Applied Microbiology and Biotechnology vol:57 issue:1-2 pages:175-81
Several yeasts, such as Candida utilis, Dekkera bruxellensis, Hanseniaspora guilliermondii, Kloeckera apiculata, Saccharomyces cerevisiae and Schizosaccharomyces pombe, were found to coaggregate with Escherichia coli, but S. pombe showed much less coflocculation than the other yeasts (Peng et al. 2001)). S. pombe is known to have galactose-rich cell walls and we investigated whether this might be responsible for its different behavior by studying the wild-type TP4-1D, with a mannose to galactose ratio of 1 to 1.2, and the glycosylation mutant gms1delta (Man:Gal=1:0). The wild-type induced very low levels of coflocculation (3%) while gms1delta induced a remarkable amount of coflocculation (48%). Coflocculation of the mutant was inhibited by mannose but not affected by galactose or glucose. The S. cerevisiae mnn2 mutant, with a mannan structure similar to gms1delta, also showed a high degree of coflocculation (40%). However, S. cerevisiae mutant mnn9, with a mature core similar to S. pombe, showed decreased coflocculation (21.3%). Both these S. cerevisae mutants were sensitive to mannose inhibition. Coflocculation of E. coli and gms1delta also could be inhibited by gms1delta mannan and plant lectins, such as HHA, GNA and NPA, specific to either alpha-1-3- or alpha-1-6-linked mannosyl units. From these results we conclude that the E. coli lectins may have specificity for alpha-1-6- and alpha-1-3-linked mannose residues either in the outer chain or in the core of S. pombe, but in wild-type strains these mannose residues are shielded by galactose residues.