Glucose promotes epithelial‐mesenchymal transitions in bladder cancer by regulating the functions of YAP1 and TAZ

Abstract Glucose levels and type 2 diabetes (T2D) are both associated with tumorigenesis and epithelial‐mesenchymal transitions (EMTs). EMTs facilitate bladder cancer (BC) metastasis development, but the mechanism by which high‐glucose levels promote these EMTs in BC remains unclear. Therefore, we sought to elucidate the mechanism underlying EMT promotion due to increased glucose levels. T24 and UMUC‐3 cells were cultured in media containing different glucose concentrations. YAP1, TAZ, GLUT1 and EMT‐associated marker expression was analysed via Western blotting and qPCR. BC cell proliferation and invasion were assessed using MTT and Transwell assays, respectively. A xenograft nude mouse model of diabetes was used to evaluate tumour growth and metastasis in vivo. T2D was positively associated with pathologic grade (P = .016) and TNM stage (P < .001) in BC. High glucose triggered BC cell proliferation and invasion in both in vitro and in vivo conditions. High‐glucose levels also promoted EMTs in BC cells and increased YAP1 and TAZ expression. YAP1 or TAZ knockdown altered EMT marker expression and decreased GLUT1 expression. Overall, our results suggest that high‐glucose levels promote EMTs in BC cells via YAP1 and TAZ regulation. These effector molecules may be promising therapeutic targets for BC cases comorbid with T2D.


| BACKG ROU N D
Bladder cancer (BC) is the fourth most common cancer among men and the second most common cause of death among urinary tract malignancies in the United States. 1 BC is a malignant disease with high morbidity and mortality and can be divided into two categories: nonmuscle-invasive bladder cancer (NMIBC) and muscle-invasive bladder cancer (MIBC). Nearly 75%-80% of newly diagnosed cases of BC are NMIBC (stages Ta, T1, or Tis) 2 ; however, more than half of these cases will relapse and progress to MIBC (T2 or greater) following transurethral resection of the bladder tumour (TURBT) and adjuvant chemoradiotherapy. 3 Epithelial-mesenchymal transitions (EMTs), the process by which epithelial cells assume a mesenchymal cell phenotype, are critical in various human cancers, 4 including BC. 5,6 Therefore, identification of the molecular mechanisms involved in BC progression relevant EMTs is essential to the development of BC treatments.
Type 2 diabetes (T2D) has been implicated in the development of several types of cancer, including liver cancer 7 and BC. 8 Previous studies have demonstrated that high-glucose levels may promote EMTs. Most of these studies focused on complications common to T2D, such as diabetic nephropathy 9,10 and macrovascular problems. 11 Related studies with regard to cancers are limited. Additionally, the expression and stability of Yes-associated protein 1 (YAP1) are positively correlated with O-GlcNAcylation in high-glucose-stimulated liver tumorigenesis. 12 YAP1 and TAZ (the transcriptional coactivator with PDZ-binding motif) are key components in the conserved Hippo pathway that controls carcinogenesis, regeneration and metabolism.
The explicit carcinogenicity of YAP1/TAZ has been observed in various human cancers, 13 and YAP1/TAZ-induced EMTs have also been reported. [14][15][16] These findings suggest a potential effect of high-glucose levels on EMT frequency via the Hippo pathway in BC cells, although the underlying mechanism remains unclear. We therefore sought to study how high-glucose levels promote EMTs through the Hippo pathway in BC.

| Immunohistochemistry
Immunohistochemistry (IHC) analyses were performed using formalin-fixed, paraffin-embedded (FFPE) tissues. The intensity of immunostaining in tumour tissues was assessed by the subjective visual scoring of brown stains by two independent evaluators.

| Quantitative real-time PCR
Total RNA was isolated from tissue samples or cells using TRIzol reagent (Invitrogen), and cDNA was synthesized according to the manufacturer's recommended protocols (Thermo Scientific). Quantitative real-time PCR (qPCR) was performed using a standard SYBR Green PCR Kit (Sangon Biotech) and an Applied Biosystems 7500 Real-Time PCR System (Applied Biosystems). We used glyceraldehyde-3-phosphate dehydrogenase (GAPDH) for normalization. Primer sequences are summarized in Table S1.

| Western blotting
Tissues and cells were lysed and analysed via Western blotting as previously described. 17 The following primary and secondary anti-

| Cell viability assays (MTT)
Cell viability was assessed using an MTT Cell Proliferation and Cytotoxicity Assay Kit (Bio Basic). BC cells were seeded in 96-well plates at a density of 5 × 10 3 cells per well. Viability was assessed at various time points (0, 24, 48, 72 and 96 hours) by measuring the absorbance of formazan at 570 nm using a microplate reader (Thermo Scientific).

| Transwell cell invasion assays
Cell invasion assays were conducted as follows: 1 × 10 5 cells suspended in 200 μL of serum-free medium were seeded on membranes precoated with 24 mg/mL Matrigel (BD Biosciences) and inserted into the Transwell apparatus (Corning Life Sciences). Medium containing 20% foetal bovine serum was used as a chemoattractant in the lower chamber. Cells were incubated at 37°C for 24 hours to allow for invasion. Cells on the lower surface were fixed in 4% paraformaldehyde and stained with Wright Giemsa (Sangon). In each replicate, cells were counted under a microscope (Olympus; ×200) in six predetermined fields. Assays were independently repeated at least three times.

| Enzyme-linked immunosorbent assay
The LKB1 activities and lactate concentration in the culture medium were detected using enzyme-linked immunosorbent assay (ELISA) kits (Laibio) according to the manufacturer protocol. A microplate reader (Thermo Scientific) was used to read the absorbance of each well at 450 nm.

| Xenograft model
To evaluate whether high-glucose levels would affect BC tumour growth and metastasis in vivo, we established a streptozotocin (STZ)induced xenograft nude mouse model of diabetes. 18 Mice were randomly divided into two groups (7 mice per group): The control group was fed a normal diet and injected with insulin to maintain blood glucose at levels lower than 16.7 mmol/L and was considered euglycemic; the treatment group was fed a high-glucose diet to maintain blood glucose at levels greater than or equal to 30.0 mmol/L and was considered hyperglycaemic. Overall, 5 × 10 6 T24 cells were implanted in the mice; these mice were monitored every four days.
Tumour volumes were estimated using the following formula: Volum e = width × length × (width + length)/2. The mice were killed on day 28, and the tumours were dissected out.
In the metastasis experiment, mice were randomly separated into two groups and subcutaneously implanted with 0.1 mL of T24 cells by lateral tail vein (caudal vena) injection. Metastatic progression was monitored weekly. After 28 days, the mice were killed and their lungs were surgically removed.
In the Kaplan-Meier survival curve experiment, 20 nude mouse models of diabetes were randomly separated into two groups, euglycemic group and a hyperglycaemic group. One nude mouse in the euglycemic group soon died of an infection, which occurred during the establishment process of diabetes model and was excluded. T24

| Statistical analysis
Statistical analyses were performed using SPSS18.0. The results are shown as means ± the standard error of the mean (SEM). Significance was indicated by P < .05. Table 1 shows the basic characteristics of BC patients based on their blood glucose levels. Among these 200 patients, the mean HbA1c level was 5.2% in the normal glucose tolerance (NGT) group and 8.3% in the T2D group. Fasting glucose was 5.5 nmol/L in the NGT group and 9.1 nmol/L in the T2D group. The data listed were acquired during the patient's first hospitalization due to BC. The aver- Although the average age and gender were similar in both groups, there were significant differences in some baseline characteristics, including tumour grade and pathological tumour-node-metastasis (pTNM) stage. The T2D group showed significantly reduced differentiation (P = .016), significantly higher T classification (P < .001) and

| Characteristics of BC patients
significantly higher N classification (P = .003) than the NGT group.
No significant differences in M classification and tumour size were found between the two groups (P > .05).

| YAP1 and TAZ are up-regulated in BC tissues
qPCR assessment confirmed the significant up-regulation of YAP1 and TAZ in BC tissues relative to that in adjacent normal tissues ( Figure 1A). IHC further confirmed that both YAP1 and TAZ protein expression co-ordinately increased the pathologic grade of BC ( Figure 1B). Furthermore, YAP1 and TAZ expression was higher in BC patients with T2D than in those without ( Table 2). qPCR analysis demonstrated that T24 and UMUC-3 BC cells treated with a high concentration glucose solution for 48 hours exhibited markedly higher YAP1 and TAZ mRNA levels than did other classical oncogenes ( Figure 1C).

| High-glucose concentration facilitates EMTs by regulating YAP1 and TAZ
To assess the effects of glucose on the functions of YAP1 and TAZ, we used two culture media with differing glucose concentrations:

| YAP1 and TAZ regulate GLUT1 in high-glucose culture media
To investigate the role of YAP1 and TAZ in glucose metabolism, YAP1, TAZ or both (YAP1 and TAZ) were knocked down in T24 and UMUC-3 cells via RNA Silencing. GLUT1 expression was obviously decreased upon YAP1 or TAZ inhibition ( Figure 3A,B), with an even more obvious decrease observed in cells cotransfected with siYAP1 and siTAZ (P < .05). These findings indicate that YAP1 and TAZ may be involved in glucose metabolism via GLUT1 regulation.

| MET impairs high-glucose-induced EMTs in BC cells
Our results showed that high-glucose levels altered EMT-associated

| High-glucose levels promote the invasion and proliferation of BC cell lines
The number of UMUC-3 and T24 cells that invaded through the Matrigel was determined using Transwell assays under low

| Hyperglycaemia promotes tumour growth and metastasis in vivo
In our STZ-induced xenograft nude mouse model of diabetes, we observed that the weight and volume of tumours were markedly higher in the hyperglycaemic mice than in the euglycemic mice ( Figure 6A  randomized controlled trials (RCTs) suggested that newly diagnosed cases of BC occur in a greater proportion of patients treated with dapagliflozin compared to the incidence in patients receiving placebo or a comparative treatment. 23 However, this conclusion was weak and controversial 24 and most studies have demonstrated that T2D itself, rather than antidiabetic drugs, is a risk factor for cancer.

| D ISCUSS I ON
Metastasis and proliferation play essential roles in cancer development and are considered hallmarks of the condition. 25 In BC, distant metastasis generally precedes an extremely poor prognosis. 26 EMT, a cellular programme crucial for embryogenesis and wound healing, is a pivotal process in cancer metastasis. 25 High-glucose levels have also been confirmed to promote EMTs in a differing set of benign diseases. 9,10 This may be partial because high-glucose levels and dysregulated glucose metabolism result in an acidic microenvironment, which promotes extracellular matrix degradation and impairs cell adhesion, consequently facilitating EMTs. 27 Because of the substantial glucose consumption of cancer cells, some studies have focused on the positive effects of high-glucose levels in cancer F I G U R E 4 MET impaired high-glucoseinduced EMTs in BC cells. In T24 (A) and UMUC-3 (B) cell lines, MET at least partially reduced EMTs and the expression of YAP1, TAZ and GLUT1 induced by high-glucose levels. The expression of YAP1, pYAP1, TAZ, E-cadherin, vimentin, N-cadherin, fibronectin and GLUT1 was assessed using Western blotting and qPCR. *P < .05 development. 28,29 Our study demonstrates that high-glucose levels up-regulated vimentin, N-cadherin and fibronectin expression, and down-regulated E-cadherin expression. As a result, cell proliferation and invasion in vitro, and tumour growth and metastasis in vivo were notably increased by high-glucose levels (hyperglycaemia). As a standard antidiabetic drug, MET partially suppressed the pro-EMT effects of high-glucose levels. Together, these results suggest that the pro-EMT effects induced by high-glucose levels may contribute to BC progression, while targeting glucose metabolism with molecular interventions may be a viable strategy for BC therapy, particularly in BC patients with T2D. Furthermore, among BC patients with T2D, tight glucose control may reduce metastasis and improve outcomes.  34 We confirmed that high-glucose levels promote EMT by increasing YAP1 and TAZ activity and that MET could partially suppress this effect.
As a first-line pharmacotherapy for T2D, previous studies have confirmed that MET also inhibits cancer progression. This anticancer effect has been associated with its antidiabetic function and the mTOR pathway involved in its progression. 35,36 In addition to high-glucose levels affecting YAP1 and TAZ expression, the Hippo pathway also regulates glucose metabolism. 32 In our study, GLUT1 (a transporter responsible for glucose uptake in cells) levels were found to be positively associated with YAP1 and TAZ expression.
In summary, our results indicate that T2D is positively cor- were found to regulate GLUT1 expression ( Figure 7). Therefore, the Hippo signalling pathway may act as a switch that controls the carcinogenic effect of dysfunctional glucose metabolism in BC.
Thus, targeting this pathway could preferentially be exploited as a therapeutic strategy for BC.

| CON CLUS IONS
In this study, we investigated the function and effects of high-glucose levels on EMTs in BC. Our study demonstrated that YAP1 and TAZ are key targets of high-glucose-induced EMTs in BC. These novel findings suggest that diabetes management could have a significant impact in cases of BC and T2D comorbidity and that targeting the key effector molecules of high-glucose-induced EMTs may be of therapeutic value in these patients.

ACK N OWLED G EM ENTS
This study was financially supported by the National Natural

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated or analysed during this study are included in this published article. This suggests that large amounts of YAP1 and TAZ are translocated into the nucleus where they enhance TEAD-mediated transcription and perform their functions. C, EMTs were more common and GLUT1 levels were up-regulated in BC cell lines, facilitating their invasion and proliferation