Long noncoding RNA HOTTIP promotes the metastatic potential of ovarian cancer through the regulation of the miR ‐615‐3p/ SMARCE1 pathway

Abstract

Upregulation of lncRNA HOXA transcript at the distal tip (HOTTIP) plays important roles in cancer progression. Nevertheless, its functions in the growth and metastasis of ovarian carcinoma are unknown. In this study, we demonstrated overexpression of HOTTIP in ovarian cancer cell lines and clinical tissues. Further, we showed that higher level of HOTTIP was associated with poor survival of ovarian cancer patients. Notably, HOTTIP silencing restrained proliferation, migration, and invasiveness of ovarian carcinoma cells. On the other hand, upregulation of HOTTIP remarkably exacerbated the aggressive traits of ovarian carcinoma cells. In addition, HOTTIP served as a sponge of miR‐615‐3p to upregulate SMARCE1 level. Further, upregulation of miR‐615‐3p or downregulation of SMARCE1 reversed the carcinogenic impacts of HOTTIP in ovarian cancer. HOTTIP and miR‐615‐3p expression levels in ovarian cancer cells were negatively correlated, whereas HOTTIP and SMARCE1 expression levels were positively correlated. In nude mice, downregulation of HOTTIP reduced cell growth in vivo. In summary, lncRNA HOTTIP promotes the growth and metastatic phenotypes of ovarian cancer via regulating miR‐615‐3p/SMARCE1 pathway.

INTRODUCTION

Ovarian carcinoma is the second leading cause of cancer‐related deaths in women.¹ Although many advances, such as tumor screening, imaging technology, and novel treatment strategies, have been made in the management of ovarian cancer, the clinical prognosis of ovarian cancer patients is still unsatisfactory.² Therefore, exploring new biomarkers and new therapeutic targets is essential to improve the prognosis of ovarian cancer patients.

Previous studies report that long noncoding RNAs (lncRNAs), a class of noncoding RNA transcripts; markedly distinguish cancer tissues from corresponding normal tissues.^3,4 For instance, LINC00511 is upregulated in human clear cell renal cell carcinoma (ccRCC) and promotes the malignant phenotype of ccRCC via sponging microRNA‐625 and increasing cyclin D1 expression.⁵ Further, LncRNA UICLM promotes the metastasis of colorectal cancer cells by acting as a ceRNA for miRNA‐215 to regulate ZEB2 expression.⁶ In ovarian cancer, several lncRNAs have been identified as critical cancer suppressors or oncogenes. LncRNA RP11‐552M11.4 induces growth and progression of ovarian cancer by targeting BRCA2.⁷ In addition, LncRNA MLK7‐AS1 promotes the progression of ovarian cancer by regulation of miR‐375/YAP1 axis.⁸ These studies indicate that lncRNAs are upregulated in ovarian carcinoma and promote growth and metastatic ability of cancer cells.

Previous studies report that upregulation of HOTTIP frequently occurs in human malignancies and contributes to oncogenesis.^9–11 For example, lncRNA HOTTIP promotes BCL‐2 expression and induces chemoresistance in small cell lung cancer by sponging miR‐216a.⁹ In prostate cancer, HOTTIP facilitates the growth and mobility of cancer cells by sponging miR‐216a‐5p.¹² In ovarian cancer, HOTTIP enhances IL‐6 level and causes immune escape of ovarian carcinoma cells by raising PD‐L1 levels in neutrophils.¹³ Currently, the role of lncRNA HOTTIP in regulating metastasis of ovarian cancer is not clear. This research aimed to explore the underlying mechanisms of lncRNA HOTTIP in the regulation of ovarian cancer growth and metastasis.

MATERIALS AND METHOD

Ovarian cancer tissues

A total of 42 pairs of human ovarian cancer samples and paracancerous samples (normal tissues) were obtained from Weifang Maternal and Child Health Care Hospital. This study was approved by the Ethics Committee (No: 20151109) of Weifang Maternal and Child Health Care Hospital and written consent was obtained from all patients. None of the patients received chemotherapy or radiotherapy before this study.

Cell culture and transfections

Human ovarian cancer cell lines (SOV3, OVCAR3, and A2780) and normal ovarian cell line (HcerEpic) were obtained from Nanjing KeyGen Biotech. Inc. (Nanjing, Jiangsu, China). SiRNA targeting HOTTIP (si‐HOTTIP), siRNA negative control (si‐NC), siRNA targeting (si‐SMARCE1), miR‐615‐3p mimics, and miRNA negative control (miR‐NC) were purchased from GenePharma (Shanghai, China). HOTTIP expression vector (pcDNA3.1‐HOTTIP) was synthesized and constructed by RiboBio (Guangzhou, Guangdong, China). pcDNA3.1 empty vector was used as the control. Specific shRNA oligonucleotides targeting HOTTIP (sh‐HOTTIP) and negative control (sh‐Con) were obtained from Shanghai Genepharma Co., Ltd. (Shanghai, China). SKOV3 or OVCAR3 cells were transfected with miRNAs or siRNA using Lipofectamine 2000 (Thermo Fisher Scientific, Waltham, Massachusetts) for 24 hours.

Quantitative reverse transcription PCR assay

Total RNA in cells was extracted using TRIzol kit (Thermo Fisher Scientific). Total RNAs (500 ng) were reverse transcribed to cDNA with a PrimeScript Reverse Transcriptase Kit (Takara, Dalian, China). Further, miR‐615‐3p and HOTTIP were detected by quantitative reverse transcription PCR (qRT‐PCR) using a Bulge‐Loop TM miRNA qRT‐PCR Starter Kit (Applied RiboBio Biotechnology, Guangzhou, Guangdong, China) and a SYBR Green PCR Kit (Takara), respectively. The primer sequences were as follows: HOTTIP—Forward: 5′‐CCTAAAGCCACGCTTCTTTG‐3′; HOTTIP—Reverse 5′‐TGCAGGCTGGAGATCCTACT‐3′; miR‐615‐3p—Forward: 5′‐ACACTCCAGCTGGGTCCGAGCCTGGGTCTC‐3′ miR‐615‐3p—Reverse: 5′—TGGTGTCGTGGAGTC‐3′; SMARCE1—Forward: 5′‐CCTGCAACACAAATGCCCAG‐3′ SMARCE1—Reverse: 5′‐TTTTGGAATCGTGATACCAGAGG‐3′; GAPDH—Forward: 5′‐CTCACCGGATGCACCAATGTT‐3′ GAPDH—Reverse: 5′‐CGCGTTGCTCACAATGTTCAT‐3′; U6—Forward: 5′‐CTCGCTTCGGCAGCACATATACT‐3′ U6—Reverse: 5′‐ACGCTTCACGAATTTGCGTGTC‐3′. GAPDH and U6 were used as internal controls. RNA expression was normalized against GAPDH using the 2^−△△Ct method.

Kaplan‐Meier plotter database

The Kaplan‐Meier (KM) plotter (http://kmplot.com/analysis) is an online analysis tool based on GEO as well as TCGA database. It assesses the association of genes on survival in several types of cancer samples, including ovarian cancer. To analyze the overall survival (OS) of patients with low or high HOTTIP, patients were split by auto select best cutoff. GSE30161 and GSE63883 were assigned cutoff values of 42 and 28, respectively. Hazard ratio (HR; and 95% confidence intervals) and log rank P were calculated and displayed on the webpage.

Cell proliferation

SKOV3 or OVCAR3 cells were plated into a 96 well plate and cultured for 1, 2, 3, or 4 days, respectively. The proliferation of ovarian cancer cell was analyzed by Cell Counting Kit‐8 (CCK‐8) kit (Beyotime Biotechnology, Nanjing, Jiangsu, China). A volume of 10 μL of CCK8 solution was added into 96‐well plates and the OD value detected using a microplate reader at 450 nm.

Colony formation

SKOV3 or OVCAR3 cells (1 × 10³) were cultured into 6‐well plates. After 2 weeks, cells in plates were fixed by methanol and were stained by 1% crystal violet. The number of cell colonies (>50 cells) was counted.

Migration assay

SKOV3 or OVCAR3 cells were cultured overnight into 6‐well plates. The cells were scratched using a sterile 100‐μL pipette tip. Cells were cultured for 24 hours. The wound width was observed after 0 hour and 24 hours under an inverted microscope.

Invasion assay

The basement membrane of Transwell upper chamber was precoated with Matrigel. The cell suspension (1 × 10⁵) was added into the upper chamber and 600 μL medium was plated into the Transwell lower chamber. After 18 hours, the invaded ovarian cancer cells were fixed with formaldehyde and then stained by crystal violet (1%). The photograph was taken using an inverted microscope.

Luciferase reporter gene assay

To explore the relationship between miR‐615‐3p and HOTTIP, the wild type (wt) or mutant (mut) HOTTIP was inserted into the pmirGLO vector (Promega, Madison, Wisconsin). To explore the relationship between miR‐615‐3p and SMARCE1, the wt or mut SMARCE1 3′‐untranslated region (3′‐UTR) was inserted into the pmirGLO vector (Promega). SOV3 and OVCAR3 cells were cotransfected with reporter plasmid and miR‐615‐3p mimics for 48 hours. Luciferase activity in cells was measured using the luciferase assay system (Promega).

Immunoblotting

Total proteins were extracted from cells. Twenty‐five microgram proteins were electrophoresed using 8% SDS‐PAGE and transferred to PVDF membrane (Millipore, Braunschweig, Germany). The PVDF membrane was blocked in 5% BSA and then incubated with SMARCE1 or GAPDH antibody overnight at 4°C. Further, the PVDF membrane was incubated with the secondary antibody (1:5000) for 2 hours. The bands were visualized using ECL reagent (Thermo Fisher Scientific).

Ovarian cancer cell growth in vivo

BALB/c nude mice were randomly divided into two groups (n = 6). Mice were injected subcutaneously with 1 × 10⁶ sh‐HOTTIP or sh‐Con transfected SKOV3 cells to construct a xenograft model. Tumor length and width were measured once a week. After 5 weeks, the nude mice were sacrificed. Tumor size was represented as tumor = 0.5 × length×width². The study was approved by the animal Ethics Committee of Weifang Maternal and Child Health Care Hospital.

Statistical analysis

All data were analyzed using SPSS19.0 and presented as mean ± SD. Analysis of variance (ANOVA) was used for statistical analysis and differences between groups compared using Student t test. The relationship between two variables was analyzed by Spearman's correlation analysis. OS curves were plotted using the Kaplan‐Meier method and Kaplan‐Meier plotter (http://kmplot.com/analysis/index.php?p=service&cancer=ovar). P < .05 was considered as statistically significant.

RESULTS

HOTTIP is upregulated in ovarian carcinoma

HOTTIP expression levels in ovarian carcinoma samples were determined by qRT‐PCR assay. As shown in Figure 1A, HOTTIP was highly overexpressed in ovarian carcinoma compared with adjacent normal tissues. HOTTIP expression levels were significantly high in the ovarian cancer cells (SKOV3, OVCAR3, and A2780) compared with the levels in normal ovarian cell line, HcerEpic (Figure 1B). Notably, survival analysis using Online Kaplan‐Meier Plotter (http://kmplot.com/analysis/index.php?p=service&cancer=ovar) indicated that high HOTTIP expression level was associated with poor OS and progression‐free survival (PFS) of ovarian carcinoma patients (Figure 1C,D). Further, we analyzed a publicly available gene expression profiling data set (GEO database‐GSE14001) for differential expression of genes in three normal vs 20 ovarian carcinoma tissues. As shown in Figure 1E, HOTTIP was significantly upregulated in ovarian carcinoma tissue compared with normal. In addition, we calculated the prognostic value of HOTTIP in ovarian carcinoma and analyzed ovarian carcinoma prognosis using GEO database‐GSE30161 and GSE63885. As shown in Figure 1F,G, patients with higher HOTTIP levels exhibited poor OS than patients with lower HOTTIP levels. These findings indicated that HOTTIP was upregulated in ovarian carcinoma and was negatively correlated with the clinical outcome of ovarian carcinoma patients.

1 FIGURE. Expression of HOTTIP in ovarian cancer. A, Expression levels of HOTTIP in ovarian cancer tissues and adjacent normal tissues (n = 42). **P < .01 compared with normal. B, Expression levels of HOTTIP in normal ovarian cell lines and ovarian cancer cell lines. **P < .01 compared with HcerEpic. C, Kaplan‐Meier plot of ovarian cancer patients with high HOTTIP expression levels had poor OS. D, Kaplan‐Meier overall survival curve analysis revealed that high expression levels of HOTTIP was associated with poor PFS of ovarian cancer patients. E, Expression of HOTTIP in normal (n = 3) and ovarian carcinoma (n = 20) tissues. Data represent log2 expression values. F, Kaplan‐Meier estimates of the overall survival (OS) of ovarian carcinoma patients in GSE30161 data set. G. Kaplan‐Meier estimates of the OS of ovarian carcinoma patients in GSE63885 data set

HOTTIP silencing restrains ovarian carcinoma cell growth

To suppress HOTTIP expression, siRNA targeting lncRNA HOTTIP (si‐HOTTIP) or siRNA control (si‐NC) was transfected into ovarian carcinoma cell, SKOV3, and OVCAR3. As shown in Figure 2A, si‐HOTTIP significantly reduced HOTTIP expression levels in both SKOV3 and OVCAR3 cell lines. Further, the proliferation rate of HOTTIP silencing ovarian cancer cell was significantly inhibited (Figure 2B). In addition, the numbers of cell colonies in SKOV3 and OVCAR3 cell were reduced considerably by si‐HOTTIP (Figure 2C). The migration ability and invasiveness of cells were detected using the wound closure and Transwell invasion assay. As shown in Figure 2D,E, the migration and invasion capacities were remarkably low in HOTTIP silencing group compared with the si‐NC group. To further validate the effect of HOTTIP on the aggressive traits of ovarian carcinoma cell, SKOV3, and OVCAR3 cell lines were transfected with pcDNA3.1 empty vector or pcDNA3.1 carrying HOTTIP (HOTTIP). As shown in Figure 3A, HOTTIP levels were significantly high in cells transfected with HOTTIP. These findings show that overexpression of HOTTIP accelerated the growth of SKOV3 and OVCAR3 cells as determined by colony formation assay (Figure 3B). Similarly, overexpression of HOTTIP enhanced the migration and invasion of SKOV3 and OVCAR3 cells in vitro (Figure 3C,D). These data suggested that HOTTIP promoted the growth and metastatic phenotypes of ovarian carcinoma cell.

2 FIGURE. Downregulation of HOTTIP inhibits the malignant behavior of ovarian cancer cell. A, The levels of HOTTIP in SKOV3 and OVCAR3 cells were detected using qRT‐PCR. B, The effect of HOTTIP on cell viability was determined by CCK‐8 assay. C, Colony formation assay for the effect of HOTTIP on cell growth. D, The migration of SKOV3 and OVCAR3 cells was measured using wound healing. E, Effect of HOTTIP on the invasion of SKOV3 and OVCAR3 cells. **P < .01 compared with si‐NC

3 FIGURE. HOTTIP promotes ovarian carcinoma cell growth, migration, and invasion. A, The levels of HOTTIP in SKOV3 and OVCAR3 cell lines transfected with vector or HOTTIP were detected using qRT‐PCR. B, Colony formation assay for the effect of HOTTIP on cell growth. C, The migration of SKOV3 and OVCAR3 cells was measured using wound healing. D, Effect of HOTTIP on the invasion of SKOV3 and OVCAR3 cells. **P < .01 compared with vector

HOTTIP serves as a sponge of miR‐615‐3p

By using bioinformatics analysis tool (http://starbase.sysu.edu.cn/starbase2/mirLncRNA.php), we observed that miR‐615‐3p was a potential target for HOTTIP (Figure 4A). Luciferase reporter gene assay indicated that miR‐615‐3p inhibited luciferase activities in SKOV3 and OVCAR3 cells that were transfected with pmirGLO containing wt‐HOTTIP. However, miR‐615‐3p showed no inhibitory effect on luciferase activities in SKOV3 and OVCAR3 cells transfected with pmirGLO containing mut‐HOTTIP (Figure 4B). In addition, miR‐615‐3p levels higher in si‐HOTTIP transfected SKOV3 and OVCAR3 cell compared to that in the si‐NC group (Figure 4C). Further, miR‐615‐3p expression levels in ovarian cancer samples were determined using qRT‐PCR analysis. As presented in Figure 4D, miR‐615‐3p was downregulated in ovarian carcinoma tissue compared with that in normal tissue. Notably, a negative correlation was observed for HOTTIP and miR‐615‐3p expression levels in clinical ovarian carcinoma tissue (Figure 4E).

4 FIGURE. HOTTIP regulates miR‐615‐3p expression in ovarian cancer. A, Starbase predicts putative targeting sites for HOTTIP and miR‐615‐3p. B, Analysis of luciferase activity in SKOV3 and OVCAR3 cells cotransfected with miR‐615‐3p mimic and pmirgLO‐HOTTIP‐WT or pmirgLO‐HOTTIP‐Mut vector. **P < .01 compared with miR‐NC + wt‐HOTTIP. C, The levels of miR‐615‐3p in SKOV3 and OVCAR3 cells with HOTTIP knockdown were detected using qRT‐PCR analysis. **P < .01 compared with si‐NC. D, Expression level of miR‐615‐3p in ovarian cancer tissues and adjacent normal tissues was measured by qRT‐PCR assay. **P < .01 compared with normal. E, Spearman correlation analysis of HOTTIP and miR‐615‐3p in ovarian cancer tissues

HOTTIP upregulates SMARCE1 expression

SMARCE1 was predicted as the target gene for miR‐615‐3p using bioinformatics tools (starbase2) (Figure 5A). Luciferase reporter gene assay was performed to validate this finding. As shown in Figure 5B, transfection of miR‐615‐3p reduced luciferase activities of cells transfected with pmirGLO containing wt‐SMARCE1. On the contrary, no suppressive effects on the luciferase activities were observed for cells transfected with pmirGLO containing mut‐SMARCE1. miR‐584‐5p transfection resulted in lower mRNA levels of SMARCE1 in SKOV3 and OVCAR3 cells, whereas the expressions of SMARCE1 in cells cotransfected with miR‐584‐5p and HOTTIP were rescued (Figure 5C,D). This implied that HOTTIP inhibited miR‐615‐3p effect on SMARCE1 expression. Finally, we found that HOTTIP level was positively correlated with SMARCE1 level in ovarian cancer tissues (Figure 5E). Western blot analysis was carried out to detect protein expression levels of SMARCE1 in clinical ovarian cancer samples and paracancerous tissues from the same patients. As shown in Figure 5F, SMARCE1 levels were significantly higher in ovarian cancer tissues compared with paracancerous samples (Figure 5F). These findings implied that HOTTIP regulated SMARCE1 expression by modulating miR‐584‐5p.

5 FIGURE. HOTTIP upregulated SMARCE1 expression by sponging miR‐615‐3p. A, Putative binding sites for miR‐615‐3p and SMARCE1 3′‐UTR. B, Analysis of luciferase activity in SKOV3 and OVCAR3 cells cotransfected with miR‐615‐3p mimic and pmirgLO SMARCE13′‐UTR‐wt or pmirgLO‐SMARCE1 3′‐UTR‐mut vector. **P < .01 compared with miR‐NC + wt‐ SMARCE1. C, The mRNA levels of SMARCE1 in different groups. **P < .01 compared with miR‐NC, ##P < .01 compared with miR‐615‐3p. D, The protein expression levels of SMARCE1 in SKOV3 and OVCAR3 cells were detected using western blotting assay. E, Spearman correlation analysis of HOTTIP and SMARCE1 in ovarian cancer tissues. F, The protein expression levels of SMARCE1 in clinical ovarian cancer tissues and paracancerous samples were detected using western blotting assay

MiR‐615‐3p or SMARCE1 reverses HOTTIP‐induced promotion of ovarian cancer aggressiveness

To assess the activity of miR‐615‐3p or SMARCE1 in HOTTIP‐induced promotion of ovarian cancer aggressiveness, several rescue experiments were carried out. First, SKOV3 or OVCAR3 cell was transfected with HOTTIP alone or cotransfected with HOTTIP and miR‐615‐3p mimics (Figure 6A). Increased HOTTIP levels promoted colony formation and metastatic‐related traits of SKOV3 and OVCAR3, whereas cotransfection of HOTTIP with miR‐615‐3p mimics decreased the aggressive phenotypes of ovarian cancer cell (Figure 6B,C). Second, SKOV3 or OVCAR3 cell was transfected with HOTTIP alone or cotransfected with HOTTIP and si‐SMARCE1 (Figure 6D). Similarly, we observed that HOTTIP promoted colony formation and metastatic‐related traits of SKOV3 and OVCAR3 whereas cotransfection of HOTTIP with si‐SMARCE1 decreased the aggressive phenotypes of ovarian cancer cell (Figure 6E,F). The results suggested that overexpression of miR‐615‐3p or SMARCE1 silencing impaired HOTTIP growth promotion effect on ovarian cancer cell.

6 FIGURE. MiR‐615‐3p overexpression or SMARCE1 silencing reverses HOTTIP‐induced ovarian cancer progression. A, SMARCE1 levels in SKOV3 and OVCAR3 cell that was transfected with HOTTIP alone or cotransfected with HOTTIP and miR‐615‐3p was detected using qRT‐PCR assay. B, Colony formation assay for colony proliferation. C, Transwell for cell invasion. D, SMARCE1 level in SKOV3 and OVCAR3 cell that was transfected with HOTTIP alone or cotransfected with HOTTIP and si‐SMARCE1 was detected using qRT‐PCR assay. E, Colony formation assay for colony proliferation. F, Transwell for cell invasion. **P < .01 compared with vector, ##P < .01 compared with HOTTIP

HOTTIP promotes ovarian carcinoma cell growth in vivo

Further, lncRNA HOTTIP effect on ovarian cancer cell growth in vivo was investigated. Transfected tumor model was constructed via injection with sh‐NC or sh‐HOTTIP transfected SKOV3 cells. As shown in Figure 7A,B, HOTTIP knockdown significantly reduced both tumor weight and tumor volumes in nude mice injected with sh‐HOTTIP transfected SKOV3 cells compared with sizes in the sh‐NC group. Furthermore, miR‐615‐3p levels were higher in sh‐HOTTIP transfected group compared with sh‐NC group (Figure 7C). In addition, sh‐HOTTIP lowered SMARCE1 expression levels in tumor tissue (Figure 7D). These findings showed that downregulation of HOTTIP inhibited ovarian carcinoma cell growth and modulated miR‐615‐3p/SMARCE1 axis.

7 FIGURE. HOTTIP promotes ovarian cancer growth in vivo. A, Representative images of the tumor tissues in sh‐con and sh‐HOTTIP group are shown (upper panel). The growth curves of xenograft tumors are presented (lower panel). B,Tumor weight of tumor tissues in each group. C, miR‐615‐3p level was detected using qRT‐PCR assay. D, Protein expression level of SMARCE1 in sh‐Con and sh‐HOTTIP group (n = 3 in each group). **P < .01 compared with sh‐Con

DISCUSSION

Ovarian carcinoma is characterized by a short 5‐year survival rate and high rate of metastasis. Therefore, investigating the molecules implicated in ovarian cancer metastasis is paramount. LncRNA HOTTIP has been shown to participate in the progression of several cancers. HOTTIP facilitates the migration, invasion, and epithelial‐mesenchymal transition (EMT) process of osteosarcoma cell by positively regulating c‐Myc.¹¹ In addition, HOTTIP promotes hypoxia‐induced glycolysis by targeting miR‐615‐3p/HMGB3 axis in non‐small‐cell lung cancer cells.¹⁰ HOTTIP overexpression is positively correlated with the poor prognostic and poor clinicopathological features in prostate cancer patients.¹⁴ Consistent with previous investigations, our study revealed that HOTTIP is significantly upregulated in ovarian cancer samples as well as ovarian cancer cell lines. Further, higher HOTTIP levels were positively correlated with shorter OS of patients with ovarian cancer.

Previous studies have reported that dysregulation of lncRNAs is closely related to the metastasis of cancers.^15,16 HOTTIP promotes migration, invasiveness, and EMT of breast cancer cells.¹⁷ In addition, HOTTIP promotes proliferation and migration of prostate cancer cells in vitro.¹² In the current study, we observed that HOTTIP silencing reduced ovarian cancer cell growth and colony formation. Furthermore, downregulation of HOTTIP significantly lowered migration and invasion abilities of ovarian cancer cells. On the contrary, upregulation of HOTTIP significantly exacerbated the aggressive traits of ovarian carcinoma cell. in vivo experiments showed that HOTTIP silencing reduced tumor growth of ovarian cancer cells. All these observations imply that HOTTIP plays a role as an oncogene to promote the development of ovarian carcinoma.

One main mechanism of lncRNA is serving as a molecular sponge for miRNA. Although HOTTIP has been proved as a valuable indicator of the prognosis of ovarian cancer patients and enhances the proliferation and invasion of ovarian cancer cells,¹⁸ the specific roles of HOTTIP in the metastasis of ovarian cancer need to be explored further. In this study, we used bioinformatic tools and luciferase reporter gene analysis, to demonstrate that HOTTIP directly binds to miR‐615‐3p and lowers miR‐615‐3p levels in ovarian cancer cells. Previous reports suggest that miR‐615‐3p plays an important role in neoplasms. For instance, miR‐615‐3p acts as a tumor suppressor in NSCLC and miR‐615‐3p overexpression inhibits the growth of NSCCL cells.¹⁹ Further, high miR‐615‐3p levels are used to predict adverse clinical outcomes and miR‐615‐3p promotes the growth and migration ability of prostate cancer cells.²⁰ These findings suggest that depending on the levels, miR‐615‐3p can serve as tumor suppressor or an oncogene in cancers. Our study provides the first evidence that miR‐615‐3p promotes ovarian cancer cell growth, migration, and invasion. In addition, miR‐615‐3p expression is negatively correlated with HOTTIP expression in ovarian cancer. Rescue experiments demonstrated that overexpression of miR‐615‐3p reverses the antimetastatic ability of HOTTIP.

We also investigated the downstream target of miR‐615‐3p in ovarian cancer. By using bioinformatic tools and luciferase reporter gene analysis, we showed that SMARCE1 was a downstream target for miR‐615‐3p in ovarian cancer cells. In addition, we observed that miR‐615‐3p significantly decreased SMARCE1 expression levels in ovarian cancer cells, while HOTTIP and miR‐615‐3p cotransfection inhibited the effect of HOTTIP or miR‐615‐3p through the expression of SMARCE1. HOTTIP level and SMARCE1 level in ovarian cancer tissues were negatively correlated. Finally, miR‐615‐3p overexpression and SMARCE1 silencing inhibited the proliferation and metastatic effects of HOTTIP in ovarian cancer cells. These findings suggest that HOTTIP serves as miR‐615‐3p sponge to modulate the level of SMARCE1 in ovarian cancer.

CONCLUSION

HOTTIP facilitates growth and metastatic phenotypes of ovarian carcinoma cells by acting as a miR‐615‐3p sponge thus upregulating SMARCE1.

CONFLICT OF INTERESTS

The authors declare no conflict of interests.