Determination of lipophilic metabolites for species discrimination and quality assessment of nine leafy vegetables
Determination of lipophilic metabolites for species discrimination and quality assessment of nine leafy vegetables
复制标题
DOI：
10.1007/s13765-015-0119-6
复制
发表时间：
2015-12-01
影响因子：
--
通讯作者：
Kim, Jae Kwang
中科院分区：
文献类型：
--
作者：
Kim, Tae Jin;Lee, Kyoung Bok;Kim, Jae Kwang
韩国食用的9种蔬菜中的亲脂性化合物具有植物化学含量的多样性。我们还分析了这些化合物之间的含量关系。采用主成分分析（PCA）、Pearson相关分析、层次聚类分析（HCA）和偏最小二乘判别分析（PLS-DA）等数据挖掘方法对18种亲脂性化合物进行了分析。这些物种可以通过PCA的结果进行区分。这些植物化学物质的HCA产生的簇源于密切相关的生化途径。PLS-DA分析结果显示，紫花苋科、菊科、芸苔科和锦葵科4个科的提取物具有显著的分离性。在PLS-DA模型中，促进分化的主要代谢物是油菜甾醇、β -谷甾醇、β -amyrin、豆甾醇、胆固醇、八糖醇（c28）、α -amyrin和六糖醇（c26）。甜菜中三康醇（c30）含量较高，与c28含量呈正相关（r = 0.746, p < 0.0001）。菊科、莴苣和橡树等菊科植物的豆甾醇、α -amyrin和β -amyrin含量高于其他植物科。这些结果证明了代谢谱分析与多变量分析相结合，在蔬菜品种鉴别和食品质量评价方面的实用性。
The lipophilic compounds in nine vegetables consumed in Korea were characterized for diversity in phytochemical content. We also analyzed the relationships among these compounds in terms of their contents. The profiles of 18 lipophilic compounds in the leaves were subjected to data-mining processes, including principal component analysis (PCA), Pearson's correlation analysis, hierarchical clustering analysis (HCA), and partial least squares discriminant analysis (PLS-DA). These species could be distinguished by means of the PCA results. HCA of these phytochemicals resulted in clusters derived from closely related biochemical pathways. PLS-DA showed significant separation among extracts from the following four families: Amaranthaceae, Asteraceae, Brassicaceae, and Malvaceae. The major metabolites that facilitated differentiation in the PLS-DA model were campesterol, beta-sitosterol, beta-amyrin, stigmasterol, cholesterol, octacosanol (c28), alpha-amyrin, and hexacosanol (c26). Chard contained high levels of triacontanol (c30), which was positively correlated with the c28 content (r = 0.746, p < 0.0001). Stigmasterol, alpha-amyrin, and beta-amyrin contents were higher in Asteraceae species including chicon, lettuce, and oak than other plant families. These results demonstrate the utility of metabolic profiling, combined with multivariate analysis, for discrimination of vegetable species as well as evaluation of food quality.