MiR-454-3p-Mediated Wnt/β-catenin Signaling Antagonists Suppression Promotes Breast Cancer Metastasis

Abstract

The Wnt/β-catenin pathway is constitutively active and promotes multiple tumor processes, including breast cancer metastasis. However, the underlying mechanism by which the Wnt/β-catenin pathway is constitutively activated in breast cancer metastasis remains unclear. Inhibition of Wnt antagonists is important for Wnt/β-catenin signaling activation, and post-transcriptional regulation of these antagonists by microRNAs (miRNAs) might be a possible mechanism underlying signaling activation. Regulation of nuclear pre-mRNA domain-containing 1A (RPRD1A) is a known inhibitor of cell growth and Wnt/β-catenin signaling activity, but the function and regulatory mechanism of RPRD1A in breast cancer have not been clarified. The aim of this study was to understand how regulators of the Wnt/β-catenin pathway may play a role in the metastasis of this cancer. Methods: RPRD1A expression and its association with multiple clinicopathological characteristics was analyzed immunohistochemically in human breast cancer specimens. miR-454-3p expression was analyzed using real-time PCR. RPRD1A or miR-454-3p knockdown and overexpression were used to determine the underlying mechanism of their functions in breast cancer cells. Xenografted tumor model, 3D invasive culture, cell migration and invasion assays and sphere formation assay were used to determine the biofunction of RPRD1A and miR-454-3p in breast cancer. Electrophoretic mobility shift assay (EMSA), luciferase reporter assay, and RNA immunoprecipitation (RIP) were performed to study the regulation and underlying mechanisms of RPRD1A and miR-454-3p expression and their correlation with the Wnt/β-catenin pathway in breast cancer. Results: The Wnt/β-catenin signaling antagonist RPRD1A was downregulated and its upstream regulator miR-454-3p was amplified and overexpressed in metastatic breast cancer, and both were correlated with overall and relapse-free survival in breast cancer patients. The suppression by miR-454-3p on RPRD1A was found to activate Wnt/β-catenin signaling, thereby promoting metastasis. Simultaneously, three other negative regulators of the Wnt/β-catenin pathway, namely, AXIN2, dickkopf WNT signaling pathway inhibitor (DKK) 3 and secreted frizzled related protein (SFRP) 1, were also found to be targets of miR-454-3p and were involved in the signaling activation. miR-454-3p was found to be involved in early metastatic processes and to promote the stemness of breast cancer cells and early relapse under both in vitro and in vivo conditions. Conclusions: The findings indicate that miR-454-3p-mediated suppression of Wnt/β-catenin antagonist RPRD1A, as well as AXIN2, DKK3 and SFRP1, sustains the constitutive activation of Wnt/β-catenin signaling; thus, miR-454-3p and RPRD1A might be potential diagnostic and therapeutic targets for breast cancer metastasis.

Keywords: RPRD1A; Wnt/β-catenin signaling; breast cancer; metastasis; miR-454-3p.

Figure 1

RPRD1A is downregulated in breast cancer and correlates with breast cancer metastasis. (A-B) Western blotting analysis of RPRD1A in 12 breast cancer cells (A) and 8 clinical breast cancer samples paired with their adjacent non-tumor tissues (B). α-tubulin served as the loading control. ANT: adjacent non-tumor; T: tumor. (C) IHC staining and statistical analyses of the average MOD of RPRD1A expression in normal breast tissues and breast cancer tissues without (M0; n=223) or with (M1; n=9) distant metastasis. Magnification, ×400 (left); enlarged images (right). The MOD value was determined with the Image-Pro Plus software. MOD: mean optical density. (D) Kaplan-Meier curves for breast cancer patients with low and high expression of RPRD1A. (E) Bioluminescence images of subcutaneous tumors showing distant metastasis signals (left, upper). Representative bright field images (left, middle) and quantification (right) of metastases in the lungs (arrows indicate surface metastatic nodules). Lung metastases in the mice were confirmed by H&E staining (left, lower). (F) Kaplan-Meier curves for RPRD1A-overexpressing mice (P = 0.002) compared with control mice (log-rank test). (G) Luciferase assay of TCF/LEF transcriptional activity in the indicated cells. Each bar represents the mean ± SD value from three independent experiments. *P < 0.05.

Figure 2

miR-454-3p directly suppresses RPRD1A in breast cancer cells. (A) Expression of RPRD1A in human breast cancer clinical specimens from the TCGA mRNA expression array data. (B) Illustration of miR-454-3p selection. Luciferase activity of reporters containing the 3ʹ-UTR of RPRD1A in the indicated miRNA-transfected cells was assessed for the selection of target genes (right). (C) Western blotting analysis of RPRD1A in the indicated cells. (D) Luciferase activities of reporters containing the 3ʹ-UTR of RPRD1A in miR-454-transduced cells, miR-454-3p-silenced cells, control cells or miR-454-3p-mutant-transfected cells. (E) Luciferase assay of TCF/LEF transcriptional activity in the indicated cells. (F) Subcellular β-catenin localization in the indicated cells was assessed by immunofluorescence staining. Magnification, ×200. (G) IHC staining of β-catenin in tissues with low and high miR-454-3p expression. Percentage of specimens (right) showing low or high miR-454-3p expression in relation to subcellular β-catenin localization. Magnification, ×200. (H) The DNA-binding activity of β-catenin as determined by the EMSA assay and the correlation between the DNA-binding activity of β-catenin and miR-454-3p expression in clinical breast cancer samples (n = 8) in comparison with normal breast tissues (n = 2). Each bar represents the mean ± SD value from three independent experiments. *P < 0.05.

Figure 3

miR-454-3p suppresses multiple key components in the Wnt/β-catenin cascade in breast cancer cells. (A) Luciferase assay of TCF/LEF transcriptional activity in the indicated cells. (B) Illustration of miR-454-3p downstream target gene selection. (C) miRNP immunoprecipitation assay revealed that AXIN2, DKK3, RPRD1A and SFRP1 mRNAs were recruited to miRNP complexes following immunoprecipitation with Ago2. IgG immunoprecipitation was used as a negative control. (D) Western blotting analysis of AXIN2, DKK3 and SFRP1 in the indicated cells. (E) Luciferase activities of reporters containing the 3ʹ-UTRs of AXIN2, DKK3 and SFRP1 in miR-454-3p-transduced cells, miR-454-3p-silenced cells, control cells or miR-454-3p-mutant-transfected cells. (F) Luciferase assay of TCF/LEF transcriptional activity in AXIN2-, RPRD1A-, DKK3- or SFRP1-expressing cells. (G) IHC staining of RPRD1A, AXIN2, DDK3 and SFRP1 in tissues with low and high miR-454-3p expression. Percentage of specimens (right) showing low or high miR-454-3p expression in relation to RPRD1A, AXIN2, DDK3 and SFRP1 expression. Each bar represents the mean ± SD value from three independent experiments. *P < 0.05.

Figure 4

miR-454-3p overexpression correlates with shorter survival in breast cancer patients and promotes in vivo metastasis of breast cancer. (A) Expression of miR-454-3p in breast cancer cell lines. (B) Expression of miR-454-3p in stage I-IV human breast cancer clinical specimens. (C-D) Kaplan-Meier curves for breast cancer patients with low and high expression of miR-454-3p. (E-F) Bioluminescence images of subcutaneous tumors showing distant metastasis signals (left). Representative bright field images (middle) and quantification of metastases (right) in the lungs (arrows indicate surface metastatic nodules). Lung metastases in the mice were confirmed by H&E staining (middle, lower). (G) Kaplan-Meier curves for the indicated mice (log-rank test). (H) Expression of miR-454-3p in the peripheral blood circulation of the indicated mice. (I) Expression of miR-454-3p in the serum of breast cancer patients with (M1) or without (M0) metastasis. Each bar represents the mean ± SD value from three independent experiments. *P < 0.05.

Figure 5

miR-454-3p contributes to early events in the breast cancer metastasis cascade. (A) H&E staining of the borders of primary tumors in the indicated mice. (B) 3D spheroid invasion assays of the indicated cells. (C-D) Immunofluorescence staining (C) and western blotting analysis (D) of EMT markers in the indicated cells. Magnification, ×400. (E) Colonies from peripheral blood samples of the indicated mice. Each bar represents the mean ± SD value from three independent experiments. *P < 0.05.

Figure 6

miR-454-3p promotes early distant relapse in breast cancer. (A) Images (left) and quantification (right) of tumor spheres formed from the indicated cells. (B) Bioluminescence images of distant metastasis signals in mice bearing the indicated cells injected through the tail vein (upper). Lung metastases in the mice were confirmed by H&E staining (lower image: arrows indicate surface metastatic nodules). The number of lung tumor nests in each group was counted under low-power fields (LPFs) (right). NOD/SCID mice were used in the study. Each bar represents the mean ± SD value from three independent experiments. *P < 0.05.

Figure 7

miR-454-3p is amplified in breast cancer. (A) Analysis of miR-454 copy number variation (CNV) in the TCGA dataset. (B) Expression of miR-454 corresponding to different amplification levels in the TCGA dataset. (C) Kaplan-Meier curves for miR-454 in miR-454 amplification and non-amplification groups (miR-454 amplification groups: CNV gain and amplification; miR-454 non-amplification groups: diploid CNV; P = 0.019, log-rank test). (D) Analysis of miR-454 CNV in non-metastatic (M0) and metastatic (M1) cases from the TCGA dataset.