Circular RNA hsa_circ_0051246 acts as a microRNA-375 sponge to promote the progression of gastric cancer stem cells via YAP1

Microarray data analysis

As presented in our previous research (Liu et al., 2020), to explore circRNAs with potential regulatory effects, we screened and downloaded GC-related microarray data from the GEO database (http://ncbi.nlm.nih.gov/geo/), obtaining the circRNA expression microarray GSE83521 and the miRNA expression microarray GSE78091. Screening conditions were |log₂(FC)| ≥ 2.0 and P-value < 0.05. We identified 26 DE circRNAs and 151 DE miRNAs from the GEO database using the interactive online tool GEO2R. Subsequently, we used the circular RNA interactome website (https://circinteractome.nia.nih.gov/index.html) (Dudekula et al., 2016) to predict 196 miRNAs that may interact with 26 DE circRNAs. A Venn diagram (via OmicShare tools, https://www.omicshare.com/tools/Home/Soft/venn) was used to obtain the intersection of 151 DE miRNAs (from GSE78091) and 196 miRNAs (from predictions), and ten miRNAs were identified. Additionally, we used gene expression profiling interactive analysis (GEPIA) (Tang et al., 2017) to determine YAP1 expression in stomach adenocarcinoma.

Cell culture

SGC-7901 (C6795, Beyotime, China) and AGS (iCell-h016, iCell Bioscience, China) cell lines (human gastric adenocarcinoma cells), and the GES-1 (GES) cell line (iCell-h062, iCell Bioscience, China) (human gastric mucosal epithelial cells) were cultured in RPMI 1640 (R20161, Yuanye, China) containing 10% fetal bovine serum (FBS; 10099141C, Gibco, USA) in 10% CO₂ at 37 °C.

CSCs sorting and flow cytometry

SGC-7901 CSCs were obtained using immunomagnetic bead sorting (Dalerba et al., 2007). Human CD44 (20 µL; 130-095-194) and CD326 (EpCAM) (20 µL; 130-061-101) microbeads were added into 80 µL cell suspension (1 × 10⁷cells) and incubated in the dark for 15 min at 4 °C. Microbeads were purchased from Miltenyi Biotec (Bergisch Gladbach, Germany). Cells were then centrifuged (300 × g, 10 min) and resuspended using autoMACS running buffer (130-091-221; Miltenyi Biotec) to separate the CD44⁺ EpCAM⁺ SGC-7901 cells using a MidiMACS separator (Miltenyi Biotec, Bergisch Gladbach, Germany).

Flow cytometry was used to detect the expression of cell biomarkers. Phenol red and EDTA-free trypsin (15090046; Gibco, USA) were used to digest the cells. The centrifuged cells were resuspended and incubated with antibodies against CS44-FITC (ab30405; Abcam, Cambridge, UK) and EpCAM-PE (ab239311; Abcam, Cambridge, UK) antibodies in the dark and on ice for 30 min. The results were measured and analyzed using a flow cytometer (C6; BD Biosciences) and FlowJo10.0 software (BD Biosciences, Franklin Lakes, NJ, USA). Additionally, apoptosis was measured by flow cytometry using 1 × 10⁶cells/mL cells and the Annexin V-FITC kit (556547; BD Biosciences) as described by Bieszczad et al. (2020).

Cell transfection and grouping

Lentiviruses carrying short hairpin RNA (shRNA) were used to target circ_0051246 (sh-circ_0051246 #1, #2, and #3) and targeted YAP1 (sh-YAP1) and puromycin (1 µg/mL; ST551; Beyotime, Jiangsu, China) was administered for three weeks to establish knockdown cells. Additionally, plasmids carrying circ_0051246, YAP1 cDNA, hsa-miR-375-5p mimics, inhibitors, and agomir were used with Lipofectamine 3000 (Invitrogen, Waltham, MA, USA) for 48 h to establish cells with circ_0051246 and YAP1 overexpression, miR-375 activation, or suppression. The lentivirus, plasmid and their empty vectors were provided by Jikai Gene Chemical Technology Co., Ltd. (Shanghai, China).

First, SGC-7901 CSCs were divided into four groups, namely sh-NC, sh-circ_0051246, circ-NC, and circ_0051246, to explore the effects of circ_0051246 on CSCs in GC. Next, to explore the effects of miR-375 on GC and SGC-7901 GSCs, GSCs were divided into control, miR-375 mimic, miR-375 inhibitor, and circ_0051246 + miR-375 mimic groups. To explore the effects of YAP1 and the effect of miR-375 on YAP1, the CSCs were divided into the control, the YAP1, the sh-YAP1, and the miR-375 mimic + YAP1 groups. Finally, the SGC-7901 CSCs were divided into sh-NC, sh-circ_0051246, agomiR-NC, and agomiR-375 groups to establish a xenograft model.

Quantitative real-time PCR (qRT-PCR)

The supernatant from the lysed cell suspension was used to isolate RNA precipitates using chloroform and isopropanol successively. Subsequently, the precipitates were rinsed with 75% absolute ethanol. The precipitation was dissolved in 20 µL DEPC water (R0601; Thermo Fisher Scientific, Waltham, MA, USA) and standby. MiScript II RT Kit (218161; Qiagen, Hilden, Germany) and HiFiScript cDNA Synthesis Kit (CW2569; Cwbio, Beijing, China) were used for miRNA and mRNA reverse transcription, respectively. Finally, the SYBR Premix Ex Taq II (RR3090R; Takara, Shiga, Japan) was used for PCR amplification. GAPDH and U6 were used as internal references. All qRT-PCR experiments were independently repeated thrice. The relative quantitative method (2−^ΔΔCt) was used to analyze the data. Table 1 shows information regarding primers used.

Immunofluorescence (IF) staining

IF assays were performed on CSCs in 24-well plates, as previously described (Tang et al., 2020a). After 24 h, the cells were sequentially treated with 4% formaldehyde (1004965000, Sigma, Japan), 5% bovine serum albumin (ST023; Beyotime, China), and 0.5% Triton X-100 (X100; Sigma-Aldrich, St. Louis, MO, USA). Then, cells were incubated with anti-CD44 (ab254530, Abcam, UK) and anti-EpCAM antibodies (#46403; Cell Signaling, Danvers, MA, USA) at 4 °C overnight. The next day, cells were incubated with anti-mouse IgG H&L (ab150113; Abcam) antibody at 25 °C for 1 h. Finally, after incubation with DAPI (1:1000, BS097; Biosharp; Heifei city, China), the cells were observed under an inverted fluorescence inverted microscope (Ts2-FC, Nikon, Japan).

Cell counting kit-8 (CCK-8) assay

CCK-8 kits (HY-K0301; MedChemExpress, Monmouth Junction, NJ, USA) were used to measure cell viability in 96-well plates. After the cells were processed according to the experimental requirements, 5000 cells/well were incubated with CCK-8 reagent (10 µL/well) for 1 h. Optical density (OD) was measured using a microplate reader (450 nm) (CMaxPlus; Molecular Devices, San Jose, CA, USA). The value of the relative experimental OD/ relative control OD was used as the evaluation indicator. The relative OD value was calculated as the OD value minus that of the blank control.

Tumor spheroid experiment

Single-cell suspensions (500 cells) were cultured for seven days using 6-well plates with low-adhesion and the five mL of FBS-free culture medium. Tumor spheres were photographed under a microscope (Olympus, Tokyo, Japan) to count tumor sphere numbers.

Cell colony formation

CSCs (1000 cells) were treated according to grouping for 48 h and incubated in per-well of the 6-well plates with RPMI 1640 culture medium (30% FBS) for ten days. The cell culture medium was replaced with 4% paraformaldehyde to fix cells at 4 °C for 60 min. The cells were then washed and incubated with crystal violet (HY-B324A, MCE, USA) for 2 min.

5-Ethynyl-2′-deoxyuridine (EdU) assay

CSCs at 90% confluence (1 × 10⁶ cells/well) were cultured using 12-well plates. The cells were processed using the EdU cell proliferation kit (BeyoClick™, C0078s; Beyotime), observed, and photographed using an inverted fluorescence microscope.

Transwell assay

Matrigel was used to observe invasion, was diluted at a ratio of 3:1 in serum-free culture medium and was applied to a Transwell chamber (3422; Coring, Corning, NY, USA). The cell suspension (200 µL; 2 × 10⁵ cells) was placed in the upper chamber. After 24 h, the fixed cells were stained with a Crystal Violet solution (E607309; Sangon Biotech, Shanghai, China). Finally, the number of deep purple cells that invaded the matrigel was used as an evaluation indicator under a microscope. The migration assay was similar to the invasion assay except that Matrigel was not added.

Wound healing assays

In 6-well plates, CSCs (1 × 10⁵cells/mL) were cultured in 2% FBS until they reached 90% confluence. A 10 µL pipette tip was used to scratch cells in each well. After washing, the cells were cultured in serum-free medium. The scratch spacing was photographed under a microscope with the same eyepiece and objective lens at different time points (0, 24 and 48 h). Axio Vision (version 4.8; Zeiss, San Diego, CA, USA) was used to evaluate the migratory ability of CSCs. Data were analyzed in triplicate.

Dual-luciferase reporter assay

According to a previous study (Dai et al., 2021), Lipofectamine 3000 was used to transfect wild-type (WT) and mutant (Mut) recombinant plasmids (SV40-FIREFLY_Luciferase-MCS; Promega, Madison, WI, USA). SGC-7901 CSCs were divided into four groups, including circ_0051246 WT + miR-NC/miR-375 and circ_0051246 Mut + miR-NC/miR-375 groups, to explore the interaction between miR-375 and circ_0051246. After lysis, the cells were collected and processed using a dual-luciferase reporter gene assay kit (RG027, Beyotime, China). A microplate reader was used to measure the ratio of firefly luciferase activity to Renilla luciferase activity.

Fluorescence in situ hybridization (FISH)

Cy3-labeled probes for circ_0051246 and FAM-labeled probes for miR-375 were synthesized (Geneseed, Guangzhou, China). A FISH probe in a hybridization buffer (GenePharma, Shanghai, China) was used to incubate the CSCs for 16 h. Cell nuclei were stained with DAPI, and images were captured using a fluorescence microscope.

Orthotopic xenograft tumor model

The BALB/c male nude mice (5 weeks, 19 ± 2 g) (Beijing Vital River Laboratory Animal Technology Co., Ltd.) were used to establish the orthotopic xenograft tumor model. Mice were raised under pathogen-free conditions at 22–25 °C and kept on a 12-hour light-dark cycle. Mice had free access to standard laboratory feed and water. After one-week of adaptive feeding, the nude mice were anesthetized with isoflurane (4% for induction and 2% for maintenance). Transfected CSCs with green fluorescent protein (1 × 10⁶ cells/100 µL) were subcutaneously injected into the serosal layer of the greater curvature side of the stomach of the corresponding nude mice (Yin et al., 2019). According to the 3R principle, mice (n = 6 per group) were divided into sh-NC, sh-circ_0051246, AgomiR-NC, and AgomiR-375 groups using a random table method. Mice were euthanized using CO2 if they experience weight loss exceeding 20%, had tumors with diameters larger than 5% of their body weight, or if the tumor ruptured. Since none of these criteria were met during the experiment, no mice were euthanized in advance. D-luciferin was administered intraperitoneally at 28 days after tumor cell inoculation in mice. The mice were anesthetized with isoflurane and placed on an intravital fluorescence imaging system for small animals (AniView 100; BLT, Guangzhou, China) to capture the tumor situation and the luminescence information. All mice were euthanized using CO₂ (100% CO₂ at a filling rate of 30% per min). Tumors were removed, fixed, and embedded in paraffin. The mice exhibited no evident discomfort; therefore, supplementary analgesia was not required. No animals were excluded, and all animal experiments were performed by professionals blinded to the group assignment, according to the Guidelines for the Humane Treatment of Laboratory Animals. All animal experiments were performed in accordance with the Animal Experimentation Ethics Committee of Zhejiang Eyong Pharmaceutical Research and Development Center (ZJEY-20220301-02).

Hematoxylin-eosin (HE) staining and immunohistochemistry

Tumor tissue paraffin blocks were cut into 4 µm tumor paraffin slices for HE staining using an HE kit (ab245880; Abcam, Cambridge, UK). The sections were then dehydrated, made transparent, sealed, and observed under a microscope. The degree of inflammatory cell infiltration and the proportion of metastatic tumor cells are tumor histological evaluation indicators (Kajioka et al., 2021; Xiao et al., 2021). The higher the score, the more severe was the degree of infiltration and metastasis. After inactivation and antigen repair, the blocked-dewaxed slices were incubated with anti-Ki67 (1:200, ab16667; Abcam) and anti-YAP1 (1:500, ab205270, Abcam) antibodies at 37 °C for 2 h. Thereafter, the samples were treated with Rabbit IgG H&L antibody (1:5000, ab97080; Abcam) and incubated with hematoxylin (Bry-0001-01/03/04; Runnerbio, China) for 4 min at 25 °C. Sections were observed using a microscope (Eclipse E100; Nikon, Tokyo, Japan).

Western blot (WB)

The total protein from each group was diluted to the same concentration. Proteins were denatured, electrophoresed, and then transferred to polyvinylidene fluoride membranes (1620177; Bio-Rad, Hercules, CA, USA). Blots were blocked with 5% BSA and washed. The membrane was then incubated with the following antibodies overnight. The antibodies were anti-YAP1 (1:1000, ab205270; Abcam), anti-neurogenic locus notch homolog protein 1 (Notch 1) (1:1000, ab52627; Abcam), anti-jagged canonical notch ligand 1 (Jagged 1) (1:1000, ab7771, Abcam), anti-GAPDH (1:10000, AF7021; Affinity, Shanghai, China), anti-Ki67 (1:1000, ab16667, Abcam), anti-PCNA (1:2000, ab92552; Abcam, UK), anti-Bcl-2-associated X protein (Bax) (1:5000, ab32503; Abcam), anti-Bcl-2 (1:5000, ab32124, Abcam), anti-N-cadherin (1:5000, ab245117; Abcam), anti-Vimentin (1:1000; Abcam), and anti-E-cadherin (1:2000, ab40772; Abcam) antibodies. After washing, the membranes were incubated with anti-rabbit (7074, CST, USA) or anti-mouse (7076, CST) IgG HRP-linked antibodies. Proteins were visualized using a chemiluminescence imaging analysis system (610020-9Q; Qinxiang, China) and ImageJ software (NIH, Bethesda, MD, USA).

Statistical analysis

All statistical analyses were performed using SPSS Statistics (version 25.0.; IBM, USA) and Graphpad Prism 9.0 (GraphPad Software, La Jolla, CA, USA) was used to present the statistical analysis results. Multigroup data that were normally distributed and conformed to the homogeneity of variance tests were analyzed using the one-way analysis of variance (ANOVA) followed by the Turkey honest significant difference test, and two groups of data were analyzed using the t-test. The data were analyzed using a non-parametric Kruskal-Wallis rank sum test in cases where a normal distribution was not observed and the variance was not homogeneous. All data are expressed as the mean ± standard deviation (SD), P < 0.05 was used as the threshold for significant differences.

Circ_0051246 in SGC-7901 CSCs with a high expression level

Based on the analysis of GSE83521 and GSE78091 (Figs. 1A–1D), we have found the potential correlation between miR-375 and GCs, and circ_0051246 was predicted to be related to miR-375 (Liu et al., 2020). To the best of our knowledge, studies on the effects of circ_0051246 on CSCs derived from GCs are lacking. Therefore, we performed qRT-PCR to observe circ_0051246 expression levels and found that it had a higher expression level in SGC-7901 and AGS cells compared to human gastric mucosal epithelial cells (GES group) (P < 0.01) (Fig. 2A). After separating and collecting SGC-7901 CSCs (CD44⁺ EpCAM⁺ cells) using flow cytometry, we found that circ_0051246 was overexpressed in CSCs (P < 0.05) (Fig. 2B). The cell microstructure, actin in the cytoskeleton, and biomarkers of SGC-7901 CSCs were identified (Figs. 2C–2E). The cell structure was clear and actin was densely distributed in the cytoplasm. Additionally, CD44 and EpCAM expression were also observed on the cell membrane (Fig. 2E). We selected SGC-7901 CSCs for subsequent experiments.

Circ_0051246 promotes the progression of SGC-7901 CSCs

To observe the effects of circ_0051246 on SGC-7901 CSCs, we used the cells with sh-circ_0051246 #1, #2, #3, or circ_0051246 treatment. Circ_0051246 expression in SGC-7901 CSCs was reduced to 47% after treatment with sh-circ_0051246 #3 compared to that in the sh-NC group. Conversely, overexpression of circ_0051246 in CSCs led to a 72% increase in its expression (P < 0.01) (Fig. 3A). Accordingly, the sh-circ 0051246 #3 was used to establish circ_0051246 knockdown cells. We observed the proliferation and self-renewal of CSCs using CCK-8 assays, sphere formation, colony formation assays, and EdU cell proliferation imaging analysis. We found that circ_0051246 knockdown inhibited cell viability, the number of spheroids, colony number, and the proportion of EdU-positive cells (Figs. 3B, 3D–3F) (P < 0.05), whereas circ_0051246 overexpression increased the above levels in CSCs cells (P < 0.01) (Figs. 3C–3E and 3G).

Furthermore, we examined the role of circ_0051246 in CSC apoptosis using flow cytometry. Silencing circ_0051246 increased the apoptosis level and overexpression of circ_0051246 decreased apoptosis compared to their control cells (P < 0.05, P < 0.01) (Fig. 4A). Transwell and wound-healing assays were used to detect cell migration and invasion abilities. Circ_0051246 knockdown decreased migrated and invaded cell numbers and inhibited the migration distance percentage at 24 h and 48 h (P < 0.01) (Figs. 4B, 4D–4E). In contrast, circ_0051246 overexpression increased these indicators (P < 0.05) (Figs. 4C–4E).

Western blot was used to measure the proliferation-related proteins Ki67 and PCNA, the apoptosis-related Bax and Bcl-2, the migration and invasion-related proteins N-cadherin, Vimentin, and E-cadherin, and the YAP1, Notch 1, and Jagged 1 proteins levels (Figs. 4F–4H). The oncoprotein YAP1 is targeted by miR-375 and can be inactivated via the Hippo tumor suppressor pathway (Liu et al., 2022; Xu et al., 2019). Compared to their control cells, silencing circ_0051246 decreased Ki67, PCNA, Bcl-2, YAP1, Notch1, Jagged 1, N-cadherin, and Vimentin expression levels, but silenced circ_0051246 raised the Bax and E-cadherin levels (P < 0.05) (Fig. 4G). Conversely, circ_0051246 overexpression increased Ki67, PCNA, Bcl-2, YAP1, Notch1, Jagged 1, N-cadherin, and Vimentin expression levels and inhibited the Bax and E-cadherin levels (P < 0.05) (Fig. 4H). This suggests that circ_0051246 can regulate proliferation, apoptosis, migration, invasion, and the YAP1-related signaling pathway.

Circ_0051246 inhibited miR-375 in SGC-7901 CSCs

To explore the interaction between circ_0051246 and miR-375, we observed the co-localization of circ_0051246 and miR-375 using FISH (Fig. 5A). The interaction between circ_0051246 and miR-375 was assessed using a dual-luciferase reporter assay. The relative fluorescence intensity was lower in circ_0051246 cells treated with miR-375 treatment (P < 0.01) (Fig. 5B), suggesting that circ_0051246 act as a sponge for miR-375. Furthermore, the miR-375 levels were increased in circ_0051246 knockdown CSCs, whereas miR-375 levels were decreased in CSCs overexpressing circ_0051246 (P < 0.01) (Fig. 5C). Circ_0051246 treatment inhibited the miR-375 levels in miR-375 mimics-treated CSCs (P < 0.05) (Fig. 5D).

To determine whether circ_0051246 inhibits miR-375 to regulate the malignant phenotype of SGC-7901 CSCs, we performed CCK-8, sphere formation, colony formation assays, and EdU cell proliferation imaging analysis. Treatment with miR-375 mimics inhibited cell viability, number of spheroids, colony number, and EdU-positive cell proportion (P < 0.05) (Figs. 5E–5H), while miR-375 inhibitors increased these levels in CSCs (P < 0.01) (Figs. 5E–5H). MiR-375 mimics increased CSC apoptosis, whereas miR-375 inhibitors inhibited it (P < 0.05) (Fig. 5I). Circ_0051246 antagonized the effects of miR-375 mimics on the proliferation, self-renewal, and apoptosis of CSCs (Figs. 5E–5I).

In addition, treatment of miR-375 mimics on SGC-7901 CSCs decreased the number of migrated and invaded cells and decreased the migration distance at 24 h and 48 h (P < 0.05) (Figs. 6A–6B). In contrast, treatment with the miR-375 inhibitors increased these indicators (P < 0.01) (Figs. 6A–6B). Moreover, miR-375 mimics decreased the Ki67, PCNA, Bcl-2, N-cadherin, Vimentin, YAP1, Notch1, and Jagged 1 expression levels and increased Bax and E-cadherin protein levels (P < 0.05) (Figs. 6C–6F). MiR-375 inhibitors induced opposing effects on the abovementioned proteins in SGC-7901 CSCs were completely opposite (P < 0.05) (Figs. 6C–6F). Moreover, circ_0051246 antagonized the effects of miR-375 mimics on the migration and invasion of SGC-7901 CSCs (P < 0.05) (Figs. 6A–6B). In addition, circ_0051246 antagonized the regulation of miR-375 mimics on the expression levels of proteins associated with proliferation, apoptosis, migration, invasion, and the YAP1-related signaling pathway, suggesting that circ_0051246 acts as a sponge for miR-375 to promote the malignant phenotype of SGC-7901 CSCs.

MiR-375 suppresses the progression of SGC-7901 CSCs by inhibiting YAP1

We observed that circ_0051246 overexpression increased YAP1 mRNA level in SGC-7901 CSCs (P < 0.05), while circ_0051246 knockdown decreased YAP1 (P < 0.01) (Fig. 7A). In addition, miR-375 mimics inhibited YAP1 mRNA levels, whereas miR-375 inhibitors markedly increased YAP1 mRNA level (P < 0.05) (Fig. 7B). Circ_0051246 overexpression antagonized the inhibition of miR-375 mimics on YAP1 expression (P < 0.05) (Fig. 7B). We hypothesized that miR-375 acts directly on YAP1 expression. Therefore, we used the YAP1 overexpression and knockdown in SGC-7901 CSCs.

MiR-375 mimics inhibited YAP1 mRNA level in YAP1-overexpresed CSCs (Fig. 7C). A dual-luciferase reporter assay was used to observe the interaction between miR-375 and YAP1. SGC-7901 CSCs harboring a mutant YAP1 gene (YAP1 Mut) were used. The relative fluorescence intensity decreased in the YAP1 WT group with miR-375 mimic treatment (P < 0.01) (Fig. 7D).

The proliferation and self-renewal levels of YAP1-silenced and -overexpressed CSCs were measured using the CCK-8 assay, sphere formation, colony formation assay, and EdU cell proliferation imaging analysis. YAP1 overexpression increased cell viability, spheroid number, colony number, and EdU-positive cell proportion in CSCs (P < 0.05) (Figs. 6F–6L), while YAP1 silencing decreased the above levels (P < 0.05) (Figs. 6F–6L). YAP1 overexpression led to a decrease in CSC apoptosis, while YAP1 silencing increased it (P < 0.01) (Fig. 6M). Furthermore, miR-375 mimics antagonized the effects of YAP1 overexpression in CSCs (P < 0.05) (Figs. 6F–6M).

Overexpression of YAP1 heightened the number of migrated and invaded cells and led to a higher percentage of migration at both 24 h and 48 h (P < 0.05) (Figs. 8A and 8B). In contrast, YAP1 silencing decreased the levels of these indicators (P < 0.01) (Figs. 8A–8B). Moreover, YAP1 overexpression increased N-cadherin, vimentin, YAP1, Notch1, and Jagged 1 expression levels in CSCs and inhibited the E-cadherin expression (P < 0.05), while YAP1 silencing induced the opposite effect (P < 0.05) (Figs. 8C and 8D). Moreover, miR-375 mimics treatment antagonized the effects of YAP1 overexpression on SGC-7901 CSCs (P < 0.05) (Figs. 8A–8D).

Circ_0051246 knockdown and miR-375 activation inhibits the tumorigenicity of CSCs in vivo

An in vivo imaging system was used to observe the fluorescence signal intensity in an orthotopic xenograft tumor model with circ_0051246 knockdown or miR-375 activation. Circ_0051246-silenced and miR-375-activated cells inhibited the average signal strength of tumor fluorescence (P < 0.05) (Fig. 9A). Moreover, circ_0051246-silenced and miR-375-activated tumors had more pathological damage and inflammatory cell infiltration, and they had higher HE scores (P < 0.05) (Fig. 9B). In addition, Ki67 and YAP1 expression levels were observed using immunohistochemistry in orthotopic xenograft tumor mice. Circ_0051246-silenced and miR-375-activated cells showed decreased levels of Ki67 and YAP1 in the tumors (P < 0.01) (Figs. 9C and 9D). Circ_0051246-silenced tumors displayed lower circ_0051246 levels (P < 0.01), but miR-375-activated tumors did not show a significantly change in circ_0051246 levels (Fig. 9E). Additionally, silencing circ_0051246 and activating miR-375 markedly increased the miR-375 levels in the tumors (P < 0.01) (Fig. 9E). Furthermore, circ_0051246 knockdown and miR-375 activation inhibited YAP1 mRNA and protein expression (Figs. 9E–9G). Moreover, circ_0051246 knockdown and miR-375 activation inhibited Notch 1 and Jagged 1 protein levels in tumors (P < 0.01) (Figs. 9G and 9H). These results suggested that circ_0051246 inhibits miR-375 to promote the tumorigenicity of CSC cells in vivo.