Journal of Chromatographic Science 2015;53:161– 166 doi:10.1093/chromsci/bmu034 Advance Access publication May 28, 2014
Article
Fingerprinting and Simultaneous Determination of Alkaloids and Limonins in Phellodendri Amurensis Cortex From Different Locations by High-Performance Liquid Chromatography with Diode Array Detection Lihong Wang, Guangli Yan, Aihua Zhang, Hui Shi, Hui Sun and Xijun Wang* National TCM Key Lab of Serum Pharmacochemistry, Key Lab of Metabolomics and Chinmedomics, Heilongjiang University of Chinese Medicine, Heping Road 24, Harbin 150040, China *Author to whom correspondence should be addressed. Email: [email protected] Received 19 August 2013; revised 19 November 2013
A sensitive high-performance liquid chromatography method coupled with diode array detection (HPLC-DAD) was developed for the quality control of Phellodendri amurensis cortex (PAC), the quality control included the simultaneous determination of seven major constituents, namely phellodendrine, magnoflorine, jatrorrhizine, palmatine, berberine, obaculactone and obacunone. The chromatographic separation was accomplished on a Diamonsil-C18 column (4.6 mm 3 200 mm, 5 mm) with acetonitrile and 0.1% phosphoric acid (0.02 mol sodium dihydrogen phosphate per liter) by linear gradient elution. The established method was successfully validated by acceptable linearity, limits of detection and quantitation, precision, repeatability, stability and accuracy. The HPLC-DAD fingerprint chromatograph under 220 nm consisting of 21 peaks was constructed for the evaluation of the 11 batches of PAC. The HPLC fingerprints were analyzed by similarity analysis, hierarchical clustering analysis and principal component analysis. The results indicated that the combination of multicomponent determination method and chromatographic fingerprint analysis could be employed for the quantitative analysis and identification of PAC, as well as pharmaceutical products containing this herbal material.
Introduction Phellodendri amurensis cortex (PAC) is a well-known traditional Chinese medicine, known as Guanhuangbai, originated from the dried bark of Phellodendron amurense Rupr. Modern pharmacological researches show that PAC possess a variety of biological activities including anti-tumor (1, 2), anti-inflammatory (3), antimicrobial (4), anti-oxidant (4), anti-herpes simplex virus (4), hypoglycemic and neuroprotective (5), and the prevention of prostate cancer (6) activities is better than Phellodendri Chinensis Cortex (commonly called Huangbai). Multiple constituents contribute for the effects of PAC; however, only palmatine (Pal) and berberine (Ber) were determined according to the current version of 2010 Chinese Pharmacopoeia and most of the studies (7). Although they are the main effective components of PAC, and they are also found in some other species, such as Berberis and Ranunculaceae Thalictrum. Therefore, it is essential to establish integral quality control methods for evaluating the quality of PAC to ensure the efficacy of the drug. Previous pharmacological studies demonstrated its main chemical constituents in blood except for several alkaloids, and there were also some limonoids (8). Thus, this paper select alkaloids including phellodendrine (Phe), magnoflorine (Mag), jatrorrhizine (Jat), Pal, Ber and limonoids consisting of obaculactone (Ob1) and obacunone (Ob2), seven ingredients as indicators for the simultaneous
determination method for the quality control of PAC. In recent years, high-performance liquid chromatography coupled with diode array detection (HPLC-DAD) has become a convenient, frequently used, and powerful tool for the identification and qualification of herbal medicines (9 – 11). The paper aimed to demonstrate the quantification of multi-ingredients and chromatographic fingerprint for the quality control of PAC. Materials and methods Chemical reagents and materials Acetonitrile was of HPLC grade and supplied by Merck Company, Inc. (Merck, Darmstadt, Germany). Ultrapure water was prepared by a Milli-Q50 SP Reagent Water System (Millipore Corporation, MA, USA). All other reagents were of analytical grade. Authentic standards of Phe, Mag, Jat, Pal, Ber, Ob1 and Ob2 were provided by Sichuan WeiKeqi Bio-Tech Co., Ltd (Sichuan, China). All the standard compounds have over 98% purity. Their structures are shown in Figure 1. The samples of PAC were collected from Heilongjiang (H), Jilin (J) and Liaoning (L) Provinces of China in November 2012 and were authenticated by Prof. Xijun Wang. Standard solutions and sample preparation A mixed stock solution containing 35.5 mg mL21 Phe, 97.8 mg mL21 Mag, 5.8 mg mL21 Jat, 72.5 mg mL21 Pal, 252 mg mL21 Ber, 146.6 mg mL21 Ob1 and 57.4 mg mL21 Ob2 was prepared with methanol. Working solutions were prepared by diluting the mixed stock solution with methanol to give different concentrations for the establishment of calibration curves. All the standard solutions were stored in the refrigerator at 48C and filtered through a 0.45-mm membrane (Automatic Science, Tianjin, Instrument Co., LET) before HPLC analysis. The precisely weighed powder (0.5 g, 40-mesh) was transferred into Erlenmeyer flask and extracted with 50 mL of 60% methanol in an ultrasonic bath at room temperature for 45 min (250 W, 40 kHz). Then, additional 60% methanol was added to make up the lost. Finally, the resulting solution was filtered through a 0.45-mm membrane prior to HPLC injection. All samples were prepared for analysis in duplicate. HPLC apparatus and chromatographic conditions The HPLC analysis was carried out on Waters 2695 Alliance HPLC system (Waters Corp., Milford, MA, USA) consisting of a quaternary pump, an on-line degasser, an autosampler and a diode array
# The Author [2014]. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Figure 1. The chemical structures of seven active components in PAC.
UV/Vis multiwavelength detector (DAD) (Waters 996, USA). The data were collected and processed with the Millennium 32 Software. The separations were performed on a Diamonsil-C18 column (4.6 mm 200 mm, 5 mm), and the column temperature was set at 308C. The mobile phase was composed of (A) acetonitrile – (B) 0.1% phosphoric acid (0.02 mol sodium dihydrogen phosphate per liter) using a gradient elution of 4 – 6% A at 0 – 5 min, 6 – 10% A at 5 – 10 min, 10 – 14% A at 10 – 30 min, 14 – 32% A at 30 – 50 min, 32 – 70% A at 50 – 60 min and 70 – 90% A at 60 –75 min. The solvents were filtered and degassed prior to use. The solvent flow rate was 1 mL/min, and the injection volume was 10 mL. The detection wavelength for content determination was set at the maximal UV absorption peak of each compound, they were 207 nm for compound Phe and Ob1, 213 nm for compound Ob2, 223 nm for compound Mag and 348 nm for compound Jat, Pal and Ber and the detection wavelength for chromatogram fingerprint was 220 nm. The absorption spectra were recorded within 200–400 nm. Method validation According to the guideline of International Conference on Harmonization (ICH) and some studies about determination, analysis was validated by its linearity, limits of detection (LODs) and quantitation (LOQs), precision, repeatability, stability and recovery.
Results Optimization of chromatographic conditions Initially, different compositions of mobile phase (acetonitrile – 0.05% phosphoric acid, acetonitrile–0.1% phosphoric acid, acetonitrile – 0.2% phosphoric acid, acetonitrile – 0.1 mol L21 NH4Cl, acetonitrile–0.2 mol L21 NH4Cl, acetonitrile–0.4 mol L21 NH4Cl, acetonitrile – 0.1% phosphoric acid, acetonitrile – 0.1% formic acid, acetonitrile–0.2% triethylamine solution) were tested. As a 162 Wang et al.
result, acetonitrile –0.1% phosphoric acid in the gradient mode was adopted because of the stable baseline, the better separation and the more characteristic peaks in the chromatograms. In addition, four column temperatures, 25, 30, 35 and 408C, and two chromatographic columns (Diamonsil-C18 column, 4.6 mm 200 mm, 5 mm), Symmetry Shied-RP18 column (4.6 mm 150 mm, 5 mm), were also tested, eventually, the optimal condition was determined as 2.3. Detection wavelength for fingerprint was set at 220 nm, where all compounds could be detected with adequate adsorption. According to maximum absorption in the range of 200 – 400 nm, the detection wavelength for determination was set. The optimized wavelength of Phe and Ob1 was detected at 207 nm, Ob2 at 213 nm, Mag at 223 nm and Jat, Pal and Ber at 348 nm, respectively. Optimization of sample extraction In terms of the extraction efficiency, extraction solvent, sample –solvent ratio, extraction method and extraction time were all investigated. The solvent of methanol was found to have much more peaks with higher response and better peak shape than ethanol. Then, various solvent ratios of methanol (60, 80 and 100%) were evaluated, and 60% methanol was choosed for its highest extraction yields. Finally, ultrasonication (30, 45, 60 and 90 min) and heat reflux (30, 60 and 90 min) extraction methods were examined. The results showed that ultrasonication 45 min was the most effective by comparing the peak areas of analytes in HPLC chromatograms. So, 60% methanol ultrasonic extracting for 45 min was confirmed to be the optimal extraction methods. Analytical method validation Calibration curves, LODs and LOQs Methanol stock solutions containing the seven analytes were prepared and diluted to appropriate concentrations for plotting
the calibration curves. Six different concentrations of the seven analytes solution were analyzed, and then the calibration curves were generated by plotting peak area (Y) versus the concentrations of each analyte (X). The calibration curves and ranges of the seven components were presented in Table I. And, all of the analytes achieved good linearity (R 2 . 0.999) within test ranges. The LODs and LOQs were determined at signal-to-noise (S/N) ratios of 3 and 10, respectively.
Precision, repeatability, stability and accuracy Injection precision was assessed by analyzing the same sample six consecutive times within 1 day. The results were expressed with relative standard deviations (RSD). The results appeared that the RSD of relative retention time (RRT) and relative peak area (RPA) did not exceed 0.66 and 1.68%, respectively. To evaluate the repeatability of the assay, six independent working solutions were prepared and analyzed as described above. The RSDs of RRT and RPA were not more than 0.27 and 2.14% for all analytes, respectively. Stability was tested with one sample solution at room temperature and analyzed at 0, 2, 4, 8, 12 and 24 h within a day. The RSD. values of the RRTs were ,0.80% and the RPAs was ,1.88% for all analytes, respectively. The similarity of these results showed that the sample solution kept stable within 24 h at room temperature. The accuracy of the analytical method was evaluated by recovery tests. The recoveries were performed by spiking accurately known contents of the mixed standard solution of seven analytes into 0.25 g of the PAC and then extracted, processed and analyzed with the established procedures. The recovery of all
seven tested compounds was within the range of 97.22 – 101.06%, with RSD ,2.68%. The results of the recovery test indicated that the established method was reliable and acceptable. Quantitative determination of the seven compounds extracted from PAC The proposed method was applied for the simultaneous determination of seven compounds in PAC from different locations of China under the optimized HPLC conditions. Nine different PAC samples were analyzed, and each sample was analyzed in duplicate to determine the mean content and the results were listed in Supplementary data, Table S1. The HPLC-DAD chromatograms of standards and samples were shown in Figure 2. And, the peaks that correspond to each chemical were separated. The amounts of the seven compounds in the nine PAC samples were found to be different. The variations in the contents of these main effective components might be result from these factors, such as cultivation year, geographical location and storage conditions. However, the contents of Pal and Ber in all investigated samples complied with the Chinese Pharmacopoeia (State Pharmacopoeia Commission of People’s Republic of China, 2010). These data indicated that the proposed HPLC-DAD method could be used for the simultaneous determination of the seven compounds in PAC. Fingerprint analysis of PAC The HPLC fingerprints of the samples were all generated at the UV absorption of 220 nm. Peaks that exist in all chromatograms
Table I Regression data, LODs and LOQs for the components determined (n ¼ 6) Components
Monitoring wavelength (nm)
Regression equation
Correlation coefficient (R 2)
Linear range (mg/mL)
LOD (mg/mL)
LOQ (mg/mL)
Phellodendrine Magnoflorine Jatrorrhizine Palmatine Berberine Obaculactone Obacunone
207 223 348 348 348 207 213
Y ¼ 139,038,302.1X þ 76,955 Y ¼ 74,938,711.1X-13,218 Y ¼ 49,005,312.9X-456 Y ¼ 45,891,531.4X þ 31,385 Y ¼ 39,582,070.7X þ 124,575 Y ¼ 7,551,783X-1,274 Y ¼ 2,466,601.2X þ 8,707
0.9995 0.9997 0.9993 0.9998 0.9995 0.9990 0.9998
3.55 –35.5 9.78 –97.8 0.579 8 –5.798 7.25 –72.5 25.2 –252 14.66 –146.6 5.738 –57.38
0.27 0.14 0.08 0.43 1.01 0.44 0.49
1.06 0.42 0.28 1.45 3.78 1.47 1.72
Figure 2. The HPLC-DAD chromatograms of mixed standards (A) and PAC samples at 207 (B), 213 (C), 223 (D) and 348 nm (E), respectively. (1) Phellodendrine (28.6 min); (2) magnoflorine (33.5 min); (3) jatrorrhizine (48.3 min); (4) palmatine (51.6 min); (5) berberine (52.3 min); (6) obaculactone (59.8 min) and (7) obacunone (62.1 min).
Fingerprinting and Determination of Alkaloids and Limonins in Phellodendri Amurensis Cortex 163
of the samples were assigned as the common peak. A total of 21 common peaks were found in the chromatograms among all 11 batches of PAC samples, and seven peaks were identified by comparing their retention times and UV spectra (Figure 3) with those of the corresponding standards, the peaks of number 7, 9, 14, 15, 16, 17 and 19 were corresponding to compound Phe, Mag, Jat, Pal, Ber, Ob1 and Ob2, respectively (Figure 4). The correlation coefficients and cosine of the angle of the samples were calculated as the similarity evaluation of the fingerprints. The average chromatogram of the 11 batches of PAC samples was assumed to be the common pattern of PAC. So, the similarity analysis was carried out by comparing with the common pattern, and the results are shown in Supplementary data, Table S2. The similarity values of 11 samples were . 0.93, meaning that samples from different batches were similar in general. Hierarchical clustering analysis was used to sort samples into groups following the basis of 21 peak values of the HPLC fingerprints. In this study, different samples of PAC were analyzed by using the SPSS 19.0 software, the between-group linkage method was applied and Euclidean distance as a measure to classify the 11 samples, and they were grouped as in Figure 5. These samples were divided into four main clusters, that is, Group A (H2, J2, H5), Group B (H1, J3), Group C (H4, J1, L1, L2, H3) and Group D (L3), based on the absorption intensity of common peaks, Groups B and D were thought to be quality product, Groups A and C were thought to be general quality.
Figure 3. The UV spectrum of seven components in PAC.
164 Wang et al.
Discussion In recent years, HPLC-DAD has become a convenient, frequently used, and powerful tool for the identification and qualification of herbal medicines. The paper integrated the quantification of multi-ingredients and chromatographic fingerprint for the quality control of PAC. For further analyzing, principal component analysis was performed based on peak areas of 21 common chromatographic peaks from the HPLC fingerprints. The samples were clustered into four domains. With each domain being formed by samples H4, J1, L1, L2, H3, samples H2, J2, H5, samples H1, J3, and sample L3, respectively. The results of the PCA were shown in Figure 6. Here, the grouping was identical to that of the analysis by HCA. On the basis of the results above, similarity analysis, HCA and PCA, could be combined for identifying and controlling the quality evaluation of PAC. In this research, a simple and reliable analytical method was developed for the simultaneous analysis of seven major compounds in PAC, namely Phe, Mag, Jat, Pal, Ber, Ob1 and Ob2. The fingerprints of PAC were also established. Eleven batches of PAC from three genuine producing areas were assessed by chromatographic fingerprint analysis with similarity analysis, hierarchical clustering analysis and principal component analysis. The method obtained good linearity, precision, repeatability, stability and recovery. Thus, this study provided an example for quality evaluation of PAC by using HPLC-DAD coupled with multiple compounds determination and HPLC fingerprint.
Figure 4. (A) The HPLC chromatogram of mixed standards; (B) common pattern of HPLC fingerprint of PAC sample and (C) HPLC characteristic fingerprints of 11 PAC samples. All chromatograms were detected at the wavelength of 220 nm. The peaks marked with 1– 21 in the chromatogram represent 21 common peaks.
81173500, 81102556, 81373930, 81302905, 90709019), National Key Technology Research and Development Program of the Ministry of Science and Technology of China (grant no. 2011BAI03B03, 2011BAI03B06, 2011BAI03B08), Key Science and Technology Program of Heilongjiang Province, China (grant no. GC06C501, GA08C303, GA06C30101), Foundation of Heilongjiang University of Chinese Medicine (grant no. 201209) and Key Project of Chinese Ministry of Education (grant no. 212044). Conflict of Interest statement: The authors have declared that they have no competing interests.
Figure 5. The cluster analysis results of 11 batches of PAC.
References
Figure 6. Score plot obtained from the PCA of 11 PAC samples.
Supplementary data Supplementary data are available at Journal of Chromatographic Science online.
Funding This work was supported by grants from the Key Program of Natural Science Foundation of State (grant no. 81202639,
1. James, M.A., Fu, H., Liu, Y., Chen, D.R., You, M.; Dietary administration of berberine or Phellodendron amurense extract inhibits cell cycle progression and lung tumorigenesis; Molecular Carcinogenesis, (2011); 50: 1–7. 2. Kumar, R., Das, M., Ansari, K.M.; Nexrutine(R) inhibits tumorigenesis in mouse skin and induces apoptotic cell death in human squamous carcinoma A431 and human melanoma A375 cells; Carcinogenesis, (2012); 33: 1909–1918. 3. Kim, J.H., Huh, J.E., Baek, Y.H., Lee, J.D., Choi, D.Y., Park, D.S.; Effect of Phellodendron amurense in protecting human osteoarthritic cartilage and chondrocytes; Journal of Ethnopharmacology, (2011); 134: 234–242. 4. Wang, W., Zu, Y., Fu, Y., Reichling, J., Suschke, U., Nokemper, S., et al.; In vitro antioxidant, antimicrobial and anti-herpes simplex virus type 1 activity of Phellodendron amurense Rupr. from China; The American Journal of Chinese Medicine, (2009); 37: 195– 203. 5. Xian, Y.F., Lin, Z.X., Ip, S.P., Su, Z.R., Chen, J.N., Lai, X.P.; Comparison the neuropreotective effect of Cortex Phellodendri Chinensis and Cortex Phellodendri Amurensis against beta-amyloid-induced neurotoxicity in PC12 cells; Phytomedicine, (2013); 20: 187– 193. 6. Kumar, A.P., Graham, H., Robson, C., Thompson, I.M., Ghosh, R.; Natural products: potential for developing Phellodendron amurense bark extract for prostate cancer management; Mini Reviews in Medicinal Chemistry, (2010); 10: 388–397. 7. Chan, C.-O., Chu, C.-C., Mok, D.K., Chau, F.-T.; Analysis of berberine and total alkaloid content in Cortex Phellodendri by near infrared spectroscopy (NIRS) compared with high-performance liquid chromatography coupled with ultra-visible spectrometric detection; Analytica Chimica Acta, (2007); 592: 121– 131.
Fingerprinting and Determination of Alkaloids and Limonins in Phellodendri Amurensis Cortex 165
8. Wang, H., Yan, G., Zhang, A., Li, Y., Wang, Y., Sun, H., et al.; Rapid discovery and global characterization of chemical constituents and rats metabolites of Phellodendri amurensis cortex by ultra-performance liquid chromatography-electrospray ionization /quadrupole – timeof-flight mass spectrometry coupled with pattern recognition approach; Analyst, (2013); 138: 3303– 3312. 9. Tang, X., Huang, Z., Chen, Y., Liu, Y., Liu, Y., Zhao, J., et al.; Simultaneous determination of six bioactive compounds in evodiae fructus by high-performance liquid chromatography with diode array detection; Journal of Chromatographic Science, (2014); 52: 149–156.
166 Wang et al.
10. Mayer, M., Benko, A., Husza´r, A., Sipos, K., Lajtai, A., Lakatos, A., et al.; Simultaneous determination of 4-substituted cathinones (4-MMC, 4-MEC and 4-FMC) in human urine by HPLC-DAD; Journal of Chromatographic Science, (2013); 51: 861–866. 11. Branquinho, R.T., Mosqueira, V.C., Kano, E.K., de Souza, J., Dorim, D.D., Sau´de-Guimara˜es, D.A., et al.; HPLC-DAD and UVspectrophotometry for the determination of lychnopholide in nanocapsule dosage form: validation and application to release kinetic study; Journal of Chromatographic Science, (2014); 52: 19– 26.
Article
Fingerprinting and Simultaneous Determination of Alkaloids and Limonins in Phellodendri Amurensis Cortex From Different Locations by High-Performance Liquid Chromatography with Diode Array Detection Lihong Wang, Guangli Yan, Aihua Zhang, Hui Shi, Hui Sun and Xijun Wang* National TCM Key Lab of Serum Pharmacochemistry, Key Lab of Metabolomics and Chinmedomics, Heilongjiang University of Chinese Medicine, Heping Road 24, Harbin 150040, China *Author to whom correspondence should be addressed. Email: [email protected] Received 19 August 2013; revised 19 November 2013
A sensitive high-performance liquid chromatography method coupled with diode array detection (HPLC-DAD) was developed for the quality control of Phellodendri amurensis cortex (PAC), the quality control included the simultaneous determination of seven major constituents, namely phellodendrine, magnoflorine, jatrorrhizine, palmatine, berberine, obaculactone and obacunone. The chromatographic separation was accomplished on a Diamonsil-C18 column (4.6 mm 3 200 mm, 5 mm) with acetonitrile and 0.1% phosphoric acid (0.02 mol sodium dihydrogen phosphate per liter) by linear gradient elution. The established method was successfully validated by acceptable linearity, limits of detection and quantitation, precision, repeatability, stability and accuracy. The HPLC-DAD fingerprint chromatograph under 220 nm consisting of 21 peaks was constructed for the evaluation of the 11 batches of PAC. The HPLC fingerprints were analyzed by similarity analysis, hierarchical clustering analysis and principal component analysis. The results indicated that the combination of multicomponent determination method and chromatographic fingerprint analysis could be employed for the quantitative analysis and identification of PAC, as well as pharmaceutical products containing this herbal material.
Introduction Phellodendri amurensis cortex (PAC) is a well-known traditional Chinese medicine, known as Guanhuangbai, originated from the dried bark of Phellodendron amurense Rupr. Modern pharmacological researches show that PAC possess a variety of biological activities including anti-tumor (1, 2), anti-inflammatory (3), antimicrobial (4), anti-oxidant (4), anti-herpes simplex virus (4), hypoglycemic and neuroprotective (5), and the prevention of prostate cancer (6) activities is better than Phellodendri Chinensis Cortex (commonly called Huangbai). Multiple constituents contribute for the effects of PAC; however, only palmatine (Pal) and berberine (Ber) were determined according to the current version of 2010 Chinese Pharmacopoeia and most of the studies (7). Although they are the main effective components of PAC, and they are also found in some other species, such as Berberis and Ranunculaceae Thalictrum. Therefore, it is essential to establish integral quality control methods for evaluating the quality of PAC to ensure the efficacy of the drug. Previous pharmacological studies demonstrated its main chemical constituents in blood except for several alkaloids, and there were also some limonoids (8). Thus, this paper select alkaloids including phellodendrine (Phe), magnoflorine (Mag), jatrorrhizine (Jat), Pal, Ber and limonoids consisting of obaculactone (Ob1) and obacunone (Ob2), seven ingredients as indicators for the simultaneous
determination method for the quality control of PAC. In recent years, high-performance liquid chromatography coupled with diode array detection (HPLC-DAD) has become a convenient, frequently used, and powerful tool for the identification and qualification of herbal medicines (9 – 11). The paper aimed to demonstrate the quantification of multi-ingredients and chromatographic fingerprint for the quality control of PAC. Materials and methods Chemical reagents and materials Acetonitrile was of HPLC grade and supplied by Merck Company, Inc. (Merck, Darmstadt, Germany). Ultrapure water was prepared by a Milli-Q50 SP Reagent Water System (Millipore Corporation, MA, USA). All other reagents were of analytical grade. Authentic standards of Phe, Mag, Jat, Pal, Ber, Ob1 and Ob2 were provided by Sichuan WeiKeqi Bio-Tech Co., Ltd (Sichuan, China). All the standard compounds have over 98% purity. Their structures are shown in Figure 1. The samples of PAC were collected from Heilongjiang (H), Jilin (J) and Liaoning (L) Provinces of China in November 2012 and were authenticated by Prof. Xijun Wang. Standard solutions and sample preparation A mixed stock solution containing 35.5 mg mL21 Phe, 97.8 mg mL21 Mag, 5.8 mg mL21 Jat, 72.5 mg mL21 Pal, 252 mg mL21 Ber, 146.6 mg mL21 Ob1 and 57.4 mg mL21 Ob2 was prepared with methanol. Working solutions were prepared by diluting the mixed stock solution with methanol to give different concentrations for the establishment of calibration curves. All the standard solutions were stored in the refrigerator at 48C and filtered through a 0.45-mm membrane (Automatic Science, Tianjin, Instrument Co., LET) before HPLC analysis. The precisely weighed powder (0.5 g, 40-mesh) was transferred into Erlenmeyer flask and extracted with 50 mL of 60% methanol in an ultrasonic bath at room temperature for 45 min (250 W, 40 kHz). Then, additional 60% methanol was added to make up the lost. Finally, the resulting solution was filtered through a 0.45-mm membrane prior to HPLC injection. All samples were prepared for analysis in duplicate. HPLC apparatus and chromatographic conditions The HPLC analysis was carried out on Waters 2695 Alliance HPLC system (Waters Corp., Milford, MA, USA) consisting of a quaternary pump, an on-line degasser, an autosampler and a diode array
# The Author [2014]. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Figure 1. The chemical structures of seven active components in PAC.
UV/Vis multiwavelength detector (DAD) (Waters 996, USA). The data were collected and processed with the Millennium 32 Software. The separations were performed on a Diamonsil-C18 column (4.6 mm 200 mm, 5 mm), and the column temperature was set at 308C. The mobile phase was composed of (A) acetonitrile – (B) 0.1% phosphoric acid (0.02 mol sodium dihydrogen phosphate per liter) using a gradient elution of 4 – 6% A at 0 – 5 min, 6 – 10% A at 5 – 10 min, 10 – 14% A at 10 – 30 min, 14 – 32% A at 30 – 50 min, 32 – 70% A at 50 – 60 min and 70 – 90% A at 60 –75 min. The solvents were filtered and degassed prior to use. The solvent flow rate was 1 mL/min, and the injection volume was 10 mL. The detection wavelength for content determination was set at the maximal UV absorption peak of each compound, they were 207 nm for compound Phe and Ob1, 213 nm for compound Ob2, 223 nm for compound Mag and 348 nm for compound Jat, Pal and Ber and the detection wavelength for chromatogram fingerprint was 220 nm. The absorption spectra were recorded within 200–400 nm. Method validation According to the guideline of International Conference on Harmonization (ICH) and some studies about determination, analysis was validated by its linearity, limits of detection (LODs) and quantitation (LOQs), precision, repeatability, stability and recovery.
Results Optimization of chromatographic conditions Initially, different compositions of mobile phase (acetonitrile – 0.05% phosphoric acid, acetonitrile–0.1% phosphoric acid, acetonitrile – 0.2% phosphoric acid, acetonitrile – 0.1 mol L21 NH4Cl, acetonitrile–0.2 mol L21 NH4Cl, acetonitrile–0.4 mol L21 NH4Cl, acetonitrile – 0.1% phosphoric acid, acetonitrile – 0.1% formic acid, acetonitrile–0.2% triethylamine solution) were tested. As a 162 Wang et al.
result, acetonitrile –0.1% phosphoric acid in the gradient mode was adopted because of the stable baseline, the better separation and the more characteristic peaks in the chromatograms. In addition, four column temperatures, 25, 30, 35 and 408C, and two chromatographic columns (Diamonsil-C18 column, 4.6 mm 200 mm, 5 mm), Symmetry Shied-RP18 column (4.6 mm 150 mm, 5 mm), were also tested, eventually, the optimal condition was determined as 2.3. Detection wavelength for fingerprint was set at 220 nm, where all compounds could be detected with adequate adsorption. According to maximum absorption in the range of 200 – 400 nm, the detection wavelength for determination was set. The optimized wavelength of Phe and Ob1 was detected at 207 nm, Ob2 at 213 nm, Mag at 223 nm and Jat, Pal and Ber at 348 nm, respectively. Optimization of sample extraction In terms of the extraction efficiency, extraction solvent, sample –solvent ratio, extraction method and extraction time were all investigated. The solvent of methanol was found to have much more peaks with higher response and better peak shape than ethanol. Then, various solvent ratios of methanol (60, 80 and 100%) were evaluated, and 60% methanol was choosed for its highest extraction yields. Finally, ultrasonication (30, 45, 60 and 90 min) and heat reflux (30, 60 and 90 min) extraction methods were examined. The results showed that ultrasonication 45 min was the most effective by comparing the peak areas of analytes in HPLC chromatograms. So, 60% methanol ultrasonic extracting for 45 min was confirmed to be the optimal extraction methods. Analytical method validation Calibration curves, LODs and LOQs Methanol stock solutions containing the seven analytes were prepared and diluted to appropriate concentrations for plotting
the calibration curves. Six different concentrations of the seven analytes solution were analyzed, and then the calibration curves were generated by plotting peak area (Y) versus the concentrations of each analyte (X). The calibration curves and ranges of the seven components were presented in Table I. And, all of the analytes achieved good linearity (R 2 . 0.999) within test ranges. The LODs and LOQs were determined at signal-to-noise (S/N) ratios of 3 and 10, respectively.
Precision, repeatability, stability and accuracy Injection precision was assessed by analyzing the same sample six consecutive times within 1 day. The results were expressed with relative standard deviations (RSD). The results appeared that the RSD of relative retention time (RRT) and relative peak area (RPA) did not exceed 0.66 and 1.68%, respectively. To evaluate the repeatability of the assay, six independent working solutions were prepared and analyzed as described above. The RSDs of RRT and RPA were not more than 0.27 and 2.14% for all analytes, respectively. Stability was tested with one sample solution at room temperature and analyzed at 0, 2, 4, 8, 12 and 24 h within a day. The RSD. values of the RRTs were ,0.80% and the RPAs was ,1.88% for all analytes, respectively. The similarity of these results showed that the sample solution kept stable within 24 h at room temperature. The accuracy of the analytical method was evaluated by recovery tests. The recoveries were performed by spiking accurately known contents of the mixed standard solution of seven analytes into 0.25 g of the PAC and then extracted, processed and analyzed with the established procedures. The recovery of all
seven tested compounds was within the range of 97.22 – 101.06%, with RSD ,2.68%. The results of the recovery test indicated that the established method was reliable and acceptable. Quantitative determination of the seven compounds extracted from PAC The proposed method was applied for the simultaneous determination of seven compounds in PAC from different locations of China under the optimized HPLC conditions. Nine different PAC samples were analyzed, and each sample was analyzed in duplicate to determine the mean content and the results were listed in Supplementary data, Table S1. The HPLC-DAD chromatograms of standards and samples were shown in Figure 2. And, the peaks that correspond to each chemical were separated. The amounts of the seven compounds in the nine PAC samples were found to be different. The variations in the contents of these main effective components might be result from these factors, such as cultivation year, geographical location and storage conditions. However, the contents of Pal and Ber in all investigated samples complied with the Chinese Pharmacopoeia (State Pharmacopoeia Commission of People’s Republic of China, 2010). These data indicated that the proposed HPLC-DAD method could be used for the simultaneous determination of the seven compounds in PAC. Fingerprint analysis of PAC The HPLC fingerprints of the samples were all generated at the UV absorption of 220 nm. Peaks that exist in all chromatograms
Table I Regression data, LODs and LOQs for the components determined (n ¼ 6) Components
Monitoring wavelength (nm)
Regression equation
Correlation coefficient (R 2)
Linear range (mg/mL)
LOD (mg/mL)
LOQ (mg/mL)
Phellodendrine Magnoflorine Jatrorrhizine Palmatine Berberine Obaculactone Obacunone
207 223 348 348 348 207 213
Y ¼ 139,038,302.1X þ 76,955 Y ¼ 74,938,711.1X-13,218 Y ¼ 49,005,312.9X-456 Y ¼ 45,891,531.4X þ 31,385 Y ¼ 39,582,070.7X þ 124,575 Y ¼ 7,551,783X-1,274 Y ¼ 2,466,601.2X þ 8,707
0.9995 0.9997 0.9993 0.9998 0.9995 0.9990 0.9998
3.55 –35.5 9.78 –97.8 0.579 8 –5.798 7.25 –72.5 25.2 –252 14.66 –146.6 5.738 –57.38
0.27 0.14 0.08 0.43 1.01 0.44 0.49
1.06 0.42 0.28 1.45 3.78 1.47 1.72
Figure 2. The HPLC-DAD chromatograms of mixed standards (A) and PAC samples at 207 (B), 213 (C), 223 (D) and 348 nm (E), respectively. (1) Phellodendrine (28.6 min); (2) magnoflorine (33.5 min); (3) jatrorrhizine (48.3 min); (4) palmatine (51.6 min); (5) berberine (52.3 min); (6) obaculactone (59.8 min) and (7) obacunone (62.1 min).
Fingerprinting and Determination of Alkaloids and Limonins in Phellodendri Amurensis Cortex 163
of the samples were assigned as the common peak. A total of 21 common peaks were found in the chromatograms among all 11 batches of PAC samples, and seven peaks were identified by comparing their retention times and UV spectra (Figure 3) with those of the corresponding standards, the peaks of number 7, 9, 14, 15, 16, 17 and 19 were corresponding to compound Phe, Mag, Jat, Pal, Ber, Ob1 and Ob2, respectively (Figure 4). The correlation coefficients and cosine of the angle of the samples were calculated as the similarity evaluation of the fingerprints. The average chromatogram of the 11 batches of PAC samples was assumed to be the common pattern of PAC. So, the similarity analysis was carried out by comparing with the common pattern, and the results are shown in Supplementary data, Table S2. The similarity values of 11 samples were . 0.93, meaning that samples from different batches were similar in general. Hierarchical clustering analysis was used to sort samples into groups following the basis of 21 peak values of the HPLC fingerprints. In this study, different samples of PAC were analyzed by using the SPSS 19.0 software, the between-group linkage method was applied and Euclidean distance as a measure to classify the 11 samples, and they were grouped as in Figure 5. These samples were divided into four main clusters, that is, Group A (H2, J2, H5), Group B (H1, J3), Group C (H4, J1, L1, L2, H3) and Group D (L3), based on the absorption intensity of common peaks, Groups B and D were thought to be quality product, Groups A and C were thought to be general quality.
Figure 3. The UV spectrum of seven components in PAC.
164 Wang et al.
Discussion In recent years, HPLC-DAD has become a convenient, frequently used, and powerful tool for the identification and qualification of herbal medicines. The paper integrated the quantification of multi-ingredients and chromatographic fingerprint for the quality control of PAC. For further analyzing, principal component analysis was performed based on peak areas of 21 common chromatographic peaks from the HPLC fingerprints. The samples were clustered into four domains. With each domain being formed by samples H4, J1, L1, L2, H3, samples H2, J2, H5, samples H1, J3, and sample L3, respectively. The results of the PCA were shown in Figure 6. Here, the grouping was identical to that of the analysis by HCA. On the basis of the results above, similarity analysis, HCA and PCA, could be combined for identifying and controlling the quality evaluation of PAC. In this research, a simple and reliable analytical method was developed for the simultaneous analysis of seven major compounds in PAC, namely Phe, Mag, Jat, Pal, Ber, Ob1 and Ob2. The fingerprints of PAC were also established. Eleven batches of PAC from three genuine producing areas were assessed by chromatographic fingerprint analysis with similarity analysis, hierarchical clustering analysis and principal component analysis. The method obtained good linearity, precision, repeatability, stability and recovery. Thus, this study provided an example for quality evaluation of PAC by using HPLC-DAD coupled with multiple compounds determination and HPLC fingerprint.
Figure 4. (A) The HPLC chromatogram of mixed standards; (B) common pattern of HPLC fingerprint of PAC sample and (C) HPLC characteristic fingerprints of 11 PAC samples. All chromatograms were detected at the wavelength of 220 nm. The peaks marked with 1– 21 in the chromatogram represent 21 common peaks.
81173500, 81102556, 81373930, 81302905, 90709019), National Key Technology Research and Development Program of the Ministry of Science and Technology of China (grant no. 2011BAI03B03, 2011BAI03B06, 2011BAI03B08), Key Science and Technology Program of Heilongjiang Province, China (grant no. GC06C501, GA08C303, GA06C30101), Foundation of Heilongjiang University of Chinese Medicine (grant no. 201209) and Key Project of Chinese Ministry of Education (grant no. 212044). Conflict of Interest statement: The authors have declared that they have no competing interests.
Figure 5. The cluster analysis results of 11 batches of PAC.
References
Figure 6. Score plot obtained from the PCA of 11 PAC samples.
Supplementary data Supplementary data are available at Journal of Chromatographic Science online.
Funding This work was supported by grants from the Key Program of Natural Science Foundation of State (grant no. 81202639,
1. James, M.A., Fu, H., Liu, Y., Chen, D.R., You, M.; Dietary administration of berberine or Phellodendron amurense extract inhibits cell cycle progression and lung tumorigenesis; Molecular Carcinogenesis, (2011); 50: 1–7. 2. Kumar, R., Das, M., Ansari, K.M.; Nexrutine(R) inhibits tumorigenesis in mouse skin and induces apoptotic cell death in human squamous carcinoma A431 and human melanoma A375 cells; Carcinogenesis, (2012); 33: 1909–1918. 3. Kim, J.H., Huh, J.E., Baek, Y.H., Lee, J.D., Choi, D.Y., Park, D.S.; Effect of Phellodendron amurense in protecting human osteoarthritic cartilage and chondrocytes; Journal of Ethnopharmacology, (2011); 134: 234–242. 4. Wang, W., Zu, Y., Fu, Y., Reichling, J., Suschke, U., Nokemper, S., et al.; In vitro antioxidant, antimicrobial and anti-herpes simplex virus type 1 activity of Phellodendron amurense Rupr. from China; The American Journal of Chinese Medicine, (2009); 37: 195– 203. 5. Xian, Y.F., Lin, Z.X., Ip, S.P., Su, Z.R., Chen, J.N., Lai, X.P.; Comparison the neuropreotective effect of Cortex Phellodendri Chinensis and Cortex Phellodendri Amurensis against beta-amyloid-induced neurotoxicity in PC12 cells; Phytomedicine, (2013); 20: 187– 193. 6. Kumar, A.P., Graham, H., Robson, C., Thompson, I.M., Ghosh, R.; Natural products: potential for developing Phellodendron amurense bark extract for prostate cancer management; Mini Reviews in Medicinal Chemistry, (2010); 10: 388–397. 7. Chan, C.-O., Chu, C.-C., Mok, D.K., Chau, F.-T.; Analysis of berberine and total alkaloid content in Cortex Phellodendri by near infrared spectroscopy (NIRS) compared with high-performance liquid chromatography coupled with ultra-visible spectrometric detection; Analytica Chimica Acta, (2007); 592: 121– 131.
Fingerprinting and Determination of Alkaloids and Limonins in Phellodendri Amurensis Cortex 165
8. Wang, H., Yan, G., Zhang, A., Li, Y., Wang, Y., Sun, H., et al.; Rapid discovery and global characterization of chemical constituents and rats metabolites of Phellodendri amurensis cortex by ultra-performance liquid chromatography-electrospray ionization /quadrupole – timeof-flight mass spectrometry coupled with pattern recognition approach; Analyst, (2013); 138: 3303– 3312. 9. Tang, X., Huang, Z., Chen, Y., Liu, Y., Liu, Y., Zhao, J., et al.; Simultaneous determination of six bioactive compounds in evodiae fructus by high-performance liquid chromatography with diode array detection; Journal of Chromatographic Science, (2014); 52: 149–156.
166 Wang et al.
10. Mayer, M., Benko, A., Husza´r, A., Sipos, K., Lajtai, A., Lakatos, A., et al.; Simultaneous determination of 4-substituted cathinones (4-MMC, 4-MEC and 4-FMC) in human urine by HPLC-DAD; Journal of Chromatographic Science, (2013); 51: 861–866. 11. Branquinho, R.T., Mosqueira, V.C., Kano, E.K., de Souza, J., Dorim, D.D., Sau´de-Guimara˜es, D.A., et al.; HPLC-DAD and UVspectrophotometry for the determination of lychnopholide in nanocapsule dosage form: validation and application to release kinetic study; Journal of Chromatographic Science, (2014); 52: 19– 26.
