Effect of glycosylation inhibitors on N-glycosylpeptides and on invasion of malignant mouse MO₄ cells in vitro

Cell surface glycans are believed to play a role in tumour invasion and metastasis. Yet, we have previously shown that the inhibitors of N-linked glycan processing swainsonine (SW) and 1-deoxynojirimycin (dNM) did not prevent invasion of chick heart fragments by MO₄ murine fibrosarcoma cells in organ culture. We now present biochemical evidence that these and other inhibitors of processing were indeed effective in remodeling glycans, including those expressed at the cell surface. After metabolic labeling with tritiated mannose or fucose, glycosylpeptides were obtained by Pronase treatment of material released from intact cells by trypsin. Glycosylpeptides were separated by Biogel P-10 chromatography. With all drugs tested, there was a shift towards lower molecular weight of the glycan chains. There were, however, major quantitative differences between the different drugs and also, for monensin (MON; 0.1 μg ml^-1), between fucose-labeled and mannoselabeled chains. The shift in apparent molecular weight affected mainly fucose-labeled peptides after treatment of MO₄ cells with SW (0.4 μg ml_-1). The shift induced by dNM (10mM)+SW (0.4 μg ml^-1) in both fucosylated and mannosylated chains was much larger than that induced by SW given alone. 1-Deoxymannojirimycin (dMM; 1 mM) had major effects on both mannose and fucose-labeled structures and so did N-methyl-1-deoxynojirimycin (MdNM; 2 mM) and castanospermine (CS; 100 μg ml^-1). With the latter drugs, incorporation of fucose in complex-type glycosylpeptides was dramatically reduced. The effect of SW on fucose-labeled glycosylpeptides of embryonic chick heart was similar to that observed on MO₄ cells. After removal of sialic acid, control and SW-treated glycosylpeptides from both MO₄ and embryonic chick heart cells had similar gel-chromatographic profiles, suggesting that a decrease in cell surface sialic acid accounts to a large extent for the difference between glycans from control and SW-treated cells. Additional biological experiments were done with dMM (1mM), MdNM (2mM), CS (100 μg ml^-1), 2,5-dihydroxymethy1-3,4-dihydroxypyrrolidine (DMDP; 250 μg ml^-1) and SW (0.4 μ gml^-1)+dNM (10mM). All these compounds or combinations failed to inhibit invasion. The observation that inhibitors of N-linked glycan processing did not interfere with invasion, although they clearly modified the glycosylation of cell proteins, indicated that the integrity of glycans including those of the cell surface might not be a prerequisite for invasion of MO₄ cells into living embryonic tissue in vitro.