高级检索

    基于指纹图谱和化学模式识别法的墨染樱花质量评价以及化学成分的鉴定

    Quality Evaluation of Flowers of Prunus Lannesiana Var. Speciosa Cv. Shin-Sumizome Based on Fingerprints Combined with Chemical Pattern Recognition and Identification of Chemical Components

    • 摘要: 取干燥墨染樱花瓣,粉碎,过筛,分取0.500 g,加入25 mL 70%(体积分数,下同)甲醇溶液,超声30 min。冷却至室温,用70%甲醇溶液补足重量,摇匀后过 0.22 μm滤膜,续滤液按照仪器工作条件测定。以高效液相色谱法采集10批不同采收期的样品的指纹图谱,并结合相似度分析和化学模式识别法对10批样品质量进行评价。基于超高效液相色谱-线性离子阱/静电场轨道阱质谱法鉴定样品中的化学成分。结果显示:比对10批样品的指纹图谱,筛选出8个共有成分,通过对照品比对确定了2个共有成分绿原酸和芦丁;10批样品的指纹图谱与对照指纹图谱的相似度均大于0.980,说明不同采收期样品质量较稳定;将8个共有峰的峰面积导入SPSS 21.0软件和SIMCA 14.1软件,系统聚类分析、主成分分析、正交偏最小二乘-判别分析(OPLS-DA)模型均将10批样品分为4类,且分类结果完全一致;以变量重要性投影分析筛选引起质量差异的标志物,共筛选出3个成分;通过质谱库和文献比对,鉴定出48个化学成分,包括黄酮类19个,有机酸类18个,萜类4个,其他类7个。

       

      Abstract: Dried samples of Prunus lannesiana var. speciosa cv. Shin-sumizome were taken, crushed, and sieved, and an aliquot (0.500 g) was taken into 25 mL of 70% (volume fraction, the same below) methanol solution. The mixture was sonicated for 30 min, and cooled to room temperature. The weight of the above solution was made up by 70% methanol solution. After shaking evenly, the mixed solution was passed through a 0.22 μm filter membrane. The subsequent filtrate was determined according to the operating conditions of the instrument. Fingerprints of 10 batches of samples from different harvesting periods were collected by high performance liquid chromatography, and quality evaluation of the 10 batches of samples was made by combining similarity analysis and chemical pattern recognition. Identification of chemical components in samples was made based on ultra-high performance liquid chromatography-linear ion trap/electrostatic orbitrap mass spectrometry. It was shown that 8 common components were screened by comparing the fingerprints of 10 batches of samples, and 2 common components, chlorogenic acid and rutin, were determined through comparison with reference standards. The similarity between the fingerprints of 10 batches of samples and the reference fingerprint was greater than 0.980, indicating that the quality of samples from different harvesting periods was relatively stable. Importing peak areas of 8 common peaks into SPSS 21.0 software and SIMCA 14.1 software, 10 batches of samples were classified into 4 categories by systematic clustering analysis, principal component analysis, and orthogonal partial least squares discriminant analysis (OPLS-DA) models, and the classification results were completely consistent. Using variable importance projection analysis to screen for biomarkers that caused quality differences, a total of three components were identified. Through mass spectrometry library and literature comparison, 48 chemical components were identified, including 19 flavonoids, 18 organic acids, 4 terpenes, and 7 other classes.

       

    /

    返回文章
    返回