{"paper_id":"3a73055b-2afe-47b5-a8f8-e7d438e00186","body_text":"Traditional Chinese medicines, playing an irreplaceable role in the health of people, have attracted more and more attention around the world. Some traditional Chinese medical formulations contain two herbs from the same family, such as Moutan cortex ( M ) and Paeoniae radix alba ( P ) in GuizhiFuling Wan (GZFLW), Asparagi radix and Ophiopogonis in Erdong Gao, Mastic and Myrrha in Qili Capsule, as well as Citrireticulataepericarpium and Aurantiifructusimmaturs in Ermuningke Wan [ 1 ]. Although the two herbs may have similar appearance and constituents, they may differ in their pharmacodynamic action. Within our knowledge, almost all the quality of traditional Chinese medical formulations is controlled through major or active constituents [ 2 , 3 , 4 , 5 ]. The methods for detecting quality are sensitive and available when the formulation has no common components in any of the constituent herbs, but how to determine the real ratio of the two herbs along with the prescription remains a difficult question. Therefore, effective and comprehensive analytical methods have to be developed, therefore controlling the ratio of the each herb to agree with the prescription. In this paper, an available and accurate method, high performance liquid chromatography method with photodiode array detector (HPLC-DAD), was applied for detecting the contents of  M  and  P  in GZFLW.\nGZFLW was first described in Jin Kui Yao Lue written by theeminent Chinese physician Zhang Zhongjing during the Han Dynasty (220 A.D.). In pharmacological research, GZFLW has been reported to protect against brain ischemia-reperfusion injuries and NO-mediated neuronal death [ 6 , 7 ], to improve blood circulation and arteriosclerosis [ 8 ], regulate immunology in endometriosis [ 9 ], and inhibit cervical cancer [ 10 ]. Clinically, GZFLW has been applied to treat gynecological diseases, such as uterine fibroids, endometriosis, pelvic inflammatory disease, ovarian cysts, dysmenorrheal, and “oketsu” syndrome, the so-called blood stasis syndrome in postmenopausal women [ 11 , 12 ].\nGZFLW is composed of five herbs (1:1:1:1:1, g/g), namely  M ,  P , Cinnamomi ramulus, Poria and Persicae semen.  M  and  P , both from the Paeoniaceae family and the  Paeonia  genus, contain similar constituents including galloylpaeoniflorin, paeoniflorin and benzoylpaeoniflorin [ 13 , 14 , 15 ], but have different pharmacodynamic actions, whereby galloylpaeoniflorin has apronounced radical scavenging effect [ 16 ] and inhibits phenylhydroquinone-induced oxidative DNA cleavage [ 17 ], but paeoniflorin could attenuate Aβ25-35-induced neurotoxicity in PC12 cells [ 18 ], and inhibit tumor invasion and metastasis in human hepatocellular carcinoma cells [ 19 ]. Mudanpioside C can be used for treating and preventing cardiovascular diseases [ 20 ].while benzoylpaeoniflorin inhibited the replication of hepatitis B Virus [ 21 ] and has lipoxygenase inhibitory and antioxidant activities [ 22 ].\nAs to the quality control of GZFLW [ 23 , 24 , 25 ], several analytical methods have been only used to analyze the total contents of the common constituents from the two herbs, but not the ratio of the two herbs in the prescription. However, the contents of  M  and  P  in the prescription may exert a large influence on the clinical effect. Therefore, it is essential to determine the contents of  M  and  P  in GZFLW through analysis of four common monoterpene glycosides by HPLC-DAD. The structures of the four common monoterpene glycosides are shown in  Figure 1 . As we know, traditional Chinese medicine contents always vary according to the geographical origins, cultivation and harvesting methods, as well as post-harvesting processes, thereby the original medicinal materials of the proprietary Chinese medicine should be determined according to the analytical method.\nChemical structures of galloylpaeoniflorin ( 1 ), paeoniflorin ( 2 ), mudanpioside C ( 3 ) and benzoylpaeoniflorin ( 4 ).\n\nTo determine the best extraction method for the four monoterpene glycosides with high recovery and no interference at the retention time, various solvents (methanol, acetone and ethyl acetate), methods (ultrasonic extraction and Soxhlet extraction) and times of Soxhlet extraction (1, 2, 3 and 4 h) were applied. As a result, the Soxhlet extraction with methanol for 3 h was selected as the optimum method. The results are shown in  Table 1 .\nOptimization of extraction method of GZFLW.\nTo achieve symmetric peak shapes and short run times for the simultaneous analysis of the four compounds, chromatographic conditions were optimized through different trials. In this respect, the column choice had a great influence on the compound separation, which was essential for the success of the method. During the development of methods, three reversed-phase columns, SunFire TM  C18 (5 μm, 4.6 × 150 mm), Agela Technologies Lnc. C18 (5 μm, 4.6 × 200 mm) and Kromasil ODSI C18 (5 μm, 4.6 × 250 mm), were tested with different mobile phase compositions. UV-Visible detection was applied over the wavelength range of 200–600 nm. Consequently, the selected wavelengths were 270 nm for  1  and  2 , 254 nm for  3  and 230 nm for  4 , and the SunFire TM  column with a gradient elution, methanol, and 0.3% phosphoric acid as the mobile phase, thereby providing the best balance of peak shape, sensitivity and retention time for each monoterpene glycoside. Examples of typical chromatograms wreshown in  Figure 2 ,  Figure 3  and  Figure 4 .\nStacked view of different detector wavelength HPLC chromatograms of mixed reference standards (from topto bottom: 230, 254, 270 nm). Column: SunFire TM  C18 (4.6 mm × 150 mm, 5 μm), temperature of 30 °C.\nStacked view of different detector wavelength HPLC chromatograms of GZFLW (from top to bottom: 230, 254, 270 nm).\nStacked view of HPLC chromatograms (wavelength: 270 nm) of (A) GZFLW; (B) Moutan cortex; (C) Paeoniae radix alba; (D) mixed reference standards; (E) negative control; (F) Blank solvent (from up to down).\nThe specificity of the method was tested by comparing the chromatograms of a blank solution, the mixed working standard solution, the control solution, and the sample solution.\nLinearity was established by the injection of 2, 6, 10, 14, 18 and 22 μL of the mixed working standard solutions. It was also assessed by analyzing calibration curves with the least square linear regression of the integrated peak area (Y)  versus  monoterpeneglycoside content (X). All obtained correlation coefficients were above 0.999. The limits of detection (LOD) and quantification (LOQ) of each monoterpeneglycoside were determined at a signal-to-noise ratio (S/N) of 3 and 10, respectively. Detailed information regarding calibration curves, linear ranges, LOD and LOQ is presented in  Table 2 .\nLinearity, LOD, LOQ of tested compounds determined by the current method.\nThe precision was evaluated by injecting 10 μL of the mixed working reference solution in six replicates in one day. Stability was tested with a GZFLW sample solution over 24 h (the time points of the injections were at 0, 2, 4, 6, 8, 12, and 24 h). The variations were reported as relative standard deviations (RSD in %). The reproducibility of the method was assessed within six independently prepared sample solutions and evaluated by the RSD value of each monoterpeneglycoside content in GZFLW. Recovery was tested in a set of six replicates by spiking the appropriate stock standard solutions into untreated GZFLW at the same concentration. As shown in  Table 3 , validation studies of the method proved that it had good precision and reproducibility, with RSD ranging from 1.13% to 1.69%, and 0.89% to 2.31%, respectively. It was also found that the four monoterpene glycosides in the GZFLW sample solution were all stable for 24 hours with a RSD of 1.20%–1.84%. The results of the recovery test indicated that the recoveries of the four monoterpeneglycosides were satisfactory, between 97.9% and 103.4% with RSD of 0.77%–2.42%.\nPrecision, stability, recovery and reproducibility of the assay method.\nRSD refers to relative standard deviation.\nThe contents of  1 ,  2 ,  3  and  4  in  M ,  P  and GZFLW with different weight ratios of  M  and  P  were quantified by the simultaneous determination of multiponents. The contents of each monoterpeneglycoside in  M ,  P  and GZFLW with different ratios of  M  and  P  were shown in  Table 4 . These results could be used to calculate the weight of  M  and  P  in GZFLW.\nContents of  1 ,  2 ,  3 ,  4  in  M ,  P  and GZFLW (mg/g).\nND (not detected).\nThe contents of  M  and  P  in GZFLW were calculated according to following equations:\n (1) \n (2) \n (3) \n C  ( i )  M  ×  W M  +  C  ( i )  P  ×  W P  =  m i (4) \nwhere  I  was the number of the monoterpene glycoside, also known as  1 ,  2 ,  3  and  4 .  m i  (mg) was the weight of each monoterpene glycoside in GZFLW with the corresponding ratio of M and P.  Y i  was each monoterpene glycoside after accounting for the ratio of four monoterpene glycosides.  C(i) M  and  C(i) P  were the content of each monoterpene glycoside (mg/g) in  M  and  P , while  W M  and  W P  were the weight of  M  and  P , respectively. In these equations,  m i ,  C(i) M  and  C(i) P  were determined as in  Section 2.3.4 .  Y i  could be calculated according to Equation (1). Equation (2) transformed into Equation (3).  W M  and  W P  were calculated using Equations (3) and (4).\nBased on each monoterpene glycoside,  Y i  could be obtained, and according to the equations, the  W M  and  W P  in GZFLW was calculated. Taking  1  as an example, the results are shown in  Table 5 . The accuracy was described by related error (RE), all within 10%. When the weight of  M  and  P  in GZFLW were calculated through all four monoterpeneglycosides, the results would be more available and effective. The results of the method validation indicated that the method developed above could be successfully applied for the analysis of the weight of  M  and  P  in GZFLW.\nThe result of analysis of the weight of  M  and  P  in GZFLW (g).\nCal: calculated value, Act: actual value, RE: related error.\n\nThe reference standard of  1  was isolated from  P  in our laboratory, while  3  and  4  were isolated from  M . The structures were determined by spectral methods, including MS, 1H-NMR and 13C-NMR. The data were consistent with those reported in the literature [ 26 , 27 , 28 ]. The purity of the reference standard was found to be above 98%, based on a peak area normalization method using HPLC-DAD and HPLC-ESI-TOF-MS. The extractions and isolations were as follows:\nThe air-dried  P  (500 g) was extracted with EtOH-H 2 O (70:30, 3 × 5 L) under refluxconditions for 3 h. The combined EtOH extracts were concentrated  in vacuo  to generate a crude residue (82.7 g) that was suspended in H 2 O (200 mL). The suspension was extracted with  n -BuOH-EtOAc (4:1, 5 × 1 L). The combined  n -BuOH-EtOAc portion (41.8 g) was separated by a silica gel CC eluted with CH 2 Cl 2 -MeOH (from 100:0 to 0:100) to yield 24 fractions (A-X). Fraction N (0.3 g) was separated over preparative RP-HPLC (MeCN-H 2 O, 44:56,  v / v ) to afford  1  (34.2 mg).\nCrushed air-dried Moutan cortex (500 g) was extracted with EtOH-H 2 O (70:30, 3 × 5 L) under reflux conditions for 3 h. The combined EtOH extracts were concentrated  in vacuo  to generate a crude residue (43.6 g) that was suspended in H 2 O (150 mL). The suspension was extracted with  n -BuOH-EtOAc (4:1, 5 × 1 L). The combined  n -BuOH-EtOAc portion (20.5 g) was subjected to silica gel column chromatography with a gradients of CH 2 Cl 2 -MeOH (from 100:0 to 0:100) to afford 22 fractions (F 1 –F 22 ). Fraction 14 (0.3 g) was separated by preparative RP-HPLC, with MeCN/H 2 O (40:60,  v / v ) as mobile phase, to afford  3  (27.6 mg). Fraction 10 (0.2 g) was separated via preparative RP-HPLC (MeCN-H 2 O, 40:60,  v / v ) to obtain  3  (24.8 mg).\nThe reference standard of  2  was purchased from the Chinese National Institute for Control of Pharmaceutical and Biological Products (Beijing, China). Cinnamomiramulus, Poriacocos, Moutan cortex, Paeoniae radix alba, and Persicae semen were provided by Jiangsu KanionPharmaceut Co. Ltd. (Lianyungang, Jiangsu, China). The five crude dried parent plants were pulverized and sifted through 24 mesh sieve before analysis. HPLC grade methanol was obtained from Mallinckrodt Baker Inc. (Phillipsburg, PA, USA). Phosphoric acid (HPLC grade) was purchased from Tianjin Kemiou Chemical Reagent Co. Ltd. (Tianjin, China). HPLC grade water was prepared with redistilled water equipment (Shanghai, China) in this study. Acetone and other solvents of analytical grade were obtained from Tianjin Fuyu Chemical Reagent Co. Ltd. (Tianjin, China).\nGZFLW and the negative control (without  M  and  P ) were prepared in accordance with the process stated in the Chinese Pharmacopeia (2010 Edition) [ 1 ]. The negative control was used to prove that the otherthree herbs didn’t contain theanalyzed compoundsgalloylpaeoniflorin ( 1 ), paeoniflorin ( 2 ), mudanpioside C ( 3 ) and benzoylpaeoniflorin ( 4 ).\nThe accurately weighed GZFLW (5 g), negative control (3 g),  M  (1 g) and  P  (1 g) were extracted by methanol for 3 h in 60 mL soxhlet flask, respectively. After cooling, the solutions were removed and diluted with methanol to 100 mL in a volumetric flask, and then filtered through a 0.22 μm Millipore filter.\nStock solutions of  1 ,  2 ,  3  and  4  were prepared by dissolving the appropriate amount of each standard compound in methanol. A mixed working standard solution was prepared by diluting a mixture of each reference compound stock solution with methanol. The concentrations of the mixed working standard solutions were 22.05 μg/mL for  1 , 332.75 μg/mL for  2 , 5.95 μg/mL for  3  and 45.6 μg/mL for  4 , respectively. These solutions were stored in a refrigerator at −20 °C and brought to room temperature before analysis.\nWaters-2690 Alliance HPLC instrument (Waters Corporation, Milford, MA, USA) was used in this study, equipped with an online degasser, an auto injector, a column heater and a 2996 photodiode array detector (DAD). UV-Visible detection was achieved over the wavelength range of 200–600 nm. The detection wavelength was 270 nm for  1  and  2 , while  3  and  4  were monitored at 254 nm and 230 nm, respectively. Chromatographic separation was performed at 30 °C on a SunFire TM  C18 reverse phase column (5.0 μm, 150 mm×4.6 mm I.D.), with (A) methanol and (B) 0.3% phosphoric acid as the mobile phase. Gradient programming was performed with linear gradient (5%–8% A at 0–8 min, 8%–24% A at 8–13 min, 24%–29% A at 13–30 min, 29%–42% A at 30–35 min, 42%–51% A at 35–60 min). The flow rate was 1.0 mL/min, and the injection volume was 10 μL. Therefore, the monoterpene glycosides were well separated in the above chromatographic conditions.\n\nIn our study, a new and reliable analytic method was developed and validated to detect the weight of  M  and  P  in GZFLW by HPLC-DAD through the analysis of the content of  1 ,  2 ,  3  and  4  in  M ,  P  and GZFLW with different ratios of  M  and  P . The method was simple, but has been demonstrated to be of excellent precision and accuracy. We successfully used the method by measuring the weight of  M  and  P  in GZFLW. All relative errors (RE) of the weight of  M  and  P  in GZFLW were within 10%. The result has shown that the method could besuccessfully applied for the content analysis of  M  and  P  inGZFLW. Moreover, it may be widely used to control the quality of proprietary Chinese medicines containing herbs of the same genus or family in industrial production.","source_license":"CC-BY-4.0","license_restricted":false}