Abstract

The quantitative analysis of andrographolides in Andrographis paniculata plant materials is essential for pharmaceutical factories. This analysis cannot be done for all samples due to the conventional process using the extraction and HPLC methods requires a long analysis time and sample destruction. Therefore, near-infrared spectroscopy (NIRS) was employed to classify the class of A. paniculata and to determine the content of two active ingredients, andrographolide (AP1) and dehydroandrographolide (AP3) in A. paniculata, rapidly and non-destructively. One hundred twenty dried powder samples were obtained from aerial parts, branches, leaves, and branches mixed with leaves. The NIR absorption scans were collected from a broad spectral region (1000–2500 nm). Then, the scanned samples were extracted and analyzed for their AP1 and AP3 contents using an HPLC reference method. The success classification model based on AP1 level was developed using the second derivative pretreated NIR spectra of the entire wavelength region using the Partial Least Squares-Discriminant Analysis (PLS-DA) method. The NIR calibration models were developed and tested for quantitative analysis with 50 independent samples. The models were identified for the analysis of the AP1 content with excellent performance (correlation coefficient (R) = 0.98; standard error of validation (SEV) = 0.24%) and for the analysis of the AP3 content at a good level of efficiency (R = 0.93; SEV = 0.15%). This study showed that NIR spectroscopic method offers rapid analysis for the selection of A. paniculata that meets the requirement in bioactive amount.

Graphic Abstract

中文翻译：

---

基于主要生物活性的近红外光谱快速筛选穿心莲药材

摘要

穿心莲植物材料中穿心莲内酯的定量分析是制药厂必不可少的。由于使用萃取和 HPLC 方法的常规过程需要较长的分析时间和样品破坏，因此无法对所有样品进行这种分析。因此，近红外光谱法（NIRS）被用于将类分类的穿心莲并确定两种活性成分的含量，穿心莲内酯（AP1）和脱水（AP3）的穿心莲，快速且无损。从地上部分、树枝、树叶和树枝与树叶混合获得一百二十个干粉样品。NIR 吸收扫描是从宽光谱区域（1000-2500 nm）收集的。然后，使用 HPLC 参考方法提取和分析扫描样品的 AP1 和 AP3 含量。基于 AP1 级别的成功分类模型是使用偏最小二乘判别分析 (PLS-DA) 方法使用整个波长区域的二阶导数预处理 NIR 光谱开发的。NIR 校准模型是针对 50 个独立样品的定量分析而开发和测试的。模型被确定用于分析具有优异性能的 AP1 含量（相关系数（R) = 0.98; 验证标准误差 (SEV) = 0.24%) 并以良好的效率水平分析 AP3 含量 ( R  = 0.93; SEV = 0.15%)。本研究表明，近红外光谱法为筛选符合生物活性量要求的大枣提供了快速分析。

图形摘要