MicroRNA-96-5p represses breast cancer proliferation and invasion through Wnt/β-catenin signaling via targeting CTNND1

Low miR-96-5p expression is characteristic of many cancers but its role in breast cancer (BCa) remains poorly defined. Here, the role of miR-96-5p in BC development was assessed. We demonstrate that exogenously expressing miR-96-5p inhibits the proliferative, migratory and invasive capacity of BCa cells. Mechanistically, miR-96-5p in BCa cells was found to target and downregulate catenin delta 1 (CTNND1) leading to decreased β-catenin expression, a loss of WNT11 signaling, reduced cyclin D1 levels and lower MMP7 expression. Exogenously expressing CTNND1 alleviated these effects. In summary, we are the first to reveal that miR-96-5p inhibits the proliferative, invasive and migratory phenotypes of BCa cells the targeting of CTNND1 and subsequent Wnt/β-catenin signaling. These data highlight miR-96-5p as a novel target for BC treatment.

Cell proliferation assay. BCa   www.nature.com/scientificreports www.nature.com/scientificreports/ were used to assess clinical features relative to miR-96-5p levels. Kaplan Meier (KM) curves were used for survival assessments and compared via log-rank tests. Cox-regression models were used for univariate/multivariate comparisons. P-values < 0.05 were considered significant (n = 3 for all experiments). MiR-96-5p expression correlates with BCa phenotypes. To characterize BCa phenotypes according to miRNA expression, two miR-96-5p subgroups were formed (high vs. low groups). As shown in Table 2, a significant correlation of reduced miR-96-5p expression in BCa with distant metastasis (P = 0.003) and TNM stage (P < 0.001) were evident suggesting a direct link to BCa progression. We further examined the association of miR-96-5p with BCa survival. In the low miR-96-5p group, recurrent-free survival (RFS) was 34.40 months compared to 43.40 months in the high-expression group, which significantly differed (low group vs. high group; P = 0.036; 95% CI: 1.03 to 2.32; Fig. 1C). Similarly, significant differences in the media overall survival (OS) of low-and high-expression groups were also observed (50.20 vs 76.90 months, respectively; low group vs. high group; P = 0.012; 95% CI: 1.13 to 2.74; Fig. 1D). Cox regression analysis of the 155 patients in the study cohort showed that both TNM staging and miR-96-5p levels could predict the OS and RFS of BCa patients (Table 3). Collectively, these data inferred a causative role for the loss of miR-96-5p expression for malignant BCa.

Results
MiR-96-5p suppresses BC cell metastasis. We next restored the levels of the miRNA through the transfection of miRNA mimics and assessed their effects on BCa cells. Successful miR-96-5p overexpression was confirmed by q-RTPCR (P < 0.05 in both cell lines, Fig. 2A) which markedly inhibited the proliferation of BCa cells (CCK-8 assays, P < 0.05 in both cell lines, Fig. 2B). Moreover, miR-96-5p overexpression inhibited the rates of colony formation (P < 0.05 in both cell lines, Fig. 2C) and reduced the migratory and invasive phenotypes of BCa cells (P < 0.05, Fig. 2D,E).
MiR-96-5p targets CTNND1 in BCa cells. Using TargetScan (http://www.targetscan.org/vert_72/) we screened a range of endogenous genes for potential miR-96-5p binding. The performed query identified that the well-characterized and important BC oncogene CTNND1 contains a complementary sequence for miR-96-5p (Fig. 3A). In view of this finding, the levels of CTNND1 were assessed in human BCa samples and adjacent  www.nature.com/scientificreports www.nature.com/scientificreports/ non-tumor tissues. As shown in Fig. 3B, CTNND1 mRNA levels were markedly higher in BC than in normal tissue (P < 0.05). Restoring the miRNA in BCa cell-lines downregulated CTNND1 (P < 0.05; Fig. 3C,D). Luciferase reporter assays confirmed that wt CTNND1 3′-UTR cells showed reduced luciferase levels upon the restoration of miR-96-5p (P < 0.05, Fig. 3E). In control experiments, modifying the levels of miR-96-5p had no effects on luciferase activity in mt CTNND1 3′-UTR cells (Fig. 3E). This confirmed CTNND1 as the target of miR-96-5p in BCa cells.

Discussion
Previous study have confirmed that miRNAs play key role in the progression of BCa 18 . The multiple effects exerted by miRNAs in BCa cells include the regulation of proliferative, apoptotic, autophagic, migratory, invasive, metastatic, epithelial-mesenchymal transition (EMT), angiogenic and drug resistant phenotypes 19 . Dysregulated miR-96-5p expression occurs in numerous cancers. Here, we reveal that the downregulation of this miRNA in BCa cells and tissues that is directly associated with low RFS and OS. Low levels of the miR-96-5p were further www.nature.com/scientificreports www.nature.com/scientificreports/ shown to correlate with TNM and stage and importantly, distant metastasis. This highlights its anti-BCa properties and utility as a biomarker for the prediction of both survival and metastasis in BCa.
Additionally, we show that the restoration of miR-96-5p levels inhibited BCa metastatic phenotypes, highlighting its tumor suppressor function in BCa. MiRNAs bind to the 3′-UTRs of target mRNAs to reduce post-transcriptional gene expression 20 . Previous studies using various cell types have documented several targets of miR-96-5p including Caveolae1 21 , CCDC67 22 , PTEN 23 , and caspase-9 24 . The current study expanded the repertoire of its targets through the identification of CTNND1. This conclusion was supported by several experiments. In BCa tissue, CTNND1 mRNA was inversely related to miR-96-5p. Earlier reports have shown that CTNND1 is regulated by various species of miRNAs. For example, miR-145 and miR-29c target CTNND1 and prevent metastatic phenotypes in gastric cancer cells 25,26 . Direct regulation of CTNND1 by miR-409c has also been documented in osteosarcoma 27 . In BC, high expression of CTNND1 is required for tumor growth and metastasis 17 . The function of CTNND1 as an oncoprotein is dependent on its ability to indirectly activate the Wnt/β-catenin pathway, known to induce BCa progression 17,28 . In this regard, we documented that both CTNND1 silencing and miR-96-5p overexpression inhibited the Wnt/β-catenin cascade and inhibit its downstream targets such as cyclin www.nature.com/scientificreports www.nature.com/scientificreports/ D1 and MMP7. Of these, Cyclin D1 is required for the promotion of BC progression by Wnt/β-catenin signaling 29 , and MMP7 activity enables migration and invasion of BC cells 30 . Consistent with this notion, exogenous CTNND1 expression reversed the miR-96-5p-induced inactivation of Wnt and β-catenin, promoting metastatic phenotypes in miR-96-5p-overexpressing BCa cell lines. Together, these data highlight the critical importance of the miR-96-5p/CTNND1 axis in BCa cells. Some study limitations should be noted: (1) The study lacked in vivo data for further confirmation of the role of miR-96-5p on BCa; (2) We did not explore whether miR-96-5p silencing could increase MCF-10A cell migration, invasion, and proliferation through CCTNND1/Wnt-β-catenin signaling; (3) We did not explore whether miR-96-5p inhibitors produced the opposite effects to miRNA mimics in the BCa cell lines. These factors now warrant further investigation.
In summary, we demonstrate that BCa is characterized by the downregulation of miR-96-5p, and that this alteration contributes to tumor progression. MiR-96-5p post-transcriptionally suppresses CTNND1 expression