Magnesium Isoglycyrrhizinate Suppresses the Progression of Bladder Cancer by Modulating miR-26b/Nox4 Axis

Background: Magnesium Isoglycyrrhizinate (MI), a magnesium salt of 18α-GA stereoisomer, has been reported to exert ecient hepatoprotective activity. However, its effect in bladder cancer remain unclear. Methods: The effect of MI in the growth, colony formation, apoptosis, invasion and migration of bladder cancer cells (HTB9 and BIU87 cells) was evaluated in vitro. Typical apoptotic changes of bladder cancer cells such as nuclear concentration and fragmentation was observed by Hoechst staining. The effect of MI in the expression of miR-26b, Nox4, NF-κB and HIF-1α was detected by qRT-PCR and western blot in vitro. Targetscan was used to predict the potential targets of miR-26b, then their interaction was determined by the luciferase reporter assay. Finally, the xenograft model of mice was established to evaluate the anti-tumor effects of MI in vivo. Results: MI signicantly suppressed the proliferation, colony formation, invasion, migration and induced apoptosis of human bladder cancer cells. Nox4 was identied to be a direct target of miR-26b, and MI signicantly increased the expression of miR-26b. MiR-26b mimics signicantly decreased the relative luciferase activity of wild type (WT) Nox4, while exhibited no obvious change in mutant type (MUT) Nox4. Meanwhile, MI markedly downregulated the expression of Nox4, NF-κB and HIF-1α both in vitro and in vivo. Moreover, MI could eciently inhibit the growth of xenograft tumor in vivo, and also obviously decreased the expression of Nox4, NF-κB and HIF-1α. Conclusion: In summary, MI showed a potent anti-tumor effect against bladder cancer partially through modulating miR-26b/Nox4 axis.


Background
Bladder cancer has becoming a common cancer worldwide with higher morbidity and mortality, and also with an estimated 430 000 new cases diagnosed every year 1, 2 . The incidence of bladder cancer is higher in old men than young, suggesting the geographic variation and is mostly an environmental disease 3 .
Even worse, there are approximately one-third of patients with muscle-invasive bladder cancer already have undetected metastases at the time of treatment 4 . Although considerable progress has been achieved in surgical techniques and adjuvant chemotherapies, the mortality of bladder cancer is still higher 5 . A large number of previous studies have demonstrated that the occurrence and development of bladder cancer is a complex process, which can be caused by abnormal genetic changes or epigenetic abnormalities 6 . Therefore, a better understanding of speci c mechanism in the progression of bladder cancer contributes to identify e cient drug targets and even expends the window period of treatment in bladder cancer.
Magnesium isoglycyrrhizinate (MI), a magnesium salt of 18α-GA stereoisomer, has been known as an e cient hepatoprotective agent [7][8][9] . Expect for protective effect on the liver, MI has been identi ed to exert potential protective effect in various human diseases. For instance, MI can prevent drug-induced liver damage after the initial chemotherapy for those patients with early stage gastrointestinal cancer 10 . One previous study reported that the liver toxicities induced by paclitaxel plus cisplatin chemotherapy can potentially decrease the ability of hepatic elimination and increase system exposure of paclitaxel, while MI e ciently help to restore hepatic clearance of paclitaxel 11 . MI can protect against renal-ischemia-reperfusion injury in a rat model through anti-in ammation, anti-oxidation and anti-apoptosis capacity 12 . MI has been demonstrated to inhibit the myocardial hypertrophy by inactivating the TLR4/NF-κB signaling pathway in mice 13 . In addition, MI has also reported that can ameliorate fructose-induced podocyte apoptosis by downregulating the expression of miR-193a to upregulate Wilms' tumor protein (WT1) expression 14 . However, the effect of MI in bladder cancer has not been studied.
MicroRNAs (miRNAs) are a group of small and non-coding RNAs with approximately 22 nucleotides in length, and often regulate gene expression post-transcriptionally though binding to the 3'-untranslated region (3'-UTR) of the target mRNAs 15 . Recently, miR-26b has been a focus of interest for its role as a tumor suppressor in several cancer types. In colorectal cancer, upregulation of miR-26b promotes chemosensitivity of cancer cells through targeting P-glycoprotein (Pgp) 16 . In esophageal squamous cancer, miR-26b can inhibit the cell proliferation by suppressing c-MYC signaling pathway, and suggested that miR-26b might be a potential target of prevention and treatment of esophageal squamous cancer 17 . In lung cancer, miR-26b has been identi ed that can inhibit the invasion and migration of cancer cells by directly targeting hENT1 depending on RhoA/ROCK-1 signaling pathway 18 . Although previous studies have revealed the crucial anti-tumor effect in various cancers, its function in bladder cancer remains unclear. NAD(P)H oxidase 4 (Nox4), a substrate of NADPH that can generate H 2 O 2 reactive oxygen species, has been reported to be highly expressed in human tumors 19 . NOX4 can support glycolysis and promote glutamine metabolism in non-small cell lung cancer (NSCLC) cells 20 . NOX4-driven ROS formation has been found to regulate the proliferation and apoptosis of gastric cancer cells via the GLI1 pathway 21 . In addition, Nox4 has been identi ed to be closely correlated with gastric cancer progression and predicts a poor prognosis 22 . Although Nox4 also play essential roles in the development of bladder cancer 23 , the regulatory axis of Nox4 in bladder cancer has not been well studied.
In this study, we explore the role and molecular mechanisms of MI in bladder cancer, and rstly reported that MI could e ciently inhibit the progression of bladder cancer both in vitro and in vivo, speci cally, MI could suppress the activation of Nox4/ NF-κB/HIF-1α signaling pathway through upregulating the expression of miR-26b. Our results indicated that MI might a potential anti-tumor agent against bladder cancer.

Materials
Magnesium isoglycyrrhizinate (MI) (purity > 98%, monohydrate) was obtained from Zhengda Tianqing Pharmaceutical Co., Ltd (Jiangsu, China). And all of other chemicals and solvents were commercially available. The structural formula of MI was shown in Fig. 1.

Cell culture
Human bladder cancer cell lines HTB9 and BIU87 cells were purchased from the American Type Tissue Culture Collection (ATCC), and cultured in RPMI-1640 medium (Solarbio, Beijing, China) containing 10% FBS (Gibco BRL, Grand Island, NY, USA), 100 unit/mL penicillin and 0.1 mg/mL streptomycin at 37 °C with 5% CO 2 .

Cell transfection
Cell transfection was performed by using the Lipofectamine RNAi MAX kit (invitrogen) according to the instructions. Of which, miR-26b mimics, miR-26b inhibitor and corresponding negative controls (miR-NC and inhibitor NC) were purchased from Sigma Aldrich.

Luciferase reporter assay
The fragments of wild type (WT) or mutant type (MUT) 3'UTR of Nox4 was ampli ed from mouse cDNA and cloned into pRL reporter plasmids. HTB9 and BIU87 cells were co-transfected with luciferase reporter plasmid containing WT or MUT 3'UTR of Nox4 and miR-26b inhibitor or inhibitor NC by using the Lipofectamine RNAi MAX kit. After transfection for 48 h, cells were collected, lysed, and the relative luciferase activity was detected by Dual luciferase reporter system.

MTT assay
Cell viability was evaluated using the MTT assay as previously described 24 . Brie y, approximately 2×10 5 HTB9 or BIU87 cells were deeded into 96-well plates and cultured for 24 h. Then cells were treated with different concentrations of MI such as 0, 1. 2, 5, 10 and 20mg/ml for 24 h. Finally, 10μL MTT solution (5mg/mL, Sigma) was added to each well and incubated for another 5h. After medium was aspirated, DMSO (150 μL; Sigma) was added to dissolve formazan crystals and the absorbance at 490 nm was detected using a microplate reader (Biotek, SYNERGY HTX, VT, USA).

Colony formation assay
Colony formation assay was performed as previously described 6 . In brief, HTB9 and BIU87 cells treated with or without 3.82mg/ml or 2.85 MI were seeded into 6-well plates at a density of 1,000 cells per well and cultured for two weeks. After washing with PBS for twice, cells were xed with methanol, stained with 0.1% crystal violet for 25 min, and then the colonies were imaged and counted.
Hoechst 33258 staining assay The changes in cellular morphology of HTB9 and BIU87 cells were observed by using Hoechst 33258 staining (Sigma) as previously described 25 . Brie y, HTB9 and BIU87 cells treated with or without 3.82mg/ml or 2.85 MI were seeded into 6-well plates and cultured overnight. Then cells were xed with 4% formaldehyde for 15 min, and stained in Hoechst 33258 (10mg/ L) for another 1 h. After washing with PBS for twice, cells were subjected to uorescence microscopy (Olympus, Tokyo, Japan). Meanwhile, the morphological changes including reduction in the volume and nuclear chromatin condensation were observed.

Transwell assay
The invasion and migration capacities of HTB9 and BIU87 cells was evaluated by transwell assay by using 24-well Transwell™ plates with or without 8.0-μm-pore Matrigel™-coated membranes previously described 26 . Brie y, approximately 1 × 10 5 cells in serum-free medium were seeded into the upper chamber, and the lower chamber was lled with medium containing 20% FBS. The migration assay was performed similarly without coating the membranes with Matrigel™. After the incubation for 24 h, cells were xed by 4% paraformaldehyde and stained with 0.1% crystal violet. Finally, the numbers of invaded and migrated cells were counted in ve randomly selected elds by a microscope.
qRT-PCR Total RNA of cultured cells was extracted by using the TRIzol reagent (Invitrogen) according to the manufacturer's instructions. Approximately 1.2 ug RNA was reversely transcribed into cDNA by using the Prime Script RT Master Mix (TaKaRa, Japan) and the qRT-PCR analysis was performed by the SYBR Premix Ex Taq II kit (Takara, Otsu, Japan) based on a Quantstudio™ DX system (Applied Biosystems, Singapore). The relative expression change of targets was analyzed by 2 -∆∆CT method 27 . With GAPDH and U6 as considered as the internal reference. The primers used in this study as follows:

Flow cytometer assay
Cell apoptosis was evaluated by using the Annexin V-FITC/PI apoptosis detection kit (Keygen Biotech, Nanjing, China) according to the manufacturer's instructions. Brie y, cells treated with or without 3.82mg/ml or 2.85 MI were seeded into 6-well plates at a density of approximately 1×10 5 cells/ml. Then cells were collected, harvested, washed and re-suspended in binding buffer. After stained with Annexin V-FITC at room temperature for 20 min in dark, PI was added into each well and incubated for another 10 min. The apoptosis rate of cells was analyzed by a FACSCalibur ow cytometer within 1 h after supravital staining.

Animal model
The female BALB/c nude mice (6-week-old, weighing approximately 20.0 ± 2.0 g) were purchased from The First A liated Hospital of Zhengzhou University. Mice were kept at room temperature with 12/12-h light-dark cycle, and all animal handling procedures were performed in strict according to the PR China legislation of the use and care of laboratory animals. This study was approved by The First A liated Hospital of Zhengzhou University HTB9 tumor-bearing mice model was established as previously described with minor modulation28. In brief, HTB9 cells were intraperitoneally injected into the abdominal cavity of BALB/c nude mice, and the ascites were removed from mice and diluted to 1×10 7 cells/ml with physiological saline. Then 0.2 mL of the cells were subcutaneously injected into the right armpit of mice to establish the solid tumor model. To determine the effect of MI in vivo, mice were randomly divided into two groups (n = 10), model group and MI group. MI groups were intragastrically given the certain dose (0.1 mL/10 g body weight.) of MI once daily for 5 weeks, and model groups were given the same volume of normal saline, simultaneously. Tumor volume was calculated by using the following formula: (length× width 2 )/2. When experiments were nished, mice were sacri ced, and the excised tumors were removed, weighted, and photographed.

Statistical analysis
All data were presented as the mean ± SE method which was derived from at three independent experiments. The difference between groups was determined by Two-tailed Student's t-test or one-way ANOVA. Statistical analysis was performed by the SPSS software v18.0 with p < 0.05 as the signi cant threshold.

MI inhibited the growth and colony formation of bladder cancer cells in vitro
To explore the effect of MI in bladder cancer, bladder cancer cell lines HTB9 and BIU87 cells were treated with different concentrations of MI such as 0, 1. 2, 5, 10 and 20 mg/ml for 24 h. MTT assay indicated that MI obviously decreased the cell viability in a dose dependent manner and exhibited the signi cant decrease at 5, 10 and 20 mg/ml both in HTB9 and BIU87 cells (p < 0.01) ( Fig. 2A). Further, The IC50 value calculated by SPSS software was 3.82 mg/mL in HTB9 cells and 2.85 mg/mL in BIU87 cells, which was selected for the subsequent experiments. Then the colony formation assay was performed and the results showed that MI signi cantly inhibited the colony formation both in HTB9 (p < 0.01) and BIU87 cells (p < 0.01). These results suggested that MI could e ciently inhibit the growth of bladder cancer cells, and had high selectivity to tumor cells in vitro.

MI promoted the apoptosis of bladder cancer cells in vitro
Further, the apoptosis of bladder cancer cell lines HTB9 and BIU87 cells was observed by Hoechst 33258 staining assay (Fig. 3). From the morphological changes of cell apoptosis, the condensation of chromosome and nuclear fragmentation were obviously observed both in HTB9 (Fig. 3A) and BIU87 cells (Fig. 3B) when cells exposed to MI, and apoptotic cells were approximately between 50-80%. These results suggested that MI could signi cantly promote the condensation of chromosome and nuclear fragmentation in bladder cancer cells in vitro.

MI inhibited the invasion, migration and induced apoptosis of bladder cancer cells in vitro
Next, transwell assay was performed to explore the effect of MI on cell invasion and migration, and the results indicated that MI signi cantly inhibited the migration capacity of both HTB9 (p < 0.01) and BIU87 cells (p < 0.01) (Fig. 4A). Similarly, MI also markedly suppressed the invasion capacity of both HTB9 (p < 0.01) and BIU87 cells (p < 0.01) (Fig. 4B). Meanwhile, the effect of MI in the apoptosis rate of HTB9 and BIU87 cells was evaluated by ow cytometry, and the results revealed that MI signi cantly promote the apoptosis of both HTB9 (p < 0.01) and BIU87 cells (p < 0.01) (Fig. 4C). These results suggested that MI could e ciently inhibit the invasion, migration and induce apoptosis of bladder cancer cells in vitro.

Nox4 Was A Direct Target Of Mir-26b
Interestingly, we found that MI signi cantly increased the expression of miR-26b both in HTB9 (p < 0.01) and BIU87 cells (p < 0.01) (Fig. 6A). To explore the speci c mechanisms of miR-26b in response to MI, Targetscan was used to predict the potential targets of miR-26b, and the results showed that there was a putative binding site between miR-26b and Nox4 (Fig. 6B), suggesting that Nox4 might be a target of miR-26b. Then luciferase reporter assay was performed and revealed that miR-26b mimics signi cantly decreased the relative luciferase of Nox4 WT vector compared with miR-NC control (p < 0.01), while exhibited no obvious change in Nox4 MUT vector (Fig. 6C). Meanwhile, miR-26b inhibitor signi cantly increased the relative luciferase of Nox4 WT vector compared with inhibitor NC control (p < 0.01), while showed no obvious change in Nox4 MUT vector (Fig. 6D). In addition, miR-26b mimics markedly decreased the protein expression of Nox4 compared with miR-NC control (p < 0.01), while miR-26 inhibitor signi cantly increased the protein level of Nox4 compared with inhibitor NC control (Fig. 6E). These results indicated that Nox4 was a target of miR-26b, and the effect of MI in bladder cancer was mediated by miR-26b/Nox4 axis.

MI e ciently repressed the progression of bladder cancer in vivo
Finally, the xenograft rat model was established to determine the anticancer effect of MI in bladder cancer. The representative images of xenograft tumors from model group and MI group was shown in Fig. 7A. As expected, the tumor weight was signi cantly decreased in MI group compared with model group (p < 0.01) (Fig. 7B). Meanwhile, the tumor volume was obviously decreased in a time dependent manner in MI group compared with model group (p < 0.01) (Fig. 7C). In addition, Ki-67 staining assay showed that the numbers of positive cells in MI group were markedly decreased compared with model group (Fig. 7D), suggesting that MI e ciently inhibited the proliferation of bladder cancer cells. Moreover, compared with model group, the protein level of Nox4 (p < 0.01), NF-κB (p < 0.01) and HIF-1α (p < 0.01) was signi cantly decreased compared with model group (Fig. 7E). These results demonstrated that MI could e ciently repress the progression of bladder cancer in vivo.

Discussion
Due to its higher recurrence rate, bladder cancer brings to patients physical agony and high therapy costs to the patients' family and society 29,30 . Hence, a lot of attentions have focused on the diagnosis or treatment in bladder cancer 31 . In the last decades, a series of natural or unnatural components have been identi ed and demonstrated to show a certain anti-tumor activity for bladder cancer. For example, sulforaphane, a natural agent that was abundant in cruciferous vegetables, has been demonstrated to suppress the proliferation of non-muscle invasive bladder cancer cells via blocking HIF-1α-mediated glycolysis in hypoxia 32 . Inoue et al. reported a 5-Aminolevulinic acid-mediated photodynamic therapy for bladder cancer, and some favorable outcomes have been achieved clinically 33 . It was also reported that curcumin, a yellow substance belonging to the polyphenols superfamily, has been revealed that can attenuate the progression of bladder cancer partially through suppressing the Sp-1 activity 34 . In addition, amygdalin, a natural compound, has been demonstrated to e ciently inhibit the growth of bladder cancer cells in vitro through diminishing cyclin A and cdk2 35 . Despite many anti-tumor agents for bladder cancer have been identi ed, the identi cation of speci c and e cient anti-bladder cancer agents is still urgent. In this study, we explore the effect of a hepatoprotective agent MI in bladder cancer, and revealed that MI could signi cantly inhibit the growth, invasion, migration and induce apoptosis of bladder cancer cell lines including HTB9 and BIU87 cells in vitro, suggesting a protective role of MI in bladder cancer development.
MiRNAs have been identi ed to play essential roles in tumorigenesis and metastasis through negatively regulating gene expression by directly targeting the 3'-UTR in bladder cancer. MiR-217 has been reported to inhibit the proliferation and migration of bladder cancer cells through targeting KMT2D 36 . MiR-556-3p can exacerbate the capacity of proliferation, migration and invasion of human bladder cancer cells via negatively regulating the expression of DAB2IP 37 . Xie et al. have revealed that downregulation of miR-532-5p can signi cantly promote the proliferation and invasion of bladder cancer cells by activating the HMGB3/Wnt/β-catenin signaling pathway 38 . MiR-124-3p has also been found that can suppress cell migration and invasion of bladder cancer cells through targeting ITGA3 39 . MiR-129-5p suppresses gemcitabine resistance and promotes the apoptosis of bladder cancer cells via targeting Wnt5a 40 . In addition, there are many other miRNAs including miR-223 41 , miR-1265 42 , miR-125a-5p 43 , miR-155 44 , and so on which were signi cantly involved in the progression of bladder cancer. In the present study, we found that MI could signi cantly increase the expression of miR-26b in bladder cancer cell lines such as HTB9 and BIU87 cells, suggesting that the protective effect of MI in bladder cancer was mediated by miR-26b.
To explore the speci c mechanisms of miR-26b induced by MI in bladder cancer, targetscan was applied to predict the potential targets of miR-26b and found that Nox4 might a direct target of miR-26b. Meitzler et al. found that the expression of Nox4 was signi cantly upregulated in the carcinoma of bladder cancer patients compared with normal controls 45 . Recently, NAD(P)H oxidase 4 (Nox4) have been revealed to play important roles in the invasion of bladder cancer cells 46 . One previous study reported that knockdown of NOx4 exhibits a signi cant inhibitory effect on survival, and induces apoptosis of bladder cancer cells 47 . These reports all con rmed a tumorigenic role of Nox4 in bladder cancer. To determine the interaction between miR-26b and Nox4, the luciferase reporter assay was performed and showed that miR-26b mimics signi cantly decreased the relative luciferase of Nox4 WT vector, while miR-26b inhibitor signi cantly increased the relative luciferase of Nox4 WT vector. Meanwhile, downregulation of miR-26b signi cantly increased the expression of Nox4, and upregulation of miR-26b reversely decreased Nox4 expression. In addition, MI could obviously decrease the expression of Nox4 both in vitro and in vivo. These results demonstrated that Nox4 was a direct target of miR-26, and the effect of MI in bladder cancer was mediated by miR-26b/Nox4 axis.
Cell proliferation is one of the main features of solid tumor progression, and the rapid tumor cell growth usually results in hypoxia because of the low oxygen environment 32 . A previous study indicated that one of the key factors regulating the response to hypoxia is the heterodimer hypoxia-inducible factor-1α (HIF-1α) 48 . A series of previous studies have demonstrated that HIF-1α can exacerbate the progression of bladder cancer including promoting EMT process 49 , growth 50 , and conferring the chemo-resistance to cisplatin of bladder cancer cells 51 . It has been reported that NF-κB inhibits the apoptosis and promotes the proliferation of bladder cancer cells through upregulating the expression of survivin both in vitro and in vivo 52 . Here, our results showed that MI could markedly decreased the expression of NF-κB and HIF-1α both in in vitro and in vivo. Moreover, the experiments in vivo demonstrated that MI could e ciently inhibit the development of bladder cancer, suggesting that MI might be a potential anti-tumor agent for bladder cancer. However, there was few limit in this study: although our results suggested that MI inhibited the progression of bladder cancer through suppressing Nox4/NF-κB/HIF-1α signaling, the Nox4/NF-κB/HIF-1α signaling-related inhibitor (upregulation of Nox4, NF-κB or HIF-1α) whether reversed the protective effect of MI in bladder cancer needed to be determined in the further.

Conclusion
In summary, MI could e ciently inhibit the progression of bladder cancer both in in vitro and in vivo through inhibiting Nox4/NF-κB/HIF-1α signaling pathway by directly upregulating the expression of miR-26b, suggesting that MI might be a potential and safe anticancer bioactive agent for bladder cancer.

Declarations
Ethics approval and consent to participate The present study was approved by The First A liated Hospital of Zhengzhou University. The research has been carried out in accordance with the World Medical Association Declaration of Helsinki. All patients and healthy volunteers provided written informed consent prior to their inclusion within the study.

Consent for publication
All authors have read and approved the nal manuscript.

Availability of data and materials
The analyzed data sets generated during the study are available from the corresponding author on reasonable request.

Competing interests
There was no any con ict of interest.   (B) HTBP cells were treated with 3.82 mg/mL MI and BIU87 cells were treated with 2.85 mg/mL MI, and the cell growth was evaluated by colony formation assay. N = 6, ** P < 0.01.    Nox4 was a target of miR-26b. (A) HTB9 were treated with 3.82mg/mL and BIU87 cells were treated with 2.85mg/mL. The mRNA level of miR-26b was evaluated by qRT-PCR. (B) The interaction between miR-26b and Nox4 was predicted by Targetscan. (C and D) HTB9 cells were co-transfected with luciferase reporter plasmids containing WT or MUT Nox4 and miR-26b mimics/miR/NC (C), or miR-26b inhibitor/inhibitor NC (D), and the relative luciferase activity was detected by dual luciferase reporter system. (E) HTB9 cells were co-transfected with miR-26b mimics, miR-NC, miR-26 inhibitor, inhibitor NC, and the protein level of Nox4 was evaluated by western blot. N = 6, ** P < 0.01. Full-length blots/gels are presented in Supplementary

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. originalwesternblot.docx