Study On The High-Throughput Analysis Of Ephedra Sinica Plants With Diffuse Reflectance FT-NIR And Chemometrics

Ephedra sinica (Ephedraceae), which has been recorded in thePharmacopoeia of the People’s Republic of China (2010edition) as one ofthe herb Ephedra (Ma Huang), is the main source of Medicinal Ephedra.Due to the small differences in its morphology, histological structure, andsecondary metabolites, the accurately identification of Ephedra sinica ofdifferent habitats, especially different picking times, is difficult by themorphological or histological identification methods; While thechromatography has some disadvantages, such as, complex samplepreparation, long analysis time, and consumption of chemical reagents. Onthe other hand, Ephedra alkaloids are the major active constituents ofEphedra herbs, so the determination of ephedrine and pseudoephedrine inEphedra sinica is also important in the research of Ephedra herbs.Diffuse reflectance Fourier transform near infrared spectroscopy(diffuse reflectance FT-NIR) which has the advantages of without specialsample preparation, nor the use of chemical reagents, pollution-free, small sample amount, rapid and easy to use, combined with chemometrics, canachieve the rapid high-throughput analysis of complex samples. In this study,a high-throughput analytical method of Ephedra sinica by near-infrareddiffuse reflectance spectra was established. Combined with thecounter-propagation artificial neural network (CP-ANN), different habitatsand different picking times of Ephedra sinica were identified; and combinedwith the partial least square (PLS), the ephedrine and pseudoephedrinecontents were determined. Prediction accuracies of the validation samplewere used for the evaluation of habitats and picking times; root mean squareerror of cross validation (RMSECV), root mean square error of prediction(RMSEP) and correlation coefficient (R) were used for ephedrine andpseudoephedrine contents.OBJECTIVES1. Establish a method to discriminate different habitats and differentpicking times of Ephedra sinica plants with diffuse reflectance FT-NIR.2. Establish a method to determine ephedrine and pseudoephedrinecontents of Ephedra sinica plants with diffuse reflectance FT-NIR.METHODS1. Measurements of near infrared diffuse reflectance spectra ofEphedra sinica.2. Determination of the reference values of ephedrine and pseudoephedrine contents of Ephedra sinica.3. Calibration and validation of the prediction models for the analysisof Ephedra sinica(1) The habitats and picking time models of diffuse reflectance FT-NIRwere respectively established by counter-propagation artificial neuralnetwork (CP-ANN) and evaluated with the prediction accuracies of thevalidation samples.(2) The ephedrine and pseudoephedrine content models of diffusereflectance FT-NIR were respectively established by partial least square(PLS) and evaluated with RMSECV, RMSEP and R.RESULTS1. The prediction accuracies of the CP-ANN models for the validationsamples were100.0%for the two habitats and80.0%for the two pickingtimes.2. The RMSECV of the CP-ANN models for ephedrine andpseudoephedrine were2.62201and1.33701, RMSEP were1.12and0.236,the R of the prediction and reference values were0.9721and0.9309.CONCLUSIONSThe proposed approach combined CP-ANN and PLS couldsimultaneously identify habitats and picking times of Ephedra sinica plantsand determine ephedrine and pseudoephedrine in the plants, which achieve the high-throughput analysis of Ephedra sinica plants.