RESUMO
OBJECTIVE：To provide reference for the identification and proces sing end-point determination of raw Morus alba and its processed products （honey-processed M. alba ）. METHODS ：UPLC method was adopted. The determination was performed on Waters BEH Shield RP C 18 column with mobile phase consisted of acetonitrile- 0.1％ phosphoric acid solution （gradient elution ） at the flow rate of 0.30 mL/min. The column temperature was set at 30 ℃. The program wavelengths were set at 280 nm（0-4 min） and 320 nm（4-35 min）. Similarity Evaluation System for Chromatogram Fingerprint of TCM （2012 edition）was used to establish UPLC fingerprint and carry out similarity evaluation of 13 batches of M. alba and honey-processed M. alba . The chromatographic peaks were identified with reference substance fingerprint. The colorimetric value （L，a，b） of 13 batches of M. alba and honey-processed M. alba powder were determined ，and average total colorimetric value （E）was calculated. OPLS-DA and cluster analysis were adopted to analyze the differences in fingerprints and colorimetric values of M. alba before and after processing. At the same time ，the dynamic change rule of fingerprint and colorimetric value of honey-processed M. alba at different processing time points were analyzed to determine the processing end-point. RESULTS ：There were obvious differences in fingerprints before and after processing ，and the similarity of 13 batches of M. alba and honey-processed M. alba were all higher than 0.9. Totally 21 common peaks were calibrated for M. alba ，and 23 common peaks for honey-processed M. alba ；peak 1 and peak 2 were newly produced compounds of honey-processed M. alba . Peak 2，peak 7，peak 14 and peak 19 were identified as 5-hydroxymethylfurfural， mulberry glucoside A ，oxidized resveratrol ，mulberry flavonoids G. Results of OPLS-DA showed that the peak area-sample quantity ratio of peak 1，peak 2，peak 18，peak 20 and the chromaticity values （L，a，b）were the most important factors affecting the difference of raw and processed products of M. alba . When the E ranged 75.84-80.88 as the processing end-point of honey-processed M. alba ，the processing time was determined as 22-34 min. CONCLUSIONS ： The established UPLC fingerprint and colorimetric value determination method can be used to identify the raw and processed products of M. alba as well as determine the processing end-point of honey-processed M. alba .
RESUMO
OBJECTIVE：To establish HP LC ch aracteristic ch romatogram of different medicinal parts of Cirsium japonicum ， and to compare the difference of chemical components in different medicinal parts of C. japonicum according to chemical identification method ，and to provide reference for quality control and evaluation of C. japonicum . METHODS ：Medicinal material （overground part ），leaves，flower，main stem and lateral stem of C. japonicum were determined by HPLC. According to the TCM Chromatographic Fingerprint Similarity Evaluation System （2012A edition ），the chromatograms were matched to generate the HPLC characteristic chromatogram of each medicinal part. The differences of common characteristic peak area were analyzed according to variance analysis of single factor. The chromatographic peaks were identified by comparison of reference substance. Meanwhile，the chemical pattern recognition was performed to research the different medicinal parts of C. japonicum according to principal component analysis （PCA）and cluster analysis. RESULTS ：HPLC characteristic chromatograms of medicinal material ， leaves，flower，main stem and lateral stem from C. japonicum were established respectively ，and 15 common peaks were confirmed for medicinal material ，leaves and flower of C. japonicum ；11 common peaks were confirmed in chromatograms of main stem and lateral stem from C. japonicum （absence of No. 7，9，12，13 peak）. The contents of chemical components were different greatly among different medicinal parts. No. 1，2，3，10，11 peaks were identified as neochlorogenic acid ，chlorogenic acid ， cryptochlorogenic acid ，linarin and pectolinarin. Results of PCA and cluster analysis showed that chemical pattern recognition and clustering of the flower and stem of C. japonicum were distinct and can be clustered into one category respectively. However ，the leaves distribution of C. japonicum was relatively scattered ，so it was difficult to cluster . CONCLUSIONS ：Established HPLC characteristic chromatogram-chemical pattern recognition can reflect the differences of different medicinal parts of C. japonicum integrally， comprehensively and truly ， which has vital significance for origin indentification ， quality control and overall evaluation of C. japonicum .