Notch3 inhibits epithelial-mesenchymal transition by activating Kibra-mediated Hippo/YAP signaling in breast cancer epithelial cells

Abstract

Invasion, metastasis and chemoresistance are leading causes of death in breast cancer patients. A vital change of epithelial cells, epithelial-mesenchymal transition (EMT), is involved in these processes. Unfortunately, the molecular mechanisms controlling EMT remain to be elucidated. Our previous studies have shown that ectopic N3ICD expression inhibits EMT in MDA-MB-231, a triple-negative breast cancer (TNBC) epithelial cell line. To decipher the mechanism, we performed in-depth studies. Specifically, we found that overexpressing N3ICD transcriptionally upregulated the expression of Kibra, an upstream member of the Hippo pathway. Correspondingly, we also observed that phosphorylated Hippo pathway core kinases, including Lats1/2 and MST1/2, were increased and decreased by overexpressing and knocking down Notch3, respectively. Furthermore, we found that the oncogenic transcriptional coactivator yes-associated protein (YAP), which is negatively regulated by the Hippo pathway, was inhibited by overexpressing N3ICD in breast cancer epithelial cells. The ability of Kibra to inhibit EMT has been previously reported. We thus speculated that Notch3 inhibition of EMT is mediated by upregulated Kibra. To verify this hypothesis, a rescue experiment was performed. Evidently, the ability of Notch3 to inhibit EMT can be countered by knocking down Kibra expression. These data suggest that Notch3 inhibits EMT by activating the Hippo/YAP pathway by upregulating Kibra in breast cancer epithelial cells, and Kibra may be a downstream effector of Notch3. These findings deepen our understanding of EMT in both development and disease, and will undoubtedly help to provide new therapeutic strategies for interfering with cancer invasion and metastasis, especially for TNBC.

Figure 1

Notch3 expression in various breast cancer cell lines and the correlation between Notch3 and EMT. (a) Notch3 expression levels in various human breast cell lines, including MCF-7, T47D, SKBR3, MDA-MB-231 and BT549, were tested via western blotting. (b) Confirmation that Notch3 positively regulates E-cadherin expression via western blotting by overexpressing or knocking down Notch3 in MDA-MB-231 cells. (c) Confirmation that ectopically overexpressed Notch3 inhibits EMT by qRT-PCR. (d) Confirmation that knock down of Notch3 promotes EMT by qRT-PCR. Data are presented as the mean±s.d. of three experiments. *P<0.05, **P<0.01 and ***P<0.001 (Student's t-test) as compared with control cells.

Figure 2

Notch3 can reverse EMT induced by TGF-β1. (a) Morphological changes of MCF-7 cells before and after TGF-β1 treatment as well as TGF-β1 treatment combined with N3ICD overexpression. (b) RT-PCR confirms that MCF-7 cells undergo EMT after TGF-β1 treatment. (c) Western blotting confirms that MCF-7 cells undergo EMT after TGF-β1 treatment. (d) RT-PCR confirms that ectopically overexpressed N3ICD can reverse EMT induced by TGF-β1 in MCF-7 cells. (e) Western blotting confirms that ectopically overexpressed N3ICD can reverse EMT induced by TGF-β1 in MCF-7 cells. Data are presented as the mean±s.d. of three experiments. *P<0.05, **P<0.01 and ***P<0.001 (Student's t-test) as compared with control cells.

Figure 3

Kibra is regulated by Notch3 in a classical Notch signaling activation pattern. (a) RT-PCR confirms that ectopically overexpressed N3ICD can regulate upstream components of the Hippo/YAP signaling pathway. (b) Western blotting confirms that ectopically overexpressed N3ICD or Notch3 knockdown can activate or inhibit the core kinase cassette, respectively. (c) RT-PCR confirms that ectopically overexpressed N3ICD can activate the canonical signaling pathway. Data are presented as the mean±s.d. of three experiments. *P<0.05, **P<0.01 and ***P<0.001 (Student's t-test) as compared with control cells.

Figure 4

YAP target genes are downregulated when the Hippo/YAP signaling pathway is activated by ectopically overexpressed N3ICD. (a and b) RT-PCR confirms that ectopically overexpressed N3ICD-induced upregulation of Kibra expression is epithelial cell specific, as this phenomenon does not occur in the U87 glioblastoma cell line. (c and d) Western blotting and immunofluorescent staining confirm that non-phosphorylated YAP is excluded from the nucleus (the white arrows show non-phosphorylated YAP in the nucleus by western blot) and then is phosphorylated and degraded in MDA-MB-231 cells ectopically overexpressing N3ICD as compared with control cells. (e) RT-PCR confirms that YAP target genes are downregulated once the Hippo/YAP signaling pathway is activated by ectopically overexpressing N3ICD. Data are presented as the mean±s.d. of three experiments. *P<0.05, **P<0.01 and ***P<0.001 (Student's t test) as compared with control cells.

Figure 5

ChIP and luciferase assays confirm that Kibra is regulated by Notch3 at the transcriptional level. (a) The ChIP assay used normal IgG (IgG) or anti-Notch3 antibody to determine whether Notch3 can bind the RBP-Jk binding site in the Kibra promoter in MDA-MB-231 cells. After ChIP, PCR shows that the Notch3/RBP-Jk does not bind to the RBP-Jk binding site in the sense orientation. (b) After ChIP, PCR reveals that the Notch3/RBP-Jk complex binds to the RBP-Jk binding site in the antisense orientation. (c) MDA-MB-231 cells were co-transfected with pCLE, pCLE/N3, pGL-Wt-Luc or pGL-Mut-Luc. All cells were also co-transfected with a Renilla luciferase plasmid. Luciferase activity was normalized to that of Renilla. *P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001 (Student's t-test) as compared with control cells. Data are presented as mean±s.e.m. (n=3).

Figure 6

Confirming the ectopic overexpression and knockdown efficiency of Kibra. (a) The Kibra overexpression construct was engineered, and its successful transfection is shown by GFP expression. (b) RT-PCR confirms that the expression of Kibra mRNA is upregulated after MDA-MB-231 cells are transfected with pCMV plasmids. (c) Western blotting confirms that the expression of Kibra protein is upregulated after MDA-MB-231 cells are transfected with pCMV plasmids. (d) The Kibra knockdown construct was engineered and its successful transfection is shown by GFP expression. (e) RT-PCR confirms that the expression of Kibra mRNA is downregulated after MDA-MB-231 cells are transfected with each of the four knockdown plasmids, shKibra #5–#8. (f) Western blotting shows that Kibra protein in MDA-MB-231 cells is reduced after each of the four knockdown plasmids is transfected. Data are presented as the mean±s.d. of three experiments. *P<0.05, **P<0.01 and ***P<0.001 (Student's t-test) as compared with controls.

Figure 7

The effect of Notch3 inhibiting EMT can be counteracted by knocking down Kibra with shRNA. (a–d) After MDA-MB-231 cells were first transfected with pCLE/N3 for 48 h, Kibra expression was knocked down through a second transfection with shKibra #5–#8. RT-PCR results reveal that the effect of Notch3 inhibiting EMT was significantly attenuated at the transcriptional level as demonstrated by the expression of Notch3, Kibra, E-cadherin and slug mRNA. (e) MDA-MB-231 cells were subjected to the above-mentioned co-transfection. Kibra expression was knocked down through transfection with shKibra #5–#8. The effect of Notch3 inhibiting EMT was significantly attenuated at the protein level as demonstrated via western blotting to test the expression of Notch3, Kibra, E-cadherin and vimentin. *P<0.05, **P<0.01 and ***P<0.001 (Student's t-test) as compared with control cells. Data are presented as mean±s.e.m. (n=3)

Figure 8

Functional relationship between Notch3 and Kibra in breast cancer epithelial cells. (a) After shKibra constructs are transfected into MCF-7 cells, the expression level of Notch3 is significantly reduced. Correspondingly, significantly elevated vimentin and decreased E-cadherin expression levels are seen as compared with those of MCF-7 cells transfected with shNC. However, the expression of Notch1 is unchanged. (b) RT-PCR shows that the expression levels of Notch3 and E-cadherin mRNA are downregulated in MDA-MB-231 cells after shKibra#6 and #8 are transfected, whereas the expression level of vimentin was upregulated. (c) Nevertheless, when Kibra overexpression constructs are transfected into MDA-MB-231 cells, the expression of Notch3 is slightly upregulated (or unchanged), and the expression levels of vimentin and E-cadherin have no significant changes. The expression of Notch1 is only slightly inhibited. (d) RT-PCR analysis reveals that the expression levels of Notch3 and E-cadherin mRNA are upregulated in MDA-MB-231 cells, whereas the expression of vimentin mRNA is reduced when Kibra overexpression constructs are transfected into MDA-MB-231 cells. Data are presented as the mean±s.d. of three experiments. *P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001 (Student's t-test) as compared with control cells.

Figure 9

A working model of Notch3 and Kibra co-regulation in breast cancer cells. There is a complex inter-related and inter-dependent correlation between Kibra and Notch3. Loss of Notch3 leads to the decreased Kibra expression at both the transcriptional and protein levels. Downregulated Kibra, through the canonical Hippo kinase cascade, inhibits YAP phosphorylation, contributes to the nuclear accumulation of YAP, initiates target gene transcription and promotes EMT. Furthermore, loss of Kibra leads to downregulation of Notch3 expression. Conversely, ectopically overexpressed Notch3 upregulates Kibra expression, followed by an activated core kinase cassette, inhibited target gene transcription and finally suppressed EMT in breast cancer epithelial cells. In addition, ectopically overexpressed Notch3 upregulates the expression of the apical transmembrane protein Crb (Crumbs), which also interacts with Ex and modulates its localization and stability.