SNHG17, as an EMT‐related lncRNA, promotes the expression of c‐Myc by binding to c‐Jun in esophageal squamous cell carcinoma

Abstract Dysregulation of long noncoding RNA SNHG17 is associated with the occurrence of several tumors; however, its role in esophageal squamous cell carcinoma (ESCC) remains obscure. The present study demonstrated that SNHG17 was upregulated in ESCC tissues and cell lines, induced by TGF‐β1, and associated with poor survival. It is also involved in the epithelial‐to‐mesenchymal transition (EMT) process. The mechanism underlying SNHG17‐regulated c‐Myc was detected by RNA immunoprecipitation, RNA pull‐down, chromatin immunoprecipitation, and luciferase reporter assays. SNHG17 was found to directly regulate c‐Myc transcription by binding to c‐Jun protein and recruiting the complex to specific sequences of the c‐Myc promoter region, thereby increasing its expression. Moreover, SNHG17 hyperactivation induced by TGF‐β1 results in PI3K/AKT pathway activation, promoting cells EMT, forming a positive feedback loop. Furthermore, SNHG17 facilitated ESCC tumor growth in vivo. Overall, this study demonstrated that the SNHG17/c‐Jun/c‐Myc axis aggravates ESCC progression and EMT induction by TGF‐β1 and may serve as a new therapeutic target for ESCC.

Long noncoding RNAs (lncRNAs) are RNA molecules with length exceeding 200 nucleotides (nt) and account for >76% of noncoding RNAs (ncRNAs). 7,8 The lncRNAs have been implicated in a wide repertoire of cellular processes, including cell differentiation, cell cycle regulation, cell apoptosis, and epithelial-to-mesenchymal transition (EMT). 9,10 Recent studies have shown that lncRNAs serve as master regulators in tumor progression and play a key role in encouraging or combating carcinogenesis, including ESCC. 11,12 The oncogenic role of a lncRNA, small nucleolar RNA host gene 17 (SNHG17), has been reported, in which it annotates cancer-associated lncRNA functionally and is mechanistically linked to disease pathogenesis in colon cancer, 13 prostate cancer, 14 gastric cancer, 15 and non-small-cell lung cancer. 16 SNHG17 acts as a prognostic factor, promoting tumor proliferation and metastasis and is related to poor prognosis in these cancers. Nonetheless, how SNHG17 participates in the progress of ESCC is yet unknown.
EMT is a cellular process that regulates cancer cell invasion, metastatic potential, and resistance to radiation and chemotherapy. TGFβ acts as a potent EMT-inducing cytokine in promoting cancer cell invasion and metastasis. 17 Moreover, TGFβ-induced lncRNAs have been shown to play crucial roles in various carcinogenic processes. For instance, long noncoding RNA activated by transforming growth factorbeta (lncRNA-ATB) may be induced by TGFβ treatment to facilitate the malignant progression of hepatocellular carcinoma and breast cancer cells. 18,19 TGFβ-activated lncRNA UCA1 promotes the proliferation of bladder transitional carcinoma cells through the non-canonical PI3K-AKT pathway. 20 In prostate cancer, knockdown of SNHG17 weakened the EMT progression of cancer cells. 14 Together, these studies suggested that lncRNA plays a major role in TGFβ-induced tumor invasion and metastasis. However, few studies have addressed the mechanism of interaction between lncRNA and TGFβ classical and non-classical pathways in ESCC. Additionally, SNHG17 promotes tumor invasion and metastasis in oral squamous cell carcinoma, 21 gastric cancer, 22 and other gastrointestinal tumors. 23 Therefore, we speculated that SNHG17 is related to TGFβ ESCC; and whether TGFβ-regulated SNHG17 is involved in EMT needs further exploration.
In the current study, we aimed to investigate the biological function of SNHG17 and its role in TGF-β1-induced EMT in ESCC.
Therefore, we unveiled the molecular mechanism underlying the pathogenesis of ESCC. These findings indicated that SNHG17 acts as a potential therapeutic strategy for ESCC.

| Microarray data analysis
The human microarray dataset (GSE20347) was downloaded from the public GEO database (http://www.ncbi.nlm.nih.gov/geo). Primary tumor samples (17 ESCC patients from a high-risk region of China) were used to analyze these datasets. The DEGs were screened out with a cut-off value of P < .05 and log 2 (fold change) > 2. For RNAsequencing, total RNA was obtained from Eca109 cells transfected with sh-SNHG17 or sh-NC and subjected to sequencing.

| Human ESCC tissue specimens
In total, 128 matched ESCC and paired para-cancerous tissues (at least 5 cm distal to tumor lesion) were collected from the Fourth Affiliated Hospital of Hebei Medical University, Shijiazhuang, China from May 2012 to December 2015. Clinical and survival data of patients were provided by the clinical data registry (Table S1). Ethical approval for this study was obtained from the Ethics Committee of the Fourth Affiliated Hospital, Hebei Medical University.

| Quantitative real-time polymerase chain reaction (qRT-PCR) analysis
Total RNA was extracted from ESCC tissues and cells using TRIzol reagent (Invitrogen). Then, 1 μg of the total RNA was reverse transcribed into cDNA as detailed by the manufacture (Roche).
Quantitative real-time PCR assay was performed using the SYBR qPCR Master Mix (Life Technology). Relative expression levels of lncRNA/mRNA were normalized to that of the housekeeping gene β-actin and quantitated by the comparative cycle threshold (C t ) method (2 −ΔΔCt ). Primers sequences and reaction assays for this study are displayed in Table S2.

| Plasmid construction and cell transfection
Gene-specific shRNAs targeting SNHG17, control shRNAs, small interfering RNAs (siRNAs) targeting c-Jun, and siNC were synthe-  For the colony forming assay, Eca109, Kyse150 and TE1 cells were seeded into six-well plates at a density of 3000-5000 cells/well and cultured for 7 d. The surviving colonies were fixed in 4% paraformaldehyde and stained with 0.4% crystal violet solution and counted.

| ISH and immunohistochemistry (IHC)
The same paraffin-embedded sample was serially sectioned, one for ISH, the other IHC staining. Detection of SNHG17 in ESCC was per- IHC assay was performed as described previously. 24 The paraffinembedded sections were dewaxed and hydrated in a xylene bath and absolute ethanol solution. Primary antibodies used for the study were as follows: c-Myc (ab32795), Ki-67(bs-23103R). The ISH and IHC results were evaluated as follows: 0 (no staining, no positive cells), 1 (1%-25% positive cells), 2 (26%-50% positive), 3 (51%-75% positive) and 4 (>75% positive). Tissues with scores of 3 and 4 were classified as high expression, and those with scores of 0, 1, and 2 were deemed low expression.
To eliminate the subjective evaluation error of percentage, ImageJ software (National Institutes of Health, USA) was used to score signals for inconsistencies in the percentages of the IHC and ISH signals.

| Isolation of cytoplasmic/nuclear RNA
The nuclear and cytoplasmic fractions of Eca109 and Kyse150 cells were isolated according to the protocol of the PARIS™ Kit Protein and RNA Isolation System (Invitrogen). Total RNA was isolated simultaneously from independent cytoplasmic and nuclear fractions.
Finally, qRT-PCR was used to determine the relative expression of SNHG17 to estimate the relative ratios of SNHG17 in nuclear and cytoplasmic RNA concentrations.

| RNA-binding protein immunoprecipitation assay (RIP)
RIP was performed for Eca109 and Kyse150 cells using the Magna RIP™ RNA-Binding Protein Immunoprecipitation Kit (Millipore).
Eca109 and Kyse150 cells were lysed in complete RIP lysis buffer (Millipore). The lysates were incubated with magnetic beads (Millipore) conjugated to anti-c-Jun antibody (Cell Signaling Technology) and control IgG antibody. Finally, total RNA was extracted for subsequent qRT-PCR analysis.

| RNA pull-down assay
Biotin-labeled RNA pull-down assays were performed on Eca109

cells as instructed by the supplier for Pierce™ Magnetic RNA-Protein
Pull-Down Kit (Thermo Fisher Scientific). SNHG17 and antisense-SNHG17 were transcribed in vitro from linearized constructs using the RiboMAX™ Large-Scale RNA Production Systems (Promega).
Streptavidin magnetic beads (Promega) were applied to isolate the RNA-binding protein complexes. Finally, proteins were detected by western blot assay.

| Chromatin immunoprecipitation (ChIP)
ChIP assay was conducted in Eca109 cells. Cross-linked chromatin DNA was sheared into fragments (200-1000 bp) using ultrasound, and immunoprecipitated with anti-c-Jun (Cell Signaling Technology).
After washing the beads, purified DNA was isolated for qRT-PCR.

| Statistical analysis
All statistical analyses were performed using the SPSS 22.0 software.

Student t test or paired Student t test was conducted to compare
between the 2 groups. Chi-square (χ 2 ) test was applied to analyze the association between SNHG17 expression and clinicopathological features. All experiments data were presented as mean ± SD from 3 independent experiments performed in duplicate. P < .05 indicated statistical significance.   ×100). F, G, qRT-PCR and western blot were performed to examine the expression levels of Ecadherin and N-cadherin in ESCC cells. Error bars are shown as mean ± SD, **P < .01

| SNHG17 is elevated in ESCC tissues and cell lines and associated with a poor prognosis
Based on searches of GEO datasets (GSE20347) and GEPIA (http://gepia2.cance r-pku.cn/#index), the analysis showed that SNHG17 was markedly upregulated in esophagus cancer samples  (Figures 1C and S1). Moreover, 128 pairs of ESCC and adjacent normal tissues were evaluated. Patients were divided into 2 groups based on the median value: high SNHG17 expression (fold change ≥ median, n = 63) and low SNHG17 expression (fold change ≤ median, n = 65) groups. As shown in Figure 1D, SNHG17 expression in cancer tissues was higher than that in the adjacent normal tissues. Upregulated SNHG17 expression was related to TNM stage, depth of invasion, tumor differentiation, lymph node metastasis, and mortality (Table 1). Additionally, Kaplan-Meier survival analysis indicated that patients with highly expressed SNHG17 had a significantly poor overall survival ( Figure 1E). Univariate and multivariate Cox regression analyses demonstrated that upregulated SNHG17 was an independent prognostic marker for ESCC patients (Table 2).

| SNHG17 boosts the proliferation, lessens apoptosis, and elevates migration, invasion, and EMT of ESCC cells
To address the biological functions of SNHG17 in ESCC cell lines,

| SNHG17 is upregulated in TGFβ-induced Eca109 cells and involved in TGF-β1-mediated EMT
EMT is a key player in promoting cancer cell migration and invasion. 25 Moreover, since TGF-β1 is required for EMT, we further es- its expression following TGFβ1 treatment in Eca109 cells ( Figure 3A).
The expression of Eca was significantly downregulated, while the levels of N-cadherin, Twist1, and SNAI2 expression in Eca109 cells was increased following TGF-β1 treatment ( Figure 3B). As SNHG17 was induced by TGF-β1, we further investigated whether silencing SNHG17 abrogated TGFβ-mediated EMT. As shown in Figure 3C, Taken together, these data indicated that SNHG17 may be an EMTrelated lncRNA and contribute to TGFβ-induced EMT in ESCC cells.

| SNHG17 upregulates the expression of c-Myc and medicates the PI3K/AKT pathway in ESCC
To elucidate the potential mechanism by which SNHG17 promotes the progression of ESCC, we performed an RNA-sequencing assay  Figure 4D). c-Myc was confirmed to having a prominent role in sustaining tumor growth in many tumor types, including ESCC. 26 As shown by ISH and IHC, we found that tumor samples for high expression of SNHG17 showed an increased expression frequency of c-Myc protein, and positive staining occurred in the nucleus ( Figure 4E). Therefore, we speculated that SNHG17 positively regulated c-Myc expression.

| SNHG17 binds to transcription factor c-Jun
The dysregulation of SNHG17 decreased the expression of c-Myc dramatically. Furthermore, overexpression of SNHG17 increased the transcriptional activity of the c-Myc promotor, indicating that SNHG17 potentially regulated c-Myc by acting on its promoter ( Figure 5A). Subsequently, to explore the molecular mechanism by which SNHG17 modulates c-Myc expression, we further examined the localization of SNHG17 through subcellular fractionation in Eca109 and Kyse150 cells, illustrating that SNHG17 was predominantly localized in the nucleus ( Figure 5B). According to a previous study, lncRNAs located in the nucleus regulated the transcriptional process through binding to transcription factors (TFs), which in turn degenerated the DNA sequence motifs on the promoter in specific oncogenes in cancer cells. 27 Therefore, we proposed a tentative mechanism in which SNHG17/TF may activate c-Myc expression through binding to the promoter region. First, we predicted the TF binding to the c-Myc promoter region using PROMO and University of California, Santa Cruz (UCSC) genome browser ( Figure 5C). Then, we evaluated the possibility of SNHG17 interaction with the TFs via the catRAPID algorithm. Notably, the interaction between SNHG17 and c-Jun was higher ( Figures 5D and S4A,B). CatRAPID fragments was used to estimate individual interaction propensities of polypeptide and nucleotide sequence fragments, further suggesting that the 100-400 nucleotide positions of the SNHG17 sequence may bind to the amino acid residues of the c-Jun protein with high propensity ( Figure 5E). c-Jun is also a TF that regulates TGF-β1 at the transcriptional level. 28 We also found that depletion of c-Jun attenuated the ectopic expression of SNHG17 induced by TGF-β1 ( Figure S4C).
Therefore, TF c-Jun was selected for further research.
To evaluate the interaction between SNHG17 and c-Jun, both whole cell lysate and nuclear fraction were incubated with biotinylated SNHG17 or its antisense RNA transcribed in vitro, the results of the RNA pull-down assay revealed that SNHG17 bound  Figure S4D).

| SNHG17 recruits c-Jun, which regulates the expression of c-Myc by binding its promoter region
Next, we analyzed the effect of c-Jun on c-Myc in ESCC cells.

| SNHG17/c-Jun/c-Myc axis promotes ESCC cell growth and metastasis in vitro and in vivo
Rescue assays were performed to observe whether SNHG17 boosted Also, qRT-PCR analysis showed that overexpression of SNHG17 increased the expression of SNHG17 and c-Myc in xenograft tissues ( Figure 7F). To further elucidate the tumorigenic ability of SNHG17, the data of western blot indicated that the levels of c-Myc, N-cadherin, Twist1, SNAI2, and Ki-67 were enhanced by SNHG17 upregulation in xenografts ( Figure 7G). Consistently, IHC assay showed that the xenograft tissues formed by SNHG17 upregulation showed higher positive staining for c-Myc, N-cadherin and Ki-67 than that in the control group ( Figure 7H). Together, these findings suggested that SNHG17 facilitates ESCC tumor growth in vivo.

| D ISCUSS I ON
In the present study, we confirmed that SNHG17 was upregulated in ESCC tissues and cell lines and was a poor prognosis factor for ESCC EMT is characterized by epithelial cells acquiring the mesenchymal cell phenotype, which is widely implicated in tumor invasion and metastasis. TGFβ, a potent EMT-inducing cytokine, promotes cancer cell invasion and metastasis. [38][39][40] Previous studies have indicated that TGFβ induced several lncRNAs, such as lncRNA-ELIT-1, 41 lncRNA-smad7, 42 lncRNA UCA1, 43 lncRNA-ATB, 44 ln-cRNA NKILA, 45 MALAT1, 46 and MEG3, 47 which are crucial for carcinogenic processes orchestrated by TGFβ. lncRNA-ELIT-1, lincRNA-ATB, lncRNA NKILA, MALAT1, and MEG3 are involved in EMT. In the current study, SNHG17 expression was upregulated by TGF-β1 treatment in Eca109 cells, which created an effect on EMT, including E-cadherin and N-cadherin. Furthermore, we found that depletion of SNHG17 prominently restrained the original cobble-stone shape to a spindle-like shape, spindle-like formation in Eca109 cell lines, and its depletion also largely overrode the inhibition of E-cadherin and the induction of N-cadherin by TGFβ treatment. Therefore, SNHG17 was deemed to be an EMT-related lncRNA and positively contributed to TGFβ-induced EMT in ESCC cells.
c-Myc is a vital tumor promoter, and its expression can be con-