Metastasis is a multi-step, inefficient process that accounts for approximately 90% of cancer-related deaths. The process of cancer metastasis can be briefly summarized as cancer cells escaping their initial sites, surviving in blood and lymph transfer, and developing new distal tumor sites. Sialylation modifies the conformation of essential proteins to promote cancer cell proliferation, invasion, and migration. α2-6-Sialylation of epidermal growth factor receptor (EGFR) regulates the epithelial-to-mesenchymal transition (EMT) of cancer cells and sustains its membrane retention, regulating integrin tension, focal adhesion, and cell motility.
During the formation of pulmonary metastasis, interactions between P/L-selectin and blood components permit and initiate the formation of a metastatic milieu, while E-selectin is responsible for the local stimulation of lung microenvironment endothelial cells. Reduction of P/L-selectin, not E-selectin, significantly reduces lung metastasis
[1]. However, E-selectin upregulates FAK in lung vasculature, hence facilitating the homing of cancer cells to lung
[2]. Selectins also engage in endothelial activation caused by complicated interactions between tumor cells, platelets, and leukocytes, and then upregulate the expression of C-C chemokine ligand 5 (CCL5), which promotes the survival of tumor cells and leads to local lung metastasis
[3]. Through these cancer cell stimulations, CCL5 derived from mesenchymal stem cells further boosts cancer cells’ motility, invasion, and metastasis
[4]. Additionally, E-selectin interacts with its high-affinity ligands, sialyl Lewis X (sLeX) and sialyl Lewis A (sLeA), increasing cancer cell adherence and lung metastasis
[5]. As ligands of selectins, sialic acids contribute to the negative charges of tumor cell surfaces. Cancer cell motility can be reduced by neutralizing the negative charges caused by sialoglycans overexpressed on cancer cells while keeping the glycan moieties
[6]. The sialic acid blockade has been reported to reduce lung metastasis in a murine melanoma model
[7]. These studies indicate that interrupting the process of sialylation shows the potential to reduce lung metastasis and improve the outcomes of cancer patients.
Tumor-associated mucins typically have shortened core structures and sialylated epitopes
[8]. Sialyl Tn (STn) on MUC5AC synthesized by ST6GalNAc1, which is upregulated by mutant p53
R175H, promotes the liver metastasis of lung cancer
[9]. Sialylated structures of MUC16 bind to E- and L-selectin to promote the metastatic ability of pancreatic cancer cells
[10]. The interaction of MUC1 and galectin-3 promotes cancer spread by altering tumor cell surface polarization and increasing the exposure of cell surface adhesion molecules
[11]. MUC1-positive gastric cancer cells with sialyl Tn (sTn) antigen exhibit more metastatic potential than MUC1-positive cells alone
[12]. MUC13 is upregulated in the liver metastasis of metastatic colon cancers, indicating its role in cancer migration and metastasis
[13].
Additionally, sialylation of integrin β4 enhances FAK and ERK1/2 pathway signaling to facilitate the liver metastasis of colon cancer cells
[14]. The activation of ERK by E-selectin enhances the extravasation and transendothelial migration of colon cancer cells by activating Src kinase and dissociating the VE-cadherin/beta-catenin complex
[15]. E-selectin enhances the adhesion of cancer cells to the sinusoidal endothelial cells, promoting metastasis to the liver
[16]. Thus, liver metastasis of cancer cells is mucin- or E-selectin-dependent. Understanding the mechanisms of liver metastasis may pave the way for overcoming the relevant cancer therapeutic obstacles.
Bone vascular niche E-selectin binding with disseminated tumor cells triggers both their mesenchymal-epithelial transition (MET) and cancer stem cell traits
[17]. E-selectin interacts with sLeX and sLeA involved in EMT of colon cancer cells
[18]. Breast cancer cells enter bone sinusoidal niches via E-selectin contacts and are anchored by stromal cell-derived factor 1/C-X-C chemokine receptor type 4 (SDF1/CXCR4) interactions to develop dormant micrometastases
[19]. The sLeX of estrogen receptor alpha-positive breast cancer contributes to its bone metastasis
[20]. Breast cancer stem cells also exploit sialic acid interactions to induce immune tolerance and distant metastasis
[21]. In bone metastatic prostate cancer cells, α2,3-sialylation of α2 subunit of integrin α2β1 is upregulated and plays a crucial role in their initial adhesion capacity
[22]. Furthermore, the bone-homing behavior, adhesion, and migration of multiple myeloma cells are also impacted on α2,3-sialylation regulated by ST3Gal6
[23]. According to the roles of sialylation in MET, EMT, and adhesion capacity, sialylation is crucial for the initiation of bone metastasis. Additionally, medicinal approaches that interfere with sialylation may hold tremendous promise for preventing bone migration.
High sialylated
N-glycans facilitate breast cancer cells transit through the blood-brain barrier, hence promoting breast cancer brain metastasis
[24][25][26]. ST6GalNAc5 and ST6Gal1, which mediate α2,6-sialylation, are upregulated in the brain metastasis process of breast cancer
[26][27]. The expression of ST6GalNAc5 in breast cancer cells enhances their ability to adhere to brain endothelial cells and breach the blood-brain barrier
[27]. Similarly, α2,6-sialylated 4G8 (an IgG antibody drug) affects blood-brain barrier penetration via competitively inhibiting neonatal Fc-receptor-mediated transport
[28]. Therefore, sialylation of the cell surface plays a role in breaching the blood–brain barrier and subsequent colonization, and its disruption offers therapeutic opportunities.
Interestingly, sialylation has been shown in some labs to play distinct roles in cancer invasion and spread. Although sialyl-Tn is associated with a poor outcome for breast cancer
[29][30][31][32], ST6GalNAc2 was identified as a metastasis suppressor of breast cancer by in vivo RNA interference (RNAi) screen combined with next-generation sequencing
[33]. High ST6GanNAc2 in estrogen receptor (ER)-positive breast cancers reduces galectin-3 binding and metastasis by increasing the sialylation of core 1 antigen, whereas low ST6GanNAc2 in ER-negative breast cancers shows high endothelial cell adhesion and metastasis via galectin-3 binding
[34]. These suggest caution when using ST6GalNAc2 as a possible biomarker for predicting metastases in ER-negative breast cancers
[33]. Moreover, sialic acid-containing GM3 has been reported to reduce phosphoinositide-3 kinase/serine/threonine protein kinase B (PI3K/Akt) signaling to increase breast and colon cancer migration and invasion via inhibiting EGFR phosphorylation, upregulating phosphatase and tensin homolog (PTEN) expression, and interacting with integrins
[35]. On the contrary, downregulation of ST3Gal4 is associated with malignant progression
[36][37] in part by activating PI3K/Akt pathway in renal cell carcinoma
[36]. Similarly, certain ganglioside modifications (including the increase of GM3) mediated by ST6Gal1 and ST6GalNAc5 can inhibit glioma invasion
[38]. In addition, ST6Gal1 promotes the exosome-mediated exporting of the metastasis suppressor Kang-Ai 1 (KAI1, also known as CD82), thereby reducing KAI1-mediated suppression of integrin signaling in human metastatic colorectal cancer cells
[39]. In bladder cancer, ST8SIA1 is reported to decrease proliferation, invasion, and migration by inhibiting the phosphorylation of JAK2 and STAT3, thus downregulating their target genes’ transcription
[40].
These conflicting roles of STs in cancer invasion and metastasis may excite the interests of researchers and may pave the way for the uncovering of underlying mechanisms. Prevailing opinions hold that sialylation of cell membrane proteins/lipids increases tumor spread. However, it has been found that highly expressed STs hinder cancer invasion and migration via inhibiting the PI3K/Akt and JAK2/STAT3 pathways
[36][40]. The downstream signaling of PI3K/Akt is activated by the phosphorylation of two key conserved sites (Thr308 and Ser473) of Akt
[41][42]. Since O-GlcNAcylation, adds β-D-N-acetylglucosamine to protein serine or threonine residues, O-GlcNAcylations at or around Thr308 and Ser473 may compete with phosphorylation and exert distinct effects. The O-GlcNAcylations at Thr305 and Thr312 are reported to inhibit the phosphorylation at Thr308, by disrupting its interaction with phosphoinositide-dependent kinase 1 (PDK1)
[43]. The increased O-GlcNAcylation at Ser473 hinders its phosphorylation and enhances the apoptosis of murine pancreatic β cells
[44]. Moreover, O-GlcNAc is reported to be galactosylated and then sialylated
[45][46][47], which may be the mechanism behind the regulation of intracellular protein activities by some STs. How specific sialylation levels and glycoforms trigger signal transduction to switch on/off cancer spread remains a mystery. Overall, sialylation, as a type of post-transcriptional modification, is sure to affect the protein functions; yet its complex impacts on cancer invasion and metastasis are still obscure.