Glycosyl linkage characteristics and classifications of exo-polysaccharides of some regionally different strains of Lentinula edodes by amplified fragment length polymorphism assay and cluster analysis.

We report here the first combined amplified fragment length polymorphism (AFLP) analysis of genomic DNA fingerprinting data and cluster analysis of the exo-polysaccharide glycosyl linkage data of 10 regionally different strains of Lentinula edodes to compare their genetic and structural similarities and differences. In addition, the monosaccharide compositions, molecular weights, glycosyl structural linkages were investigated for the exo-polysaccharides extracted from these different phylogenetic groups of regionally different L. edodes. All exo-polysaccharides had similar molecular weight distribution between 1x10(4) and 3x10(6) Da and the monosaccharide composition analysis revealed the presence of heterogeneous materials containing glucose, mannose, xylose, galactose, fucose, rhamnose and arabinose in different ratios. Among these monosaccharides, the glucose contents are the highest for all but one strain, indicating that glucose probably is the building block of the backbones of these exo-polysaccharides. The AFLP assay data helped to classify the 10 L. edodes strains into three distinct genetic groups. Gas chromatographic and mass spectrometric (GC-MS) data revealed five different glycosyl linkage types for these exo-polysaccharides. Most of the exo-polysaccharide backbone structures contain (1-->4)-linked-D-glucopyranosyl and (1-->6)-linked-D-glucopyranosyl moieties. Arabinose 1-->4 linkages and mannose 1-->2 linkages also exist in all strains. The only differences among these linkages are their monosaccharide compositions leading to different degree of backbone and branch formations. Cluster analyses of the GC-MS data of the exo-polysaccharides of the 10 strains resulted in 10 dendrograms. However, four of the 10 dendrograms were identical and were obtained using the average, Ward and weighted linkage type method of Manhattan distance and using the Ward method of Euclidean distance. The results of cluster analyses were not very much different from that of the AFLP assay and allowed the comparison of genetic and structural similarities and differences.