2789
J. Sep. Sci. 2015, 38, 2789–2796
Weiwei Guo1 Yun Jiang2 ∗ Xiaoqing Chen1 Ping Yu1 Meng Wang1 Xia Wu1 Dayong Zhang3 1 Beijing
Key Lab of TCM Collateral Disease Theory Research (School of Traditional Chinese Medicine, Capital Medical University), Beijing, China 2 Institute of Chinese Medical Sciences, University of Macau, Macau, China 3 Sichuan New Lotus Traditional Chinese Herb Limited Company, Chengdu, China Received March 19, 2015 Revised April 23, 2015 Accepted May 19, 2015
Research Article
Identification and quantitation of major phenolic compounds from penthorum chinense pursh. by HPLC with tandem mass spectrometry and HPLC with diode array detection Penthorum chinense Pursh. is a traditional Chinese herbal medicine used for the treatment of various ailments specially related to liver. Gansu Granule, the medicine made from the extract of P. chinense, has been widely used in the clinical setting. But the information about its active ingredients is lacking. In this paper, the extract of P. chinense was analyzed by high performance liquid chromatography with electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry. Among the 27 compounds that were identified based on their mass spectrometry data, ten were reported for the first time from P. chinense. Chromatographic fingerprints generated by high-performance liquid chromatography by analyzing 21 batches of P. chinense, displayed six common peaks. Finally, four major compounds were identified namely; gallic acid, brevifolin carboxylic acid, 2,6-dihydroxyacetophenone-4-O--D-glucoside, and pinocembrin-7-O--D-glucoside. The average content of each compound was 24.58, 109.6, 15.52, and 18.81 mg/g, respectively. In addition, this study also suggests that the qualitative liquid chromatography with mass spectrometry and the quantitative high-performance liquid chromatography analytical methods using monolithic columns are simple, rapid, accurate, and reproducible and have the potential to be used for the comprehensive quality control of P. chinense. Keywords: Chromatographic fingerprints / Liquid chromatography / Mass spectrometry / Penthorum chinense Pursh. / Phenolic Compounds / Quantification DOI 10.1002/jssc.201500303
1 Introduction Penthoru chinense Pursh. (P. chinense) is a traditional Chinese herbal medicine, which is frequently used to treat jaundice, edema, amenorrhea, metrorrhagia, leucorrhea, bruises, and some other ailments. People usually drink the tea brewed the leaf of P. chinense in Gulin County of Sichuan Province where the incidence rate of liver disease is very low [1]. Gansu Granule, made from P. chinense extract, is widely used in the clinic due to its curative effects on various ailments of liver [2]. Earlier reports on chemical composition of P. chinense have revealed that it contains many active components including alkaloids, flavonoids, volatile oils, and glycosides [3]. In our previous study, we have isolated and identified 13 compounds from the aerial parts of P. chinense, including a new flavanone, 5-methoxy-pinocembrin-7-O--D-glucoside [4]. The
Correspondence: Prof. Xia Wu, School of Traditional Chinese Medicine, Capital Medical University, 10 Xitoutiao, Youanmen, Beijing 100069, China E-mail: [email protected] Fax: +86-10-8391-1627
Abbreviations: DAD, diode array detection; HHDP, hexahydroxydiphenoyl C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
emergence of highly sensitive HPLC–ESI-MS/MS methodology has allowed the possibility of detecting some additional compounds even if they are in trace amounts [5]. In previous reports, quercetin, gallic acid, quercitrin, pinocembrin7-O--D-glucoside, and -sitosterol have been reported as index components for quantification by HPLC methods [6–9]. However, in our HPLC analysis, we find low peak heights for quercetin and quercitrin. But gallic acid, brevifolin carboxylic acid, 2,6-dihydroxyacetophenone-4-O--D-glucoside, and pinocembrin-7-O--D-glucoside displayed large peak areas. Among these four compounds, pinocembrin, and aglycone of pinocembrin-7-O--D-glucoside has been reported to exhibit various biological activities, like antimicrobial, antiinflammatory, antioxidant, and antihepatocarcinoma activities as well as neuroprotective effects [10, 11]. In addition, brevifolin carboxylic acid can induce apoptosis in A549 human lung cancer cells [12]. Therefore, due to the important beneficial effects of these compounds derived from P. chinense, there is an absolute requirement to establish the ∗ Additional correspondence: Dr.Yun Jiang, Institute of Chinese Medical Sciences, University of Macau, Av. Padre Tom´as Pereira, Taipa, Macau, China Email: [email protected] Fax: +853-2884-1358.
www.jss-journal.com
2790
J. Sep. Sci. 2015, 38, 2789–2796
W. Guo et al.
quality standards to meet the QC consistency and efficacy of the herbal medicine. In this study, 27 compounds were identified in the P. chinense extract by HPLC combined with ESI Fourier transform ion cyclotron resonance MS (HPLC–ESI-FTICRMS/MS) method. The four main compounds mentioned above were quantified by HPLC with diode array detection (DAD). Also fingerprints were generated by using a monolithic column for the first time, which can get rapid separation by increasing flow rate [13, 14]. The results thus promote the QC of P. chinense and will be a guiding principle for further research.
2 Materials and methods 2.1 Chemicals and reagents Gallic acid (1) was purchased from National Institutes for Food and Drug Control (Purity = 90.1%). Brevifolin carboxylic acid (4), 2,6-dihydroxyacetophenone-4-O--Dglucoside (5) and pinocembrin-7-O--D-glucoside (17) were isolated by the Chemistry department of Chinese Material Medicine Laboratory at the Capital Medical University (Beijing, China). The purities were determined to be >98% by HPLC–UV analysis. HPLC-grade acetonitrile and methanol were obtained from Fisher (NJ, USA). Ultra-pure water was prepared by Thermo Scientific Barnstead Pure Water Systems (D11931, MA, USA). All the other solvents were purchased from Beijing Chemical Factory (Beijing, China).
2.2 Sample preparation Twenty-one batches of P. chinense extract were provided by Sichuan Longlife Pharmaceutical, Chengdu, Sichuan, in January 2013 and October 2014. All the samples of P. chinense were authenticated by Zhang Li, Senior Engineer of this company, and voucher specimens (P1-P21) were deposited in the School of Traditional Chinese Medicine, Capital Medical University. The aerial parts of P. chinense herb (10 kg) were cut into pieces and decocted with water three times for 2 h each. After filtration, the filtrate was evaporated to yield a clear paste (relative density 1.15–1.18; 60–65⬚C). The paste was cooled and 60% ethanol was added. After stirring, standing, and filtering, the precipitate was washed three times with ethanol. The washing liquid and the filtrate were combined and evaporated to yield an extract (relative density 1.30–1.32) [4]. The extracts were dried in a vacuum drying oven at 60⬚C for 4 h. Then 50 mg of the dried extract was dissolved in 10 mL of 60% methanol. Through ultrasonic treatment for 10 min at 25⬚C, then the sample was brought to room temperature and finally 60% methanol was added to compensate for the loss of weight. All the sample solutions were filtered through a 0.45 m membrane and stored at 4⬚C and brought to room temperature before analysis. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2.3 Standard solutions preparation A stock solution containing four standards (gallic acid, brevifolin carboxylic acid, 2,6-dihydroxyacetophenone-4-O--Dglucoside, and pinocembrin-7-O--D-glucoside) was prepared by dissolving them in 60% methanol at concentration of 900, 600, 440, and 600 g/mL, respectively. Each stock standard solution was diluted and taken at the correct volume to make a standard mixture that was used to establish the calibration curves. All solutions were filtered by 0.45 m membrane filter before use. 2.4 HPLC–ESI-FTICR-MS/MS conditions LC–MS analyses were carried out on an Agilent series 1260 HPLC instrument (Agilent, Waldbronn, Germany) coupled with a solarix FTICR mass spectrometer (Bruker Daltonik, MA, USA) consisting of an ESI/MALDI dualion source. The sample was separated on a Hibar250–4, 6 column (4.6 × 250 mm, 5 m). The mobile phase was acetonitrile (B) and 0.2% v/v formic acid in water (A) using a step gradient elution of 10–50% B at 0–65 min, 50–10% B at 65–70 min. The analysis was performed at 25⬚C with a flow rate of 1 mL/min. The analytes were determined in the positive ion mode. The operation parameters were as follows: Dry gas flow rate, 4 L/min; Dry gas temperature, 180⬚C; Nebulizer, 1.0 bar; capillary voltage, 4 kV and end plate offset voltage, –500 V. The data acquisition software was Ftms Control 2.0 and data process software was Data Analysis 4.1.
2.5 HPLC–DAD conditions HPLC analysis was performed on an Agilent series 1260 HPLC instrument (Agilent, Germany) consisting of a quaternary pump, a DAD, an autosampler and a column compartment. Chromatographic separations were performed on a Merck Chromolith Perfomance RP-C18 column (4.6 × 100 mm, 2 m, Merck). The mobile phase consisted of acetonitrile (B) and 0.1% v/v phosphoric acid in water (A) with a gradient of 5–12% B for 0–5 min, 12–40% B for 5–20 min, and 40–50% B for 20–30 min. The flow rate was 2 mL/min and column temperature was set to 25⬚C. The detection wavelength was 280 nm. An aliquot of 2 L was injected for HPLC. The data were processed with Agilent Technologies ChemStation Revision B.04.03 software. 2.6 Chromatogram similarity analysis The HPLC chromatograms from 21 different batches of P. chinense were analyzed by “Similarity Evaluation System for Chromatographic Fingerprint of Traditional Chinese Medicine” software (Version 2004A, Beijing, China) [15] composed by Chinese Pharmacopoeia Commission and recommended by China Food and Drug Administration (CFDA). The standardized HPLC fingerprint of P. chinense was www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2015, 38, 2789–2796
2791
Table 1. Identification and speculation of phenolic compounds in P. chinense extract by LC–MS
Noa)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
Proposed compound
Gallic acid Ethyl gallate Epicatechin Brevifolin-carboxylic acid 2,6-dihydroxyacetophenone-4-O-β-D-glucoside Quercetin di-O-glycoside Rutin Kaempferol-3-O-rutinoside Quercetin-3-O-xyloside Quercetin-3-O-arabinopyranoside Quercetin-3-O-rhamnoside Quercetin Quercetin-3 -O-rhamnoside 5-methoxy-pinocembrin-7-O-glucoside Kaempferol Kaempferol-3-O-rhamnoside Pinocembrin-7-O-glucoside Pinocembrin-7-O-galloyl-D-glucoside isomer
Pinocembrin-7-O-[4 ,6 -HHDP]-glucoside Pinocembrin dihydrochalcone-7-O-[4 ,6 - HHDP]-glucoside Pinocembrin-7-O-[3 -O-galloyl-4 ,6 -HHDP]-glucoside Pinocembrin dihydrochalcone-7-O-[3 -O- galloyl-4 ,6 -HHDP]-glucoside Pinocembrin
tR (min)
4.84 6.28 11.29 12.08 16.68 17.26 21.16 24.41 25.69 25.76 26.34 26.36 26.40 29.46 29.74 30.63 39.76 41.38 42.16 43.49 47.38 48.40 49.83 55.55 55.84 60.90 62.03
[M+H]+
MS/MS (m–z)
Detected
Expected
Error (ppm)
171.02876 199.06000 291.08632 293.02902 331.10208 627.15426 611.15965 595.16476 435.09174 435.09168 449.10722 303.04975 449.10722 433.14877 287.05481 433.11247 419.13316 571.14363 571.14357 571.14370 571.14361 571.14365 721.13813 723.15326 873.14865 875.16428 257.08073
171.02880 199.06010 291.08631 293.02919 331.10236 627.15558 611.16066 595.16575 435.09219 435.09219 449.10784 303.04993 449.10784 433.14931 287.05501 433.11292 419.13366 571.14462 571.14462 571.14462 571.14462 571.14462 721.13993 723.15558 873.15088 875.16654 257.08084
0.3 0.5 0.6 0.6 0.8 2.1 1.7 1.7 1.0 1.2 1.2 0.6 1.4 1.2 0.7 1.1 1.2 1.7 1.8 1.6 1.8 1.7 2.4 2.9 2.6 2.5 0.4
247, 219, 191, 163 169, 151 535, 303 303 287 303, 257, 229 303, 285, 257, 229 303, 285, 257, 229 285, 257, 229 303, 285, 257, 229 271, 229, 167 287 257, 215, 153 257, 153 257, 153 153 257, 153 257, 153 303, 277, 257, 153 303, 277, 259 493, 303, 257, 153 303, 153 153
a) The notation for peaks refers to Fig. 1.
generated and the similarity indices of these batches were calculated.
2.7 Validation of HPLC method
and stability. The recovery was carried out by spiking known amounts of mixed standard solutions into samples at three concentration levels (low, medium, and high). The average recoveries were calculated by the formula: recovery (%) = (amount found−original amount)/amount spiked × 100.
2.7.1 Calibration curves, LOD, and LOQ
3 Results and discussion The calibration curves were constructed by plotting the peak areas (y axis) versus the concentration (x axis, g/mL) of each standard solution, by analyzing a series of dilute solution with seven concentrations. The LOD and LOQ under the present chromatographic conditions were determined at S/N = 3 and 10, respectively. 2.7.2 Precision, repeatability, stability, and recovery The precision of HPLC was tested by analyzing the same sample (P1) six times. Six working solutions prepared from the same extract (P1) were analyzed to examine the repeatability. The stability was established by analyzing the same sample (P1) at 0, 2, 4, 8, 16, and 24 h. The RSD (%) was taken as a measure to express the precision, repeatability, C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3.1 Optimization HPLC–DAD conditions To achieve a better chromatographic behavior for HPLC fingerprints and determination, different mobile phases, column temperatures and column temperatures were optimized. The combination of acetonitrile and 0.1% v/v formic acid aqueous solution, methanol and 0.1% v/v formic acid aqueous solution, acetonitrile and 0.1% v/v phosphoric acid aqueous solution were investigated, the last one was selected as a mobile phase to reach a satisfactory peak shape. Meanwhile, different column temperatures of 20, 25, and 30⬚C were compared. The results showed that the column temperatures had little influence on separation of the investigated compounds. At 25⬚C, baseline separation of the www.jss-journal.com
2792
W. Guo et al.
analytes could be achieved. To achieve good accuracy and sensitivity for quantification, different detection wavelengths of 245, 280, 301, and 365 nm were investigated by using the DAD. As a result, at 280 nm, maximum UV absorption for each analyte was selected as detection wavelength.
3.2 LC–MS analysis for identification A total of 27 compounds were detected from extract of P. chinense (Table 1 and Fig. 1, m/z error ࣘ 5 ppm). Among them, 13 compounds (1, 4, 5, 7, 9, 12–15, 17, 23, 24, and 27) were identified by comparing with reference substances, and 14 other compounds (2, 3, 6, 8, 10, 11, 16, 18–22, 25, and 26) were tentatively identified based on their MS and MS2 spectra (Fig. 2A). Ten of these compounds (6, 8, 10, 18–24) are reported for the first time from P. chinense. 3.2.1 Identification of phenolic acids Compounds 1, 2, and 3 eluting at 4.84, 6.28 and 11.29 min, respectively, showed [M+H]+ at m/z 171.02876, 199.06000 and 291.08632 respectively. The three compounds matched the elemental composition of C14 H14 O10 with an accuracy of 0.25 ppm, C18 H22 O10 with 0.48 ppm and C15 H14 O6 with 0.55 ppm. They were identified as gallic acid, ethyl gallate and epicatechin, respectively [16, 17]. Compound 4 at m/z 293.02902 (C13 H8 O8 ) showed fragment ions at m/z 247 ([M+H–COOH]+ ), 219 (C11 H7 O6 ), 191 (C10 H7 O4 ) and 177 (C9 H5 O4 ). It was identified as brevifolin-carboxylic acid [18]. For compound 5, the precursor m/z 331.10208 (C14 H19 O9 ) yielded fragments at m/z 169 ([M+H–C6 H10 O5 ]+ ) and 151 ([M+H–C6 H10 O5 –H2 O]+ ), indicating the loss of a glucose residue and consecutive dehydration. It was identified as 2,6dihydroxyacetophenone-4-O--D-flucoside [19]. 3.2.2 Identification of quercetin derivatives Compound 12 produced a MS/MS spectrum [M+H]+ of m/z 303.04975 (C15 H11 O7 ) and fragmented to yield m/z 285.03924 [M+H–H2 O]+ , 257.04440 [M+H–CO–H2 O]+ and 229.04935 [M+H–2CO–H2 O]+ , which was identified to be quercetin [20]. Similarly, a number of derivatives of quercetin were detected by analyzing the characteristic fragmentation patterns of compounds 6, 7, 9, 10, 11, and 13 at m/z 303. Compound 6 yielded [M+H]+ at m/z 627.15426, generating [M+H]+ ion at m/z 303.04962, indicating loss of two glucose moiety (m/z 162, C6 H10 O5 ). Thus, it was tentatively identified as quercetin di-O-glycoside [21]. Compound 7 was identified as rutin because it has the same retention time and UV spectrum as rutin, and its [M+H]+ at m/z 611.15965 indicated the loss of a disaccharide moiety (C12 H20 O9 ) to generate the fragment at m/z 303.04960 [22]. Both compounds 9 and 10 yielded [M+H]+ ion at m/z 435 (C20 H19 O11 ). The presence of m/z at 303 [M+H–132]+ of compounds 9 and 10 indicated the loss of a xylose or arabinose moiety (C5 H8 O4 ). Hence, based on the previous studies [20], compound 9 was identified as quercetin-3-O-xyloside C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2015, 38, 2789–2796
and compound 10 was tentatively identified as quercetin3-O-arabinopyranoside. For both compounds 11 (ions at m/z 449.10728) and 13 (ions at m/z 449.10722), molecular formulas of C21 H21 O11 were predicted. In MS/MS, both compounds further produced the characteristic fragments at m/z 303 ([M+H–C6 H10 O4 ]+ , C15 H11 O7 ), 285 (C15 H9 O6 ), 257 (C14 H9 O5 ) and 229 (C13 H9 O4 ), which were very similar to quercetin. Therefore, compound 11 was tentatively identified as quercetin-3-O-rhamnoside and compound 13 as quercetin-3 -O-rhamnoside [4, 23]. 3.2.3 Identification of kaempferol derivatives The EIC of m/z 287 (C15 H11 O6 ) in positive ion mode exhibited three peaks (8, 15, and 16). Compound 15 yielded [M+H]+ ion at m/z 287.05481 and was identified as kaempferol according to previous report [24]. Compounds 8 and 16 are derivatives of compound 15 and the ion of compound 8 (m/z 595.16476) consecutively lost a disaccharide residue to produce the peak of kaempferol 287.05481 [M+H–C12 H20 O9 ]+ . It was thus, tentatively identified as kaempferol-3-O-rutinoside that was reported in P. chinense herb for the first time. Compound 16 was identified as kaempferol-3-O-rhamnoside because its [M+H]+ ion was at m/z 433.14877 and it produced m/z 287.05479 [M+H–C6 H10 O4 ]+ by elimination of a rhamnose residue [24]. 3.2.4 Identification of pinocembrin derivatives Compound 27 was identified as pinocembrin and gave [M+H]+ at m/z 257.08073, which produced 153.01808 (C7 H5 O4 ) [25]. Compound 17 yielded [M+H]+ at m/z 419.13316 (C21 H23 O9 ) and also consisted of m/z 257.08069 [M+H–C6 H10 O5 ]+ and 153.01815. It was identified as pinocembrin-7-O-glucoside that has been reported to be the main component of P. chinense [4]. Compound 14 yielded [M+H]+ ion at m/z 433.14877 (C22 H25 O9 ) and subsequently produced fragments at m/z 271 and 167, thus indicating the substitution of hydrogen by methyl group (CH3 ). So compound 14 was identified as 5-methoxy-pinocembrin-7O-glucoside and this is a new compound isolated from P. chinense in our laboratory [4]. In addition, five [M+H]+ ions at m/z 571 (C28 H27 O13 ) were detected from 40 to 50 min in full-scan mass spectra. Among these, four compounds produced the fragments at m/z 257 and 153, and hold the same fragment composition. Another one compound only produced ion at m/z 153. The presence of m/z 257 indicated the loss of a galloyl residue (C7 H4 O4 ) and a glucose moiety (C6 H10 O5 ). Thus, compounds 18–22 were tentatively identified as isomers of pinocembrin-O-galloyl-glucoside. Compounds 23 and 24 were identified as pinocembrin-7O-[4 ,6 -hexahydroxydiphenoyl]-glucoside and pinocembrin dihydrochalcone-7-O-[4 ,6 -hexahydroxydiphenoyl]-glucoside (Thonningianins B) respectively and yielded ions at m/z 721.13813 and 723.15326 [26]. The characteristic fragments of compounds 23 (m/z 257.08072, m/z 153.01822) and 24 (m/z 259.09680) indicated the loss of a glucose moiety (C6 H10 O5 ) and a hexahydroxydiphenoyl (HHDP) residue, www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2015, 38, 2789–2796
2793
Figure 1. Chemical structures of 27 compounds identified from P. chinense extract. Table 2. Regression equations, linear range, LOD, and LOQ of investigated compounds by HPLC
Analytes
Regression equationa)
Linear range (g/ml)
r2
LOD (ng)b)
LOQ (ng)c)
Gallic acid (1) Brevifolin carboxylic acid (4) 2,6-Dihydroxyacetophenone-4-O-β–D-glucoside (5) Pinocembrin-7-O-β-D-glucoside (17)
y = 2.6186x–6.3521 y = 0.4606x–5.9059 y = 2.1282x–2.2826 y = 2.0865x–1.1575
9.0-900.0 6.0-600.0 4.4-440.0 6.0-600.0
0.9999 0.9996 0.9999 0.9999
0.45 1.50 0.88 0.60
1.35 6.00 2.20 3.00
a) y represents peak area, x represents concentration of compounds per injection volume (g/mL). b) LOD represents limit of detection (S/N = 3). c) LOQ represents limit of quantification (S/N = 10). Table 3. Precision, repeatability, and stability of compounds 1, 4, 5, and 17
Analytes
Gallic acid (1) Brevifolin carboxylic acid (4) 2,6-Dihydroxyacetophenone-4-O-β-D-glucoside (5) Pinocembrin-7-O-β-D-glucoside (17)
Precision (n = 6)
Repeatability (n = 6)
Stability (n = 6)
Mean(mg/g)
RSD(%)a)
Mean (mg/g)
RSD (%)a)
Mean (mg/g)
RSD (%)a)
27.42 108.75 15.69 22.19
0.30 0.62 0.64 1.12
26.99 105.86 15.38 21.60
2.13 1.19 1.78 2.51
27.50 108.89 15.73 22.37
0.45 0.75 0.81 1.53
a) RSD(%) = (SD/mean) × 100%.
which were consistent with that of their reference substance. Compound 25 and 26 yielded [M+H]+ ion at m/z 873.14865 (C42 H33 O21 ) and 875.16428 (C42 H35 O21 ). They have common fragments at m/z 303 and 153, thus indicating the loss of the HHDP and galloyl moiety. Therefore, compounds 25 and 26 were tentatively identified as pinocembrin-7-
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
O-[3 -O-galloyl-4 ,6 -hexahydroxydiphenoyl]-glucoside and pinocembrin dihydrochalcone-7-O-[3 -O-galloyl-4 ,6 -hexahydroxydiphenoyl]- glucoside (Thonningianins A), which have been reported as active components of P. chinense herb [26, 27]. Compounds 27, 17, 23, and 25 had the same parent nucleus (pinocembrin), which in turn, linked with
www.jss-journal.com
2794
W. Guo et al.
J. Sep. Sci. 2015, 38, 2789–2796
Figure 2. The HPLC–UV chromatogram of P. chinense extract (A) and MS2 spectra of 27, 17, 23, and 25 in positive ion modes (B).
Figure 3. HPLC chromatograms of 21 batches of P. chinense extracts (A) and common peaks identified on reference spectrum (B): (1) gallic acid; (4) brevifolin carboxylic acid; (5) 2,6-dihydroxyacetophenone-4-O--D-glucoside; (17) pinocembrin-7-O--D-glucoside; (25) pinocembrin-7-O-[3 -O-galloyl-4 , 6 -hexahydroxydiphenoyl]-glucoside; (26) pinocembrin dihydrochalcone-7-O-[3 -Ogalloyl-4 ,6 -hexahydroxydiphenoyl]glucoside.
different residues of glucosyl, galloyl, and HHDP. The characteristic spectra of these four compounds were shown in Fig. 2B.
3.3 Method validation The calibration curves of four analytes showed good linearity (r2 > 0.9995) within the test ranges. The LOD and LOQ for each compound was less than 1.5 and 6 ng (Table 2). As summarized in Table 3, the precision (RSD) of four analytes was found to be in the range of 0.30–1.12%. The RSD values of repeatability ranged from 1.19% to 2.51%, which suggests high C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
repeatability. All the analytes showed good stability within 24 h (RSD values of four compounds < 2%). The four compounds showed good recovery rates in the range of 97.45– 103.6%. The results indicate that the method was suitable for the simultaneous determination of four major compounds in P. chinense.
3.4 HPLC fingerprint and quantitation analysis In our study, HPLC chromatograms from 21 batches of P. chinense extract were submitted to the Similarity Evaluation www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2015, 38, 2789–2796
2795
Table 4. The contents (mg/g) of compounds 1, 4, 5, and 17 in 21 samples of P. chinense analyzed by HPLC
Numbers
P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 P13 P14 P15 P16 P17 P18 P19 P20 P21 Mean
Contents (mg/g)a)
Similarity index
1
4
5
17
27.43 26.82 29.40 28.65 26.14 26.63 26.20 25.99 26.23 25.36 25.08 24.55 19.20 18.38 21.92 21.49 22.78 22.49 23.66 23.74 24.01 24.58
107.78 91.41 96.40 74.73 94.71 84.37 97.07 102.96 90.93 84.07 101.39 96.98 98.51 146.39 135.87 132.07 113.35 135.74 115.83 150.71 150.36 109.60
15.52 15.85 16.27 14.83 15.47 14.58 16.30 16.77 15.88 14.66 16.96 16.33 14.54 19.62 15.35 15.47 13.15 14.75 12.84 15.45 15.25 15.52
22.35 21.67 22.41 20.61 20.54 17.46 19.55 18.59 19.37 18.33 18.60 17.85 16.89 18.52 17.25 17.63 16.21 14.78 16.64 19.79 19.88 18.81
0.990 0.990 0.980 0.970 0.986 0.988 0.995 0.990 0.990 0.990 0.986 0.986 0.989 0.962 0.985 0.986 0.987 0.978 0.988 0.984 0.985
a) The notation for analytes refers to Fig. 1.
System for Chromatographic Fingerprint of TCM software, to generate standardized fingerprints (Fig. 3A) [28, 29]. Each sample was processed with a monolithic column within 30 min; otherwise it usually takes more than 60 min with a common column (Fig. 2). The monolithic column results in a rapid analysis instead of UHPLC, which was first used to analyze P. chinense extract. A total of six peaks were defined as common among all the observed peaks (Fig. 3B). Moreover, the similarity index of each batch was not less than 0.96 in comparison to the reference fingerprint. This revealed that different samples had similar and consistent chromatographic patterns. The developed HPLC method was used for simultaneous quantitation of four compounds in extracts from 21 batches of P. chinense. The contents of four major compounds from all batches are shown in Table 4. The average content of gallic acid (1), brevifolin carboxylic acid (4), 2,6-dihydroxyacetophenone-4-O--D-glucoside (5) and pinocembrin-7-O--D-glucoside (17) in P. chinense are 24.58, 109.6, 15.52, and 18.81 mg/g, respectively. The content of brevifolin carboxylic acid was observed to be the highest (average 109.60 mg/g), but it varied significantly from 74.73 to 150.71 mg/g.
the 27 compounds identified from this plant by the highly sensitive LC–MS method, ten have been reported for the first time. Thus, the LC–MS method can be used to identify compounds of P. chinense with high sensitivity. In addition, HPLC fingerprints of the 21 batches showed high similarity and displayed six common peaks. Moreover, the contents of four major compounds were simultaneously determined by HPLC. The average contents of gallic acid, brevifolin carboxylic acid, 2,6-dihydroxyacetophenone-4-O--D-glucoside, and pinocembrin-7-O--D-glucoside in P. chinense were 24.58, 109.6, 15.52, and 18.81 mg/g, respectively. The established HPLC method using a monolithic column is simple, rapid, and effective, which can be used for the comprehensive QC of P. chinense. The authors have declared no conflict of interest.
5 References [1] Xiang, H. H., Yang, J. H., Fu, X.C., Zhang, Y. Y., Wan, H. T., J. Exp. Tradit. Med. Form. 2013, 19, 50–52. [2] Chen, X. R., Yao, H., Jiang, Y., Wu, S. M., Zhu, X. F., Zhou, X. Q., Wang, L. T., Chin. J. Hepatol. 2004, 12, 50.
4 Concluding remarks
[3] Zeng, Q. H., Zhang, X. W., Xu, X. L., Jiang, M. H., Xu, K. P., Piao, J. H., Zhu, L., Chen, J., Jiang, J. G., Food Funct. 2013, 4, 1811–1818.
A simple, fast, and reliable method for the chemical analysis of P. chinense extract was developed in our study. Out of
[4] Wang, M., Jiang, Y., Liu, H. L., Chen, X. Q., Wu, X., Zhang, D. Y., Nat. Prod. Res. 2014, 28, 70–73.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
2796
W. Guo et al.
J. Sep. Sci. 2015, 38, 2789–2796
[5] Xiao, X. Y., Sun, B. H., Ye, X. T., Zhang, G. Y., Hou, P. Y., Gao, H. Y., J. Sep. Sci. 2014, 37, 1533–1545.
[18] Chen, Y. W., Ren, L. J., Chin. Tradit. Herb Drug 1997, 28, 198–202.
[6] Feng, C. G., Wang, H. W., Ren, Q. S., Chin. Pharm. J. 2004, 39, 97–98.
[19] Zhang, X., Yang, M., J. Chengdu Univ. Tradit. Chin. Med. 2002, 25, 46.
[7] He, X. H., Xu, L., Du, F. L., Zeng, J. G., Cent. South. Pharm. 2008, 6, 717–719.
[20] Zhang, M., Pan, D. R., Zhou, Y. F., Zhu, Q. Q., Wang, S. M., J. Anhui Agri. Sci. 2012, 40, 5211–5214.
[8] Du, X., Tang, B., Zhang, Q. B., J. Environ. Health 2011, 28, 358–359.
[21] Qiao, X., He, W. N., Xian, C., Han, J., Wu, L. J., Guo, D. A., Ye, M., Phytochem. Anal. 2011, 22, 475–483.
´ D. C., [9] Stashenko, E. E., Mart´ınez, J. R., Cala, M. P., Duran, Caballero, D., J. Sep. Sci. 2013, 36, 192–202.
[22] Wang, Z. W., Master’s Thesis of Shenyang Pharmaceutical University 2005.
[10] Rasul, A., Millimouno, F. M., Eltayb, W. A., Ali, M., Li, J., Li, X. M., Biomed. Res. Int. 2013, 2013, 379850.
[23] Wang, M., Wu, X., Jiang, Y., Zhang, D. Y., Food Drug 2013, 15, 202–205.
[11] Kapoor, S., Agric, J., Food Chem. 2013, 61, 1416. [12] Lou, K. Q., Master’s Thesis of Jinan University 2013.
[24] Fu, M., Wei, L., Yu, J., Hu, Z. T., Chin. Pharm. J. 2013, 48, 1911–1914.
[13] Pellati, F., Calo, S., Benvenuti, S., J. Chromatogr., A 2007, 1149, 56–65.
[25] Li, Y. Y., Shangguan, Y. Y., Chin. J. Mod. Appl. Pharm. 2012, 29, 159–163.
[14] Atia, N. N., York, P., Clark, B. J., J. Sep. Sci. 2009, 32, 931–938.
[26] Ohtani, I. I., Gotoh, N., Tanaka, J., Higa, T., Gyamfi, M. A., Aniya, Y., J. Nat. Prod. 2000, 63, 676–679.
[15] Kang, J., Zhou, L., Sun, J.H., Han, J., Guo, D. A., J. Pharm. Biomed. Anal. 2008, 47, 778–785.
[27] Lu, Q., Jiang, M. H., Jiang, J. G., Zhang, R. F., Zhang, M. W., J. Agric. Food Chem. 2012, 60, 11097–11103.
[16] Wang, Y., Master’s Thesis of Second Military Medical University 2012.
[28] Liu, Y. S., Meng, Q. H., Jiang, S. M., Hu, Y., Chin. J. Chromatogr. B 2005, 23, 158–163.
[17] Fu, F. Y., Shong, T. M., Xu, Z. P., Acta. Pham. Sinic. 1983, 29, 555–557.
[29] Zhou, W., Xie, M. F., Zhang, X. Y., Liu, T. T., Yu, Y. J., Duan, G. L., J. Sep. Sci. 2011, 34, 1123–1132.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
J. Sep. Sci. 2015, 38, 2789–2796
Weiwei Guo1 Yun Jiang2 ∗ Xiaoqing Chen1 Ping Yu1 Meng Wang1 Xia Wu1 Dayong Zhang3 1 Beijing
Key Lab of TCM Collateral Disease Theory Research (School of Traditional Chinese Medicine, Capital Medical University), Beijing, China 2 Institute of Chinese Medical Sciences, University of Macau, Macau, China 3 Sichuan New Lotus Traditional Chinese Herb Limited Company, Chengdu, China Received March 19, 2015 Revised April 23, 2015 Accepted May 19, 2015
Research Article
Identification and quantitation of major phenolic compounds from penthorum chinense pursh. by HPLC with tandem mass spectrometry and HPLC with diode array detection Penthorum chinense Pursh. is a traditional Chinese herbal medicine used for the treatment of various ailments specially related to liver. Gansu Granule, the medicine made from the extract of P. chinense, has been widely used in the clinical setting. But the information about its active ingredients is lacking. In this paper, the extract of P. chinense was analyzed by high performance liquid chromatography with electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry. Among the 27 compounds that were identified based on their mass spectrometry data, ten were reported for the first time from P. chinense. Chromatographic fingerprints generated by high-performance liquid chromatography by analyzing 21 batches of P. chinense, displayed six common peaks. Finally, four major compounds were identified namely; gallic acid, brevifolin carboxylic acid, 2,6-dihydroxyacetophenone-4-O--D-glucoside, and pinocembrin-7-O--D-glucoside. The average content of each compound was 24.58, 109.6, 15.52, and 18.81 mg/g, respectively. In addition, this study also suggests that the qualitative liquid chromatography with mass spectrometry and the quantitative high-performance liquid chromatography analytical methods using monolithic columns are simple, rapid, accurate, and reproducible and have the potential to be used for the comprehensive quality control of P. chinense. Keywords: Chromatographic fingerprints / Liquid chromatography / Mass spectrometry / Penthorum chinense Pursh. / Phenolic Compounds / Quantification DOI 10.1002/jssc.201500303
1 Introduction Penthoru chinense Pursh. (P. chinense) is a traditional Chinese herbal medicine, which is frequently used to treat jaundice, edema, amenorrhea, metrorrhagia, leucorrhea, bruises, and some other ailments. People usually drink the tea brewed the leaf of P. chinense in Gulin County of Sichuan Province where the incidence rate of liver disease is very low [1]. Gansu Granule, made from P. chinense extract, is widely used in the clinic due to its curative effects on various ailments of liver [2]. Earlier reports on chemical composition of P. chinense have revealed that it contains many active components including alkaloids, flavonoids, volatile oils, and glycosides [3]. In our previous study, we have isolated and identified 13 compounds from the aerial parts of P. chinense, including a new flavanone, 5-methoxy-pinocembrin-7-O--D-glucoside [4]. The
Correspondence: Prof. Xia Wu, School of Traditional Chinese Medicine, Capital Medical University, 10 Xitoutiao, Youanmen, Beijing 100069, China E-mail: [email protected] Fax: +86-10-8391-1627
Abbreviations: DAD, diode array detection; HHDP, hexahydroxydiphenoyl C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
emergence of highly sensitive HPLC–ESI-MS/MS methodology has allowed the possibility of detecting some additional compounds even if they are in trace amounts [5]. In previous reports, quercetin, gallic acid, quercitrin, pinocembrin7-O--D-glucoside, and -sitosterol have been reported as index components for quantification by HPLC methods [6–9]. However, in our HPLC analysis, we find low peak heights for quercetin and quercitrin. But gallic acid, brevifolin carboxylic acid, 2,6-dihydroxyacetophenone-4-O--D-glucoside, and pinocembrin-7-O--D-glucoside displayed large peak areas. Among these four compounds, pinocembrin, and aglycone of pinocembrin-7-O--D-glucoside has been reported to exhibit various biological activities, like antimicrobial, antiinflammatory, antioxidant, and antihepatocarcinoma activities as well as neuroprotective effects [10, 11]. In addition, brevifolin carboxylic acid can induce apoptosis in A549 human lung cancer cells [12]. Therefore, due to the important beneficial effects of these compounds derived from P. chinense, there is an absolute requirement to establish the ∗ Additional correspondence: Dr.Yun Jiang, Institute of Chinese Medical Sciences, University of Macau, Av. Padre Tom´as Pereira, Taipa, Macau, China Email: [email protected] Fax: +853-2884-1358.
www.jss-journal.com
2790
J. Sep. Sci. 2015, 38, 2789–2796
W. Guo et al.
quality standards to meet the QC consistency and efficacy of the herbal medicine. In this study, 27 compounds were identified in the P. chinense extract by HPLC combined with ESI Fourier transform ion cyclotron resonance MS (HPLC–ESI-FTICRMS/MS) method. The four main compounds mentioned above were quantified by HPLC with diode array detection (DAD). Also fingerprints were generated by using a monolithic column for the first time, which can get rapid separation by increasing flow rate [13, 14]. The results thus promote the QC of P. chinense and will be a guiding principle for further research.
2 Materials and methods 2.1 Chemicals and reagents Gallic acid (1) was purchased from National Institutes for Food and Drug Control (Purity = 90.1%). Brevifolin carboxylic acid (4), 2,6-dihydroxyacetophenone-4-O--Dglucoside (5) and pinocembrin-7-O--D-glucoside (17) were isolated by the Chemistry department of Chinese Material Medicine Laboratory at the Capital Medical University (Beijing, China). The purities were determined to be >98% by HPLC–UV analysis. HPLC-grade acetonitrile and methanol were obtained from Fisher (NJ, USA). Ultra-pure water was prepared by Thermo Scientific Barnstead Pure Water Systems (D11931, MA, USA). All the other solvents were purchased from Beijing Chemical Factory (Beijing, China).
2.2 Sample preparation Twenty-one batches of P. chinense extract were provided by Sichuan Longlife Pharmaceutical, Chengdu, Sichuan, in January 2013 and October 2014. All the samples of P. chinense were authenticated by Zhang Li, Senior Engineer of this company, and voucher specimens (P1-P21) were deposited in the School of Traditional Chinese Medicine, Capital Medical University. The aerial parts of P. chinense herb (10 kg) were cut into pieces and decocted with water three times for 2 h each. After filtration, the filtrate was evaporated to yield a clear paste (relative density 1.15–1.18; 60–65⬚C). The paste was cooled and 60% ethanol was added. After stirring, standing, and filtering, the precipitate was washed three times with ethanol. The washing liquid and the filtrate were combined and evaporated to yield an extract (relative density 1.30–1.32) [4]. The extracts were dried in a vacuum drying oven at 60⬚C for 4 h. Then 50 mg of the dried extract was dissolved in 10 mL of 60% methanol. Through ultrasonic treatment for 10 min at 25⬚C, then the sample was brought to room temperature and finally 60% methanol was added to compensate for the loss of weight. All the sample solutions were filtered through a 0.45 m membrane and stored at 4⬚C and brought to room temperature before analysis. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2.3 Standard solutions preparation A stock solution containing four standards (gallic acid, brevifolin carboxylic acid, 2,6-dihydroxyacetophenone-4-O--Dglucoside, and pinocembrin-7-O--D-glucoside) was prepared by dissolving them in 60% methanol at concentration of 900, 600, 440, and 600 g/mL, respectively. Each stock standard solution was diluted and taken at the correct volume to make a standard mixture that was used to establish the calibration curves. All solutions were filtered by 0.45 m membrane filter before use. 2.4 HPLC–ESI-FTICR-MS/MS conditions LC–MS analyses were carried out on an Agilent series 1260 HPLC instrument (Agilent, Waldbronn, Germany) coupled with a solarix FTICR mass spectrometer (Bruker Daltonik, MA, USA) consisting of an ESI/MALDI dualion source. The sample was separated on a Hibar250–4, 6 column (4.6 × 250 mm, 5 m). The mobile phase was acetonitrile (B) and 0.2% v/v formic acid in water (A) using a step gradient elution of 10–50% B at 0–65 min, 50–10% B at 65–70 min. The analysis was performed at 25⬚C with a flow rate of 1 mL/min. The analytes were determined in the positive ion mode. The operation parameters were as follows: Dry gas flow rate, 4 L/min; Dry gas temperature, 180⬚C; Nebulizer, 1.0 bar; capillary voltage, 4 kV and end plate offset voltage, –500 V. The data acquisition software was Ftms Control 2.0 and data process software was Data Analysis 4.1.
2.5 HPLC–DAD conditions HPLC analysis was performed on an Agilent series 1260 HPLC instrument (Agilent, Germany) consisting of a quaternary pump, a DAD, an autosampler and a column compartment. Chromatographic separations were performed on a Merck Chromolith Perfomance RP-C18 column (4.6 × 100 mm, 2 m, Merck). The mobile phase consisted of acetonitrile (B) and 0.1% v/v phosphoric acid in water (A) with a gradient of 5–12% B for 0–5 min, 12–40% B for 5–20 min, and 40–50% B for 20–30 min. The flow rate was 2 mL/min and column temperature was set to 25⬚C. The detection wavelength was 280 nm. An aliquot of 2 L was injected for HPLC. The data were processed with Agilent Technologies ChemStation Revision B.04.03 software. 2.6 Chromatogram similarity analysis The HPLC chromatograms from 21 different batches of P. chinense were analyzed by “Similarity Evaluation System for Chromatographic Fingerprint of Traditional Chinese Medicine” software (Version 2004A, Beijing, China) [15] composed by Chinese Pharmacopoeia Commission and recommended by China Food and Drug Administration (CFDA). The standardized HPLC fingerprint of P. chinense was www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2015, 38, 2789–2796
2791
Table 1. Identification and speculation of phenolic compounds in P. chinense extract by LC–MS
Noa)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
Proposed compound
Gallic acid Ethyl gallate Epicatechin Brevifolin-carboxylic acid 2,6-dihydroxyacetophenone-4-O-β-D-glucoside Quercetin di-O-glycoside Rutin Kaempferol-3-O-rutinoside Quercetin-3-O-xyloside Quercetin-3-O-arabinopyranoside Quercetin-3-O-rhamnoside Quercetin Quercetin-3 -O-rhamnoside 5-methoxy-pinocembrin-7-O-glucoside Kaempferol Kaempferol-3-O-rhamnoside Pinocembrin-7-O-glucoside Pinocembrin-7-O-galloyl-D-glucoside isomer
Pinocembrin-7-O-[4 ,6 -HHDP]-glucoside Pinocembrin dihydrochalcone-7-O-[4 ,6 - HHDP]-glucoside Pinocembrin-7-O-[3 -O-galloyl-4 ,6 -HHDP]-glucoside Pinocembrin dihydrochalcone-7-O-[3 -O- galloyl-4 ,6 -HHDP]-glucoside Pinocembrin
tR (min)
4.84 6.28 11.29 12.08 16.68 17.26 21.16 24.41 25.69 25.76 26.34 26.36 26.40 29.46 29.74 30.63 39.76 41.38 42.16 43.49 47.38 48.40 49.83 55.55 55.84 60.90 62.03
[M+H]+
MS/MS (m–z)
Detected
Expected
Error (ppm)
171.02876 199.06000 291.08632 293.02902 331.10208 627.15426 611.15965 595.16476 435.09174 435.09168 449.10722 303.04975 449.10722 433.14877 287.05481 433.11247 419.13316 571.14363 571.14357 571.14370 571.14361 571.14365 721.13813 723.15326 873.14865 875.16428 257.08073
171.02880 199.06010 291.08631 293.02919 331.10236 627.15558 611.16066 595.16575 435.09219 435.09219 449.10784 303.04993 449.10784 433.14931 287.05501 433.11292 419.13366 571.14462 571.14462 571.14462 571.14462 571.14462 721.13993 723.15558 873.15088 875.16654 257.08084
0.3 0.5 0.6 0.6 0.8 2.1 1.7 1.7 1.0 1.2 1.2 0.6 1.4 1.2 0.7 1.1 1.2 1.7 1.8 1.6 1.8 1.7 2.4 2.9 2.6 2.5 0.4
247, 219, 191, 163 169, 151 535, 303 303 287 303, 257, 229 303, 285, 257, 229 303, 285, 257, 229 285, 257, 229 303, 285, 257, 229 271, 229, 167 287 257, 215, 153 257, 153 257, 153 153 257, 153 257, 153 303, 277, 257, 153 303, 277, 259 493, 303, 257, 153 303, 153 153
a) The notation for peaks refers to Fig. 1.
generated and the similarity indices of these batches were calculated.
2.7 Validation of HPLC method
and stability. The recovery was carried out by spiking known amounts of mixed standard solutions into samples at three concentration levels (low, medium, and high). The average recoveries were calculated by the formula: recovery (%) = (amount found−original amount)/amount spiked × 100.
2.7.1 Calibration curves, LOD, and LOQ
3 Results and discussion The calibration curves were constructed by plotting the peak areas (y axis) versus the concentration (x axis, g/mL) of each standard solution, by analyzing a series of dilute solution with seven concentrations. The LOD and LOQ under the present chromatographic conditions were determined at S/N = 3 and 10, respectively. 2.7.2 Precision, repeatability, stability, and recovery The precision of HPLC was tested by analyzing the same sample (P1) six times. Six working solutions prepared from the same extract (P1) were analyzed to examine the repeatability. The stability was established by analyzing the same sample (P1) at 0, 2, 4, 8, 16, and 24 h. The RSD (%) was taken as a measure to express the precision, repeatability, C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3.1 Optimization HPLC–DAD conditions To achieve a better chromatographic behavior for HPLC fingerprints and determination, different mobile phases, column temperatures and column temperatures were optimized. The combination of acetonitrile and 0.1% v/v formic acid aqueous solution, methanol and 0.1% v/v formic acid aqueous solution, acetonitrile and 0.1% v/v phosphoric acid aqueous solution were investigated, the last one was selected as a mobile phase to reach a satisfactory peak shape. Meanwhile, different column temperatures of 20, 25, and 30⬚C were compared. The results showed that the column temperatures had little influence on separation of the investigated compounds. At 25⬚C, baseline separation of the www.jss-journal.com
2792
W. Guo et al.
analytes could be achieved. To achieve good accuracy and sensitivity for quantification, different detection wavelengths of 245, 280, 301, and 365 nm were investigated by using the DAD. As a result, at 280 nm, maximum UV absorption for each analyte was selected as detection wavelength.
3.2 LC–MS analysis for identification A total of 27 compounds were detected from extract of P. chinense (Table 1 and Fig. 1, m/z error ࣘ 5 ppm). Among them, 13 compounds (1, 4, 5, 7, 9, 12–15, 17, 23, 24, and 27) were identified by comparing with reference substances, and 14 other compounds (2, 3, 6, 8, 10, 11, 16, 18–22, 25, and 26) were tentatively identified based on their MS and MS2 spectra (Fig. 2A). Ten of these compounds (6, 8, 10, 18–24) are reported for the first time from P. chinense. 3.2.1 Identification of phenolic acids Compounds 1, 2, and 3 eluting at 4.84, 6.28 and 11.29 min, respectively, showed [M+H]+ at m/z 171.02876, 199.06000 and 291.08632 respectively. The three compounds matched the elemental composition of C14 H14 O10 with an accuracy of 0.25 ppm, C18 H22 O10 with 0.48 ppm and C15 H14 O6 with 0.55 ppm. They were identified as gallic acid, ethyl gallate and epicatechin, respectively [16, 17]. Compound 4 at m/z 293.02902 (C13 H8 O8 ) showed fragment ions at m/z 247 ([M+H–COOH]+ ), 219 (C11 H7 O6 ), 191 (C10 H7 O4 ) and 177 (C9 H5 O4 ). It was identified as brevifolin-carboxylic acid [18]. For compound 5, the precursor m/z 331.10208 (C14 H19 O9 ) yielded fragments at m/z 169 ([M+H–C6 H10 O5 ]+ ) and 151 ([M+H–C6 H10 O5 –H2 O]+ ), indicating the loss of a glucose residue and consecutive dehydration. It was identified as 2,6dihydroxyacetophenone-4-O--D-flucoside [19]. 3.2.2 Identification of quercetin derivatives Compound 12 produced a MS/MS spectrum [M+H]+ of m/z 303.04975 (C15 H11 O7 ) and fragmented to yield m/z 285.03924 [M+H–H2 O]+ , 257.04440 [M+H–CO–H2 O]+ and 229.04935 [M+H–2CO–H2 O]+ , which was identified to be quercetin [20]. Similarly, a number of derivatives of quercetin were detected by analyzing the characteristic fragmentation patterns of compounds 6, 7, 9, 10, 11, and 13 at m/z 303. Compound 6 yielded [M+H]+ at m/z 627.15426, generating [M+H]+ ion at m/z 303.04962, indicating loss of two glucose moiety (m/z 162, C6 H10 O5 ). Thus, it was tentatively identified as quercetin di-O-glycoside [21]. Compound 7 was identified as rutin because it has the same retention time and UV spectrum as rutin, and its [M+H]+ at m/z 611.15965 indicated the loss of a disaccharide moiety (C12 H20 O9 ) to generate the fragment at m/z 303.04960 [22]. Both compounds 9 and 10 yielded [M+H]+ ion at m/z 435 (C20 H19 O11 ). The presence of m/z at 303 [M+H–132]+ of compounds 9 and 10 indicated the loss of a xylose or arabinose moiety (C5 H8 O4 ). Hence, based on the previous studies [20], compound 9 was identified as quercetin-3-O-xyloside C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2015, 38, 2789–2796
and compound 10 was tentatively identified as quercetin3-O-arabinopyranoside. For both compounds 11 (ions at m/z 449.10728) and 13 (ions at m/z 449.10722), molecular formulas of C21 H21 O11 were predicted. In MS/MS, both compounds further produced the characteristic fragments at m/z 303 ([M+H–C6 H10 O4 ]+ , C15 H11 O7 ), 285 (C15 H9 O6 ), 257 (C14 H9 O5 ) and 229 (C13 H9 O4 ), which were very similar to quercetin. Therefore, compound 11 was tentatively identified as quercetin-3-O-rhamnoside and compound 13 as quercetin-3 -O-rhamnoside [4, 23]. 3.2.3 Identification of kaempferol derivatives The EIC of m/z 287 (C15 H11 O6 ) in positive ion mode exhibited three peaks (8, 15, and 16). Compound 15 yielded [M+H]+ ion at m/z 287.05481 and was identified as kaempferol according to previous report [24]. Compounds 8 and 16 are derivatives of compound 15 and the ion of compound 8 (m/z 595.16476) consecutively lost a disaccharide residue to produce the peak of kaempferol 287.05481 [M+H–C12 H20 O9 ]+ . It was thus, tentatively identified as kaempferol-3-O-rutinoside that was reported in P. chinense herb for the first time. Compound 16 was identified as kaempferol-3-O-rhamnoside because its [M+H]+ ion was at m/z 433.14877 and it produced m/z 287.05479 [M+H–C6 H10 O4 ]+ by elimination of a rhamnose residue [24]. 3.2.4 Identification of pinocembrin derivatives Compound 27 was identified as pinocembrin and gave [M+H]+ at m/z 257.08073, which produced 153.01808 (C7 H5 O4 ) [25]. Compound 17 yielded [M+H]+ at m/z 419.13316 (C21 H23 O9 ) and also consisted of m/z 257.08069 [M+H–C6 H10 O5 ]+ and 153.01815. It was identified as pinocembrin-7-O-glucoside that has been reported to be the main component of P. chinense [4]. Compound 14 yielded [M+H]+ ion at m/z 433.14877 (C22 H25 O9 ) and subsequently produced fragments at m/z 271 and 167, thus indicating the substitution of hydrogen by methyl group (CH3 ). So compound 14 was identified as 5-methoxy-pinocembrin-7O-glucoside and this is a new compound isolated from P. chinense in our laboratory [4]. In addition, five [M+H]+ ions at m/z 571 (C28 H27 O13 ) were detected from 40 to 50 min in full-scan mass spectra. Among these, four compounds produced the fragments at m/z 257 and 153, and hold the same fragment composition. Another one compound only produced ion at m/z 153. The presence of m/z 257 indicated the loss of a galloyl residue (C7 H4 O4 ) and a glucose moiety (C6 H10 O5 ). Thus, compounds 18–22 were tentatively identified as isomers of pinocembrin-O-galloyl-glucoside. Compounds 23 and 24 were identified as pinocembrin-7O-[4 ,6 -hexahydroxydiphenoyl]-glucoside and pinocembrin dihydrochalcone-7-O-[4 ,6 -hexahydroxydiphenoyl]-glucoside (Thonningianins B) respectively and yielded ions at m/z 721.13813 and 723.15326 [26]. The characteristic fragments of compounds 23 (m/z 257.08072, m/z 153.01822) and 24 (m/z 259.09680) indicated the loss of a glucose moiety (C6 H10 O5 ) and a hexahydroxydiphenoyl (HHDP) residue, www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2015, 38, 2789–2796
2793
Figure 1. Chemical structures of 27 compounds identified from P. chinense extract. Table 2. Regression equations, linear range, LOD, and LOQ of investigated compounds by HPLC
Analytes
Regression equationa)
Linear range (g/ml)
r2
LOD (ng)b)
LOQ (ng)c)
Gallic acid (1) Brevifolin carboxylic acid (4) 2,6-Dihydroxyacetophenone-4-O-β–D-glucoside (5) Pinocembrin-7-O-β-D-glucoside (17)
y = 2.6186x–6.3521 y = 0.4606x–5.9059 y = 2.1282x–2.2826 y = 2.0865x–1.1575
9.0-900.0 6.0-600.0 4.4-440.0 6.0-600.0
0.9999 0.9996 0.9999 0.9999
0.45 1.50 0.88 0.60
1.35 6.00 2.20 3.00
a) y represents peak area, x represents concentration of compounds per injection volume (g/mL). b) LOD represents limit of detection (S/N = 3). c) LOQ represents limit of quantification (S/N = 10). Table 3. Precision, repeatability, and stability of compounds 1, 4, 5, and 17
Analytes
Gallic acid (1) Brevifolin carboxylic acid (4) 2,6-Dihydroxyacetophenone-4-O-β-D-glucoside (5) Pinocembrin-7-O-β-D-glucoside (17)
Precision (n = 6)
Repeatability (n = 6)
Stability (n = 6)
Mean(mg/g)
RSD(%)a)
Mean (mg/g)
RSD (%)a)
Mean (mg/g)
RSD (%)a)
27.42 108.75 15.69 22.19
0.30 0.62 0.64 1.12
26.99 105.86 15.38 21.60
2.13 1.19 1.78 2.51
27.50 108.89 15.73 22.37
0.45 0.75 0.81 1.53
a) RSD(%) = (SD/mean) × 100%.
which were consistent with that of their reference substance. Compound 25 and 26 yielded [M+H]+ ion at m/z 873.14865 (C42 H33 O21 ) and 875.16428 (C42 H35 O21 ). They have common fragments at m/z 303 and 153, thus indicating the loss of the HHDP and galloyl moiety. Therefore, compounds 25 and 26 were tentatively identified as pinocembrin-7-
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
O-[3 -O-galloyl-4 ,6 -hexahydroxydiphenoyl]-glucoside and pinocembrin dihydrochalcone-7-O-[3 -O-galloyl-4 ,6 -hexahydroxydiphenoyl]- glucoside (Thonningianins A), which have been reported as active components of P. chinense herb [26, 27]. Compounds 27, 17, 23, and 25 had the same parent nucleus (pinocembrin), which in turn, linked with
www.jss-journal.com
2794
W. Guo et al.
J. Sep. Sci. 2015, 38, 2789–2796
Figure 2. The HPLC–UV chromatogram of P. chinense extract (A) and MS2 spectra of 27, 17, 23, and 25 in positive ion modes (B).
Figure 3. HPLC chromatograms of 21 batches of P. chinense extracts (A) and common peaks identified on reference spectrum (B): (1) gallic acid; (4) brevifolin carboxylic acid; (5) 2,6-dihydroxyacetophenone-4-O--D-glucoside; (17) pinocembrin-7-O--D-glucoside; (25) pinocembrin-7-O-[3 -O-galloyl-4 , 6 -hexahydroxydiphenoyl]-glucoside; (26) pinocembrin dihydrochalcone-7-O-[3 -Ogalloyl-4 ,6 -hexahydroxydiphenoyl]glucoside.
different residues of glucosyl, galloyl, and HHDP. The characteristic spectra of these four compounds were shown in Fig. 2B.
3.3 Method validation The calibration curves of four analytes showed good linearity (r2 > 0.9995) within the test ranges. The LOD and LOQ for each compound was less than 1.5 and 6 ng (Table 2). As summarized in Table 3, the precision (RSD) of four analytes was found to be in the range of 0.30–1.12%. The RSD values of repeatability ranged from 1.19% to 2.51%, which suggests high C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
repeatability. All the analytes showed good stability within 24 h (RSD values of four compounds < 2%). The four compounds showed good recovery rates in the range of 97.45– 103.6%. The results indicate that the method was suitable for the simultaneous determination of four major compounds in P. chinense.
3.4 HPLC fingerprint and quantitation analysis In our study, HPLC chromatograms from 21 batches of P. chinense extract were submitted to the Similarity Evaluation www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2015, 38, 2789–2796
2795
Table 4. The contents (mg/g) of compounds 1, 4, 5, and 17 in 21 samples of P. chinense analyzed by HPLC
Numbers
P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 P13 P14 P15 P16 P17 P18 P19 P20 P21 Mean
Contents (mg/g)a)
Similarity index
1
4
5
17
27.43 26.82 29.40 28.65 26.14 26.63 26.20 25.99 26.23 25.36 25.08 24.55 19.20 18.38 21.92 21.49 22.78 22.49 23.66 23.74 24.01 24.58
107.78 91.41 96.40 74.73 94.71 84.37 97.07 102.96 90.93 84.07 101.39 96.98 98.51 146.39 135.87 132.07 113.35 135.74 115.83 150.71 150.36 109.60
15.52 15.85 16.27 14.83 15.47 14.58 16.30 16.77 15.88 14.66 16.96 16.33 14.54 19.62 15.35 15.47 13.15 14.75 12.84 15.45 15.25 15.52
22.35 21.67 22.41 20.61 20.54 17.46 19.55 18.59 19.37 18.33 18.60 17.85 16.89 18.52 17.25 17.63 16.21 14.78 16.64 19.79 19.88 18.81
0.990 0.990 0.980 0.970 0.986 0.988 0.995 0.990 0.990 0.990 0.986 0.986 0.989 0.962 0.985 0.986 0.987 0.978 0.988 0.984 0.985
a) The notation for analytes refers to Fig. 1.
System for Chromatographic Fingerprint of TCM software, to generate standardized fingerprints (Fig. 3A) [28, 29]. Each sample was processed with a monolithic column within 30 min; otherwise it usually takes more than 60 min with a common column (Fig. 2). The monolithic column results in a rapid analysis instead of UHPLC, which was first used to analyze P. chinense extract. A total of six peaks were defined as common among all the observed peaks (Fig. 3B). Moreover, the similarity index of each batch was not less than 0.96 in comparison to the reference fingerprint. This revealed that different samples had similar and consistent chromatographic patterns. The developed HPLC method was used for simultaneous quantitation of four compounds in extracts from 21 batches of P. chinense. The contents of four major compounds from all batches are shown in Table 4. The average content of gallic acid (1), brevifolin carboxylic acid (4), 2,6-dihydroxyacetophenone-4-O--D-glucoside (5) and pinocembrin-7-O--D-glucoside (17) in P. chinense are 24.58, 109.6, 15.52, and 18.81 mg/g, respectively. The content of brevifolin carboxylic acid was observed to be the highest (average 109.60 mg/g), but it varied significantly from 74.73 to 150.71 mg/g.
the 27 compounds identified from this plant by the highly sensitive LC–MS method, ten have been reported for the first time. Thus, the LC–MS method can be used to identify compounds of P. chinense with high sensitivity. In addition, HPLC fingerprints of the 21 batches showed high similarity and displayed six common peaks. Moreover, the contents of four major compounds were simultaneously determined by HPLC. The average contents of gallic acid, brevifolin carboxylic acid, 2,6-dihydroxyacetophenone-4-O--D-glucoside, and pinocembrin-7-O--D-glucoside in P. chinense were 24.58, 109.6, 15.52, and 18.81 mg/g, respectively. The established HPLC method using a monolithic column is simple, rapid, and effective, which can be used for the comprehensive QC of P. chinense. The authors have declared no conflict of interest.
5 References [1] Xiang, H. H., Yang, J. H., Fu, X.C., Zhang, Y. Y., Wan, H. T., J. Exp. Tradit. Med. Form. 2013, 19, 50–52. [2] Chen, X. R., Yao, H., Jiang, Y., Wu, S. M., Zhu, X. F., Zhou, X. Q., Wang, L. T., Chin. J. Hepatol. 2004, 12, 50.
4 Concluding remarks
[3] Zeng, Q. H., Zhang, X. W., Xu, X. L., Jiang, M. H., Xu, K. P., Piao, J. H., Zhu, L., Chen, J., Jiang, J. G., Food Funct. 2013, 4, 1811–1818.
A simple, fast, and reliable method for the chemical analysis of P. chinense extract was developed in our study. Out of
[4] Wang, M., Jiang, Y., Liu, H. L., Chen, X. Q., Wu, X., Zhang, D. Y., Nat. Prod. Res. 2014, 28, 70–73.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
2796
W. Guo et al.
J. Sep. Sci. 2015, 38, 2789–2796
[5] Xiao, X. Y., Sun, B. H., Ye, X. T., Zhang, G. Y., Hou, P. Y., Gao, H. Y., J. Sep. Sci. 2014, 37, 1533–1545.
[18] Chen, Y. W., Ren, L. J., Chin. Tradit. Herb Drug 1997, 28, 198–202.
[6] Feng, C. G., Wang, H. W., Ren, Q. S., Chin. Pharm. J. 2004, 39, 97–98.
[19] Zhang, X., Yang, M., J. Chengdu Univ. Tradit. Chin. Med. 2002, 25, 46.
[7] He, X. H., Xu, L., Du, F. L., Zeng, J. G., Cent. South. Pharm. 2008, 6, 717–719.
[20] Zhang, M., Pan, D. R., Zhou, Y. F., Zhu, Q. Q., Wang, S. M., J. Anhui Agri. Sci. 2012, 40, 5211–5214.
[8] Du, X., Tang, B., Zhang, Q. B., J. Environ. Health 2011, 28, 358–359.
[21] Qiao, X., He, W. N., Xian, C., Han, J., Wu, L. J., Guo, D. A., Ye, M., Phytochem. Anal. 2011, 22, 475–483.
´ D. C., [9] Stashenko, E. E., Mart´ınez, J. R., Cala, M. P., Duran, Caballero, D., J. Sep. Sci. 2013, 36, 192–202.
[22] Wang, Z. W., Master’s Thesis of Shenyang Pharmaceutical University 2005.
[10] Rasul, A., Millimouno, F. M., Eltayb, W. A., Ali, M., Li, J., Li, X. M., Biomed. Res. Int. 2013, 2013, 379850.
[23] Wang, M., Wu, X., Jiang, Y., Zhang, D. Y., Food Drug 2013, 15, 202–205.
[11] Kapoor, S., Agric, J., Food Chem. 2013, 61, 1416. [12] Lou, K. Q., Master’s Thesis of Jinan University 2013.
[24] Fu, M., Wei, L., Yu, J., Hu, Z. T., Chin. Pharm. J. 2013, 48, 1911–1914.
[13] Pellati, F., Calo, S., Benvenuti, S., J. Chromatogr., A 2007, 1149, 56–65.
[25] Li, Y. Y., Shangguan, Y. Y., Chin. J. Mod. Appl. Pharm. 2012, 29, 159–163.
[14] Atia, N. N., York, P., Clark, B. J., J. Sep. Sci. 2009, 32, 931–938.
[26] Ohtani, I. I., Gotoh, N., Tanaka, J., Higa, T., Gyamfi, M. A., Aniya, Y., J. Nat. Prod. 2000, 63, 676–679.
[15] Kang, J., Zhou, L., Sun, J.H., Han, J., Guo, D. A., J. Pharm. Biomed. Anal. 2008, 47, 778–785.
[27] Lu, Q., Jiang, M. H., Jiang, J. G., Zhang, R. F., Zhang, M. W., J. Agric. Food Chem. 2012, 60, 11097–11103.
[16] Wang, Y., Master’s Thesis of Second Military Medical University 2012.
[28] Liu, Y. S., Meng, Q. H., Jiang, S. M., Hu, Y., Chin. J. Chromatogr. B 2005, 23, 158–163.
[17] Fu, F. Y., Shong, T. M., Xu, Z. P., Acta. Pham. Sinic. 1983, 29, 555–557.
[29] Zhou, W., Xie, M. F., Zhang, X. Y., Liu, T. T., Yu, Y. J., Duan, G. L., J. Sep. Sci. 2011, 34, 1123–1132.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
