Hypoxic Regulation of Glycosylation in Bladder Cancer

Hypoxic Regulation of Glycosylation in Bladder Cancer Andreia Filipa Ferreira Peixoto Mestrado em Bioquímica Departamento de Química e Bioquímica 2014 Orientador José Alexandre Ferreira, Doutor, Universidade de Aveiro Co-orientador Maria José Oliveira, Professor Doutor, Faculdade de Ciências Todas as correções determinadas pelo júri, e só essas, foram efetuadas. O Presidente do Júri, Porto, ______/______/_________ FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer i Agradecimentos A palavra que expressa a admiração, respeito e carinho pelos meus professores é agradecimento. Este ano tive um sem número de professores, aqueles que me ensinam a técnica e aqueles que me limam a personalidade, a forma de encarar o desafio e a forma de me resguardar do excesso. Penso que esta é a oportunidade ideal para agradecer por tudo aquilo que fizeram por mim e por este projeto. Devo indubitavelmente começar pelo mais precioso elemento deste grupo de professores, o Doutor Alexandre Ferreira. Com a sua excelência, conseguiu fazer-me descobrir um fantástico caminho de conhecimento e novidades que me deslumbraram a cada passo. Retirou vários entraves desse meu caminho à custa de muito esforço pessoal e, por isso, estou-lhe muito grata. Considero-o mestre, mestre do saber e da busca incessante pelo conhecimento e também um exemplo que me acompanhará em novas fases, dentro e fora da Ciência. Obrigada por me deixar acompanha-lo nesta viagem e obrigada por me abrir horizontes e me garantir perspetivas de Futuro. A si, um profundo obrigada! À Professora Doutora Maria Oliveira, que muito me ensinou e acarinhou também dedico um profundo agradecimento. Obrigada por todos os conhecimentos que me transmitiu e obrigada pelos preciosos conselhos. Também a Maria constitui um muito importante exemplo de bem-fazer e a pessoa especialíssima que é transparece em tudo a que se dedica. Obrigada por este maravilhoso ano. Á minha querida amiga Elisabete Fernandes agradeço tudo aquilo que me ensinou e agradeço o exemplo de profissionalismo e rigor que me proporcionou. Agradeço ainda a oportunidade de me ter deixado crescer com ela nesta fase a que me propus e agradeço o facto de ter estado comigo a cada passo e de ter sido sempre um grande apoio. Ao Professor Doutor Luís Lima devo também um agradecimento pelos conhecimentos técnicos que me transmitiu mas acima de tudo pela incrível capacidade de fazer tudo parecer mais simples sem nunca perder o rigor. Obrigada por me deixar contar consigo. Ao meu colega Manuel Neves agradeço a calma que me transmite e a prontidão com que se propõe a ajudar. Tenho certeza que será uma enorme mais-valia nesta nova fase em que trabalharemos juntos. Á minha colega Beatriz Parreira agradeço pela simpatia com que sempre me recebeu. Às minhas colegas Cátia Monteiro e Marta Pinto agradeço pelos FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer ii conhecimentos que me transmitiram e pela disponibilidade com que sempre me receber. Agradeço ainda ao Professor Doutor Carlos Palmeira pela incansável disponibilidade e amabilidade com que me recebeu e acompanhou e pela oportunidade que me concedeu para explorar novos horizontes. Finalmente, devo um agradecimento ao Professor Doutor Lúcio Lara Santos por me ter permitido fazer parte do Grupo de Patologia e Terapêutica Experimental e trabalhar com tão maravilhosas pessoas. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer iii Sumário O cancro de bexiga músculo-invasivo figura um dos tipos de cancro genito-urinários mais comuns e com maior taxa de mortalidade. A principal terapêutica utilizada consiste em regimes quimio-terapêuticos baseados em cisplatina, que falham na prevenção da recorrência e disseminação da doença. Assim, são necessários novos bio-marcadores que permitam a estratificação dos doentes e novas estratégias terapêuticas. Mais de 70% dos tumores de bexiga musculo-invasivos expressam o antigénio Sialil-Tn (STn) em resultado da terminação prematura do processo de elongação das cadeias de Oglicanos das proteínas da superfície celular. O antigénio STn demonstrouse promotor da invasão celular, da evasão ao sistema imunitário e possivelmente da quimio-resistência, constituindo um bio-marcador importante dos fenótipos celulares mais agressivos de cancro de bexiga. Contudo, o conhecimento acerca dos acontecimentos que promovem esta profunda desregulação dos padrões de glicosilação proteica é escasso. Uma vez que a hipóxia é também uma característica proeminente dos tumores de bexiga avançados, este trabalho visa compreender de que forma os níveis de oxigénio influenciam o glicofenótipo das células tumorais de bexiga, enfatizando a expressão do antigénio STn. Três linhas celulares tumorais de bexiga com backgrounds genéticos e moleculares distintos (T24, 5637 e HT1376) foram submetidas a hipóxia (0.1% O2). Com o objetivo de identificar processos mediados pelo HIF-1α, as experiências foram conduzidas simultaneamente na presença do estabilizador do HIF-1α Mesilato de Deferoxamina (DFX). Em ambas as condições, todas as linhas celulares sobreexpressam o HIF-1α e a AC IX, uma proteína altamente HIF-1α – regulada, bem como aumentam a biossíntese de lactato, denotando a conversão de metabolismo celular aeróbio para anaeróbio. A hipóxia compromete a proliferação celular e potencia a mobilidade/ invasão celular em matrigel. Todavia, não se verificam diferenças significativas na atividade das MMP avaliadas em hipóxia para todas as linhas celulares, sugerindo estratégias alternativas de mobilidade/ invasão. Fizeram-se observações idênticas aquando da exposição das células ao DFX, sugerindo que se tratam de eventos HIF-1αmediados. A análise de um painel de 21 genes associados a fenótipos estaminais, epiteliais, de transição epitélio-mesenquimal e mesenquimal por qRT-PCR revelou que a hipóxia promove a ativação de programas de FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer iv genes de transição epitélio-mesenquimal e estaminais. Adicionalmente, todas as linhas celulares sobreexpressam o antigénio STn de forma HIF-1αdependente. Porém, o aumento dos níveis de STn não está associado à sobreexpressão da principal enzima responsáveis pela sua biossíntese, a ST6GalNAC.I, mas sim com a sub-expressão das enzimas envolvidas na elongação das cadeias O-glicosídicas. A exposição de células em hipóxia ao anticorpo monoclonal anti-STn TKH2 promove um decréscimo dramático na mobilidade/ invasão celular, corroborando um papel chave do STn na disseminação da doença. A associação entre a hipóxia, evidenciada pela sobre-expressão de HIF-1α, e a sobre-expressão de STn foi clinicamente confirmada em tecidos tumorais por imunohistoquímica. No seu conjunto, estes dados indicam que a hipóxia não só favorece a aquisição de um fenótipo mesenquimal mas também compromete a O-glicosilação proteica como uma forma de conferir ás células maior mobilidade/ capacidade invasiva. Por fim, uma análise glicoproteómica preliminar por western blot, sob o ponto de vista da expressão de STn, revelou que células em hipóxia sobreexpressam proteínas STn positivas de baixo peso molecular (<50 kDa) em relação à normóxia, o que poderá permitir atingir seletivamente células em hipóxia prevenindo a disseminação da doença. Em resumo, o presente trabalho reafirma a relevância clinica do antigénio STn no cancro de bexiga musculo-invasivo. Apresenta-se ainda um novo mecanismo, baseado na modulação da O-glicosilação proteica e na expressão de STn, através do qual a hipóxia contribui para a agressividade dos tumores de bexiga. Adicionalmente, este trabalho realça a necessidade de se explorar o glicoproteoma de células hipóxicas visando o desenvolvimento de biomarcadores capazes de melhorar a gestão do cancro de bexiga. Palavras-chave Cancro de bexiga; Hipóxia; Sialil-Tn; Glicosilação e metabolismo; Glicosilação proteica, Glicanas associados a cancro, HIF-1α FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer v Abstract Muscle invasive bladder cancer (MIBC) is amongst the most common and deadliest genitourinary cancers. The mainstay treatment is surgery and cisplatin-based regimens, which fail in avoiding tumour relapse and disease progression, urging for novel biomarkers for accurate patient stratification and new therapeutics. Over 70% of MIBC express the cell-surface tumour-associated carbohydrate antigen Sialyl-Tn (STn) that stems from a premature stop in the O-glycosylation of cell surface proteins. STn was found to be a promoter of cell invasion, immune escape, and possibly chemotherapy resistance, making it an attractive biomarker for aggressive cellular phenotypes. However, there is scarce information about the events underlying this profound deregulation in protein glycosylation. Based on the fact that hypoxia is also a salient feature of advanced stage bladder tumours, this work devotes to understanding how oxygen levels influences the glycophenotype of bladder cancer cells, with emphasis on the STn antigen. Three bladder cancer cell lines with distinct genetic and molecular backgrounds (T24, 5637 and HT1376) were submitted to hypoxia (0.1% O2). In an attempt to determine HIF-1α-mediated events, experiments were also conducted in the presence of the HIF-1α stabilizer Deferoxamine Mesilate (DFX). In both conditions all cell lines overexpressed HIF-1α and its regulated protein CA IX, and increased lactate biosynthesis, denoting a shift to an anaerobic metabolism. Hypoxia also impaired cell proliferation and enhanced motility/invasion in matrigel; however no significant differences were observed in MMP activity in hypoxia for all cell lines, denoting alternative motility/invasion-promoting strategies. Similar observations were made in the presence of DFX, suggesting these are HIF-1α-mediated events. The analysis of a panel of 21 genes associated with stem, epithelial, epithelia-to-mesenchymal (EMT) and mesenchymal phenotypes by quantitative polymerase chain reaction (qPCR) showed that hypoxia led to the activation of EMT/stem cell programs. Concomitantly, all cell lines overexpressed the STn antigen, in an HIF-1αdependent manner. However, STn elevations did not associate with an overexpression of its main biosynthesis enzyme ST6GalNAC.I, but with a significant down-regulation of the enzymes involved in O-glycan elongation. The exposure of hypoxic cells to anti-STn FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer xii Abbreviations BC Bladder cancer BCG Bacillus Calmette-Guérin CA IX Carbonic anhydrase IX CDH1 E-cadherin (epithelial) CDH2 N-cadherin (neuronal) CIS Carcinoma in situ DC Denditic cells DFX Deferoxamine mesylate salt DNA DSBs DNA double-strand breaks DNA SSBs DNA single-strand breaks DPCs DNA–protein crosslinks DSP Desmoplakin EMT Epithelial-mesenchymal transition EPCAM Epithelial cell adhesion molecule FGFR3 Fibroblast Growth Factor 3 FIH-1 Factor Inhibiting HIF-1 FN1 Fibronectin 1 GalNAc N-acetiylgalactosamine residue GAPDH Glyceraldehyde 3-phosphate dehydrogenase FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer xii GlcNAc N-acetylglucosamine residue GlcNAc-T GnT N-acetylglucosaminyltransferases HIF Hypoxia-inducible factor HPRT Hypoxanthine-guanine phosphoribosyltransferase HRAS Harvey rat sarcoma viral oncogene homolog iPSCs Induced pluripotent stem cells IR ionizing radiation KLF Kruppel-like factor gene LIN28A Lin-28 homolog A gene MAPK mitogen-activated protein kinase MET Mesenchymal-epithelial transition MHC Major histocompatibility complex MIBC Muscle invasive bladder cancers MMP Matrix metalloproteinase MYC V-myc avian myelocytomatosis viral oncogene homolog NANOG Homeobox transcription factor NANOG NMIBC Non-muscle invasive bladder cancer OCT-4 (POU5F1) POU class 5 homeobox 1 gene PERK Endoplasmic reticulum kinase PHDs Prolyl hydroxylase domain enzymes PI3K/ AKT phosphatidylinositol-3-OH kinase/ Protein kinase B PI3KCA phosphoinositide-3-kinase ppGalNAc-Ts N-acetylgalactosaminyltransferases FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer xii PTM Post-translational modification pVHL Von Hippel-Lindau protein qPCR Quantitative polymerase chain reaction RB Retinoblastoma RUNX Runt-related transcription factor SNAI Snail family zinc finger SOX-2 SRY (sex determining region Y)-box 2 gene SPARC Secreted protein, acidic, cysteine-rich (osteonectin) ST6GalNac.I GalNAc α-2,6-sialyltransferase or ST6 (alpha-N-acetyl-neuraminyl-2,3beta-galactosyl-1,3)-N-acetylgalactosaminide alpha-2,6-sialyltransferase STn Sialyl-Tn antigen T antigen Thomsen-Friedenreich antigen TAMs Tumour-associated macrophages TB-α α-Tubulin TCC Transitional cell carcinoma TIS Tumours in situ TNM Tumour-node-metastasis staging system TUR Transurethral resection TWIST Twist family bHLH transcription factor UBC Urinary bladder cancer UPR unfolded protein response VHL Von Hippel-Lindau tumour suppressor E3 ligase complex VIM Vimentin FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer xii VNTR Variable number tandem repeat WHO World Health Organization ZEB Zinc finger E-box binding homeobox β3/4 Gal-Ts β3/4-galactosyltransferases β3/4 Gn-Ts N-β3/4-acetylglucosaminyltransferases FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 1 I. Introduction 1.1. Bladder cancer epidemiology and disease information 1.1.1. Epidemiology Bladder cancer (BC) is the most common malignancy of the urinary tract and the 7th most common cancer in men and the 17th in women (1). The incidence of BC increases with age, reaching a peak between 50 and 70 years, and is three times more common in men than in women (2). Environmental factors are thought to play a significant role in BC initiation. Chemical or environmental/occupational exposures and chronic inflammation are well known risk factors for the development of BC and they may lead to genetic and molecular changes, which irreversibly convert a normal urothelial cell to one with a malignant phenotype. Chemical and environmental exposures include aromatic amines, aniline dyes, nitrites and nitrates, acrolein, coal, and arsenic. Other causal factors include indwelling catheters, Shistosomiasis (Schistosoma haematobium infection), and pelvic irradiation (3). Notwithstanding all these factors, in Western world cigarette smoking is the most relevant risk factor, accounting for approximately 50% of BC cases (1).Genetic predisposition also has a significant influence on BC, especially via its impact on susceptibility to other risk factors (4). 1.1.2. Pathophysiology and progression of Bladder Cancer Urothelial cancers arise through two distinct but somewhat overlapping pathways that are driven by different genetic changes: papillary and non-papillary. Approximately 80% to 85% of urothelial cancers are papillary lesions, which arise from hyperplastic epithelium with the expansion of a preneoplastic clone, which shows minimal phenotypic FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 2 deviation from normal urothelium. The continuous growth of such clone results in the development of low-grade superficial papillary tumours that harbour frequent mutations in the Fibroblast Growth Factor 3 (FGFR3) (~70%), Harvey rat sarcoma viral oncogene homolog (HRAS) (30-40%) and phosphoinositide-3-kinase (PI3KCA) genes (5). Nonpapillary and invasive tumours usually arise from severe dysplasia or carcinoma in situ (CIS). In a non-papillary pathway, a successive subclone of the initial hyperplasia develops genetic instability with frequent loss of major tumour suppressor genes as Retinoblastoma (RB) and p53 (~ 50%). The vast majority of invasive bladder cancers occur in patients without a prior history of papillary tumours (6). Macroscopic (frank or gross) haematuria is the commonest presenting sign of BC, occurring in about 75% of patients. More than 90% of diagnosed bladder cancers are transitional cell carcinomas (currently classified as high-grade urothelial cell carcinomas according to WHO guidelines), 5% are squamous cell carcinomas, and less than 2% are adenocarcinomas. Of all newly diagnosed cases of urothelial carcinomas, about 70% are superficial tumours (stages Ta, T1, or tumours in situ [Tis]), but as many as 50–70% of those superficial tumours will recur and roughly 10–20% will progress to muscularis propria invasive disease (T2–4) usually with metastasis (as well as localized persistent disease) within a median of 2 years if managed only by transurethral resection and intravesical therapy (7). Consequently, non-muscle invasive bladder cancer (NMIBC) is a chronic disease with varying oncologic outcomes requiring frequent follow-up and repeated treatments, making the cost per patient from diagnosis to death the highest of all cancers (8). Increasing tumour grade, stage, size and multifocality have been associated with an increased risk of progression. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 3 Figure 1. Transitional cell carcinoma staging by the tumour-node-metastasis (TNM) staging system, in which the stage of the primary tumour (T) is based on the extent of penetration or invasion into the bladder wall (9).Tis, Tumour in situ: ‘‘flat tumour’’; Ta, Non-invasive papillary carcinoma; T1, Tumour invades subepithelial connective tissue; T2, Tumour invades muscle; T2a, Tumour invades superficial muscle (inner half); T2b, Tumour invades deep muscle (outer half); T3, Tumour invades perivesical tissue; T4, Tumour invades any of the following: prostate, uterus, vagina, pelvic or abdominal wall. 1.1.2.1. Diagnosis, management and therapeutics of BC Primary treatment of Ta/T1 transitional cell carcinoma is endoscopic resection of the bladder tumour. Further management is based upon risk factors discussed above and stage/grade of the disease. Depending on these, a patient may need only surveillance, single-installation of intravesical chemotherapy, a course of intravesical chemotherapy or immunotherapy or further surgery. The most commonly used and recommended intravesical therapy is the administration of the live attenuated bacillus Calmette-Guérin (BCG) causing an extensive local inflammatory reaction in the bladder wall. While the precise mechanism of BCG action is unclear, studies have implicated natural killer cells and T lymphocytes as critical mediators of the anti-tumour immune response (10). However, only two thirds of patients respond to BCG and one third of the responders will have recurrent disease. Recurrence after BCG treatment is associated with a poor prognosis (11). FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 4 In patients with low grade Ta disease, the 15-year progression-free survival is 95% with no cancer-specific mortality. Patients with high grade Ta tumours have a progressionfree survival of 61% and a disease-specific survival of 74%, whereas patients with T1 disease have a progression-free survival of 44% and a disease-specific survival of 62%, lending support to the view that invasion of the lamina propria is a prognostic indicator for risk of disease progression and reduced survival (3). Once transitional cell carcinoma invades the muscle of the bladder wall (T2), perivesical tissue (T3) or adjacent pelvic organs (T4), it cannot be controlled by endoscopic resection and intravesical treatment alone (12). The standard surgical approach to treatment of muscle-invasive bladder cancer is radical cystectomy. For patients who undergo total or partial resection of the bladder, treatment with adjuvant (after surgery) or neoadjuvant (before surgery) chemotherapy is also an option (12). Approximately 30% of patients with urothelial cancer present muscle-invasive disease, and about half relapse after radical cystectomy, depending on the pathological stage of the primary tumour and the presence of loco-regional or distant metastasis. Local recurrence accounts for approximately 30% of relapses, whereas distant metastasis are more common. Ten to fifteen percent of patients are already metastatic at diagnosis. Before the development of effective chemotherapy, patients with metastatic urothelial cancer rarely had a median survival that exceeded 3-6 months (13). Advanced bladder cancer may metastasize to lymph nodes (usually pelvic), or via vascular spread to the liver, lung, bone and intestines or other organs. Patients with good renal function may be suitable for systemic chemotherapy with cisplatin-based combinations (methotrexate, vinblastine, cisplatin, doxorubicin, i.e. the MVAC regime or gemcitabine/cisplatin, GC). MVAC and GC prolonged survival up to 14.8 and 13.8 months, respectively, compared to monotherapy and other combinations. Response rates were 46% and 49% for MVAC and GC, respectively. The major difference between the abovementioned combinations is toxicity. The lower toxicity of GC has resulted in it becoming a new standard regimen. Even though chemotherapy is efficient against highly proliferative malignant cells that form the tumor bulk, the five-year overall survival does not exceed 25% and many patients die prematurely from adverse drug reactions (14) urging for effective and safe targeted therapeutics. Modest disease control rates, with sporadic marked chemotherapy responses has led to the investigation of biomarkers for assessment of postoperative prognosis and the potential value of perioperative chemotherapy, and as predictors of response to FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 5 chemotherapy or its monitoring. Most of the biomarkers are associated with tumour angiogenesis. Small studies, usually retrospective, have investigated microvessel density, altered p53 tumour expression, serum vascular endothelial growth factor, urinary and tissue basic fibroblast growth factor, urinary (wild-type and mutant) and tissue fibroblast growth factor receptor-3, and more recently, thrombospondin-1, circulating tumour cells, and multidrug resistance gene expression. More recently, it was also reported that over 75% of MIBC express the sialyl-Tn antigen that stems from profound alterations in the Oglycosylation (linked to Ser/Thr residues) of cell-surface proteins that appears to have implications in cell behaviour towards invasion and migration (15). Although some biomarkers have shown predictive potential, none has sufficient evidence to support its routine clinical use in diagnostics or to develop novel therapeutics. Despite the several studies on pathways associated to invasion and metastasis for prognostic and predictive biomarkers of response to chemotherapy, few have addressed the influence of microenvironment factors. In particular, the role of hypoxia, a known promoter of drug resistance, invasion and metastasis in solid tumours (16), remains poorly understood in the context of bladder cancer. 1.2. Hypoxia: an hallmark of cancer progression and dissemination 1.2.1. HIFs and cellular O2 sensing Hypoxia is defined as a reduction of oxygen tension available to a cell, tissue or organism. At the atmospheric pressure (150 mm Hg), ambient air consists of 21% O2 however in most mammalian tissues O2 percentage is around 2%–9% (on average 40 mm Hg). Hypoxia is usually defined as ≤ 2% O2 and severe hypoxia (or anoxia) is defined as ≤ 0.02% O2. Depending on regional and temporal status of blood flow through tortuous vessels, hypoxia can vary from moderate to severe, acute to chronic, and intermittent to persistent. Low O2 tensions are normally associated to intense inflammation or necrotic regions as well as to less vascularized regions of the bone marrow. In solid tumours, hypoxia results from of an inadequate supply of oxygen, due to exponential cellular proliferation and an inefficient vascular structure. Both acute and chronic hypoxia co-exist FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 12 chemotherapeutic agents in vivo. Some factors include; i) decreased drug action in the absence of O2, ii) decreased effect of agents in hypoxic cells that are poorly proliferating or have altered pH gradients, iii) induction of gene amplification, and iv) an overall decreased drug diffusion and delivery to cells distant from functional vasculature (47). Moreover, HIF1α is able to activate the multidrug resistance 1 (MDR1) gene that encodes a membraneresident P-glycoprotein (P-gp) that belongs to the family of ATP-binding cassette (ABC) transporters. On its turns, P-gp decreases the intracellular concentration of several chemotherapeutic drugs by acting as drug efflux pump. Multidrug-resistance-associated protein 1 (MDRP1) is another ABC transporter related with HIF-1α-mediated drug resistance (28). Earlier studies also reported that transient hypoxia induces DNA overreplication and amplification of a drug resistant gene encoding dihydrofolate reductase. HIF-1-mediated alteration in cell proliferation and survival represent other pivotal causes for drug resistance. The relative importance of each factor might depend on the duration of hypoxic exposure (such as acute versus chronic) (27,32). Genetic instability The cause of human cancers is imputed to the genetic alterations at nucleotide and chromosomal levels of ill-fated cells. It has long been recognized that genetic instability is responsible for the cellular changes that confer progressive transformation on cancerous cells (48). Hypoxia is one of the most important microenvironmental factors that can drive genetic instability and its role in gene amplification, DNA breaks at chromosomal fragile sites and disruption of DNA damage repair is well established (49). Mutation in human DNA mismatch repair (MMR) genes such as MLH1, MSH2, and MSH6 are associated with the development of both hereditary and sporadic cancers and it’s well known that hypoxia/ ischemia impairs the MMR system by inhibiting gene expression (50,51). Moreover, it has been reported that MutSα, MSH2 and MSH6 genes responsible for MMR initiation, are specifically down-regulated by hypoxia via the HIF-1αMyc pathway (52). Hypoxic stress can also suppress the nucleotide excision repair (NER) pathway. It is hypothesized that the unique metabolic conditions induced by hypoxia (e.g., decreased adenosine triphosphate [ATP] production) may impair the enzymatic activity of specific NER proteins, leading to a functional decrease in NER repair under these conditions (53). In addition to genetic changes at the nucleotide levels, chromosomal instability is more commonly observed in cancers. Moreover, cancer-associated hypoxia FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 13 induces additional chromosomal breakage and rearrangement during replication stress at regions called chromosomal fragile sites (54). Therefore, hypoxia-induced DSBs may contribute to a later stage of tumour progression, by inducing chromosomal instability mainly due to impairment of DSBs repair (49). DSBs are mostly repaired by homologous recombination (HR) or non-homologous end-joining (NHEJ) pathways in a cell cycledependent manner. The proteins RAD51, BRCA1/2 and the MRN complex (MRE11, RAD50, NBS1) together regulate HR during S and G2 phases of the cell cycle and proteins such as KU70/80, DNA-PKcs and DNA-ligase IV function in NHEJ across all phases of the cell cycle (55,56). The majority of HR proteins are repressed by chronic hypoxia through decreased transcription, translation, miRNA modulation and epigenetic silencing (57). Altogether HR pathway repression is associated with substantial suppression of recombinational repair activity in hypoxic cells. Other chromatin responses to hypoxia include global deacetylation and methylation of histones. Moreover, histone deacetylases (HDACs), implicated in alteration of chromatin assembly and tumorigenesis, are activated by hypoxia (58,59). Premature condensation of chromosomes and abnormal chromosome mis-segregation are also common features of hypoxic cells (27). Glycolysis and pH regulation As already stated, cells undergo a variety of biological responses towards hypoxic conditions, however the earliest recognized pathway was that hypoxic cells undergo a shift from aerobic to anaerobic metabolism. Metabolic adaptation represents a canonical response to hypoxia, which includes a more or less dramatic shift toward the glycolytic metabolism that enables the sustained, although less efficient production of energy (60). Hypoxic cancer cells use glycolysis as a primary mechanism of ATP production without the need for O2-dependent oxidative phosphorylation, and HIFs are master regulators of glucose metabolism during hypoxia. HIF-1α activates transcription of genes encoding glucose transporters (GLUT1 and GLUT3), which mediate cellular glucose uptake, and other glycolytic enzymes such as lactate dehydrogenase A (LDHA), phosphoglycerate kinase 1 (PGK-1), and hexokinase 1 (HK1) and thus plays an important role in the glycolytic switch (61,62). Hypoxia and HIF-1α increase glucose/carbon flux through the glycolytic pathway while minimizing input into the tricarboxylic acid cycle (TCA) and oxidative phosphorylation. The up-regulation of pyruvate dehydrogenase kinase (PDK1) which in turn deactivates pyruvate dehydrogenase (PDH), the enzyme responsible for FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 14 conversion of pyruvate to acetyl-CoA, redirects glucose away from the oxidative phosphorylation pathway (61,63). HIF-1α has also been shown to collaborate with c-Myc oncogene to active hexokinase 2 (HK2) as well as PDK1, resulting in increased conversion of glucose to lactic acid (64). Moreover, glycolysis also provides substrates for biosynthetic reactions (e.g. nucleotides, amino acids, and lipids) and thereby facilitates cell proliferation in the process of tumour expansion (65). In addition, hypoxia leads to selection of inherently glycolytic cells developed through oncogene-mediated, hypoxia-independent HIF-1α stabilization (66). These are the reasons why tumour cells maintain glycolytic metabolism even in the presence of oxygen and why they do not switch back to oxidative phosphorylation under aerobiosis. Nevertheless, some tumour cells strongly rely on uptake of glutamine and glutaminolysis, which supports the mitochondrial TCA cycle and pentose phosphate pathway and thereby facilitates the synthesis of fatty acids, nonessential amino acids and nucleosides (67). Although HIF-1α strongly links aerobic glycolysis to carcinogenesis, it would be premature to conclude that the glycolytic phenotype in cancer cells is invariably due to deregulation of the HIF system. In addition, glycolysis is thought to be the main mechanism by which tumours lower their pH, through generation of lactic acid. Carbonic Anhydrases (CA), which are metalloenzymes that catalyse the reversible hydration of carbon dioxide to bicarbonate ions and protons, might also be involved. CA IX is one of the 15 human isoforms of the carbonic anhydrase family and it is not expressed in the majority of normal tissues. Actually, it is abundant only in the stomach and gallbladder epithelia. On the other hand, it is very often and strongly expressed in tumours, generally in the more aggressive forms (68). HIF-1α is the essential transcription factor of the CA9 gene driving its transcription in response to intra-tumour hypoxia or inactivating mutation of the VHL tumour suppressor gene. Recently, the hypoxia-induced CA9 transcription was found to be dependent on the cooperation between HIF-1α and an intracellular domain of Notch3 (NICD3), a transcription factor involved in cell-fate determination, morphogenesis, and oncogenesis (69). Importantly, hypoxia also regulates CA9 expression by post-transcriptional mechanisms through splicing and mRNA stabilization. Accumulating experimental evidence supports the direct participation of CA IX in many hypoxiaand acidosis-induced features of tumour phenotype, including increased adaptation of tumour cells to microenvironmental stresses, resistance to therapy, increased tumour cell migration and invasiveness. Moreover, increased focal adhesion during cell spreading, destabilization of intercellular contacts, maintenance of stem cell phenotype, tumour–stroma crosstalk, FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 15 signal transduction and possibly other cancer-related phenomena (22). For example, extracellular acidosis mediated by CA can activate the proteolytic enzymes (some of them being transcriptionally induced by hypoxia) that degrade extracellular matrix and facilitate invasion of tumour cells across the basal membrane and through the surrounding normal tissue (70,71). CA IX also contributes to therapy resistance in a way that tumours expressing high CA IX levels are less responsive to experimental therapy, and inhibition of CA IX catalytic activity significantly improves their chemoor radiosensitivity (68). Hypoxia and extracellular acidosis also contribute to cancer cell dedifferentiation toward the stem cell-like phenotype and CA IX was shown to be involved in this phenomenon in breast (72), bladder (73) and esophageal carcinoma models (73). Expression of CA IX correlates with the expression of a stem cell marker CD44 and CA IX targeting by inhibition of its catalytic activity results in the inhibition of breast cancer stem cell expansion in hypoxia, indicating that pH-regulating function of CA IX is required for the maintenance of stemness (22,74). Figure 3. Genes that are involved in many essential processes are direct HIF-1α targets (75). Four groups of direct HIF-1 target genes particularly relevant to cancer encode: angiogenic factors, glucose transporters and glycolytic enzymes, survival factors and invasion factors. ADM, adrenomedullin; ALDA, aldolase A; AMF, autocrine motility factor; CATHD, cathepsin D; EG-VEGF, endocrine gland-derived VEGF; ENG, endoglin; ET1, endothelin-1; ENO1, enolase 1; EPO, erythropoietin; FN1, fibronectin 1; GLUT1, glucose transporter 1; GLUT3, glucose transporter 3; GAPDH, glyceraldehyde-3P-dehydrogenase; HK1, hexokinase 1; HK2, hexokinase 2; IGF2, insulin-like growth-factor 2; KRT14, keratin 14; KRT18, keratin 18; KRT19, keratin 19; LDHA, lactate dehydrogenase A; LEP, leptin; MMP2, matrix metalloproteinase 2; PFKBF3, 6phosphofructo-2-kinase/fructose-2,6-biphosphatase-3; PFKL, phosphofructokinase L; PGK 1, phosphoglycerate kinase 1; TGF-α, transforming growth factor-α; TGF-β3, transforming growth factor-β3; TPI, triosephosphate isomerase; VEGF, vascular endothelial growth factor; UPAR, urokinase plasminogen activator receptor; VEGFR2, VEGF receptor-2; VIM, vimentin. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 16 1.2.4. Epithelial-mesenchymal transition, hypoxia and acquisition of stem cell traits The epithelial-mesenchymal transition (EMT) is a multistep process characterized by loss of homotypic adhesion and cell polarity in which epithelial cells lose their epithelial characteristics and gain mesenchymal characteristics, such as motility and invasive properties (76). Type 3 or oncogenic EMT occurs in neoplastic cells that have previously undergone genetic and epigenetic changes, specifically in genes that favour clonal outgrowth and the development of localized tumours (77). As already described, tissue invasion and metastasis are prognostic indicators for risk of disease progression and reduced survival in bladder cancer, thus understanding EMT may provide insight for the development of novel tumour markers or new therapeutic strategies. In BC and other malignancies, the hallmarks of EMT are the loss of epithelial markers such as E-cadherin (calcium dependent transmembrane glycoproteins found at adherent junctions and responsible for cell-cell adhesion in epithelial tissues), the increase of matrix metalloproteinases (MMPs) involved in basement membrane degradation, the activation of the Rac/ Rho/Cdc42 small GTPase family implicated in cytoskeleton reorganization, and the nuclear translocation of transcription factors. The down-regulation of E-cadherin is associated with the release of β-catenin, which then migrates to the nucleus and activates WNT signalling, thereby resulting in the EMT and metastasis (78).There are multiple mechanisms resulting in E-cadherin mediated cell adhesion inactivation in cancer, such as “cadherin switching” in which the normal expression of Ecadherin is replaced by the abnormal expression of P-cadherin and N-cadherin or Ncadherin expression is increased and E-cadherin levels do not change significantly (79). However, one of the main mechanisms remains the transcriptional repression mediated by transcription factors such as ZEB-1, ZEB-2 (SIP1), Snail-1, Snail-2 (Slug), and TWIST1 as well as microRNAs (80,81). These transcriptional repressors result in epigenetic silencing of the Ecadherin promoter of E-cadherin-encoding gene (CDH1) by DNA hypermethylation. miR-200 family members participate in the EMT in a way that the loss of miR-200 expression leads to the accumulation of Zeb-1 and Zeb-2, which is sufficient to silence CDH1 and promote EMT and tumour invasion. In BC, expression of P-cadherin and Ncadherin, correlates with late stage, high grade disease and increased expression of MMP-9 was associated with a poor clinical outcome in patients with urothelial tumours. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 17 Additionally, vimentin, an intermediate filament protein that induces epithelial cell changes, including their adoption of a mesenchymal shape and their increased motility, is characteristically up-regulated in cells undergoing EMT and was mainly detected in invasive BC (31% in MIBC vs. 7% in NMIBC) and was positively associated with tumour grade and stage (78). Moreover, EMT also seems to have a role in drug sensitivity of primary patient bladder tumours (44,45). Although both hypoxia and the EMT are separately considered as crucial events favouring the invasion and metastasis of many cancer cells, the two processes are long known to involve few common molecular mechanisms (83). Recently, an increasing amount of evidence demonstrates that alterations in microenvironmental oxygen levels and activation of hypoxic signalling through HIFs are important triggers and modulators of the EMT, which is now thought to take a key role as the convergence point between hypoxia and cancer (84). Under hypoxic conditions, the tumour microenvironment generates and sustains major EMT-triggering pathways for facilitating tumour growth and metastasis such as transforming growth factor (TGF) β and Notch signalling pathway as well as the nuclear factor kappa B (NFκB) pathway (85,86). At the early stages of tumour formation, TGFβ arrests cell proliferation and induces apoptosis. At later stages of tumorigenesis, TGFβ acts as a tumour promoter by increasing tumour cell proliferation, survival, motility and invasion probably by up-regulation of Snail and Slug that are potent initiators of the EMT in cancer cells (78). HIF-1α and TGFβ co-regulate gene products whose expression is increased as a result of the EMT process. Additionally, HIF-1α interacts with the Notch intracellular domain and increases its transcriptional activity. Moreover, the interplay between HIF-1α and Notch appears to be important for stem cell maintenance under hypoxia. Furthermore, NFκB is a key transcriptional activator of HIF1α, and basal NFκB activity is required for HIF-1α protein accumulation under hypoxia in cultured cells and in the livers and brains of hypoxic animals. Finally, there is growing experimental evidence that HIFs modulate the EMT by regulating the expression and activity of major transcription factors including those already referred TWIST, SNAIL, SLUG, SIP1 and ZEB1 (83). Up until now, EMT has been implicated in two of the most important processes responsible for cancer-related mortality: development of distant metastasis and acquisition of therapeutic resistance. Both of these processes may be linked, in turn, to a third: the generation by EMT of cancer cells with stem cell-like characteristics. As described before, hypoxia has the potential to regulate cell differentiation, which has led to the emergence of FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 18 a new paradigm that tumour hypoxia may facilitate the maintenance of cancer stem cell characteristics in a HIF-dependent manner and thus allow a tumour cell with self-renewal potential to accumulate a multitude of genetic and epigenetic changes over a long period of time in order to become increasingly aggressive and resistant (87,88). Within a tumour, stem-like cells reside in defined microenvironments termed “stem cell niches” and hypoxia may be an important feature of those microenvironments. HIF-2α is an upstream regulator of Pou5f1 (Oct4), which is one of the main transcription factors used to generate the first induced pluripotent stem cells (iPSCs). It has been shown that knock-down of HIF-2α but not HIF-1α, leads to a decrease in the expression of Oct4, Nanog and Sox2, which are important stem cells markers (89). In several models of cancer, the induction of EMT potentiates self-renewal and the acquisition of cancer stem-cell properties. Consequently, a common notion is that EMT may be a general feature of cancer stem or progenitor populations. This notion associates local invasiveness with the ability to colonize distant organs as expressions of two tightly interdependent gene programs borne by the same tumour cells. However, other models of neoplasia have found an inverse correlation between local invasiveness and the ability of tumour cells to colonize distant organs. This suggests a dichotomy between these two critical features of the metastatic process, possibly expressed by separate tumour cell subpopulations in which tumour cells that display a strong epithelial phenotype are endowed with the strongest capacity to survive in circulation and to establish distant metastases (76,90). It is important to note that EMT is usually not an irreversible transition, cells can return to their epithelial phenotype once they have reached the metastatic sites for reestablishing barrier function and for enabling cell proliferation in a process called mesenchymal-epithelial transition (MET). Thus, in the absence of the EMT-inducing signals received from the ‘activated’ stroma, metastatic cancer cells might simply fall back to an epithelial state through an MET. For example, E-cadherin expression is re-induced in metastases, a change that is accompanied by the demethylation of the CDH1 promoter (77). However, sustained activation of EMT leads to progressive epigenetic alterations in cells, inducing heritable effects that maintain the mesenchymal state even after EMTinitiating signals are no longer present. Hence, under certain conditions, EMTs can yield stable changes in the phenotype and thus lineage identity of cells (77). FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 19 Figure 4. Invasion of the basement membrane is the defining characteristic of epithelial cancers. A. Epithelial cells are normally constrained by cell–cell contacts and by the basement membrane (BM). B. Cancer cells produce proteases, including the urokinase-type plasminogen-activator receptor (uPAR) and matrix metalloproteinase-2 (MMP2), which digest the basement membrane/extracellular matrix (ECM). C. Degraded ECM is replaced by fibronectin and other ECM proteins that are recognized by integrins that are expressed on cancer cells. D. An epithelial-to-mesenchymal transformation occurs in which intermediate-filament production is switched from keratin subtypes, which are characteristic of fixed epithelial cells, to keratins and vimentin, which promote the fluid structure that is required for motility and which is also stimulated by expression of secreted factors such as autocrine motility factor (AMF) and transforming growth factor-α (TGF-α), and surface receptors such as the c-MET tyrosine kinase. HIF-1 target genes that regulate invasion are listed (75). 1.3. Glycosylation and cancer 1.3.1. General features of glycosylation Glycosylation is the most frequent, complex and plastic posttranslational modification (PTM) of membrane-bound and secreted proteins and results from a coordinated action of nucleotide sugar transporters, glycosyltransferases and glycosidases in the endoplasmatic reticulum and the Golgi apparatus. Glycans play a key role in protein folding, trafficking and stability and act as mediators of cell-cell adhesion, cell differentiation, migration, cell signaling pathways, immune recognition and host-pathogen interactions (91–94). Glycosylation increases the diversity of the proteome to a level unmatched by any other PTM, due to the diversity in sugar compositions, glycosidic linkages, chain length and substitution patterns. Glycosylation is also thought to be the most complex PTM due to the large number of enzymatic steps involved, along with the fact that unlike other cell FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 20 processes, such as transcription or translation, glycosylation is non-templated, and thus, all of these steps do not necessarily occur during every glycosylation event. On the other hand, despite all the heterogeneity and enzymes involved, glycosylation mechanisms are highly-ordered and normally individual enzyme activity is dependent upon the completion of the previous enzymatic reactions. As said before, this PTM is characterized by various glycosidic linkages, including N- , Oand C-linked glycosylation, glypiation (GPI anchor attachment) and phosphoglycosylation (95–99). Concerning the different types of glycosylation, they depend on multiple factors such as enzyme availability, amino acid sequence and protein conformation (100). Focusing on the most common types of glycosylation, two main classes of glycans can be found at the cell-surface, namely N-glycans and O-glycans. Although glycosylation has been so far characterized as a post-translational modification, N-glycosylation often occurs co-translationally during the translation and transport of proteins into the ER. Nglycans are covalently attached to protein asparagine residues by N-glycosidic bonds, of which GlcNAcβ1-Asn is the most common. Precursor N-glycan synthesis begins on the cytosolic face of the ER and is further elongated after the structure is flipped into the ER lumen. In proteins, the candidates for receiving an N-glycan are called Asn-X-Ser/Thr “sequons”, with “X” being any amino acid residue except proline (92). Oligosaccharide transferase (OSTase) scans the nascent protein polypeptides for this consensus sequence and then transfers the precursor glycan (Glc3Man9GlcNac2-) from dolichol pyrophosphate to the Asn residue. To this point, all N-linked glycoproteins have the same precursor glycan structure. Glycan processing to diversify the glycans on individual glycoproteins occurs in the Golgi and combines both trimming and adding sugars to the structures in a step-wise fashion (101–103). Mature N-glycans chains can be modified by the action of fucosyl and sialyltransferases, yielding sialic acids, Lewis (Le) blood group related antigens (Lea, SLea, Lex, SLex, Leb and Ley) or ABO(H) blood group determinants as terminal structures. Other sugar modification may include phosphorylation, O-acetylation of sialic acids and O-sulfation of galactose and N-acetylglucosamine residues, thereby increasing the structural complexity of the glycome. N-glycosylation does not preclude the second most common type of glycosylation, O-glycosylation, from happening, as O-glycosylation commonly takes place on glycoproteins previously N-glycosylated in the ER. O-glycosylation occurs posttranslationally by covalently α-linking a GalNAc moiety from a sugar donor UDP-GalNAc to protein serine or threonine residues and is controlled by UDPGalNAc-polypeptide N- FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 21 acetylgalactosaminyltransferases (ppGalNAc-Ts). As opposed to N-glycosylation there is no consensus sequence for the activity of N-acetylgalactosamine transferase and following the first sugar addition a highly variable number of sugars are consecutively added to the growing glycan chain such as galactose, N-acetylglucosamine, fucose or sialic acid but not mannose, glucose or xylose residues. A second level of complexity in O-glycosylation is the processing of carbohydrate chains by other glycosyltransferases. After the first glycan (GalNAc) is added forming the Tn antigen, the core 1 structure is synthesised by Galtransferase (β(1-3)-galactosyltransferase, C1Gal-T1 or T-synthase) , which adds Gal to GalNAc. The core 1 structure may be also termed T antigen or Thomsen-Friedenreich antigen (Galβ1-3GalNAcα-O-Ser/Thr). Alternatively, Tn and T antigens can be sialylated by sialyltransferases forming the sialyl-Tn, sialyl-T and disialyl-T antigens and the formation of the sialyl-Tn antigen stops any further processing of the oligosaccharide chain. Core 1 may function as a precursor of other core structures (from core 2 to 8), by the addition of different monosaccharides, such as galactose, N-acetylgalactosamine, Nacetylglucosamine and sialic acids. Furthermore, cores 1-4 are the most common in humans. The extension of core units provides a vast array of glycan structures, and is catalysed by N-β3/4-acetylglucosaminyltransferases (β3/4 Gn-Ts) and/or β3/4galactosyltransferases (β3/4 Gal-Ts), leading to the formation of side chains designated type-1 (Galβ1-3GlcNAc-R) and type-2 (Galβ1-4GlcNAc-R) chains. These chains present a ubiquitous expression, and therefore are widely expressed among epithelial tissues. Mature O-glycans may present terminal structures similar to the ones found in N-glycans (100). This type of glycosylation is particularly found and modulates the biological role of mucins, a family of high molecular weight glycoproteins rich in repetitive sequences of serine and threonine residues termed tandem repeat domains (VNTR), which are ubiquitously present in mucous secretions, on cell surfaces as transmembrane glycoproteins and in body fluids. Mucins that span the plasma membrane are known to be involved in signal transduction, to mediate cell-cell adhesion or to have an anti-adhesive function (104). Mucins have also been shown to have roles in fertilization and immune responses. Their presence shield the epithelial surfaces against physical and chemical damage and protects against infection by pathogens. The expression of mucin genes is regulated by a large number of cytokines and growth factors, differentiation factors and bacterial products. Recently, a precision mapping of human O-GalNAc glycoproteome has revealed over 6000 glycosites in more than 600 O-glycoproteins, the majority of which FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 28 As already stated, hyaluronic acid (HA) and its degrading enzyme hyaluronidase (HAase) are intricately involved in tumour growth and metastasis. HA is synthesized by the HA synthases HAS1, HAS2, and HAS3. HAS1 expression in tumour tissues is a predictor of bladder cancer recurrence and treatment failure. HA promotes tumour metastasis and is an accurate diagnostic marker for bladder cancer. Moreover, the measurement of HA and HAase (the HA-HA test) has been applied to the screening of bladder cancer. It has been shown recently that HAS1 regulates bladder cancer growth and progression by modulating HA synthesis and HA receptor levels (149). The galectin protein family was also implicated in bladder cancer. Galectin-1 mRNA levels were markedly increased in most high-grade bladder cancers compared with normal bladder or low-grade cancers. Galectin-3 mRNA levels were also increased in most tumours compared with normal urothelium, but levels were comparable among tumours of different histological grade. Moreover, galectin-7 expression may be related to the chemosensitivity of urothelial cancer (150). Glycosphingolipids (GSLs), including gangliosides (acidic glycosphingolipids which contain sialic acids), interact with specific membrane proteins, such as growth factor receptors, integrins, tetraspanins (TSPs), and non-receptor cytoplasmic kinases (e.g., Src family kinases and small G proteins), to form glycosynaptic microdomains controlling GSLdependent or -modulated cell adhesion, growth, and motility. The glycosphingolipid composition of human bladder cancer tissue has been assessed and the results have demonstrated that large amounts of ganglioside GM3 accumulate in superficial bladder tumours, compared with invasive bladder tumours and that exogenous GM3 inhibits the invasive potential of bladder tumour cells. Furthermore, the GM3 overexpression system was applied to bladder tumour therapy and exogenous GM3 inhibited bladder cancer cell invasion as well as locally injected GM3 (151). Additionally, GM3 synthase gene transfection had an antitumor effect on the murine bladder cancer MBT-2 cell line. It has also been reported that the expression of GM2, GM3, or GM2/GM3 complexes affect cell motility and growth in bladder cancer (152). Finally, increased levels of simple-mucin type O-GalNAc glycans have also been observed in bladder cancer. Several reports associate the presence of Tn and T antigens with recurrence and metastasis suggesting these antigens may be surrogate markers of profound cellular alterations (153,154). There is also growing evidences linking the overexpression of the sialyl-T antigen and ST3Gal.I, the enzyme responsible by T antigen sialylation, with bladder cancer aggressiveness and recurrence (155). Adding to these FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 29 findings, Ferreira et al. has recently demonstrated that over 70% of high-grade NMIBC and MIBC expressed the STn antigen, whereas 80% of low-grade NMIBC and the healthy urothelium do not. The STn antigen was mostly expressed by cells in non-proliferative tumour areas, known for their high resistance to cytostatic agents currently used to improve the overall survival of advanced stage bladder cancer patients. Studies in vitro have further demonstrated that this antigen plays an important role in bladder cancer cell migration and invasion trough mechanisms so far unexplored (15). Also, previous works of our group have demonstrated the presence of STn in metastized ganglia and distant metastasis, suggesting that the expression of STn may reflect on the mobility of cancer cells and the capability to metastasize. Other studies in STn-expressing bladder cancer cells shown that STn has the ability to down-regulate the anti-cancer immune-response through different mechanisms. First, it hinders the expression of MHC-II and co-stimulatory molecules by dendritic cells (DCs), resulting in impaired ability to present cancer-associated antigens to T cells and making DCs unresponsive to successive activation stimuli. Second, it hinders the expression of inflammatory, Th1-inducing cytokines in DCs, which may result in an attenuation of the Th1 microenvironment and reduced ability to activate and polarize T cells towards the Th1 phenotype. Altogether, these results highlight the expression of STn by cancer cells as a crucial event in the establishment of the tolerogenic microenvironment which allows cancers to escape from the attack of innate and adaptive immunity (156). Despite the key role of the STn in bladder cancer progression and dissemination, to this date, the promoters leading its biosynthesis remain yet unknown. Nevertheless, preliminary data from our group is showing a correlation between the levels of STn and hypoxic marker HIF-1 in bladder tumours sections, suggesting that hypoxia might be the missing link responsible by STn expression in bladder cancer. More in depth studies are needed to validate these preliminary observations. 1.3.4. Hypoxic regulation of glycosylation in cancer Glycosylation is kinetically regulated by dynamically changing the portfolio of glycosyltransferases, nucleotide sugars, and nucleotide sugar transporters, which together form a part of what is currently referred to as the “Glycan cycle”. Hypoxic conditions dramatically change gene expression profiles, by activating HIF-1, which mediates FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 30 adaptive cellular responses, which have been extensively addressed in the previous section. As already mentioned, HIF-1α induces the expression of glucose transporters and various types of glycolytic enzymes, leading to shifts in glucose metabolic patterns. This fact strongly suggests that hypoxic conditions are an important factor modulating various nucleotide sugar biosynthetic pathways. Hypoxia-induced glycosyltransferases and nucleotide sugar transporters have also been shown to modulate glycosylation patterns that are part of the mechanism associated with cancer metastasis. For example, hypoxia has an impact in the UDP-GalNAc cycle, as well as in the NAcetylglucosaminyltransferases V (GnT-V) activity by reducing the activity of a portion of the GlcNAc cycle, including intracellular UDP-GlcNAc levels and GnT-V activity in a HIF1α-dependent manner (157). Hypoxia also influences galactose metabolism as exemplified by enhanced expression of genes involved in transport, phosphorylation, and transfer of galactose, and this seems to trigger drastic alteration of other monosaccharides including sialic acid (24). The influence of hypoxia on the metabolism of monosaccharides other than glucose, however, is not fully elucidated. Hypoxia increases the levels of cell surface SLex and Slea determinants in colon cancer cell lines, as well as the transcription of fucosyltransferase VII (FUT7), sialyltransferase ST3Gal-I and UDP-Gal transporter 1 in colon cancer tissues, which are involved in the synthesis of the carbohydrate ligands of endothelial E-selectin (157–159). Some qualitative changes of glycans are also induced by tumour hypoxia. For example, a part of the sialic acid residues in glycans is replaced by N-glycolyl sialic acid in cancers, while normal glycans usually carry N-acetyl sialic acid residues. This turned out to be due to induction of a gene for sialic acid transporter, Sialin, in cancers by tumour hypoxia. For instance, such cancer cells will frequently express N-glycolyl-SLea and N-glycolyl-SLex (24). Among various kinds of glycosaminoglycans (GAGs), hyaluronic acid (HA) and its degrading enzyme hyaluronidase (HAase) are intricately involved in tumour growth and metastasis. Metabolism of hyaluronan is also markedly influenced by hypoxia. Hyaluronan serves as a specific ligand for CD44, and the cell adhesion mediated by the CD44⁄ hyaluronan interaction is heavily involved in cancer cell motility (24). Tumour hypoxia also leads to expression of useful glycolipid tumour markers, such as gangliosides having Nglycolyl sialic acid. Moreover, hypoxia affects not only the glycan moiety of glycolipids, but also their ceramide moiety (160). Nevertheless, despite these findings, no study has FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 31 addressed the influence of hypoxia in the expression of simple mucin-type O-GalNAc glycans in cancer. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 32 II. Background, Aims and Experimental outline Muscle invasive bladder cancer (MIBC) is considered a neglected neoplasia, nevertheless, over 54,000 deaths have been reported in Europe and around 165,100 deaths worldwide in 2012, placing it amongst the deadliest genitourinary cancers (161). Cisplatin-based regimens are the only available therapeutics for invasive and metastatic cases, nevertheless, due to treatment failure, the five-year overall survival does not exceed 25% and many patients die prematurely from adverse drug reactions (2), urging for effective and safe targeted therapeutics. Even though conventional chemotherapy is efficient against highly proliferative malignant cells that form the tumour bulk, non-proliferative tumour areas, characterized by high hypoxia, harbour chemoresistant clones responsible by disease relapse, progression and dissemination. Preliminary findings from our group suggest that these areas express the STn antigen, which stems from a premature stop in protein glycosylation. This epitope has been found to favour disease dissemination and is also responsible by generating distinct protein signatures at the cell-surface, providing means to target aggressive cells. Nevertheless, the association between these STn expression and hypoxia remain yet to be elucidated. In the present work, hypoxia and STn are the main targets explored, especially because changes in the glycosylation of cell-surface and growth under oxygen deficiency (hypoxia) are salient features of solid tumours that often correlate with advanced stages of malignancy such as invasion and metastasis. As shown by several studies, hypoxia induces a wide range of biological changes, such as decreased cell proliferation, increased expression of drug-resistance genes, selection of apoptosis-resistant clones, facilitation of tumour invasion and metastasis, reduced expression of DNA repair genes, and increased genomic instability. These mechanisms undoubtedly contribute to the evolution of malignant tumour cells. However, it remains to be fully understood why hypoxic tumour cells tend to be more aggressive and more resistant to treatment than nonhypoxic tumour cells within the same tumour, despite their similar genetic background. Furthermore, there is a lack of specific hypoxia-associated cell-surface biomarkers to guide therapeutics to these cell. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 33 Based on this background, the present work aims to explore a possible association between STn expression and hypoxia, envisaging theragnostic biomarkers capable of improving bladder cancer overall survival. The project experimental outline includes the characterization of the effect of hypoxia on three bladder cancer cell lines with different molecular and genetic profiles regarding the activation of stem-cell and/or epithelial-to-mesenchymal transition programs, degree of proliferation, invasion capacity and STn expression, followed by validation in human tumour samples. Studies including the hypoxia-mimetic Deferoxamine Mesilate that stabilizes HIF-1α through the inhibition of Prolyl Hydroxylases (PHDs) activity by the chelation of Fe2+, will also be included to disclose HIF-1-mediated alterations. The generated information is regarded of primary importance to expand the knowledge about the clinical relevance of the STn antigen in bladder cancer and create the rationale for a STn-based therapy. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 34 III. Materials and Methods 3.1. Cell Lines This study was conducted on three urinary bladder cancer cell models, namely T24, 5637 and HT1376 comprising the main genetic alterations commonly found in human urinary bladder cancer (UBC), which allows for the study of the most representative human bladder tumours. The accumulated data have shown that FGFR3 mutations are characteristic for low grade and low stage tumours whereas p53 and RB1 mutations are characteristic for invasive tumours. This has led to the suggestion that UBC develop through at least two molecular pathways, one related to FGFR3 and other related to p53 and RB1 alterations (162). The cell lines used have different histological origins, the 5637 cell line is derived from a grade II carcinoma, the HT1376 cell line from a grade III carcinoma and the T24 cell line from a transitional cell carcinoma (TCC). 5637 and HT1376 represent the E2F3/RB1 pathway with loss of one copy of RB1 and mutation of the remaining copy. Additionally, HT1376 exhibit deletion of PTEN gene and no alteration of PIK3CA, which in combination with the inactivation of p53 grants a more invasive and metastatic potential to this cell line. In contrast, the 5637 cell line does not present any loss of PTEN and loses PIK3CA gene, which gives it a less-invasive phenotype. The T24 cell line belongs to the alternative pathway of FGFR3/CCND1 by presenting a mutated HRAS and over-represented CCND1 (cyclin D1) (163). 3.2. Cell culture conditions Cells were cultured in RPMI Medium 1640 (1X) + GlutaMAXTM-I (Gibco ®; Life Technologies) supplemented with 10% heat-inactivated FBS (Gibco ® Life Technologies) and 1% Pen Strep (10,000 Units/mL Penicilin, 10,000 µg/mL Streptomycin) (Gibco ® Life Technologies). Cell lines were cultured as a monolayer at 37 ºC in a 5% CO2 humidified atmosphere (normoxia), and were routinely subcultured after trypsinization. The cells were also grown under hypoxic atmosphere for 6, 24, 48 and 72 hours (as further indicated) in a BINDER C 150 incubator with 0.1% O2, 5% CO2 and the balance N2. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 35 Additionally, cells were exposed to 500 µM Deferoxamine Mesilate CRS (DFX, SigmaAldrich) to stabilize HIF-1α. All experiments were performed at 70-80% of cell confluence. 3.3. L-Lactate assay A fluorometric L-Lactate assay kit (Abcam) was used to determine the concentration of the intermediary metabolism product L-Lactate in culture media. Lactate is produced in proliferating cells under anaerobic conditions, with L-lactate being the major stereo-isomer formed. In Abcam’s L–Lactate Assay Kit, lactate is oxidized by lactate dehydrogenase to generate a product which interacts with a probe producing color (570 nm) and fluorescence (at Ex/Em= 535/587nm). The reaction product fluorescence was measured at Ex/Em=535/590nm in a microplate reader (SynergyTM Mx, BioTek). 3.4. Cell Proliferation Assay A colorimetric BrdU cell proliferation ELISA kit (ab126556 Abcam) was used to estimate cell proliferation. The test allows a quantitative measurement of the incorporation of Bromodeoxyuridine (5-bromo-2'-deoxyuridine, BrdU), a synthetic nucleoside analogue to thymidine, into newly synthesized DNA of actively proliferating cells. Briefly, T24 (0.35x105 cells/mL), 5637 (0.15x105 cells/mL) and HT1376 (0.25x105 cells/mL) cells were cultured in 96 well plates and BrdU was added to the wells during the final 24 hours of culture. To enable antibody binding to incorporated BrdU, cells were fixed, permeabilized and the DNA denatured. This was all done in one step by treatment with a kit provided Fixing Solution. Subsequently, the anti-BrdU monoclonal antibody was pipetted into the wells and allowed to incubate for one hour, binding to any incorporated BrdU. Unbound antibody was washed and horseradish peroxidase-conjugated goat anti-mouse antibody was then added. The horseradish peroxidase catalysed the conversion of the chromogenic substrate tetra-methylbenzidine (TMB) from a colourless solution to a blue solution (or yellow after the addition of stopping reagent), the intensity of which is proportional to the amount of incorporated BrdU in the cells. The coloured reaction product was quantified using a single absorbance read at 450 nm by a microplate reader (SynergyTM Mx, FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 36 BioTek) allowing the assessment of the population of cells which are actively synthesizing DNA. 3.5. Cell Invasion Assay Invasion assays were performed using BD BioCoat MatrigelTM invasion chambers, comprised by an 8 µm diameter pore size filter coated with a thin layer of Matrigel. Prior to each experiment, filters were re-hydrated in serum-free RPMI medium for 1 h at 37 ºC. After detachment of confluent cells by trypsinization, cells were suspended in complete culture medium, counted and seeded on the upper side of the matrigel-coated filter at a density of 5x104 cells/ well. After incubation for 24h at 37 ºC under normoxic or hypoxic conditions, invasive cells were fixed with 4% Paraformaldehyde and non-invading cells, present on the upper side, were completely removed, to facilitate analysis. Cells that invaded the underside of the filters were mounted in Vectashield Mounting Medium with DAPI (Vectashiels®, Vector Laboratories), and visualized through a Leica DM2000 fluorescence microscope (Leica Microsystems). Invasive cells were scored in at least 12 microscopic fields (20x objective). 3.6. Flow Cytometry analysis STn expression was determined by flow-cytometry. Approximately 106 cells/ml were seeded in 6-well plates and allowed to adhere for 24h. Subsequently, the medium was removed and the cells were repeatedly washed with PBS and dissociated into single-cell suspensions with Versene 1:5000 (1X) (Gibco) at 4 ºC, followed by filtration using a 70 µm Nylon cell strainer (BD Falcon). Detached cells were fixed with 4% Paraformaldehyde, suspended in 500 µL PBS and the anti-STn primary antibody was used according with table I indications. Finally, the cells were incubated with polyclonal rabbit anti-mouse Immunoglobulins/FITC (1:84) secondary antibody. After being washed, the cells were analysed by flow cytometry in a BeckMan Coulter FC500 Cytometer. Each independent experiment was performed in triplicate. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 43 where it plays a pivotal role in targeting HIF-1 for proteasomal degradation, was used in an attempt to highlight HIF-1-mediated events. Western blot data showed HIF-1α protein has two bands at approximately 92 and 110 kDa (Figure 5A). The two bands for HIF-1α may result from post-translational modifications, splice variants or isorforms or multimer formation. Recent studies suggest that phosphorylation of HIF-1α by mitogen-activated protein (MAP) kinases is a common modification that increases its transcriptional activity independently of the effects of hypoxia, but the site(s) of phosphorylation and the mechanism by which transcriptional activity is enhanced have not been established yet (166). Additionally, splice variants of HIF-1α created by alternative splicing of the same gene, lacking several exons or displaying different exons than the wild type protein have already been reported and also may account for the two bands pattern for HIF-1α protein (167). Some of these isoforms encode cytoplasmic HIF-1α protein or proteins with altered transcriptional activity compared to the wild type protein. Regarding the N-terminal domain, two HIF-1α isoforms with different first exons have been identified: HIF1α1.2, a protein with a different first and second exons which is 59 amino acids shorter than wild type HIF-1α and encodes a cytoplasmic protein specifically expressed in the human testis (168); HIF1α1.3, which is present in activated T-lymphocytes and encodes a functional protein with weaker transcriptional activity that the wild type protein (169). At the C-terminal domain, isoforms lacking either exon 12 (170) or exons 11 and 12 (171) have been reported. Both isoforms have been shown to be stable cytoplasmic proteins and inhibit the function of full length HIF-1α. Another shorter HIF-1α isoform lacking exon 14 has also been reported (172). This isoform was shown to be 3-fold less active than full-length HIF1α. Furthermore, multimers may also contribute to this feature. Although multimers are usually prevented in reducing conditions, strong interactions can result in the appearance of higher bands. Altogether, post-translation modifications, splice variants and multimer formation contribute to the two bands pattern, and for quantification purposes the sum of the two consecutive bands was taken into account. The identification of the proteins in these bands by mass spectrometry should, in the future, allow confirming these hypotheses. The CA IX and TBα proteins appear in a single band pattern and have observed band sizes of 55 kDa and 50 kDa, respectively (Figure 5A). Figure 5B shows that for T24 and 5637 cell lines, HIF-1α expression is higher at 24h, while for HT1376 cell line it is significantly increased at 6h in hypoxia and DFX conditions in relation to normoxia. The CA IX proteins follow the same tendency as HIF-1α (Figure FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 44 5B), in accordance with the fact that the transcription of this protein is strongly regulated by HIF-1. An increase in the exposure time of T24 and 5637 cells to hypoxia and DFX (48 and 72h) results in a decrease in HIF-1 and CA IX protein levels. Such differences may reflect a response shift from acute hypoxia (≤ 24h) to longer-term or chronic hypoxia (> 24h), which has been described to affect the transcription and translation of several proteins that might indirectly influence the expression of these markers (22). In addition, miRNAs may also directly influence HIF-1α responses to prolonged hypoxia. As part of the RNA-induced silencing complex (RISC), miRNAs negatively regulate gene transcription by annealing to the 3′ untranslated region of specific mRNA targets to repress translation, enhance mRNA degradation, or both (26,173). Hypoxia initially induces increased expression of HIF-1α, but during sustained hypoxia, the amount of HIF-1α mRNA is reduced. Instead, there is considerable overexpression of the HIF-1α natural antisense transcript (aHIF), which encodes the antisense template of the 3’-untranslated region of HIF-1α mRNA. This represents a negative feedback loop, in which aHIF inhibits the translation of HIF-1α mRNA after chronic hypoxia, clarifying the pattern observed (174,175). Furthermore, hypoxia has been shown to induce the expression of a number of other miRNAs, which have been termed “hypoxamirs” that regulate HIF-1α expression (17,21,105,106,176) Finally, the molecular differences observed between cell lines overtime might be explained by different genetic characteristics of each cell line. In tumour cells, loss of p53 activity results in increased HIF-1α expression and increased transcription of downstream target genes such as CA IX. Furthermore, PTEN negatively regulates the PI3K pathway and, therefore, loss of PTEN activity leads to increased HIF-1α expression. So, PTEN mutations might promote tumour growth by synergistically promoting HIF-mediated responses (20). In HT1376 cells, PTEN gene is deleted which in combination with the inactivation of p53 potentiates expression of HIF-1α. This particular features may explain the reason why HT1376 cell line shows significant levels of HIF-1α protein after only 6h of exposure to hypoxic conditions in comparison with T24 and 5637 under the same microenvironmental challenge. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 45 Figure 5. Urinary bladder cancer cell lines exposed to hypoxia and hypoxia mimetic DFX show differential expression of hypoxia markers over time. All cell lines were cultures for 24h before exposure to any experimental conditions. A. Western Blot analysis of HIF-1α and CA IX proteins, 24h (T24 and 5637 cell lines) and 6h (HT1376 cell line) after treatment. TB-α was used as loading control. Molecular weight markers are shown to the right and are expressed in kDa. The western blot samples appear in the following order: T24, 5637 and HT1376 Normoxic conditions (N); T24, 5637 and HT1376 Hypoxic conditions (H); T24, 5637 and HT1376 DFX treatment (D); B. Protein expression time course of HIF-1α and CA IX was measured by western blot analysis. Graphs represent average value of at least three independent experiments, flags correspond to SD and ***p < 0.001; **p < 0.01; *p < 0.05 The metabolic shift of the cells from aerobic to anaerobic metabolism, a critical event underlying hypoxia, was also confirmed by increased lactate levels in hypoxic cells culture mediums (Figure 6). Similar behaviour was also observed upon DFX exposure, which suggests that HIF-1 may be the key mediator for these changes. Hypoxic tumour cells primarily use glucose for glycolytic energy production and release lactic acid, creating a lactate gradient that mirrors the oxygen gradient in the tumour. Although lactate is generally considered a waste product, Sonveaux et al. showed that it is a prominent FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 46 substrate that fuels the oxidative metabolism of oxygenated tumour cells (177). There is therefore a symbiosis in which glycolytic and oxidative tumour cells mutually regulate their access to energy metabolites. The same authors further identified monocarboxylate transporter 1 (MCT1) as the prominent path for lactate uptake by a human cervix squamous carcinoma cell line that preferentially utilized lactate for oxidative metabolism (177). The expression of MCT1 will also be assessed in the future in the bladder cancer cell models. Interestingly, HT1376 cell line shows significantly elevated basal levels of lactate in normoxic conditions in comparison to the other cell lines (Figure 6). A number of studies have shown that malignant transformation is associated with an increase in glycolytic flux and in anaerobic and aerobic cellular lactate excretion, which is consistent to the basal levels of lactate in normoxic conditions for all cell lines (178). Nevertheless, other mechanisms could account for the HT1376 particular case. The HT1376 elevated basal levels of lactate may correlate with elevated basal levels of HIF-1α, since lactate also stimulate the accumulation of HIF-1α, independently of hypoxia (27). In various tumour types investigated, high concentrations of lactate were correlated with a high incidence of distant metastasis already in an early stage of the disease. Moreover, recent reports show various biological activities of lactate that can enhance the malignant behaviour of cancer cells. These mechanisms include the activation of hyaluronan synthesis by tumourassociated fibroblasts, up-regulation of vascular endothelial growth factor and HIF-1α itself, and direct enhancement of cellular motility that generates favourable conditions for metastatic spread. Thus, lactate accumulation not only mirrors but also actively enhances the degree of tumour malignancy, which correlate to the characteristics of grade III carcinoma HT1376 cells. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 47 Figure 6. When exposed to hypoxic conditions all UBC cell lines actively switch to an anaerobic metabolism. All cell lines were cultures for 24h before exposure to any experimental conditions. After a hypoxia (0.1% O2) or DFX (500µM) exposure period of 24h (T24 and 5637) and 6h (HT1376), lactate levels of the cultures conditioned mediums were measured. Graph represents average value of three replicates, flags correspond to SD and ***p < 0.001; **p < 0.01; *p < 0.05. Altogether, bladder cancer cell lines T24 and 5637 overexpress hypoxia biomarkers HIF-1α and CA IX after 24h of exposure to oxygen deprivation, while for HT1376 this effect was already observable at 6h. Concomitantly, these cells shifted towards an anaerobic metabolism. The stabilization of HIF-1α with DFX resulted in similar behaviours suggesting that this transcription factor might regulate these events. 4.2. Morphological characterization of urinary bladder cancer cells In normoxic conditions T24 cells show a heterogeneous appearance, i.e. cells on the periphery of the islands mostly present a polyhedral morphology but peripheral elongated cells are also present. T24 cells are pleomorphic and have more centroid nucleus than the other cell lines (Figure 7A). The growth of T24 cells in tissue culture was characterized by a disorderly pattern of growth in one or more layers and by mixed epithelioid-fibroblastoid morphology. These observations are consistent with the original characterization of T24 cell line (128,129). In the course of hypoxic treatment, cells acquire a more elongated semblance as well as a larger intercellular space between them. These morphological T24 5637 HT1376 0.00 0.02 0.04 0.06 0.08 0.5 1.0 1.5 2.0 2.5 Normoxia Hipoxia DFX ** *** * *** *** *** Lactate (ng/g Protein) FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 48 changes are more pronounced during DFX treatment, suggesting this might be an HIF-1αmediated phenomena. Concerning the 5637 cell line, in normoxic conditions 5637 cells present polyhedral morphology with the nucleus in a lateral position. Even in the inner cells of the island the nucleus is not the centroid of the cell. Peripheral normoxic cells also present extensive areas of membrane ruffling (MR). In hypoxic conditions polyhedral and elongated cells are present as well as membrane ruffling. This morphology can be considered as intermediate between normoxic and DFX conditions. When subjected to DFX treatment, cells do not form cohesive but dispersed islands with hook-shaped cells (HSC) in the extremities, which along with increased intercellular distances suggests possible cell scattering (Figure 7B). The 5637 cell line scattering has already been reported by other authors in other conditions than hypoxia (130). Normoxic HT1376 cells present polyhedral morphology with peripheral nucleus and grows in islands. The existence of giant cells, large vacuoles and cytoplasmatic granules could be noticed in cultured cells. These cells have an extensive membrane area without membrane ruffling. Intercellular spaces are evident and emphasized with higher amplification, highlighting the cell junctions (CJ) (Figure 7C). When subjected to hypoxia and DFX treatment, HT1376 cells have a very similar appearance to that observed in normoxic conditions, suggesting these cells retain their original morphology under oxygen privation or that 6h treatments may not be enough to see morphological changes. The epithelial morphology of HT1376 cell line and the presence of citoplasmatic vacuoles were also reported by other authors (131). FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 49 Figure 7. Hypoxic conditions modify morphological traits of urinary bladder cancer cell lines. All cell lines were cultured for 24h before exposure to experimental conditions. T24 (A) and 5637 (B) cells were then exposed to 500 µM DFX or 0.1% O2 for 24h, while HT1376 (C) cells were exposed to the same conditions for 6h. Images were acquired using a high-resolution inverted microscope. CJ, cell junctions, HSC, hook-shaped cells, MR, membrane ruffling; V, vacuole. Scale bars correspond to 100 and 50 µm. 4.3. Hypoxic modulation of EMT and Stemness As previously described, the epithelial-to-mesenchymal transition (EMT) is a process by which epithelial cells lose their polarity and are converted to a mesenchymal phenotype, which is regarded as a critical event in morphogenetic changes during embryonic development, wound healing and cancer metastasis (82).The hypoxic microenvironment emerges as an important factor in the induction of pathological EMT, which is a key link in cancer progression (83). According with this observations, to explore the events mentioned above in UBC models, the cells were further characterized in relation to the expression of panel of 21 genes associated with stem cell (NANOG, LIN28A, POU5F1, KLF9, KLF4, SOX2), epithelial (CDH1, DSP, EPCAM), epithelial-tomesenchymal transition (SNAI1, SNAI2, TWIST1, TWIST2, ZEB1, ZEB2, RUNX1, RUNX2), and mesenchymal (FN1, CDH2, VIM, SPARC) programs. The GAPDH gene was FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 50 selected as reference gene, since it was expressed at constant levels in all experimental condition and all cell lines. According to the map in Figure 8A, under normoxia the T24 and 5637 cells presented a significant down-regulation of epithelial markers (CDH1, EPCAM and DSP) accompanied by an up-regulation of mesenchymal genes (CDH2, FN1, SPARC, and VIM) when compared to HT1376. The up-regulation of mesenchymal markers is particularly evident in the 5637 cell line whereas the T24 cell line showed a more marked stem cell character. A comparison between these two cell lines highlighted a significant (p ≤ 0.005) down-regulation of KLF4, CDH1, DSP and CDH2 and up-regulation of NANOG, LIN28A, TWIST1 and ZEB2. An integrative analysis of these differences using ClueGo and CluePedia showed that T24 cell line presents a significant up-regulation of pathways involved in stem cell development and also a profound deregulation of cell-cell adhesion, translated by a negative regulation of cell-cell contact and adherent junction organization, suggesting a higher migration potential. In resume, gene expression analysis highlights that both T24 and 5637 present a pronounced mesenchymal properties, with T24 exhibiting a more pronounced stem nature. On the other hand, the HT1376 cell line showed a marked epithelial nature which was not found in the other two cell lines. Noteworthy, is that 5637 cells are close related with HT1376 from the genetic point of view (163); however this cell line was found more similar to T24 based on the studied gene expression patterns. Cells were then grown under hypoxia and in the presence of DFX to determine the influence of oxygen and HIF-1α on the gene expression pattern. According to Figure 8B, T24 and 5637 presented similar variations in terms of genes expression from hypoxia to normoxia, while HT1376 had a distinct gene expression pattern, like previously observed in normoxia. The most string difference between the cell lines was the up-regulation of epithelial-to-mesenchymal and mesenchymal markers (6/11 studied genes) in HT1376, which was not evident in the other cells that already presented this phenotype in normoxia. However, both similarities and differences could be observed between the cell lines based on the most significant gene variations (p ≤ 0.005) as highlighted by the Venn diagram in Figure 8C. In particular, at the 95% confidence level, T24 and 5637 cells significantly down-regulated SPARC and SNAI1 while overexpressing SNAI2 and RUNX1. On the other hand, T24 and HT1376 cells shared the up-regulation of FN1 and KLF4, while 5637 and HT1376 both up-regulated ZEB2. The T24 cells distinguished from the other cell lines by also down-regulating KLF9 while up-regulating RUNX2. An increase in POUF1 was FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 51 only observed in 5637 and TWIST1 overexpression was only found in HT1376. The integrative analysis of these profiles using ClueGo and CluePedia (Figure 9) showed that T24, 5637 and HT1376 cells in hypoxia presented a positive regulation of stem cell development, i.e. up-regulation of any process that increases the rate, frequency or extent of stem cell maturation, which does not include the steps involved in committing a cell to a specific fate. The T24 cell line also showed up-regulated stem cell differentiation processes, which comprehends events that drive relatively unspecialized cells to acquire specialized features of a stem cell. In addition, 5637 and HT1376 cells presented an upregulation of biological pathways committed to mesenchymal development. Altogether, this data suggests that hypoxia drives bladder cancer stem cell establishment and mesenchymal phenotypes. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 52 Figure 8. Heatmaps showing normoxic (A) and hypoxic (B) modulation of the transcript levels of a small group of EMT, Stem and mesenchymal markers. All cell lines were cultured for 24h before exposure to any experimental conditions. After hypoxia (0.1% O2) or DFX (500µM) exposure period of 24h (T24 and 5637) or 6h (HT1376), the transcript levels of a 21 genes tailored panel were quantified by qRT-PCR C. Venn diagrams showing statistical significant data, focusing on overlapping patterns of gene expression in each cell line. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 59 (189). On the other hand, DFX-induced apoptosis is also mediated by the p38 pathway and a caspase-8-dependent Bid-Bax pathway (190). Viable but hypoxic and non-proliferating tumour cells are of particular interest, because it is presumed that they comprise a substantial fraction of the cells in solid tumours and could be the major source of chemotherapy resistant clones. Figure 12. Oxygen deprivation and DFX treatment have in vitro growth-inhibitory effects on bladder cancer cells. All cell lines were cultures for 24h before exposure to any experimental conditions. BrdU was added to the cultured cells medium and after a hypoxia (0.1% O2) or DFX (500µM) exposure period of 24h (T24 and 5637) and 6h (HT1376), the BrdU incorporation was assessed. Graph represents average value of three replicates, flags correspond to SD and ***p < 0.001; **p < 0.01; *p < 0.05. 4.6. Invasion and proteolytic activity of urinary bladder cancer models As previously described, a critical event in tumour cell invasion is degradation of the extracellular matrix (ECM), a complex network of extracellular macromolecules such as collagen, proteoglycans, fibronectin, laminin and many other glycoproteins that act as a barrier to the invasion of cancer cells. Although several different proteases are implicated in ECM degradation, a special group of metalloproteinases, matrix metalloproteinases (MMPs), a family of zinc and calcium-dependent proteolytic enzymes, have been given special attention in this work, mostly because of their important roles in several cancersupporting cellular processes besides extracellular matrix (ECM) remodelling, such as T24 5637 HT1367 0 1 2 3 4Normoxia Hypoxia DFX ** *** *** *** * *** Proliferation (Abs 450nm) FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 60 angiogenesis, apoptosis, epithelial-to-mesenchymal transition and cell proliferation (148). Moreover, hypoxia has been described to potentiate the migration/invasion of cancer cells, which is a critical step for disease progression and dissemination. Based on these observations, the three bladder cancer cell lines were first characterized in relation to their migration/invasive potential on matrigel. In normoxia T24 and 5637 cells presented higher invasion than HT1376; however exposure to hypoxia increased this property in all cell lines in comparison to normoxia (Figure 13A). Conversely, growth under oxygen deprivation in the presence of an anti-STn antibody significantly lowered invasion, reinforcing our previous observations supporting a role for STn in bladder cancer invasion (Figure 13B) (15,120). A gelatin zymography assay (Figure 13C) was then used to evaluate if these observations stemmed from higher MMPs activity or were a consequence of the higher cell motility presented by cells in hypoxia. First it was observed that MMPs activity was similar between the three cell lines grown under normoxia (Figure 13D), leading to conclude that T24 and 5637 cells crossed matrigel due to their higher migratory capacity. Such observations further reinforce the mesenchymal character of these cell lines previously suggested by morphological and gene expression analyses. Exposure to hypoxia did not promote a statistically significant increase in MMPs activity, again reinforcing a main role for motility in the context of bladder cancer hypoxic cells dissemination (Figure 13D). Altogether, these findings show that hypoxia enhances cell motility irrespectively of their morphological and genetic background in, what appears to be, a STn-expression dependent process. The ClueGo and CluePedia analysis, revealed in all cell lines a decreased rate, frequency or extent of cell adhesion molecule production by GCNT1 down-regulation, further reinforcing that these events contribute to negatively regulate cell adhesion. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 61 Figure 13. Hypoxia influences invasive ability but not MMP proteolytic activity in urinary bladder cancer cells. A. The invasive cells from 24 h hypoxic cultures and aerobic control cultures were visualized through a Leica DM2000 fluorescence microscope and scored in at least 12 microscopic fields (20x objective) when DAPI-counterstained nuclei passed through the filter pores. B. A 24h invasion assay was conducted under oxygen deprivation conditions in the presence of the anti-STn antibody TKH2 and the number of invasive cells visualized as previously described. C. Representative Gelatinase zymogram of bladder cancer cells show two bands, one corresponding to pro-MMP-2 (72 kDa) and other corresponding to pro-MMP-9 (92 kDa). MWM are shown to the left and are expressed in kDa. D. Bar graph showing greater expression of pro-MMP-2 over pro-MMP-9 in all cell lines with no significant changes between normoxic and hypoxic conditions. Graphs represent average value of three different experiment, flags correspond to SD and ***p < 0.001; **p < 0.01; *p < 0.05. 4.7. STn glycoproteomic profilling The STn antigen is a post-translational modification that may be putatively expressed by all glycoproteins at the cell surface (105) as translated by the smear in the western blots presented in Figure 14. However, two high-molecular weight bands at approximately 250 and 200 kDa and four low molecular weight bands at 50, 37, 25 and 15 kDa are distinguishable for all the cell lines and experimental conditions (normoxia, hypoxia, DFX; Figure 14A and 14B). The specificity of STn expression was confirmed by the disappearance of the bands upon treatment with a -neuraminidase (Figure 14C). The relative quantification of the main bands highlighted a common profile for 5637 and FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 62 HT1376 cell lines in hypoxia, characterized by an increased expression of all evaluated proteins in relation to normoxia (Figure 14D). Similar behaviour was observed in the T24 cell line, but only for low MW bands. The exposure to DFX appears to modulate the overexpression of STn glycoproteins in T24, specifically for low molecular weight bands. However, for 5637 and HT1376, DFX only promoted the overexpression of high MW bands observed, again denoting a common behaviour at the protein level for these cell lines. The different profiles presented by the cells in hypoxia and DFX also suggest that not all glycoprotein STn expression is mediated by HIF-1; however these findings warrant more in depth evaluation. Altogether, these findings demonstrate that exposure to hypoxia translates into an alteration in the glycoprofile of bladder cancer cells, that may be explored to selectively target these cells. More studies are necessary to identify the STnexpressing proteins and to determine if the alterations found by flow cytometry and western blot are influenced solely by alterations in glycosylation pathways and/or by the relative abundance of the glycoproteins carrying this antigen. This information is expected to translate into further understanding on the role of glycosylation on the biological and clinical behaviour of bladder cancer. Nevertheless, based on previous findings, the high MW bands are concordant with the presence of mucins. Namely, MUC1, a high-molecular weight transmembrane protein previously reported to be heavily O-glycosylated and one of the main carriers of STn antigen, which lead us to hypothesize that the intense staining band at 250 kDa present in the blot refers to MUC1 mucin (127,191,192). Other mucins were reported to have a roll in bladder cancer such as MUC2, MUC4 and MUC7, which may account for more than one high MW bands (193,194). MUC1 is expressed on the apical surface or in umbrella cells of the normal non-neoplastic bladder urothelium and strong expression of MUC1 was also observed in urothelial carcinoma. Moreover, according to the literature, it is suggested that high levels of expression of sialylated MUC1 are associated with an aggressive phenotype, which together with its hypoxia-mediated overexpression may contribute to a more malignant phenotype of 5637 and HT1376 cell lines (195). Since, the decrease in high MW proteins is accompanied by an increase in low MW proteins in T24 cell line, one can hypothesise that the low MW species observed could result from the proteolysis of high MW proteins/ mucins. This hypothesis lines with the identification of a pathway of EGF (epidermal growth factor)-dependent metastasis that requires a Src (tyrosine-protein kinase)-mediated MUC1 proteolysis (196,197). The authors report that EGF stimulation induces MUC1 cleavage, leading the MUC1 cytoplasmic domain transmembrane subunit FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 63 (MUC1.CD) (15kDa) to translocate to the nucleus, where it promotes the expression of a metastatic gene signature associated with epithelial-to-mesenchymal transition. Those results showed that EGFR and Src activity, and consequently proteolytic cleavage of MUC1, contribute to tumour metastasis. Nevertheless, these observations need to be validated in a large number of patients and efforts should be conducted to identify these glycoproteins and disclose their contribution to malignancy. Other works correlate STn expression with invasion and metastatic potential in bladder cancer which associated with this observations could suggest an important role of sialylated MUC1 in the metastatic potential of our models as well (15).The other low MW bands might be a result of similar cleavage processes of high MW mucins (198,199). As previously described, 5637 and HT1376 cell lines belong to the same molecular pathway of bladder cancer carcinogenesis which may account for the similarities in STn pattern even in response to hypoxia. On the other hand, T24 cell line follows another molecular pathway of carcinogenesis and has a somewhat different STn pattern. In addition, the differential STn pattern between cell lines and experimental conditions vouch the importance of glycosylation as biomarker reflecting not only genetic differences but environmental pressures as well. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 64 Figure 14. STn expression pattern in urinary bladder cancer cells, as analysed by western blotting. A. Using a 10µg protein input, two intense bands at approximately 250 and 200 kDa are present in all cell lines and experimental conditions, suggesting an association between STn expression, mucins and malignancy. The molecular weights/ kDa are shown on the left. All cell lines and all conditions presented low-molecular weight bands between 50 kDa and 15kDa. The optical density (OD) of the bands was estimated and normalized in relation to GAPDH, previously observed to be a stable housekeeping gene in this bladder cancer cell lines. The bands were then classified in relation to their intensity for comparison purposes. B. To better analyse the STn profile regarding the low molecular weight bands, a new WB using a 20µg protein input was carried out. C. To determine the specificity of the low-molecular weight bands a Neuraminidase enzyme treatment was applied. Because Neuraminidase cleaves the glycosidic linkages of neuraminic/sialic acids, the STn structure is disrupted preventing TKH2 linkage. The western blot samples appear in the following order: T24, 5637 and HT1376 Normoxic conditions (N); T24, 5637 and HT1376 Hypoxic conditions (H); T24, 5637 and HT1376 DFX treatment (D). D. Bar graph showing differences in optical densities of the bands between experimental conditions. 4.8. HIF-1α, CA IX and STn expression in bladder tumours The expression of hypoxia markers HIF-1α and CA IX and cancer-associated glycan STn was evaluated by immunohistochemistry in a series including 30 NMIBC tumours and 43 MIBC tumours, randomly selected, representing all stages of the disease (15 pTa, 15 pT1, 13 pT2, 15 pT3, 15 pT4), to disclose associations between altered glycosylation and hypoxia. All studied tumours were positive for HIF-1α and approximately 70% presented an extensive expression (> 20% of the tumour area), irrespectively of their stage. HIF-1α was predominantly detected in the tumour cells cytoplasm but also in the nucleus, where it acts has a transcription factor. However, 43% of NMIBC (13/30) and 85% of MIBC (37/43), presented both cytoplasmic and nuclear HIF-1α expressions, demonstrating an association between the presence of the protein in the nucleus and advanced stage bladder cancer (p < 0.005). The nuclear HIF-1α positive tumours were further re-classified according to the degree of STn expression (Table IV). Table III shows that high nuclear expression was present in only 43% of the NMIBC (pTa and pT1) and in 72.1% of the MIBC (pT2, pT3 and pT4), reinforcing previous association of nuclear HIF-1α expression with muscle invasion (p < 0.05). Altogether, these findings strongly suggest an association FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 65 between high hypoxia levels and muscle invasive bladder cancer. The CA IX, a cellsurface protein up-regulated by HIF-1α under hypoxic conditions, was highly expressed in all tumours (> 30% of the tissue). However, as expected, CA IX expression was significantly overexpressed (> 50%) in tumours presenting high levels of nuclear HIF-1α. Regarding the evaluation of STn antigen, table IV shows the frequency of STn presence along the different stages of the disease. Higher STn expression was observed for MIBC (60%) when compared to NMIBC (30%), demonstrating an association of the antigen with muscle invasion (p= 0.03), in accordance with previous observations. The expression of HIF-1α and STn was further evaluated and it was found that all the tumour areas in which STn is expressed presented HIF-1α nuclear expression. Moreover, high expression of STn antigen co-localized with high nuclear HIF-1α expression in 75% of the cases, irrespectively of their stage. Table III – Correlation between tumour stage and nuclear HIF-1α expression Table IV – Correlation between STn and HIF-1α expression Tumour stage High HIF-1α nuclear expression Ta 4/15 (26.7%) T1 9/15 (60%) T2 9/13 (69%) T3 10/15 (66.7%) T4 12/15 (80%) Total 44/73 (60.3%) NMIBC 13/30 (43%) MIBC 31/43 (72.1%) P<0.05 Tumour stage STn+Tumours with STn/ HIF-1α nuclear positive areas Ta 3/15 (20%) 3/4 (75%) T1 6/15 (40%) 4/6 (67%) T2 6/12 (50%) 3/4 (75%) T3 10/15 (67%) 7/9 (78%) T4 9/15 (60%) 2/3 (67%) Total 33/72 (47%) P=0.03 19/26 (73%) NMIBC 9/30 (30%) MIBC 25/42 (60%) FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 67 V. Concluding remarks and Future perspectives The STn antigen is considered a pancarcinoma antigen, based on the fact that it is expressed by the majority of advanced stage solid tumours, while absent from the corresponding healthy tissues. It is also far from being an innocent bystander in the disease outcome, since it was found to be a promoter of invasion and disease dissemination in many models, including stomach, pancreas and breast (120,165) carcinomas. In agreement with these observations our group has recently demonstrated that STn is expressed by over 70% of advanced stage bladder cancers, in poorly proliferative tumour areas and invasion fronts. Moreover, we found that STn expression enhanced the motility and invasive potential of cancer cells in vitro and are now detecting the antigen in lymph node and distant bladder cancer metastasis (unpublished data). Subsequent studies concluded that STn-expressing cancer cells impair DC maturation and promote a tolerogenic function, limiting their capacity to trigger protective anti-tumour T cell responses (156). Altogether these findings suggest that STn positive cells are endowed with the capability to invade and disseminate throughout the organ and to distant locations, while avoiding immune surveillance. Based on these observations we envisage that STn antigens and, in particular, STn expressing glycoproteins are potential targets to circumventing tumour-induced tolerogenic mechanisms and to avoid disease progression. Nevertheless, and despite the key role of STn expression in cancer, the events responsible by this premature stop in elongation of O-glycans by sialylation of the Tn antigen are largely unknown. To this date two main mechanisms have been described: i) the overexpression of ST6GalNAc.I, which is commonly observed; ii) loss-of-function mutations in C1GALT1 chaperone Cosmc, described by few reports. This work now demonstrates, for the first time, that hypoxia, a salient feature of solid tumours, is the main promoter of STn overexpression in bladder cancer. We started by submitting three bladder cancer cell lines showing different genetic backgrounds to hypoxia and observed an overexpression of HIF-1α, CA IX and increased lactate levels in the culture media, denoting a shift to anaerobic metabolism. All cells responded similarly, by showing altered cellular morphology, characterized by increased intercellular spaces, impaired proliferation and enhanced cell migration. This was also true for cells growing the presence of DFX, highlighting these are HIF-1α-mediated events. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 68 However, a gene expression analysis of genes associated with stem cell, epithelial cell, epithelial-to-mesenchymal and mesenchymal cell programs highlighted significant similarities between T24 and 5637 in normoxia as well as hypoxia; interestingly 5637 cells were regarded as genetically related to HT1376. Based on these observations we concluded that T24 and 5637 cells present a more mesenchymal nature, with T24 being markedly more immature, whereas HT1376 presents a more epithelial nature. Moreover, hypoxia up-regulated SNAI2, KLF4 and FN1 genes in a HIF-1αmediated process, thus activating EMT and stem gene programs in all cell lines. Concomitantly, and despite morphological, genetic and molecular differences, all cells overexpressed the STn antigen in hypoxia, which translated in higher cell migration/invasion mediated by this antigen. More interestingly, STn overexpression was not linked to an overexpression of ST6GalNAc.I but to a striking down-regulation of the enzymes that participate in the downstream elongation of O-glycans. This highlights a new mechanism by which bladder cancer cells, probably of stem cell nature, acquire migration capacity in response to the hypoxic challenge. The effect was reverted by reoxygenation of the cells, denoting an onoff switch which endows bladder cancer cells with the capability to escape hypoxic niches and ultimately colonize distant locations. These observations may acquaint for the focal expression of STn in bladder tumours and in the metastasis (unpublished data). Cells grown in the presence of DFX also overexpressed the STn antigen, suggesting that HIF1α may play an active role in the modulation of the glycophenotype by direct or indirect modulation of O-glycosyltransferases transcription. Studies on a retrospective series of bladder tumours comprehending all stages of the disease has confirmed the associations between hypoxia, translated by the nuclear overexpression of HIF-1α, and elevated STn. The high STn and nuclear HIF-1α phenotypes was predominantly observed in MIBC, also in accordance with in vitro studies. In previous studies we also reported that STn positive cells are mainly present in non-proliferative niches, which present high resistance to cisplatin-based regimens used in bladder cancer treatment. Here we observed that STn overexpression under hypoxia was accompanied by a decrease in cell proliferation. Imagiological studies by imagestream flow cytometry are ongoing to determine the cell cycle phase of STn positive cells envisaging understanding its susceptibility to chemotherapy. Furthermore, we aim to also determine if the presence of this antigen contributes positively to the mitigation of cell proliferation. It would also be important to confirm the capability of STn positive cells to form metastasis and recapitulate tumour heterogeneity in distant locations. Given the pancarcinoma nature of the STn antigen, FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 69 studies should be conducted to understand if hypoxia promotes similar alterations in cell lines from other organs. Finally, a preliminary glycoproteomic study has showed that bladder cancer cells under hypoxia overexpress low molecular weight STn-positive proteins (< 50 kDa). The future identification of the proteins responsible by this distinct pattern by Mass Spectrometry-based proteomics are expected to provide highly specific theranostic biomarkers to improve bladder cancer management. Furthermore, this will bring more insights about the biological significance of STn expression in bladder cancer but also other advanced stage tumours that also overexpress this antigen. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 75 66. Jose C, Bellance N, Rossignol R. Choosing between glycolysis and oxidative phosphorylation: a tumor’s dilemma? Biochim Biophys Acta. Elsevier B.V.; 2011;1807(6):552–61. 67. Wise DR, Thompson CB. Glutamine Addiction: A New Therapeutic Target in Cancer. Trends Biochem Sci. 2010;35(8):427–33. 68. Sedlakova O, Svastova E, Takacova M, Kopacek J, Pastorek J, Pastorekova S. Carbonic anhydrase IX, a hypoxia-induced catalytic component of the pH regulating machinery in tumors. Front Physiol. 2014;4(January):400. 69. Shareef MM, Udayakumar TS, Sinha VK, Saleem SM, Griggs WW. Interaction of HIF-1α and Notch3 Is Required for the Expression of Carbonic Anhydrase 9 in Breast Carcinoma Cells. Genes Cancer. 2013;4(11-12):513–23. 70. Gilkes DM, Semenza GL, Wirtz D. Hypoxia and the extracellular matrix: drivers of tumour metastasis. Nat Rev Cancer. Nature Publishing Group; 2014;14(6):430–9. 71. Trastour C, Benizri E, Ettore F, Ramaioli A, Chamorey E, Pouysségur J, et al. HIF1alpha and CA IX staining in invasive breast carcinomas: prognosis and treatment outcome. Int J Cancer. 2007;120(7):1451–8. 72. Lock FE, McDonald PC, Lou Y, Serrano I, Chafe SC, Ostlund C, et al. Targeting carbonic anhydrase IX depletes breast cancer stem cells within the hypoxic niche. Oncogene. Nature Publishing Group; 2013;32(44):5210–9. 73. Klatte T, Seligson DB, Rao JY, Yu H, de Martino M, Kawaoka K, et al. Carbonic anhydrase IX in bladder cancer: a diagnostic, prognostic, and therapeutic molecular marker. Cancer. 2009;115(7):1448–58. 74. Kim Y, Lin Q, Glazer PM, Yun Z. Hypoxic Tumour Microenvironment and Cancer Cell Differentiation. Curr Mol Med. 2010;9(4):425–34. 75. Semenza GL. Targeting HIF-1 for cancer therapy. Nat Rev Cancer. 2003;3(10):721–32. 76. Celià-terrassa T, Meca-cortés Ó, Mateo F, Paz AM De, Rubio N, Arnal-estapé A, et al. Epithelial-mesenchymal transition can suppress major attributes of human epithelial tumor-initiating cells. J Clin Invest. 2012;122(5):1849-68. 77. Polyak K, Weinberg R a. Transitions between epithelial and mesenchymal states: acquisition of malignant and stem cell traits. Nat Rev Cancer. 2009;9(4):265–73. 78. Yun SJ, Kim W-J. Role of the epithelial-mesenchymal transition in bladder cancer: from prognosis to therapeutic target. Korean J Urol. 2013;54(10):645–50. 79. Bryan RT, Tselepis C. Cadherin switching and bladder cancer. J Urol. Elsevier Inc.; 2010;184(2):423–31. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 76 80. Wallerand H, Robert G, Pasticier G, Ravaud A, Ballanger P, Reiter RE, et al. The epithelial-mesenchymal transition-inducing factor TWIST is an attractive target in advanced and/or metastatic bladder and prostate cancers. Urol Oncol. Elsevier Inc.; 2010;28(5):473–9. 81. Yu Q, Zhang K, Wang X, Liu X, Zhang Z. Expression of transcription factors snail, slug, and twist in human bladder carcinoma. J Exp Clin Cancer Res. 2010;29:119. 82. McConkey DJ, Choi W, Marquis L, Martin F, Williams MB, Shah J, et al. Role of epithelial-to-mesenchymal transition (EMT) in drug sensitivity and metastasis in bladder cancer. Cancer Metastasis Rev. 2009;28(3-4):335–44. 83. Jiang J, Tang Y, Liang X. EMT: A new vision of hypoxia promoting cancer progression. Cancer Biol Ther. 2011;11(8):714–23. 84. Haase VH. Oxygen regulates epithelial-to-mesenchymal transition: insights into molecular mechanisms and relevance to disease. Kidney Int. 2009;76(5):492–9. 85. Zhiwei Wang, Yiwei Li, Dejuan Kong and FHS. The Role of Notch Signaling Pathway in Epithelial-Mesenchymal Transition (EMT) During Development and Tumor Aggressiveness. Curr Drug Targets. 2010;11(6):745–51. 86. Huber MA, Azoitei N, Baumann B, Grünert S, Sommer A, Pehamberger H, et al. NFκ B is essential for epithelialmesenchymal transition and metastasis in a model of breast cancer progression. J Clin Invest. 2004;114(4):569–81. 87. Jaenisch R, Young R. Stem cells, the molecular circuitry of pluripotency and nuclear reprogramming. Cell. 2008;132(4):567–82. 88. Cojoc M, Mäbert K, Muders MH, Dubrovska A. A role for cancer stem cells in therapy resistance: Cellular and molecular mechanisms. Semin Cancer Biol. Elsevier Ltd; 2014 89. Szablowska-Gadomska I, Zayat V, Buzanska L. Influence of low oxygen tensions on expression of pluripotency genes in stem cells. Acta Neurobiol Exp (Wars). 2011;71(1):86–93. 90. Lin Q, Yun Z. Impact of the hypoxic tumor microenvironment on the regulation of cancer stem cell characteristics. Cancer Biol Ther . 2010;9(12):949–56. 91. Rabinovich G a, Toscano M a. Turning “sweet” on immunity: galectin-glycan interactions in immune tolerance and inflammation. Nat Rev Immunol. 2009;9(5):338–52. 92. Helenius AAM. Intracellular Functions of N-linked Glycans. Science. 2001;291:2364-69. 93. Kannagi R. Carbohydrate-mediated cell adhesion involved in hematogenous metastasis of cancer. Glycoconj J. 1997;14(5):577–84. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 77 94. Häuselmann I, Borsig L. Altered tumor-cell glycosylation promotes metastasis. Front Oncol. 2014;4(February):28. 95. Trombetta ES. The contribution of N-glycans and their processing in the endoplasmic reticulum to glycoprotein biosynthesis. Glycobiology. 2003;13(9):77R – 91R. 96. Doucey M a, Hess D, Cacan R, Hofsteenge J. Protein C-mannosylation is enzymecatalysed and uses dolichyl-phosphate-mannose as a precursor. Mol Biol Cell. 1998;9(2):291–300. 97. Kezawa HI. Glycosylphosphatidylinositol ( GPI ) -Anchored Proteins. Biol. Pharm. Bull.2002; 25(4) 409—417. 98. Kodukula K, Gerber LD, Amthauer R, Brink L, Udenfriend S. Biosynthesis of glycosylphosphatidylinositol (GPI)-anchored membrane proteins in intact cells: specific amino acid requirements adjacent to the site of cleavage and GPI attachment. J Cell Biol. 1993;120(3):657–64. 99. Haynes P. Phosphoglycosylation: a new structural class of glycosylation? Glycobiology. 1998;8(1):1–5. 100. Spiro RG. Protein glycosylation: nature, distribution, enzymatic formation, and disease implications of glycopeptide bonds. Glycobiology. 2002 Apr;12(4):43R – 56R. 101. Burda P, Aebi M. The dolichol pathway of N-linked glycosylation. Biochim Biophys Acta. 1999 6;1426(2):239–57. 102. Dempski RE, Imperiali B. Oligosaccharyl transferase: gatekeeper to the secretory pathway. Curr Opin Chem Biol. 2002;6(6):844–50. 103. Lederkremer GZ, Glickman MH. A window of opportunity: timing protein degradation by trimming of sugars and ubiquitins. Trends Biochem Sci. 2005;30(6):297–303. 104. Tran DT, Ten Hagen KG. Mucin-type O-glycosylation during development. J Biol Chem. 2013;288(10):6921–9. 105. Steentoft C, Vakhrushev SY, Joshi HJ, Kong Y, Vester-Christensen MB, Schjoldager KT-BG, et al. Precision mapping of the human O-GalNAc glycoproteome through SimpleCell technology. EMBO J. 2013;32(10):1478–88. 106. Hakomori S. Glycosylation defining cancer malignancy: new wine in an old bottle. Proc Natl Acad Sci U S A. 2002;99(16):10231–3. 107. Dall’Olio F. Protein glycosylation in cancer biology : an overview. J Clin PatholMol Pathol. 1996;49:126–35. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 78 108. Hakomori S. Tumor Malignancy Defined by Aberrant Glycosylation and Sphingo (glyco) lipid Metabolism Tumor Malignancy. Cancer Res 1996;56:5309-18. 109. Hakomori S. Tumour-associated carbohydrate antigens defining tumor malignancy: Basis for development of anti-cancer vaccines. In: Albert M; Wu KAP, editor. 2001. p. 369–402. 110. Almaraz RT, Tian Y, Bhattarcharya R, Tan E, Chen S-H, Dallas MR, et al. Metabolic flux increases glycoprotein sialylation: implications for cell adhesion and cancer metastasis. Mol Cell Proteomics. 2012;11(7):1-34. 111. Rillahan CD, Antonopoulos A, Lefort CT, Sonon R, Azadi P, Ley K, et al. Global metabolic inhibitors of sialyland fucosyltransferases remodel the glycome. Nat Chem Biol. 2012;8(7):661–8. 112. Martín-satué M, Marrugat R, Cancelas JA, Cells LA, Martã-n-satuã M, Cancelas A, et al. Enhanced Expression of α (1,3) -Fucosyltransferase Genes Correlates with Eselectin-mediated Adhesion and Metastatic Potential of Human Lung Adenocarcinoma Cells. Cancer Res. 1998;58:1544–50. 113. Ito H, Hiraiwa N, Sawada-Kasugai M, Akamatsu S, Tachikawa T, Kasai Y, et al. Altered mRNA expression of specific molecular species of fucosyland sialyltransferases in human colorectal cancer tissues. Int J Cancer. 1997;71(4):556–64. 114. Kannagi R. Regulatory roles of carbohydrate ligands for selectins in the homing of lymphocytes. Curr Opin Struct Biol. 2002;12(5):599–608. 115. Reis C a, Osorio H, Silva L, Gomes C, David L. Alterations in glycosylation as biomarkers for cancer detection. J Clin Pathol. 2010;63(4):322–9. 116. Marcos NT, Bennett EP, Gomes J, Magalhaes A, Gomes C, David L, et al. ST6GalNAc-I controls expression of sialyl-Tn antigen in gastrointestinal tissues. Front Biosci. 2011;3:1443–55. 117. Sewell R, Bäckström M, Dalziel M, Gschmeissner S, Karlsson H, Noll T, et al. The ST6GalNAc-I sialyltransferase localizes throughout the Golgi and is responsible for the synthesis of the tumor-associated sialyl-Tn O-glycan in human breast cancer. J Biol Chem. 2006;281(6):3586–94. 118. Ju T, Lanneau GS, Gautam T, Wang Y, Xia B, Stowell SR, et al. Human tumor antigens Tn and sialyl Tn arise from mutations in Cosmc. Cancer Res. 2008;68(6):1636–46. 119. Yoo, N.J.; Kim, M.S.; Lee SH. Absence of COSMC gene mutations in breast and colorectal carcinomas. APMIS. 2008;116:154–5. 120. Pinho S, Marcos NT, Ferreira B, Carvalho AS, Oliveira MJ, Santos-Silva F, et al. Biological significance of cancer-associated sialyl-Tn antigen: modulation of malignant phenotype in gastric carcinoma cells. Cancer Lett. 2007;249(2):157–70. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 79 121. Itzkowitz SH, Yuan M, Montgomery CK, Kjeldsen T, Takahashi HK, Bigbee WL, et al. Expression of Tn , Sialosyl-Tn , and T Antigens in Human Colon Cancer. Cancer Res. 1989;49:197–204. 122. Cazet A, Julien S, Bobowski M, Burchell J, Delannoy P. Tumour-associated carbohydrate antigens in breast cancer. Breast Cancer Res. 2010;12(3):204. 123. Kumar SR, Sauter ER, Quinn TP, Deutscher SL. Thomsen-Friedenreich and Tn antigens in nipple fluid: carbohydrate biomarkers for breast cancer detection. Clin cancer Res. 2005;11(19 Pt 1):6868–71. 124. Takamiya R, Ohtsubo K, Takamatsu S, Taniguchi N, Angata T. The interaction between Siglec-15 and tumor-associated sialyl-Tn antigen enhances TGF-β secretion from monocytes/macrophages through the DAP12-Syk pathway. Glycobiology. 2013;23(2):178–87. 125. Ghazizadeh M, Ogawa H, Sasaki Y, Araki T, Aihara K. Mucin carbohydrate antigens (T, Tn, and Sialyl-Tn) in human ovarian carcinomas: Relationship with histopathology and prognosis. Hum Pathol. 1997;28(8):960–6. 126. Uemura T, Shiozaki K, Yamaguchi K, Miyazaki S, Satomi S, Kato K, et al. Contribution of sialidase NEU1 to suppression of metastasis of human colon cancer cells through desialylation of integrin beta4. Oncogene. 2009;28(9):1218–29. 127. Pinho S, Marcos NT, Ferreira B, Carvalho AS, Oliveira MJ, Santos-Silva F, et al. Biological significance of cancer-associated sialyl-Tn antigen: modulation of malignant phenotype in gastric carcinoma cells. Cancer Lett. 2007;249(2):157–70. 128. Cao Y, Stosiek P, Springer GF, Karsten U. Thomsen-Friedenreich-related carbohydrate antigens in normal adult human tissues: a systematic and comparative study. Histochem Cell Biol. 1996;106(2):197–207. 129. Guadagni F, Roselli M, Amato T, Roseli M, Cosimeli M, Ferri P, et al. CA 72-4 Measurement of Tumor-associated Glycoprotein 72 ( TAG-72 ) as a Serum Marker in the Management of Gastric Carcinoma. Cancer Res. 1992;72:1222–7. 130. Motoo Yoshiharu, Kawakami Hiroyasu, Watanabe Hiroyuki SY. Serum Sialyl-Tn Antigen Levels in Patients with Digestive Cancers. Oncology. 1991;48:321–6. 131. Gomes C, Almeida A, Ferreira JA, Silva L, Santos-Sousa H, Pinto-de-Sousa J, et al. Glycoproteomic analysis of serum from patients with gastric precancerous lesions. J Proteome Res. 2013;12(3):1454–66. 132. Marrelli D, Stefano A De. Prognostic Significance of CEA , CA 19-9 and CA 72-4 Preoperative Serum Levels. Oncology. 1999;57:55–62. 133. Levels CA, Joypaul B, Browning M, Newman E, Byrne D, Cuschieri A. Comparison of Serum CA 72-4 and CA19-9 levels in Gastric Cancer patients and correlation with recurrence. Am J Surg. 1995;169:595–9. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 80 134. Louhimo J, Alfthan H, Stenman U-H, Haglund C. Serum HCG beta and CA 72-4 are stronger prognostic factors than CEA, CA 19-9 and CA 242 in pancreatic cancer. Oncology. 2004;66(2):126–31. 135. Miles D, Papazisis K. Rationale for the Clinical Development of STn-KLH (Theratope®) and Anti—MUC-1 Vaccines in Breast Cancer. Clin Breast Cancer. Elsevier Inc.; 2003;3(February):S134–8. 136. Julien S, Picco G, Sewell R, Vercoutter-Edouart-S, Tarp M, Miles D, et al. Sialyl-Tn vaccine induces antibody-mediated tumour protection in a relevant murine model. Br J Cancer. Nature Publishing Group; 2009;100(11):1746–54. 137. Ibrahim NK, Murray JL, Zhou D, Mittendorf E a, Sample D, Tautchin M, et al. Survival Advantage in Patients with Metastatic Breast Cancer Receiving Endocrine Therapy plus Sialyl Tn-KLH Vaccine: Post Hoc Analysis of a Large Randomized Trial. J Cancer. 2013;4(7):577–84. 138. Holmberg LA, Oparin D V, Gooley T, Lilleby K, Bensinger W, Reddish MA. Clinical outcome of breast and ovarian cancer patients treated with high-dose chemotherapy, autologous stem cell rescue and THERATOPE ® STn-KLH cancer vaccine. Bone Marrow Transplantation. 2000;25:1233–41. 139. Chihara Y, Sugano K, Kobayashi A, Kanai Y, Yamamoto H, Nakazono M, et al. Loss of blood group A antigen expression in bladder cancer caused by allelic loss and/or methylation of the ABO gene. Lab Invest [Internet]. 2005 Jul [cited 2014 Oct 18];85(7):895–907. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15880137 140. Sheinfeld J, Reuter VE, Fair WR, Cordon-Cardo C. Expression of blood group antigens in bladder cancer: Current concepts. Semin Surg Oncol. 1992;8(5):308– 15. 141. Langkilde NC, Wolf H, Meldgard P. Frequency and mechanism of Lewis antigen expression in human urinary bladder and colon carcinoma patients. Br J Cancer. 1991;63(4):583–6. 142. Thorpe SJ, Abel P, Slavin G, Feizi T. Blood group antigens in the normal and neoplastic bladder epithelium. J Clin Pathol. 1983;36(8):873–82. 143. Ørntoft TF, Wolf H, Watkins WM, Torben FÃ. Activity of the Human Blood Group ABO , Se , H , Le , and X Gene-encoded Glycosyltransferases in Normal and Malignant Bladder Urothelium. Cancer Res. 1988;48:4427–33. 144. Cordon-cardo C, Reuter VE, Lloyd KO, Determinants L, Sheinfeld J, Fair WR, et al. Blood Group-related Antigens in Human Urothelium : Enhanced Expression of Precursor , Le X , and Le Y Determinants in Urothelial Carcinoma. Cancer Res. 1988;48:4113–20. 145. Limas C, Lange PH. Lewis Antigens in Normal and Neoplastic Urothelium. AJP. 1985;121(1):176–83. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 81 146. Numahata K, Satoh M, Handa K, Saito S, Ohyama C, Ito A, et al. Sialosyl-Le(x) expression defines invasive and metastatic properties of bladder carcinoma. Cancer. 2002;94(3):673–85. 147. Nagao K, Itoh Y, Fujita K, Fujime M. Evaluation of urinary CA19-9 levels in bladder cancer patients classified according to the combinations of Lewis and Secretor blood group genotypes. Int J Urol. 2007;14(9):795–9. 148. Ohyama C. Glycosylation in bladder cancer. Int J Clin Oncol. 2008;13(4):308–13. 149. Lokeshwar VB, Schroeder GL, Selzer MG, Hautmann SH, Posey JT, Duncan RC, et al. Bladder tumor markers for monitoring recurrence and screening comparison of hyaluronic acid-hyaluronidase and BTA-Stat tests. Cancer. 2002;95(1):61–72. 150. Cindolo L, Benvenuto G, Salvatore P, Pero R, Salvatore G, Mirone V, et al. Galectin-1 and Galectin-3 Expression in Human Bladder Transitional-Cell Carcinomas. Int J Cancer. 1999;84(1):39–43. 151. Kawamura S, Ohyama C, Watanabe R, Satoh M, Saito S, Hoshi S, et al. Glycolipid composition in bladder tumor: a crucial role of GM3 ganglioside in tumor invasion. Int J Cancer. 2001;94(3):343–7. 152. Todeschini AR, Dos Santos JN, Handa K, Hakomori S. Ganglioside GM2/GM3 complex affixed on silica nanospheres strongly inhibits cell motility through CD82/cMet-mediated pathway. Proc Natl Acad Sci U S A. 2008;105(6):1925–30. 153. Langkilde NC, Wolf H, Clausen H. Nuclear Volume and Expression and Tn-Antigen in Carcinoma of the Human Bladder Relation to Tumor Recurrence and Progression. Cancer. 1992;69(1):219–27. 154. Summers JL, Coon JS, Ward RM, Falor WH, Miller a W, Weinstein RS. Prognosis in carcinoma of the urinary bladder based upon tissue blood group ABH and Thomsen-Friedenreich antigen status and karyotype of the initial tumor. Cancer Res. 1983;43(2):934–9. 155. Videira P a, Correia M, Malagolini N, Crespo HJ, Ligeiro D, Calais FM, et al. ST3Gal.I sialyltransferase relevance in bladder cancer tissues and cell lines. BMC Cancer. 2009;9:357. 156. Carrascal M a, Severino PF, Guadalupe Cabral M, Silva M, Ferreira JA, Calais F, et al. Sialyl Tn-expressing bladder cancer cells induce a tolerogenic phenotype in innate and adaptive immune cells. Mol Oncol. 2014;8(3):753–65. 157. Shirato K, Nakajima K, Korekane H, Takamatsu S, Gao C, Angata T. Hypoxic regulation of glycosylation via the N acetylglucosamine cycle. J Clin Biochem Nutr. 2011;48(1):20–5. 158. Fabio Dall’Olio, Nadia Malagolini, Marco Trinchera MC. Mechanisms of cancerassociated glycosylation changes. Front Biosci. 2012;17:670–699. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 82 159. Koike T, Kimura N, Miyazaki K, Yabuta T, Kumamoto K, Takenoshita S, et al. Hypoxia induces adhesion molecules on cancer cells: A missing link between Warburg effect and induction of selectin-ligand carbohydrates. Proc Natl Acad Sci U S A. 2004;101(21):8132–7. 160. Yin J, Miyazaki K, Shaner RL, Merrill AH, Kannagi R. Altered sphingolipid metabolism induced by tumor hypoxia - new vistas in glycolipid tumor markers. FEBS Lett. 2010;584(9):1872–8. 161. Ferlay J, Soerjomataram I, Ervik M, Dikshit R, Eser S, Mathers C, Rebelo M, Parkin DM, Forman D, Bray F. GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 11. 2012. p. xiii. 162. Lindgren D, Sjödahl G, Lauss M, Staaf J, Chebil G, Lövgren K, et al. Integrated genomic and gene expression profiling identifies two major genomic circuits in urothelial carcinoma. PLoS One. 2012;7(6):e38863. 163. Pinto-Leite R, Carreira I, Melo J, Ferreira SI, Ribeiro I, Ferreira J, et al. Genomic characterization of three urinary bladder cancer cell lines: understanding genomic types of urinary bladder cancer. Tumour Biol. 2014;35(5):4599–617. 164. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001;25(4):402–8. 165. Julien S, Adriaenssens E, Ottenberg K, Furlan a, Courtand G, Vercoutter-Edouart aS, et al. ST6GalNAc I expression in MDA-MB-231 breast cancer cells greatly modifies their O-glycosylation pattern and enhances their tumourigenicity. Glycobiology. 2006;16(1):54–64. 166. Richard DE, Berra E, Gothie E, Roux D, Pouyssegur J. p42/p44 Mitogen-activated Protein Kinases Phosphorylate Hypoxia-inducible Factor 1 (HIF-1) and Enhance the Transcriptional Activity of HIF-1. J Biol Chem. 1999;274(46):32631–7. 167. Lim J-H, Lee E-S, You H-J, Lee JW, Park J-W, Chun Y-S. Ras-dependent induction of HIF-1alpha785 via the Raf/MEK/ERK pathway: a novel mechanism of Rasmediated tumor promotion. Oncogene. 2004;23(58):9427–31. 168. Depping R, Hägele S, Wagner KF, Wiesner RJ, Camenisch G, Wenger RH, et al. A dominant-negative isoform of hypoxia-inducible factor-1 alpha specifically expressed in human testis. Biol Reprod. 2004;71(1):331–9. 169. Sitkovsky DLM. Preferential expression of the novel alternative isoform I.3 of Hypoxia-Inducible Factor 1α in activated human T lymphocytes. Hum Immunol. 2008;69(7):421–5. 170. Chun Y, Choi E, Yeo E, Lee JH, Kim M, Park J. A new HIF-1 alpha variant induced by zinc ion suppresses HIF-1-mediated hypoxic responses. J Cell Sci. 2001;144:4051–61. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 83 171. Chun Y, Choi E, Kim T, Kim M, Park J. A dominant-negative isoform lacking exons 11 and 12 of the human hypoxia-inducible factor-1α gene. Biochem J. 2002;362:71– 9. 172. Gothie E, Richard DE, Berra E, Pages G, Pouyssegur J. Identification of Alternative Spliced Variants of Human Hypoxia-inducible Factor-1α. J Biol Chem. 2000;275(10):6922–7. 173. Loscalzo J. The cellular response to hypoxia: tuning the system with microRNAs. J Clin Invest. 2010;120(11):3815–7. 174. Span PN, Rao JU, Oude Ophuis SBJ, Lenders JWM, Sweep FCGJ, Wesseling P, et al. Overexpression of the natural antisense hypoxia-inducible factor-1alpha transcript is associated with malignant pheochromocytoma/paraganglioma. Endocr Relat Cancer. 2011;18(3):323–31. 175. Uchida T, Rossignol F, Matthay M a, Mounier R, Couette S, Clottes E, et al. Prolonged hypoxia differentially regulates hypoxia-inducible factor (HIF)-1alpha and HIF-2alpha expression in lung epithelial cells: implication of natural antisense HIF1alpha. J Biol Chem. 2004;279(15):14871–8. 176. Bruning U, Cerone L, Neufeld Z, Fitzpatrick SF, Cheong A, Scholz CC, et al. MicroRNA-155 promotes resolution of hypoxia-inducible factor 1alpha activity during prolonged hypoxia. Mol Cell Biol. 2011;31(19):4087–96. 177. Sonveaux P, Végran F, Schroeder T, Wergin MC, Verrax J, Rabbani ZN, et al. Targeting lactate-fueled respiration selectively kills hypoxic tumor cells in mice. J Clin Invest. 2008;118(12):3930–42. 178. Walenta S, Mueller-Klieser WF. Lactate: mirror and motor of tumor malignancy. Semin Radiat Oncol. 2004;14(3):267–74. 179. Ju T, Lanneau GS, Gautam T, Wang Y, Xia B, Stowell SR, et al. Human tumor antigens Tn and sialyl Tn arise from mutations in Cosmc. Cancer Res. 2008;68(6):1636–46. 180. Clément M, Rocher J, Loirand G, Le Pendu J. Expression of sialyl-Tn epitopes on beta1 integrin alters epithelial cell phenotype, proliferation and haptotaxis. J Cell Sci. 2004;117(Pt 21):5059–69. 181. Picco G, Julien S, Brockhausen I, Beatson R, Antonopoulos A, Haslam S, et al. Over-expression of ST3Gal-I promotes mammary tumorigenesis. Glycobiology. 2010;20(10):1241–50. 182. Wouters A, Pauwels B, Lardon F, Vermorken JB. Review: implications of in vitro research on the effect of radiotherapy and chemotherapy under hypoxic conditions. Oncologist. 2007;12(6):690–712. FCUP Hypoxic Regulation of Glycosylation in Bladder Cancer 84 183. Krtolica A, Ludlow JW. Hypoxia Arrests Ovarian Carcinoma Cell Cycle Progression , but Invasion Is Unaffected. Cancer Res. 1996;56:1168–73. 184. Weinmann M, Marini P, Jendrossek V, Betsch A, Goecke B, Budach W, et al. Influence of hypoxia on TRAIL-induced apoptosis in tumor cells. Int J Radiat Oncol. 2004;58(2):386–96. 185. Wilson WR, Hay MP. Targeting hypoxia in cancer therapy. Nat Rev Cancer. 2011;11(6):393–410. 186. Cazzola, Mario; Bergamaschi Gaetano; Dezza Laura; Arosio P. Manipulations of cellular iron metabolisms for modulating normal and malignant cell proliferation: Achievements and prospects. J Am Soc Hematol. 1990;75(10):1903–19. 187. Hileti D, Panayiotidis P, Hoffbrand a V. Iron chelators induce apoptosis in proliferating cells. Br J Haematol. 1995;89(1):181–7. 188. Porreca E, Ucchino S, Di Febbo C, Di Bartolomeo N, Angelucci D, Napolitano a. M, et al. Antiproliferative effect of desferrioxamine on vascular smooth muscle cells in vitro and in vivo. Arterioscler Thromb Vasc Biol. 1994;14(2):299–304. 189. Kim BM, Choi JY, Kim YJ, Woo HD, Chung HW. Desferrioxamine (DFX) has genotoxic effects on cultured human lymphocytes and induces the p53-mediated damage response. Toxicology. 2007;229(3):226–35. 190. Kim B-M, Chung H-W. Desferrioxamine (DFX) induces apoptosis through the p38caspase8-Bid-Bax pathway in PHA-stimulated human lymphocytes. Toxicol Appl Pharmacol. 2008;228(1):24–31. 191. Cancer G, Karsten U, Otto G, Krebsforschungszentrum D. Expression of MUC1, Thomsen-Friedenreich antigen, Tn, sialosyl-Tn and α 2 , 6-linked sialic acid in hepatocellular carcinomas and preneoplastic hepatocellular lesions. Virchows Arch. 1999;434:503–9. 192. Ozaki H, Matsuzaki H, Ando H, Kaji H, Nakanishi H, Ikehara Y, et al. Enhancement of metastatic ability by ectopic expression of ST6GalNAcI on a gastric cancer cell line in a mouse model. Clin Exp Metastasis. 2012;29(3):229–38. 193. Retz M, Lehmann J, Röder C, Ret M, Roder C, Eggers J, et al. Differential Mucin MUC7 Gene Expression in Invasive Bladder Carcinoma in Contrast to Uniform MUC1 and MUC2 Gene Expression in Both Normal Urothelium and Bladder Carcinoma. Cancer Res. 1998;58:5662–6. 194. Kaur S, Momi N, Chakraborty S, Wagner DG, Horn AJ, Lele SM, et al. Altered expression of transmembrane mucins, MUC1 and MUC4, in bladder cancer: pathological implications in diagnosis. PLoS One. 2014;9(3):e92742. 195. Taylor-Papadimitriou J, Burchell J, Miles DW, Dalziel M. MUC1 and cancer. Biochim Biophys Acta. 1999;1455(2-3):301–13.