Blocking endothelin-1-receptor/β-catenin circuit sensitizes to chemotherapy in colorectal cancer

Abstract

The limited clinical response to conventional chemotherapeutics observed in colorectal cancer (CRC) may be related to the connections between the hyperactivated β-catenin signaling and other pathways in CRC stem-like cells (CRC-SC). Here, we show the mechanistic link between the endothelin-1 (ET-1)/ET-1 receptor (ET-1R) signaling and β-catenin pathway through the specific interaction with the signal transducer β-arrestin1 (β-arr1), which initiates signaling cascades as part of the signaling complex. Using a panel of patient-derived CRC-SC, we show that these cells secrete ET-1 and express ET_AR and β-arr1, and that the activation of ET_AR/β-arr1 axis promotes the cross-talk with β-catenin signaling to sustain stemness, epithelial-to-mesenchymal transition (EMT) phenotype and response to chemotherapy. Upon ET_AR activation, β-arr1 acts as a transcription co-activator that binds β-catenin, thereby promoting nuclear complex with β-catenin/TFC4 and p300 and histone acetylation, inducing chromatin reorganization on target genes, such as ET-1. The enhanced transcription of ET-1 increases the self-sustained ET-1/β-catenin network. All these findings provide a strong rationale for targeting ET-1R to hamper downstream β-catenin/ET-1 autocrine circuit. Interestingly, treatment with macitentan, a dual ET_AR and ET_BR antagonist, able to interfere with tumor and microenvironment, disrupts the ET-1R/β-arr1-β-catenin interaction impairing pathways involved in cell survival, EMT, invasion, and enhancing sensitivity to oxaliplatin (OX) and 5-fluorouracil (5-FU). In CRC-SC xenografts, the combination of macitentan and OX or 5-FU enhances the therapeutic effects of cytotoxic drugs. Together, these results provide mechanistic insight into how ET-1R coopts β-catenin signaling and offer a novel therapeutic strategy to manage CRC based on the combination of macitentan and chemotherapy that might benefit patients whose tumors show high ET_AR and β-catenin expression.

Figure 1

Expression of ET-1/β-arr1 axis in CRC-SC. (a) In a panel of patient-derived CRC-SC (CC09, CSC5, CSC2 and CSC1), expression of ET-1, ET_AR, ET_BR and β-arr1 mRNA copy number was analyzed by real-time PCR (qPCR) normalized using endogenous cyclophilin-A. Values are shown as mean±S.D. from three independent experiments repeated in triplicates. (b) ET_AR, ET_BR and β-arr1 protein expression was analyzed by immunoblotting (IB). Tubulin was used as loading control. (c) ET-1 secretion evaluated by ELISA in 24 h cell-conditioned media. Values are shown as mean±S.D. from three independent experiments repeated in triplicates. (d) Time-dependent effect of treatment with ET-1 (100 nM) and/or macitentan (MAC) (1 μM) on cell growth of CC09 cells transfected with siRNA negative control (SCR) or with β-arr1 siRNA (si-β-arr1). Values are shown as mean±S.D. from three independent experiments repeated in triplicates (*P<0.001 versus CTR; **P<0.001 versus ET-1) (e) Sphere formation assay of CSC5 cells, treated with ET-1 (100 nM) and MAC (1 μM) alone or in combination for 7 days. Values are shown as mean±S.D. from three independent experiments repeated in triplicates (*P<0.05 versus CTR; **P<0.05 versus ET-1). Representative images of tumor spheres were shown in the right panel

Figure 2

ET-1R/β-arr1 axis drives EMT process in CRC-SC. (a) Lysates from CC09 cells were analyzed by IB for the expression of epithelial (E-cadherin) and mesenchymal (N-cadherin, Snail, and vimentin) markers after ET-1 (100 nM) stimulation for 24 h. Tubulin was used as loading control. (b) Lysates from CC09 cells transfected with SCR or si-β-arr1 and treated for 24 h with ET-1 (100 nM) and/or MAC (1 μM) were analyzed by IB for the expression of E-cadherin, N-cadherin Snail and Twist. Tubulin was used as loading control. (c) Snail, E-cadherin, and N-cadherin expression in CC09 cells upon ET-1 (100 nM) and/or MAC (1 μM) treatment evaluated by qPCR, normalized using endogenous cyclophilin-A. Values are shown as mean±S.D. from three independent experiments repeated in triplicates (*P<0.002 versus CTR; **P<0.005 versus ET-1). (d) E-cadherin promoter activity and Snail promoter activity (e) evaluated in CC09 cells transfected with SCR or si-β-arr1 and treated for 24 h with ET-1(100 nM) and/or MAC (1 μM). Values are shown as mean±S.D. from three independent experiments repeated in triplicates (*P<0.01 versus CTR; **P<0.001 versus ET-1 in SCR-transfected cells). (f) Lysates from CC09 cells transfected with SCR or si-β-arr1 treated with ET-1 (100 nM) and/or MAC (1 μM) for 24 h were analyzed for MMP-2 and -9 by IB. Tubulin was used as loading control. (g) Conditioned media collected from CCO9 treated as in (f) were used to determine the secretion and activity of MMP-2 and -9 by gelatin zymography. (h) Cell invasion assay of CC09 cells transfected with SCR or si-β-arr1 and exposed to ET-1 (100 nM) and/or MAC(1 μM) for 24 h. Values are shown as mean±S.D. from three independent experiments repeated in triplicates (*P<0.002 versus CTR; **P<0.001 versus ET-1 in SCR-transfected cells). Representative images of invading cells were shown in the right panel

Figure 3

ET-1R/β-arr1 axis induces β-catenin pathway in CRC-SC. (a) β-catenin localization evaluated by immunofluorescence staining (green) in CC09 cells stimulated for 30 min with ET-1 (100 nM) alone or in combination with MAC (1 μM). Nuclei were counterstained with DAPI (blue) (scale bar, 10 μm). (b) IB analysis for β-catenin and β-arr1 protein expression in cytoplasmic and nuclear extract of CC09 cells treated with ET-1(100 nM) for the indicated times. HSP-70 and PCNA were used as cytoplasmic and nuclear loading control, respectively. (c) IB analysis for β-catenin and β-arr1 protein expression in nuclear extracts of CC09 cells transfected with SCR or si-β-arr1 and treated for 30 min with ET-1 (100 nM) and/or MAC (1 μM). Histone H3 (H3) was used as loading control. (d) Nuclear extracts of CC09 cells, treated for 30 min with ET-1 (100 nM) and/or MAC (1 μM), were immunoprecipitated (IP) with anti-β-arr1 or with irrelevant IgG (IgG) and immunoblotted with anti-β-catenin and anti-β-arr1. H3 was used as loading control. (e) β-catenin transcriptional activity evaluated in CC09 cells transfected with SCR or si-β-arr1 and treated for 24 h with ET-1 (100 nM) and/or MAC (1 μM). Values are shown as mean±S.D. from three independent experiments repeated in triplicates (*P<0.001 versus CTR; **P<0.002 versus ET-1 in SCR-transfected cells). (f) ET-1 promoter activity evaluated in CC09 cells treated as in (e). Values are shown as mean±S.D. from three independent experiments repeated in triplicates (*P<0.001 versus CTR; **P<0.001 versus ET-1 in SCR-transfected cells). ChIP analysis performed in CC09 cells treated for 30 min with ET-1 (100 nM) and/or MAC (1 μM) (g) or in CC09 cells transfected with SCR or si-β-arr1 and treated for 30 min with ET-1 (100 nM) and/or MAC (1 μM) (h). Chromatin was incubated with β-arr1, β-catenin, p300, H3K27ac and TCF4 Abs and analyzed by PCR analysis by using specific primers for ET-1 promoter. Non-specific immunoglobulin G (IgG) was used as irrelevant Ab (IRR). The input DNA lane represents one-twentieth of the precleared chromatin used in each ChIP reaction. (i) ET-1, Cyclin D1 and Axin 2 mRNA expression in CC09 cells stimulated for 24 h with ET-1 (100 nM) alone or in combination with MAC (1 μM) evaluated by qPCR, normalized using endogenous cyclophilin-A. Values are shown as mean±S.D. from three independent experiments repeated in triplicates (*P<0.01 versus CTR; **P<0.05 versus ET-1)

Figure 4

ET-1R blockade sensitizes CRC-SC to standard chemotherapeutic drugs. (a) Lysates from CC09 cells treated for 1 h with ET-1 (100 nM) and/or MAC (1 μM) and transfected with SCR or with si-β-arr1 were immunoblotted with anti-pp42/44MAPK, anti-p42/44MAPK, anti-pAkt and anti-Akt. (b) Effect of exposure to different concentrations of oxaliplatin (OX) and 5-fluorouracil (5-FU) after 24 h on cell vitality of CC09 cells. (c) Time-dependent effect of treatment with MAC (1 μM) or OX (100 μM) or 5-FU (50 μg/ml) alone and combination, for 24 h on cell growth of CC09 cells transfected with SCR or with si-β-arr1. Values are shown as mean±S.D. from three independent experiments repeated in triplicates (*P<0.001 versus CTR; **P<0.001 versus OX; ***P<0.002 versus 5-FU). (d) Effect of treatment with OX (100 μM) or 5-FU (50 μg/ml), for 24 h on cell growth of CC09 cells transfected with empty vector (Mock) or with β-arr1-FLAG. Values are shown as mean±S.D. from three independent experiments repeated in triplicates (*P<0.002 versus CTR; **P<0.002 versus chemotherapy-treated cells). (e) IB analysis of PARP cleaved form (cl PARP) in CC09 cells treated for 24 h with MAC (1 μM) or OX (100 μM) or 5-FU (50 μg/ml) alone and combination. Tubulin was used as loading control. (f) IB analysis of Bcl-XL in CC09 cells treated for 24 h with ET-1 (100 nM) or MAC (1 μM) or OX (100 μM) alone and combination. Tubulin was used as loading control

Figure 5

ET-1R blockade by macitentan inhibits tumor growth and restores sensitivity to oxaliplatin in CRC-SC xenografts. (a) CC09 cells (5 × 10⁵) were injected s.c. into the flank of nude mice. When tumors were detected, mice were treated with vehicle (CTR), or MAC (30 mg/Kg/oral daily), or OX (0.25 mg/Kg/i.p. once a week), or MAC (30 mg/Kg/oral daily) with OX (0.25 mg/Kg/i.p. once a week) combination for 4 weeks. The comparison of the time course of tumor growth curves by two-way ANOVA with group-by-time interaction for tumor growth was statistically significant (P<0.02). Data points, averages±S.D. The upper panels represented the hematoxylin–eosin staining of transplanted tumor xenografts (scale bar, 50 μm) or the images of tumors from each treatment group. (b) Expression of pp42/44MAPK, p42/44MAPK, pAkt and Akt as evaluated by IB on total extracts from tumors of CC09 xenografts. (c) E-cadherin, N-cadherin, Snail and vimentin, evaluated by IB of total extracts from tumors of CC09 xenografts

Figure 6

Macitentan inhibits tumor growth and restores sensitivity to 5-FU in CRC-SC xenografts. CC09 cells (5 × 10⁵) were injected s.c. into the flank of nude mice. When tumors were detected, mice were treated with vehicle (CTR), or MAC (30 mg/Kg/oral daily), or 5-FU (15 mg/Kg/i.p daily), or MAC (30 mg/Kg/oral daily) with 5-FU (15 mg/Kg/i.p daily) combination for 4 weeks. The comparison of the time course of tumor growth curves by two-way ANOVA with group-by-time interaction for tumor growth was statistically significant (P<0.05). Data points, averages±S.D.