利用分群分析香菇菌種類源與多醣結構的特色及利用多樣統計分析香菇多醣結構與巨噬細胞活性關係之研究

Abstract

本研究利用深層培養十株不同的香菇菌種，將菌種培養至靜止期後萃取其發酵濾液中的多醣體進行單醣組成、分子量分布、結構型態、免疫調節活性及抗腫瘤活性的測定。測定出十株菌種的多醣體皆alpha-(1-4)-glucan鍵結beta-(1-6)-glucan複雜巨分子，分子量分布在1x102 至 3x103 kDa，並且各由不同比例的七種單醣所構成，包括了:葡萄糖、甘露糖、木糖、半乳糖、黑藻糖、鼠李糖及阿拉伯糖。由於固態的多醣體具有不易回溶水的特性，因此我們開發以超臨界流體系統之高壓水方式，在10.1MPa、28°C、70分鐘的條件下提升多醣體的回溶率可達90%，而且此萃取方法依然能保alpha-(1-4)-glucan鍵結beta-(1-6)-glucan的結構及其生物的活性。我們也發現alpha-(1-4)-glucan鍵結beta-(1-6)-glucan多醣體亦能引發RAW264.7巨噬細胞的細胞與細胞間之訊息傳遞。 另外，以多變量迴歸分析十種多醣體中單醣組成與RAW264.7巨噬細胞活性的關聯及建立結構-活性預測模式；以主成分分析法和因素分析法探就出阿拉伯糖、甘露糖、半乳糖及黑藻糖為多醣結構中的組成特色。最後再利用擴增片段長度多形性分析菌種的類源及利用統計的群集分析醣鍵結的構型關係，進一步比較基因分群與結構分群的相同或相異之處。多醣體的結構與其引發的生物活性相當複雜，我們的研究結果能助於開啟醣類資訊學的新興研究的方法，也能有助於簡化醣類領域的複雜問題。

In this research, we have made major efforts to isolate polysaccharides from the submerged cultures of ten different strains of Lentinula edodes and have determined their monosaccharide compositions, molecular weights, structural linkage types, immuno-modulating and anti-cancer activities. The isolates had a molecular weight distribution between 1x102 and 3x103 kDa. The monosaccharide composition analysis revealed that these polysaccharides were heterogeneous and contained glucose, mannose, xylose, galactose, fucose, rhamnose and arabinose in different ratios. A major form of polysaccharide linkage with a backbone of alpha-(1-4)-glucan and side chains of beta-(1-6)-glucan has been identified. A method using the pressurized water extraction (PWE) technique (10.1 MPa, 28 °C, 70 min) to improve the extraction efficiency and to retain the general structures and bioactivities of the polysaccharides has also been developed. These polysaccharides were also found to stimulate macrophage cell-cell interactions. The use of multiple linear regression analysis allows the deduction of the correlation between the monosaccharide composition ratios and in vitro macrophage (RAW264.7) stimulatory activities of polysaccharides obtained from different mushroom strains. Arabinose, xylose, mannose, fucose and galactose were identified as the monosaccharides that could be related to macrophage stimulatory activities. In addition, principal component analysis and factor analysis methods were used to find polysaccharide structural characteristics of these L. edodes. Furthermore, investigations by amplified fragment length polymorphism analysis (AFLP) and by statistical cluster analyses of GC-MS linkage data were compared to understand the relationships of genotypes and structural features of these species. The structures and biological activities of these polysaccharides were quite complex and remained to be delineated in the future at the molecular level. The novel feature of using statistical methods to treat limited but complex experimental data in this work represents a new approach in the bioinformatics area and allows the simplification of research problems in the glycobiology field.