Circ‐SERPINE2 promotes the development of gastric carcinoma by sponging miR‐375 and modulating YWHAZ

Abstract Objectives Circular RNAs (circRNAs) exist extensively in the eukaryotic genome. The study aimed to identify the role of hsa_circ_0008365 (Circ‐SERPINE2) in gastric carcinoma (GC) cells and its downstream mechanisms. Materials and methods Gene Expression Omnibus (GEO) database was applied to screen differentially expressed circRNAs. CircInteractome, TargetScan and miRecords websites were used to predict target relationships. qRT‐PCR and RNase R treatment were utilised to detect molecule expression and confirm the existence of circ‐SERPINE2. RNA pull‐down assay and dual‐luciferase reporter assay were performed for interaction between circRNA and miRNA or mRNA. EdU assay, colony formation assay, and flow cytometry for apoptosis and cell cycle detections were utilised to assess cell function. Western blot and immunohistochemistry (IHC) assays were applied for detection of proteins in tissues or cells. Results Circ‐SERPINE2 and YWHAZ were upregulated, and miR‐375 was downregulated in GC tissues and cells. Circ‐SERPINE2 and YWHAZ targetedly bound to miR‐375. Circ‐SERPINE2 promoted cell proliferation and cell cycle progress and inhibited cell apoptosis by sponging miR‐375 and regulating YWHAZ expression in vitro. Circ‐SERPINE2 repressed solid tumour growth through enhancing miR‐375 expression and reducing YWHAZ expression in vivo. Conclusions Circ‐SERPINE2 is a novel proliferative promoter through the regulation of miR‐375/YWHAZ. Circ‐SERPINE2/miR‐375/YWHAZ axis might provide a novel therapeutic target of GC.

and distribution in various tissues suggest that circRNAs have a vital impact on human disease. 6 Studies have shown that circRNAs have a powerful regulatory function in cancer, 4 especially in GC. The study of Zhao et al 7 showed that the hsa_circ_0000181 expression level was much higher in GC tissues than that in normal control tissues.
Its expression in GC was related to tumour diameter, lymphatic metastasis and distal metastasis. Furthermore, Xie et al 8 found that the hsa_circ_0074362 expression was observably increased in cells and tissues of GC, and the gastric cancer lymphatic metastasis was positively correlated with the hsa_circ_0074362 expression level. microRNAs (miRNAs) are a kind of non-coding RNA with a length of 18-25 nucleotides, and the abnormal expression of miRNA is closely related to various diseases, especially cancer. 9 Several researches have shown that miRNAs are involved in cell proliferation and apoptosis in many malignant tumours, such as colon cancers, 10 melanoma 11 and renal cancer. 12 Chang et al 13  YWHAZ, also known as 14-3-3 zeta, belongs to one member of 14-3-3 protein family, which is highly conservative in all eukaryotes and can regulate many cell events, such as signal transduction, cell cycle and apoptosis. 15 Recently, YWHAZ has been identified as a clinical prognostic indicator for some tumours. 16 The study of Bergamaschi et al 17 showed that 14-3-3 zeta was a key predictor for risk of failure in endocrine therapy and was valuable in recovering endocrine sensitivity and decreasing recurrence risk in breast cancer. Nishimura et al illustrated that the YWHAZ protein expression was linked to the tumour size, lymphatic and venous invasion, tumour depth, pathological stage and recurrence rate. Besides, the higher level of YWHAZ protein was significantly correlated with the lower level of miR-375 expression. 16 However, the research on the upstream molecular mechanism of YWHAZ in GC was limited, which prompts us to conduct a more in-depth study on the molecular mechanism of YWHAZ in GC.
In tumour development, circRNA-miRNA-mRNA interaction networks may be involved and play a crucial role. Here, we aimed to explore how hsa_circ_0008365 (circ-SERPINE2) exerts its function in GC through exploring the correlation among circ-SERPINE2, miR-375 and YWHAZ and their expressions in GC tissues and cells.
Furthermore, the findings from in vitro experiments were verified by in vivo experiments. This study might provide a new molecular marker for the molecular therapy of GC. and tumour-nodes-metastasis (TNM) staging system were utilised to make a definite diagnosis. Meanwhile, GC was identified by histopathology. Human Research Ethics Committee approved this study.

| Patient samples
Besides, this study was supported with written informed consent obtained from all enrolled subjects.

| Cell culture
Human gastric smooth muscle cells (HGSMC) and human GC cells

| Plasmid construction and stable transfection
Human circ-SERPINE2 cDNA was synthesised and provided by TSINGKE and then cloned into pcD-ciR vector (Geneseed Biotech) to construct circ-SERPINE2 overexpression plasmids. Then, transfection was implemented with these plasmids and GC cells. Then, the transfected cells were selected with G418 (Life Technologies).

| Oligonucleotide transfection
Short interfering RNAs (siRNAs) for hsa_circ_0008365 (circ-SERPINE2) knockdown were designed by CircInteractome web and synthesised by RiboBio. The siRNA against for YWHAZ (si-YWHAZ), miR-375 mimics, miR-375 inhibitor, universal siRNA negative control (si-NC), mimics control (mimics-NC) and inhibitor control (inhibitor-NC) were provided by RiboBio. The sequences of siRNAs were displayed in Table 2. After treated with DNase I (Roche), the lysate was then diluted with 900 μL RIP immunoprecipitation buffer and incubated with antibody-coupled magnetic beads (anti-AGO2, Abcam). RNA immunoprecipitation mixture (10 μL) was saved as input. Beads were subsequently washed six times with RIP wash buffer and treated with proteinase K at 37°C for 30 minutes. RNA was extracted using TRIzol reagent (Invitrogen) according to the manufacturer′s instructions.
Biotinylated-circ-SERPINE2 probe was synthesised and provided by RiboBio. The circ-SERPINE2 probe was incubated at 25°C for 2 hour with C-1 magnetic beads (Life Technologies), and the cell lysates were incubated at 4°C overnight with oligo probe or circ-SERPINE2 probe. The RNA complexes combined with the beads were extracted using RNeasy Mini Kit (QIAGEN) for RT-PCR.

| Pull-down assay with biotinylated miR-375
Biotinylated miR-375 mimics and its mutant were synthesised by RiboBio. The stably expressed circ-SERPINE2 GC cells were transfected with 50 nM of biotinylated miR-375 mimics or mutant using Lipofectamine RNAiMax (Life Technologies), harvested after transfection for 48 hour and followed by sonication. The cell lysates (50 μL) were aliquot for input, and the remanent cell lysates were incubated at 4°C for 3 hour with C-1 magnetic beads (Life Technologies) and Abbreviation: NC, negative control. then washed by wash buffer. The combined RNAs were purified with RNeasy Mini Kit (QIAGEN) for RT-PCR.

| Luciferase reporter assay
The circ-SERPINE2 or YWHAZ 3′UTR containing miR-375 binding sites were synthesised and subcloned into a pGL2-Base vector (Promega). HEK-293T cells were plated into 96-well plates and were co-transfected with a mixture of firefly luciferase reporter, pRL-CMV Renilla luciferase reporter, and miR-375 mimics or its mutant control using Lipofectamine 2000 (Invitrogen). After 48 hour of incubation, the firefly and Renilla luciferase activities were quantified with a dual-luciferase reporter assay system (Promega). The Renilla luciferase activity was considered as internal control.

| Western blot
The

| Colony formation assay
Within the colony formation assays, the transfected cells were trypsinised, plated in 24-well plates and incubated routinely at 37°C for 14 days. Colonies were fixed using methanol for 10 minutes and dyed with 0.1% crystal violet for 15 minutes. Cell colonies were then counted and analysed.
Then, the sections were washed with TBS wash buffer and incubated with secondary antibody (goat anti-rabbit). The 14-3-3 zeta immunoreactivity was scored by three pathologists based on staining intensity and staining distribution. Scores 0, 1, 2 and 3 were for negative staining, faint yellow staining, sandy beige staining and dark brown staining, respectively. Meanwhile, scores 1, 2, 3 and 4 were for 0%-25% staining, 25%-50% staining, 50%-75% staining and 75%-100% staining, respectively. The final score was the sum of staining intensity and distribution score. were sacrificed and tumours were separated for further analyses.

| Statistical analysis
GraphPad Prism (Version 6.0) was utilised to perform the statistical analyses, in which data were documented as means ± standard deviation. Student′s t test or one-way ANOVA was applied to examine the statistical significance for the comparison of two or more groups with P < 0.05.

| Circ-SERPINE2 is upregulated in the microarray analysis of human GC
According to the GEO database of GC, two microarray chips, namely GSE78092 and GSE93541, were selected to filtered differentially expressed circRNAs. With the criteria of log|fold change| > 1 and adjusted P value < 0.05, top 100 differentially expressed circRNAs between the GC and the normal control samples in GSE78092 and in GSE93541 were displayed by heat map in Figure 1A, respectively.
The common differentially expressed circRNAs in GSE78092 and in GSE93541 were revealed using Venn diagram. Two circRNAs with expression upregulation in GC (namely hsa_circ_0008365/ circ-SERPINE2 and hsa_circ_0067127/circ-ALDH1L1) and the hsa_ circ_0000507 (circ-CUL4A) with expression downregulation in GC were found ( Figure 1B). The detailed log|fold change| and adjusted P value of three circRNAs in GSE78092 and GSE93541 are shown in Table S1. Among the 3 screened circRNAs, only circ-SERPINE2 is predicted to bind to AGO2 (Table S2)

| Circ-SERPINE2 is upregulated and can sponge miR-375 in GC tissues and cells
GC cell lines were used to perform the further experiments, including GES 1, HGSMC and gastric cancer cell lines (HGC-27, AZ521 and MGC-803). The qRT-PCR analysis showed that circ-SERPINE2 has higher expression in GC cells than in normal control cells (Figure 2A).
Clinical GC patients′ tissues and adjacent tissues confirmed circ-SERPINE2 was upregulated in GC ( Figure 2B). To verify the head-totail splicing, on the one hand, the divergent primers and convergent primers were designed to amplify circular RNA and linear RNA based on cDNA and gDNA from GC tissues. As shown in Figure 2C, circ-SERPINE2 was amplified in cDNA by two primers, but not in gDNA with the divergent primer. On the other hand, RNase R treatment was conducted. After RNase R treatment, the RNAs of circular form resisted to RNase R, but the linear RNAs was significantly reduced ( Figure 2D).
CircInteractome website was used to predict target miRNAs of circ-SERPINE2, and miRNAs related to GC reported in HMDD v3.0 were revealed. As shown in Figure 2E, the overlapping of foregoing two types of miRNAs contained hsa-miR-375, hsa-miR-194 and hsa-miR-203. To assess whether circ-SERPINE2 acts as a sponge for miR-NAs, we performed AGO2 immunoprecipitation using the lysates of

HGC-27 and MGC-803 cells. RNA immunoprecipitation (RIP) assays
showed that circ-SERPINE2 was enriched in AGO2 immunoprecipitation, suggesting that circ-SERPINE2 might have interactions with miRNAs ( Figure 2F). Then, a 3′-terminal-biotinylated-circ-SERPINE2 probe was designed to address which miRNAs latently interact with circ-SERPINE2. The probe was confirmed to pull-down circ-  was observed compared with the group of mutant miR-375 (biotin-miR-375-mut) ( Figure 2M). Furthermore, binding site between circ-SERPINE2 and miR-375 predicted in CircInteractome is shown in Figure 2N. In addition, dual-luciferase reporter assay further verified circ-SERPINE2 could directly bind to miR-375 ( Figure 2O). Above all, these findings indicated circ-SERPINE2 was upregulated and could sponge miR-375 in GC tissues and cells.

| Circ-SERPINE2 promotes GC cells development by sponging miR-375
QRT-PCR suggested circ-SERPINE2 expressions were effectively decreased by si-circ-1 not si-circ-2 and increased by ov-circ ( Figure 3A,B). To explore the effects of circ-SERPINE2 on proliferation and apoptosis, EdU assay ( Figure  Furthermore, si-YWHAZ repressed cell proliferation ( Figure 6A,B) and cell cycle progress ( Figure S3) and accelerated cell apoptosis

| D ISCUSS I ON
In the current study, we found that the downregulation of circ- CircRNAs may function as competing endogenous RNAs (ceR-NAs) that could sponge miRNAs to regulate miRNAs expression.
CircRNA ciRS-7 was reported to be obviously increased in GC and led to a more aggressive oncogenic phenotype by antagonising the miR-7-mediated PTEN-PI3K/AKT pathway. 22 The circHIPK3 was revealed to exert a negative regulatory effect on miR-124/miR-29b expression and related to Ming′s classification and T stage in GC. 23 Besides, circular RNA_LARP4 was suggested to restrain cell proliferation and invasion in GC. 4 Similarly, we found that the circ-SERPINE2 expression level was high in GC and improved tumour progress by sponging miR-375 and regulating YWHAZ expression.
miRNAs were believed to potentially act as valuable predictor for GC. 24  and breast cancer. 29 Specifically, YWHAZ was inhibited by miR-375 to suppress tumour properties, such as EMT, cell cycle, invasion and migration. 32 In addition, YWHAZ was reported as a promoter of cisplatin and paclitaxel resistance in ovarian cancer. 28  HW revising the work critically; All authors gave final approval of the work.

E TH I C A L A PPROVA L
All procedures performed in studies involving human participants were in accordance with the ethical standards of The Second Hospital of Shandong University committee. Informed consent was obtained from all individual participants included in the study. All procedures involving animals were performed in compliance with guidelines of The Second Hospital of Shandong University.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.