Chemometrics shown to fix complexity of herbal mix

Skip to Navigation

Ezine

  • Published: Dec 14, 2017
  • Author: Ryan De Vooght-Johnson
  • Channels: Laboratory Informatics / Chemometrics & Informatics
thumbnail image: Chemometrics shown to fix complexity of herbal mix

Mixed herbal medicine, Lianqiao Baidu, requires new analysis methods

Lianqiao Baidu pills (LQBDPs) are a traditional Chinese medicine (TCM) consisting of a mixture of 19 herbs. They are used to treat constipation, skin ulcers, rheumatism, herpes and colds. LQBDP has been included in the Chinese National Standard for patent medicine, but only microscopic and TLC methods of analysis are currently used. Other QC studies have merely looked at identifying one or just a few compounds, rather than giving an overall picture of the mixture.

The Shenyang researchers used HPLC and chemometric methods to obtain measures of LQBDP quality. Fifteen marker compounds were examined by HPLC. The data were processed using the limited-ratio quantified fingerprint method (LRQFM), which is a method of comparing sample fingerprints (HPLCs in this case) with reference fingerprints. Principal component analysis (PCA) and a partial least squares (PLS) model were also employed.

HPLC and chemometric methods applied to Lianqiao Baidu pills

Pulverised LQBDPs were extracted with ethanol and then methanol. The combined extracts were evaporated, and the residue was dissolved in methanol. Standard samples of the 15 marker compounds were also dissolved in methanol. HPLC was carried out using a Kromosil C18 column with gradient elution, using an aqueous mobile phase containing 0.2% phosphoric acid (3 mM sodium 1-heptanesulfonate) and an organic phase of 1% acetic acid in acetonitrile. The proportion of the organic phase was increased from 3 to 90% in a series of gradients. The extraction conditions, column and gradient were optimised to give the maximum value of S, the total entropy of the HPLC fingerprint (a measure of the amount of information given by the HPLC trace). Twenty-eight samples of LQBDP were run on the HPLC at five different wavelengths, using DAD (PDA) detection.

The limited-range quantified fingerprint method (LRQFM) was applied to the HPLC data using in-house software. A reference fingerprint was generated from the average results from 20 of the samples that came from a high-quality supplier, and the sample fingerprints were compared with it. Macro-qualitative similarity values (Sm), macro-quantitative similarity values (Pm) and fingerprint-levelling coefficients (α), all measures of the difference between the sample and the reference fingerprint, were determined for the samples. These values were used to place the samples into eight quality grades, ranging from 8 (lowest) to 1 (highest). Samples of grades 1 to 5 were considered to be qualified by the method; two samples were classed as grade 6 (fail), with the remaining 26 being classed in grades 1 to 5.

A three-component PCA analysis based on the HPLC traces was applied. Apart from two outliers (which were classed as grades 4 and 5 by the LRQFM assessment), the samples could be divided into two groups: a main group of 24 samples and another group that consisted of the two samples that were classed as grade 6 by the LRQFM assessment.

The antioxidant activity of the samples was measured by noting their ability to quench 2,2-diphenyl-1-picrylhydrazyl (DPPH) radicals. A partial least squares (PLS) method was employed, plotting antioxidant activity against a measure of HPLC areas. Four samples were excluded as outliers, leaving 24 samples (these were the same ones that were in the main group in the PCA). Nineteen samples were used as a training group, while five samples acted as a validation group. Good linearity was achieved, and the actual quench values for the validation group were in line with those predicted by the PLS model.

Chemometric methods help assess herbal mixture

The use of chemometric methods allowed the assessment of the quality of a multi-component herbal mixture. The LRQFM gave roughly similar results to the more commonly used PCA method; it remains to be seen whether the former method will be widely used in the future. It was also possible to link the HPLC results to antioxidant activity using PLS analysis.

Related Links

Journal of Separation Science, 2017, Early View paper. Chen et al. Quantitative fingerprinting based on the limited-ratio quantified fingerprint method for an overall quality consistency assessment and antioxidant activity determination of Lianqiao Baidu pills using HPLC with a diode array detector combined with chemometric methods.

Wikipedia, Partial Least Squares Regression

Wikipedia, Principal Component Analysis

Article by Ryan De Vooght-Johnson

The views represented in this article are solely those of the author and do not necessarily represent those of John Wiley and Sons, Ltd.

Follow us on Twitter!

Social Links

Share This Links

Bookmark and Share

Microsites

Suppliers Selection
Societies Selection

Banner Ad

Click here to see
all job opportunities

Copyright Information

Interested in spectroscopy? Visit our sister site spectroscopyNOW.com

Copyright © 2018 John Wiley & Sons, Inc. All Rights Reserved