515
J. Sep. Sci. 2014, 37, 515–526
Liang Zhao1 Shanshan Liang1,2∗ Lei Lv1 Hai Zhang1 Guang guo-Tan3 Yifeng Chai3 Guoqing Zhang1 1 Department
of Pharmacy, Eastern Hepatobiliary Surgery Hospital, Shanghai, China 2 Instrumental Analysis & Research Center, Shanghai Institute of Pharmaceutical Industry, Shanghai, China 3 School of Pharmacy, Second Military Medical University, Shanghai, China Received September 21, 2013 Revised December 11, 2013 Accepted December 13, 2013
Research Article
Screening and analysis of metabolites in rat urine after oral administration of Apocynum venetum L. extracts using HPLC–TOF-MS HPLC with diode array detection and ESI-TOF-MS was used for the study of the constituents in Apocynum venetum L. extracts and the metabolites in rat urine after oral administration of A. venetum L. extracts. A formula database of the known constituents in A. venetum L. was established, and 21 constituents were rapidly identified by accurately matching their molecular masses with the formulae of the compounds in the database. Furthermore, 34 metabolites were detected and elucidated in the rat urine. The scientific and plausible biotransformation pathways of the flavonoid components in A. venetum L. were also proposed together with the presentation of clues for potential mechanisms of bioactivity. This specific and sensitive HPLC–ESI-TOF-MS method can be used to identify the chemical components in the extracts of A. venetum L. and their metabolites in rat urine. This method can also be used to reveal the possible metabolic mechanisms of action of the extract components in vivo. Keywords: Apocynum venetum L. / HPLC–TOF-MS / Identification / Metabolites / Rat urine DOI 10.1002/jssc.201301036
1 Introduction The dried leaves of Apocynum venetum L. (Apocynaceae family), referred to as Luobuma (LBM) in China, are among the most popular traditional Chinese herbal medicines (TCHMs). Currently, LBM is not only used as an effective medicine but also as an important health tea in people’s daily lives [1]. Previous pharmacological research has demonstrated that LBM has a wide variety of activities, including the lowering of blood pressure [2] and antidepressant [3], antinephritis [4], and antineurasthenia [5] effects. Recently, increasing attention has been devoted to metabolic studies of TCHMs as they help to elucidate drug efficacy [6], screen for bioactive constituents [7], evaluate potential toxic metabolites [8], clarify herb–drug interactions [9], and identify drug degradation pathways [10]. In the past few decades, several analytical methods, such as LC–MS [11], NMR spectroscopy [12], countercurrent chromatography [13, 14], and microdialysis [15], have been used for screening the bioactive constituents in LBM. However, little attention has been devoted to the metabolites of LBM in vivo. With its widespread usage and an increasing knowledge about the efficacy of LBM on human health, it is urgent Correspondence: Professor Guoqing Zhang, Department of Pharmacy, Eastern Hepatobiliary Surgery Hospital, Second Military Medical University, Shanghai 200438, People’s Republic of China E-mail: [email protected]
Abbreviations: LBM, Luobuma; RT, retention time; TCHM, traditional Chinese herbal medicine; TIC, total ion chromatogram C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
to gain insight into the metabolic fates and metabolites of the constituents in LBM. A method that could simultaneously identify most of the components and their metabolites of this crude herb is urgently needed. It is known that the in vivo biotransformation of components in TCHMs is complex, not only because of their complicated parent constituents but also because of various reactions, such as reduction, cyclization, glucuronide conjugation, and sulfation [16–18]. In addition, many metabolites, which are biomarkers of interest, will often be present only in low abundances, and ion suppression from some endogenous metabolites of high abundance may prevent their response [19]. Therefore, the comprehensive detection and qualification of metabolites undoubtedly and thoroughly from a biological sample is a major challenge in current metabolomic research. Fortunately, the development of modern analytical technology can overcome these limitations and provide an efficient way to perform metabolomics research on TCHMs. HPLC coupled with diode array detection or MS has often been used for the determination of the constituents in TCHMs and their metabolites in biological samples such as plasma, bile, urine, and other tissues [20]. The HPLC coupled with diode array detection method can only be used to determine the constituents and their metabolites with standards, which limits its usage [21–25]. HPLC–MS provides more information on molecular formulae and fragment ions for the identification and quantification of the constituents and their metabolites [26]. HPLC–TOF-MS, which can provide accurate ∗ This
author contributed equally to this work.
www.jss-journal.com
516
J. Sep. Sci. 2014, 37, 515–526
L. Zhao et al.
mass measurements and formulae of nontarget compounds, is able to tentatively identify nontarget compounds and provide valuable structural insights for the characterization of drug metabolites [27, 28]. Now, it has been successfully used for the separation and identification of metabolites in vivo after the administration of TCHMs extract [29, 30]. In this work, we established a simple, selective, and specific method using HPLC–TOF-MS for the screening of the major constituents and analysis of their metabolism in rat urine after LBM extract was orally administered to rats. This method can provide more comprehensive insights into the preclinical pharmacokinetic studies and a better understanding of the mechanism of the pharmacological action of LBM.
2 Materials and methods 2.1 Chemicals and reagents Chlorogenic acid, rutin, hyperoside, isoquercetin, and quercetin standards were purchased from the National Institute for the Control of Pharmaceutical & Biological Products (NICPB, Beijing, China). Their purity was shown to be over 99.0%. The crude herbs of LBM were purchased from Leiyunshan Medicine Corporation (Shanghai, China) in 2010 and were identified by Professor Lianna Sun (Department of Pharmacognosy, School of Pharmacy, Second Military Medical University, Shanghai, China). Acetonitrile (Merker, Germany) and methanol (Merker) were of HPLC grade, and ultrapure water was prepared using a Milli-Q system (Millipore, Bedford, MA, USA). All of the other reagents were of analytical grade. 2.2 Animals and treatments All of the animal studies followed the relevant national legislation and local guidelines. The permission number of the animal experiments was SCXK (Hu) 2007-0005. Five male Sprague–Dawley rats, weighing 205.3 ± 10.9 g, were purchased from Shanghai Laboratory Animal (SLAC, Shanghai, China) and maintained under a standard 12 h light–dark cycle with water and food provided ad libitum. The standard diet was withheld for one day before the experiment, however, water was provided freely. 2.3 Sample preparation 2.3.1 Preparation of the LBM extracts Dry leaves of LBM were powdered to a homogeneous particle size with a DFT-200 mill (Wenzhou, Zhejiang, China) and were passed through a 40 mesh sieve before extraction. Aliquots of 100 g of powder were weighed and soaked with 1.0 L of deionized water for 60 min. Then, the decoction pieces were extracted in 70% ethanol for 1 h using a SK2200H sonicator (59 kHz, 90 W; KUDOS, Shanghai, China), and the C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
extraction solution was filtered through a funnel with four layers of gauze. Then, the drugs were extracted once again for 60 min with 70% ethanol, and the extraction solution was filtered again with the method described above. Afterwards, the two successive decoctions were merged and evaporated to 100 mL. Next, 200 mL of ethanol was added to the residue with an extracted solution/ethanol ratio of 1:2 v/v to precipitate the polysaccharides and proteins. The mixture was left to sit overnight at 4⬚C and filtered through three layers of filter paper the following day. The filtered solution was evaporated to dryness under reduced pressure with a rotary evaporator at 60⬚C. The residue was dissolved in deionized water to obtain a solution with the concentration of 3.6 g (decoction pieces)/mL. 2.3.2 Urine sample preparation LBM extract (36 g/kg) was administered by gavage to five rats, and urine (0–12 h) was collected from rats individually housed in stainless-steel wire-mesh cages. Control urine samples were similarly collected from each rat at least 12 h prior to the administration of the LBM extract. All of the urine samples were pretreated by SPE before the LC–MS analysis. Supelclean LC-18 SPE columns (1 mL/100 mg volume, Supelco, USA) were first preconditioned with 3 mL of methanol and were then equilibrated with 3 mL of deionized water. The urine samples (1 mL) were loaded onto the preconditioned SPE columns directly after being washed with 6 mL of deionized water. The SPE columns were eluted using 1 mL of 8:2 methanol/water v/v solution, and then the eluted solution was evaporated to dryness at 45⬚C under a nitrogen stream. The residues were dissolved in 200 L of the mobile phase and vortexed for 1 min followed by centrifugation at 14 000 × g at 4⬚C for 10 min. A 5 L aliquot of the supernatant was automatically injected into the LC–ESI-TOF-MS system.
2.4 LC–TOF-MS analysis The 5 L aliquots of each sample were separated on a Shiseido Capcell Pak C18 MG column (3.0 × 100 mm, 3.0 m) at 30⬚C using an Agilent 1100 series LC system (Santa Clara, CA, USA) with a gradient mobile phase composed of 0.1% formic acid (A) and acetonitrile containing 0.1% formic acid (B). A linear gradient was optimized as follows: 0–5 min, 5–10% A; 5–14 min, 10–15% A; 14–19 min, 15–24% A; 19–24 min, 24-26% A; 24–28 min 26–40% A; and 28–32 min, 40–90% A. Afterwards, the column was rinsed for 5 min with 90% A, and the content of A was lowered to 10% over 2 min. Then, the column was reequilibrated for 10 min. The flow rate was kept at 0.8 mL/min, and a postcolumn split was used to maintain a flow rate of 0.2 mL/min into the mass spectrometer source to obtain good nebulization efficiency. MS was performed on an Agilent 6220 TOF–MS operating in the positive ion mode, and the operating software was MassHunter Workstation Software (version B.02.00). The capillary and skimmer voltages were set to 4000 and 60 V, respectively. www.jss-journal.com
J. Sep. Sci. 2014, 37, 515–526
Liquid Chromatography
517
Figure 1. TOF–MS spectra in the positive ion mode, (A) LBM extracts; (B) drug-containing urine; (C) blank urine.
The gas temperature was 350⬚C, the drying gas flow rate was 9 L/min, and the nebulizer was 40 psig. The MS spectra were acquired in the full scan analysis mode over a m/z range of 100–1000 using an extended dynamic range and were stored in the centroid mode. The fragmentor voltages were set at 120, 300, and 400 V for the metabolic profile,fingerprint acquisition, and identification of interesting components, respectively. To maintain mass accuracy during the run-time, a reference mass solution, which contained reference ions with m/z values of 121.0508 and 922.0097, was used in the positive ionization mode.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2.5 Construction of the LBM chemical library The library was created using the Formula-DatabaseGenerator Agilent software. The library contains one table with ten searchable fields, including structure, formula, accurate mass, name, Chinese name, Chemical Abstracts service registry number, UV spectral data, mass spectral data, references, and notes. Records of 43 compounds were put into the library according to the phytochemical and pharmacological literature of LBM and the Combined Chemical Dictionary.
www.jss-journal.com
518
J. Sep. Sci. 2014, 37, 515–526
L. Zhao et al.
Table 1. HPLC–ESI-TOF-MS accurate masses of [M+H]+ and [M+Na]+ ions of constituents in the LBM extract
No.
Compound
RT (min)
C1
Gallocatechin or epigallocatechin
6.928
C2
Chlorogenic acid
7.727
C3
9.438
C4
Quercetin-3-O--D-glucosyl-D-glucopyranoside Catechin or epicatechin
10.845
C5
Scopoletin
14.542
C6 C7
Myricetin Rutin
15.976 18.023
C8
Hyperoside
19.030
C9
Isoquercetin
19.046
C10
Trifolin
20.165
C11
Malonated hyperoside
20.261
C12
Astragalin
20.804
C13
Acetylated hyperoside
21.204
C14
Malonated isoquercetin
21.396
C15
Malonated trifolin
21.572
C16
Acetylated isoquercetin
21.828
C17
Malonated astragalin
22.227
C18
Acetylated trifolin
23.634
C19
Quercetin
26.528
C20
Cymarin
28.670
C21
Kaempferol
29.405
Experimental m/z
M+X
Formula
Calculated m/z
ppm error
307.0820 329.0641 355.1035 377.0853 731.1823 627.1545 649.1360 291.0877 313.0671 193.0487 215.0323 319.0435 611.1608 633.1423 465.1046 487.0869 951.1755 465.1050 487.0870 951.1851 449.1079 471.0901 919.1934 551.1050 573.0826 449.1064 471.0883 919.1920 507.1123 529.0955 551.1048 573.0826 535.1066 557.0924 507.1113 529.0939 535.1082 557.0934 491.1174 513.0984 303.0502 325.0328 549.3050 571.2873 287.0539 309.0372
M+H M+Na M+H M+Na 2M+Na M+H M+Na M+H M+Na M+H M+Na M+H M+H M+Na M+H M+Na 2M+Na M+H M+Na 2M+Na M+H M+Na 2M+Na M+H M+Na M+H M+Na 2M+Na M+H M+Na M+H M+Na M+H M+Na M+H M+Na M+H M+Na M+H M+Na M+H M+Na M+H M+Na M+H M+Na
C15 H15 O11 C15 H14 O11 Na C16 H19 O9 C16 H18 O9 Na C32 H36 O18 Na C27 H31 O17 C27 H30 O17 Na C15 H15 O6 C15 H14 O6 Na C10 H9 O4 C10 H8 O4 Na C15 H11 O8 C27 H31 O16 C27 H30 O16 Na C21 H21 O12 C21 H20 O12 Na C42 H40 O24 Na C21 H21 O12 C21 H20 O12 Na C42 H40 O24 Na C21 H21 O11 C21 H20 O11 Na C42 H40 O22 Na C21 H23 O15 C24 H22 O15 Na C21 H21 O11 C21 H20 O11 Na C42 H40 O22 Na C23 H23 O13 C23 H22 O13 Na C21 H23 O15 C24 H22 O15 Na C24 H23 O14 C24 H22 O14 Na C23 H23 O13 C23 H22 O13 Na C24 H23 O14 C24 H22 O14 Na C23 H23 O11 C23 H22 O11 Na C15 H11 O7 C15 H10 O7 Na C30 H45 O9 C30 H44 O9 Na C15 H11 O6 C15 H10 O6 Na
307.0812 329.0632 355.1024 377.0843 731.1794 627.1556 649.1376 291.0863 313.0683 193.0495 215.0315 319.0448 611.1607 633.1426 465.1028 487.0847 951.1802 465.1028 487.0847 951.1802 449.1078 471.0898 919.1903 551.1031 573.0851 449.1078 471.0898 919.1903 507.1133 529.0953 551.1031 573.0851 535.1082 557.0902 507.1133 529.0953 535.1100 557.0911 491.1184 513.1003 303.0499 325.0319 549.3058 571.2878 287.0550 309.0370
2.61 2.74 3.10 2.65 3.97 −1.75 −2.46 4.81 −3.83 −4.14 3.72 −4.07 0.16 −0.47 3.87 4.52 −4.94 4.73 4.72 5.15 0.22 0.64 3.37 3.45 −4.36 −3.12 −3.18 1.85 −1.97 0.38 3.08 −4.36 −2.99 3.95 −3.94 −2.65 −3.36 4.13 −2.04 −3.70 0.99 2.77 −1.46 −0.88 −3.83 0.65
3 Results and discussion 3.1 Optimization of the HPLC and TOF–MS conditions and the identification of compounds in the LBM extracts Given the acidity of flavonoid compounds, it was found that good chromatographic behavior could be achieved after opti C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
mizing the mobile phase. The chromatographic conditions, especially the composition of the mobile phase, were optimized through several trials to achieve good resolution and symmetrical peak shapes for the analytes as well as a short run-time. Modifiers, such as formic acid alone or in combination with other additives, in different concentrations were added. It was found that a mixture of 0.1% formic acid and acetonitrile containing 0.1% formic acid could achieve this www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2014, 37, 515–526
519
Figure 2. The structure of the chemical compounds in the extracts of Apocynum venetum L.
purpose, and this mixture was finally adopted as the mobile phase. The percentage of formic acid was optimized to maintain symmetrical peak shapes while still being consistent with good ionization and fragmentation in the mass spectrometer. The retention time (RT) was somewhat affected by temperature, and the best separation was achieved at 30⬚C. It was C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
also found that the best separation was achieved at a flow rate of 0.8 mL/min and a detection wavelength of 360 nm. To acquire the maximum sensitivity for most compounds, TOF-MS parameters such as the capillary voltage, nebulizer gas pressure, drying gas flow rate, gas temperature, fragmentor voltage, and skimmer voltage were optimized. It was found that the sensitivity in the positive ion mode was five- to
www.jss-journal.com
520
J. Sep. Sci. 2014, 37, 515–526
L. Zhao et al.
Table 2. HPLC–ESI-TOF-MS accurate masses of [M+H]+ and [M+Na]+ ions of the parent compounds in rat urine
No.
P1 P2 P3 P4 P5 P6 P7 P8
RT (min)
7.70 14.54 19.03 19.05 20.17 20.80 26.53 29.41
[M+H]+ m/z Detected
Expected
Error (ppm)
355.1035 193.0503 465.1050 465.1049 449.1097 449.1061 303.0488 287.0558
355.1024 193.0495 465.1028 465.1028 449.1078 449.1078 303.0499 287.0550
3.10 4.14 4.73 4.52 4.23 −3.79 −3.63 2.79
tenfold more than that in the negative ion mode for most of the compounds with TOF-MS. The fragmentor voltage in TOF-MS is of crucial importance for the efficient transmission of the ion to obtain the best balance between the sensitivity and ion fragmentation for the structural identification of LC-separated biomolecules [31, 32]. Adjustment of the fragmentor voltage can provide characteristic fragment ions in the MS spectra resulting from in-source collision-induced dissociation. Typically, a voltage setting of approximately 120 V or lower provides minimal fragmentation and maximum molecular ion intensity of most of the components, while fragmentor voltages of 300 and 400 V lead to extensive fragmentation. For the analysis of LBM in rat urine, the metabolic profile of the urine was obtained at a low fragmentor voltage (120 V), which enabled not only the sensitive detection of the metabolites but also as much of a reduction in the excess information from fragmentation as possible. However, high fragmentor voltages (300 and 400 V) were used to acquire characteristic fragment ions to help in the identification of interesting components.
3.2 Means of detection and identification of parent compounds and metabolites By a comparison of the total ion chromatogram (TIC) of postdose rat urine, predose rat urine, and the extracts of LBM by LC–ESI-TOF-MS in the positive ion mode (Fig. 1), a rat urinary metabolite profile was obtained and analyzed. The extracts of LBM were directly injected into the TOF mass spectrometer to optimize the ESI conditions. Their mass spectra acquired by HPLC–TOF-MS in the positive ion mode are shown in Fig. 1A. The manifested ion patterns in the spectra were [M+H]+ and [M+Na]+ , which enabled us to discover more about the molecular weights and molecular formulae of the compounds. Twenty-one compounds in the TIC from TOF–MS were characterized by searching against the formula database established by the Agilent software, 16 of which were flavonoid compounds. Five of these were unambiguously identified by comparison with their related RT, characteristic UV absorption, and mass spectra according to chemical standards of chlorogenic acid, rutin, hyperoside, C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[M+Na]+ m/z
Formula
Compound name
377.0843 215.0315 487.0847 487.0847 471.0898 471.0898 325.0453 309.0476
C16 H18 O9 C10 H8 O4 C21 H20 O12 C21 H20 O12 C21 H20 O11 C21 H20 O11 C15 H10 O7 C15 H10 O6
Chlorogenic acid Scopoletin Hyperoside Isoquercetin Trifolin Astragalin Quercetin Kaempferol
isoquercetin, and quercetin. These compounds can be identified according to calculated formulae, obtained diagnostic ions, and fragmentation information. The protonated molecular weights of all of the target compounds were calculated within an error of 5 ppm. The accuracy of the mass determination results of the compounds in the LBM extract are shown in Table 1, and their structures are shown in Fig. 2. Based on the means applied for the identification of compounds in the extracts of LBM summarized above, ten parent compounds were tentatively identified in the positive ion mode, four of which were unambiguously identified by comparison of their RTs and TOF-MS data with those of standards, which were chlorogenic acid, hyperoside, isoquercetin, and quercetin. The accuracy of the mass determination results of the protonated molecules of the parent constituents in the positive ion mode are summarized in Table 2. It is known that discriminating ions of metabolites from the endogenous compounds is very difficult because of endogenous interferences from complex biological matrices. Moreover, on account of the complex constituents and many unknown compounds, no suitable software is available for the mass spectrometric analysis of drug-related metabolites. However, with the high selectivity of the extracted ion chromatogram, the metabolites of LBM in rat urine showed high specificities. By comparing the TIC of postdose rat urine (Fig. 1B) with that of predose rat urine (Fig. 1C) to discriminate endogenous interferences from drug-related metabolites, 34 peaks (M1–M34) were tentatively predicted to be the metabolites of LBM (Table 3). For the identification of the metabolites, the most probable molecular formulae of the metabolites were determined using different criteria including a mass accuracy
J. Sep. Sci. 2014, 37, 515–526
Liang Zhao1 Shanshan Liang1,2∗ Lei Lv1 Hai Zhang1 Guang guo-Tan3 Yifeng Chai3 Guoqing Zhang1 1 Department
of Pharmacy, Eastern Hepatobiliary Surgery Hospital, Shanghai, China 2 Instrumental Analysis & Research Center, Shanghai Institute of Pharmaceutical Industry, Shanghai, China 3 School of Pharmacy, Second Military Medical University, Shanghai, China Received September 21, 2013 Revised December 11, 2013 Accepted December 13, 2013
Research Article
Screening and analysis of metabolites in rat urine after oral administration of Apocynum venetum L. extracts using HPLC–TOF-MS HPLC with diode array detection and ESI-TOF-MS was used for the study of the constituents in Apocynum venetum L. extracts and the metabolites in rat urine after oral administration of A. venetum L. extracts. A formula database of the known constituents in A. venetum L. was established, and 21 constituents were rapidly identified by accurately matching their molecular masses with the formulae of the compounds in the database. Furthermore, 34 metabolites were detected and elucidated in the rat urine. The scientific and plausible biotransformation pathways of the flavonoid components in A. venetum L. were also proposed together with the presentation of clues for potential mechanisms of bioactivity. This specific and sensitive HPLC–ESI-TOF-MS method can be used to identify the chemical components in the extracts of A. venetum L. and their metabolites in rat urine. This method can also be used to reveal the possible metabolic mechanisms of action of the extract components in vivo. Keywords: Apocynum venetum L. / HPLC–TOF-MS / Identification / Metabolites / Rat urine DOI 10.1002/jssc.201301036
1 Introduction The dried leaves of Apocynum venetum L. (Apocynaceae family), referred to as Luobuma (LBM) in China, are among the most popular traditional Chinese herbal medicines (TCHMs). Currently, LBM is not only used as an effective medicine but also as an important health tea in people’s daily lives [1]. Previous pharmacological research has demonstrated that LBM has a wide variety of activities, including the lowering of blood pressure [2] and antidepressant [3], antinephritis [4], and antineurasthenia [5] effects. Recently, increasing attention has been devoted to metabolic studies of TCHMs as they help to elucidate drug efficacy [6], screen for bioactive constituents [7], evaluate potential toxic metabolites [8], clarify herb–drug interactions [9], and identify drug degradation pathways [10]. In the past few decades, several analytical methods, such as LC–MS [11], NMR spectroscopy [12], countercurrent chromatography [13, 14], and microdialysis [15], have been used for screening the bioactive constituents in LBM. However, little attention has been devoted to the metabolites of LBM in vivo. With its widespread usage and an increasing knowledge about the efficacy of LBM on human health, it is urgent Correspondence: Professor Guoqing Zhang, Department of Pharmacy, Eastern Hepatobiliary Surgery Hospital, Second Military Medical University, Shanghai 200438, People’s Republic of China E-mail: [email protected]
Abbreviations: LBM, Luobuma; RT, retention time; TCHM, traditional Chinese herbal medicine; TIC, total ion chromatogram C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
to gain insight into the metabolic fates and metabolites of the constituents in LBM. A method that could simultaneously identify most of the components and their metabolites of this crude herb is urgently needed. It is known that the in vivo biotransformation of components in TCHMs is complex, not only because of their complicated parent constituents but also because of various reactions, such as reduction, cyclization, glucuronide conjugation, and sulfation [16–18]. In addition, many metabolites, which are biomarkers of interest, will often be present only in low abundances, and ion suppression from some endogenous metabolites of high abundance may prevent their response [19]. Therefore, the comprehensive detection and qualification of metabolites undoubtedly and thoroughly from a biological sample is a major challenge in current metabolomic research. Fortunately, the development of modern analytical technology can overcome these limitations and provide an efficient way to perform metabolomics research on TCHMs. HPLC coupled with diode array detection or MS has often been used for the determination of the constituents in TCHMs and their metabolites in biological samples such as plasma, bile, urine, and other tissues [20]. The HPLC coupled with diode array detection method can only be used to determine the constituents and their metabolites with standards, which limits its usage [21–25]. HPLC–MS provides more information on molecular formulae and fragment ions for the identification and quantification of the constituents and their metabolites [26]. HPLC–TOF-MS, which can provide accurate ∗ This
author contributed equally to this work.
www.jss-journal.com
516
J. Sep. Sci. 2014, 37, 515–526
L. Zhao et al.
mass measurements and formulae of nontarget compounds, is able to tentatively identify nontarget compounds and provide valuable structural insights for the characterization of drug metabolites [27, 28]. Now, it has been successfully used for the separation and identification of metabolites in vivo after the administration of TCHMs extract [29, 30]. In this work, we established a simple, selective, and specific method using HPLC–TOF-MS for the screening of the major constituents and analysis of their metabolism in rat urine after LBM extract was orally administered to rats. This method can provide more comprehensive insights into the preclinical pharmacokinetic studies and a better understanding of the mechanism of the pharmacological action of LBM.
2 Materials and methods 2.1 Chemicals and reagents Chlorogenic acid, rutin, hyperoside, isoquercetin, and quercetin standards were purchased from the National Institute for the Control of Pharmaceutical & Biological Products (NICPB, Beijing, China). Their purity was shown to be over 99.0%. The crude herbs of LBM were purchased from Leiyunshan Medicine Corporation (Shanghai, China) in 2010 and were identified by Professor Lianna Sun (Department of Pharmacognosy, School of Pharmacy, Second Military Medical University, Shanghai, China). Acetonitrile (Merker, Germany) and methanol (Merker) were of HPLC grade, and ultrapure water was prepared using a Milli-Q system (Millipore, Bedford, MA, USA). All of the other reagents were of analytical grade. 2.2 Animals and treatments All of the animal studies followed the relevant national legislation and local guidelines. The permission number of the animal experiments was SCXK (Hu) 2007-0005. Five male Sprague–Dawley rats, weighing 205.3 ± 10.9 g, were purchased from Shanghai Laboratory Animal (SLAC, Shanghai, China) and maintained under a standard 12 h light–dark cycle with water and food provided ad libitum. The standard diet was withheld for one day before the experiment, however, water was provided freely. 2.3 Sample preparation 2.3.1 Preparation of the LBM extracts Dry leaves of LBM were powdered to a homogeneous particle size with a DFT-200 mill (Wenzhou, Zhejiang, China) and were passed through a 40 mesh sieve before extraction. Aliquots of 100 g of powder were weighed and soaked with 1.0 L of deionized water for 60 min. Then, the decoction pieces were extracted in 70% ethanol for 1 h using a SK2200H sonicator (59 kHz, 90 W; KUDOS, Shanghai, China), and the C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
extraction solution was filtered through a funnel with four layers of gauze. Then, the drugs were extracted once again for 60 min with 70% ethanol, and the extraction solution was filtered again with the method described above. Afterwards, the two successive decoctions were merged and evaporated to 100 mL. Next, 200 mL of ethanol was added to the residue with an extracted solution/ethanol ratio of 1:2 v/v to precipitate the polysaccharides and proteins. The mixture was left to sit overnight at 4⬚C and filtered through three layers of filter paper the following day. The filtered solution was evaporated to dryness under reduced pressure with a rotary evaporator at 60⬚C. The residue was dissolved in deionized water to obtain a solution with the concentration of 3.6 g (decoction pieces)/mL. 2.3.2 Urine sample preparation LBM extract (36 g/kg) was administered by gavage to five rats, and urine (0–12 h) was collected from rats individually housed in stainless-steel wire-mesh cages. Control urine samples were similarly collected from each rat at least 12 h prior to the administration of the LBM extract. All of the urine samples were pretreated by SPE before the LC–MS analysis. Supelclean LC-18 SPE columns (1 mL/100 mg volume, Supelco, USA) were first preconditioned with 3 mL of methanol and were then equilibrated with 3 mL of deionized water. The urine samples (1 mL) were loaded onto the preconditioned SPE columns directly after being washed with 6 mL of deionized water. The SPE columns were eluted using 1 mL of 8:2 methanol/water v/v solution, and then the eluted solution was evaporated to dryness at 45⬚C under a nitrogen stream. The residues were dissolved in 200 L of the mobile phase and vortexed for 1 min followed by centrifugation at 14 000 × g at 4⬚C for 10 min. A 5 L aliquot of the supernatant was automatically injected into the LC–ESI-TOF-MS system.
2.4 LC–TOF-MS analysis The 5 L aliquots of each sample were separated on a Shiseido Capcell Pak C18 MG column (3.0 × 100 mm, 3.0 m) at 30⬚C using an Agilent 1100 series LC system (Santa Clara, CA, USA) with a gradient mobile phase composed of 0.1% formic acid (A) and acetonitrile containing 0.1% formic acid (B). A linear gradient was optimized as follows: 0–5 min, 5–10% A; 5–14 min, 10–15% A; 14–19 min, 15–24% A; 19–24 min, 24-26% A; 24–28 min 26–40% A; and 28–32 min, 40–90% A. Afterwards, the column was rinsed for 5 min with 90% A, and the content of A was lowered to 10% over 2 min. Then, the column was reequilibrated for 10 min. The flow rate was kept at 0.8 mL/min, and a postcolumn split was used to maintain a flow rate of 0.2 mL/min into the mass spectrometer source to obtain good nebulization efficiency. MS was performed on an Agilent 6220 TOF–MS operating in the positive ion mode, and the operating software was MassHunter Workstation Software (version B.02.00). The capillary and skimmer voltages were set to 4000 and 60 V, respectively. www.jss-journal.com
J. Sep. Sci. 2014, 37, 515–526
Liquid Chromatography
517
Figure 1. TOF–MS spectra in the positive ion mode, (A) LBM extracts; (B) drug-containing urine; (C) blank urine.
The gas temperature was 350⬚C, the drying gas flow rate was 9 L/min, and the nebulizer was 40 psig. The MS spectra were acquired in the full scan analysis mode over a m/z range of 100–1000 using an extended dynamic range and were stored in the centroid mode. The fragmentor voltages were set at 120, 300, and 400 V for the metabolic profile,fingerprint acquisition, and identification of interesting components, respectively. To maintain mass accuracy during the run-time, a reference mass solution, which contained reference ions with m/z values of 121.0508 and 922.0097, was used in the positive ionization mode.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2.5 Construction of the LBM chemical library The library was created using the Formula-DatabaseGenerator Agilent software. The library contains one table with ten searchable fields, including structure, formula, accurate mass, name, Chinese name, Chemical Abstracts service registry number, UV spectral data, mass spectral data, references, and notes. Records of 43 compounds were put into the library according to the phytochemical and pharmacological literature of LBM and the Combined Chemical Dictionary.
www.jss-journal.com
518
J. Sep. Sci. 2014, 37, 515–526
L. Zhao et al.
Table 1. HPLC–ESI-TOF-MS accurate masses of [M+H]+ and [M+Na]+ ions of constituents in the LBM extract
No.
Compound
RT (min)
C1
Gallocatechin or epigallocatechin
6.928
C2
Chlorogenic acid
7.727
C3
9.438
C4
Quercetin-3-O--D-glucosyl-D-glucopyranoside Catechin or epicatechin
10.845
C5
Scopoletin
14.542
C6 C7
Myricetin Rutin
15.976 18.023
C8
Hyperoside
19.030
C9
Isoquercetin
19.046
C10
Trifolin
20.165
C11
Malonated hyperoside
20.261
C12
Astragalin
20.804
C13
Acetylated hyperoside
21.204
C14
Malonated isoquercetin
21.396
C15
Malonated trifolin
21.572
C16
Acetylated isoquercetin
21.828
C17
Malonated astragalin
22.227
C18
Acetylated trifolin
23.634
C19
Quercetin
26.528
C20
Cymarin
28.670
C21
Kaempferol
29.405
Experimental m/z
M+X
Formula
Calculated m/z
ppm error
307.0820 329.0641 355.1035 377.0853 731.1823 627.1545 649.1360 291.0877 313.0671 193.0487 215.0323 319.0435 611.1608 633.1423 465.1046 487.0869 951.1755 465.1050 487.0870 951.1851 449.1079 471.0901 919.1934 551.1050 573.0826 449.1064 471.0883 919.1920 507.1123 529.0955 551.1048 573.0826 535.1066 557.0924 507.1113 529.0939 535.1082 557.0934 491.1174 513.0984 303.0502 325.0328 549.3050 571.2873 287.0539 309.0372
M+H M+Na M+H M+Na 2M+Na M+H M+Na M+H M+Na M+H M+Na M+H M+H M+Na M+H M+Na 2M+Na M+H M+Na 2M+Na M+H M+Na 2M+Na M+H M+Na M+H M+Na 2M+Na M+H M+Na M+H M+Na M+H M+Na M+H M+Na M+H M+Na M+H M+Na M+H M+Na M+H M+Na M+H M+Na
C15 H15 O11 C15 H14 O11 Na C16 H19 O9 C16 H18 O9 Na C32 H36 O18 Na C27 H31 O17 C27 H30 O17 Na C15 H15 O6 C15 H14 O6 Na C10 H9 O4 C10 H8 O4 Na C15 H11 O8 C27 H31 O16 C27 H30 O16 Na C21 H21 O12 C21 H20 O12 Na C42 H40 O24 Na C21 H21 O12 C21 H20 O12 Na C42 H40 O24 Na C21 H21 O11 C21 H20 O11 Na C42 H40 O22 Na C21 H23 O15 C24 H22 O15 Na C21 H21 O11 C21 H20 O11 Na C42 H40 O22 Na C23 H23 O13 C23 H22 O13 Na C21 H23 O15 C24 H22 O15 Na C24 H23 O14 C24 H22 O14 Na C23 H23 O13 C23 H22 O13 Na C24 H23 O14 C24 H22 O14 Na C23 H23 O11 C23 H22 O11 Na C15 H11 O7 C15 H10 O7 Na C30 H45 O9 C30 H44 O9 Na C15 H11 O6 C15 H10 O6 Na
307.0812 329.0632 355.1024 377.0843 731.1794 627.1556 649.1376 291.0863 313.0683 193.0495 215.0315 319.0448 611.1607 633.1426 465.1028 487.0847 951.1802 465.1028 487.0847 951.1802 449.1078 471.0898 919.1903 551.1031 573.0851 449.1078 471.0898 919.1903 507.1133 529.0953 551.1031 573.0851 535.1082 557.0902 507.1133 529.0953 535.1100 557.0911 491.1184 513.1003 303.0499 325.0319 549.3058 571.2878 287.0550 309.0370
2.61 2.74 3.10 2.65 3.97 −1.75 −2.46 4.81 −3.83 −4.14 3.72 −4.07 0.16 −0.47 3.87 4.52 −4.94 4.73 4.72 5.15 0.22 0.64 3.37 3.45 −4.36 −3.12 −3.18 1.85 −1.97 0.38 3.08 −4.36 −2.99 3.95 −3.94 −2.65 −3.36 4.13 −2.04 −3.70 0.99 2.77 −1.46 −0.88 −3.83 0.65
3 Results and discussion 3.1 Optimization of the HPLC and TOF–MS conditions and the identification of compounds in the LBM extracts Given the acidity of flavonoid compounds, it was found that good chromatographic behavior could be achieved after opti C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
mizing the mobile phase. The chromatographic conditions, especially the composition of the mobile phase, were optimized through several trials to achieve good resolution and symmetrical peak shapes for the analytes as well as a short run-time. Modifiers, such as formic acid alone or in combination with other additives, in different concentrations were added. It was found that a mixture of 0.1% formic acid and acetonitrile containing 0.1% formic acid could achieve this www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2014, 37, 515–526
519
Figure 2. The structure of the chemical compounds in the extracts of Apocynum venetum L.
purpose, and this mixture was finally adopted as the mobile phase. The percentage of formic acid was optimized to maintain symmetrical peak shapes while still being consistent with good ionization and fragmentation in the mass spectrometer. The retention time (RT) was somewhat affected by temperature, and the best separation was achieved at 30⬚C. It was C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
also found that the best separation was achieved at a flow rate of 0.8 mL/min and a detection wavelength of 360 nm. To acquire the maximum sensitivity for most compounds, TOF-MS parameters such as the capillary voltage, nebulizer gas pressure, drying gas flow rate, gas temperature, fragmentor voltage, and skimmer voltage were optimized. It was found that the sensitivity in the positive ion mode was five- to
www.jss-journal.com
520
J. Sep. Sci. 2014, 37, 515–526
L. Zhao et al.
Table 2. HPLC–ESI-TOF-MS accurate masses of [M+H]+ and [M+Na]+ ions of the parent compounds in rat urine
No.
P1 P2 P3 P4 P5 P6 P7 P8
RT (min)
7.70 14.54 19.03 19.05 20.17 20.80 26.53 29.41
[M+H]+ m/z Detected
Expected
Error (ppm)
355.1035 193.0503 465.1050 465.1049 449.1097 449.1061 303.0488 287.0558
355.1024 193.0495 465.1028 465.1028 449.1078 449.1078 303.0499 287.0550
3.10 4.14 4.73 4.52 4.23 −3.79 −3.63 2.79
tenfold more than that in the negative ion mode for most of the compounds with TOF-MS. The fragmentor voltage in TOF-MS is of crucial importance for the efficient transmission of the ion to obtain the best balance between the sensitivity and ion fragmentation for the structural identification of LC-separated biomolecules [31, 32]. Adjustment of the fragmentor voltage can provide characteristic fragment ions in the MS spectra resulting from in-source collision-induced dissociation. Typically, a voltage setting of approximately 120 V or lower provides minimal fragmentation and maximum molecular ion intensity of most of the components, while fragmentor voltages of 300 and 400 V lead to extensive fragmentation. For the analysis of LBM in rat urine, the metabolic profile of the urine was obtained at a low fragmentor voltage (120 V), which enabled not only the sensitive detection of the metabolites but also as much of a reduction in the excess information from fragmentation as possible. However, high fragmentor voltages (300 and 400 V) were used to acquire characteristic fragment ions to help in the identification of interesting components.
3.2 Means of detection and identification of parent compounds and metabolites By a comparison of the total ion chromatogram (TIC) of postdose rat urine, predose rat urine, and the extracts of LBM by LC–ESI-TOF-MS in the positive ion mode (Fig. 1), a rat urinary metabolite profile was obtained and analyzed. The extracts of LBM were directly injected into the TOF mass spectrometer to optimize the ESI conditions. Their mass spectra acquired by HPLC–TOF-MS in the positive ion mode are shown in Fig. 1A. The manifested ion patterns in the spectra were [M+H]+ and [M+Na]+ , which enabled us to discover more about the molecular weights and molecular formulae of the compounds. Twenty-one compounds in the TIC from TOF–MS were characterized by searching against the formula database established by the Agilent software, 16 of which were flavonoid compounds. Five of these were unambiguously identified by comparison with their related RT, characteristic UV absorption, and mass spectra according to chemical standards of chlorogenic acid, rutin, hyperoside, C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[M+Na]+ m/z
Formula
Compound name
377.0843 215.0315 487.0847 487.0847 471.0898 471.0898 325.0453 309.0476
C16 H18 O9 C10 H8 O4 C21 H20 O12 C21 H20 O12 C21 H20 O11 C21 H20 O11 C15 H10 O7 C15 H10 O6
Chlorogenic acid Scopoletin Hyperoside Isoquercetin Trifolin Astragalin Quercetin Kaempferol
isoquercetin, and quercetin. These compounds can be identified according to calculated formulae, obtained diagnostic ions, and fragmentation information. The protonated molecular weights of all of the target compounds were calculated within an error of 5 ppm. The accuracy of the mass determination results of the compounds in the LBM extract are shown in Table 1, and their structures are shown in Fig. 2. Based on the means applied for the identification of compounds in the extracts of LBM summarized above, ten parent compounds were tentatively identified in the positive ion mode, four of which were unambiguously identified by comparison of their RTs and TOF-MS data with those of standards, which were chlorogenic acid, hyperoside, isoquercetin, and quercetin. The accuracy of the mass determination results of the protonated molecules of the parent constituents in the positive ion mode are summarized in Table 2. It is known that discriminating ions of metabolites from the endogenous compounds is very difficult because of endogenous interferences from complex biological matrices. Moreover, on account of the complex constituents and many unknown compounds, no suitable software is available for the mass spectrometric analysis of drug-related metabolites. However, with the high selectivity of the extracted ion chromatogram, the metabolites of LBM in rat urine showed high specificities. By comparing the TIC of postdose rat urine (Fig. 1B) with that of predose rat urine (Fig. 1C) to discriminate endogenous interferences from drug-related metabolites, 34 peaks (M1–M34) were tentatively predicted to be the metabolites of LBM (Table 3). For the identification of the metabolites, the most probable molecular formulae of the metabolites were determined using different criteria including a mass accuracy
