Simultaneous Quantification of Nine Target Compounds in Traditional Korean Medicine, Bopyeo-Tang, Using High-Performance Liquid Chromatography–Photodiode Array Detector and Ultra-Performance Liquid Chromatography–Tandem Mass Spectrometry

Bopyeo-tang (BPT) is composed of six medicinal herbs (Morus alba L., Rehmannia glutinosa (Gaertn.) DC., Panax ginseng C.A.Mey., Aster tataricus L.f., Astragalus propinquus Schischkin, and Schisandra chinensis (Turcz.) Baill.) and has been used for the treatment of lung diseases. This study focused on establishing an analytical method that can simultaneously quantify nine target compounds (i.e., hydroxymethylfurfural, mulberroside A, chlorogenic acid, calycosin-7-O-glucoside, 3,5-dicaffeoylquinic acid, quercetin, kaempferol, schizandrin, and gomisin A) from a BPT sample using high-performance liquid chromatography with a photodiode array detector (HPLC–PDA) and ultra-performance liquid chromatography with tandem mass spectrometry (UPLC–MS/MS). The separation of compounds in both analyses was performed on a C18 reversed-phase column using the gradient elution of water–acetonitrile as the mobile phase. In particular, the multiple reaction monitoring mode was applied for quick and accurate detection in UPLC–MS/MS analysis. As a result of analyzing the two methods, HPLC–PDA and UPLC–MS/MS, the coefficient of determination of the regression equation for each compound was ≥0.9952, and recovery was 85.99−106.40% (relative standard deviation (RSD) < 9.58%). Precision testing of the nine compounds was verified (RSD < 10.0%). The application of these analytical assays under optimized conditions for quantitative analysis of the BPT sample gave 0.01–4.70 mg/g. Therefore, these two assays could be used successfully to gather basic data for clinical research and the quality control of BPT.


Results and Discussion
2.1.HPLC-PDA Analysis 2.1.1.Selection of Target Compounds in BPT for Simultaneous Quantification by HPLC-PDA For the selection of target compounds for the quality assessment of BPT, 17 candidate components were compared with the samples (BPT, M. alba, R. glutinosa, P. ginseng, A. tataricus, A. propinquus, and S. chinensis samples).Specifically, the 17 components to be compared were the following: mulberroside A, rutin, isoquercetin, and resveratrol of M. alba; hydroxymethylfurfural of R. glutinosa; ginsenoside Rb 1 and ginsenoside Rg 1 of P. ginseng; chlorogenic acid, 3,4-dicaffeoylquinic acid, 3,5-dicaffeoylquinic acid, quercetin, and kaempferol of A. tataricus; astragaloside IV and calycosin-7-O-glucoside of A. propinquus; and schizandrin, gomisin A, and gomisin N of S. chinensis.Comparison HPLC chromatograms for each sample and the candidate components are shown in Figure S1.Following a comparison of the results, among the 17 candidate components, nine compounds were finally detected in the BPT sample, which were then selected as target compounds in BPT for simultaneous quantification by HPLC-PDA.

HPLC Operating Conditions for Simultaneous Quantification of BPT
Various parameters, such as the type of column and the temperature of the column oven, and the acid added to the mobile phase were compared to determine the optimal HPLC analytical conditions for the simultaneous quantification of the nine targets selected (Figure S2) from a BPT sample.As a first step, reverse-phase C 18 columns from different manufacturers were compared to select an appropriate column for the separation of the target compounds.Columns included the following: SunFire TM (Waters), Capcell Pak UG80 (Shiseido, Tokyo, Japan), and Gemini (Phenomenex, Torrance, CA, USA).The columns were identical in length (250 mm), inner diameter (4.6 mm), and particle size (5 µm).As shown in Figure S3, seven and eight components were detected on the Gemini column (Figure S3B) and the Capcell Pak UG80 column (Figure S3D), respectively, while nine components were detected on the Waters SunFire TM column (Figure S3F).The latter column was therefore considered the most preferable for further work.
As a second step, the effects of the acid(s) (i.e., formic acid, phosphoric acid, trifluoroacetic acid, and acetic acid) added to the mobile phase on the separation of the nine target compounds in the first determined column were compared.As a result, as shown in Figure S4, when trifluoroacetic acid and phosphoric acid were added, the 3,5-dicaffeoylquinic acid of both acids overlapped with the peak of an unknown peak, and it was detected (Figure S4B,D).Also, when acetic acid was added, the calycosin-7-O-glucoside overlapped with the unknown peak, and it was detected (Figure S4F).However, in the case of formic acid, the nine target compounds were well separated without interference from neighboring components (Figure S4H).Therefore, formic acid was selected as the acid of choice to be added to the mobile phase.Different column temperatures (30,35, and 40 • C) were considered, and 30 • C was determined to be the most suitable (Figure S5).
The following were then established as the optimal conditions for the simultaneous analysis of the nine target compounds from a BPT sample: Waters SunFire TM column, distilled water-acetonitrile mobile phase (both containing 0.1% (v/v) formic acid), and 30 • C column temperature.Table S1 summarizes the optimized analytical conditions and gradient elution conditions of the mobile phase in more detail.Under the established optimal analysis conditions, all target compounds were completely eluted within 45 min with a resolution of ≥10.70.Representative HPLC chromatograms are shown in Figure 1.
Molecules 2024, 29, x FOR PEER REVIEW 3 of 14 As a second step, the effects of the acid(s) (i.e., formic acid, phosphoric acid, trifluoroacetic acid, and acetic acid) added to the mobile phase on the separation of the nine target compounds in the first determined column were compared.As a result, as shown in Figure S4, when trifluoroacetic acid and phosphoric acid were added, the 3,5-dicaffeoylquinic acid of both acids overlapped with the peak of an unknown peak, and it was detected (Figure S4B,D).Also, when acetic acid was added, the calycosin-7-O-glucoside overlapped with the unknown peak, and it was detected (Figure S4F).However, in the case of formic acid, the nine target compounds were well separated without interference from neighboring components (Figure S4H).Therefore, formic acid was selected as the acid of choice to be added to the mobile phase.Different column temperatures (30,35, and 40 °C) were considered, and 30 °C was determined to be the most suitable (Figure S5).
Suitable results were found for all the verification parameters, which confirmed that the established analytical method was suitable for the simultaneous quantification of the nine target compounds selected from BPT.

UPLC-MS/MS Simultaneous Analysis 2.2.1. UPLC-MS/MS Multiple Reaction Monitoring (MRM) Method for Simultaneous Analysis
Simultaneous determination of target components in BPT by UPLC-MS/MS was conducted on the nine compounds selected in the HPLC-PDA analysis assay.As a result of detecting these components, using the ESI mode, two components (i.e., chlorogenic acid and quercetin) were detected in negative ion mode, and the other seven components (i.e., hydroxymethylfurfural, mulberroside A, calycosin-7-O-glucoside, 3,5dicaffeoylquinic acid, kaempferol, schizandrin, and gomisin A) were detected in positive ion mode (Figures 2 and S6).
The MRM transitions (precursor ion (Q1) and product ion (Q3)) of each compound are shown in Table 5 and Figure S7.Briefly, hydroxymethylfurfural and schizandrin were set to m/z 109.0 and 415.0, which are the ions generated by the removal of a water molecule from Q1, respectively [20,21].The Q3 peak of mulberroside A was set at m/z 244.9, where two glucopyranosyl groups were removed [22].In the case of calycosin-7-O-glucoside, one glucose molecule was eliminated, and the peak generated at m/z 284.9 was designated as Q3 [23], while in chlorogenic acid and 3,5-dicaffeoylquinic acid, the m/z 162.9 of the caffeoyl group was set as the Q3 peak [24].The flavonols, kaempferol and quercetin, were produced by the cleavage of the C-ring, and m/z 152.9 and 150.9 were set as the Q3 peak, respectively [25].The Q3 peak of gomisin A was set at m/z 341.0, which is the ion generated by the removing the water molecule, CH 2 O, and CO groups from Q1 [26,27].

Plant Materials
The six raw herbal medicines (see Table S4) were purchased from Kwangmyungdang Pharmaceutical (Ulsan, Republic of Korea).Prior to use, they were subjected to morphological sensory tests by Dr. Goya Choi, Korea Institute of Oriental Medicine (KIOM, Daejeon, Republic of Korea).Scientific names were verified from the World Folra Online Plant List (www.wfoplantlist.org;21 November 2023) [28].Six crude herbs (CA05-1 to CA05-6) were stored in the KM Science Research Division, KIOM.

Chemicals and Reagents
The reference target compounds used in this simultaneous quantification were purchased from specialized natural product manufacturing companies: hydroxymethylfurfural and chlorogenic acid from Merck KGaA (Darmstadt, Germany); mulberroside A from Ensol BioSciences (Daejeon, Republic of Korea); calycosin-7-O-glucoside, 3,5-dicaffeoylquinic acid, kaempferol, and gomisin A from Shanghai Sunny Biotech (Shanghai, China); quercetin from ChemFaces Biochemical (Wuhan, China); and schizandrin from Biopurify Phytochemicals (Chengdu, China).Detailed information on the structures of these compounds is given in Table S5 and Figure S2, respectively.For analysis, all solvents (i.e., methanol, acetonitrile, and distilled water) and reagents (i.e., formic acid, trifluoroacetic acid, phosphoric acid, and acetic acid) were either HPLC or LC-MS grade.They were purchased from JT Baker (Phillipsburg, NJ, USA), Merck (Darmstadt, Germany), or Thermo Fisher Scientific (Cleveland, OH, USA).

Preparation of the BPT Sample
Preparation of the BPT sample was conducted at KIOM following previously reported preparation protocols [29][30][31].Briefly, after mixing the amounts as shown in Table S4 (each at 1500 g; M. alba and R. glutinosa, each at 500 g; P. ginseng, A. tataricus, A. propinquus, and S. chinensis), 50 L of distilled water was added, and the mixture was boiled at 100 • C for 2 h using a COSMOS-660 heating extractor (Kyungseo E&P, Incheon, Republic of Korea).The extract was lyophilized to obtain a powder sample (1600 g, yield 32.0%).The lyophilized sample was stored at −20 • C until it was required for use.

Equipment and Analytical Conditions for HPLC-PDA Simultaneous Quantification
A Prominence LC-20A series HPLC system (Shimadzu, Kyoto, Japan) was used to analyze nine target components from a BPT sample simultaneously.The system comprised two mobile phase delivery units (i.e., pumps), an online degasser, a column oven with forced air circulation, an autosampler with cooling, and a photodiode array detector.These systems were controlled using LC solution software (version 1.24; Shimadzu).The nine targets (i.e., hydroxymethylfurfural, mulberroside A, chlorogenic acid, calycosin-7-O-glucoside, 3,5-dicaffeoylquinic acid, quercetin, kaempferol, schizandrin, and gomisin A) were separated, without any other interfering peaks, using a Waters SunFire TM reversephase analytical column (250 mm length × 4.6 mm inner diameter, particle size 5 µm; Waters, Milford, MA, USA) and a distilled water-acetonitrile (both containing 0.1% (v/v) formic acid) gradient elution condition.Further details of the HPLC analysis conditions are given in Table S1.

Equipment and Analytical Conditions for UPLC-MS/MS Simultaneous Quantification
The simultaneous quantification of nine target compounds in a BPT sample was performed using a UPLC-MS/MS system comprising a Waters Acquity UPLC H-Class PLUS system and a TQ-S micro-MS system (Xevo, Milford, MA, USA).The operation conditions are given in Table S6.Various parameters for UPLC-MS/MS MRM analysis of targets are given in Table 5.These include the ion mode, MRM transition, cone voltage, and collision energy.

Validation of Established Assays in HPLC-PDA and UPLC-MS/MS Systems
Based on guidelines from the International Conference on Harmonization [32], linearity, sensitivity, such as the LOD and LOQ, accuracy, and precision were evaluated to verify the established assays.Briefly, in both methods, the linearity was evaluated by the coefficient of determination (r 2 ) value in the regression equation for each analyte.In the case of the HPLC-PDA method, the LOD and LOQ concentrations of each compound were calculated using the following equation: where σ is the standard deviation (SD) of the y-intercept, and S is the slope of the regression equation.
On the other hand, in the UPLC-MS/MS method, the LOD and LOQ concentrations were calculated using signal-to-noise ratios of 3:1 and 10:1, respectively.
Determination of the recovery was conducted using the standard addition method with three different concentrations (i.e., low, medium, and high) of the target compounds.The recovery parameter was calculated based on the following equation: Recovery(%) = found amount spiked amount × 100 target components in a BPT sample by HPLC-PDA; Table S2: System suitability for the simultaneous analysis of the nine target components by HPLC-PDA; Table S3: Repeatability of retention time and peak area of the nine targets by HPLC (n = 6); Table S4: Composition of and information on Bopyeotang; Table S5: Information on the nine reference standard compounds; Table S6: UPLC-MS/MS MRM conditions for the simultaneous analysis of nine target components in BPT.

Figure 2 .
Figure 2. Representative total ion chromatograms of the standard solution (A) and the BPT sample (B) using the UPLC−MS/MS MRM method.Hydroxymethylfurfural (1), mulberroside A

Table 2 .
Recovery (%) of the selected nine target compounds in the established HPLC-PDA method.

Table 3 .
Precision test of the nine target compounds in the established HPLC-PDA method.

Table 4 .
Amounts (mg/g) of the nine target compounds in the BPT sample by the established HPLC-PDA assay.

Table 5 .
Optimized parameters for the UPLC-MS/MS MRM simultaneous analysis of the nine analytes in BPT.

Table 6 .
Retention time, the linear range, regression equation, r 2 , LOD, and LOQ of the nine compounds by the UPLC-MS/MS MRM analytical method.

Table 8 .
Intra-and inter-day precision data of the nine compounds evaluated by the developed UPLC-MS/MS MRM analytical method (n = 3).

Table 9 .
Amounts (mg/g) of the nine target compounds in a BPT sample, as evaluated by the developed UPLC−MS/MS MRM assays.