Galectin-4 Is Involved in the Structural Changes of Glycosphingolipid Glycans in Poorly Differentiated Gastric Cancer Cells with High Metastatic Potential.

Gastric cancer with peritoneal dissemination is difficult to treat surgically, and frequently recurs and metastasizes. Currently, there is no effective treatment for this disease, and there is an urgent need to elucidate the molecular mechanisms underlying peritoneal dissemination and metastasis. Our previous study demonstrated that galectin-4 participates in the peritoneal dissemination of poorly differentiated gastric cancer cells. In this study, the glycan profiles of cell surface proteins and glycosphingolipids (GSLs) of the original (wild), galectin-4 knockout (KO), and rescue cells were investigated to understand the precise mechanisms involved in the galectin-4-mediated regulation of associated molecules, especially with respect to glycosylation. Glycan analysis of the NUGC4 wild type and galectin-4 KO clones with and without peritoneal metastasis revealed a marked structural change in the glycans of neutral GSLs, but not in N-glycan. Furthermore, mass spectrometry (MS) combined with glycosidase digestion revealed that this structural change was due to the presence of the lacto-type (ß1-3Galactosyl) glycan of GSL, in addition to the neolacto-type (ß1-4Galactosyl) glycan of GSL. Our results demonstrate that galectin-4 is an important regulator of glycosylation in cancer cells and galectin-4 expression affects the glycan profile of GSLs in malignant cancer cells with a high potential for peritoneal dissemination.

Suppression of galectin-4 attenuates peritoneal metastasis of poorly differentiated gastric cancer cells.

BACKGROUND: Peritoneal dissemination, most often seen in metastatic and/or recurrent gastric cancer, is an inoperable condition that lacks effective treatment. The use of molecular targeted drugs is also limited; therefore, identifying novel therapeutic targets and improving our understanding of this metastatic cancer are an urgent requirement. In this study, we focused on galectin-4, which is specifically expressed in poorly differentiated cells with high potential for peritoneal dissemination. METHODS: We knocked out the galectin-4 gene in NUGC4 cells using CRISPR/Cas9-mediated genome editing. Proliferation and peritoneal cancer formation in knockout cells were compared with those in wild-type and galectin-4 re-expressing cells. Western blotting and proximity ligation assays were performed to identify associated molecules affected by the expression of galectin-4. The effect of galectin-4 knockdown on cell proliferation and peritoneal metastasis was studied using a specific siRNA. Expression of galectin-4 in peritoneal metastatic tumors from 10 patients with gastric cancer was examined by immunohistochemistry. RESULTS: Suppression of galectin-4 expression reduced proliferation and peritoneal metastasis of malignant gastric cancer cells. Galectin-4 knockout and knockdown reduced the expression of activated c-MET and CD44. Galectin-4 was found to interact with several proteins on the cell surface, including CD44 and c-MET, via its carbohydrate-binding ability. Immunohistochemistry showed galectin-4 expression in peritoneal metastatic tumor cells in all patients examined. CONCLUSIONS: We clarified the role of galectin-4 in the development of peritoneal dissemination of poorly differentiated gastric cancer cells. Our data highlight the diagnostic and therapeutic potential of galectin-4 in the peritoneal dissemination of gastric cancer.

Preparation and biological activities of anti-HER2 monoclonal antibodies with multibranched complex-type N-glycans.

Immunoglobulin G (IgG) has a conserved N-glycosylation site at Asn297 in the fragment crystallizable (Fc) region. Previous studies have shown that N-glycosylation of this site is a critical mediator of the antibody's effector functions, such as antibody-dependent cellular cytotoxicity. While the N-glycan structures attached to the IgG-Fc region are generally heterogenous, IgGs engineered to be homogenously glycosylated with functional N-glycans may improve the efficacy of antibodies. The major glycoforms of the N-glycans on the IgG-Fc region are bi-antennary complex-type N-glycans, while multibranched complex-type N-glycans are not typically found. However, IgGs with tri-antennary complex-type N-glycans have been generated using the N-glycan remodeling technique, suggesting that more branched N-glycans might be artificially attached. At present, little is known about the properties of these IgGs. In this study, IgGs with multibranched N-glycans on the Fc region were prepared by using a combination of the glycosynthase/oxazoline substrate-based N-glycan remodeling technique and successive reactions with glycosyltransferases. Among the IgGs produced by these methods, the largest N-glycan attached was a bisecting N-acetylglucosamine containing a sialylated penta-antennary structure. Concerning the Fc-mediated effector functions, the majority of IgGs with tri- and tetra-antennary N-glycans on their Fc region showed properties similar to IgGs with ordinary bi-antennary N-glycans.

Novel endo-ß-N-acetylglucosaminidases from Tannerella species hydrolyze multibranched complex-type N-glycans with different specificities.

Endo-ß-N-acetylglucosaminidases are enzymes that hydrolyze the N,N'-diacetylchitobiose unit of N-glycans. Many endo-ß-N-acetylglucosaminidases also exhibit transglycosylation activity, which corresponds to the reverse of the hydrolysis reaction. Because of these activities, some of these enzymes have recently been used as powerful tools for glycan remodeling of glycoproteins. Although many endo-ß-N-acetylglucosaminidases have been identified and characterized to date, there are few enzymes that exhibit hydrolysis activity toward multibranched (tetra-antennary or more) complex-type N-glycans on glycoproteins. Therefore, we searched for novel endo-ß-N-acetylglucosaminidases that exhibit hydrolysis activity toward multibranched complex-type N-glycans in this study. From database searches, we selected three candidate enzymes from Tannerella species-Endo-Tsp1006, Endo-Tsp1263 and Endo-Tsp1457-and prepared them as recombinant proteins. We analyzed the hydrolysis activity of these enzymes toward N-glycans on glycoproteins and found that Endo-Tsp1006 and Endo-Tsp1263 exhibited hydrolysis activity toward complex-type N-glycans, including multibranched N-glycans, preferentially, whereas Endo-Tsp1457 exhibited hydrolysis activity toward high-mannose-type N-glycans exclusively. We further analyzed substrate specificities of Endo-Tsp1006 and Endo-Tsp1263 using 18 defined glycopeptides as substrates, each having a different N-glycan structure. We found that Endo-Tsp1006 preferred N-glycans with galactose or α2,6-linked sialic acid residues in their nonreducing ends as substrates, whereas Endo-Tsp1263 preferred N-glycans with N-acetylglucosamine residues in their nonreducing ends as substrates.

LacdiNAcylation of N-glycans in MDA-MB-231 human breast cancer cells results in changes in morphological appearance and adhesive properties of the cells.

We demonstrated previously that the expression of the disaccharide, GalNAcß1 → 4GlcNAc (LacdiNAc), on N-glycans of cell surface glycoproteins in MDA-MB-231 human breast cancer cells suppresses their malignant properties such as tumor formation in nude mice. Here, we report changes in the morphological appearance and adhesive properties of two kinds of clonal cells of MDA-MB-231 cells overexpressing ß4-N-acetyl-galactosaminyltransferase 4. The clonal cells exhibited a cobble stone-like shape as compared to a spindle-like shape of the mock-transfected cells and the original MDA-MB-231 cells. This was associated with an increased expression of cell surface E-cadherin, a marker of epithelial cells, and a decreased expression of N-cadherin, vimentin, α-smooth muscle actin and ZEB1, markers of mesenchymal cells. In addition, the clonal cells showed a lower migratory activity compared to the mock-transfected cells by wound-healing assay. These results suggest that mesenchymal-epithelial transition may be occurring in these clonal cells. Furthermore, increased adhesion to extracellular matrix proteins such as fibronectin, collagen type I, collagen type IV, and laminin was observed. The clonal cells spread and enlarged, whereas the mock-transfected cells demonstrated poor spreading on laminin-coated plates in the absence of fetal calf serum, indicating that expression of LacdiNAc on cell surface glycoproteins results in changes in cell adhesive and spreading properties particularly to laminin.

Discovery of Novel Indazole Derivatives as Orally Available ß3-Adrenergic Receptor Agonists Lacking Off-Target-Based Cardiovascular Side Effects.

We previously discovered that indazole derivative 8 was a highly selective ß3-adrenergic receptor (ß3-AR) agonist, but it appeared to be metabolically unstable. To improve metabolic stability, further optimization of this scaffold was carried out. We focused on the sulfonamide moiety of this scaffold, which resulted in the discovery of compound 15 as a highly potent ß3-AR agonist (EC50 = 18 nM) being inactive to ß1-, ß2-, and α1A-AR (ß1/ß3, ß2/ß3, and α1A/ß3 > 556-fold). Compound 15 showed dose-dependent ß3-AR-mediated responses in marmoset urinary bladder smooth muscle, had a desirable metabolic stability and pharmacokinetic profile (Cmax and AUC), and did not obviously affect heart rate or mean blood pressure when administered intravenously (3 mg/kg) to anesthetized rats. Thus, compound 15 is a highly potent, selective, and orally available ß3-AR agonist, which may serve as a candidate drug for the treatment of overactive bladder without off-target-based cardiovascular side effects.

Discovery of Novel Indazole Derivatives as Highly Potent and Selective Human ß3-Adrenergic Receptor Agonists with the Possibility of Having No Cardiovascular Side Effects.

Novel indazole derivatives were prepared and evaluated for their biological activity and cardiovascular safety profile as human ß3-adrenergic receptor (AR) agonists. Although the initial hit compound 5 exhibited significant ß3-AR agonistic activity (EC50 = 21 nM), it also exhibited agonistic activity at the α1A-AR (EC50 = 219 nM, selectivity: α1A/ß3 = 10-fold). The major metabolite of 5, which was an oxidative product at the indazole 3-methyl moiety, gave a clue to a strategy for improvement of the selectivity for ß3-AR agonistic activity versus α1A-AR agonistic activity. Thus, modification of the 3-substituent of the indazole moiety effectively improved the selectivity to develop compound 11 with potent ß3-AR agonistic activity (EC50 = 13 nM) and high selectivity (α1A/ß3 = >769-fold). Compound 11 was also inactive toward ß1 and ß2-ARs and showed dose dependent ß3-AR mediated relaxation of marmoset urinary bladder smooth muscle, while it did not obviously affect heart rate or blood pressure (iv, 3 mg/kg) in anesthetized rats.

Dlx5 and mef2 regulate a novel runx2 enhancer for osteoblast-specific expression.

Runx2 is essential for osteoblast differentiation and chondrocyte maturation. The expression of Runx2 is the first requisite step for the lineage determination from mesenchymal stem cells to osteoblasts. Although the transcript from Runx2 distal promoter is majorly expressed in osteoblasts, the promoter failed to direct green fluorescent protein (GFP) expression to osteoblasts. To find the regulatory region, we generated GFP reporter mice driven by a bacterial artificial chromosome (BAC) of Runx2 locus, and succeeded in the reproduction of endogenous Runx2 expression. By serially deleting it, we identified a 343-bp enhancer, which directed GFP expression specifically to osteoblasts, about 30 kb upstream of the distal promoter. The sequence of the 343-bp enhancer was highly conserved among mouse, human, dog, horse, opossum, and chicken. Dlx5, Mef2c, Tcf7, Ctnnb1, Sp7, Smad1, and Sox6, which localized on the enhancer region in primary osteoblasts, synergistically upregulated the enhancer activity, whereas Msx2 downregulated the activity in mouse osteoblastic MC3T3-E1 cells. Msx2 was predominantly bound to the enhancer in mouse multipotent mesenchymal C3H10T1/2 cells, whereas Dlx5 was predominantly bound to the enhancer in MC3T3-E1 cells. Dlx5 and Mef2 directly bound to the enhancer, and the binding sites were required for the osteoblast-specific expression in mice, whereas the other factors bound to the enhancer by protein-protein interaction. The enhancer was characterized by the presence of the histone variant H2A.Z, the enrichment of histone H3 mono- and dimethylated at Lys4 and acetylated at Lys18 and Lys27, but the depletion of histone H3 trimethylated at Lys4 in primary osteoblasts. These findings indicated that the enhancer, which had typical histone modifications for enhancers, contains sufficient elements to direct Runx2 expression to osteoblasts, and that Dlx5 and Mef2, which formed an enhanceosome with Tcf7, Ctnnb1, Sp7, Smad1, and Sox6, play an essential role in the osteoblast-specific activation of the enhancer. © 2014 American Society for Bone and Mineral Research.