Sialylation is a process in which sialic acid (SA) epitopes bind to certain saccharides under the catalysis of specific sialyltransferases (SiaTs) (Li and Ding, 2019). This process is altered significantly in many cancer tissues compared to normal tissues and plays a crucial role in human health and disease (Varki, 2008). In recent years, a number of articles have reported a close association between sialylation and digestive cancers, including gastric cancer (GC), colorectal cancer (CRC), pancreatic ductal adenocarcinoma (PDAC), and hepatocellular carcinoma (HCC) (Sakuma et al., 2012, Gretschel et al., 2003, Itai et al., 1991, Perez-Garay et al., 2010, Chung et al., 2014, Ito et al., 2001, Hosono et al., 1998, Futamura et al., 2000, Marcos et al., 2011, Ohuchi et al., 1986, Itzkowitz et al., 1989, Itzkowitz et al., 1990, Kim et al., 2002). We found that the altered expression levels of sialylated glycoproteins and SiaTs are strongly correlated with the prognosis of these cancers, and they may serve as valuable targets for diagnosis and treatment. Therefore, it is imperative to understand the exact modification process of sialylated glycoproteins in digestive cancers.

SA is a family of 9-carbon carboxylated sugars that are typically found as terminal monosaccharides in glycoconjugates. This family includes approximately 40 derivatives (Varki, 1992), all of which are derived from a nine-carbon sugar, namely, neuraminic acid (Traving and Schauer, 1998). Neuraminic acid acts as the mother molecule of the SA family. From this molecule, three major derivatives are produced, and these three derivatives are prevalent in most sialylation processes. The three major deratives include N-acetylneuraminic acid (Neu5Ac), N-glycolylneuraminic acid (Neu5Gc), and 2-keto-3-deoxy-D-glycero-D-galacto-nononic acid (Kdn) (Schauer and Kamerling, 2018). Neu5Ac and Neu5Gc are found in mammals. However, due to the loss of the cytidine monophosphate-N-acetylneuraminic acid hydroxylase (CMAH) gene, which encodes a key enzyme for synthesizing Neu5Gc from Neu5Ac, only one type of SA, Neu5Ac, is found in humans (Varki et al., 2011, Varki, 2010). Therefore, the SA mentioned in the following text specifically refers to Neu5Ac.

The sialylation process mainly exerts its effects on organisms by modifying glycoproteins and glycolipids through the catalysis of SiaTs (Julien et al., 2012). Although abnormal sialylation of glycolipids has been reported in diseases, its role may not be as vital as the altered sialylation level of glycoproteins in digestive cancers. Therefore, this review focuses on the relationship between abnormal sialylation of glycoproteins and digestive cancers.

Glycoproteins can be formed by covalently attaching saccharides to a polypeptide backbone via N-linkage or O-linkage. The saccharide linked to the specific peptide sequence Asn-X-Ser/Thr in proteins via N-linkage is also known as N-glycan (X refers to any amino acid). Similarly, saccharides linked to Ser/Thr in polypeptide sequences via O-linkage are also known as O-glycans (Spiro, 2002, Moremen et al., 2012) (Fig. 1). N-glycans share a common pentasaccharide core region, which can be further diversified by terminal structures such as mannose (Man), N-acetylgalactosamine (GalNAc), galactose (Gal) and Neu5Ac into oligomannose or other complex types (Pinho and Reis, 2015, Gao et al., 2019) (Fig. 1b).

O-glycan is also an essential substrate for sialylation. The first monosaccharide of glycans attached to serine or threonine residues via O-linkage is commonly GalNAc. Therefore, the typical type of protein O-glycosylation is initiated by GalNAc. This form of protein O-glycosylation is also known as mucin-type O-glycosylation. Mucin (MUC) is a glycoprotein commonly modified by sialylation, and mucin-type O-glycosylation is one of the most abundant forms of protein glycosylation in animals (Bennett et al., 2012). On the basis of O-linked GalNAc, four cores of O-glycans are formed by the corresponding four glycosyltransferases (Pinho and Reis, 2015) (Fig. 1a). Mucin-type O-glycans are frequently found in secreted or membrane-associated glycoproteins (Clausen and Bennett, 1996). There are a variety of MUCs expressed on the gastrointestinal mucosa, including MUC1, which is expressed on the mucosal surface. MUC5AC and MUC6 are secreted by deep glands. MUC2 is secreted by intestinal epithelial cells (Teixeira et al., 2002). MUCs are usually carriers of sialylated structures. For example, superficial gastric mucosa expresses MUC1 and secretes MUC5AC, which are modified by type 1 Lewis antigens. Deep glands of the gastric mucosa secrete MUC6, which is modified by type 2 Lewis antigens (De Bolos et al., 1995). MUC1 and MUC2 have been identified as the major carriers of the sialyl-Tn (sTn) antigen, which is a truncated glycoprotein structure expressed on intestinal and gastric mucosa with a sialyl-GalNAc-β-R sequence (Pinho et al., 2007, Conze et al., 2010).

Lewis-type blood group antigens are lacto- or neolacto-series oligosaccharides that are commonly found as part of glycoproteins or glycolipids on the eukaryotic cell surface. Lewis blood group antigens can be classified into two types: type 1 Lewis antigens, including LewisA (LeA), LewisB (LeB), and sialyl-LewisA (sLeA); and type 2 Lewis antigens, including LewisX (LeX), LewisY (LeY), and sialyl-LewisX (sLeX) (Guberman et al., 2019). The distribution of Lewis blood group antigens exhibits a coregulated pattern with MUC in normal gastric mucosa. Specifically, MUC1, MUC2, and MUC5AC expressed by cells in the superficial gastric epithelium are carriers of type 1 Lewis antigens, including LeA, LeB, and sLeA, whereas cells in deep glands predominantly express MUC6, which is associated with type 2 Lewis antigens, including LeX and LeY (De Bolos et al., 1995, Mollicone et al., 1985). Among these Lewis antigens, only the LeA and LeX antigens can be sialylated to form sLeA (also known as CA19–9) and sLeX, respectively. Both sLeA and sLeX are crucially involved in the pathogenesis of inflammation and cancer in various digestive diseases (Kannagi, 2007, Renkonen et al., 1997, Koprowski et al., 1982).

Generally, the glycoproteins modified by SiaTs are complex and diverse. This diversity can be manifested in three aspects. (i) The monosaccharides that compose the polysaccharide are diverse. (ii) There are different types of linkages between monosaccharides, such as α2,6-linkage, β1,3-linkage and β1,4-linkage. (iii) The linkage between the glycan and its aglycone protein part is also different and can be classified into N-glycans and O-glycans. This diversity of glycoproteins provides an abundance of different types of substrates for SiaTs.

SiaTs are enzymes that catalyze the linkage between SAs and polysaccharides. They modify the polysaccharides with SAs and prevent the overextension of polysaccharides (Julien et al., 2012). Based on their catalytic activities toward different substrates, such as Gal and GalNAc, or different chemical bonds, including α2,3-sialylation, α2,6-sialylation and α2,8-sialylation, that they finally form, SiaTs can be categorized into four distinct families, each of which is further subdivided into various subfamilies (Rodriguez et al., 2021, Hugonnet et al., 2021, Harduin-Lepers et al., 2005) (Table 1).

ST3GAL is an abbreviation for β-galactoside α2,3-sialyltransferase. The ST3GAL family is a large family that has six subfamilies (from ST3GAL1 to ST3GAL6). All members of this enzyme family facilitate the attachment of Neu5Ac residues to the terminal Gal residues of N-glycans, O-glycans, or glycolipids through an α2,3-linkage [39,40] (Fig. 2).

The primary role of ST3GAL1 is to catalyze the attachment of Neu5Ac and Thomsen-Friedenreich (TF) antigens (Gal-β1,3-GalNAc), resulting in the production of sialylated TF antigens (Dalziel et al., 2001). ST3GAL1 also participates in the sialylation of LeA, leading to the generation of sLeA. sLeA is commonly known as CA19–9 and is a well-known carcinoma-associated carbohydrate antigen (Rodriguez et al., 2021). ST3GAL2 catalyzes the sialylation process of mucin-type glycoproteins and some gangliosides. However, until now, there has been no direct evidence of the link between ST3GAL2 and digestive diseases (Chandrasekaran et al., 2011). ST3GAL3 exhibits a preference for modifying the Lewis antigen type 1 chain (Gal-β1,3-GlcNAc), which encompasses two polysaccharides, namely, LeA and LeB. However, only the LeA antigen can be sialylated to form sLeA. In contrast, ST3GAL4 exhibits the highest activity toward Lewis antigen type 2 chain (Gal-β1,4-GlcNAc), which also encompasses two polysaccharides, namely, LeX and LeY. However, only the LeX antigen can be modified to generate sLeX. Interestingly, ST3GAL3 also slightly modifies sLeX, and ST3GAL4 also has moderate affinity toward sLeA (Kono et al., 1997, Carvalho et al., 2010). Similar to ST3GAL4, ST3GAL6 also catalyzes the sialylation of the Lewis antigen type 2 chain and generates sLeX (Okajima et al., 1999). ST3GAL5, also known as GM3 synthase, shows a specific catalyzing affinity toward glycolipids instead of glycoproteins. Although it does not modify glycoproteins, there is a report demonstrating that ST3GAL5 is associated with normal digestive system functions. Namely, it participates in the cholesterol uptake process of intestines (Inokuchi et al., 2022).

Generally, the catalytic activity of SiaTs in the ST3GAL family exhibits significant diversity, as a single SiaT is capable of catalyzing the synthesis of various sialylated glycans. For example, ST3GAL1 could catalyze not only the formation of sLeA but also the synthesis of sialyl-TF antigen (Sakuma et al., 2012, Dalziel et al., 2001). In addition, the same sialylation process can also be catalyzed by different SiaTs, but different enzymes have different catalytic activities. For example, the synthesis of sLeX can be catalyzed by ST3GAL3, ST3GAL4, or ST3GAL6, but ST3GAL4 has the highest catalytic activity among them. The synthesis of sLeA can be catalyzed by ST3GAL3, ST3GAL4, or ST3GAL1, but ST3GAL3 has the highest catalytic activity among them (Sakuma et al., 2012, Kono et al., 1997, Okajima et al., 1999) (Table 1).

ST6GalNAc is an abbreviation for β-N-acetylgalactosaminide α2,6-sialyltransferase, which facilitates the attachment of Neu5Ac residues to GalNAc residues via an α2,6-linkage. The ST6GalNAc family has six subfamilies (from ST6GalNAc1 to ST6GalNAc6). ST6GalNAc1 and ST6GalNAc2 exhibit high activity toward O-linked glycoproteins that are widely distributed in many kinds of gastric tissues (Harduin-Lepers et al., 2005) (Fig. 2). ST6GalNAc1 is also known as the sTn synthase, which preferentially catalyzes the generation of the sTn antigen from the Tn antigen (GalNAc-β-peptides) (Marcos et al., 2004). This process can prevent polysaccharide chains from excessive extension (Julien et al., 2012). ST6GalNAc2 mainly catalyzes the sialylation of Gal-β1,3-GalNAc-β-peptides, also known as TF antigens, to form sialyl-TF antigens (Marcos et al., 2004). ST6GalNAc2 also modifies sialylated TF antigen to form disialylated TF antigen (Schneider et al., 2001). Compared with ST6GalNAc1 and ST6GalNAc2, the SiaTs ST6GalNAc3, ST6GalNAc4, ST6GalNAc5, and ST6GalNAc6 exhibit a more restricted substrate specificity, as they only utilize the sialylated structures Neu5Ac-α2,3-Gal-β1,3-GalNAc-R as their acceptor substrates, which is found either in glycoproteins or in glycolipids (Harduin-Lepers et al., 2005). In contrast to ST6GalNAc3, ST6GalNAc4 exhibits a preference for O-glycans over glycolipids and catalyzes the synthesis of disialyl-TF antigen from sialyl-TF antigen (Lee et al., 1999). Additionally, ST6GalNAc6 has been reported to be responsible for synthesizing the 2,3/2,6-disialyl-LewisA determinant that is distributed on glycolipids (Tsuchida et al., 2003) (Table 1).

In conclusion, ST6GalNAc1 and ST6GalNAc2 catalyze the sialylation of O-linked glycoproteins, while ST6GalNAc3 and ST6GalNAc4 catalyze the sialylation of both O-linked glycoproteins and glycolipids. ST6GalNAc5 and ST6GalNAc6 catalyze only the sialylation of glycolipids.

ST6GAL is an abbreviation for β-galactoside α2,6-sialyltransferase. The ST6GAL family has two subfamilies (ST6GAL1 and ST6GAL2), both of which are capable of catalyzing the sialylation of N-glycosylated proteins and show a specific affinity toward Gal-β1,4-GlcNAc-R structures (Harduin-Lepers et al., 2005) (Fig. 2). ST6GAL1 is an enzyme that is widely distributed in many tissues and plays a crucial role in the pathobiology of many cancers (Gc et al., 2022). Here, we summarize the ST6GAL1-affected pathways or targets that are associated with digestive cancers, including GC, CRC, PDAC, and HCC.

ST6GAL1 and α2,6-sialylation are responsible for the development of GC. The presence of α2,6-sialylated HER2 activates both the Akt and ERK pathways, leading not only to resistance against trastuzumab but also to the promotion of GC proliferation (Liu et al., 2018). Additionally, upregulation of ST6GAL1 and α2,6-sialylation level impedes TNF-induced apoptosis, thereby conferring immortality to GC cells (Alexander et al., 2020).

In the progression of CRC, ST6GAL1-mediated sialylation of integrin β1 facilitates activation of the signaling molecules paxillin and AKT (Lee et al., 2010). Furthermore, the sialylation function of ST6GAL1 hinders Fas death receptor-mediated apoptosis, thereby conferring CRC cells with resistance to programmed cell death (Swindall and Bellis, 2011, Swindall et al., 2013). Moreover, the sialylation of EGFR confers resistance to some antitumor agents in CRC cells (Park et al., 2012). These functions of ST6GAL1-mediated sialylation significantly increase the malignancy of CRC.

In the progression of PDAC, ST6GAL1 overexpression mediates α2,6-sialylation of EGFR, which subsequently activates the downstream molecules Wnt and β-catenin, which are responsible for the epithelial to mesenchymal transition (Britain et al., 2021). It also enhances the activity of hypoxia-inducible factor 1α (HIF-1α) and its signaling pathway, thereby conferring protection for tumor cells against hypoxic stress in the tumor microenvironment (Jones et al., 2018). Moreover, the sialylation function of ST6GAL1 hinders tumor necrosis factor receptor 1 (TNFR1)-mediated apoptosis. Consequently, it confers the immortality of PDAC cells (Holdbrooks et al., 2018).

In the progression of HCC, ST6GAL1-mediated sialylation of integrin β1 facilitates activation of the downstream signaling molecule FAK, which mediates cell motility and migration signaling in HCC (Han et al., 2018). Additionally, ST6GAL1 promotes the immune escape of HCC cells in the tumor microenvironment by increasing CD147 or MMP levels (Wang et al., 2019a).

In summary, ST6GAL1 is a pivotal molecule that triggers multiple signaling pathways across various tissues, and the digestive system is not an exception (Gc et al., 2022). We posit that comprehending the carcinogenicity of ST6GAL1 in the digestive system will facilitate the development of novel anticancer strategies.

ST6GAL2 is an enzyme with very limited distribution. The ST6GAL2 gene is mainly expressed in brain and fetal tissues, but its expression level is extremely low in testes, lungs, and other tissues (Choi et al., 2018). Compared with ST6GAL1, ST6GAL2 is far less impressive. Until now, there has been no direct evidence of the link between ST6Gal2 and digestive diseases (Takashima et al., 2002).

ST8SIA is an abbreviation for α2,8-sialyltransferase. Members of the ST8SIA family facilitate sialylation by transferring a single SA molecule to another in an α2,8-linkage, thereby forming linear chains of polysialic acids (Harduin-Lepers et al., 2005). There are six members in this family, from ST8SIA1 to ST8SIA6. ST8SIA transfers SA to a terminal α2,3-linked or α2,6-linked SA in α2,8-linkage (Fig. 2). This process generally occurs on N-glycans (Stanley et al., 2022). The function of this SiaT family is simple, namely, they just target SA. However, the intrinsic structure of this SiaT family is complicated, and it is responsible for its different affinities toward different receptors (Huang et al., 2017).

Generally, SiaTs show different affinities to different glycan structures, some of which show high affinity toward N-glycans and some toward O-glycans (Table 1). The sialylated glycoproteins and SiaTs compose an elegant network. Because the digestive tract is a natural channel that directly communicates with the outside environment, this network can be easily disrupted by many outside interferences, such as host-microbial interactions, viral infection or exposure to chemical substances. Many studies have reported that the expression levels of SAs, sialylated glycoproteins, and SiaTs are markedly changed in the development of many digestive diseases, especially digestive cancers (Sakuma et al., 2012, Gretschel et al., 2003, Itai et al., 1991, Perez-Garay et al., 2010, Chung et al., 2014, Ito et al., 2001, Hosono et al., 1998, Futamura et al., 2000, Marcos et al., 2011, Ohuchi et al., 1986, Itzkowitz et al., 1989, Itzkowitz et al., 1990, Kim et al., 2002). Although there are many different mechanisms underlying the occurrence of digestive cancers, there is no doubt that altered expression levels of sialylated glycoproteins are indispensable. These sialylated glycoproteins usually interact with their receptors to play oncogenic roles. For example, sialic acid-binding immunoglobulin-like lectin (Siglec) is a common receptor of SA that could regulate the immune system and act as an immune checkpoint. It thus plays an essential role in the progression of inflammation and the killing program of cancer cells, including digestive cancer cells (Duan and Paulson, 2020). These receptors also include E-selectin, which is a molecule that is commonly distributed in endothelial cells. Cancer cells overexpressing sialylated glycans, such as sLeA and sLeX, show a high affinity for endothelial cells via E-selectins and can easily enter blood circulation to achieve distant metastasis (Kannagi et al., 2004, Takada et al., 1993, Irimura et al., 1993).

Hence, it is worth determining how these sialylated glycoproteins and SiaTs change and what the changes are in digestive cancers. Here, from the perspective of altered sialylation patterns of glycoproteins, we summarize the mechanisms underlying the occurrence and development of gastric, colonic, pancreatic, and hepatic cancers.