The long non‐coding RNA SNHG1 promotes bladder cancer progression by interacting with miR‐143‐3p and EZH2

Abstract The long non‐coding RNA (lncRNA) SNHG1 has been shown to be implicated in the progression of multiple human carcinomas. Nevertheless, the biological functions and potential mechanism of SNHG1 in bladder cancer (BC) are uncharacterized. In the present study, SNHG1 was found to be substantially up‐regulated in BC tissues and cells and was intimately correlated with the TNM stage, lymphatic invasion, metastasis and recurrence‐free survival in BC patients. Down‐regulation of SNHG1 dramatically attenuated the proliferation, migration and invasion of BC cells, whereas the ectopic overexpression of SNHG1 had the opposite effects in vitro. The in vivo experimental results also indicated that SNHG1 down‐regulation hampered the tumour growth and metastasis of BC cells. Mechanistic investigations revealed that SNHG1 enhances HK2 expression by serving as an endogenous sponge to regulate miR‐143‐3p in the cytoplasm of BC cells. In the nucleus, SNHG1 could interact with EZH2 and regulate the histone methylation of the CDH1 promoter, altering the biological behaviours of BC cells. Overall, these findings elucidate an oncologic role of SNHG1 in BC and provide a new therapeutic strategy against BC.


| INTRODUC TI ON
Bladder cancer (BC), one of the most lethal neoplasms of the genitourinary system worldwide, has shown a trend towards increasing incidence and mortality rates in recent years. 1 Despite substantial progress in understanding the pathophysiology of BC, the exact mechanisms underlying bladder carcinogenesis remain unclear. Therefore, identifying novel molecular targets and more effective therapeutic approaches is of crucial importance to improve the treatment and prognosis of BC.
Currently, emerging evidence has indicated that non-coding RNAs (ncRNAs) are involved in the pathogenesis of human carcinomas, providing novel insights into mechanistic research on the occurrence and development of malignant neoplasms. 2 miRNAs are a group of small ncRNAs with an average length of 20-24 nucleotides that mediate the translational suppression or degradation of mRNAs by binding to their 3′-untranslated regions (3′-UTRs). 3 Additionally, lncRNAs are a subgroup of ncRNAs that have lengths greater than 200 bp. It has been generally recognized that lncRNAs have key roles in the numerous processes, including epigenetic modification, transcription, post-transcriptional modification and translation. 4 Notably, increasing emphasis has been placed on mechanistic research of lncRNAs that affect the tumorigenesis and development of numerous human carcinomas, including BC. 1,5,6 For example, the enhanced expression of ln-cRNA MALAT1 was characterized in BC and shown to promote metastasis by interacting with SUZ12. 7 UCA1 was also observed to be up-regulated in BC and has been identified as a sensitive and specific marker for BC. 8 In addition, the aberrant expression of lncRNA H19 was demonstrated to be a risk marker for the development of non-muscle-invasive BC. 9 Other important ln-cRNAs such as BLACAT2, GAS5 and TUG1 have also been shown to be involved in the tumorigenesis and progression of BC. [10][11][12] These findings demonstrated that lncRNAs are valuable diagnostic and prognostic biomarkers in BC. Nevertheless, further investigations are needed to elucidate the biological functions and underlying mechanisms by which lncRNAs influence the progression of BC.
Small nucleolar RNA host gene 1 (SNHG1) is an lncRNA located on chromosome 11q12.3. Recent reports have suggested that SNHG1 is aberrantly expressed and contributes to tumorigenesis in several types of human malignances, such as glioma, 13 lung carcinoma, 14 cholangiocarcinoma, 15 breast cancer, 16 hepatocarcinoma 17 and prostate cancer. 18 Knockdown of SNHG1 in cancer cells induces their proliferation or growth and contributes to the diminishment of migration and invasion, indicating that SNHG1 has crucial roles in the maintenance of the above malignances. However, to our knowledge, whether SNHG1 has a strong impact on BC cell function remains unelucidated. In the present study, after demonstrating that SNHG1 up-regulation is a characteristic feature of human BC, the biological role and downstream molecular mechanism of SNHG1 in BC progression were investigated. SNHG1 was ultimately identified as a competing endogenous RNA (ceRNA) that modulates the expression of HK2, which was further confirmed to be targeted by miR-143-3p in BC cells. Additionally, SNHG1 has been reported to bind histone methyltransferase EZH2, an essential enzymatic subunit of polycomb repressor complex 2 (PRC2), which catalyses the trimethylation of histone H3 lysine 27 (H3K27me3) in the promoters of target genes to inhibit target genes expression. 19 The CDH1 gene encodes E-cadherin, a well-studied protein involved in maintaining the epithelial phenotype and modulating tissue homoeostasis by regulating various signalling pathways. 20 E-Cadherin has been characterized as a potent suppressor gene in numerous human carcinomas, including breast cancer, 21 colorectal cancer, 22 hepatocarcinoma 23 and BC. 24 In the present study, SNHG1 was shown to recruit EZH2 protein to the promoter region of the CDH1 gene and epigenetically suppressed the expression of E-cadherin in BC cells, modulating the biological behaviour of BC cells. Taken together, our results elucidated the mechanism associated with SNHG1 function in BC, indicating that SNHG1 can serve as an oncogenic marker in BC.

| Hierarchical clustering analysis and tissue sample collection
The Cancer Genome Atlas (TCGA) data were used to investigate the differences in SNHG1 expression between BC tissues (n = 414) and non-tumour tissues (n = 19) using hierarchical cluster analysis.
Sixty BC and paired non-tumour tissue specimens were acquired from patients who received radical resection of tumour and were pathologically diagnosed as BC at Wuhan No.

| Cell culture
The BC-derived cell lines (T24 and 5637) and the immortalized normal human uroepithelial cell line (SV-HUC-1) were acquired from the American Type Culture Collection (ATCC

| Subcellular fractionation assay
To isolate the cytosolic and nuclear fractions, the PARIS Kit (Invitrogen) was utilized. Then, the assay was performed based on the manufacturer's recommendation.

| Quantitative reverse transcription PCR (RT-qPCR)
Total RNA in BC samples and cell lines was harvested with TRIzol reagent (Invitrogen). For RT-qPCR assays, 2 μg total RNA treated with DNase I was subjected to cDNA synthesis, with a PrimeScript RTpolymerase kit (Takara). A StepOnePlus TM RT-qPCR System (Applied Biosystems) and SYBR Green (Takara) were used for RT-qPCR analysis. The mRNA level of each gene was normalized to that of GAPDH.
The quantity of miRNA was normalized to U6 expression. The specific primers for GAPDH, SNHG1, CDH1, HK2 and other genes were obtained from Sangon (Shanghai, China), while those for miRNAs and U6 were acquired from RiboBio (Guangzhou, China). The relative expression of each gene was assessed using the 2 −ΔΔCT method.
The primers are shown in Table S1.

| Fluorescence in situ hybridization (FISH)
Fluorescence in situ hybridization (FISH) experiments were conducted to determine the subcellular location of SNHG1 in BC cells.
SNHG1 antisense RNA probes labelled with fluorescein were obtained from GenePharma (Shanghai, China), and the sense RNA probe for SNHG1 was used as a control. The assays were conducted in accordance with the protocol.  and at least three random fields were imaged.

| Luciferase reporter assay
For luciferase analysis, the sequences of wild-type or mutant SNHG1 were amplified and cloned into the luciferase vector pMIR-Report. For miR-143-3p and HK2, the sequences of the luciferase vector 3′-UTRs of wild-type or mutant HK2 were generated and inserted into the pMIR-Report. Subsequently, these plasmids and synthetic oligonucleotides were cotransfected into BC cells, and the cells were harvested 48h later. Relative luciferase activity was assessed with a Dual-Luciferase Reporter Assay Detection kit (Promega).

| Western blotting, haematoxylin-eosin (HE) staining and immunohistochemistry (IHC)
Proteins from BC tissues and cells were isolated using RIPA buffer

| RNA immunoprecipitation (RIP) experiments
RNA immunoprecipitation (RIP) experiments were conducted with an EZ-Magna RIP™ RNA-Binding Protein Immunoprecipitation kit (Millipore). In brief, the BC cells were harvested and lysed in RIP lysis buffer according to the manufacturer's recommendations.
Subsequently, pre-incubated magnetic beads coated with indicated antibodies were immunoprecipitated with the supernatant of cell lysate for 6 hours at 4°C. Subsequently, the purified RNA was checked by RT-qPCR.
Premier Primer 5.0 was used to design the CDH1 promoter primers that bind adjacent to the transcriptional start site, and the ChIP primers were obtained from RiboBio. The primers used for the ChIP assay are available in Table S1. A Magnetic RNA-Protein Pull-Down kit (Pierce, USA) was utilized to conduct RNA pull-down experiments following the manufacturer's protocol. The biotinylated RNA used to detect SNHG1 was obtained from RiboBio (Guangzhou).

| Animal experiments
To investigate the effects of SNHG1 silencing on BC growth in vivo, EJ cells (6 × 10 6 ) stably transfected with the sh-NC or sh-SNHG1 vectors were implanted into BALB/c nude mice (n = 5 for each group). Five weeks later, the mice were killed, and tumour volume and weight were measured. IHC assays were performed to examine protein levels in xenograft tumour specimens. In another assay, the stably transfected cells (3 × 10 6 ) were intravenously injected into the nude mice (5 weeks old, n = 5 per group). After 7 weeks, the mice were killed, and their lungs were dissected for further examination. All animal assays were approved by the Animal Care Committee of Tongji Medical College.

| Statistical analysis
All statistical analyses were performed using GraphPad Prism 6.0 (La Jolla, USA). The data are expressed as the means ± SD. Student's t test, Chi-square test, one-way ANOVA and Spearman correlation analysis were applied to estimate the statistical differences among various groups, where P < .05 was considered significant.

| SNHG1 expression is up-regulated in BC and is associated with disease progression
An analysis of TCGA data (containing 19 normal bladder specimens and 414 BC specimens) revealed that the abundance of SNHG1 in the BC samples was higher than that observed in the non-tumour bladder tissues ( Figure 1A,B). Consistent expression was observed for the 60 paired BC specimens and non-tumorous bladder tissues via RT-qPCR. As shown in Figure  In the patient samples, SNHG1 expression was classified into lower SNHG1 and higher expression groups using the 25th percentile and 75th percentile expression levels as cut-off points. A higher recurrence rate was observed in patients with high SNHG1 abundance compared with low SNHG1 expression ( Figure 1H). Interestingly, the results of the clinicopathological analysis indicated that increased SNHG1 expression was closely correlated with the tumour size, tumour stage, invasion and metastasis (P < .05), whereas no significant correlation was observed between SNHG1 and other clinicopathological features, including sex, age and histologic grade (Table 1). We also investigated the expression levels of SNHG1 on the basis of clinical features using the TCGA databases. SNHG1 expression levels were increased in BC tissues compared with that observed in normal bladder tissues in the subgroup analyses, including tumour stage and nodal metastasis status ( Figure S1A,B). However, no significant difference in the expression levels of SNHG1 was observed among different molecular subtypes (neuronal, basal, luminal, luminal infiltrated and luminal papillary) ( Figure S1C). Taken together, these findings revealed that SNHG1 is most likely implicated in tumorigenesis and development of human BC.

| SNHG1 promotes the proliferation, migration and invasion of BC cells
As SNHG1 expression was observed to be substantially enhanced in BC, additional analyses were performed to determine whether it has an oncogenic effect in BC. First, we performed in vitro experiments to investigate the potential function of SNHG1. Among the four types of BC cell lines, a relatively high and low expression levels of SNHG1 were observed in EJ and T24 cells, respectively. Based on these results, the impact of SNHG1 knockdown or SNHG1 restoration in these cell lines was investigated. As shown in Figure 2A,  [25][26][27][28][29] We therefore predicted that these proteins may be useful markers to monitor biological behaviours of BC cells in SNHG1-mediated modulation. Collectively, these results revealed that SNHG1 is capable of promoting BC progression by facilitating BC cell proliferation, migration, invasion and EMT.

| SNHG1 sponges miR-143-3p in the cytoplasm of BC cells
Emerging evidence has confirmed that some lncRNAs can sponge miRNAs and regulate the expression and activity of targeted miR-NAs. 30 To determine whether SNHG1 has a similar regulatory mechanism in BC, LncBase v.2 was used to predict the potential miRNAs that may bind to SNHG1 (sheet 1 in Table S2). In addition, we searched for the down-regulated miRNAs that have been identified as tumour suppressors in BC using an online server (http://mirca ncer.ecu.edu/) (sheet 2 in Table S2). Finally, 10 candidate miRNAs were selected by analysing the overlapping prediction results of miRNAs (sheet 3 in Table S2). As indicated in Figure 4A,B, miR-143-3p was the only miRNA that was remarkably down-regulated by SNHG1 overexpression in both EJ and T24 cells. Subsequently, luciferase reporter cell lines were established to determine whether miR-143-3p was a functional target of SNHG1. The data indicated that the luciferase activity was obviously impaired in cells cotransfected with miR-143-3p mimics and wild-type SNHG1 reporter vectors (SNHG1-Wt) than in cells cotransfected with negative control (NC) mimics and SNHG1-Wt in BC cells. Furthermore, these effects were blocked by mutating the putative miR-143-3p binding sites, suggesting that miR-143-3p was targeted by SNHG1 ( Figure 4C-E). Moreover, the abundance of miR-143-3p was observed to be markedly decreased F I G U R E 2 SNHG1 silencing suppresses the proliferation, retards cell cycle progression and inhibits the migration and invasion of BC cells in vitro. A, RT-qPCR assay of SNHG1 levels in EJ and T24 cells following treatment with siRNAs against SNHG1 (si-SNHG1 #1 and si-SNHG1 #2) or an SNHG1-overexpressing plasmid (pcDNA3.1/SNHG1). B, C, The proliferation of EJ and T24 cells transfected with si-SNHG1 or pcDNA3.1/SNHG1 vectors were assessed via CCK-8 analysis. D, The proliferation of EJ and T24 cells transfected with si-SNHG1 or pcDNA3.1/SNHG1 were assayed by EdU assays. Scale bar, 50 μm. E, F, The effect of si-SNHG1 or pcDNA3.1/SNHG1 on the cell-cycle distribution of EJ and T24 cells was assessed by flow cytometry analysis. G, Representative images of the migratory and invasive BC cells transfected with si-SNHG1 or pcDNA3.1/SNHG1 vectors are presented. Scale bar, 50 μm. H, EMT-like morphological changes in EJ and T24 cells transfected with si-SNHG1 or pcDNA3.1/SNHG1. Scale bar, 20 µm. I, The protein expression of cellular proliferation, migration, invasion and EMT-related genes (p21, PCNA, cyclin D1, vimentin, E-cadherin, and MMP-9) was detected in si-SNHG1-or pcDNA3.1/SNHG1transfected BC cells via Western blotting analysis. GAPDH served as an internal control. * P < .05 and ** P < .01 in EJ, T24, 5637 and BIU-87 cells compared with that observed in SV-HUC-1 cells ( Figure 4F). As expected, the decreased expression of miR-143-3p was confirmed in BC tissues compared with that observed in non-tumorous tissues ( Figure 4G). We also investigated the potential biological function of miR-143-3p in BC cells. CCK-8 and EdU assay results indicated that miR-143-3p overexpression silencing promoted cell transition from G0/G1 to S phase in T24 cells ( Figure S2D). These data revealed that SNHG1 targets miR-143-3p, which plays a tumour suppressive role in the modulation of proliferation in BC cells. Additionally, to further investigate whether SNHG1 directly interacts with miR-143-3p, RIP assays were conducted on EJ cell extracts. As shown in Figure 4H, Ago2 protein was obviously F I G U R E 5 HK2 is targeted by SNHG1-regulated miR-143-3p. A, B, RT-qPCR analysis of mRNA expression levels of predicted miR-143-3p targets in miR-143-3p-overexpressing EJ and T24 cells. C-F, The levels of HK2 mRNA and protein were examined by RT-qPCR and Western blotting analysis in cotransfected BC cells, respectively. G, The putative binding sites of miR-143-3p in the wild-type/mutant HK2 mRNA 3′-UTR sequence. H-I, The luciferase activity assay results indicated that miR-143-3p overexpression or SNHG1 knockdown could remarkably inhibit the luciferase activity of the wild-type HK2 3′-UTR, but not the mutant HK2 3′-UTR. And the decreased luciferase activity caused by SNHG1 silencing could be restored by miR-143-3p suppression in EJ and T24 cells. J, Western blotting analysis of HK2 protein expression in EJ cells following cotransfection with HK2 siRNAs and NC inhibitors or cotransfection with HK2 siRNAs and miR-143-3p inhibitors. K, A positive association was shown between the abundance of SNHG1 and HK2 mRNA (r = 0.6308, P < .0001). L, The cellular proliferation of EJ cells cotransfected with the SNHG1 overexpression vector and si-Control or cotransfected with SNHG1 overexpression vector and HK2 siRNAs was assessed by EdU assays. M, IHC assessment of HK2 protein expression in xenograft tumour tissues. Scale bar, 50 μm. N, IHC assessment of HK2 protein in BC and paired non-tumour bladder tissues. * P < .05 and ** P < .01. Scale bar, 50 μm immunoprecipitated from the extracts of EJ cells using an anti-Ago2 antibody. Furthermore, RT-qPCR analysis of SNHG1 and miR-143-3p abundances in the immunoprecipitates showed a higher enrichment of SNHG1 and miR-143-3p in the Ago2 immunoprecipitate group compared with that in the control group ( Figure 4I). SNHG1 was also demonstrated to interact with AGO2 protein in BC cells via RNA pull-down assays ( Figure 4J). Notably, SNHG1 was shown to have a negative association with miR-143-3p expression in BC tissues (R = −0.7221, P < .0001; Figure 4K). Furthermore, RNA FISH experiments indicated that SNHG1 and miR-143-3p were co-localized in the cytoplasm of BC cells ( Figure 4L). These results demonstrated that SNHG1 functions as a miR-143-3p sponge in BC.

| HK2 is a direct target of miR-143-3p that is modulated by SNHG1
As mentioned above, SNHG1 can sponge and inhibit the expression of miR-143-3p in BC. We next explored the underlying targets of miR-143-3p in BC. The potential targets of miR-143-3p were predicted using TargetScan and miRanda. A total of 197 genes were identified overlapping the prediction results from the two programs (Table S3) silencing was restored by cotransfection with si-HK2 ( Figure 5J). A positive association was observed between SNHG1 and HK2 in 60 BC tissues ( Figure 5K). Notably, HK2 knockdown repressed SNHG1 overexpression-mediated increase in BC cell proliferation ( Figure 5L).
IHC results indicated that HK2 expression was remarkably downregulated by SNHG1 knockdown in xenograft tumours formed from a stable EJ cell line ( Figure 5M). Additionally, increased protein levels of HK2 were observed in BC specimens compared with those observed in adjacent non-tumour tissues ( Figure 5N). Taken together, these findings revealed that SNHG1 functions as a ceRNA to sponge miR-143-3p, restoring the capability of miR-143-3p to bind the 3′-UTR of the target gene HK2 and inhibit its expression.

| SNHG1 epigenetically silences CDH1 transcription via EZH2-mediated H3K27 trimethylation
We also evaluated the potential functions of SNHG1 in the nucleus.
A number of lncRNAs have been shown to bind proteins that regulate transcription and alter the expression of downstream genes. 5,15 We used the online RNA-protein binding prediction site catRAPID (http://servi ce.tarta glial ab.com/page/catra pid_omics_group) to predict the ability of SNHG1 to bind to such proteins, and the results indicated that 579 proteins could potentially interact with SNHG1 (Table S4). Of note, previous reports revealed that lncRNAs located in the nucleus can recruit PRC2 to their target genes to regulate target genes expression. 31 EZH2, the catalytic subunit of PRC2, mediates the trimethylation of histone H3K27, epigenetically repressing target genes in human cancers. 32 Recent studies identified EZH2 as a non-canonical RNA-binding protein that is a key regulator in the development of tumours. 31,33 In the present study, the RNAprotein interaction prediction tool (RPISeq, http://pridb.gdcb.iasta te.edu/RPISe q/index.html) was also used to evaluate the potential interaction between SNHG1 and candidate RNA-binding proteins.
EZH2 and the top 10 predicted RNA binding proteins (Table S4 and Figure 6A) were subjected to further investigation through RIP assays, with the results indicating that only EZH2 could bind SNHG1 in both assayed BC cell lines ( Figure 6B). Moreover, RNA pull-down assay results also demonstrated that SNHG1 can specifically bind to EZH2 in BC cells ( Figure 6C). We further detected the potential downstream targets of EZH2 in BC cell lines. Ten genes (WNT1, STAB1, MYT1, CDKN1A, CDKN1B, CDKN1C, KLF2, RUNX3, CDH1 and CNR1), which have been reported to be targeted by EZH2 in human tumours, were selected as candidate genes in BC. As shown in Figure 6D,E, CDH1 was demonstrated to be obviously up-regulated by EZH2 knockdown in both BC cell lines. Subsequently, we investigated whether SNHG1 knockdown affected PRC2-mediated epigenetic inhibition in BC cells. As shown in Figure 6F Figure 6M). Additionally, IHC results indicated that E-cadherin expression was reduced in BC tissues, whereas EZH2 expression was elevated in BC tissues ( Figure 6N). Taken together, these findings illustrated the regulatory mechanism of SNHG1 in BC cell progression, by which SNHG1 functions as a platform for recruiting EZH2 to the promoter region of CDH1 that is epigenetically repressed by EZH2-induced H3K27me3 formation. In contrast, cytosolic lncRNAs have been shown to regulate the stability of mRNA, the localization of protein and to function as microRNA sponges. 35 In the present study, SNHG1 was observed to be located in both the cytoplasm and nucleus of BC cells.

| D ISCUSS I ON
Therefore, we further investigated the possible mechanism by which SNHG1 exerts its function in the BC cell cytoplasm and nucleus. Increasing evidence has shown that lncRNAs can post-transcriptionally modulate the protein expression of miRNA target genes by sponging and inhibiting miRNAs in the cytoplasm. 30,36 For example, the lncRNA ARNILA was confirmed to sponge miR-204 and to enhance breast cancer cell invasion and metastasis in breast cancer. 37 It has been also demonstrated that lncRNA H19 sponges miR-141 and contributes to tumour development and chemoresistance in human colorectal cancer. 38  was shown to directly target HK2, which exerts crucial functions in the modulation of cell proliferation. As a rate-determining enzyme, HK2 catalyses the first essential step of glycolysis and is characterized as a metabolic hallmark of human carcinoma cells. 44 The up-regulation of HK2 has been reported in several types of human neoplasms, including BC, and the increased protein expression of HK2 was shown to be required for glucose metabolism and cancer cell progression. 45,46 Recent research has highlighted that some lncRNAs are involved in the reactivation of HK2 inhibited by miRNAs in human cancers. [47][48][49][50] Notably, we showed that SNHG1 acts as an endogenous ceRNA that regulates HK2 expression via sponging miR-143-3p in BC cells. In addition, SNHG1 expression was identified to be negatively associated with miR-143-3p expression, whereas SNHG1 expression was shown to be positively correlated with HK2 expression in BC specimens. These data were in line with our hypothesis and the previously reported findings, 51 indicating the existence of a SNHG1/miR-143-3p/HK2 signalling axis in the cytoplasm of BC cells.
Evidence from a number of studies has implicated lncRNAs in the epigenetic modulation of gene expression via DNA methylation. [52][53][54] The histone methyltransferase EZH2, the catalytic subunit of PRC2, has been confirmed to promote the trimethylation of H3K27 (H3K27me3), which is involved in the epigenetic silencing of target genes during cancer development. In our experiments, we showed that SNHG1 directly binds to EZH2 and is involved in the EZH2-mediated epigenetic silencing of the EMT marker CDH1. As a member of the cadherin superfamily, CDH1 (E-cadherin) expression is decreased and behaves as a tumour suppressor to block the migration, invasion and metastasis of numerous human carcinomas. 21,22 Additionally, the results of previous studies indicated that EZH2-mediated hypermethylation of the CDH1 promoter induces the down-regulation of CDH1 expression in human cancers, including BC. 55,56 In accordance with previous findings, our data confirmed that SNHG1 recruits EZH2 to the promoter region of CDH1 and then epigenetically suppresses the transcription of CDH1. Notably, the elevated levels of H3K27me3 induced by EZH2 were shown to be involved in this process. Moreover, the loss-and gain-of-function analysis results further proved that SNHG1 and the change in EZH2 and CDH1 expression are crucial in the modulation of BC cell migration, invasion, EMT and metastasis.
Thus, in the nucleus, SNHG1 and its downstream target genes form a regulatory network that regulates the progression of BC.
Taken together, the results of our present work highlights the oncogenic role of SNHG1 in human BC. In the cytoplasm of BC cells, SNHG1 sponges miR-143-3p to alleviate the inhibitory effects of miR-143-3p on HK2, thereby promoting the proliferation of BC cells.
Additionally, SNHG1 was shown to interact with EZH2 and facilitates BC cell migration, invasion and metastasis via the epigenetic silencing of CDH1 in the nucleus (Figure 7). The present study not only reveals the diverse regulatory mechanisms of SNHG1 in different subcellular locations, but also represents a promising therapeutic option for inhibiting BC progression.

CO N FLI C T S O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated or analysed during this study are included in this article.