TrkC promotes colorectal cancer growth and metastasis

Abstract

The current work reveals that TrkC receptor is crucial to many aspects of tumorigenicity and metastasis of cancer. However, with only a few exceptions, such as colorectal cancer (CRC), where suppressing tumorigenic and metastatic ability via expression of TrkC as tumor suppressor have been proposed. These diverse lines of evidence led us to investigate whether TrkC is involved in CRC progression. By using mouse models and molecular biology analyses, we demonstrate that TrkC acts as an activator in tumorigenicity and metastasis of colorectal cancer. In this study, TrkC was frequently overexpressed in CRC cells, patients' tumor samples and an azoxymethane/dextran sulphate sodium-induced mouse model of colitis-associated CRCs. TrkC expression was associated with a high-grade CRC phenotype, leading to significantly poorer survival. Also, TrkC expression promoted the acquisition of motility and invasiveness in CRC. Moreover, TrkC increased the ability to form tumor spheroids, a property associated with cancer stem cells. Importantly, knockdown of TrkC in malignant mouse or human CRC cells inhibited tumor growth and metastasis in a mouse xenograft model. Furthermore, TrkC enhanced metastatic potential and induced proliferation by aberrant gain of AKT activation and suppression of transforming growth factor (TGF)-β signalling. Interestingly, TrkC not only modulated the actions of TGF-β type II receptor, but also attenuated expression of this receptor. These findings reveal an unexpected physiological role of TrkC in the pathogenesis of CRC. Therefore, TrkC is a potential target for designing effective therapeutic strategies for CRC development.

Keywords: EMT program; TrkC; colorectal cancer; metastasis; tumorigenicity.

Figure 1. Correlation of TrkC with CRC pathogenesis and patient survival

(A) Box-and-whisker (Tukey) plots of the mean expression of TrkC and NT-3 in CRC patients. TrkC and NT-3 levels were extracted from the Skrzypczak microarray dataset (GSE20916) and averaged in each tumor. Points below and above the whiskers are drawn as individual dots. P < 0.05 was considered to indicate significance in ANOVA. (B) TrkC expression is correlated with the stages of CRC. Mean expression of TrkC and NT-3, obtained through RNA-sequence analysis of 629 CRC patients in the TCGA dataset, were plotted as box plots according to the tumor stages. TrkC and NT-3 levels were extracted from the dataset and averaged in each tumor. Points below and above the whiskers are drawn as individual dots. P < 0.05 was considered to indicate significance in ANOVA. NS, not significant. (C) TrkC expression is correlated with recurrence in CRC patients, but NT-3 expression is not. Mean expression of TrkC and NT-3, obtained by RNA-sequence analysis of 629 CRC patients in the TCGA dataset, was plotted as box plots according to the disease-free status of CRC patients. TrkC and NT-3 levels were extracted from the dataset and averaged in each tumor. Points below and above the whiskers are drawn as individual dots. The Student's t-test was performed to assess statistical significance (*P < 0.05). (D) Mean methylated TrkC expression, obtained by analysis of the Infinium Human Methylation 450 BeadChip array (HM450) of 331 CRC patients in the TCGA dataset, was plotted as box plots. TrkC levels were extracted from the dataset and averaged in each tumor. Points below and above the whiskers are drawn as individual dots. P < 0.05 was determined by the Student's t-test. NS, not significant. (E, F) In total, 629 CRC patients from the TCGA dataset were divided into high and low TrkC or NT-3 expressers, and overall (E) and recurrence-free (F) survival were compared. P values correspond to the log-rank test comparing the survival curves.

Figure 2. Expression patterns of TrkC in human CRC cells and samples

(A) Schematic overview of the AOM/DSS CRC model. (B) Representative image of colon adenocarcinomas formation on day 91 of AOM/DSS-induced colon cancer. (C) The relative levels of TrkC expression in the distal colon from five AOM/DSS-treated and five control mice were assessed by TaqMan real-time quantitative PCR analysis. The endogenous 18S mRNA level was measured as the internal control. The Student's t-test was performed to assess statistical significance (*P < 0.05). (D) Western blot analysis of TrkC expression in the distal colon from five AOM/DSS-treated and five control mice. (E) Expression of TrkC mRNA in a panel of human normal colon (CCD841 CoN and CCD112 CoN) or CRC cells was examined by TaqMan real-time quantitative PCR analysis. The endogenous 18S mRNA level was measured as the internal control. Error bars represent the mean ± SD of triplicate experiments. P < 0.05 was considered to indicate significance in ANOVA. (F) The relative levels of TrkC expression of individual human 26 normal or 26 CRC samples were assessed by TaqMan real-time quantitative PCR analysis. Expression was compared with that in healthy tissue. The endogenous 18S mRNA level was measured as the internal control. The Student's t-test was performed to assess statistical significance (*P < 0.05). (G) Immunohistochemical analysis of TrkC protein levels in human colon normal, colon adenocarcinoma and metastatic colon adenocarcinoma in lymph nodes. E-cadherin was measured as the epithelial maker.

Figure 3. Contribution of TrkC to the metastatic ability of CRC

(A) CT26 cells infected with the indicated shRNAs grown for 7 days in ultra-low cluster plates and visualized with bright-field microscopy. (B) Western blot analysis of expression of phospho-AKT, phospho-MEK1/2 and cyclin D1 in CT26, WiDr and SW480 cells infected with the indicated shRNAs. β-actin was used as a loading control. (C) Colony-forming assay of CT26, WiDr and SW480 cells infected with the indicated shRNAs (n = 3). The Student's t-test was performed to assess statistical significance (P < 0.05). (D) Migration assay of control-shRNA- or shTrkC-treated CT26, WiDr and SW480 cells. Cells that migrated to the bottom of the chamber were counted in five fields (n = 3). The Student's t-test was performed to assess statistical significance (P < 0.05). (E) Wound healing assay of WiDr and SW480 cells infected with the indicated shRNAs. Wound closures were imaged at 0, 12 and 24 h after wounding.

Figure 4. Suppression of TrkC expression inhibits tumorigenicity and metastasis of CRCs

(A) Tumor formation by WiDr cells infected with the indicated shRNAs. In total, 1.0 × 10⁵ cells were implanted into the mammary fat pads of mice (n = 7). The Student's t-test was performed to assess statistical significance (P < 0.0005). (B) Representative images of tumors from mice harboring WiDr cells infected with the indicated shRNAs. (C) Tumor weights from mice harbouring WiDr cells expressing control-shRNA or TrkC-shRNA. (n = 7). The Student's t-test was performed to assess statistical significance (P < 0.0005). (D) Tumor formation by SW480 cells infected with the indicated shRNAs. In total, 1.0 × 10⁵ cells were implanted into the mammary fat pads of mice (n = 7). The Student's t-test was performed to assess statistical significance (P < 0.0005). (E) Representative images of tumors from mice harbouring SW480 cells infected with the indicated shRNAs. (F) Tumor weights from mice harbouring SW480 cells infected with the indicated shRNAs (n = 7). The Student's t-test was performed to assess statistical significance (P < 0.0005). (G) Representative images of lungs and immunohistochemical images of haematoxylin and eosin staining in sections of lungs from individual mice harbouring CT26 cells infected with the indicated shRNAs. (H) Representative images and total number of lung metastatic nodules in each mouse in each group after tail vein injection of SW480 control-shRNA or TrkC-shRNA cells (n = 7). The Student's t-test was performed to assess statistical significance (P < 0.0005).

Figure 5. TrkC as a key regulator of the EMT program and maintenance of the CSC state

(A) mRNA expression levels of E-cadherin, N-cadherin, fibronectin and vimentin in WiDr and SW480 cells infected with the indicated shRNAs. 18S mRNA was used to normalise variability in template loading. The Student's t-test was performed to assess statistical significance (P < 0.05). (B) Western blot analysis of the expression of E-cadherin, N-cadherin, fibronectin and vimentin proteins in SW480 cells infected with the indicated shRNAs. β-actin was used as a loading control. (C) Tumor spheroids formation assay of CT26, WiDr and SW480 cells infected with the indicated shRNAs. Cells were counted in five fields (n = 3). The Student's t-test was performed to assess statistical significance (*P < 0.001). (D) Tumor spheroids formation assay of control or RIE-1-TrkC cells. Cells were counted in five fields (n = 3). The Student's t-test was performed to assess statistical significance (*P < 0.001).

Figure 6. Knockdown of TrkC restores TGF-β signalling

(A) Western blot analysis of the expression of phospho-Smad2, phospho-Smad3, Smad2 and Smad3 proteins in CT26 cells infected with the indicated shRNAs after stimulation with TGF-β1 (5 ng/mL). β-actin was used as a loading control. (B) Western blot analysis of the expression of phospho-Smad2, phospho-Smad3, Smad2 and Smad3 proteins in WiDr cells infected with the indicated shRNAs after stimulation with TGF-β1 (5 ng/mL). β-actin was used as a loading control. (C) Western blot analysis of the expression of phospho-Smad2, phospho-Smad3, Smad2 and Smad3 proteins in SW480 cells infected with the indicated shRNAs after stimulation with TGF-β1 (5 ng/mL). β-actin was used as a loading control. (D, E) Luciferase reporter assay of TGF-β1-responsive SBE (D) or 3TP (E) in CT26 cells infected with the indicated shRNAs. Luciferase activity was measured 24 h after treatment with TGF-β1. **Control versus treatment with TGF-β1, P < 0.001. n = 3. (F, G) Luciferase reporter assay of TGF-β1-responsive SBE (F) or 3TP (G) in WiDr cells infected with the indicated shRNAs. Luciferase activity was measured 24 h after treatment with TGF-β1. **Control versus treatment with TGF-β1, P < 0.001. n = 3. (H, I) Luciferase reporter assay of TGF-β1-responsive SBE (H) or 3TP (I) in SW480 cells infected with the indicated shRNAs. Luciferase activity was measured 24 h after treatment with TGF-β1. *Control versus treatment with TGF-β1, P < 0.001. n = 3.

Figure 7. TrkC blocks TβRII/TβRI complex formation via TrkC/TβRII interaction

(A) Identification of TrkC/TβRII complexes in control and RIE-1-TrkC cells. Cell lysates were subjected to immunoprecipitation using an anti-TβRII antibody followed by immunoblotting with the indicated antibodies. β-actin was used as a loading control. (B) Identification of endogenous TrkC/TβRII complexes in RIE-1 and CT26 cells. Cell lysates were subjected to immunoprecipitation using anti-IgG and anti-TβRII antibodies followed by immunoblotting with the indicated antibodies. β-actin was used as a loading control. (C) Identification of endogenous TrkC/TβRII complexes in CT26 cells infected with the indicated shRNAs. Cell lysates were subjected to immunoprecipitation using an anti-TβRII antibody followed by immunoblotting with the indicated antibodies. β-actin was used as a loading control. (D) Identification of TrkC/TβRII complexes with or without TGF-β1. Immunoblot analysis of whole-cell lysates and immunoprecipitates derived from 293T cells transfected with V5-TrkC and Myc-TβRII constructs as indicated. (E) Immunoblot analysis of whole-cell lysates and immunoprecipitates derived from 293T cells transfected with V5-TrkC deletion constructs and Myc-TβRII constructs as indicated.