LC/MS/MS for identification of in vivo and in vitro metabolites of jatrorrhizine

Copyright © 2008 John Wiley & Sons, Ltd. Biomed. Chromatogr. 22: 1360–1367 (2008)DOI: 10.1002/bmc

1360 Y. Zhang et al.ORIGINAL RESEARCH ORIGINAL RESEARCH

Copyright © 2008 John Wiley & Sons, Ltd.

BIOMEDICAL CHROMATOGRAPHYBiomed. Chromatogr. 22: 1360–1367 (2008)Published online 23 July 2008 in Wiley InterScience(www.interscience.wiley.com) DOI: 10.1002/bmc.1066

LC/MS/MS for identification of in vivo and in vitrometabolites of jatrorrhizine

Yi Zhang, Wenhua Wu, Fengmei Han and Yong Chen*

Hubei Province Key Laboratory of Biotechnology of Chinese Traditional Medicine, Hubei University, Wuhan 430062, People’s Republic of China

Received 31 December 2007; revised 17 March 2008; accepted 18 March 2008

ABSTRACT: The in vivo and in vitro metabolism of jatrorrhizine has been investigated using a specific and sensitive LC/MS/MSmethod. In vivo samples including rat feces, urine and plasma collected separately after dosing healthy rats with jatrorrhizine(34 mg/kg) orally, along with in vitro samples prepared by incubating jatrorrhizine with rat intestinal flora and liver microsome,respectively, were purified using a C18 solid-phase extraction cartridge. The purified samples were then separated with a reversed-phase C18 column with methanol–formic acid aqueous solution (70:30, v/v, pH3.5) as mobile phase and detected by on-line MS/MS. The structural elucidation of the metabolites was performed by comparing their molecular weights and product ions withthose of the parent drug. As a result, seven new metabolites were found in rat urine, 13 metabolites were detected in rat feces,11 metabolites were detected in rat plasma, 17 metabolites were identified in intestinal flora incubation solution and ninemetabolites were detected in liver microsome incubation solution. The main biotransformation reactions of jatrorrhizine werethe hydroxylation reaction, the methylation reaction, the demethylation reaction and the dehydrogenation reaction of parentdrug and its relative metabolites. All the results were reported for the first time, except for some of the metabolites in rat urine.Copyright © 2008 John Wiley & Sons, Ltd.

KEYWORDS: jatrorrhizine; metabolite; MS; MS/MS; HPLC

INTRODUCTION

Identification of metabolites plays a key role in drugmetabolic studies. In general, the content of metabo-lites in biological samples is very low, and the chemicalsynthesis of metabolites is time-consuming and difficult.Therefore, a sensitive and specific assay method is veryimportant for the identification of metabolites.

LC/MS/MS has been proved to be a modern power-ful tool for the identification of drug metabolites inbiological matrices owing to the high sensitivity andselectivity (Thevis et al., 2006; Chen et al., 2007; Chanet al., 2003), especially for the identification of ther-molabile, highly polar and non-volatile metabolites dueto the application of soft-ionization techniques such aselectrospray ionization (ESI) technique (Xing et al.,2007; Smyth, 2003, 2005). Because metabolites retainthe basic structure of the parent drug after biotrans-

formation, the mass spectral fragmentation behavior ofthe parent drug can be used as a structural template forinterpreting the structures of metabolites (Appolonovaet al., 2004; Chung et al., 2004).

Jatrorrhizine (Fig. 1), as one of the protoberberinealkaloids derived from the plants Coptidis Rhizoma andMahonia bealei Carr., has important physiological ac-tivities such as antibiotic and hypoglycemic activity (Fuet al., 2005a, b). To our best knowledge, only one reportfrom our group was concerned with its metabolites inrat urine after dosing healthy rats with jatrorrhizine(Han et al., 2006; Zuo et al., 2006). In this work, thein vivo metabolites in rat feces, urine and plasmaafter dosing healthy rats with jatrorrhizine, along within vitro metabolites in rat intestinal flora incubationsolution and liver microsome incubation solution, wereinvestigated to further clarify the metabolic pathways.

EXPERIMENTAL

Chemicals and reagents

Jatrorrhizine hydrochloride was purchased from NICPBP(National Institute for the Control of Pharmaceutical andBiological Products, China). Methanol was HPLC-grade fromFisher Chemical Co., Inc. (CA, USA). Water was deionizedand double-distilled. All other reagents were of analyticalreagent grade.

*Correspondence to: Yong Chen, Hubei Province Key Laboratory ofBiotechnology of Chinese Traditional Medicine, Hubei University,Wuhan 430062, People’s Republic of China.E-mail: [email protected]

Abbreviations used: CID, collision-induced dissociation; ESI, electro-spray ionization.

Contract/grant sponsor: National Natural Science Foundation ofChina; Contract/grant number: 30630075.Contract/grant sponsor: Ministry of Education Key Laboratoryfor the Synthesis and Application of Organic Functional Molecules;Contract/grant number: 2007-KL-006.


Identification of in-vivo and in-vitro metabolites of jatrorrhizine 1361ORIGINAL RESEARCH

For the anaerobic culture solution, 37.5 mL of solution A(0.78% K2HPO4), 37.5 mL of solution B [0.47% KH2PO4,1.18% NaCl, 1.2% (NH4)2SO4, 0.12% CaCl2 and 0.25%MgSO4·H2O] and 50 mL of solution C (8% Na2CO3) weremixed along with 0.5 g L-cysteine, 2 mL 25% L-ascorbic acid,1 g beef extract, 1 g tryptone and 1 g nutrient agar dissolvedin 1 L of distilled water, and adjusted to pH 7.4 with 2 mol/LHCl (Chen et al., 2007).

Instrumentation

LC/MS (MS/MS) experiments were performed on an LCQDuo quadrupole ion trap mass spectrometer (Thermo-Finnigan Corp., San Jose, CA, USA) with an Agilent 1100Series G1311A Quat pump and a G1313A autosampler usingpositive electrospray as the ionization process. The softwareXcalibur version 1.2 (Thermo-Finnigan Corp., San Jose, CA,USA) was applied for the system operation and data collec-tion. A refrigerated centrifuge (2K15C, Sigma, Germany) andan ultracentrifuge (LE-80K, Beckman Coulter, USA) wereused to centrifuge samples. The samples were purified by useof a C18 solid-phase extraction (SPE) cartridge (3 mL, 200 mg,AccuBond, Agilent). The intestinal incubation experimentswere performed in anaerobic incubation bags (AnaeroPouch™-Anaero 08G05A-23, Mitsubishi Gas Chemical Company).

Chromatographic and mass spectrometricconditions

A reversed-phase column (Zorbax Extend-C18, 2.1 × 50 mmi.d., 3.5 μm, Agilent, USA) was connected to a guard column(4.6 × 12.5 mm cartridge, 5 μm, Agilent) containing the samepacking material to separate jatrorrhizine and its metabo-lites. The column temperature was set at 25°C. The mobilephase was 70:30 (v/v) methanol–formic acid aqueous solution(pH 3.5) with a flow rate of 0.1 mL/min during the wholechromatographic run time of 5 min. An aliquot of 20 μL ofthe purified sample was injected automatically into the HPLCsystem for LC/MS/MS analysis.

MS/MS analyses were conducted in positive ion detectionmode. The optimum conditions were a source spray potentialof 4.5 kV, a capillary voltage of 3 V and a capillary tempera-ture of 25°C. The flow rates of sheath gas (N2) and auxiliarygas were 40 and 20 units (corresponding to 0.6 and 6 L/minapproximately), respectively. Other parameters, including thepotentials of octapole offset and the tube lens offset, werealso optimized for maximum abundance of the ions of inter-est by the automatic tuning procedure of the instrument.

The MS/MS product-ion spectra were produced by collision-induced dissociation (CID) of the molecular ions [M]+ ofall analytes. The collision energy for each ion transition wasoptimized to produce the maximum abundance of the selectedion. The optimized CID energy was 32–38% (relative colli-sion energy) for MS/MS works. Data acquisition was per-formed in full-scan LC/MS and LC/MS/MS mode.

Sample preparation

In vivo samples. Five Wistar rats (male 180 ± 5 g, HubeiExperimental Animal Research Center, China) were housedin metabolic cages for the collection of urine, feces and plasma.The rats were fasted for 24 h but with access to water, andthen administered a single dose of jatrorrhizine (34 mg/kg) byoral gavage (Yu et al., 2007). Urine and feces were collectedindividually during the time period 0–48 h and stored at−20°C until further processing. Aliquots of 1 mL of hepar-inized blood samples were collected from the ophthalmicveins of rats by sterile capillary tube respectively at 1, 2, 4 and8 h, then shaken and centrifuged at 10,000g for 10 min. Thesupernatants were mixed and immediately frozen at −20°Cuntil processing.

A mixed 0–48 h urine sample was centrifuged at 2000gfor 10 min. The supernatant was loaded onto a C18 SPEcartridge which had been preconditioned with 1 mL methanoland 2 mL water. The SPE cartridge was then washed with2 mL water to elute the impurity and 1 mL methanol to elutethe analytes in turn. The eluent was filtered through a0.45 μm membrane, and an aliquot of 20 μL was injected forLC/MS/MS analysis.

A mixed 0–48 h feces sample was homogenized with waterin the ratio of 1 g:2 mL and then centrifuged at 2000g for10 min. The supernatant was also extracted by the SPE proce-dures mentioned above. The eluent was filtered through a0.45 μm membrane, and an aliquot of 20 μL was injected forLC/MS/MS analysis.

The mixed blood sample was purified by the SPE proce-dures described above. The eluent was filtered through a0.45 μm membrane, and an aliquot of 20 μL was injected forLC/MS/MS analysis.

In vitro samples. Fresh feces of Wistar rats were homo-genized with normal saline in the ratio 1 g:4 mL immediatelyand then centrifuged at 1000g for 5 min. An aliquot of1.5 mL of the supernatant was mixed with 10 mL of theanaerobic culture solution. Jatrorrhizine was added to theabove intestinal flora cultural solution to a final concentration

Figure 1. The structure of jatrorrhizine.


1362 Y. Zhang et al.ORIGINAL RESEARCH

Table 1. MS/MS product ions of jatrorrhizine and its metabolites

Analyte [M]+ tR (min) Product ions (m/z, relative abundance/%) A B C D E

M0 338 3.31 323(100) 294(10) 308(5) 306(4) + + + + +M1 336 3.43 321(100) 292(13) 306(6) + + + + +M2 322 3.29 307(100) 278(9) + + + + +M3 294 3.37 279(100) 250(6) + + + + +M4 350 3.86 335(100) 306(8) + + + + –M5 354 3.15 339(100) 310(10) 324(4) + + + + +M6 370 3.29 355(100) 326(8 ) 338(5) + + + + –M7 386 3.00 371(23) 342(9) 354(100) + + – + –M8 352 3.83 337(100) 308(8) + + + + +M9 368 3.63 353(100) 324(6) + + + + +M10 384 3.59 369(14) 354(100) 352(8) + + + + –M11 400 3.87 385(24) 356(6) 368(100) + – – + –M12 366 3.93 351(100) 322(14) + + + + –M13 382 3.84 367(100) 338(18) 352(16) 350(33) + – – + –M14 398 3.46 383(100) 366(44) + – – + –M15 324 3.27 309(100) 280(8) + + + + +M16 310 3.07 295(100) 266(5) + + – + +M17 296 3.36 281(100) + – – + +

A, urine; B, feces; C, plasma; D, intestinal bacteria incubation; E, liver microsome incubation; +, found; −, not found; M0, jatrorrhizine.

of 0.1 mg/mL and cultured in anaerobic incubation bags at37°C in a shaking water bath for 24 h. After the incubation,the incubation solution was centrifuged at 10,000g for 15 min.The supernatant was purified by the SPE procedures men-tioned above, and an aliquot of 20 μL was injected for LC/MS/MS analysis.

Wistar rats were starved overnight prior to being sacrificedby cervical dislocation. Livers were weighted and washedwith ice-cold normal saline. In order to minimize degradationof CYP450, all the preparation procedures of the liver micro-some were performed at 4°C. The minced liver tissue washomogenized with 0.1 mol/L PBS buffer (pH 7.4) in the ratio1 g:4 mL within 5 min and subsequently centrifuged at 10,000gfor 20 min. The supernatant was ultracentrifuged at 100,000gfor 60 min to obtain the liver microsome. The microsomewas resuspended in 30 mL of 0.1 mol/L PBS buffer (pH 7.4,containing 30% glycerine), in which the total protein concen-tration, determined by the method of Coomassie BrilliantBlue (Bradford, 1976) with bovine serum albumin as the pro-tein standard, was 6.08 mg/mL, and stored at −80°C until use.Incubation was performed in a final volume of 0.5 mL PBSbuffer (0.1 mol/L pH 7.4) containing 0.25 mg/mL jatrorrhizinehydrochloride, 1.0 mmol/mL NADPH and 1 mg/mL liver micro-somal protein. The incubation mixture was preincubated at37°C for 3 min prior to adding NADPH to initiate the meta-bolic reaction. Incubation experiment was performed at 37°Cfor 1 h and terminated by adding an equal volume of metha-nol. The incubation mixture was then centrifuged at 10,000gfor 10 min and the supernatant was evaporated at 37°C undera gentle stream of nitrogen. The residues were reconstitutedin 0.5 mL of methanol and then centrifuged at 10,000g for10 min. An aliquot of 20 μL was injected for LC/MS/MSanalysis.

The structural analysis method of the metabolites

The electrospray ionization mass spectral behavior of jatror-rhizine was reported in the previous publication from our

group (Han et al., 2006). The characteristic product ions ofjatrorrhizine (Table 1 and Fig. 2) were the most importantinformation for the identification of its metabolites. By com-paring the product ions of the metabolites with those ofthe parent drug, the structures of metabolites may be rapidlycharacterized even if no standard for the metabolite is avail-able. According to the structure of parent drug and knowncommon metabolic pathways, the possible structures of metabo-lites were postulated first. Then the full-scan mass spectra ofthe samples from drug-treated animals and the correspondingcontrol animals were compared with each other to find thepossible metabolites, and the product ions of the possiblemetabolite were obtained by CID in SRM mode by isolatingthe molecular ion or the fragment ion of interest in the iontrap. Finally, the structures of the metabolites were elucidatedby comparing their molecular weights and product ions withthose of the parent drug.

RESULTS

Metabolites in intestinal flora incubation solution

There were 17 metabolites of the parent drug foundin rat intestinal flora incubation solution with theirmolecular ions at m/z 294, 296, 310, 322, 324, 336, 350,352, 354, 366, 368, 370, 382, 384, 386, 398 and 400,respectively. Their MS/MS ions along with theirrelative abundance and retention time (tR), which wereperformed by selective ion monitoring (SIM), arepresented in Table 1 and product ions are alsopresented in Fig. 2.

The molecular ion at m/z 336 (M1) and its productions at m/z 321, 292 and 306 were all 2 Da less than themolecular ion of the parent drug and its main fragmentions at m/z 323, 294 and 308. Therefore, M1 may be the



dehydrogenation product of the parent drug, and thedehydrogenated position was at the unique saturatedC—C bond of the B-ring. The molecular ions ofM2 (m/z 322) and M3 (m/z 294), along with the corre-sponding fragment ions of them, were all 14 and 42 Daless than the molecular ion of M1 (m/z 366) and itsmain fragment ions at m/z 321 and 292, respectively.These results indicated that M2 and M3 were themono-demethylation and tri-demethylation products ofM1, respectively.

The molecular ion of M4 (m/z 350) and its main frag-ment ions at m/z 335 and 306 were all increased by14 Da compared with the molecular ion of M1 and itsmain fragment ions at m/z 321 and 292. Therefore, M4can be confirmed as the monomethylation product ofM1, and the methylated position was at the uniquehydroxyl of the A-ring.

The molecular ions of M5 (m/z 354), M6 (m/z 370)and M7 (m/z 386), along with the corresponding mainfragment ions of them, were increased by 16, 32 and

48 Da respectively, compared with those of the parentdrug. The results indicated that M5, M6 and M7may be the mono-hydroxylate, di-hydroxylate and tri-hydroxylate of jatrorrhizine, respectively. According tothe structure of the parent drug, the saturated carbonatom at B-ring is near to an ortho sp2 hybridization C-atom at the A-ring and can be oxidated more easilythan the carbon atoms at aromatic ring. Therefore, thefirst hydroxylation position may be the meta-position ofthe N-atom at the B-ring, the secondary hydroxylationposition may be the saturated C-atom at the D-ringconsidering the isomer of the parent drug (Fig. 1), andthe third hydroxylation position may be the ortho posi-tion of the N-atom at the B-ring.

The molecular ion of M8 (m/z 352) and its mainfragment ions at m/z 337 and 308 were all 14 Da morethan the molecular ion of the parent drug and its mainfragment ions at m/z 323 and 294. Therefore, M8 wasconfirmed as the methylated product of the parentdrug, for which the methylation position was at the

Figure 2. MS2 product-ion spectra of jatrorrhizine and its metabolites.



unique hydroxyl at the A-ring. The molecular ions ofM9 (m/z 368), M10 (m/z 384) and M11 (m/z 400), alongwith the corresponding main fragment ions of them,were increased by 16, 32 and 48 Da, respectively, com-pared with those of M8. Therefore, M9 (m/z 368),M10 (m/z 384) and M11 (m/z 400) may be the mono-hydroxylate, di-hydroxylate and tri-hydroxylate of M8,respectively, and the hydroxylation positions of M9,M10 and M11 were considered the same as those ofM5, M6 and M7, respectively.

The molecular ions of M12 (m/z 366), M13 (m/z 382)and M14 (m/z 398), along with their correspondingmain fragment ions, were decreased by 2 Da comparedwith those of M9 (m/z 368), M10 (m/z 384) and M11(m/z 400) respectively. Therefore, M12, M13 and M14may be the dehydrogenation products of M9, M10 andM11 respectively.

The molecular ions of M15 (m/z 324), M16 (m/z 310)and M17 (m/z 296), along with their corresponding main

fragment ions, were decreased by 14, 28 and 42 Da,respectively, compared with that of the parent drug.Therefore, M15, M16 and M17 may be the mono-,di-, and tri-demethylated products of the parent drug,respectively.

Metabolites in liver microsome incubationsolution

Nine metabolites, including seven phase I metabolites(M1, M2, M3, M5, M15, M16 and M17) and two phaseII metabolites (M8 and M9) were identified in livermicrosome incubation solution.

Metabolites in rat urine, feces and plasma

Seventeen metabolites were found in rat urine. Amongthem, seven new metabolites including phase I metabo-lites (M2 and M3) and phase II metabolites (M4, M11,

Figure 2. (Continued)



M12, M13 and M14) were detected in rat urine for thefirst time. The SIM chromatograms of jatrorrhizineand the new metabolites in urine are shown in Fig. 3.Thirteen metabolites, including phase I metabolites(M1, M2, M3, M5, M6, M7, M15 and M16) and phaseII metabolites (M4, M8, M9, M10 and M12) were de-tected in rat feces. Eleven metabolites, including phaseI metabolites (M1, M2, M3, M5, M6 and M15) andphase II metabolites (M4, M8, M9, M10 and M12) weredetected in plasma. The proposed metabolic pathwaysof jatrorrhizine are shown in Fig. 4.

CONCLUSION

In this work, the metabolites of jatrorrhizine wereidentified on the basis of their MS/MS fragmentationinformation after in vivo and in vitro biotransformation.As a result, 17 metabolites were found in rat urine, 13

metabolites were detected in rat feces, 11 metaboliteswere detected in rat plasma, 17 metabolites were iden-tified in intestinal flora incubation solution and ninemetabolites were detected in liver microsome incuba-tion solution. Hydroxylation, methylation, dehydro-genation and demethylation were the main metabolicpathways of jatrorrhizine. All the results were reportedfor the first time, except for some of the metabolites inrat urine.

Acknowledgements

This work was supported by National Natural ScienceFoundation of China (grant no. 30630075) and theMinistry of Education Key Laboratory for the Synthesisand Application of Organic Functional Molecules (HubeiUniversity no. 2007-KL-006). The authors would liketo thank their colleagues for their valuable technicalassistance.

Figure 3. SIM chromatograms of jatrorrhizine and its new metabolites in urine.



Figure 4. The proposed metabolic pathways of jatrorrhizine.



REFERENCES

Appolonova SA, Shpak AV and Semenov VA. Liquid chromatography–electrospray ionization ion trap mass spectrometry for analysis ofmesocarb and its metabolites in human urine. Journal of chromato-graphy B 2004; 800: 281–289.

Bradford MM. A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein–dye binding. Analytical Biochemistry 1976;�72: 248–254.

Chan SA, Chen MJ, Liu TY, Fuh MR, Deng JF, Wu ML and HsiehSJ. Determination of aristolochic acids in medicinal plant andherbal product by liquid chromatography–electrospray–ion trapmass spectrometry. Talanta 2003; 60: 679–685.

Chen HX, Chen Y, Du P and Han FM. LC-MS for identification andelucidation of the structure of in-vivo and in-vitro metabolites ofatropine. Chromatographia 2007; 65: 413–418.

Chung EW,�Ho ENM,�Leung DKK, Tang FPW, Yiu KCH and WanTSM. Detection of anti-ulcer drugs and their metabolites in horseurine by liquid chromatography–mass spectrometry. Chromatographia2004; 59: S29–38.

Fu Y, Hu B, Tang Q, Zhang Q and Xiang J. Effect of jatrorrhizine,berberine, Huanglian decoction and compound-mimic prescriptionon blood glucose in mice. Chinese Traditional and Herbal Drugs2005a; 36: 548–551.

Fu Y, Hu B, Tang Q, Fu Q and Xiang J. Hypoglycemic activityof jatrorrhizine. Journal of Huazhong University of Science andTechnology. Medical Sciences 2005b; 25: 491–493.

Han FM, Zhu MM, Chen HX and Chen Y. Liquid chromatography–tandem electrospray ionization ion trap mass spectrometric assay

for the metabolites of jatrorrhizine in rat urine. Acta PharmaceuticaSinica 2006; 41: 846–851.

Smyth WF. Electrospray ionisation mass spectrometric behaviour ofselected drugs and their metabolites. Analytica Chimica Acta 2003;492: 1–16.

Smyth WF. Recent studies on the electrospray ionisation massspectrometric behaviour of selected nitrogen-containing drugmolecules and its application to drug analysis using liquid chro-matography–electrospray ionisation mass spectrometry. Journal ofChromatography B 2005; 824: 1–20.

Thevis M, Fubholler G, Geyer H, Rodchenkov G, Mareck U,Sigmund G, Koch A, Thomas A and Schanzer W. Detection ofstanozolol and its major metabolites in human urine by liquid chro-matography–tandem mass spectrometry. Chromatographia 2006;64: 441–446.

Xing J, Xie CF and Lou HX. Recent applications of liquid chro-matography–mass spectrometry in natural products bioanalysis.Journal of Pharmaceutical and Biomedical Analysis 2007; 44: 368–378.

Yu Sen, Pang X, Deng Y, Liu L, Liang Y, Liu X, Xie L, Wang Gand Wang X. A sensitive and specific liquid chromatography massspectrometry method for simultaneous determination of berberine,palmatine, coptisine, epiberberine and jatrorrhizine from CoptidisRhizoma in rat plasma. International Journal of Mass Spectrometry2007; 268: 30–37.

Zuo F, Nakamura N, Akao T and Hattori M. Pharmacokinetics ofberberine and its main metabolites in conventional and pseudogerm-free rats determined by liquid chromatography/ion trap massspectrometry. Drug Metabolism and Disposition 2006; 34: 2064–2072.

Documents

LC/MS/MS for identification of in vivo and in vitro metabolites of jatrorrhizine