MicroRNA-223 Targeting STIM1 Inhibits the Biological Behavior of Breast Cancer

Breast cancer is the highest incidence of cancer among women in China [1] and United States [2]. In recent years, great achievements have been made in early detection and individual treatment, however, recurrence and metastasis are still the main reasons for the majority of cancer-related death, which accounts for about 30% of these patients dying of metastatic disease within five years [3].

Stromal interaction molecule 1 (STIM1) is an endoplasmic reticulum Ca²⁺ sensor protein which maintains intracellular calcium homeostasis and regulates intracellular calcium concentrations. Ca²⁺ is a crucial second messenger modulating various cellular processes，thereby STIM1 is also involved in a variety of biological processes including immune activation [4], cardiovascular damage [5], pulmonary disease [6] and so on. Moreover, it is demonstrated that STIM1 is also associated with the development and metastasis of several cancers [7-10] and has emerged as a target for cancer therapeutics [11]. Yang and his colleagues [12] were the first to report that STIM1 mediated breast cancer cells migration and metastasis in 2009. Two subsequent studies also confirmed the role of STIM1 in breast cancer metastasis [13] and proliferation [14]. These studies suggest that STIM1 may be a potential target to block the progression of breast cancer.

MicroRNAs (miRNAs) are endogenous, small non-coding RNAs with less than 25 nucleotides. They usually negatively regulate a large of genes expression by binding to the 3’-untranslated region (3’-UTR) of target mRNAs [15]. Several studies have shown that miRNAs act as a means of anticancer and play an important role in regulating the invasion and migration of cancer [16-18]. Notably, recent studies revealed that a variety of miRNAs, including miR-152 [19], miR-381 [20], miR-211 [21], miR-223 [22] and so on, could inhibit tumor proliferation, invasion and metastasis in breast cancer by targeting various genes. However, the miRNAs targeting STIM1 in breast cancer are still poorly characterized. Hence, understanding which miRNA targeting STIM1to drive tumor proliferation and metastasis remains of great interest.

In this study, we are committed to find out the miRNAs that regulate STIM1 expression and function in breast cancer. Our results demonstrate that miR-223 interacts with the STIM1 3’UTR and negatively regulates endogenous STIM1 expression. Meanwhile, overexpression of miR-223 and downregulation of STIM1could attenuate breast cancer cellular proliferation, invasion and decrease patient’s recurrence. Taken together, our findings uncover a new regulatory mechanism of STIM1 by miR-223 which may be a therapeutic intervention against breast cancer.

MCF-10A, MCF-7, T-47D, SKB-R3, MDA-MB-231 and MDA-MB-453 cells were purchased from the Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, (Chinese Academy of Science, Shanghai, China). Those cells were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum and maintained at 37°C in humidified air with 5% CO₂. Once the density of the cell growth reached 80%, it would be handed from generation to generation.

For the transfection assay, cells were transfected with miR-223 mimics (artificially synthesized miR-223 mimics) and anti-miR-223 (miR-223 inhibitor), meanwhile their non-targeting sequence were used as control groups (NC group). For STIM1 rescue experiments, the coding domain sequence of human STIM1 mRNA was inserted into pcDNA 3.1 vector. All the RNAs were synthesized from Shanghai Gene Pharma Co., Ltd., Shanghai, China. The whole transfection experiments were mediated with Lipofectamine2000 for 6 hours, after that the cells were replaced with fresh culture medium and kept on culturing for 24, 48 or 72 hours.

We collected 165 tumor tissue specimens and 60 corresponding adjacent normal breast tissues preserved in the Tumor Tissue Bank of Tianjin Cancer Hospital from the earlier operated BC patients between January 2007 and December 2009. The protocol was approved by the Institutional Review Board of Tianjin Cancer Institute and Hospital. All patients signed a written consent for the use of their specimens and disease information for future investigations according to the ethics committee.

The total protein of cells was extracted by lysing cells with SDS lysis buffer supplemented with a protease inhibitor cocktail (Sigma). 30 µg of protein lysates were loaded onto 10% SDS-PAGE for electrophoresis and then transferred onto a nitrocellulose membrane. The membrane was subsequently incubated at 4°C overnight with rabbit anti-human STIM1 monoclonal antibody (CST) diluted 1: 1000 and GAPDH diluted 1: 5000. After further washes, the membranes were incubated with the goat anti-rabbit peroxidase–conjugated secondary antibodies (Abcam) at room temperature for 2 hours and the blots were developed using ECL (Millipore).

According to the manufacturer’s instructions, total RNA was extracted from each groups of breast cancer cells using Trizol (Invitrogen ‚ USA) and subsequently reverse transcribed to cDNA using a reverse-transcription PCR system (TaKaRa).The expression levels of miR-223 and STIM1 mRNA were analyzed by using SYBR premix real-time PCR Reagent (TaKaRa). The GAPDH gene was used as an internal control. The primer sequences are shown as follows: miR-223(forward: 5′-GCGTGTATTTGACAAGCTGAGTT-3 and reverse: 5′-GTGTCAGTTTGTCAAATACCCCA-3); STIM1(forward: 5’-TGTGTCTCCCTTGTCCATGC-3’ and reverse, 5’-CATCTGAGGTTTGGGGG-3’).

The 3′-UTR of human STIM1 mRNA was amplified by PCR and the primers sequence were: STIM1-3′UTR-F: 5′-TGC TCT AGA CAG CTT GTC CTT CCC TGG GT-3′; STIM1–3′UTR-R: 5′-ACG TCT GAG GGG CAG CAC CTC TTA GAC A-3′. The amplified product was inserted into the pmirGLO vector (Promega). MD-MA-231 cells and MCF-7 cells were utilized for the luciferase assays, which were grown in 6-well plates and transfected with 4µg of the pmirGLO-STIM1 3′-UTR reporter vector per well and 100 pmol of microRNA mimics or microRNA inhibitor and their control RNA. the Site-specific mutants of the 3′-UTR reporters for STIM1 were also tested.

MTT assay was performed to analyze cell proliferation activity by planting cells (3×10³ cells/100µl) into 96-well plates and being cultured for varying periods of time. Then, the cells were incubated in the medium containing a final concentration of 0.5mg/ml of 3-(4, 5-dimethylthylthiazol-2yl-)-2, 5-diphenyl tetrazolium bromide (MTT, Sigma, Aldrich, MO, USA) solution at 37°C for 4h. The generated formazan was dissolved in 150µl DMSO. Absorbance was measured at 490 nm using a microplate reader.

Cells (2 ×10⁵ cells) were plated on the top side of polycarbonate trans-well filter coated with Matrigel in the upper chamber and incubated at 37°C for 22h. The cells inside the upper chamber with cotton swabs were then removed. Migratory and invasive cells on the lower membrane surface were fixed, stained with hematoxylin, and counted for 10 random 100× fields per well. Cell counts are expressed as the mean number of cells per field of view.

All paraffin sections were prepared well by the pathology department, and immunostained according to the manufacturer’s instructions. Briefly, 5-µm-thick sections were subjected to antigen retrieval, endogenous peroxidase being neutralized and then incubated with primary STIM1 rat monoclonal antibody (ab57834; Abcam, UK) at 4°C overnight. Subsequently, the sections were incubated with secondary antibodies for 30 minutes at 37°C. The chromogen reaction was developed in 3, 3′-diaminobenzidine tetrahydrochloride for 5 minutes and hematoxylin was finally used as a light nuclear counterstain. Two independent observers (ZLN. and GL.) examined the slides in a blinded manner, and consensus was reached by repeated examination when results were discordant. The final scores were calculated by multiplying the scores of the intensity with the percentage in stained tumor cells as described in our previous article [23]. Two groups were divided: high expression (5-12); low expression (0-4).

Similarly, to evaluate miRNA-223 expression by ISH, the paraffin sections were incubated at 60°C for 2 hours, deparaffinized in xylene, rehydrated with graded alcohol washes, washed three times with diethyl pyrocarbonate -treated PBS, digested with 5 µg/ml proteinase K at 37°C for 30 minutes, washed and then dehydrated in graded alcohol. Slides were hybridized at 55°C for 3 hours with 10 nmol/L locked nucleic acid -modified digoxigenin-labeled probes for miRNA-223 (Boster, Wuhan, China). After stringency washes (5x, 1x, 0.2x SSC), slides were placed in blocking solution for 15 minutes followed by incubation in alkaline phosphatase conjugated anti-DIG Fab fragment solution for 1 hours. Finally, hematoxylin was used as a light nuclear counterstain. For each sample, a score was given for the percentage of positive cells as follows: 1 to 9% (0 points), 10 to 24% of the cells (1 point), 25 to 49% of the cells (2 points), 50 to 100% of the cells (3 points). Another score was given for the intensity of staining cells as follows: none (0 point), weak staining (1 points), intermediate staining (2 points), and strong staining (3 points). The staining index (SI) was calculated by multiplying the staining intensity with the proportion of positive cells and divided the samples into four grades: 0, negative (-); 1-2, low staining (+); 3-5, medium staining (++); and 6-9, high staining (+++). The following criteria were used to quantify the expression levels of miR-223: high expression (++/+++); low expression, other than high expression.

Clinicopathological correlations were tested using Fisher’s exact test and the Mann–Whitney U test for categorical and continuous data, respectively. Pearson correlation analysis was performed to evaluate the correlation between variables. Overall survival (OS) time was calculated from the date of the first curative operation to the date (2016-1) of the last follow-up or death from any cause. Survival analyses were performed by the Cox univariate proportional hazards model. For visualization purposes Kaplan–Meier analyses were used for the survival curves test (Mantel-Cox log-rank test). Multivariate Cox proportional hazard regression analysis was adjusted for relevant clinical covariates. All the tests were two-sided and the level of statistical significance was set at P<0.05. Statistical analyses were performed using GraphPad Prism version 5, and SPSS version 23.

First, the basal expression level of miR-223 was probed in five breast cancer cells (T47D, MCF-7, SKB-R3, MDA-MB-231 and MDA-MB-435) and a normal breast epithelial cell (MCF-10A) by using qRT-PCR. The results indicated that the levels of miR-223 were significantly lower in breast cancer cell lines than in the normal breast cells. Meanwhile, it was also different in various breast cancer cells, among which hormone receptor (HR)-negative cells (SKB-R3, MDA-MB-231 and MDA-MB-435) was lower than HR-positive cells (MCF-7 and T47D) (Fig. 1A).

To further confirm the role of miR-223 in the biological behavior of breast cancer, MCF-7 and MDA-MB-231 cells with high and low levels of miR-223 were transiently transfected with anti-miR-223 and miR-223 mimics respectively. MTT assay and trans-well assay were utilized to compare the ability of proliferation and migration between the two groups of cells. As shown in Fig. 1 B and C, MCF-7 cells transfected with anti-miR-223 showed significantly higher proliferation and migration activity than the control group. Differently, the proliferation and migration activity of MDA-MB-231 cells were obviously attenuated after transfection with the miR-223 mimic compared with control group (Fig. 1B and C). The differences in cell proliferation were notably observed at 48 and 72 hours after transfection with anti-miR-223 in MCF-7 cell and miR-223 mimics in MDA-MB-231 cell. Taken together, these data demonstrated a critical role of miR-223 in regulating the proliferation and migration of breast cancer cells.

To investigate whether STIM1 could be regulated directly by miRNAs, we used several web-based target analysis tools (TargetScan, miRDB, microRNA, PicTar and starBase) to identify miRNAs that potentially target STIM1. As a result, miR-223 was identified as a potential regulator of STIM1 expression (Fig. 2A).

To further understand whether STIM1 could be regulated directly by miR-223, the luciferase reporter assay was used to analyze the effect of miR-223 on gene transcription of STIM1. A wild type (WT) or mutant type (MT) 3’-UTR of the STIM1 gene was inserted into the dual luciferase vector pmirGLO (Fig. 2A). Together with the dual luciferase vector, miR-223 mimics or anti-miR-223 were cotransfected into the MDA-MB-231 or MCF-7 cells respectively. The results revealed that the miR-223 mimics significantly reduced the luciferase activity of the wild-type STIM1 but had no effect on the mutant STIM1 reporter gene in MDA-MB-231 cell (Fig. 2B, left). As expected, anti-miR-223 cotransfected with the wild-type reporter gene in MCF-7 cell, the luciferase activity of the reporter gene was elevated, while in the mutant type, the luciferase activity of the reporter gene was equivalent. (Fig. 2B, right). These results revealed that miRNA-223 directly targets STIM1.

Western blot and RT-PCR experiments were further performed to validate the relationship between miR-223 and STIM1. In MDA-MB-231 cells, transfection with the miR-223mimics induced a down-regulation of STIM1 mRNA and protein expression (Fig. 2 C above and Fig. 2 D left). In contrast, anti-miR-223 up-regulated the protein and mRNA level of STIM1 in MCF-7 cells (Fig. 2 C below and Fig. 2 D right). Altogether, these results further confirmed that STIM1 was a direct target gene of miR-223.

To further understand whether miR-223 inhibited proliferation and migration by negatively regulating STIM1, miRNA-223 mimic and pcDNA 3.1-STIM1 or null construct were cotransfected into MDA-MB-231 cell. The western blot results showed that over-expression of STIM1 rescued the miRNA-223-induced decline of STIM1 expression compared with the control group (Fig. 3A). Additionally, MTT and trans-well assay results revealed that the ability of cell proliferation and migration were significantly decreased when MDA-MB-231 cells were transfected with miRNA-223 mimic, however, pcDNA 3.1-STIM1 could partially antagonize this proliferation and migration-inhibiting effect of miRNA-223 mimic (Fig. 3B, C, D). These data suggested that miR-223 restrained cell proliferation and migration by targeting STIM1.

To explore the expression of STIM1 and miR-223 in breast cancer tissues, immunohistochemistry (IHC) and in situ hybridization (ISH) staining were respectively used in 60 breast cancer tissues and the corresponding adjacent normal tissues. Compared with adjacent non-tumor tissues, the increased expression of STIM1 (58.3%, 35/60 vs. 25%, 15/60; P＜0.001) and decreased expression of miR-223 (38.3%, 23/60 vs. 65%, 39/60; P=0.003) were found in cancer tissues (Table 1). In addition, the interaction of STIM1 and miR-223 was further explored in 165 breast cancer tissues with detailed clinicopathological information (Table 2). The results showed that STIM1 positively correlated with large tumor size (P=0.001), lymph node metastasis (P=0.005) and ER (P=0.042), while miR-223 was negatively associated with lymph node metastasis (P=0.005) and ER (P=0.042). Furthermore, Kaplan–Meier and Cox regression analysis revealed that high expressions of STIM1 and miR-223 were respectively detrimental and beneficial factor impacting patient’s DFS rather than OS (Fig. 4C and D). Moreover, Pearson correlation analysis also confirmed that STIM1 inversely correlated with miR-223 (Fig. 4E). Collectively, these findings implied that miR-223 negatively regulated STIM1 in breast cancer.

In the present study, we demonstrate that STIM1 is a direct target of miR-223 in breast cancer. By directly binding to the 3’-UTR of STIM1, miR-223 inhibits STIM1 expression, suppresses proliferation and migration of breast cancer cells and decreases breast cancer recurrence in patients.

STIM1 and Orai are two key molecules that mediate store-operated Ca²⁺ entry (SOCE). SOCE is one of the major routes of calcium entry in nonexcitable cells, in which the depletion of intracellular Ca²⁺ stores inspires activation of the endoplasmic reticulum (ER)-resident Ca²⁺ sensor protein STIM to gate and open the Orai Ca²⁺ channels in the plasma membrane [24]. Accumulating evidence indicates that STIM1 and Orai play crucial roles in promoting tumorigenesis and progression in a variety of cancers [25] in addition to regulating normal physiological functions. In 2009, Yang and his colleagues first validated the association between STIM1 and breast cancer cell migration and metastasis [12]. Recently, it was revealed that STIM1 played an important role in TGF- β -induced proliferation [14] and epithelial-mesenchymal transitions (EMT) [13] in breast cancer. Consequently, it is possible that STIM1 is a promising target for breast cancer therapeutics.

Emerging evidence indicated that miRNAs could function as oncogenes or tumor suppressors by regulating hundreds of downstream target genes. A variety of miRNAs, including miR-152 [19], miR-381 [20], miR-211 [21] and miR-203 [26], have been proved to inhibit tumor proliferation, invasion and metastasis in breast cancer by targeting various genes. In our study, miR-223 was identified as a regulator of STIM1 via qRT-PCR, western blot analysis and luciferase reporter assay. Additionally, HAX-1 [27] and Caprin-1 [22] have also been identified as the target genes of miR-223 in breast cancer.

Usually, miR-223 was considered to be a suppressor in multiple cancers, such as cervical cancer [28], bladder cancer [29], prostate cancer [30] and breast cancer [31]. Our results were consistent with these previous findings. Upregulation of miR-223 could inhibit breast cancer cell proliferation and migration, in contrast, downregulation of miR-223 caused decreased abilities of cell proliferation and migration. Besides STIM1, previous study uncovered two other target genes of miR-223, Caprin-1 [22] and HAX-1 [27], which also contributed to breast cancer proliferation, migration, and invasion. Moreover, we showed that miR-223 expression levels were significantly lower in breast cancer tissues than in the adjacent tissues, and that miR-223 was negatively correlated with some poor prognostic factors. A recent study also confirmed that miR-223 induced by radiotherapy could prevent the relapse of breast cancer [32]. However, other researchers found that overexpression of miR-223 was positively correlated with tumor recurrence and metastasis in patients with colorectal cancer [33], hepatocellular cancer [34] and non-small cell lung cancer [35]. Several in vitro studies also verified that miRNA-223 promoted lung cancer, ovarian cancer and prostate cancer cells invasion via targeting EPB41L3 [36] , SOX11 [37] and SEPT6 [38] respectively. These inconsistent results may be due to the different roles of the target genes of miRNA-223 in the tumor.

In addition to regulating tumor growth, miR-223 was also found to be an early diagnostic marker in oral cancer [39, 40], gastric cancer [41] and colorectal cancer [42] detection. Therefore, miR-223 perhaps will be a potential marker for the early diagnosis of breast cancer.

We confirmed that miR-223 could inhibit the biological behavior of breast cancer by suppressing STIM1 expression. We speculate that the miR-223/STIM1 axis is an attractive target for diagnosis and therapy breast cancer.

Funding/Support: This work was supported by a grant from Tianjin Application Foundation and Advanced Technology Research Program (grant numbers:14JCYBJC24900).

Competing interests: All authors declare that they have no conflicts of interest.

Y. Yang and Z. Jiang contributed equally to this work.