Role of protein glycosylation in cancer metastasis

Accepted Manuscript Title: Role of protein glycosylation in cancer metastasis Authors: Leticia Oliveira-Ferrer, Karen Legler, Karin Milde-Langosch PII: DOI: Reference:

S1044-579X(17)30040-8 http://dx.doi.org/doi:10.1016/j.semcancer.2017.03.002 YSCBI 1303

To appear in:

Seminars in Cancer Biology

Received date: Revised date: Accepted date:

11-1-2017 8-3-2017 13-3-2017

Please cite this article as: Oliveira-Ferrer Leticia, Legler Karen, Milde-Langosch Karin.Role of protein glycosylation in cancer metastasis.Seminars in Cancer Biology http://dx.doi.org/10.1016/j.semcancer.2017.03.002 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Role of protein glycosylation in cancer metastasis

Leticia Oliveira-Ferrer*, Karen Legler, Karin Milde-Langosch Department of Gynecology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany.

*Corresponding author: Leticia Oliveira-Ferrer Department of Gynecology University Medical Center Hamburg-Eppendorf (UKE) Martinistr. 52; Bldg. N61 20246 Hamburg Germany E-mail: [email protected] Telephone number: +49 40 74105 2559 Fax number: +49 40 74105 4103

1

Abstract Although altered glycosylation has been detected in human cancer cells decades ago, only investigations in the last years have enormously increased our knowledge about the details of protein glycosylation and its role in tumour progression. Many proteins, which are heavily glycosylated, i.e. adhesion proteins or proteases, play an important role in cancer metastasis that represents the crucial and frequently life-threatening step in progression of most tumour types. Compared to normal tissue, tumour cells often show altered glycosylation patters with appearance of new tumour-specific antigens. In this review, we give an overview about the role of glycosylation in tumour metastasis, describing recent results about O-glycans, Nglycans and glycosaminoglycans. We show that glycan structures, glycosylated proteins and glycosylation enzymes have influence on different steps of the metastatic process, including epithelial-mesenchymal transition (EMT), migration, invasion/intravasation and extravasation of tumour cells. Regarding the important role of cancer metastasis for patients survival, further knowledge about the consequences of altered glycosylation patterns in tumour cells is needed which might eventually lead to the development of novel therapeutic approaches. Keywords: glycosylation, O-glycan, N-glycan, glycosaminoglycan, metastasis.

Introduction Metastasis is the multi-step process of hematogeneous, lymphatic or peritoneal spread of tumour cells, leading to secondary tumours. A metastatic tumour cell has to go through a series of essential events, including epithelial-mesenchymal-transition (EMT), detachment from the primary tumour mass, adherence to proteins of the extracellular matrix (ECM), migration on and degradation of ECM proteins, invasion into neighboring tissue, penetration into lymphatic or blood vessels, spreading into different parts of the body, and extravasation from the vessels to form a metastatic tumour. In this review, we will give an overview and describe recent results on the role of glycosylation in tumour metastasis, with view on O-glycans, N-glycans and glycosaminoglycans. We will concentrate on the influence of glycosylation on the abovementioned steps of metastasis. Other issues, which are important for tumour progression, like proliferation, chemoresistance, angiogenesis or interactions with the immune system, will not be addressed here.

1. O-glycans 1.1 Biosynthesis and types of O-glycans The O-glycosylation is a covalent post-translational modification in which monosaccharides are transferred to serine (Ser) and/ or threonine (Thr) residues of specific proteins by an Oglycosidic bond. The synthesis of O-glycans takes place in the Golgi apparatus and is mostly initiated by the activity of polypeptide-GalNAc-transferases (pp-GalNAcTs, GALNTs) that link a single N-acetylgalactosamine (GalNAc) residue to Ser or Thr, thus forming the Tn antigen. In contrast to this glycan type which is termed mucin-type O-glycans, a wide range of nonmucin O-glycans exist, namely α-linked O-fucose, β-linked O-xylose, α-linked O-mannose, βlinked O-N-acetylglucosamine (O-GlcNAc), α- or β-linked O-galactose, and α- or β-linked O2

glucose glycans. Since there are only limited data on the functional consequences of these modifications, they will not be further addressed in this review. By elongation of the Tn antigen, four core structures named core 1 to core 4 emerge depending on the monosaccharide types and sequence. Both, synthesis of core 1 (also known as T antigen or Thomsen-Friedenreich antigen), catalysed via the glycosyltransferase C1GalT1 (T-synthase) and its chaperone COSMC, and core 2, which is formed through the action of the glycosyltransferase C2GnT, comprise galactose (Gal), whereas the latter has a -GlcNAc in addition (figure 1). The Tn antigen can be converted into core 3, containing one GlcNAc , followed by core 4, which includes two GlcNAc residues by the action of C2GnT and C3GnT. Further branching leads to more complex O-glycans including galactose, N-acetylglucosamine, fucose or sialic acid. The Tn and T antigens can be further sialylated creating sialyl-Tn antigen (via the alpha-2,6sialyltransferase ST6GalNAcI) and sialyl- of disialyl-T antigen (via alpha-2,3-sialyltransferase ST3Gal-I and ST6GalNAcI), respectively. Beside the T and Tn antigens, certain Lewis antigens are often found in tumour tissues where they enable binding of tumour cells to selectins expressed by endothelial or blood cells [1] (figure 1). Sialyl-Lewisa (SLea) is synthesised by adding β3-Gal to GlcNAc via β3Gal-transferase (β3-GalT), followed by the addition of α3-sialic acid by α3-sialyltransferase (ST3Gal) and of α4-Fuc to GlcNAc by α4-Gal-transferase (α4-FucT). In contrast, sialyl-Lewisx (SLex) is synthesised by the addition of β4-Gal to GlcNAc via β4-Gal-transferase (β4-GalT) closely followed by ST3Gal and of α3-Fuc to GlcNAc by α3-Fuc -transferase (α3-FucT). 1.2 Which proteins are O-glycosylated? Numerous proteins, including enzymes, transcription factors, receptors and structural proteins are regulated and/or modified by O-glycosylation and this posttranslational modification has been implicated repeatedly in cancer [2]. Many cell-surface or secreted proteins carry mucin-type O-glycans, which protects them against proteolytic degradation but also modulates recognition, adhesion and cell-cell-communication functions. Examples of O-glycosylated proteins are:  Nuclear phosphoprotein c-myc which regulates cell proliferation, differentiation and apoptosis  Cell surface proteins like mucins MUC1, MUC2 etc., sialomucin, CD44, integrins and other cell adhesion proteins involved in cell-cell-interaction, cell adhesion and migration  Structural proteins e.g. plakoglobin and β-catenin modulating cell-cell-adhesion and downstream signaling pathways (for example Wnt/ β-pathway)  C-type lectins (i.e. selectins) which also bind O-glycans  Receptors e.g. death receptors to control sensitivity to pro-apoptotic signals 1.3 O-glycans and metastasis Tumour-specific alterations in glycosylation can result either in loss or in alteration of carbohydrate structures. The tumour-associated Tn antigen is highly expressed in more than 90 % of breast tumours as well as 70-80% of colon, lung, bladder, cervix, ovarian, stomach and prostate carcinomas [3-5] and its presence is associated with high metastatic potential and poor prognosis [6, 7]. In histochemical studies, O-glycan antigens are frequently analyzed by use of specific lectins: In example, Helix pomatia agglutinin (HPA) specifically binds to Tn and peanut agglutinin (PNA) to core 1 and 2 structures. Binding of both lectins to breast cancer tissue has been shown to correlate with poor survival, lymphatic invasion and 3

lymph node metastasis in breast cancer patients [8]. Tn and sTn antigen are often simultaneously synthesised, and high levels of both promotes tumour growth and metastasis in vivo in breast and gastric tumours [9, 10]. The T antigen influences adhesion of tumour cells to the endothelium due to its interaction with galectin-3 which drives metastasis [11, 12]. The core 3 O-glycans also play a key role during malignant progression: In pancreatic cancer, core 3 structures are downregulated because of loss of functional core 3 synthase thereby modifying migration, invasion and metastasis [13]. SLe glycans bind to E- an P-selectins, which are mainly expressed in endothelial and immune cells, thereby promoting extravasation [14]. Diverse studies in the last years suggest that tumour cells mimic the leukocyte adhesion cascade in order to extravasate. Here, sLEselectin binding facilitates initial cell tethering and rolling which subsequently leads to firm cell adhesion, extravasation and metastasis [15, 16]. 1.4 Target proteins of O-glycosylation involved in metastasis Mucins are highly O-glycosylated proteins and aberrant glycosylation of these structures has been implicated in cancer. In breast cancer O-glycosylation of mucin 1 (MUC1) regulated via C1GALT1 promotes MUC1-C/ß-catenin signaling which leads to tumour cell growth, migration and invasion in vitro as well as tumour growth in vivo [17]. In another study the glycosyltransferase GALNT6 has been showed to glycosylate and stabilise MUC1 driving carcinogenesis via disruption of β-catenin- and E-cadherin-mediated cell adhesion [18]. O-glycosylation of the cell-surface glycoprotein CD44 is frequently altered in tumour cells [19]. In colon carcinoma cells O-glycosylated CD44 can bind endothelial E-selectin, which in turns contributes to metastasis [20]. In breast cancer cells some CD44 glycoforms possess high shear-resistant selectin-binding activity due to O-glycan- and N-glycan-based sialofucosylated terminal carbohydrate groups, whereas other CD44 variants lack this activity [21]. Integrins are also O-glycosylated which influences the attachment of tumour cells to the extracellular matrix (ECM) as well as cell-cell-interactions. In hepatocellular carcinoma cells, modification of integrin β1 via C1GALT1 regulates adhesion, migration and invasion and enhances metastasis in vivo [22]. Glycan binding proteins such as siglecs, galectins and selectins are themselves glycosylated and are highly specific for certain sugar moieties. These proteins are associated with cell-cell-recognition, cell adhesion and motility. In prostate cancer - β-galactoside– binding protein galectin-4 affects the expression of E-cadherin modulating EMT, invasion and metastasis [23]. This effect is dependent on O-glycosylation mediated by C1GALT1. Osteopontin (OPN), also known as bone sialoprotein I, has a pivotal role in bone remodeling, inflammation and cancer metastasis. Interestingly, OPN is highly O-glycosylated and its glycosylation status affects its phosphorylation and cell-adhesion activity [24]. 1.5 Enzymes involved in O-glycosylation and metastasis Overexpression or downregulation of glycosyltransferases and glycosidases affects the amount and pattern of glycans, influencing tumour progression and metastasis. The altered expression of glycosylation enzymes can be due to i) dysregulation at the transcriptional level, ii) dysregulation of chaperone function or iii) epigenetic mechanisms, i.e. methylation [25]. So far, 20 ppGalNAc-Ts (GALNTs) are noted and their under- or overexpression is frequently associated with tumour progression and metastasis. It has been shown for various

4

tumours entities that the relocation of ppGalNAc-Ts from the Golgi to the ER promoted cell adhesion, cell motility and invasiveness [26]. GALNT1 was found to be frequently up-regulated in hepatocellular carcinoma and this was correlated with poor patient survival [27]. Knockdown of GALNT1 decreased cell migration and invasion via diminishing EGFR signaling. Interestingly, in the same tumour entity GALNT2 was downregulated which was associated with vascular invasion and recurrence [28]. GALNT2 overexpression decreased EGFinduced cell growth, migration and invasion in vitro and in vivo. In gastric adenocarcinoma low GALNT2-expression induces migration, invasion and metastasis in vivo [29] and is a prognostic marker for better patient survival in neuroblastoma. Here GALNT2 affects Oglycan structures of IGF-1R and thus triggers downstream signaling events [30]. In pancreatic tumours downregulation of GALNT3 leads to increased cell motility and adhesion to the tumour endothelium [31]. Knockdown of GALNT3 increases expression of ßcatenin and E-cadherin [32]. The expression of GALNT3 has been described as a potential diagnostic and prognostic marker for pancreatic and lung cancer [33, 34]. C1GALT1 is overexpressed in hepatocellular carcinomas where its expression is associated with advanced tumour stage, metastasis and poor survival [35]. In experimental systems, C1GALT1 overexpression enhanced cell adhesion to ECM proteins, migration and invasion in vitro as well as metastasis in vivo. In breast cancer C1GALT1 is frequently up-regulated and promotes migration, invasion and growth of tumour xenografts via modulation of MUC1/ ß-catenin signaling [17]. In our own recent study, high C1GALT1 protein or mRNA expression was associated with shorter disease-free survival in breast cancer patients, and its prognostic value was clearly increased if, in addition, either GALNT1, GALNT8 or GALNT14 was highly expressed, enabling synthesis of T antigens [8]. Several other enzymes are relevant for the synthesis of selectin ligands and thus for hematogeneous metastasis. GCNT1 (C2GNT1) and GCNT4 (C2GNT3) are involved in formation of core 2 structures which serve as carriers of Lewis antigens. In a study based on microarray data of 194 breast cancer samples, both enzymes showed a significant correlation to shorter survival, even in multivariate analysis, and an increased frequency of distant metastasis [36]. In the same study, the fucosyltransfrases FUT1 was associated with poor prognosis and distant metastasis, whereas the fucosidase FUCA1 was a favorable prognostic indicator and associated with a low percentage of lung metastasis. In experimental settings, FUCA1 reduces the invasive potential of breast cancer cells and their adhesion to selectins and endothelial cells under flow conditions [37]. An additional prognostic glycosylation enzyme in breast cancer is the sialyl transferase ST3GAL6 [8] which has been shown to promote migration and metastasis in various tumour types [38, 39]. Further, the expression of ST6GalNAc-I, the major sialyl-Tn antigen synthase, correlates with invasion and metastasis. In hepatocarcinoma cells silencing of ST6GalNAc-I leads to tumour migration and invasion due to PI3K/AKT/NF-κB pathway [40]. 2 N-glycans 2.1 Biosynthesis and types of N-glycans In contrast to O-glycosylation, N-glycosylation takes place during translation of target proteins by addition of glycan structures to the amino group of asparagine (ASN) residues at the consensus motif asparagine-X-serine/threonine (NXS/T) in which X is any amino acid except proline. Instead of step-by-step addition of single sugar residues, N-glycosylation starts with synthesis of a dolichol-bound oligosaccharide precursor in the ER, consisting of 5

14 sugar moieties, among them 9 mannose residues. This oligosaccharide is then transferred to a suitable ASN residue within the nascent polypeptide by the oligosaccharyltransferase (OST) protein complex (figure 2). After this transfer, the correct folding and secretion of the glycoprotein depends on trimming of the glycan precursor in the Golgi, resulting in high-mannose glycans. Removal of part of the 9 mannose residues by Golgi mannosidases is the prerequisite of formation of complex or hybrid di-, tri- or tetraantennary glycans (figure 2). These can then undergo a variety of extensive modifications including sialylation, fucosylation, addition of galactose, GlcNAc etc., resulting in highly complex and heterogeneous structures. A specific modification is the addition of β1,4-bound GlcNAc to the beta-linked mannose of the trimannosyl core by MGAT3 resulting in the formation of “bisecting” N-glycans. Due to sterical features, this largely prevents further branching of the glycan. 2.2 Which proteins are N –glycosylated? In humans, there are more than 2000 proteins harboring an amino acid motif suitable for Nglycosylation. These are either membrane-bound or secreted, but never cytoplasmic or nuclear proteins. Examples of N-glycans are:  Adhesion proteins including members of the immunoglobulin superfamily (ALCAM, ICAM1, BCAM etc.), CD44, integrins and cadherins,  Secreted proteinases like kallikreins, cathepsins, carboxypeptidase E, matrix metalloproteinases, PSA etc.  receptors like EGFR, HER2/neu, TGFβ receptor, IGF2R etc.  ECM molecules like fibronectin, laminin etc.  Wnt family members  c-Kit, TIMP1, tetraspanins, clusterin etc. 2.3 N-glycans and metastasis In human tumours, cancer-specific N-glycan alterations include: 1. premature termination of glycan processing, leading to accumulation of high-mannose glycans in the cells; 2. reduction in bisecting glycans due to a reduced expression of MGAT3 (GnT-III); 3. increased branching due to high MGAT5 (GnT-V) expression, which facilitates formation of complex glycans, 4. increased fucosylation of the innermost GlcNAc residue (core fucosylation) and 5. terminal modifications like sialylation, fucosylation, lactose addition etc. (figure 2). These modifications also allow formation of selectin-binding structures like sLeX or sLeA, promoting extravasation of circulating tumour cells. Formation of bisecting glycans and N-glycan branching often display opposite effects on metastasis. Studies on the OvCa cell line SKOV3 and its highly metastatic derivative, SKOV3-ip, have shown up-regulation of high-mannose and complex glycans, whereas bisecting glycans were down-regulated in metastatic cells [41]. By mass spectrometry or lectin staining, increased levels of N-glycan branching with tri- and tetraantennary structures and sialylation have been detected in highly metastatic prostate cancer cell lines [42]. Breast cancer cells are characterized by dramatic increases in high-mannose and complex triantennary N-glycans compared to normal tissue [43]. In our own analysis of a breast cancer cohort, positive binding results for Galanthus nivalis lectin (GNA; detecting high-mannose structures) and Phaseolus vulgaris leucoagglutinin (PHA-L; detecting complex N-glycans with terminal Gal, GalNAc and Man) correlated with vascular invasion and, partly, lymph node involvement [8]. In another histochemical study, increased PHA-L binding in breast 6

cancer metastases compared to primary tumours and a shorter survival in PHA-L-positive patients was found, suggesting that complex β1,6-branched N-glycans promote breast cancer progression [44]. Yet, glycan changes during tumourigenesis partly vary in different tumour entities. In example, complex N-glycans were rarely found in ovarian carcinoma tissue [45], and imaging MS techniques showed that OvCa cells predominantly expressed high-mannose N-glycans whereas complex glycans were mainly detected in the tumour stroma [46]. Similarly, bisecting N-glycans and terminal GalNAc-GlcNAc (lacdiNAc) structures are elevated in OvCa cells and tissue, but low in breast cancer [47]. Sialylation is a feature of many complex glycans. In various cancer types, an increased sialylation of N-glycans has been shown which contributes to the stability and activity of target proteins (reviewed by [48]. In hepatocarcinoma cells, sialylated N-glycans promoted cell invasion and adhesion to lymph nodes as well as lymphatic metastasis in vivo, and neuraminidase treatment reverted this effect [40]. A specific modification is addition of poly N-acetyllactosamine (polylacNAc) to complex N-glycans. PolylacNAc is a high-affinity ligand for galectin-3, which is expressed constitutively in the lung including vascular epithelia, and has been reported to facilitate lung metastasis in melanoma [49]. Inhibition of the galectin3/PolylacNAc interaction blocks cancer cell adhesion and experimental metastasis [50].

2.4 Target proteins of N-glycosylation involved in metastasis Glycosylation of adhesion proteins can largely influence their binding properties, leading to altered cell-cell or cell-matrix contacts and contributing to EMT. N-glycosylation of Ecadherin has strong impact on its function as adhesion molecule, signalling protein and tumour-suppressor. Promoter analyses have shown that expression of the first enzyme of Nglycosylation, DPAG1, is activated by the Wnt/β-catenin pathway which results in strong Nglycosylation of E-cadherin and reduction of cell-cell adhesion in adherens junctions [51]. In normal epithelial cells, E-cadherin has been shown to harbor mainly bisecting N-glycans due to the activity of the MGAT3 enzyme. In carcinoma cells, MGAT3 is frequently downregulated by promoter methylation, and its counterpart MGAT5 is up-regulated leading to the formation of tri- and tetraantennary complex glycans on cadherins and other proteins. This results in E-cadherin internalization to the cytoplasm and disruption of cell-cell contacts, which compromises the associated signalling pathways, resulting in EMT, invasion and metastasis [52]. Further analysis revealed that especially the formation of complex branched glycans at Asn-554 of E-cadherin leads to disruption of E-cadherin dimers and promotes tumourigenesis in gastric cancer cells [53]. The antagonist of E-cadherin in EMT, N-Cadherin, also contains N-glycosylation sites which are relevant for homophilic adhesion: Knockdown of MGAT5 expression in fibrosarcoma cells resulted in decreased levels of complex glycans, increased cell-cell contacts, decreased migration and invasion as well as enhanced outside-in signalling as shown by ERK phosphorylation [54]. N-glycosylation has also been shown to be essential for integrins and their role in metastasis. Integrin α6β4 is dependent on N-glycosylation of the β4 subunit for proper cell adhesion, migration on laminin and signalling [55]. Overexpression of MGAT5 in HT1080 fibrosarcoma cells resulted in decreased adhesion to fibronectin, but increased migration and invasion, caused by enhanced N-glycan branching of β1-integrin [56]. MGAT5 expression is 7

stimulated by several oncogenes like src, HER2/neu, ras etc., whose activity can be partly explained by their effects on N-glycosylation. The activity of integrins is partly regulated by sialylation of their N-glycans. In colorectal cancer cells, overexpression of ST6GAL1 lead to highly sialylated integrin β1, increased adhesion to collagen I and laminin and enhanced migration [57]. Conversely, up-regulation of the neuraminidase neu1 induced desialylation of β4 integrin and reduced liver metastases in a mouse model [58]. Experiments with overexpression of either MGAT3 or MGAT5 in gastric cancer cells demonstrated that the integrin ligand laminin-332 is also N-glycosylated, and that increased N-glycan branching has an influence on integrin α1β1 clustering, cell motility and adhesion [59]. CD44 isoforms are highly O- and N-glycosylated adhesion proteins. Sialic-acid capped Nglycans of CD44 inhibit its binding to hyaluronan which has strong impact on cell migration [60]. Cancer cell invasion depends on the activity of proteolytic enzymes, which cooperate to degrade the extracellular matrix. Matrix-metalloproteinases (MMPs) degrade a variety of substrates, and their expression is partly associated with an unfavorable prognosis in various tumour types. Several MMPs harbor N-glycosylation sites within their catalytic domain, but their functional meaning is only poorly understood (summarized by [61]). In HT1080 leukemia cells, MMP1 carries highly complex glycans, which frequently end in sialyl Lewis X antigens. Thus, MMP1 might bind to the surface of activated, selectin-presenting endothelial cells, promoting invasion and metastasis. The MMP9 N-glycans are partially sialylated, corefucosylated structures, and sialidase treatment alters interaction with the MMP inhibitor TIMP1 [61]. N-glycosylation has also been reported for kallikreins [62] and cathepsins. For cathepsin V, proper N-glycosylation is required for secretion and enzyme activity in human tumour cells [63]. Similar to O-glycans, N-glycans can influence the process of adhesion to endothelial cells and extravasation since complex N-glycans might harbor selectin-binding sites like sLeX. In CRISP/Cas experiments blocking either O-glycosylation, N-glycosylation or synthesis of glycolipids in leukemia cells, N-glycans were important for E-selectin binding and rolling on endothelial cells, while O-glycans were essential for P- and L-selectin interactions [64]. Glycosylated CD44v isoforms also confer E-selectin binding of breast cancer cells under physiological shear conditions. Surprisingly, experiments with selective inhibitors of O- or Nglycosylation demonstrated that the responsible CD44 epitopes are presented by N-glycans [21].

2.5 Enzymes involved in N-glycosylation and metastasis Several glycosylation enzymes have been shown to be associated with metastasis in experimental studies or clinical tumour tissues, without further knowledge about the target proteins, which are responsible for these effects: RPN1 (ribophorin), as part of the OST complex, has been shown to facilitate N-glycosylation of specific substrates, while it is not involved in modification of other molecules [65]. Interestingly, high RPN1 mRNA levels in two breast cancer cohorts were significantly associated with shorter recurrence-free survival, even in multivariate analysis, and high RPN1 expression was associated with the presence of distant metastasis [36]. RPN2 expression as determined by IHC has also been reported to be associated with aggressive features of breast cancer, and the combination of RPN2 and p53 positivity correlated with 8

poor prognosis [66]. A prognostic impact was shown in human osteosarcoma patients, where high RPN2 expression correlates with a high frequency of metastasis after initial diagnosis, In a xenograft model using highly metastatic osteosarcoma cells, RPN2 silencing led to a reduced tumour growth and lung metastasis [67], and in colorectal carcinomas, high RPN2 mRNA or protein levels correlated with distant metastasis [68]. N33 (TUSC3) also associates with the OST complex where it regulates N-glycosylation of specific substrates. It was suggested as a potential tumour suppressor in various tumour types including prostate carcinoma, pancreatic cancer or head and neck squamous cell carcinomas [69]. In the latter tumour, loss of N33 correlated with advanced stage, lymph node involvement and poor survival. Pils et al. detected TUSC3 methylation in ovarian cancer samples, where it correlated with significantly shorter RFS and OAS. Reconstitution of N33 expression in two ovarian and pancreatic cell lines decreased adhesion to collagen [70] suggesting a role of N-glycosylation in ovarian cancer progression, and TUSC3 silencing in OvCa cell lines enhanced migration and proliferation [71]. In addition to RPN1, our own microarray study revealed additional N-glycosylation enzymes with prognostic relevance in breast cancer: The trimming enzymes GCS1 and GANAB were associated with shorter survival and distant metastasis, whereas high levels of the Golgi mannosidase MAN1A1 which degrades high-mannose glycans correlated with favorable prognosis in this tumour type [36]. MAN1A1 was also found to be down-regulated in metastatic hepatocellular cancer (HCC) cell lines and orthotopic xenograft tumours as compared to non-metastatic HCC controls in a small study on transcriptional profiling of glycogenes [22]. MGAT5 overexpression in cancer cells generally leads to an increase in β1-6 GlcNAc branching, which is necessary for formation of complex tri- and tetraantennary structures associated with the metastatic phenotype [72]. Accordingly, in immunohistochemical studies MGAT5 was associated with metastasis and poor prognosis in hepatocellular and renal carcinomas [73, 74]. ST6GAL1 which catalyzes transfer of sialic acids to terminal galactose residues of N-glycans is up-regulated in numerous cancers where it is associated with changes in adhesion, migration, invasion, and poor prognosis for patients. ST6GAL1 up-regulation led to metastatic spread in human CRC cells [75]. Similarly, high ST6GAL1 levels in mammary carcinoma cells and human anaplastic large cell lymphoma led to increased adhesion to ECM structures and increased invasiveness [76]. FUT8 catalyzes fucosylation of the innermost GlcNAc residue of N-glycans (core fucosylation). In several experimental studies, high FUT8 expression in cancer cells leads to higher motility and metastatic potential, and in breast cancer or NSCLC, it was associated with significantly shorter OAS and RFS [77, 78]. Interestingly, core fucosylation of Ecadherin leads to higher adhesion and less migration in experimental systems [79, 80]. Thus, the prognostic impact of FUT8 cannot depend on this adhesion molecule. In gastric cancer which is frequently linked to E-cadherin loss, there are less core-fucosylated structures and FUT8 is downregulated in cancer cells [81]. 3 Glycosaminoglycans (GAGs) 3.1 Biosynthesis and types of GAGs GAGs are long unbranched polysaccharides consisting of repeating disaccharide units, one of the sugars being either N-acetylglucosamine (GlcNAc) or N-acetylgalactosamine (GalNAc) 9

and the other being an uronic acid (glucuronic acid or iduronic acid) or galactose. GAGs can be classified into six major categories: hyaluronan (HA), chondroitin sulfate (CS) and dermatan sulfate (DS), heparin and heparan sulfate (HS) and keratan sulfate (figure 3A). The GAG polysaccharide backbones, excepting hyaluronan, can be modified by sulfatation and uronate epimerization, leading to a substantial degree of structural variability that in turn accounts for a wide diversity of biological activities. Sulfated GAG chains are synthetized in the Golgi apparatus by extending a preformed tetrasacharide (GlcUA-Gal-Gal-Xyl-), which is attached to a serine residue of the backbone protein, with disaccharide repeats. Subsequently, GAG chains are modified by sulfation at various positions and epimerization at GlcUA residues. A variety of enzymes including glycosyltransferases, sulfatases and epimerase are involved in this process [82]. Synthesis of CS, DS, heparin and heparan sulfate is initiated by the attachment of xylose to specific serine residues. This process, known as xylosylation, depends on UDP-xylose levels and on the activity of xylosyltransferases and regulates the number and location of the GAG chains. In contrast to the rest of GAGs, HA lacks a covalent bond to protein structures, but instead can interact non-covalently with proteoglycans via hyaluronan-binding motifs. HA is synthesized on the inner cell surface by hyaluronan synthases (HAS1, HAS2 and HAS3) and simultaneously extruded to the outside of the cell, where it is either released into the ECM, remains attached to the plasma membrane or is internalized again. Degradation of HA requires hyaluronidases (HYAL1 and HYAL2). Here, the concerted action of certain hyaluronan synthases and hyaluronidases leads to the formation of different HA-size fragments that have wide-ranging and often opposing functions. 3.2 Which proteins carry GAG chains? Proteoglycans (PG) comprise a protein backbone to which one or more GAG chains are covalently attached. Although most proteoglycans also contain O- and N-glycans, the GAG chains are much larger and dominate the chemical and functional characteristics of this class of glycoproteins. Depending on their localization, proteoglycans can be grouped into three categories: extracellular, membrane-bound and intracellular proteoglycans (figure 3B). 

Serglycin is the only proteoglycan present in the cytoplasm, where it is packaged in storage granules. In connective tissue mast cells, serglycin carries several heparin chains and therefore represents a unique proteoglycan containing this GAG type [83].



Diverse membrane-associated proteoglycans are known, mainly carrying CS (syndecan-1/-3, NG2 and betaglycan) and/or HS chains (syndecan-2/-4, betaglycan, and glypican). Syndecans consist of an intracellular domain, transmembrane and ectodomain, which allow interactions with a variety of extracellular ligands (i.e. growth factors and ECM components) leading to signal transduction events [84]. NG2 (neuron glia antigen-2) and betaglycan (TGFβ type III receptor) are transmembrane proteoglycans carrying CS and HS/CS, respectively. Further transmembrane proteoglycans are phosphacan, which is mainly expressed in the CNS and carries KS or CS chains, and certain splice forms of CD44 (CD44v3) that carry HS and/or CS and can bind hyaluronan. In

10

contrast, glypicans are bound to the plasma membrane via a GPI anchor and carry only HS chains. 

A large diversity of proteoglycans is present within the ECM. Among them, the biggest family is the small leucine-rich proteoglycan (SLRPs) group, consisting of at least 20 members, i.e. biglycan, decorin, lumican, which carry CS, DS, or KS chains. Further, the aggrecan family of proteoglycans that includes aggrecan, versican, brevican, and neurocan, mainly carrying CS, has been shown to contain additional hyaluronan binding domains. The basement membrane separates epithelial cells from the underlying lamina propia and consists mainly of laminin, proteoglycans and collagens, which can also carry GAG chains. Among the proteoglycans, perlecan, agrin, collagen XV (all carrying HS) and collagen type XVIII (carrying CS) are the best characterized.

3.3 GAGs/Proteoglycans and metastasis Within the glycosaminoglycan family only heparan sulphate, chondroitin/dermatan sulphate and hyaluronan show a well-documented role in cancer progression. Chondroitin sulfate (CS) is composed of repeating disaccharides of GalNAc and Dglucuronic acid (GlcA) with different sulfation patterns, mediated by specific carbohydrate sulfotransferases (CHSTs): CS-A [GlcA-GalNAc-4-sulfate] and CS-C [GlcA-GalNAc-6-Osulfate] contain monosulfated units, whereas CS-D [GlcA(2-O-sulfate)–GalNAc(6-O-sulfate)] and CS-E show multiple sulfation patterns [85]. The sulfation of CS strongly influences the biological function of these GAG and represent a critical factor in cancer progression [86]. A variable effect of CS/CSPGs (chondroitin sulfate proteoglycans) has been described depending on their localization. Tumour cell-associated CS/CSPGs seem to promote migration and invasion. Here, oncofetal CS chains (ofCS), which are normally restricted to the placenta, have been found in proteoglycans of tumour and tumour-infiltrated stromal cells in several cancer entities, and blocking of these structures led to reduced migration, invasion and anchorage-independent growth in vitro as well as inhibited seeding and spreading of tumour cells in vivo [87]. Also, in invasive brain tumours several CSPG core proteins have been described to promote tumour cell invasion [88-90]. In 2001, Iida et al described the key role of CS in the MMP-mediated human melanoma invasion. The same group could later show that chondroitin 4-sulfate (C4S) but not chondroitin 6-sulfate (C6S) binds and facilitates activation of pro-MMP2, strongly affecting tumour cell invasion [91, 92]. In contrast, other studies showed that CSPGs localized in the ECM strongly inhibit tumour cell invasion of human glioma cells [93]. In line with these data, decorin and chondroitin-6-sulfate inhibit B16V melanoma cell migration and invasion by acidification of the cellular suface [94]. On the contrary, and increasing the complexity of CS function in metastasis, small CS-E fragments, present in the ECM, enhance tumour cell motility via inducing CD44 shedding [95]. Tumour cell interactions with platelets and endothelial cells are of key importance for extravasation and subsequent colonization. CSPGs present on the surface of highly metastatic cancer cells mediate this essential adhesion step via binding to P-selectin, which is highly expressed on activated platelets and endothelial cells [96]. Here, the CS sulfation pattern seems to play an important role, since only CS-E and CS-B, but not CS-C/D, bind to P- and to some extent to L-selectin [97]. Further, CS-A has been described to regulate fibrosarcoma cell adhesion and migration via JNK and tyrosine kinase signalling pathways [98]. Another important aspect of CS function is the interaction with growth factors in the ECM. Thus, CS chains bind and store growth factors being able to release them 11

progressively. In example, FGF2 can be bound by CS-E and CS-A building a complex with the corresponding receptor and thereby regulating its activation [99]. Heparan sulfate (HS) is composed of alternate units of GlcNAc and uronic acid with the GlcNAc residues often being further processed by acetyl group removal and subsequent sulfatation. HS is located at the cell surface and within the ECM -as HSPG (heparan sulfate proteoglycan)- and modulates cell-cell and cell-ECM interactions. The role of diverse HSPGs (i.e. syndecans, glypicans and perlecan) on metastasis has been exhaustively analysed in the past years, whereas less information exists about the direct function of HS chains in this process. Originally, HS was considered as antitumourigenic, since reduced levels of HS chains correlated with high metastatic activity in different tumour entities, in part due to reduced cell-cell adhesion and increased invasion [100]. Recently, it has been recognized that the same HSPG might act as either inhibitor or promoter of tumour cell progression depending not only on the cancer type and stage but also on the modulation of the HS chains. A direct implication of the HS chains in syndecan-1 function has been demonstrated by analyzing the bioactivity of different syndecan-1 forms bearing mutations in one or more HS binding sites. As the number of HS chains attached to syndecan-1 decreased, its ability to inhibit invasion was also reduced [101]. Similarly, blocking of HS chains in glypican-3 with a specific antibody inhibits migration of hepatocellular carcinoma cells [102]. Hyaluronan is composed of disaccharide repeats of GlcUA and GlcNAc bound alternately by β-1,3 and β-1,4 glycosidic bonds. As described above, these polymers are hydrolysed by hyaluronidases, thereby leading to a big variety of HA chain lengths. Thus, the sizes and quantity of the HA polymers and the expression of HA-associated enzymes impact tumour progression in a manifold manner. Newly synthetized HA is located pericellularly, mainly attached to specific receptors, or in the ECM modulating the cell microenviroment. In addition, HA fragments may interact with their receptors, i.e. CD44, RHAMM, LYVE-1, acting as a signal mediator of key importance during tumour progression. By attaching to CD44, newly synthesized HA chains provide a highly hydrated environment that facilitates migration and invasion of breast cancer cells [103]. Recently, several studies described low molecular weight hyaluronan (LMW-HA) as a key player for tumour progression and metastasis. LMWHA induces cell motility via up-regulation of MMPs, invasion as well as integrin-mediated adhesion to endothelial cells [104, 105]. 3.4 Proteoglycans involved in metastasis The structural and functional diversity as well as their abundance and location predestine proteoglycans as key molecules in the metastatic cascade. Indeed, depending on the PG type, tumour entity and structure of the GAG chains, PGs have been described to either promote or inhibit carcinogenesis. Syndecan-1 loss is associated with EMT and accelerated tumour progression in some cancer types [106], whereas in others like myeloma or breast cancer it shows an antitumourigenic function, namely it raises cell-matrix adhesion and inhibits invasion. Here, a direct implication of the HS chains has been demonstrated by analyzing different syndecan-1 forms bearing mutations in one or more HS binding sites. As the number of HS chains attached to syndecan-1 decreases, the ability to inhibit invasion was also reduced [101]). Controversial functions have also been described for syndecan-4 in melanoma cells: While one study reports increased motility and decreased attachment on fibronectin [107], a different study shows reduced invasion after syndecan-4 knock down by decreased Wnt5A 12

signalling [108]. The role of glypican family members in carcinogenesis also varies depending on the cell of origin. Glypican-3 is down-regulated in some human tumours, including mesothelioma, breast and ovarian cancer. In breast cancer cells decreased glypican-3 expression lead to reduced adhesion to fibronectin and increased migration in vitro as well as metastasis in vivo. In contrast, GPC3 overexpression is associated with a poor prognosis for patients with HCC, and it may also have predictive potential for HCC invasion and metastasis [109]. A reduced expression of Glypican-5 in non-small lung cancer samples has been described compared to adjacent noncancerous tissues. Here, overexpression of glypican-5 in NSCLC cell lines significantly reduced their migration and invasion [110]. Further, both glypican-3 and glypican-5 suppress EMT in breast cancer and lung adenocarcinoma, respectively [111]. Serglycin is mainly found in secretory granules or vesicles in cells of hematopoietic origin, where it modulates the secretion of proteases, chemokines, or other cytokines. In breast and nasopharyngeal cancer, secreted serglycin has been found to promote metastasis via increasing the migratory and invasive properties of tumour cells [112, 113]. Perlecan, the most abundant HSPG of epithelial and endothelial basement membranes, has been shown to promote cell invasion [114]. Its down-regulation inhibits tumour growth and angiogenesis in vivo and suppresses the invasive behavior of melanoma cells in vitro [115, 116]. Biglycan, a SLRP family member, is secreted by tumour-associated endothelial cells and activates tumour cell migration thereby promoting metastatis [117, 118]. Versican regulates the development of peritoneal metastasis in ovarian cancer by promoting adhesion of tumour cells and tumour cell spheroids to mesothelial cells as well as subsequent spheroid disaggregation [119]. 3.5 GAG-related enzymes involved in metastasis In addition to the GAG structures themselves diverse enzymes involved in their synthesis and/or modulation play an important role for metastatic spread of tumour cells. Chondroitin sulfotransferase 11 (CHST11) has been described to be relevant for ovarian cancer progression and to play a key role in breast cancer metastasis via regulation of Pselectin ligands [85, 120]. The evidence that diverse heparan sulfate-proteases strongly influence metastasis supports indirectly the pivotal role of the HS chains themselves during this process. Interestingly, a recent study has shown that after HS chain cleavage HSPG core proteins may undergo new bindings thereby initiating downstream pro-tumourigenic signalling events [121]. The catabolic enzyme heparanase, that cleaves intact HS chains in basement membranes, the underlying ECM and on the cell surface, has been repeatedly described as a potent tumour promotor [122, 123]. Main consequences of heparanase activity are the release of growth factors and cytokines, i.e. angiogenic factors, and remodeling of the extracellular matrix. Thus, the heparanase-mediated disassembly of the tumour cell environment promotes invasion and metastasis. Syndecan-1 and perlecan are targets of heparanase, and their degradation modulates tumour cell invasion and metastasis [124]. Furthermore, heparanase expression increases cell adhesion by cleavage of HS chains of syndecans and glypicans [125-127] in various cell types reviewed by Levy-Adam et al. [128]. Also the expression of HA-related enzymes has been reported to influence metastasis by not fully clarified mechanisms [129]. Hyaluronan synthase 2 (HAS2) promotes tumour cell invasion via deregulation of TIMP-1, and HAS3 (hyaluronan synthase 3) inhibition led to a 13

decreased pericellular HA matrix thereby inhibiting anchorage-independent growth in primary colon carcinoma cells [130]. In contrast, increased expression of HAS2 and HAS3 promotes tumour cell adhesion to bone marrow endothelial cells, presumably via increasing the cell surface hyaluronan matrix [131]. Unexpectedly, hyaluronidases also correlate with tumour progression: HYAL1 promotes tumour cell proliferation, motility and invasion [132-134], and increased levels of LMW-HA, which has been previously described as pro-metastatic, are associated with overexpression of HAS2 as well as HYAL1 and HYAL2 [105]. Several enzymes involved in the synthesis and modulation of GAG correlate significantly with distant metastasis, and interestingly some of them with metastasis to specific organ sites in breast cancer. Here, high HAS2 and HYAL1 mRNA levels were associated with brain metastasis, whereas tumours metastasizing to the lung and bone showed increased expression of CSGALNACT2 (chondroitin sulfate N-acetylgalactosaminyltransferase) and XYLT2 (xylosyltransferase 2), respectively [36].

4

Conclusion and final remarks

Besides the extensive research on genome, transcriptome and proteome, a key role of glycans in cancer progression has been acknowledged within the last two decades. On the one hand, diverse glycan structures as well as changes in the glycosylation pattern of certain glycoproteins have been described as biomarkers for prognosis and monitoring in different entities [135]. On the other hand, the fact that glycoproteins are mainly located on the cell surface and in the ECM thereby modulating cell interactions with the environment, predestine glycoproteins and glycan structures as key players during metastasis. Several glycan structures, including O-glycans, N-glycans and GAGs, as well as numerous glycoproteins have been described to affect crucial steps of the metastatic cascade (Figure 4) such as EMT (E-cadherin, N-cadherin, syndecan, galectin-4, etc.), cell motility (integrin β1 and β4, laminin-322, syndecan-1, glypican-3, mucin etc.) invasion and intravasation (cathepsin V, MMPs, kallikreins, serglycin, perlecan, CS, HS, HA, mucin, core 3, etc.), and extravasation (CD44, MMP1, biglycan, CS, sLex, sLea, etc..). Deregulation of specific glycosylation enzymes in tumour cells leads to an altered glycan pattern of numerous proteins and eventually to an altered protein function that might represent an advantage for tumour cells during certain steps of the metastatic cascade and even influence the site of metastasis. Understanding the fine regulation of the tumour cell glycosylation machinery might broaden our knowledge about the mechanisms of metastasis in different tumour entities and eventually lead to the development of novel therapeutic approaches.

We apologize to all authors whose interesting results could not be mentioned due to space limitation.

The authors declare that there are no conflicts of interest.

14

References 1. 2. 3. 4. 5. 6.

7. 8.

9.

10.

11. 12.

13.

14.

15.

16.

17.

18.

Hauselmann I, Borsig L: Altered tumor-cell glycosylation promotes metastasis. Front Oncol 2014, 4:28. Bektas M, Rubenstein DS: The role of intracellular protein O-glycosylation in cell adhesion and disease. J Biomed Res 2011, 25(4):227-236. Desai PR: Immunoreactive T and Tn antigens in malignancy: role in carcinoma diagnosis, prognosis, and immunotherapy. Transfus Med Rev 2000, 14(4):312-325. Springer GF: T and Tn, general carcinoma autoantigens. Science 1984, 224(4654):1198-1206. Springer GF: Immunoreactive T and Tn epitopes in cancer diagnosis, prognosis, and immunotherapy. J Mol Med (Berl) 1997, 75(8):594-602. Inoue M, Ogawa H, Nakanishi K, Tanizawa O, Karino K, Endo J: Clinical value of sialyl Tn antigen in patients with gynecologic tumors. Obstetrics and gynecology 1990, 75(6):10321036. Numa F, Tsunaga N, Michioka T, Nawata S, Ogata H, Kato H: Tissue expression of Sialyl Tn antigen in gynecologic tumors. Journal of obstetrics and gynaecology 1995, 21(4):385-389. Milde-Langosch K, Schutze D, Oliveira-Ferrer L, Wikman H, Muller V, Lebok P, Pantel K, Schroder C, Witzel I, Schumacher U: Relevance of betaGal-betaGalNAc-containing glycans and the enzymes involved in their synthesis for invasion and survival in breast cancer patients. Breast Cancer Res Treat 2015, 151(3):515-528. Julien S, Adriaenssens E, Ottenberg K, Furlan A, Courtand G, Vercoutter-Edouart AS, Hanisch FG, Delannoy P, Le Bourhis X: 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. Ozaki H, Matsuzaki H, Ando H, Kaji H, Nakanishi H, Ikehara Y, Narimatsu H: 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):229-238. Bian CF, Zhang Y, Sun H, Li DF, Wang DC: Structural basis for distinct binding properties of the human galectins to Thomsen-Friedenreich antigen. PLoS One 2011, 6(9):e25007. Glinsky VV, Glinsky GV, Rittenhouse-Olson K, Huflejt ME, Glinskii OV, Deutscher SL, Quinn TP: The role of Thomsen-Friedenreich antigen in adhesion of human breast and prostate cancer cells to the endothelium. Cancer research 2001, 61(12):4851-4857. Radhakrishnan P, Grandgenett PM, Mohr AM, Bunt SK, Yu F, Chowdhury S, Hollingsworth MA: Expression of core 3 synthase in human pancreatic cancer cells suppresses tumor growth and metastasis. Int J Cancer 2013, 133(12):2824-2833. Kojima N, Handa K, Newman W, Hakomori S: Inhibition of selectin-dependent tumor cell adhesion to endothelial cells and platelets by blocking O-glycosylation of these cells. Biochemical and biophysical research communications 1992, 182(3):1288-1295. Gebauer F, Wicklein D, Stubke K, Nehmann N, Schmidt A, Salamon J, Peldschus K, Nentwich MF, Adam G, Tolstonog G et al: Selectin binding is essential for peritoneal carcinomatosis in a xenograft model of human pancreatic adenocarcinoma in pfp--/rag2-- mice. Gut 2013, 62(5):741-750. Wicklein D, Schmidt A, Labitzky V, Ullrich S, Valent P, Schumacher U: E- and p-selectins are essential for repopulation of chronic myelogenous and chronic eosinophilic leukemias in a scid mouse xenograft model. PLoS One 2013, 8(7):e70139. Chou CH, Huang MJ, Chen CH, Shyu MK, Huang J, Hung JS, Huang CS, Huang MC: Upregulation of C1GALT1 promotes breast cancer cell growth through MUC1-C signaling pathway. Oncotarget 2015, 6(8):6123-6135. Park JH, Nishidate T, Kijima K, Ohashi T, Takegawa K, Fujikane T, Hirata K, Nakamura Y, Katagiri T: Critical roles of mucin 1 glycosylation by transactivated polypeptide N15

19.

20. 21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

acetylgalactosaminyltransferase 6 in mammary carcinogenesis. Cancer research 2010, 70(7):2759-2769. Matsuki H, Yonezawa K, Obata K, Iwata K, Nakamura H, Okada Y, Seiki M: Monoclonal antibodies with defined recognition sequences in the stem region of CD44: detection of differential glycosylation of CD44 between tumor and stromal cells in tissue. Cancer research 2003, 63(23):8278-8283. Hanley WD, Burdick MM, Konstantopoulos K, Sackstein R: CD44 on LS174T colon carcinoma cells possesses E-selectin ligand activity. Cancer research 2005, 65(13):5812-5817. Shirure VS, Liu T, Delgadillo LF, Cuckler CM, Tees DF, Benencia F, Goetz DJ, Burdick MM: CD44 variant isoforms expressed by breast cancer cells are functional E-selectin ligands under flow conditions. Am J Physiol Cell Physiol 2015, 308(1):C68-78. Liu T, Zhang S, Chen J, Jiang K, Zhang Q, Guo K, Liu Y: The transcriptional profiling of glycogenes associated with hepatocellular carcinoma metastasis. PLoS One 2014, 9(9):e107941. Tsai CH, Tzeng SF, Chao TK, Tsai CY, Yang YC, Lee MT, Hwang JJ, Chou YC, Tsai MH, Cha TL et al: Metastatic Progression of Prostate Cancer Is Mediated by Autonomous Binding of Galectin-4-O-Glycan to Cancer Cells. Cancer research 2016, 76(19):5756-5767. Kariya Y, Kanno M, Matsumoto-Morita K, Konno M, Yamaguchi Y, Hashimoto Y: Osteopontin O-glycosylation contributes to its phosphorylation and cell-adhesion properties. Biochem J 2014, 463(1):93-102. Pangeni RP, Channathodiyil P, Huen DS, Eagles LW, Johal BK, Pasha D, Hadjistephanou N, Nevell O, Davies CL, Adewumi AI et al: The GALNT9, BNC1 and CCDC8 genes are frequently epigenetically dysregulated in breast tumours that metastasise to the brain. Clinical epigenetics 2015, 7:57. Gill DJ, Tham KM, Chia J, Wang SC, Steentoft C, Clausen H, Bard-Chapeau EA, Bard FA: Initiation of GalNAc-type O-glycosylation in the endoplasmic reticulum promotes cancer cell invasiveness. Proc Natl Acad Sci U S A 2013, 110(34):E3152-3161. Huang MJ, Hu RH, Chou CH, Hsu CL, Liu YW, Huang J, Hung JS, Lai IR, Juan HF, Yu SL et al: Knockdown of GALNT1 suppresses malignant phenotype of hepatocellular carcinoma by suppressing EGFR signaling. Oncotarget 2015, 6(8):5650-5665. Wu YM, Liu CH, Hu RH, Huang MJ, Lee JJ, Chen CH, Huang J, Lai HS, Lee PH, Hsu WM et al: Mucin glycosylating enzyme GALNT2 regulates the malignant character of hepatocellular carcinoma by modifying the EGF receptor. Cancer research 2011, 71(23):7270-7279. Liu SY, Shun CT, Hung KY, Juan HF, Hsu CL, Huang MC, Lai IR: Mucin glycosylating enzyme GALNT2 suppresses malignancy in gastric adenocarcinoma by reducing MET phosphorylation. Oncotarget 2016, 7(10):11251-11262. Ho WL, Chou CH, Jeng YM, Lu MY, Yang YL, Jou ST, Lin DT, Chang HH, Lin KH, Hsu WM et al: GALNT2 suppresses malignant phenotypes through IGF-1 receptor and predicts favorable prognosis in neuroblastoma. Oncotarget 2014, 5(23):12247-12259. Chugh S, Meza J, Sheinin YM, Ponnusamy MP, Batra SK: Loss of Nacetylgalactosaminyltransferase 3 in poorly differentiated pancreatic cancer: augmented aggressiveness and aberrant ErbB family glycosylation. Br J Cancer 2016, 114(12):13761386. Wang ZQ, Bachvarova M, Morin C, Plante M, Gregoire J, Renaud MC, Sebastianelli A, Bachvarov D: Role of the polypeptide N-acetylgalactosaminyltransferase 3 in ovarian cancer progression: possible implications in abnormal mucin O-glycosylation. Oncotarget 2014, 5(2):544-560. Dosaka-Akita H, Kinoshita I, Yamazaki K, Izumi H, Itoh T, Katoh H, Nishimura M, Matsuo K, Yamada Y, Kohno K: N-acetylgalactosaminyl transferase-3 is a potential new marker for non-small cell lung cancers. Br J Cancer 2002, 87(7):751-755. Yamamoto S, Nakamori S, Tsujie M, Takahashi Y, Nagano H, Dono K, Umeshita K, Sakon M, Tomita Y, Hoshida Y et al: Expression of uridine diphosphate N-acetyl-alpha-D16

35.

36.

37.

38.

39.

40.

41.

42.

43. 44.

45. 46.

47.

48. 49.

galactosamine: polypeptide N-acetylgalactosaminyl transferase 3 in adenocarcinoma of the pancreas. Pathobiology 2004, 71(1):12-18. Wu YM, Liu CH, Huang MJ, Lai HS, Lee PH, Hu RH, Huang MC: C1GALT1 enhances proliferation of hepatocellular carcinoma cells via modulating MET glycosylation and dimerization. Cancer research 2013, 73(17):5580-5590. Milde-Langosch K, Karn T, Schmidt M, zu Eulenburg C, Oliveira-Ferrer L, Wirtz RM, Schumacher U, Witzel I, Schutze D, Muller V: Prognostic relevance of glycosylationassociated genes in breast cancer. Breast Cancer Res Treat 2014, 145(2):295-305. Yuan K, Listinsky CM, Singh RK, Listinsky JJ, Siegal GP: Cell surface associated alpha-L-fucose moieties modulate human breast cancer neoplastic progression. Pathology oncology research : POR 2008, 14(2):145-156. Carvalho AS, Harduin-Lepers A, Magalhaes A, Machado E, Mendes N, Costa LT, Matthiesen R, Almeida R, Costa J, Reis CA: Differential expression of alpha-2,3-sialyltransferases and alpha-1,3/4-fucosyltransferases regulates the levels of sialyl Lewis a and sialyl Lewis x in gastrointestinal carcinoma cells. The international journal of biochemistry & cell biology 2010, 42(1):80-89. Perez-Garay M, Arteta B, Llop E, Cobler L, Pages L, Ortiz R, Ferri MJ, de Bolos C, Figueras J, de Llorens R et al: alpha2,3-Sialyltransferase ST3Gal IV promotes migration and metastasis in pancreatic adenocarcinoma cells and tends to be highly expressed in pancreatic adenocarcinoma tissues. The international journal of biochemistry & cell biology 2013, 45(8):1748-1757. Yu X, Wu Q, Wang L, Zhao Y, Zhang Q, Meng Q, Pawan, Wang S: Silencing of ST6GalNAc I suppresses the proliferation, migration and invasion of hepatocarcinoma cells through PI3K/AKT/NF-kappaB pathway. Tumour Biol 2016, 37(9):12213-12221. Zhang X, Wang Y, Qian Y, Wu X, Zhang Z, Liu X, Zhao R, Zhou L, Ruan Y, Xu J et al: Discovery of specific metastasis-related N-glycan alterations in epithelial ovarian cancer based on quantitative glycomics. PLoS One 2014, 9(2):e87978. Lange T, Ullrich S, Muller I, Nentwich MF, Stubke K, Feldhaus S, Knies C, Hellwinkel OJ, Vessella RL, Abramjuk C et al: Human prostate cancer in a clinically relevant xenograft mouse model: identification of beta(1,6)-branched oligosaccharides as a marker of tumor progression. Clin Cancer Res 2012, 18(5):1364-1373. Liu X, Nie H, Zhang Y, Yao Y, Maitikabili A, Qu Y, Shi S, Chen C, Li Y: Cell surface-specific Nglycan profiling in breast cancer. PLoS One 2013, 8(8):e72704. Handerson T, Camp R, Harigopal M, Rimm D, Pawelek J: Beta1,6-branched oligosaccharides are increased in lymph node metastases and predict poor outcome in breast carcinoma. Clin Cancer Res 2005, 11(8):2969-2973. Guo H, Abbott KL: Functional impact of tumor-specific N-linked glycan changes in breast and ovarian cancers. Adv Cancer Res 2015, 126:281-303. Everest-Dass AV, Briggs MT, Kaur G, Oehler MK, Hoffmann P, Packer NH: N-glycan MALDI Imaging Mass Spectrometry on Formalin-Fixed Paraffin-Embedded Tissue Enables the Delineation of Ovarian Cancer Tissues. Mol Cell Proteomics 2016, 15(9):3003-3016. Anugraham M, Jacob F, Nixdorf S, Everest-Dass AV, Heinzelmann-Schwarz V, Packer NH: Specific glycosylation of membrane proteins in epithelial ovarian cancer cell lines: glycan structures reflect gene expression and DNA methylation status. Mol Cell Proteomics 2014, 13(9):2213-2232. Vajaria BN, Patel KR, Begum R, Patel PS: Sialylation: an Avenue to Target Cancer Cells. Pathol Oncol Res 2016, 22(3):443-447. More SK, Srinivasan N, Budnar S, Bane SM, Upadhya A, Thorat RA, Ingle AD, Chiplunkar SV, Kalraiya RD: N-glycans and metastasis in galectin-3 transgenic mice. Biochem Biophys Res Commun 2015, 460(2):302-307.

17

50.

51.

52.

53.

54.

55. 56.

57.

58.

59.

60.

61.

62.

63. 64.

65. 66.

67.

Krishnan V, Bane SM, Kawle PD, Naresh KN, Kalraiya RD: Altered melanoma cell surface glycosylation mediates organ specific adhesion and metastasis via lectin receptors on the lung vascular endothelium. Clin Exp Metastasis 2005, 22(1):11-24. Sengupta PK, Bouchie MP, Nita-Lazar M, Yang HY, Kukuruzinska MA: Coordinate regulation of N-glycosylation gene DPAGT1, canonical Wnt signaling and E-cadherin adhesion. J Cell Sci 2013, 126(Pt 2):484-496. Pinho SS, Figueiredo J, Cabral J, Carvalho S, Dourado J, Magalhaes A, Gartner F, Mendonfa AM, Isaji T, Gu J et al: E-cadherin and adherens-junctions stability in gastric carcinoma: functional implications of glycosyltransferases involving N-glycan branching biosynthesis, N-acetylglucosaminyltransferases III and V. Biochim Biophys Acta 2013, 1830(3):2690-2700. Carvalho S, Catarino TA, Dias AM, Kato M, Almeida A, Hessling B, Figueiredo J, Gartner F, Sanches JM, Ruppert T et al: Preventing E-cadherin aberrant N-glycosylation at Asn-554 improves its critical function in gastric cancer. Oncogene 2016, 35(13):1619-1631. Guo HB, Johnson H, Randolph M, Pierce M: Regulation of homotypic cell-cell adhesion by branched N-glycosylation of N-cadherin extracellular EC2 and EC3 domains. J Biol Chem 2009, 284(50):34986-34997. Kariya Y, Gu J: N-glycosylation of ss4 integrin controls the adhesion and motility of keratinocytes. PLoS One 2011, 6(11):e27084. Guo HB, Lee I, Kamar M, Akiyama SK, Pierce M: Aberrant N-glycosylation of beta1 integrin causes reduced alpha5beta1 integrin clustering and stimulates cell migration. Cancer research 2002, 62(23):6837-6845. Seales EC, Shaikh FM, Woodard-Grice AV, Aggarwal P, McBrayer AC, Hennessy KM, Bellis SL: A protein kinase C/Ras/ERK signaling pathway activates myeloid fibronectin receptors by altering beta1 integrin sialylation. J Biol Chem 2005, 280(45):37610-37615. Uemura T, Shiozaki K, Yamaguchi K, Miyazaki S, Satomi S, Kato K, Sakuraba H, Miyagi T: Contribution of sialidase NEU1 to suppression of metastasis of human colon cancer cells through desialylation of integrin beta4. Oncogene 2009, 28(9):1218-1229. Kariya Y, Kato R, Itoh S, Fukuda T, Shibukawa Y, Sanzen N, Sekiguchi K, Wada Y, Kawasaki N, Gu J: N-Glycosylation of laminin-332 regulates its biological functions. A novel function of the bisecting GlcNAc. J Biol Chem 2008, 283(48):33036-33045. Faller CE, Guvench O: Terminal sialic acids on CD44 N-glycans can block hyaluronan binding by forming competing intramolecular contacts with arginine sidechains. Proteins 2014, 82(11):3079-3089. Boon L, Ugarte-Berzal E, Vandooren J, Opdenakker G: Glycosylation of matrix metalloproteases and tissue inhibitors: present state, challenges and opportunities. Biochem J 2016, 473(11):1471-1482. Guo S, Skala W, Magdolen V, Briza P, Biniossek ML, Schilling O, Kellermann J, Brandstetter H, Goettig P: A Single Glycan at the 99-Loop of Human Kallikrein-related Peptidase 2 Regulates Activation and Enzymatic Activity. J Biol Chem 2016, 291(2):593-604. Niwa Y, Suzuki T, Dohmae N, Umezawa K, Simizu S: Determination of cathepsin V activity and intracellular trafficking by N-glycosylation. FEBS Lett 2012, 586(20):3601-3607. Stolfa G, Mondal N, Zhu Y, Yu X, Buffone A, Jr., Neelamegham S: Using CRISPR-Cas9 to quantify the contributions of O-glycans, N-glycans and Glycosphingolipids to human leukocyte-endothelium adhesion. Sci Rep 2016, 6:30392. Wilson CM, High S: Ribophorin I acts as a substrate-specific facilitator of N-glycosylation. J Cell Sci 2007, 120(Pt 4):648-657. Ono M, Tsuda H, Kobayashi T, Takeshita F, Takahashi RU, Tamura K, Akashi-Tanaka S, Moriya T, Yamasaki T, Kinoshita T et al: The expression and clinical significance of ribophorin II (RPN2) in human breast cancer. Pathol Int 2015, 65(6):301-308. Fujiwara T, Takahashi RU, Kosaka N, Nezu Y, Kawai A, Ozaki T, Ochiya T: RPN2 Gene Confers Osteosarcoma Cell Malignant Phenotypes and Determines Clinical Prognosis. Mol Ther Nucleic Acids 2014, 3:e189. 18

68.

69.

70.

71.

72. 73.

74.

75.

76.

77.

78.

79.

80.

81. 82.

83. 84. 85.

Zhang J, Yan B, Spath SS, Qun H, Cornelius S, Guan D, Shao J, Hagiwara K, Van Waes C, Chen Z et al: Integrated transcriptional profiling and genomic analyses reveal RPN2 and HMGB1 as promising biomarkers in colorectal cancer. Cell Biosci 2015, 5:53. Guervos MA, Marcos CA, Hermsen M, Nuno AS, Suarez C, Llorente JL: Deletions of N33, STK11 and TP53 are involved in the development of lymph node metastasis in larynx and pharynx carcinomas. Cell Oncol 2007, 29(4):327-334. Pils D, Horak P, Vanhara P, Anees M, Petz M, Alfanz A, Gugerell A, Wittinger M, Gleiss A, Auner V et al: Methylation status of TUSC3 is a prognostic factor in ovarian cancer. Cancer 2013, 119(5):946-954. Vanhara P, Horak P, Pils D, Anees M, Petz M, Gregor W, Zeillinger R, Krainer M: Loss of the oligosaccharyl transferase subunit TUSC3 promotes proliferation and migration of ovarian cancer cells. Int J Oncol 2013, 42(4):1383-1389. Drake RR: Glycosylation and cancer: moving glycomics to the forefront. Adv Cancer Res 2015, 126:1-10. Liu H, Wu Q, Liu Y, Liu W, Zhang W, Pan D, Xu J: Prognostic significance of beta1,6-Nacetylglucosaminyltransferase V expression in patients with hepatocellular carcinoma. Jpn J Clin Oncol 2015, 45(9):844-853. Liu Y, Liu H, Liu W, Zhang W, An H, Xu J: beta1,6-N-acetylglucosaminyltransferase V predicts recurrence and survival of patients with clear-cell renal cell carcinoma after surgical resection. World J Urol 2015, 33(11):1791-1799. Park JJ, Lee M: Increasing the alpha 2, 6 sialylation of glycoproteins may contribute to metastatic spread and therapeutic resistance in colorectal cancer. Gut Liver 2013, 7(6):629641. Suzuki O, Abe M, Hashimoto Y: Sialylation by beta-galactoside alpha-2,6-sialyltransferase and N-glycans regulate cell adhesion and invasion in human anaplastic large cell lymphoma. Int J Oncol 2015, 46(3):973-980. Chen CY, Jan YH, Juan YH, Yang CJ, Huang MS, Yu CJ, Yang PC, Hsiao M, Hsu TL, Wong CH: Fucosyltransferase 8 as a functional regulator of nonsmall cell lung cancer. Proc Natl Acad Sci U S A 2013, 110(2):630-635. Yue L, Han C, Li Z, Li X, Liu D, Liu S, Yu H: Fucosyltransferase 8 expression in breast cancer patients: A high throughput tissue microarray analysis. Histol Histopathol 2016, 31(5):547555. Osumi D, Takahashi M, Miyoshi E, Yokoe S, Lee SH, Noda K, Nakamori S, Gu J, Ikeda Y, Kuroki Y et al: Core fucosylation of E-cadherin enhances cell-cell adhesion in human colon carcinoma WiDr cells. Cancer Sci 2009, 100(5):888-895. Shao K, Chen ZY, Gautam S, Deng NH, Zhou Y, Wu XZ: Posttranslational modification of Ecadherin by core fucosylation regulates Src activation and induces epithelial-mesenchymal transition-like process in lung cancer cells. Glycobiology 2016, 26(2):142-154. Zhao YP, Xu XY, Fang M, Wang H, You Q, Yi CH, Ji J, Gu X, Zhou PT, Cheng C et al: Decreased core-fucosylation contributes to malignancy in gastric cancer. PLoS One 2014, 9(4):e94536. Mizumoto S, Ikegawa S, Sugahara K: Human genetic disorders caused by mutations in genes encoding biosynthetic enzymes for sulfated glycosaminoglycans. J Biol Chem 2013, 288(16):10953-10961. Kolset SO, Tveit H: Serglycin--structure and biology. Cell Mol Life Sci 2008, 65(7-8):10731085. Choi Y, Chung H, Jung H, Couchman JR, Oh ES: Syndecans as cell surface receptors: Unique structure equates with functional diversity. Matrix Biol 2011, 30(2):93-99. Oliveira-Ferrer L, Hessling A, Trillsch F, Mahner S, Milde-Langosch K: Prognostic impact of chondroitin-4-sulfotransferase CHST11 in ovarian cancer. Tumour Biol 2015, 36(11):90239030.

19

86.

87.

88.

89.

90.

91.

92.

93.

94. 95.

96.

97.

98.

99.

100.

101. 102.

Theocharis AD, Tsolakis I, Tzanakakis GN, Karamanos NK: Chondroitin sulfate as a key molecule in the development of atherosclerosis and cancer progression. Adv Pharmacol 2006, 53:281-295. Clausen TM, Pereira MA, Al Nakouzi N, Oo HZ, Agerbaek MO, Lee S, Orum-Madsen MS, Kristensen AR, El-Naggar A, Grandgenett PM et al: Oncofetal Chondroitin Sulfate Glycosaminoglycans Are Key Players in Integrin Signaling and Tumor Cell Motility. Mol Cancer Res 2016, 14(12):1288-1299. Arslan F, Bosserhoff AK, Nickl-Jockschat T, Doerfelt A, Bogdahn U, Hau P: The role of versican isoforms V0/V1 in glioma migration mediated by transforming growth factor-beta2. Br J Cancer 2007, 96(10):1560-1568. Varga I, Hutoczki G, Szemcsak CD, Zahuczky G, Toth J, Adamecz Z, Kenyeres A, Bognar L, Hanzely Z, Klekner A: Brevican, neurocan, tenascin-C and versican are mainly responsible for the invasiveness of low-grade astrocytoma. Pathol Oncol Res 2012, 18(2):413-420. Viapiano MS, Bi WL, Piepmeier J, Hockfield S, Matthews RT: Novel tumor-specific isoforms of BEHAB/brevican identified in human malignant gliomas. Cancer research 2005, 65(15):6726-6733. Iida J, Pei D, Kang T, Simpson MA, Herlyn M, Furcht LT, McCarthy JB: Melanoma chondroitin sulfate proteoglycan regulates matrix metalloproteinase-dependent human melanoma invasion into type I collagen. J Biol Chem 2001, 276(22):18786-18794. Iida J, Wilhelmson KL, Ng J, Lee P, Morrison C, Tam E, Overall CM, McCarthy JB: Cell surface chondroitin sulfate glycosaminoglycan in melanoma: role in the activation of pro-MMP-2 (pro-gelatinase A). Biochem J 2007, 403(3):553-563. Silver DJ, Siebzehnrubl FA, Schildts MJ, Yachnis AT, Smith GM, Smith AA, Scheffler B, Reynolds BA, Silver J, Steindler DA: Chondroitin sulfate proteoglycans potently inhibit invasion and serve as a central organizer of the brain tumor microenvironment. J Neurosci 2013, 33(39):15603-15617. Stock C, Jungmann O, Seidler DG: Decorin and chondroitin-6 sulfate inhibit B16V melanoma cell migration and invasion by cellular acidification. J Cell Physiol 2011, 226(10):2641-2650. Sugahara KN, Hirata T, Tanaka T, Ogino S, Takeda M, Terasawa H, Shimada I, Tamura J, ten Dam GB, van Kuppevelt TH et al: Chondroitin sulfate E fragments enhance CD44 cleavage and CD44-dependent motility in tumor cells. Cancer research 2008, 68(17):7191-7199. Monzavi-Karbassi B, Stanley JS, Hennings L, Jousheghany F, Artaud C, Shaaf S, KieberEmmons T: Chondroitin sulfate glycosaminoglycans as major P-selectin ligands on metastatic breast cancer cell lines. Int J Cancer 2007, 120(6):1179-1191. Kawashima H, Hirose M, Hirose J, Nagakubo D, Plaas AH, Miyasaka M: Binding of a large chondroitin sulfate/dermatan sulfate proteoglycan, versican, to L-selectin, P-selectin, and CD44. J Biol Chem 2000, 275(45):35448-35456. Fthenou E, Zong F, Zafiropoulos A, Dobra K, Hjerpe A, Tzanakakis GN: Chondroitin sulfate A regulates fibrosarcoma cell adhesion, motility and migration through JNK and tyrosine kinase signaling pathways. In Vivo 2009, 23(1):69-76. Nikitovic D, Assouti M, Sifaki M, Katonis P, Krasagakis K, Karamanos NK, Tzanakakis GN: Chondroitin sulfate and heparan sulfate-containing proteoglycans are both partners and targets of basic fibroblast growth factor-mediated proliferation in human metastatic melanoma cell lines. Int J Biochem Cell Biol 2008, 40(1):72-83. Kure S, Yoshie O, Aso H: Metastatic potential of murine B16 melanoma correlates with reduced surface heparan sulfate glycosaminoglycan. Jpn J Cancer Res 1987, 78(11):12381245. Langford JK, Stanley MJ, Cao D, Sanderson RD: Multiple heparan sulfate chains are required for optimal syndecan-1 function. J Biol Chem 1998, 273(45):29965-29971. Gao W, Kim H, Ho M: Human Monoclonal Antibody Targeting the Heparan Sulfate Chains of Glypican-3 Inhibits HGF-Mediated Migration and Motility of Hepatocellular Carcinoma Cells. PLoS One 2015, 10(9):e0137664. 20

103. 104. 105. 106.

107.

108.

109. 110. 111.

112.

113.

114.

115.

116.

117.

118. 119.

120.

121.

Heldin P, Delatorre M, Ytterberg D, Bergh J: Differential synthesis and binding of hyaluronan by human breast cancer cell lines. Oncol Rep 1996, 3(6):1011-1016. Cheng XB, Kohi S, Koga A, Hirata K, Sato N: Hyaluronan stimulates pancreatic cancer cell motility. Oncotarget 2016, 7(4):4829-4840. Wu M, Cao M, He Y, Liu Y, Yang C, Du Y, Wang W, Gao F: A novel role of low molecular weight hyaluronan in breast cancer metastasis. FASEB J 2015, 29(4):1290-1298. Kato M, Saunders S, Nguyen H, Bernfield M: Loss of cell surface syndecan-1 causes epithelia to transform into anchorage-independent mesenchyme-like cells. Mol Biol Cell 1995, 6(5):559-576. Chalkiadaki G, Nikitovic D, Berdiaki A, Sifaki M, Krasagakis K, Katonis P, Karamanos NK, Tzanakakis GN: Fibroblast growth factor-2 modulates melanoma adhesion and migration through a syndecan-4-dependent mechanism. Int J Biochem Cell Biol 2009, 41(6):1323-1331. O'Connell MP, Fiori JL, Kershner EK, Frank BP, Indig FE, Taub DD, Hoek KS, Weeraratna AT: Heparan sulfate proteoglycan modulation of Wnt5A signal transduction in metastatic melanoma cells. J Biol Chem 2009, 284(42):28704-28712. Xiao WK, Qi CY, Chen D, Li SQ, Fu SJ, Peng BG, Liang LJ: Prognostic significance of glypican-3 in hepatocellular carcinoma: a meta-analysis. BMC Cancer 2014, 14:104. Yang X, Zhang Z, Qiu M, Hu J, Fan X, Wang J, Xu L, Yin R: Glypican-5 is a novel metastasis suppressor gene in non-small cell lung cancer. Cancer Lett 2013, 341(2):265-273. Castillo LF, Tascon R, Huvelle MA, Novack G, Llorens MC, Santos AF, Shortrede J, Cabanillas AM, Joffe EB, Labriola L et al: Glypican-3 induces a mesenchymal to epithelial transition in human breast cancer cells. Oncotarget 2016. Korpetinou A, Skandalis SS, Moustakas A, Happonen KE, Tveit H, Prydz K, Labropoulou VT, Giannopoulou E, Kalofonos HP, Blom AM et al: Serglycin is implicated in the promotion of aggressive phenotype of breast cancer cells. PLoS One 2013, 8(10):e78157. Li XJ, Ong CK, Cao Y, Xiang YQ, Shao JY, Ooi A, Peng LX, Lu WH, Zhang Z, Petillo D et al: Serglycin is a theranostic target in nasopharyngeal carcinoma that promotes metastasis. Cancer research 2011, 71(8):3162-3172. Iozzo RV, Cohen IR, Grassel S, Murdoch AD: The biology of perlecan: the multifaceted heparan sulphate proteoglycan of basement membranes and pericellular matrices. Biochem J 1994, 302 ( Pt 3):625-639. Adatia R, Albini A, Carlone S, Giunciuglio D, Benelli R, Santi L, Noonan DM: Suppression of invasive behavior of melanoma cells by stable expression of anti-sense perlecan cDNA. Ann Oncol 1997, 8(12):1257-1261. Sharma B, Handler M, Eichstetter I, Whitelock JM, Nugent MA, Iozzo RV: Antisense targeting of perlecan blocks tumor growth and angiogenesis in vivo. J Clin Invest 1998, 102(8):15991608. Maishi N, Ohba Y, Akiyama K, Ohga N, Hamada J, Nagao-Kitamoto H, Alam MT, Yamamoto K, Kawamoto T, Inoue N et al: Tumour endothelial cells in high metastatic tumours promote metastasis via epigenetic dysregulation of biglycan. Sci Rep 2016, 6:28039. Hu L, Duan YT, Li JF, Su LP, Yan M, Zhu ZG, Liu BY, Yang QM: Biglycan enhances gastric cancer invasion by activating FAK signaling pathway. Oncotarget 2014, 5(7):1885-1896. Desjardins M, Xie J, Gurler H, Muralidhar GG, Sacks JD, Burdette JE, Barbolina MV: Versican regulates metastasis of epithelial ovarian carcinoma cells and spheroids. J Ovarian Res 2014, 7:70. Cooney CA, Jousheghany F, Yao-Borengasser A, Phanavanh B, Gomes T, Kieber-Emmons AM, Siegel ER, Suva LJ, Ferrone S, Kieber-Emmons T et al: Chondroitin sulfates play a major role in breast cancer metastasis: a role for CSPG4 and CHST11 gene expression in forming surface P-selectin ligands in aggressive breast cancer cells. Breast Cancer Res 2011, 13(3):R58. Jung SH, Lee HC, Yu DM, Kim BC, Park SM, Lee YS, Park HJ, Ko YG, Lee JS: Heparan sulfation is essential for the prevention of cellular senescence. Cell Death Differ 2016, 23(3):417-429. 21

122.

123. 124.

125.

126.

127. 128. 129.

130.

131.

132.

133. 134. 135.

Hulett MD, Freeman C, Hamdorf BJ, Baker RT, Harris MJ, Parish CR: Cloning of mammalian heparanase, an important enzyme in tumor invasion and metastasis. Nat Med 1999, 5(7):803-809. Vlodavsky I, Friedmann Y: Molecular properties and involvement of heparanase in cancer metastasis and angiogenesis. J Clin Invest 2001, 108(3):341-347. Reiland J, Sanderson RD, Waguespack M, Barker SA, Long R, Carson DD, Marchetti D: Heparanase degrades syndecan-1 and perlecan heparan sulfate: functional implications for tumor cell invasion. J Biol Chem 2004, 279(9):8047-8055. Goldshmidt O, Zcharia E, Cohen M, Aingorn H, Cohen I, Nadav L, Katz BZ, Geiger B, Vlodavsky I: Heparanase mediates cell adhesion independent of its enzymatic activity. FASEB J 2003, 17(9):1015-1025. Levy-Adam F, Feld S, Suss-Toby E, Vlodavsky I, Ilan N: Heparanase facilitates cell adhesion and spreading by clustering of cell surface heparan sulfate proteoglycans. PLoS One 2008, 3(6):e2319. Zetser A, Bashenko Y, Miao HQ, Vlodavsky I, Ilan N: Heparanase affects adhesive and tumorigenic potential of human glioma cells. Cancer research 2003, 63(22):7733-7741. Levy-Adam F, Ilan N, Vlodavsky I: Tumorigenic and adhesive properties of heparanase. Semin Cancer Biol 2010, 20(3):153-160. Bernert B, Porsch H, Heldin P: Hyaluronan synthase 2 (HAS2) promotes breast cancer cell invasion by suppression of tissue metalloproteinase inhibitor 1 (TIMP-1). J Biol Chem 2011, 286(49):42349-42359. Bullard KM, Kim HR, Wheeler MA, Wilson CM, Neudauer CL, Simpson MA, McCarthy JB: Hyaluronan synthase-3 is upregulated in metastatic colon carcinoma cells and manipulation of expression alters matrix retention and cellular growth. Int J Cancer 2003, 107(5):739-746. Simpson MA, Reiland J, Burger SR, Furcht LT, Spicer AP, Oegema TR, Jr., McCarthy JB: Hyaluronan synthase elevation in metastatic prostate carcinoma cells correlates with hyaluronan surface retention, a prerequisite for rapid adhesion to bone marrow endothelial cells. J Biol Chem 2001, 276(21):17949-17957. Bharadwaj AG, Kovar JL, Loughman E, Elowsky C, Oakley GG, Simpson MA: Spontaneous metastasis of prostate cancer is promoted by excess hyaluronan synthesis and processing. Am J Pathol 2009, 174(3):1027-1036. Lokeshwar VB, Cerwinka WH, Isoyama T, Lokeshwar BL: HYAL1 hyaluronidase in prostate cancer: a tumor promoter and suppressor. Cancer research 2005, 65(17):7782-7789. Lokeshwar VB, Cerwinka WH, Lokeshwar BL: HYAL1 hyaluronidase: a molecular determinant of bladder tumor growth and invasion. Cancer research 2005, 65(6):2243-2250. Adamczyk B, Tharmalingam T, Rudd PM: Glycans as cancer biomarkers. Biochim Biophys Acta 2012, 1820(9):1347-1353.

22

Figure legends: Figure 1: Biosynthesis of mucin-type O-glycans. (A) A single α-linked GalNAc is transferred to Ser/Thr via the activity of ppGalNAcTs (GALNTs) forming the Tn antigen. The elongation of the Tn antigen resulted in sialyl-Tn antigen through ST6GalNAc-I or T antigen (core 1) due to T-synthase/COSMC. The T antigen can further branched in sLea/ sLex, core 2 and sialyl-T antigen through the action of certain glycosyltransferases or sialyltransferases. Core 3 is initiated via the activity of C2GnT, which in turns form core 4 and further extended structures. (B) Distribution of O-glycans in the Golgi. The synthesis of core 1 and core 2 starts in the cis-Golgi and the extension of core structures are completed in medial-Golgi and trans-Golgi. The elongation of further complex core structures (core 3, core 4 and sialyl-T antigen etc.) occurs mostly in the trans-Golgi. (ER=Endoplasmatic reticulum, TGN=Trans Golgi Network) Figure 2: Principles and steps of N-glycosylation. 1. Synthesis of an oligosaccharide precursor consisting of 14 monomers bound to a dolichol anchor in the ER. Enzymes involved are DPAG1, ALG1-14 etc. 2. Transfer of the precursor to a suitable ASN residue within the nascent polypeptide by the oligosaccharyltransferase (OST) protein complex which consists of various subunits, among them STT3A and STT3B, which form the catalytic center, ribophorin I and II (RPN1, RPN2), OST48, N33, IAP and DAD. 3. Trimming of the glycan precursor in the ER and exit to the Golgi: cleavage of 3 terminal glucose residues by DCS1 and GANAB, and one mannose by ER mannosidase resulting in high-mannose glycans. 4. Cleavage of part of the mannose residues by Golgi mannosidases, followed by branching and chain extension by different monosaccharides, i.e. GlcNAc (MGAT family), Gal (B3GALT1 etc.), Fucose (FUT’s), sialic acid (ST6GAL or ST3GAL). Figure 3: Glycosaminoglycan structure and proteoglycans. (A) Glycosaminoglycan chains consist of repeating disaccharide units. All GAGs excepting HA contain can be sulfated at diverse positions. (B) Depending on their localization, proteoglycans can be grouped into three categories: extracellular, membrane-bound and intracellular proteoglycans. Serglycin, the only proteoglycan present in the cytoplasm, is packaged in storage granules, but can be also secreted and acts pro-metastatic. Diverse membraneassociated proteoglycans are known, including syndecan, glypican, betaglycan, phosphacan and NG2, mainly carrying CS and HS chains. A large diversity of proteoglycans and GAGs are present within the ECM and basement membrane. Among them, perlecan, versican and hyaluronan has been described to impact cancer metastasis. Figure 4: Role of glycan structures and glycosylated proteins in the metastatic cascade. Numerous glycan structures, including O-glycans (red), N-glycans (blue) and GAGs (green), and glycoproteins carrying one or more than one glycan type influence crucial steps of the metastatic process such as EMT, cell motility, invasion, intravasation and extravasation.

23

Fig.1

Fig.2

24

Fig.3

Fig.4

25