Development and Validation of a HPLC–PDA Method for the Simultaneous Determination of Berberine, Palmatine, Geniposide, and Paeoniﬂorin in

: Haedoksamul-tang (HST) is a traditional medical prescription comprising eight medicinal herbs: Angelica gigas , Cnidium o ﬃ cinale , Coptis japonica , Gardenia jasminoides , Paeonia lactiﬂora , Phellodendron amurense , Rehmannia glutinosa , and Scutellaria baicalensis . HST is used to treat blood circulation disorders and has anti-inﬂammatory, hemostatic, and anticonvulsant e ﬀ ects. In this study, a high-performance liquid chromatography / photodiode array detector (HPLC–PDA) method was developed and validated for the simultaneous determination of four marker compounds in HST, namely, berberine, palmatine, geniposide, and paeoniﬂorin. Four standard solutions and HST sample solutions were analyzed using a reverse-phase SunFire ® C 18 column (4.6 × 250 mm, 5 µ m) using a 0.05% aqueous formic acid / methanol gradient. The column temperature, ﬂow rate, injection volume, and wavelengths used were 28 ± 2 °C, 1.0 mL / min, 10.0 µ L, and 230 nm and 240 nm, respectively. Calibration curves of the four marker compounds showed good linearity (r 2 ≥ 0.9994), and limits of detection (LODs) and quantiﬁcation (LOQs) were in the ranges 0.131–0.296 µ g / mL and 0.398–0.898 µ g / mL, respectively. Ranges of intra- and inter-day precisions and accuracies values were 96.74–102.53% and 97.95–100.83%, respectively, and relative standard deviation (RSD) values were all < 4%. Recoveries averaged 92.33–116.72% with RSD values < 5%. Quantitative analysis for the four marker compounds showed geniposide (10.77 mg / g) was most abundant in HST.


Plant Materials and Chemicals
Scutellaria baicalensis was purchased from Hyunjin pharmaceutical Corp., Coptis japonica from Miryung herb medicine Ltd., Gardenia jasminoides, Phellodendron amurense, and Rehmannia glutinosa from Taechang pharmaceutical Corp., Angelica gigas, Cnidium officinale, and Paeonia lactiflora from Jechen traditional herbal market. All herbal materials were authenticated by Professor Eun Ju Jeong of the Department of Agronomy and Medicinal Plant Resources at Gyeongnam National University of Science and Technology (Jinju, South Korea). Voucher specimens (PNU-0028~PNU-0035) were deposited at the College of Pharmacy, Pusan National University.

Preparation of Haedoksamul-Tang (HST) Extract
The eight herbal materials (100 g of each) were mixed and crushed for HST extraction, and then extracted for 3 h in 8 L of distilled water using an ultrasonic extractor. The extract was passed through filter paper (Advantec, Tokyo, Japan), concentrated under vacuum at 50 • C, and freeze-dried. The lyophilized HST extract was sieved to a particle size of <0.425 mm with a standard sieve and 117.6 g of HST (hereafter referred to as HST) was obtained (yield, 14.7%).

Preparation of Samples and Standard Solutions
Sample solutions contained HST (40.0 mg) dissolved in distilled water at a concentration of 20.0 mg/mL, and then sonicated for 10 min at room temperature and filtered through a 0.5 µm polytetrafluoroethylene (PTFE) syringe filter (13JP050AN, Advantec, Tokyo, Japan) prior to injection.
Standard stock solutions of the four reference standards (Figure 1), berberine, palmatine, geniposide, and paeoniflorin, were prepared at concentrations of 500.0 µg/mL, 500.0 µg/mL, 1000.0 µg/mL, and 1000.0 µg/mL, respectively, in methanol. All solutions were stored in a refrigerator at 4 • C until required for analysis.
Standard stock solutions of the four reference standards (Figure 1), berberine, palmatine, geniposide, and paeoniflorin, were prepared at concentrations of 500.0 μg/mL, 500.0 μg/mL, 1000.0 μg/mL, and 1000.0 μg/mL, respectively, in methanol. All solutions were stored in a refrigerator at 4 °C until required for analysis.

High-Performance Liquid Chromatography-Photodiode Array (HPLC-PDA) Equipment and Chromatographic Conditions
Simultaneous analysis of the berberine, palmatine, geniposide, and paeoniflorin in HST was performed using a Waters Alliance system (Waters Corporation, Milford, MA 01757, USA) consisting of an e 2695 separation module and a 2998 photodiode array (PDA) detector. All chromatographic data were recorded and analyzed using Empower three chromatography data software. The four components were separated using a reverse-phase SunFire ® C18 column (100 Å, 4.6 × 250 mm ID, 5 μm particle size, Waters), which was maintained at 28 ± 2 ℃. The mobile phase consisted of 0.05% (v/v) formic acid in water (A) and methanol (B); solvents were degassed prior to analysis. The optimum gradient elution system used was as follows: 0-30 min, 10-50% (B); 30-35 min, 50-90% (B); 35-45 min, 90-90% (B); and 45-46 min, 90-10% (B). The flow rate and injection volume used were 1.0 mL/min and 10.0 μL, and detection wavelengths for quantification and analysis were set at 230 nm and 240 nm.

Method Validation
The HPLC analytical method devised was evaluated by determining its specificity, linearity, limit of detection (LOD), limit of quantitation (LOQ), precision, and recovery performance as detailed by the International Conference on Harmonisation (ICH) guidelines [35].

Specificity
The specificity was determined by comparing chromatograms and PDA spectral patterns (λ = 200-400 nm) obtained from standard mixtures and HST samples.

High-Performance Liquid Chromatography-Photodiode Array (HPLC-PDA) Equipment and Chromatographic Conditions
Simultaneous analysis of the berberine, palmatine, geniposide, and paeoniflorin in HST was performed using a Waters Alliance system (Waters Corporation, Milford, MA 01757, USA) consisting of an e 2695 separation module and a 2998 photodiode array (PDA) detector. All chromatographic data were recorded and analyzed using Empower three chromatography data software. The four components were separated using a reverse-phase SunFire ® C18 column (100 Å, 4.6 × 250 mm ID, 5 µm particle size, Waters), which was maintained at 28 ± 2°C. The mobile phase consisted of 0.05% (v/v) formic acid in water (A) and methanol (B); solvents were degassed prior to analysis. The optimum gradient elution system used was as follows: 0-30 min

Method Validation
The HPLC analytical method devised was evaluated by determining its specificity, linearity, limit of detection (LOD), limit of quantitation (LOQ), precision, and recovery performance as detailed by the International Conference on Harmonisation (ICH) guidelines [35].

Specificity
The specificity was determined by comparing chromatograms and PDA spectral patterns (λ = 200-400 nm) obtained from standard mixtures and HST samples.

Precision
The precision of the developed method was determined by inter-day and intra-day testing at three different concentrations. Intra-day precision was measured by performing analyses three times per day, and inter-day precision was measured by repeating analyses three times per day over three consecutive days. Precisions were determined by calculating relative standard deviation (RSD) using: RSD (%) = standard deviation (SD) mean measured amount × 100

Recovery
Recovery testing was performed to verify the accuracy of the optimized HPLC-PDA method. Recovery was evaluated by spiking standard solution and HST samples with different amounts of berberine, palmatine, geniposide, and paeoniflorin. All analyses were performed in triplicate. Recovery was calculated as follows: Recovery (%) = (found amount − original amount) spiked amount × 100

Optimization of HPLC-PDA Conditions
Optimal HPLC-PDA conditions were obtained by optimizing column type, column temperature, and mobile phase variables. For efficient separation of peaks, a SunFire ® C 18 column (5 µm, 100 Å, 4.6 mm × 250 mm, Waters) was used at a column temperature of 28 ± 2°C with a water (0.05% formic acid)/methanol gradient elution system. The maximum ultraviolet (UV) absorption wavelengths of the four components were confirmed using PDA UV spectra (λ = 200-400 nm), and wavelengths of 230, 230, 240, and 230 nm were selected for berberine, palmatine, geniposide, and paeoniflorin, respectively ( Figure 3).

Precision
Precision was determined by intra-day and inter-day analysis. The intra-day and inter-day RSDs of the four compounds ranged from 0.111% to 2.857% and from 0.946% to 3.593%, respectively. Accuracy values ranged from 96.74% to 102.53% and from 97.95% to 100.83%, respectively ( Table 2). The intra-day and inter-day RSDs of berberine, palmatine, geniposide, and paeoniflorin were all <4%. Thus, the developed method was found to be both reliable and reproducible.

Recovery
Recovery testing was performed in triplicate using the method developed by spiking HST samples with reference standards. The average recoveries of berberine, palmatine, geniposide, and paeoniflorin ranged from 92.33% to 116.72% and RSD values from 0.731% to 4.705% (Table 3). Therefore, the developed assay method was found to be suitable for analyzing levels of the four compounds in HST samples.

Quantitative Analysis of Four Marker Compounds in HST
The HPLC-PDA based method showed that HST contained 2.11 mg/g of berberine, 1.54 mg/g of palmatine, 10.77 mg/g of geniposide, and 7.04 mg/g paeoniflorin, and all RSD values were <3% (Table 4). In other words, geniposide, the major compound in Gardenia jasminoides, was the most abundant, followed by paeoniflorin, the major compound in Paeonia lactiflora.

Conclusions
Summarizing, we developed and verified a HPLC-PDA-based analytical method for the simultaneous determination of berberine, palmatine, geniposide, and paeoniflorin in HST. The optimized method utilized a 0.05% formic acid in water/methanol gradient elution system and a reverse-phase C 18 column. Validation of the optimized analytical method was performed by determining specificities, linearities, limits of detection (LODs), and limits of quantification (LOQs) and using precision and recovery tests. In addition, we confirmed that geniposide (the marker compound of Gardenia jasminoides) is abundant in HST. We hope that the method developed and data acquired during the course of this study will be found useful for efficient quality control and standardization of HST.