Study on HPLC fingerprint and chemical pattern recognition of Platycodon grandiflorum from different origins
Platycodon grandiflorum (Jacq.) A.DC. is an important medicinal herb that can be used for both food and medicine. It originates from the dried roots of Platycodon grandiflorum (Jacq.) A.DC., a plant in the family Platycodon grandiflorum. It is commonly used in clinical practice for cough, phlegm, chest tightness, sore throat, hoarseness, and pus discharge from the lungs. This species is widely distributed, ranging from Guangdong and Guangxi in the south to western Sichuan, Japan and Korea in the east, and Siberia in the north. Platycodon grandiflorum is widely cultivated in Northeast, East, Central, and Southwest China, and its medicinal materials mainly come from cultivation and production. The planting area in Chifeng City, Inner Mongolia is the largest, with an annual planting area of about 50000 to 70000 mu, accounting for about 50% to 60% of the total national output. Previous studies have shown that Platycodon grandiflorum contains various types of components such as saponins, flavonoids, and phenolic acids. Among them, saponins are the main active ingredients in Platycodon grandiflorum for cough relief and phlegm elimination. Currently, more than 70 types of saponin compounds have been isolated from Platycodon grandiflorum.
Therefore, it is extremely necessary to construct a scientific and reasonable quality control method for Platycodon grandiflorus based on its main saponin components for its clinical application and new drug development. Previous studies have mostly focused on the separation of chemical components in Platycodon grandiflorum and the determination of a very small number of saponin components. Some studies have constructed fingerprint spectra using high-performance liquid chromatography equipped with evaporative light scattering detectors, but there have been no reports on the use of high-performance liquid chromatography combined with ultraviolet visible light detectors to construct fingerprint spectra and explore the quality control and evaluation of Platycodon grandiflorum using chemical pattern recognition. The study used a YMC Hydrosphere C18 analytical chromatography column (250 × 4.6 mm, 5 μ m) with water and acetonitrile as mobile phases for gradient elution, with a detection wavelength of 210 nm. Establish HPLC fingerprint spectra of 15 batches of Platycodon grandiflorum from different origins (Table 2). Chemical pattern recognition methods such as similarity evaluation, cluster analysis, principal component analysis, and orthogonal partial least squares discriminant analysis were used to analyze and evaluate the quality and control methods of Platycodon grandiflorum from different origins.

Platycodon grandiflorum is a large medicinal herb that can be used for both medicine and food, and its main active ingredient is platycodon saponins. The quality evaluation and quality control methods of this herb are of great significance. This study constructed HPLC fingerprint spectra of the main saponin components in Platycodon grandiflorus and identified 7 of them. By combining various chemical pattern recognition methods such as similarity evaluation, cluster analysis, principal component analysis, and orthogonal partial least squares discriminant analysis, compared with existing quality evaluation and quality control methods that only rely on content determination or fingerprint spectra, not only can the consistency of Platycodon grandiflorus medicinal quality be effectively evaluated, but also the main factors that contribute to the quality differences in Platycodon grandiflorus medicinal materials can be elucidated, providing richer information for the quality evaluation and quality control of Platycodon grandiflorus.

The data for chemical pattern recognition comes from the chromatographic peak area of fingerprint spectra, so the quality of fingerprint spectra directly affects the analysis results of chemical patterns. To construct a good fingerprint spectrum, this study investigated and optimized the preparation method, detection wavelength, mobile phase composition, and column temperature of the sample. As a result, it was found that the sample preparation method optimized according to the 2015 edition of the Chinese Pharmacopoeia can obtain a large number of platycodon saponin components. By examining mobile phase systems such as methanol water, acetonitrile water, acetonitrile formic acid water, and acetonitrile phosphoric acid water, as well as different chromatographic column temperatures and detection wavelength conditions, it was found that using acetonitrile water as the mobile phase and selecting a detection wavelength of 210 nm at 35 ℃ can obtain fingerprint spectra with stable baseline, good chromatographic peak separation, and overall high quality.

From the results, it can be seen that the similarity of the fingerprint spectra of 15 batches of Platycodon grandiflorus ranges from 0.927 to 0.991, indicating that there is a high similarity in quality between cultivated and wild Platycodon grandiflorus from different production areas. Cluster analysis and principal component analysis can further classify the 15 batches of Platycodon grandiflorus samples into different categories, indicating that the method constructed in this study can not only evaluate the consistency of Platycodon grandiflorus medicinal quality, but also evaluate the differences of Platycodon grandiflorus medicinal materials. OPLS-DA is a supervised discriminant analysis statistical method commonly used for discriminant analysis of sample differences. The experimental results indicate that chromatographic peaks 6, 9, 13, 18, and 20 have a significant impact on the quality differences of Platycodon grandiflorum from different origins. In this study, cluster analysis and principal component analysis clustered wild samples and cultivated varieties separately, indicating that there are certain differences in the distribution of components between wild samples and cultivated samples. According to the OPLS-DA analysis results, it can be concluded that the relative peak area of the chromatographic peak No. 6 in the fingerprint spectra of the three wild samples ranges from 0.064% to 0.088%, which is higher than that of the cultivated variety by 0.012% to 0.056%; The relative peak area of the 13th chromatographic peak ranges from 0.059% to 0.182%, which is higher than the 0.018% to 0.046% of the cultivated variety; The relative peak area of the 9th chromatographic peak ranges from 0.044% to 0.056%, which is higher than that of the cultivated variety by 0.011% to 0.032%; The chromatographic peaks of numbers 15, 18, and 20 also showed the same trend. Therefore, the slight differences in peak area of these components may be the main reason for the clustering differences between wild samples and cultivated varieties.

In summary, this study used fingerprint spectroscopy combined with chemical pattern recognition technology to explore and analyze the quality of Platycodon grandiflorus samples from different regions across the country, and screened for the characteristic components that cause differences in Platycodon grandiflorus from different regions. This provides scientific reference and experimental basis for the quality control and subsequent development of Platycodon grandiflorus.