Bladder cancer is the second most common cancer of the genitourinary system and ranks ninth among the most frequently diagnosed malignancies worldwide (1). Majority (90%) of bladder cancers are histologically classified as urothelial carcinoma of the bladder (UCB), earlier known as transitional cell carcinoma. Schistosomiasis infections are identified as the root cause of non-urothelial histologies and are found to be more prevalent at least in part, in some areas of the world including Africa and the Middle East (1). Higher rates of age-standardized incidences, multiple recurrences and heterogeneous nature of disease cause excessive burden on health care systems. The anatomical, genetical, hormonal and socio-economical status, environmental/industrial/occupational exposure to carcinogens/aromatic amines, use of tobacco, and prolonged use of arsenic-contaminated or chlorinated drinking water are the major risk factors for bladder cancer (2). These risk factors may lead to changes in urothelium, thus represent potential reasons for substantial molecular differences among clinically different types of bladder cancer.

Majority of UCB (80%) arise through the non-invasive papillary pathway, are diagnosed as papillary bladder tumors. These tumors are identified with the growth of hyperplastic urothelium towards the bladder lumen. Transurethral resection (TUR) of the bladder is the treatment of choice; however, still 50% to 90% of patients suffer frequent recurrences (3). Despite multiple recurrences, only 15%-20% of patients progress to muscle-invasive cancer. Muscle-invasive bladder cancer (MIBC) evolves from severe dysplasia or carcinoma in situ (CIS). Despite radical resection, tumors progress in 50% of such patients and the 5-year overall survival rate is reduced to 5%. Heterogeneous clinical outcomes, multifocal nature of tumor and higher rate of recurrences require attentive bladder cancer surveillance, thus makes the bladder cancer, an expensive disease to manage (4).

Detailed biological and molecular insights into the disease pathogenesis may help clinicians for adequate treatment decisions at the time of its presentation; thereby minimize the risk of tumor recurrence and shorter survival probabilities of patients (5). Nevertheless discovering molecular markers for identifying patients with invasive tumors; high recurrence rates; and development of effective therapies for locally progressed and metastatic cancers, are some of the challenges in the clinical setting (6).

Findings of genome-wide approaches and candidate gene analysis validate the significance of mitogen-activated protein kinases (MAPK) activation; increased cyclin D1 expression; oncogenic mutations in telomerase reverse transcriptase (TERT), Harvey rat sarcoma viral oncogene (HRAS), fibroblast growth factor receptor (FGFR), and phosphoinositide 3-kinase (PI3K), in the development of papillary cancer (4, 6). Increased chromosomal instability; phosphatase tensin homolog (PTEN) inactivation; loss of tumor protein 53 (TP53) function; severe disturbances in cell cycle regulators, mutations in retinoblastoma (RB1) and cyclin-dependent kinase inhibitor 2A (CDKN2A)/p16INK4A; and altered DNA methylation and its effect on the expression of cell cycle regulators, promote muscle-invasive disease (6, 7). Mutations in PI3K, tumor suppressor genes, deleted in bladder cancer 1 (DBC1), tuberous sclerosis 1 (TSC1), CDKN2A and patched are noticed in both papillary and invasive tumors (8). Gene expression, whole genome array-CGH (comparative genomic hybridization) and mutation analyses validate FGFR3, PIK3CA, KRAS, HRAS, NRAS, TP53, CDKN2A, and TSC1 genes as molecular signature tools in the classification of urothelial carcinoma into low-grade/high-grade as well as NMIBC/MIBC with high precisions and sensitivities (9) (Table 1). Gene expression profiling studies associate basal tumors with better survival, while p53-like tumors with bone metastases and chemoresistant disease thereby identify subtypes of urothelial cancer that differ in their natural history and sensitivity to chemotherapy (32). Mutational status of DNA repair genes-Excision repair cross-complementing rodent repair deficiency, complementation group 2 (ERCC2), Fanconi anaemia complementation group C (FANCC), ataxia telangiectasia mutated (ATM) and RB1provide the basis of differential clinical responses to neoadjuvant platinum-based chemotherapies/targeted therapies thereby offer opportunities for personalization of bladder cancer therapy (33).

Genetic and epigenetic changes may lead to altered molecular pathways involved in tumor development, and subsequently, drive bladder tumor cells towards epithelial-to-mesenchymal transition (EMT) (10-12, 34). EMT endows sessile, non-motile urothelial cells to undergo complex reprogramming accompanied with the loss of apicobasal polarity and gain of migratory abilities. Molecular regulatory determinants regulate EMT, allow a subset of urothelial cancer cells to gain mesenchymal traits with self-renewal abilities, evade immune mechanisms and infiltrate the surrounding basement membrane. Tumor microenvironment influences tumor cells with mesenchymal phenotype to switch to epithelial phenotype via reverse process to EMT, known as mesenchymal-to-epithelial transition (MET) (35). Urothelial tumor cells then colonize to distant organ sites, establish pre-metastatic niches, and may survive against anticancer therapies (3). The present review extends our understanding of the regulatory functions of EMT/core associates and their clinical implications in urothelial tumorigenesis, which may facilitate the individualization of therapeutic decisions.

EMT, an essential and naturally occurring physiological phenomenon, starts with alterations in cell morphology, cell-cell adhesion capabilities, apicobasal polarity and cell motility. Chain of events from the cells’ inside to extracellular matrix (ECM) triggers loss of intracellular tight junctions, stress fiber redistribution, acquisition of mesenchymal-like traits with self-renewal abilities and increased invasive potential. Based on biological functions, the EMT program is classified as type 1, type 2 and type 3 EMT. Type 1 EMT is a natural gestational orchestration and vital for morphogenesis during embryonic development. Type 2 EMT functions in wound healing or fibrosis and is majorly identified in kidney, liver, lung, and intestine. Its association with the secretion of different inflammatory components that interact with ECM, causes epithelial layers to dissociate due to basement membrane degradation. Type 3 EMT, also known as oncogenic EMT, translates early-stage tumor into invasive malignancies. EMT and MET promote cellular/epithelial plasticity by allowing tumor cells to switch between epithelial and mesenchymal phenotypes, thus induce metastatic dissemination (36).

Advancements in gene expression profiling, next-generation sequencing and the explosion of non-coding RNA provide host of information on essential roles of EMT in urothelial tumorigenesis and in its molecular subtyping. Downregulation of CDH1 (E-cadherin), zona occludens 1, claudins, occludins, cytokeratins and components of basement membrane including collagen IV and laminin 1; and upregulation of fibroblast-specific protein 1, type 1 collagen, N-cadherin (CDH2) and P-cadherin (CDH3) are essential features of EMT at cellular level (Table 1). Reduced hemophilic interactions between E-cadherin on adjacent cells are responsible for the loss in cell-cell adhesion and increased cell motility. Its attenuated expression is reported to be associated with high progression rate of bladder cancer cells (32, 37-39). Focal loss in E-cadherin facilitates the release of β-catenin, thereby contributes to reduced anchorage to the actin cytoskeleton and poor mechanical stability. Nuclear translocation of β-catenin and activation of Wnt target genes result in EMT and increased metastasis (40). Interaction of N-cadherin with FGFR1 dictates activation of MAPK/extracellular signal-regulated kinase (ERK), increases the expression of matrix metalloproteases 9 (MMP-9), and promotes tumor invasion. N-cadherin is considered as an independent marker for prognosis in low stage bladder tumors (41). Gain in N-cadherin regulates tumor invasiveness through the PI3K/protein kinase B (Akt) pathway. Recent studies document increased expression of P-cadherin as a marker of poor prognosis in bladder tumorigenesis. It is normally expressed in the basal cell layer of the urothelium. Forced expression of P-cadherin is shown to modulate catenin expression and enhance the migration of EJ and UM-UC-3/P-cadherin transfectants (42). Multivariate analysis identifies an independent association of expression levels of E-cadherin, MMP-9, and Twist; tumor size; and concomitant carcinoma in situ with intravesical recurrence-free survival of patients diagnosed with non-muscle invasive tumors (43).

Core associates of the event influence EMT/MET by the secretion of stroma-modulating ligands/ECM-degrading enzymes including VEGF, PDGF, basic FGF2, EGF receptor (EGFR), transforming growth factor β (TGFβ), tumor necrosis factor-α (TNF-α), Hedgehog, Notch and nuclear factor-kappa B (NF-kB) ligands; overexpression of EMT-activating transcription factors (EMT-ATFs); microRNAs; and oncoactivation of downstream signaling pathways; and overexpression of extracellular matrix (ECM)-degrading enzymes such as MMPs. These factors recruit mesenchymal progenitors, drive cellular plasticity, stimulate angiogenesis and tumor invasion (Figure 1).

Figure 1.

EMT and its core associates include (i) secretion of stroma-modulating growth factors; (ii) upregulation of EMT-ATFs; and (iii) activation of oncogenic microRNAs and suppression of tumor suppressor microRNAs. These factors activate oncosignaling pathways, elicit EMT/MET, recruit mesenchymal progenitors, drive cellular plasticity, stimulate angiogenesis and tumor invasion. EMT: Epithelial-to-mesenchymal transition; MET: Mesenchymal-to-epithelial transition; EMT-ATFs: EMT-activation transcription factors.

Epithelial lining of the bladder/urothelium is composed of highly differentiated, multinucleated umbrella cells towards bladder lumen; a variable number of multilayered intermediate cells; and unilayered basal cells with very high proliferative potential (Figure 2). Presence of slow cycling urothelial stem cells (USCs) in the basal cell layer contributes to high regenerative potential (73). Long life span; high colony-forming efficiency; high nuclear-cytoplasmic ratio; low granularity; and expressions of β1/β4 integrins, CD44, cytokeratins (CK-5/14, CK-17) and laminin receptor (LR) are the characteristic properties of basal USCs (72). Recent researches hypothesize the possibility of the existence of another independent pool of USCs, and support the formation of umbrella cells in p63-null mice that lack basal/intermediate cells (73). Regenerative/proliferative capacity diminishes with the differentiation of USCs into transit-amplifying cells of intermediate cell layer and then into umbrella cells. Studies on rat bladder validate the differential/maximal distribution of USCs in the caudal region of the bladder. This research suggests that maximal distribution of USCs may contribute to the relatively higher rate of incidence of carcinoma per surface area in this region (73). Mutagenic insults including genetic instability, epigenetic changes, mutations in oncogenes/tumor suppressor genes and aberrant molecular pathways transform adult stem cells into a subset of unique urothelial cancer stem cells (UCSCs). UCSCs derived from differentiated progenies acquire tumorigenic properties. These cells are characterized by their enriched ability to conserve cellular heterogeneity by inducing xenograft tumors in vivo (74). Many cell surface markers are shared by adult stem cells and mature urothelial cells. Subsets of UCSCs, also known as side-population (SP), are examined by their enriched capacity to (i) express cell surface marker proteins including 67LR at the tumor-stromal interface, aldehyde dehydrogenase 1A1 (ALDH1A1^hi), CD44, CD44 splice variant6 (CD44v6), CD133, cytokeratins (CK5, CK7, CK17), early progenitor Nestin (NES), and low carcinoembryonic antigen-related cell adhesion molecule 6 (CEACAM6^low); (ii) upregulate oncogenic genes including B lymphoma Moloney murine leukemia virus (Mo-MLV) insertion region 1 homolog (BMI1), β-catenin, GLI1, NANOG, POU domain class 5 transcription factor 1 (POU5F1)/Oct4, and signal transducer and activator of transcription 3 (STAT3); (iii) express high levels of ATP-binding cassette (ABC) transporters and multidrug resistance pumps (MDRs); and (iv) efflux of vital dye Hoechst 33342 and DyeCycle violet (3).

Figure 2.

Schematic view of cross-section of the adult bladder wall showing the four major layers: (i) urothelium towards bladder lumen; (ii) lamina propria; (iii) detrusor smooth muscle layer; and (iv) adventitia. Smooth muscle bundles are aligned in lamina propria and detrusor muscle layer.

Heterogenous populations of UroCSCs with high tumour-initiating potential, first identified in 2009, originate from normal stem cells/differentiated progenies with mutational defects, are associated with the acquisition of EMT phenotype (75). UroCSCs with hybrid epithelial/ mesenchymal (E/M) or metastable phenotype are associated with epithelial plasticity, tumor proliferation, quiescence, stemness at tumor bud level, successful colonization of distal target tissues, drug resistance, and poor survival probabilities of the patients. Powerful genome-wide screening techniques set challenges for diagnosing the two distinct types of bladder diseases, monitoring tumor recurrence/metastasis and providing prognostic risk stratification by identifying predictive markers in the clinical setting.

High-grade NMIBC compared to low-grade NMIBC are observed with higher rates of amplification of chromosome 6p22.3, SOX4, E2F3, Snail, Zeb, and Twist family genes . This supports their potential role as a regulator of bladder CSCs and validates them as predictive biomarkers for an early event in tumor progression (75). Sonic hedgehog promotes EMT induction and bladder carcinogenesis in initial tumor stages. Hh signaling contributes to drug resistance in the later clinical stages and ultimately leads to metastasis (76).

The induction of EMT, actin cytoskeleton remodeling and the overexpression of TGFβ signaling effector, cofilin, in a series of human bladder cancer specimens are significantly associated with bladder cancer progression (77). Whole genome mRNA expression profiling of xenograft models derived from human bladder cancer cell lines demonstrates the onset of reversible EMT via TGFβ/Snail pathway activation and its association with the accumulation of circulating tumor cells (CTCs) and regional/metastatic lesions (78). The findings of the experimental research performed by Liang et al. identify the mechanisms of TGFβ signaling in promoting EMT, cancer stemness and bladder cancer progression in vivo (79). Another study performed by Yu et al. demonstrates positive correlation between the loss of Nkx2.8, candidate tumor-suppressor gene, and lymph node metastasis. Overexpression of Twist1 reverses EMT inhibition by Nkx2.8 and restores the invasive phenotype of urothelial carcinoma cells (80). PI3K-deregulated p27 binds Janus kinase 2 (JAK2), drives STAT3 recruitment/activation, upregulates Twist1, induces EMT, activates AKT and promotes cancer invasion and metastasis (81). SOX4, a potential regulator of the bladder cancer stemness, serves as a biomarker of tumor progression from NMIBC to MIBC and its expression correlates with advanced cancer stages and poor survival/prognosis (75). Addition of androgen receptor (AR) contributes to the development and progression of upper urinary tract urothelial cell carcinoma by significantly expanding the population of CSCs; increasing the clonogenicity/tumor sphere formation, upregulating the expression of Oct4, Bmi1 and Nanog, altering CSC-related miRNA profile, and promoting EMT (82). The positive relationship among higher levels of lactate dehydrogenase-A(LDH-A), CSC/EMT markers, cell proliferation, invasion and migration suggests an important function of LDH-A in malignant progression in invasive bladder cell line and muscle invasive patients (83). A study by Vantaku et al. examines higher expression of ganglioside GD2, cell proliferation, mesenchymal and cancer stem cell phenotypes in high grade, muscle-invasive bladder cancer tissues (84). Experimental research by Franzen et al. provides insights into the role of exosomes in the transition of bladder cancer into the invasive disease (85). Analysis of clinical specimens identifies a significant association of CD44 staining with higher loco-regional failure rate, advanced clinical stage, and lower disease-specific survival rate for MIBC patients. The results of cellular experiments and orthotopic tumor models provide evidence on regulatory functions of activated IL-6 signaling in the induction of CD44 expression by providing a suitable microenvironment, activating STAT3, promoting EMT, and regulating programmed death ligand 1-mediated T-cell suppression (86). Exosomes isolated from T24/UMUC3 invasive bladder cancer cells and patient’s urine and bladder barbotage samples potentially modulate signaling pathways in recipient urothelial cells. Bladder cancer-shed exosomes can induce EMT and show tumorigenic effects. These exosomes increase the expression of mesenchymal proteins, α-smooth muscle actin, S100A4 and Snail, reduce the expression of E-cadherin and β-catenin levels, thereby lead to increased migration and invasion of urothelial cells (85).

Upregulated expression of circPRMT5, a novel class of non-coding circular RNAs, in serum and urine exosomes positively associates with advanced clinical stage and worse survival of UCB patients. Clinical analysis from two independent cohorts of UCB reveals the regulatory effect of circPRMT5 on cells’ EMT via sponging miR-30c and its target genes (Snail1 and E-cadherin) in UCB progression (25). MiR-200 and miR-205 silencing, DNA hypermethylation and chromatin repression are proposed as possible prognostic markers in bladder cancer (26).

EMT, a trans-differentiation process, stimulates the expansion of cancer stem/progenitor cells and is thus responsible for increased invasiveness, drug resistance and angiogenesis. Tumor-associated stromal cells and ECM provide protective cancer stem cell niches, influence the tumor growth, and develop more aggressive tumor types. Understanding the regulatory functions of EMT and its core associates not only set challenges for their diagnostic and prognostic implications in risk stratification of pathologically similar urothelial tumors, but also provide the platform for the development of novel therapy. Combinational effective therapeutic strategies based on the potential use of drug/molecules may create an inhospitable microenvironment, suppress EMT and kill cancer stem/progenitor cells.

Subcutaneous bladder tumor regeneration system, developed by recombining benign human fetal bladder stromal cells with human bladder cancer cells, recapitulates malignant human bladder tumor architecture and mimics human disease. Tumor cells in the presence of stroma exhibit EMT, cancer stemness, decreased differentiation, greater activation of oncogenic/ growth pathways, accelerated growth, and muscle invasion (87). Such efficient bladder tumor models may help elucidate the influences of stroma on tumor growth and development, and provide an accessible system for therapeutic testing.

Long-term treatment of urothelial cancer cells with drug/cisplatin does not result in enrichment of CD90 (+) cells with stemness and enhanced clonogenic characteristics. Nevertheless, these sublines display substantial phenotypic plasticity, express EMT associated markers, an altered pattern of CKs, Wnt-pathway target genes and contribute to the emergence of cisplatin resistance (88). Targeted therapies based on the use of inhibitors of urothelial plasticity by controlling EMT and its core regulators may suppress metastatic growth and tumor relapse. Epigenetic modifiers inhibit transcription factors involved in chromatin deregulation, suppress bladder cancer stem cell maintenance, and therefore, serve as the potential molecules of therapeutic significance (75). Myrtucommulone-A (MC-A), an anti-cancer novel agent, concomitantly inhibits multiple signaling pathways, ERK 1/2, p38 MAPK, and Src activity; downregulates EMT by affecting N-cadherin, β-catenin, Snail, and Slug expressions; and reduces cross-talk in bladder cancer cells (89). Enzalutamide, an AR inhibitor, has been shown to suppress its effects on tumorsphere formation. Targeting AR may lead to novel therapeutic approaches for genetically diversified urothelial carcinomas (84). Chemopreventive effects of curcumin reverse tobacco smoke-elicited activation of Wnt/β-catenin, urocystic EMT and acquisition of cancer stem cell properties (90). TGFβ receptor inhibitor treatment ablates TGFβ signaling; inhibits cancer cell proliferation, cancer stem cell population, and EMT; and suppresses invasive cancer progression (79). Silibinin, a nontoxic natural flavonoid, inhibits glycogen synthase kinase 3β (GSK3β) phosphorylation, promotes β-catenin nuclear translocation and transactivation and reverses EMT by suppressing Zeb1 levels and modulating the levels of cytokeratins, vimentin and MMP2. In addition, it suppresses the properties of UroCSCs, side population, spheroid colony formation, expression of stem cell factor CD44, thus represents a novel natural treatment in targeting bladder cancer metastasis (91). Owing to oncogenic functions of αv integrins in migration, EMT, and maintenance of ADH activity, treatment with an αv integrin antagonist or its knockdown in bladder cancer cell lines results in increased CDH1/CDH2 ratio; downregulation of Snal2, Nanog, Bmi1 and ADH; and decreased proliferative, migratory and clonogenic capacity (92).

Local administration of antagomiRNAs/antimiRNAs to suppress its oncogenic functions or miR replacement therapy to restore the loss of function, modifies stromal region, reverses EMT, and sensitizes CSCs to anti-cancer drugs. Differentiation therapies based on the stable expression of miR-200 and treatment with anti-EGFR antibody cetuximab in mesenchymal human bladder cancer cell lines reduce cell migration, modulate EMT, induce differentiation, and restore the sensitivity of cancer cells to EGFR inhibitors (3). Overexpression of TP73-AS1, a long noncoding RNA (lncRNA) with tumor suppressor functions, diminishes EMT by upregulating E-cadherin and downregulating vimentin, snail, MMP-2, and MMP-9 (27). Increased expression of lncRNA, long stress-induced non-coding transcript 5 (LSINCT5), in human bladder cancer cell lines and its correlation with poor prognosis, enhanced Wnt/β-catenin signaling activation and EMT may validate it as a potential candidate for therapeutic intervention (28). Overexpression of miR-323a-3p significantly inhibits EMT and progression of bladder cancer by regulating MET/SMAD3/SNAIL circuit (14). Anti-metastatic effects/pro-apoptotic functions of miR-34a are determined to sufficiently reverse EMT by directly targeting CD44 gene and restoring the endogenous levels of E-cadherin in cancer cell lines (29).

EMT is reported to coordinately regulate maintenance of cancer stemness, drug resistance, angiogenesis and muscle invasion/metastasis, and is documented as a dominant feature in the biology of urothelial cancer. Successful elimination of subpopulation of UroCSCs and its differentiated progenies by targeting EMT and its core associates could be clinically beneficial in reversing EMT, preventing tumor relapse and prolonging survival of patients.

One of the authors, RS is thankful to the University grant commission (UGC), Govt. of India, for providing research fellowship.