Long noncoding RNA MAGI2-AS3 inhibits bladder cancer progression through MAGI2/PTEN/epithelial-mesenchymal transition (EMT) axis

2.1 Clinical samples

The Ethics Committee at the Affiliated Hospital of Jining Medical University approved the study, which conformed to The Code of Ethics of the World Medical Association (Declaration of Helsinki, 2013 amendment). All patients provided written, informed consent to collect resected tissue samples for research use. In all, 80 Bca tissues and 30 paired para-cancerous tissues, respectively, were obtained from patients diagnosed with Bca between March 2010 and May 2015, at the Department of Urology, Affiliated Hospital of Jining Medical University. All samples were immediately stored at - 80 ∘ C. Two pathologists diagnosed the tissues based on the World Health Organization 2004 [15], and the UICC 2009 guidelines [16].

2.2 High-throughput transcriptome sequencing

Total RNA was sequentially extracted, quantified, and purified, and the integrity was confirmed. Total RNA was enriched using magnetic beads with oligo (dT) (New England BioLabs, Inc., Ipswich, MA, USA) after digesting DNA with DNase I (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Short fragments that dissociated from the total RNA were sublimated and resolved in Elution buffer (Thermo Fisher Scientific, Inc., Waltham, MA, USA). Appropriate templates were amplified using OneTaq ® DNA polymerase (New England BioLabs, Inc., MA, USA). Amplicons were evaluated using an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc., Technologies GmbH., Waldbronn, Germany) and a library was created using the ABI StepOnePlus Real-Time PCR System (Agilent Technologies, Inc.). The library was sequenced using a HiSeq™ 2000 system (Illumina, Inc., San Diego, CA, USA).

2.3 Cell culture and transfection of lentiviral plasmids

The Bca cell lines, T24, J82, 5637, RT4, and the normal urothelial cell line, SV-HUC, (Cell Bank of the Chinese Academy of Sciences, Shanghai, China) were respectively cultured in RPMI-1640 and L-15 media (both: Invitrogen, Carlsbad, CA, USA containing 1% ampicillin (100 units/mL), streptomycin (100 units/mL), and 10% (v/v) fetal bovine serum (FBS) at 37 ∘ C under a 5% CO 2 atmosphere. The MAGI2-AS3 (LV-MAGI2-AS3) and negative control (LV-NC) lentiviral plasmids (Genechem Co., Ltd., Shanghai, China) were transfected into the cells as described by the manufacturer. When the cells reached 60%–70% confluence, the media were replaced with FBS-free media supplemented with the infection enhancers, HiTransG A or HiTransG (both from Shanghai GeneChem Co., Ltd., Shanghai, China). Thereafter, the cells were incubated with lentiviral plasmids at various concentrations and multiplicities of infection (MOI), added to the media for 12 h. These infective media were then replaced with complete media after 12 h of culture. During the first 48 h, the media contained 2 g/mL puromycin to kill uninfected cells and achieve maximal infection.

2.4 Quantitative real-time PCR

Total RNA was extracted from bladder tissues and cell samples using TRIzol™ Reagent (Thermo Fisher Scientific Inc., MA, USA) as described by the manufacturer. Vimentin, MAGI2-AS3, Snail1, E-cadherin, and N-cadherin expression was quantified using RT-qPCR and SYBR select master mix kits (Thermo Fisher Scientific Inc.) Expression levels of these genes were normalized to that of glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Table S1 lists the primers for all genes. Data are presented as means of fold change compared with an internal reference.

2.5 Migration and invasion assays

Cell migration and invasion were determined using Transwell chambers. Chambers for invasion assays were inserted into Matrigel (Millipore Sigma Co., Ltd., Burlington, MA, USA), and activated under pure media at 37 ∘ C under a 5% CO 2 atmosphere for 30 min. Media (approximately 800 μ L) were added to the lower chambers, then 5 × 104 cells were mixed in 500 μ L of media, and seeded in triplicate into the upper chambers in 24-well plates. After 24 h, the cells in the upper chamber were discarded, and the cells that had migrated or invaded the lower chamber were fixed in 4% formalin and stained with 0.1% crystal violet for 30 min. The chambers were then washed twice with phosphate buffered saline (PBS), and images were acquired using a camera attached to a microscope.

2.6 Mouse model of orthotopic Bca

Six 6-week-old female NOD/SCID mice (Shanghai Laboratory Animal Center, SLAC, Shanghai, China) were anesthetized via an intraperitoneal injection of 10% Nembutal (45 mg/kg), then an indwelling catheter was placed in the mice with paraffin oil. Residual urine in bladders was removed, then the bladders were washed twice with PBS, and flushed with 30 μ L of HCl (0.1 N) for 20 s via the indwelling catheters. Thereafter, 30 μ L NAOH (0.1 N) was injected into the bladders to neutralize the acid, and the bladders were washed twice with PBS. T24 cells (1 × 107/100 μ L) were injected into the bladders, and the urethras were ligated for 2 h. The mice were monitored weekly via whole-body, live-animal bioluminescence optical imaging using D-luciferin (150 mg/kg) assay kits (Gold Biotechnology, Inc., Olivette, MO, USA) and an IVIS Lumina II imaging system (PerkinElmer, Life and Analytical Sciences Inc., Waltham, MA, USA). The mice were sacrificed at 6 weeks after tumor-cell implantation and their bladders were weighed. Major organs and whole bladders were harvested and fixed in 10% formalin (Thermo Fisher Scientific Inc.) at 3-day intervals for 8 weeks.

2.7 Western blotting

Protein expression was quantified by western blotting. All samples were adjusted to a protein concentration of 1 g/L using the bicinchoninic acid (BCA) method, then 5 μ g portions were separated by SDS-PAGE and subsequently transferred to PVDF membranes (Pall Corporation, Port Washington, NY, USA). The membranes were washed with tris-buffered saline (TBS), then non-specific binding was blocked with non-fat milk diluted in TBS buffer, followed by an overnight incubation with primary antibodies (all from Abcam, Cambridge, UK) diluted 1: 1,000: anti-GAPDH, anti-MAGI2, anti-PTEN, Snail1, E-cadherin, N-cadherin, and vimentin (1:500). All membranes were then immersed in horseradish peroxidase (HRP) conjugated anti-rabbit secondary antibodies (1:5,000, Invitrogen). Target proteins were visualized and analyzed on a ChemiDoc MP System (Bio-Rad Laboratories Inc.).

2.8 Immunohistochemistry (IHC)

Paraffin-embedded tissue samples from orthotopic Bca models were assessed by immunohistochemistry (IHC) using polyclonal antibodies against MAGI2 (diluted 1:500; Thermo Fisher Scientific Inc.) and PTEN (1:50; Abcam), followed by a standard IHC protocol. Three pathologists independently evaluated staining intensity and the proportions of positively stained tumors.

2.9 Statistical analyses

All data were statistically analyzed using GraphPad Prism version 5.0 (GraphPad Software In., San Diego, CA, USA) and SPSS version 17.0 (SPSS, Inc., Chicago, IL, USA). Data are presented as means ± standard error of the means (SEM) from at least three independent assays. Differences between groups were compared using Student t-tests, and quantitative data were analyzed by chi square tests. Correlations between MAGI2-AS3 expression and common clinicopathological parameters were analyzed using Spearman rank tests. Differences in overall survival (OS) between high and low MAGI2-AS3 expression were determined using Kaplan-Meier curves and log rank tests. Values with two-tailed 0.05 were considered statistically significant.

OPEN IN VIEWER

3.1 Expression of lncRNA MAGI2-AS3 is downregulated in Bca tissues and cell lines

The results of RNA-seq data from 10 paired of Bca and para-cancerous tissues, and 5 base tissues, showed that 36 lncRNA were significantly downregulated by at least 3-fold in Bca, compared with para-cancerous and corresponding base tissues. These downregulated lncRNA included LINC01082 (ln375190), ADAMTS9-AS2 (ln384265), and ADAMTS9-AS1(ln339386).MAGI2-AS3, defined as n410887 in the NONCODE database was also differentially expressed (Fig. 1a and b). Like that of MAGI2-AS3, expression of the sense RNA, MAGI2 (Fig. S1), and its downstream gene, PTEN, was also downregulated in Bca tissues (Fig. 1c and d). Thus, we speculated that MAGI2-AS3 correlated with its sense RNA MAGI2 in Bca. Consequently, further investigated MAGI2-AS3.

We confirmed the expression profiles of lncRNA MAGI2-AS3 in 80 Bca and 30 paired para-cancerous normal tissues by qPCR. The expression levels of MAGI2-AS3 were significantly lower in Bca, than in normal tissues (Fig. 2a). Similarly, the expression of MAGI2 and PTEN in tissues was consistent with the RNA-seq data; expression was downregulated in cancer, compared with normal tissues (Fig. 2b and c). Furthermore, expression of the three genes was obviously higher in NMIBC, than in MIBC tissues (Fig. 2d–f).

OPEN IN VIEWER

We classified the 80 patients with Bca into groups with high ( n = 40) and low ( n = 40) MAGI2-AS3 expression based on the median value. We compared age, gender, tumor number, tumor size, tumor grade, lymph node invasion, and TNM stage, and found that 31 (38.8%) of them had stage T2 cancer, whereas 49 (61.2%) 0 had stages T2–T4. We also found that 15 (15.0%) of the 80 patients had positive lymph node invasion. Therefore, we analyzed associations between MAGI2-AS3 expression and clinicopathological parameters using chi square tests. The results significantly associated downregulated MAGI2-AS3 with the number of tumors ( P = 0.021), tumor stage ( P = 0.001), and tumor grade ( P = 0.000), but not with age ( P = 0.073), gender ( P = 0.499), tumor size ( P = 0.501), and lymph node invasion ( P = 0.102). Table 1 shows details of this correlation analysis. Potential prognostic differences between the two groups were assessed as overall survival (OS) using Kaplan-Meier curves. The first and third OS rates were 89.5% and 69.9%, respectively, and 79.4% and 39.7%, respectively, in the groups with high and low MAGI2-AS3 expression (Significantly different at P = 0.029 log rank test; Fig. 3).

OPEN IN VIEWER

The expression of MAGI2-AS3 and MAGI2 mRNA significantly correlated (Fig. 4a). Moreover, expression of the latter positively correlated with that of PTEN mRNA (Fig. 4b). Subsequent qPCR results verified these correlations among the three genes (Fig. 4c and d). Analysis using the GEPIA webtool in TCGA datasets revealed that MAGI2-AS3 expression positively correlates with MAGI2 expression (Fig. 4e). The correlation between MAGI2 and PTEN was not statistically significant, as trends in expression were the same for both genes (Fig. 4f). Thus, we investigated whether MAGI2-AS3 influences Bca progression by regulating the sense strand RNA MAGI2 and its downstream gene PTEN expression.

OPEN IN VIEWER

We applied western blotting and qPCR assays to determine changes in MAGI2 and PTEN expression while overexpressing MAGI2-AS3 in T24 and J82 Bca cells in vitro. WE found that MAGI2 and PTEN expression was upregulated when MAGI2-AS3 expression was increased (Fig. 5a and b). Metastatic sites occurred in the abdomen of orthotopic Bca models after 10 weeks in the T24-LV-NC, but not in the T24-LV-MAGI2-AS3 group (Fig. 5c). The IHC findings of the resected tumors revealed more MAGI2 and PTEN protein in the T24-LV-MAGI2-AS3, than the T24-LV-NC group (Fig. 5d). Furthermore, transwell assays showed significantly higher migration and invasion capacity in T24 Bca cells overexpressing MAGI2-AS3 together with interfering MAGI2, compared with T24 Bca cells overexpressing only MAGI2-AS3 (Fig. 6a). These findings indicated that the expression of interfering MAGI2 reversibly inhibited the migration and invasion capacity induced by overexpressing MAGI2-AS3. Finally, the results of western blotting and qPCR assays showed significantly downregulated MAGI2 and PTEN expression in the co-transfected group compared with the group overexpressing only MAGI2-AS3 (Fig. 6b and c).

OPEN IN VIEWER

OPEN IN VIEWER

Inactivating PTEN induces the EMT process and tumor metastasis [17]. Furthermore, gene ontology (GO) enrichment analysis of RNA-seq data revealed that differentially expressed genes associated with MAGI2-AS3 were significantly enriched on the extracellular matrix (Fig. S2), and that protease recruitment was mostly caused by EMT activation [18]. Hence, lncRNA MAGI2-AS3 probably inhibits Bca progression by inducing EMT inactivation. We therefore selected the vital genes Vimentin, Snail1, E-cadherin, and N-cadherin that are involved in the EMT pathway, and measured their protein levels expressed in LV-MAGI2-AS3 and LV-NC groups in both cell lines by western blotting. The expression of vimentin, Snail1, and N-cadherin was much lower in the LV-MAGI2-AS3, than the LV-NC groups, whereas that of E-cadherin was obviously enhanced (Fig. 7a). The mRNA expression of the corresponding genes detected using qRT-PCR showed much more Vimentin, Snail1, and N-cadherin mRNA expression in the LV-MAGI2-AS3, than the LV-NC groups, while E-cadherin mRNA expression was significantly downregulated in the LV-MAGI2-AS3 groups (Fig. 7b and c). Changes in gene expression on the background of MAGI2-AS3 overexpression suggested inhibition of the EMT pathway.

OPEN IN VIEWER