenografted with MDA-MB-468, after treatments with NVP-LDE225 or bevacizumab in combination with paclitaxel, as described in the Methods section. Median survival differences were statistically significant median survival in the NVP-LDE225 plus paclitaxel-treated mice was significantly longer than in control mice (79.50 vs 53.50 days, P=0.0089, log-rank test). (C) Western blot analysis on total lysates from MDA-MB-468 tumour specimens of mice killed on day 21, after 2 weeks of treatment with NVP-LDE225 or bevacizumab in combination with paclitaxel. (D, E, F, G) ELISA assay for the determination of (D) hVEGF-A, (E) hTHSB1, (F) h-sVEGFR2 and (G) m-sVEGFR2 concentrations (pg ml⁻¹) in mice sera, collected on day 21. Data represent the mean (±s.d.) of three independent experiments, each performed in triplicate. Bars, s.d. Asterisks indicate statistical significance, as determined by the Student t-test (*P<0.05, **P<0.01).