Bioinformatic analysis of SNHG1 expression was performed in 1,063 breast cancer cases and 102 normal breast cases using the Human Cancer Metastasis Database (HCMDB, http://hcmdb.i-sanger.com/). The Cancer Genome Atlas Breast Invasive Carcinoma (TCGA-BRCA) dataset was selected. The prediction of the potential binding site between miR-573 and SNHG1 and LMO4 was carried out by miRDB (http://www.mirdb.org/) and miRanda software (http://www.microrna.org). The PROGgeneV2 (http://genomics.jefferson.edu/proggene/index.php) was used to study the association between LMO4 expression and the overall survival of patients with breast cancer based on the GSE42568 dataset (28).

Human breast cancer tumor tissues and matched normal breast tissues were collected from 50 patients with breast cancer at The Second Xiangya Hospital of Central South University from June 2014 to July 2017. All tissues were obtained following surgery of primary breast cancer tumors and were immediately frozen in liquid nitrogen for subsequent experiments. Prior to project initiation, written informed consent was provided by all patients enrolled in the present study and the experimental procedures were conducted under the supervision of the Ethics Committee of the Second Xiangya Hospital of the Central South University. The protocol of the experiments was approved by the Ethics Committee of the Second Xiangya Hospital of the Central South University (approval no. 2014S057).

293 cells, the human breast epithelial cell line MCF10A, the human ER⁺ breast cancer cell lines MCF7, and T47D, and the human triple-negative breast cancer (TNBC) cell lines (ER⁻/PR⁻/Her2⁻) MDA-MB-231 and MDA-MB-468 were purchased from the American Type Culture Collection (ATCC). The cell lines were used within 6 months following receipt. MCF10A cells were cultured in Mammary Epithelial Cell Growth Medium (MEGM; Lonza) supplemented with 100 ng/ml cholera toxin (Sigma-Aldrich; Merck KGaA). 293, MCF7 and T47D cells were cultured in RPMI-1640 medium (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% FBS (HyClone; GE Healthcare). MDA-MB-231 and MDA-MB-468 cells were maintained in DMEM (Gibco; Thermo Fisher Scientific, Inc.) containing 10% FBS (HyClone; GE Healthcare). All cell lines were cultured in a humidified incubator with 5% CO₂.

The full length of the LMO4 open reading frame was amplified from the cDNA of 293 cells and ligated into a pcDNA3.1 plasmid. Plasmid transfection was performed using Lipofectamine 3000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer protocol. SNHG1 siRNA and control siRNA were purchased from GenePharma. The sequence for SNHG1 siRNA was CAGCAGTTGAGGGTTTGCTGTGTAT. The transfection of SNHG1 siRNA or control siRNA sequences was achieved using LipoRNAiMax reagent (Invitrogen; Thermo Fisher Scientific, Inc.) in serum-free medium and sustained for 5 min until addition into the culture medium. miR-NC mimic, miR-573 mimic, miR-NC inhibitor and miR-573 inhibitor were purchased from Applied Biological Materials (ABM). The transfection of the mimic or inhibitor sequences was carried out using Lipofectamine 3000.

For stable knockdown of SNHG1, lentiviral particles were prepared by co-transfection of pLko.1-SNHG1 shRNA, pMD2G and pCMV-dR8.91 into 293 cells using Lipofectamine 2000. Following 72 h of culture, the medium containing lentiviral particles was obtained and filtered through a 0.45-µm filter (Millipore). The medium containing lentiviral particles was added to the MDA-MB-231 cells in 6-well plates. Following 48 h of culture, the culture medium was replaced with fresh medium containing 5 mg/ml puromycin (Solarbio) for 24 h to select the cells successfully infected with the lentivirus. For establishment of SNHG1-knockdown and LMO4-overexpression models, MDA-MB-231 cells and shSNHG1 MDA-MB-231 cells were transfected with the pcDNA3.1-LMO4 plasmid using Lipofectamine 3000. Following 24 h of cell culture, the cells were screened with 4 mg/ml G418 (Sigma-Aldrich) for the following 72 h of growth.

Total RNA from tissues and cells was prepared using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.). The separation of nuclear and cytoplasmic RNA was accomplished using a PARIS kit (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. Gene expression was detected following reverse transcription of RNA into cDNA using PrimeScript™ RT Reagent Kit (Takara Bio, Inc.). miR-573 expression was performed using a stem-loop specific primer method. Real-time RT-PCR was conducted using the SYBR Premix Ex Taq (Takara Bio, Inc.). The expression levels of genes or miRNAs was analyzed using the 2^−ΔΔCq method (29). GAPDH and U6 were used as internal controls for gene and miRNA expression analysis, respectively. The primer sequences are listed in Table I.

Protein lysates were prepared using RIPA lysis buffer (Sigma-Aldrich; Merck KGaA). The antibodies for LMO4 (cat. no. ab131030; dilution 1:2,000) and GAPDH (cat. co. AMAB91153; dilution 1:10,000) detection were purchased from Abcam and Sigma-Aldrich/Merck KGaA, respectively. Cyclin D1 (cat. no. 2978; dilution 1:2,000) and cyclin E1 (cat. no. 4129; dilution 1:2,000) antibodies were obtained from Cell Signaling Technology. Secondary antibodies for mouse (cat. no. SA00001-1; dilution 1:10,000) and rabbit (cat. no. SA00001-2; dilution 1:10,000) were obtained from Proteintech. Briefly, 25 µg proteins per lane were separated on an 8% SDS gel and transferred to a PVDF membrane. Following transfer, the membrane was blocked in 5% non-fat milk and incubated in the presence of the primary antibodies overnight at 4°C. On the next day, the membrane was incubated with secondary antibodies for an additional 1 h at room temperature. The protein bands were developed using ECL substrate (Thermo Fisher Scientific, Inc.). The intensity of the bands was quantified using ImageJ (V. 1.6.0; National Institutes of Health).

The 3′ untranslated region (3′UTR) of the pGL3 construct containing wild-type LMO4 3′UTR (LMO4 3′UTR-WT) and the pGL3 construct containing wild-type SNHG1 (SNHG1 3′UTR-WT) were prepared by PCR of the cDNA derived from 293 cells. Two site mutations were introduced into the putative seed regions of LMO4 3′UTR-WT and SNHG1 3′UTR-WT in order to produce mutant constructs. The dual luciferase assay was performed using cells that were co-transfected with either wild-type or mutant-type luciferase reporter plasmids, pRL-TK plasmid, miR-NC mimic or miR-573 mimic sequences. The transfections were performed using Lipofectamine 3000. The relative luciferase activity was measured at 24 h following transfection using a Dual-Luciferase reporter assay system (Promega Corp.) according to the manufacturer's protocol.

The cell proliferative ability was measured using a Cell Counting Kit-8 (CCK-8; Dojindo Laboratories). On the first day, the cells were seeded in each well of a 96-well plate. On the next day, the cells were transfected with the indicated siRNA, miRNA mimic or miRNA inhibitor sequences. The cell viability was subsequently examined at 24-h time points between days 1 to 4, by addition of 10 µl CCK-8 solution into the culture medium. The solution was incubated for 2 h and the medium containing CCK-8 was aspirated from the wells and added to another 96-well plate. The absorbance at 450 nm of each well was measured to estimate the cell number.

Transfected cells were harvested and fixed with ice-cold 70% ethanol at 4°C overnight. The cells were stained with propidium iodide (PI; Sigma-Aldrich; Merck KGaA) for 30 min and analyzed using flow cytometry. The percentage of the cells present in each cell cycle phase was counted.

The wound healing assay was applied to detect cell migratory activity. The cells were grown in 6-well plates at 90% confluence and subsequently transfected with siRNA for 24 h. The cell layer was scratched with a 10-µl pipette tip at the central area and subsequently washed with PBS. Serum-free DMEM was added. The images of the wound area were captured at the 0 and 24 h time points following the initial scratch. The closure areas of all wells were quantified using Image-Pro Plus (V. 6.0; Media Cybernetics).

Female, 5-week-old nude mice (BALB/c-null) were purchased from the Shanghai Laboratory Animal Center (Chinese Academy of Sciences, China) and bred under SPF conditions. The present study was approved by the Ethics Committee of the Second Xiangya Hospital of the Central South University. The mice were randomly divided into the three following groups (n=3): MDA-MB-231, shSNHG1-MDA-MB-231 and shSNHG1+LMO4 OE MDA-MB-231. The cells from each group were subcutaneously injected into the mammary armpit of the mouse fat pad. The tumor size was measured every five days with a caliper. The tumor volume was estimated according to the following formula: Volume = 0.5×LengthxWidth². The mice were sacrificed by decapitation at 35 days following cell injection and the tumors were dissected.

A total of 10 formalin-fixed, paraffin-embedded biopsy breast tumors were available. Immunohistochemistry was performed as described previously (30). The samples were stained with the LMO4 antibody (dilution, 1:100) used in the western blotting experiments. Diaminobenzidine (DAB; Boster Inc.) was used for color development. The specimens were observed via a Olympus BX51 light microscope (Olympus Corp.) at ×100 magnification.

The data presented in the present study were calculated using GraphPad Prism 7 software (GraphPad Software, Inc.) and are represented as mean ± SD. The comparison between the two groups was achieved using the paired Student's t-test. The differences among the three groups were analyzed using one-way ANOVA followed by the Student-Newman-Keul test. The comparison for more than 3 groups was performed using one-way ANOVA followed by the Tukey's test. The values were considered significantly different at P<0.05.

The expression levels of SNHG1 in breast cancer have not been previously studied. To explore the expression profile of SNHG1 in breast cancer, its expression levels were assessed in 102 normal breast and 1,063 primary breast cancer tissues derived from the TCGA-BRCA dataset. A slight yet significant elevation in SNHG1 levels was observed in breast cancer tissues compared with that noted in normal tissues (Fig. 1A). Subsequently, RT-PCR was applied to detect SNHG1 expression in 50 tissue pairs derived from breast cancer and normal breast tissues. The results demonstrated that SNHG1 expression was significantly higher in breast cancer tissues compared with that noted in normal breast tissues (Fig. 1B). In addition, the increase in the levels of SNHG1 was associated with advanced pathological stage, as well as ER⁻ and PR⁻ status (Table II). However, SNHG1 expression levels were not associated with factors, such as age, lymph node metastasis and Her2 status (Table II). Furthermore, RT-PCR indicated that SNHG1 levels were increased in TNBC MDA-MB-231 and MDA-MB-468 cell lines compared with those noted in the normal epithelial breast cell line MCF10A (Fig. 1C). SNHG1 expression levels were not increased significantly in the ER⁺ breast cancer cell lines MCF7 and T47D compared with those noted in the MCF10A cell line (Fig. 1C). These results indicated that high expression levels of SNHG1 were associated with breast cancer incidence, notably with ER⁻/PR⁻ breast cancer. Therefore, MDA-MB-231 and MDA-MB-468 cells were used for the following experiments.

The increase in the levels of SNHG1 in breast cancer cells suggest a potential oncogenic role of this lncRNA. To evaluate the function of SNHG1 in breast cancer cells, siRNA-mediated knockdown of SNHG1 was performed to assess cell proliferation of MDA-MB-231 and MDA-MB-468 cells. Transfection of the cells with SNHG1 siRNA significantly decreased SNHG1 expression in both MDA-MB-231 and MDA-MB-468 cells compared with control siRNA transfection (Fig. 2A). In MDA-MB-231 cells, knockdown of SNHG1 significantly inhibited cell growth as determined by CCK-8 cell viability assay (Fig. 2B). Although significant activation of cell apoptosis was not observed following SNHG1 downregulation (data not shown), flow cytometric analysis of cell cycle distribution demonstrated that SNHG1 knockdown resulted in cell cycle redistribution with an accumulation of cells in the S and G2/M phases (Fig. 2C and D). In the MDA-MB-468 cell line, knockdown of SNHG1 also induced cell growth arrest (Fig. 2E). Specifically, the number of MDA-MB-468 cells that accumulated in the G2/M phase was significantly increased (Fig. 2F and G). These data suggest that SNHG1 may promote breast cancer cell proliferation via the G2/M cell cycle checkpoint.

Tumor metastasis is considered a major mortality cause in breast cancer patients (31). To investigate whether SNHG1 regulates breast cancer cell migration, wound healing assays were performed to detect cell migration following SNHG1 knockdown. Knockdown of SNHG1 significantly decreased the wound closure area of the MDA-MB-231 cells (Fig. 3A and B), suggesting that SNHG1 regulates the cell migratory activity of breast cancer cells. Similarly, SNHG1 knockdown reduced the cell migratory activity of the MDA-MB-468 cells (Fig. 3C and D).

The molecular mechanism of SNHG1 was examined in breast cancer by detecting its nuclear and cytoplasmic localization. Following isolation of the nuclear and cytoplasmic RNA of MDA-MB-231 and MDA-MB-468 cells, RT-PCR analysis demonstrated that nearly half of SNHG1 was present in the cytoplasm (Fig. 4A), suggesting that it may act as an miRNA and ceRNA, as previously reported (32). Bioinformatic analysis using the miRDB software suggested a potential binding site between SNHG1 and miR-573 (Fig. 4B). Based on the prediction, it was speculated that SNHG1 could negatively regulate miR-573 in breast cancer cells. Knockdown of SNHG1 significantly increased miR-573 levels in both MDA-MB-231 and MDA-MB-468 cell lines (Fig. 4C). Moreover, overexpression of miR-573 by transfection of miR-573 mimic significantly decreased SNHG1 expression in both cell lines tested (Fig. 4D). To further validate the direct binding of SNHG1 with miR-573, luciferase plasmids containing wild-type WT and mutant Mut SNHG1 (with two site mutations in the complementary sequence) were constructed. Using dual luciferase reporter assays, co-transfection of miR-573 mimic and WT SNHG1 exhibited significantly decreased luciferase activity in the MDA-MB-231 cells (Fig. 4E). Similar results were also observed in the MDA-MB-468 cells (Fig. 4F). Therefore, SNHG1 functions as a negative regulator of miR-573 in breast cancer cells.

miRNAs regulate gene expression by binding to the 3′UTR of target gene mRNA sequences leading to their degradation or the inhibition of their translation (33). The miRanda software was used to demonstrate that the seed region of miR-573 matched the 3′UTR of LMO4 mRNA (Fig. 5A). LMO4 is an oncogene and is overexpressed in breast cancer (34). Immunohistochemical analysis was used to detect LMO4 expression in 10 breast cancer tumors. Positive expression of LMO4 was observed in the majority of breast cancer tumors (8/10) (Fig. 5B). The prognostic value of LMO4 was analyzed in breast cancer. Based on the GSE42568 (n=104) dataset, high expression levels of LMO4 were associated with reduced overall survival time of the patients with breast cancer (Fig. 5C), suggesting that LMO4 may promote breast cancer progression. In MDA-MB-231 and MDA-MB-468 cells, the transfection of the miR-573 inhibitor significantly increased LMO4 mRNA expression, while the miR-573 mimic decreased LMO4 mRNA levels (Fig. 5D). Western blot analysis further revealed that miR-573 inhibition significantly increased LMO4 protein expression, whereas transfection of the cells with miR-573 mimic significantly reduced LMO4 protein levels (Fig. 5E and F). The direct binding of miR-573 to the LMO4 3′UTR sequence was confirmed in a dual luciferase reporter assay. The overexpression of miR-573 decreased the relative luciferase activity in MDA-MB-231 cells transfected with the LMO4 3′UTR WT sequence, while the luciferase activity of the cells transfected with LMO4 3′UTR Mut sequences was not altered compared with that of the miR-573 mimic (Fig. 5G). Similarly, miR-573 overexpression reduced luciferase activity of MDA-MB-468 cells transfected with the LMO4 3′UTR WT sequence (Fig. 5H). These results indicated that miR-573 directly repressed LMO4 expression in breast cancer cells.

The aforementioned results demonstrated that SNHG1 could sponge miR-573, which in turn suppressed LMO4 expression, suggesting that SNHG1 may control LMO4 expression in breast cancer cells. As expected, SNHG1 knockdown led to a decrease in LMO4 mRNA levels in MDA-MB-231 and MDA-MB-468 cells (Fig. 6A). LMO4 is a transcription factor that activates cyclin D1 and E1 transcription and mediates cell cycle progression (35). Knockdown of SNHG1 reduced cyclin D1 and cyclin E1 mRNA levels (Fig. 6A). Moreover, the decrease in the levels of cyclin D1 and cyclin E1 was reversed by transfection of recombinant LMO4 (Fig. 6A). Western blot analysis further indicated that SNHG1 knockdown reduced LMO4, cyclin D1 and cyclin E1 protein levels. These changes were reversed by pcDNA3.1-LMO4 transfection in both cell lines (Fig. 6B and C). The RT-PCR and western blot results revealed the effects of the SNHG1/miR-573/LMO4 axis in breast cancer cell lines.

The regulation of breast cancer cell proliferation by SNHG1 was assessed by cell proliferation of MDA-MB-231 cells transfected with SNHG1 siRNA or SNHG1 siRNA in the presence of pcDNA3.1-LMO4. The increase in the levels of LMO4 partially reversed SNHG1 knockdown-mediated cell growth arrest in MDA-MB-231 cells (Fig. 7A). In addition, overexpression of LMO4 reversed the accumulation of cells at the G2/M phase following SNHG1 knockdown (Fig. 7B and C). Similarly, LMO4 overexpression reversed cell proliferation inhibition and cell cycle redistribution induction by SNHG1 knockdown in MDA-MB-468 cells (Fig. 7D-F). However, overexpression of SNHG1 did not reverse inhibition of cell migration induced by SNHG1 knockdown (data not shown), suggesting that SNHG1 could control cell migration by other mechanisms. These data indicated that LMO4 plays a major role in the SNHG1-mediated cell growth and cell cycle regulation of breast cancer cells.

In addition to the in vitro results, the regulation of breast cancer cell progression by SNHG1 was examined in vivo. MDA-MB-231 cells were infected with a lentiviral vector carrying the SNHG1 or control shRNA sequences to construct stable SNHG1-knockdown and control MDA-MB-231 cell lines. SNHG1 shRNA MDA-MB-231 cells were transfected with pcDNA3.1-LMO4 plasmid and screened with G418 to build LMO4-overexpressing and SNHG1-knockdown MDA-MB-231 cell lines. RT-qPCR indicated that SNHG1 expression was significantly downregulated in the SNHG1 shRNA MDA-MB-231 and SNHG1 shRNA+LMO4 OE MDA-MB-231 cells (Fig. 8A). Western blot analysis confirmed that LMO4 protein expression was decreased in SNHG1 shRNA MDA-MB-231 cells, whereas it was slightly elevated in SNHG1 shRNA+LMO4 OE MDA-MB-231 cells (Fig. 8B and C). Tumor growth was very slow in mice injected with SNHG1 shRNA MDA-MB-231 cells compared with mice in the control and SNHG1 shRNA+LMO4 OE groups (Fig. 8D). Following 35 days of in vivo tumor cell growth, the mice were sacrificed and the tumors were dissected. SNHG1 knockdown significantly reduced the tumor size, whereas overexpression of LMO4 reversed the inhibitory effects on breast cancer cells (Fig. 8E and F). Moreover, the expression levels of SNHG1 and LMO4 were examined in 50 breast cancer tumor tissues. The results indicated a positive correlation between SNHG1 expression and LMO4 mRNA expression in tumor tissues (Fig. 8G).