Curcumin suppresses the progression of gastric cancer by regulating circ_0056618/miR-194-5p axis

Abstract Curcumin has been demonstrated to be an anti-tumor agent in many types of cancers, including gastric cancer (GC). However, the molecular mechanisms by which curcumin performs its anti-tumor effects remain elusive. circ_0056618 and miR-194-5p are reported to be involved in GC progression, but their relationships with curcumin are unclear. In this study, circ_0056618 was elevated, and miR-194-5p was reduced in GC tissues and cells. Curcumin treatment led to a decrease in circ_0056618 level in GC cells. Overexpression of circ_0056618 promoted cell proliferation, migration, and invasion and suppressed cell cycle arrest and apoptosis in curcumin-treated GC cells. Moreover, miR-194-5p was identified as the target of circ_0056618, and its expression in GC cells increased after curcumin treatment. Overexpression of miR-194-5p reversed the promotional effect of circ_0056618 on cell progression in curcumin-treated GC cells. Additionally, curcumin treatment repressed the tumorigenesis of GC in vivo through regulating circ_0056618. Curcumin treatment delayed the development of GC partly through decreasing circ_0056618 and increasing miR-194-5p.


Introduction
As a cancer with high incidence worldwide, gastric cancer (GC) severely affects the quality of people's life [1,2]. Presently, the primary therapeutic strategies for GC include surgery, chemotherapy, and targeted therapy [3]. Unfortunately, although diagnosis and therapy methods have been improved, the prognosis remains poor due to cancer relapse and metastasis [4,5]. Thus, there is an urgent need to search for novel treatment methods for this lethal disease.
Curcumin is a natural polyphenolic compound isolated from turmeric, which has lipid-lowering, anti-tumor, anti-inflammation, and anti-oxidation effects [6,7]. Recently, studies have shown that curcumin can reduce the malignancy of several cancers, such as pancreatic cancer [8], retinoblastoma [9], osteosarcoma [10], and bladder cancer [11]. Moreover, curcumin has also been reported to have an antitumor effect on GC [12,13]. Nonetheless, the role and underlying mechanism of curcumin in GC have not been well recognized.
circular RNAs (circRNAs) are a family of non-coding RNAs (ncRNAs) that possess closed-loop structures and play a vital role in gene expression via competitive targeting microRNAs (miRNAs) [14]. circRNAs have been verified to play essential roles in the carcinogenesis of GC. For example, circ_104916 could hamper the viability and metastasis of GC cells [15]. Furthermore, circDLST contributed to the metastasis and tumorigenesis of GC by interacting with miR-502-5p [16]. As for circ_0056618, Li et al. uncovered that circ_0056618 could accelerate the malignancy of GC by reducing miR-206 [17]. However, the relationship between circ_0056618 and curcumin is barely known.
miRNAs are short ncRNAs with ∼22 nucleotides and function as important modulators in various types of cancers, including GC [18,19]. It has been reported that miR-194-5p is downregulated in GC, and miR-194-5p overexpression plays a tumor-suppressive role in GC [20,21].
Furthermore, emerging evidence has reported that miRNAs are associated with the anti-tumor properties of curcumin in GC. For example, Sun et al. declared that curcumin repressed the progression of GC by elevating miR-34a [22]. Qiang et al. suggested that curcumin treatment induces GC cell apoptosis and hampers proliferation by modulation of miR-21 [23]. However, whether miR-194-5p is implicated in the anti-tumor effect of curcumin on GC remains unclear.
In this research, we explored the function of curcumin in circ_0056618 and miR-194-5p expression in GC and then elucidated the underlying mechanisms governing the tumor-inhibitory role of curcumin in GC.

Sample acquisition
Seventy-one GC tissues and adjacent non-tumor tissues were taken from Qilu Hospital (Qingdao), Cheeloo College of Medicine, Shandong University, and preserved at −80°C prior to use. The patients enrolled in our study did not have other malignant tumors, severe infection, and cognitive impairment history. None of the patients had received any chemotherapy or radiotherapy before this study. The demographic data of 71 patients with GC are shown in Table 1.
Informed consent: Informed consent was obtained from all the individuals included in this study.
Ethical approval: The research related to human use has complied with all relevant national regulations, institutional policies and is in accordance with the tenets of the Helsinki Declaration, and has been approved by the Ethics Committee of Qilu Hospital (Qingdao), Cheeloo College of Medicine, Shandong University.
Curcumin (Sigma-Aldrich, St. Louis, MO, USA) was dissolved in dimethyl sulfoxide (DMSO; Sigma-Aldrich) at a dose of 50 mM to make stock solution. The stock solution was diluted in the culture medium at a dose of 30 µM. For curcumin treatment, GC cells were subjected to 30 µM curcumin (Sigma-Aldrich) for 24 h in the presence of FBS (Procell). The control groups were subjected to DMSO (Sigma-Aldrich) [23].

Quantitative reverse transcription polymerase chain reaction (qRT-PCR) assay
The RNA in tissues and cells was extracted using TRIzol reagent (Invitrogen) and digested with DNAse I for 1 h at  Table 2.
The expression was calculated using the 2 −ΔΔCt method [24]. GAPDH and U6 were used as internal references.

Colony formation assay
GC cells (200 cells/well) were plated into 6-well plates. After adherence and curcumin exposure, the cells were cultivated in normal medium for about 2 weeks. The medium was changed every 3 days. Next, the colonies were fixed in 4% paraformaldehyde (Sangon, Shanghai, China) and dyed with 1% crystal violet (Sangon). The colonies containing >50 cells were counted. This experiment was conducted as previously described [9].

Flow cytometry analysis
For cell cycle analysis, GC cells (1 × 10 4 cells/well) were sown into 6-well plates. On the next day, the cells were subjected to 30 µM curcumin (Sigma-Aldrich) for 24 h. Next, the cells were harvested and fixed overnight with 70% ethanol. After that, the fixed cells were washed with PBS (Solarbio, Beijing, China) and resuspended in binding buffer at a concentration of 1.0 × 10 6 cells/mL followed by incubation with propidium iodide (PI; Beyotime, Shanghai, China) for 30 min at 37°C according to manufacturers' instructions. Finally, the stained cells were analyzed with a FACScan ® flow cytometer (Beckman Coulter, Atlanta, GA, USA). For cell apoptosis, the cells (1 × 10 4 cells/well) were plated into 6-well plates and then administered with 30 µM curcumin (Sigma-Aldrich) for 24 h. After that, the cells were collected, washed, resuspended, and then mixed with 5 μL of Annexin V-fluorescein isothiocyanate (Annexin V-FITC; Beyotime) and 5 μL of PI (Beyotime) for 20 min in the dark according to manufacturers' instructions. The apoptotic cells were estimated using a FACScan ® flow cytometer (Beckman Coulter).

Wound healing assay
GC cells were sowed into 6-well plates (2 × 10 3 cells/well) and grown until 100% confluence. Then the cells were scratched utilizing a pipette tip and washed with PBS (Solarbio) to remove the detached cells followed by treatment with curcumin for 24 h (Sigma-Aldrich). The scratched areas were photographed at 0 and 48 h at a magnification of 40× and the wound closure was recorded to assess cell migration capacity. The experiment was conducted as previously reported [25].

Transwell assay
The 24-well transwell inserts (Corning Incorporated, Corning, NY, USA) coated with or without Matrigel (Solarbio) were adopted to evaluate cell invasion and migration, respectively. Briefly, after indicated transfection and treatment, HGC-27 and MKN-45 cells (2 × 10 4 cells/well) were suspended into serum-free medium and then added into

Murine xenograft model
The BALB/c nude mice (4-6 weeks old) were obtained from Vital River Laboratory (Beijing, China) and divided into four groups (n = 6/group). For control and curcumin groups, 2 × 10 6 MKN-45 cells suspended in PBS (Solarbio) were subcutaneously implanted into the flank of the nude mice, and DMSO (Sigma-Aldrich) or 30 µM curcumin (Sigma-Aldrich) was intraperitoneally administered into the mice every 7 days. For curcumin + vector and curcumin + circ_0056618 groups, circ_0056618 or vector transfected MKN-45 cells were implanted into the nude mice and then 30 µM curcumin (Sigma-Aldrich) was intraperitoneally administered into the mice every 7 days. Tumor volume was examined every 7 days and computed via the equation: (length × width 2 )/2. The mice were euthanized after 28 days through cervical dislocation and the neoplasms were weighted and preserved for qRT-PCR assay.
Ethical approval: The research related to animal use has been complied with all the relevant national regulations and institutional policies for the care and use of animals and was approved by the Ethics Committee of Animal Research of Qilu Hospital (Qingdao), Cheeloo College of Medicine, Shandong University.

Statistical analysis
Each experiment was conducted in triplicate. The data were analyzed with GraphPad Prism 7 and exhibited as mean ± standard deviation. The differences in two groups and three groups were estimated using Student's t-test and one-way analysis of variance followed by Tukey's test, respectively. Pearson's correlation coefficient analysis estimated the linear correlation between circ_0056618 and miR-194-5p in GC tissues after Agostino-Pearson, Kolmogorov-Smirnov, and Shapiro-Wilk methods were used for the analysis of normality distribution of the data. Differences were considered as significant where P-value < 0.05 is represented as * and P-value < 0.01 is represented as **.

circ_0056618 was upregulated and miR-194-5p was downregulated in GC tissues
Initially, qRT-PCR assay was conducted to test the expression levels of circ_0056618 and miR-194-5p in GC tissues and adjacent non-tumor tissues. The results showed that circ_0056618 was highly expressed and miR-194-5p was lowly expressed in GC tissues compared to non-tumor tissues (Figure 1a and b). Pearson's correlation coefficient analysis estimated that miR-194-5p level was negatively correlated with circ_0056618 level in GC tissues (Figure 1c). These results indicated that circ_0056618 and miR-194-5p might be related to GC progression.
3.2 circ_0056618 overexpression ameliorated the effects of curcumin treatment on GC cell colony formation, cell cycle process, and apoptosis As shown in Figure 2a, circ_0056618 level was elevated in GC cell lines (HGC-27 and MKN-45) compared to that in GES-1 cell line. Then HGC-27 and MKN-45 cells were subjected to 30 µM curcumin for 24 h followed by qRT-PCR assay for circ_0056618 expression level. As a result, curcumin treatment led to an inhibition in circ_0056618 level in HGC-27 and MKN-45 cells compared to control groups, but curcumin did not affect the expression of circ_0056618 in GES-1 cells (Figure 2b). Next, the transfection of circ_0056618 increased the level of circ_0056618 in HGC-27 and MKN-45 cells, suggesting the successful transfection of circ_0056618 (Figure 2c). To explore the association between curcumin and circ_0056618 in the regulation of GC progression, HGC-27 and MKN-45 cells were treated with control, curcumin, curcumin + vector, or curcumin + circ_0056618. As shown in Figure 2d, curcumin-mediated downregulation on circ_0056618 expression was reversed by circ_0056618 overexpression vector transfection. The results of colony formation assay indicated that the colony formation ability of HGC-27 and MKN-45 cells was suppressed following curcumin exposure, while circ_0056618 overexpression overturned the effect (Figure 2e-g). Flow cytometry analysis showed that the proportion of HGC-27 and MKN-45 cells was increased in the G0/G1 phase and reduced in the S phase after curcumin treatment, whereas the impacts were reversed by elevating circ_0056618 (Figure 2h-k).
In addition, flow cytometry analysis also exhibited that curcumin treatment facilitated the apoptosis of HGC-27 and MKN-45 cells, while circ_0056618 overexpression abated this effect (Figure 2l). Collectively, circ_0056618 overexpression weakened the effects of curcumin on the malignant behaviors of GC cells.

Overexpression of circ_0056618 reversed the inhibitory effects of curcumin on GC cell migration and invasion
To investigate whether curcumin affected the migration and invasion of GC cells through regulating circ_0056618, wound healing assay and transwell assay were carried out. Wound healing assay exhibited that curcumin treatment suppressed the ability of wound closure, indicating that the migration ability of HGC-27 and MKN-45 cells was suppressed, which was rescued by increasing circ_0056618 (Figure 3a and b). Moreover, transwell assay showed that the migration and invasion capacities of HGC-27 and MKN-45 cells were repressed by curcumin exposure, while circ_0056618 overexpression abrogated the effects (Figure 3c-e). Taken together, curcumin restrained GC cell migration and invasion by modulation of circ_0056618.

circ_0056618 functioned as a sponge of miR-194-5p
Through analyzing online tool Circinteractome (https:// circinteractome.nia.nih.gov), we found that miR-194-5p contained the potential binding sites of circ_0056618, indicating that miR-194-5p might be a target of circ_ 0056618 ( Figure 4a). As exhibited in Figure 4b, miR-194-5p transfection increased the level of miR-194-5p in HGC-27 and MKN-45 cells compared to miR-NC control groups. Then dual-luciferase reporter assay and RNA pull-down assay were performed to verify the interaction between circ_0056618 and miR-194-5p. Dual-luciferase reporter assay presented that miR-194-5p transfection inhibited the luciferase activity of circ_0056618 WT, but had no effect on the luciferase activity of circ_0056618 MUT in both HGC-27 and MKN-45 cells (Figure 4c and d).  (Figure 4h and i). Additionally, our results exhibited that curcumin treatment increased miR-194-5p level in HGC-27 and MKN-45 cells, whereas circ_0056618 overexpression restored this effect (Figure 4j and k). All these results suggested that circ_0056618 directly interacted with miR-194-5p to negatively regulate miR-194-5p expression in GC cells.

Curcumin blocked the tumorigenesis of GC in vivo by regulating circ_0056618
At last, the function of curcumin in GC growth in vivo was determined through constructing the murine xenograft model. As shown in Figure 7a and b, tumor size and weight were restrained in curcumin groups compared to control groups, and tumor size and weight were increased in curcumin + circ_0056618 groups compared to curcumin + vector groups. Moreover, we found that circ_0056618 level was decreased and miR-194-5p level was increased in the tumor tissues collected from curcumin groups, while the effects were rescued in curcumin + circ_0056618 groups (Figure 7c and d). These results suggested that curcumin hampered tumor growth of GC in vivo by regulating circ_0056618.

Discussion
A growing body of evidence has demonstrated that curcumin restrains the malignant biological behaviors of GC cells through different pathways [26][27][28]. Herein, we discovered that curcumin was able to inhibit GC cell growth and motility and induce apoptosis by suppressing circ_ 0056618 and elevating miR-194-5p. Previous research have verified that curcumin exerts the anti-tumor effect mainly by repressing tumor cell proliferation and motility and facilitating apoptosis. For example, curcumin treatment restrained Rb cell proliferation, invasion, and migration and accelerated apoptosis [9].
Curcumin suppressed the growth and cell cycle process and facilitated the apoptosis of GC cells [22]. In line with these reports, we demonstrated that curcumin treatment restrained cell colony formation, migration, and invasion and accelerated cell cycle arrest and apoptosis in GC cells in vitro and blocked tumorigenesis of GC in vivo, indicating that curcumin might be a candidate agent for GC therapy.
Zheng et al. suggested that circ_0056618 level was enhanced in colorectal cancer (CRC) and promoted CRC cell growth, angiogenesis, and migration by reducing miR-206 and elevating CXCR4 and VEGF-A [29]. Li et al. unraveled that circ_0056618 was elevated in GC and facilitated cell proliferation and motility in GC [17]. In this study, we also observed that there was an increase in circ_0056618 level in GC tissues and cells. Moreover, we found that curcumin treatment predominantly decreased the expression of circ_0056618 in GC cells. Thus, we further explored whether circ_0056618 could participate in regulating curcumin-mediated malignant phenotypes Figure 6: Curcumin altered the expression of CyclinD1, E-cadherin, and N-cadherin in GC cells by modulation of circ_0056618/miR-194-5p axis. (a and b) After HGC-27 and MKN-45 cells were treated with control, curcumin, curcumin + vector, curcumin + circ_0056618, curcumin + circ_0056618 + miR-NC, or curcumin + circ_0056618 + miR-194-5p, the protein levels of CyclinD1, E-cadherin, and N-cadherin were measured by western blot assay. **P < 0.01. of GC cells. As a result, overexpression of circ_0056618 ameliorated curcumin-mediated anti-proliferation, antimetastasis, and pro-apoptosis effects on GC cells, suggesting that curcumin could decelerate GC development by reducing circ_0056618 expression.
Subsequently, the underlying mechanism of how curcumin regulates GC progression was further investigated. miR-194-5p has been demonstrated to be the target of several circRNAs, such as circ_0023028 [30], circ-USP1 [31], and circ_0001971 [32]. Through bioinformatics analysis, dual-luciferase reporter assay and RNA pull-down assay, miR-194-5p was verified to be the target of circ_ 0056618 for the first time. Our results showed that miR-194-5p was weakly expressed in GC tissues and cells, and circ_0056618 could negatively regulate miR-194-5p expression in GC cells and curcumin-treated GC cells. Previous studies showed that miR-194-5p had the capacity to repress GC cell growth and metastasis [33,34]. In this study, our results exhibited that the elevation of miR-194-5p effectively reversed the impacts of circ_0056618 on cell growth, apoptosis, and metastasis in curcumin-treated GC cells, indicating that circ_0056618 promoted curcumin-mediated GC development by interacting with miR-194-5p.
In summary, curcumin treatment could repress GC cell growth and metastasis and promote apoptosis partly by regulation of circ_0056618/miR-194-5p axis. The findings facilitated our understanding on the mechanism of curcumin in GC therapy and indicated that curcumin might be a potential therapeutic drug for GC. In addition, accumulating evidence showed that curcumin might prevent GC through regulation of oncogenic pathways [35]. However, the underlying mechanisms still need further investigation. Funding information: The authors state no funding involved.

Conflict of interest:
The authors state no conflict of interest.
Data availability statement: The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.