Perturbation of mammary epithelial cell apicobasal polarity by RHBDF1-facilitated nuclear translocation of PKCζ

Abstract

Background: The establishment of apicobasal polarity in epithelial cells is of critical importance in morphogenesis of mammary gland and other secretive gland tissues. The demise of the polarity is a critical step in early stages of tumorigenesis such as in breast ductal carcinoma in situ. The underlying molecular mechanism thus warrants in-depth investigations.

Results: Protein kinase C isoform ζ (PKCζ), which is highly expressed in breast cancer cells, accumulates in the nuclei of human mammary epithelial cells overexpressing human rhomboid family-1 (RHBDF1), an endoplasmic reticulum membrane protein. Nuclear translocation of PKCζ results in the failure of the formation of the cytosolic apicobasal polarity complex Par, of which PKCζ is an essential component. Additionally, enhanced nuclear translocation of PKCζ is accompanied by an inhibition of the expression of cell tight junction and adherens junction proteins and an increase of cell mobility. Mechanistically, RHBDF1 is able to interact with importin β1 and PKCζ and promote PKCζ phosphorylation. Consistently, treatment of RHBDF1-overexpressing cells with an inhibitor of PKCζ phosphorylation leads to restoration of apicobasal polarity and cell-cell junctions, as well as suppressed cell mobility.

Conclusions: RHBDF1-facilitated nuclear translocation of PKCζ is critically responsible for the dismantlement of epithelial cell apicobasal polarity, and thus may serve as a target in the development of therapeutic approaches against early stages of breast cancer.

Keywords: Adherens junction; Apicobasal polarity; Cell invasion; PKCζ; RHBDF1; Tight junction.

Fig. 1

RHBDF1-overexpression in mammary epithelial cell leads to elevated-invasion rate and downregulated TJ/AJ markers. (A) Experimental scheme for the induction of RHBDF1 by Dox in acini before and after polarization. (B) The confocal images of DAPI-stained equatorial cross sections of TetON-RH acini treated with Dox before and after polarization (Scale bar, 20 μm). (C) Quantification of the percentage of the hollowed/no hollowed acini in TetON-RH acini (n = 100). (D) Typical images of invasion of Dox (2 µg/mL) treated TetON-RH cells for 96 h, or then removed Dox for another 96 h; scale bar, 100 μm; (-) was solvent control for 96 h (Dox-off); (+) was Dox treated for 96 h (Dox-on); (±) was treated by Dox for 96 h, then removed and cultured for another 96 h (Dox-on/off); bar graphs show the ratio of invasion cells (n = 3). (E) Typical images of E-cadherin, Occludin or ZO-1 with RHBDF1 immunofluorescence-stained TetON-RH cells; scale bar 20 μm (n = 3). (F) The protein expression of HA-RHBDF1, E-cadherin, Occludin, ZO-1 in Dox stimulated TetON-RH in 3D acini cultures (n = 3). (G-J) The mRNA expression of RHBDF1 (G), Occludin (H), ZO-1 (I) and E-Cadherin (J) in Dox stimulated TetON-RH in 3D acini cultures (n = 3). These experiments were repeated two times under identical conditions. Data are mean ± SD, ^**P<0.01, ^*P<0.05, data in (C) is analyzed by two-way ANOVA; others are analyzed by one-way ANOVA

Fig. 2

RHBDF1 facilitates PKCζ activation and accumulation in cell nucleus. (A) Boxplot of the expression of PKCζ in breast cancer from TCGA database. N: Normal tissue; T: Tumor tissue. (B) Typical images of RHBDF1 (Left) and PKCζ (Right) immunofluorescence-staining of human breast cancer and para-cancerous tissues; scale bar 50 μm (number of patients n = 6). (C) The statistical analysis of fluorescence density of RHBDF1 and PKCζ in (B). (D, E) Typical images of PKCζ with RHBDF1 immunofluorescence-stained (D) MCF10A/GFP and MCF10A/RH, and (E) Dox-stimulated TetON-RH cells; scale bar 20 μm (n = 3). (F-I) The (F) protein and (G) mRNA levels of p-PKCζ and PKCζ in (F, G) MCF10A/GFP and MCF10A/RH, and (H, I) Dox-treated TetON-RH in 3D acini cultures (n = 3). (J) The levels of HA-RHBDF1 and PKCζ in nucleus and cytoplasm of MCF10A/GFP and MCF10A/RH cells (n = 3). These experiments were repeated two times under identical conditions. Data are mean ± SD, ^**P<0.01, ^*P<0.05, data in (G) and (I) are analyzed by one-way ANOVA, others are analyzed by Student-t test

Fig. 3

Blocking PKCζ phosphorylation leads to suppression of RHBDF1-facilitated disruption of cell polarity and TJ/AJ functions. (A) The expression of related proteins in MCF10A/RH acini with CRT0066854 (4 µM) stimulation on the day 4 and cultured for another 6 days in 3D cultures. (B, C) Typical images of E-cadherin, Occludin or ZO-1 with PKCζ immunofluorescence-stained MCF10A/RH with CRT0066854 treatment; scale bar 20 μm (n = 3). (D) The confocal images of DAPI-stained equatorial cross sections of MCF-10 A/RH acini with CRT0066854; scale bar 20 μm (n = 3). (E) The mRNA expression levels of related genes in CRT0066854 treated MCF10A/RH for 3D cultures. (F) Typical images of invasion for CRT0066854 stimulated MCF10A/GFP and MCF10A/RH cells; scale bar 100 μm; bar graphs show the ratio of invasion cells (n = 3). These experiments were repeated two times under identical conditions. Data are mean ± SD, ^**P<0.01, ^*P<0.05, data are analyzed by Student-t test

Fig. 4

RHBDF1-importin β1 interaction is needed for PKCζ nuclear translocation. (A, B) The Tyrosine phosphorylation of PKCζ (Tyr-p-PKCζ) and importin β1 protein expression in (A) MCF10A/GFP, MCF10A/RH and (B) Dox-treated TetON-RH cells/acini (n = 3). (C) Co-immunoprecipitation (Co-IP) analysis of the interaction between RHBDF1, importin β1 and PKCζ in MCF10A/GFP and MCF10A/RH under 3D cultures (n = 3). These experiments were repeated two times under identical conditions. Data are mean ± SD, data are analyzed by Student-t test

Fig. 5

Schematic representation of RHBDF1-promoted destruction of mammary epithelial cell apicobasal polarity and cell-cell connections