A high-sensitivity UPLC-MS/MS method forsimultaneous determination and confirmationof triptolide in zebrafish embryosYan Zhou,a,† Ming-Fang He,b,† Franky Fung-Kei Choi,c Zhi-Heng He,d

Jing-Zheng Song,c Chun-Feng Qiao,c Song-Lin Lic and Hong-Xi Xue*

ABSTRACT: A high-sensitivity ultra-performance liquid-chromatography (UPLC) coupled with tandem mass spectrometricmethod was developed for simultaneous quantification and confirmation of triptolide in both zebrafish embryos and theaqueous-exposure solution on a tandem quadrupole mass spectrometer (TQ-MS). This was achieved by performing quantifi-cation using the multiple reaction monitoring (MRM) acquisition with simultaneous characterization of the MRM peak usingproduct ion confirmation (PIC) acquisition as it elutes from the chromatographic system. Separation was achieved on a 1.7 mmC18 UPLC column using 0.1% formic acid water–acetonitrile mobile phase with a cycle time of 6 min. The linear range of0.115–360 ng/mL, and lower limits of detection of 0.02 ng/mL and quantification of 0.064 ng/mL were established. Thismethod was successfully applied to determine the time course of triptolide absorption by zebrafish embryos and the amountof triptolide remaining in the culture medium after administration of two triptolide dosages at three time points. This coupledMRM with PIC approach could provide both qualitative and quantitative results without the need for repetitive analyses. Thisresulted in the reduction of further confirmative experiments and analytical time, and ultimately increased laboratoryproductivity. Copyright © 2010 John Wiley & Sons, Ltd.

Keywords: triptolide; zebra-fish; UPLC/MS/MS; MRM; PIC

IntroductionTriptolide, a diterpenoid triepoxide, is one of the major active andtherapeutic constituents derived from the root extracts of Triptery-gium wilfordii Hook F., which is a perennial twining vine-likemember of the Celastraceae plant family (Xia et al., 1988). In recentyears, it has attracted interest because of its important biologicalactivities such as anti-inflammatory and immunosuppressive, aswell as anticancer and antiangiogenic activities (Tao et al., 2002;Kiviharju et al., 2002; He et al., 2009, 2010). The structural formulaeof triptolide and Ginkgolide A (internal standard, IS) are shown inFig. 1A and B. Zebrafish (Danio rerio), a popular in vivo vertebratemodel, is useful for not only studying molecular genetics anddevelopmental biology of vertebrates, but also as a model organ-ism for evaluating the aquatic toxicity of environmental pollutants(Goldsmith and Harris, 2003) and for high-throughput drugscreening, drug-finding strategies as well as toxicological studies(Langheinrich, 2003; Parng, 2005). Zebrafish has also gained muchattention in drug functional and analytical research (Rose et al.,2002; Isaacson et al. 2007; Chatterjee and Gerlai, 2009). Consider-ing the high similarity between zebrafish and higher vertebratesat the cellular, anatomical and physiological levels with their shortlife cycles and high productivity of the female, it allows large-scalestudies and robust statistics to be conducted.

In our previous study, triptolide was demonstrated to possesspotent antiangiogenic activity against vessel formation inzebrafish embryo. Dose–response studies together with embryoendogenous alkaline phosphatase (EAP) staining and reversetranscription polymerase chain reaction (RT-PCR) analysis

revealed that administration of 1.2 mM triptolide into the culturemedium embryo water caused significant inhibition of angiogen-esis (He et al., 2009, 2010). However, determining the traceamount of triptolide taken up into the zebrafish embryosrequires a highly sensitive and reliable method capable ofanalyzing analyte concentration in the ng/mL range.

Various analytical methods have been developed for deter-mination of triptolide in different matrices, including

* Correspondence to: Hong-Xi Xu, Shanghai University of Traditional ChineseMedicine, Shanghai, China. E-mail: xuhongxi88@gmail.com

† Yan Zhou and Ming-Fang He contributed equally to this work.

a Chengdu Institute of Biology, Chinese Academy of Sciences, Chengdu, China

b School of Pharmaceutical Sciences and Nanjing Sino-American CancerCentre, Nanjing University of Technology, Nanjing, China

c Hong Kong Jockey Club Institute of Chinese Medicine, Shatin, New Territo-ries, Hong Kong, China

d Department of Biology, The Chinese University of Hong Kong, Shatin, NewTerritories, Hong Kong, China

e Shanghai University of Traditional Chinese Medicine, Shanghai, China

Abbreviations used: EAP, endogenous alkaline phosphatase; MRM, mul-tiple reaction monitoring; PIC, product ion confirmation; TQ-MS, tandemquadrupole mass spectrometer.

Contract/grant sponsor: Hong Kong Jockey Club Charities Trust Fund.

Short communication

Received 9 June 2010, Accepted 28 August 2010 Published online in Wiley Online Library: 06 November 2010

(wileyonlinelibrary.com) DOI 10.1002/bmc.1534

851

Biomed. Chromatogr. 2011; 25: 851–857 Copyright © 2010 John Wiley & Sons, Ltd.

high-performance liquid chromatography (HPLC) (Yao et al.,2006), gas chromatography (GC; Jin et al., 2009), capillary electro-phoresis (CE); Song et al., 2001) and liquid chromatographycoupled with mass spectrometry (LC-MS; Chang et al., 2005).Among these analytical techniques, mass spectrometry hasplayed an essential role in determining analyte in biologicalsamples due to its sensitivity, rapidity and low levels of sampleconsumption. Although high selectivity of MS can be achieved,especially in using tandem mass spectrometric methods, such asmultiple reaction monitoring (MRM), it is still a challenge to deter-mine analytes at trace levels. The target analyte may be alsoco-detected with matrix interferences, even in the MRM (blank)mode. As a result, a second qualitative experiment is requiredforfurther confirmation to verify the identity of the peak of quanti-tative interest. Moreover, the determination of trace levels ofanalyte in biological smaple can be very problematic, requiring amore sensitive method and enhanced LC-MS/MS data acquisitioncapability for quantitative analysis.

In this paper, we firstly introduced a highly sensitive and robustUPLC-MS/MS method for the determination of trace-level trip-tolide in zebrafish embryos using MRM combined with production confirmation (PIC) for simultaneous quantification and char-acterization of the target compound in a single run. Details ofmethod development and validation are also described.

Experimental

Chemical and reagents

Triptolide was purified in our laboratory and Ginkgolide A (internal stan-dard, IS) was obtained from the Chinese National Institute for Control ofPharmaceutical and Biological Products (Beijing, China). HPLC-MS aceto-nitrile and methanol were purchased from Fisher Scientific UK Ltd.(Loughborough, UK). Formic acid of spectroscopy grade was purchasedfrom Sigma-Aldrich (St Louis, MO, USA). High-purity water was purifiedin-house using a Milli-Q SP Regent Water system (Millipore, Bedford, MA,USA). Leucine-enkephalin and dimethyl sulfoxide were analytical gradeand were obtained from Sigma-Aldrich (St Louis, MO, USA). For samplepreparation an Oasis™ HLB SPE cartridge used, from Waters Corporation,Milford, MA, USA.

Instrumentation and UPLC-MS/MS conditions

Ultraperformance liquid chromatography was performed on an Acquity™UPLC™ system (Waters Corporation, Milford, MA, USA), equipped with abinary solvent delivery system and an autosampler. The chromatographicseparation was achieved on a Waters Acquity BEH C18 column (100 ¥2.1 mm i.d., 1.7 mm, Waters, Wexford, Ireland) protected by a precolumnwith a gradient elution using mobile phase composed of 0.1% formic acid

in water and acetonitrile. The column was maintained at 35°C and theflow rate was set at 0.4 mL/min. The injection volume was 7.5 mL using fullloop mode for sample injection.

Mass spectrometric detection was operated on a Waters Xevo TQ-MS(tandem quadrupole mass spectrometer; Micromass MS Technologies,Manchester, UK) using an electrospray ionization source set in positiveion mode. The quantification mode was multiple-reaction monitoring(MRM). Two MRM scans along with corresponding PIC scans were per-formed in a single UPLC run; the following precursors to production iontransitions were used for MRM scans: triptolide, precursor ion [M + H]+ tothe product ion at m/z 361 → 105 and ginkgolide A (IS) at m/z 409 → 345.A PIC scan was enabled in each MRM, where the third quadrupole wasswitched to the scanning mode to collect an enhanced product ion scanusing ScanWave daughter ion scan (DS) mode. All data collected in cen-troid mode were acquired and processed using MassLynx™ version 4.1sofware with TargetLynx™ V4.1 program (Waters Corporation, Milford,MA, USA). Detail MS parameters are listed in Table 1.

Embryo handling and drug administration

Zebra-fish embryos were generated by natural pairwise mating. Theembryos were maintained in embryo water (0.2 g/L Instant Ocean® Salt,Aquarium systems, USA) at 28.5°C. They were manually dechorionatedwith forceps at 24 h post-fertilization (hpf) immediately prior to drugtreatment. Dechorionated embryos were arrayed in a 24-well plate, and20 embryos per well were incubated with 1 mL of embryo water contain-ing various concentrations of triptolide for indicated period. The 0.2%dimethyl sulfoxide (DMSO) served as vehicle control. The ethical guide-lines described in the National Institutes of Health Guide for Care and Useof Laboratory Animals were followed throughout the experiments.

Calibration standards and quality control samples

The stock solution of triptolide was prepared precisely by dissolving 1 mgtriptolide powder in 1 mL of methanol at concentration of 1.0 mg/mL andkept at 4°C. The stock solution was serially diluted with methanol toprovide a series of standard solutions at desired concentrations. Calibra-tion standard at five concentration levels ranged from 0.115 to 360 ng/mL. The IS stock solution of 1 mg/mL was prepared in methanol, and20.0 ng/mL was applied to each working solution and samples. Qualitycontrol (QC) samples were prepared separately using the set of stocksolutions at low, medium and high standard concentration levels. Boththe calibration standard samples and the QC samples were applied in themethod validation study. All solutions were stored at 4°C before use.

Sample preparation

Embryos treated with 0.625 and 1.25 mM triptolide and their correspond-ing culture medium embryo water at three different time points, 12, 24,36 hpf, were collected individually into 1.5 mL micro-centrifuge tubes.

Figure 1. Structures of triptolide and Ginkgolide A (internal stan-dard, IS).

Table 1. Tandem mass spectrometer main workingparameters

Parameter Value

Ionization mode ESI(+)Source temperature (oC) 150Capillary voltage (kV) 1.45Cone voltage (V) 42Desolvation temperature (oC) 600Desolvation gas flow (L/h) 1000Cone gas flow (L/h) 20Collision energy (V) 35Ion transition for triptolide, m/z 361 → 105Ion transition for IS, m/z 409 → 345

852

Y. Zhou et al.

Biomed. Chromatogr. 2011; 25: 851–857wileyonlinelibrary.com/journal/bmc Copyright © 2010 John Wiley & Sons, Ltd.

Embryos were then washed three times with double-distilled water(ddH2O) and homogenized in 0.5 mL ddH2O. Embryo homogenates andtheir corresponding culture medium embryo water were freeze-dried todryness followed by storage at -80°C until sample analysis.

Dried embryo homogenates and their culture medium embryo waterat each time point were dissolved in 1.0 mL of 20% methanol separately.Then, 10 and 100 mL of ginkgolide A (internal standard, IS, 200 ng/mL)were added to embryo homogenate and the culture medium embryowater separately. After vortex-mixing for 1 min, the samples were ultra-sonicated for 10 min. Before the sample mixture was loaded onto anOasis™ HLB extraction cartridge (30 mg/1 cm3), it was pre-conditionedwith 2.0 mL of methanol followed by 1.0 mL of water. After loading ofthe samples, the cartridges were washed with 1.0 mL of 5% methanoltwice. The triptolide and IS were subsequently eluted with 2.0 mLmethanol. The eluent was collected and evaporated to dryness at 40°C.The dried residues of embryo homogenate and the culture mediumembryo water were reconstituted in 0.1 mL of methanol respectively. A7.5 mL aliquot of the resulting solutions were injected into the UPLC/MS/MS system for analysis.

Method validation

Selectivity. To investigate whether the presence of endogenous con-stituents would interfere with the assay, six blank zebra-fish embryossamples were analyzed to detect any potential interference for co-elutingwith analyte and IS. Chromatographic peaks of the analyte and IS wereidentified by comparing their retention time and MRM responses withthose of authentic standards.

Linearity. A calibration curve was prepared at five concentrationsof triptolide range from 0.115 to 360 ng/mL. The linearity ofcalibration curve was constructed by plotting the peak area ratio ofanalyte to the signal area of IS obtained from MRM as the ordinate vari-ables (y) vs the nominal concentrations of analyte (x), forming a linearregression.

Lower limits of detection and quantification. The lower limit ofdetection (LLOD) was considered as the lowest concentration producinga signal-to-noise (S/N) ratio of 3 and the lower limit of quantitation (LLOQ)was defined as the concentration producing S/N ratio of 10.

Precision and accuracy. The intra- and inter-day precision and accu-racy of the method were checked by performing six replicates of QCsamples containing three concentration levels (4, 20 and 60 ng/mL) oftriptolide within a day in order to assess the intra-day variability (n = 6)and monitoring six replicates at each concentration level during twoconsecutive validation days for inter-day variability (n = 6 series per day).Relative standard deviation (RSD) was taken as a measure of precision.The deviation of each concentration level from the nominal concentra-tion was expected to be within �15%. Similarly, the related error (RE) ofmean accuracy should not exceed 15% of the actual value.

Matrix effect. The investigation of the matrix effect on electrosprayionization (ESI) was evaluated by comparative analyses on two sets ofsamples based on the approach proposed by Matuszewski et al. (2003). Inthe first set (A), three concentration levels (low, medium and high concen-trations, 4, 20 and 60 ng/mL respectively) of neat standards were dissolvedin methanol and were analyzed in triplicate. Another set (B), blank embryohomogenate samples which were free of any significant interferences atthe retention time of analyte and IS, were extracted by SPE and then spikedwith three levels of standards. The influence of endogenous componentson LC-MS was calculated by comparing the ratio of peak areas of matrixsample after sample preparation with that of a neat standard.

Recovery. The recoveries of triptolide from zebra fish embryos sampleswere evaluated by comparing the peak areas obtained from extractedsamples spiked with known amounts of standard before extraction to

those of the standard solutions spiked after extraction. Experiments wereperformed at three concentration levels in triplicate.

Results and discussion

Optimization of mass spectrometric parameters

In order to achieve the quantification of triptolide in zebra fishembryos, the mass spectrometric parameters such as collisionenergy, cone voltage and capillary voltage were optimized toattain the maximum sensitivity for detection of the analytes.Quantification was done by MRM of the protonated molecules oftriptolide and IS. The precursor ions and product ions of triptolideand IS were clearly observed in MS and MS/MS spectra afterinfusing individual standard solutions into mass the spectrom-eter using a fluidics system at a flow rate of 10 mL/min. The MSconditions and transition channels for triptolide and IS are sum-marized in Table 1.

Coupled MRM and PIC MS method

Figure 2(B) shows the MRM chromatogram obtained from thequantification of triprolide at m/z 361 → 105. As shown in thefigure, there was no significant endogenous peak that couldinterfere with the analyte and IS (Fig. 2A, 2B). The detection speci-ficity and sensitivity were further verified by analyzing triptolideat other concentration levels. Serial dilutions were prepared toinject decreasing amounts of the tripolide standards (2, 0.2 and0.02 ng/mL) in the blank embryo samples.

Qualitative confirmation of the peak of interest was evident inthe resulting PIC spectra operated in ScanWave DS mode. In thePIC mode, the Xevo TQ-MS will switch from MRM to scan after theapex of an UPLC peak provided that a minimum intensity thresh-old is reached. The PIC scan was used to collect an enhancedproduct ion scan using ScanWave daughter ion scan (DS) mode.The ScanWave mode of operation allows ions within the collisioncell to be accumulated and then separated according to theirmass-to-charge (m/z) ratio. Synchronizing the release of theseions with the scanning of the second quadrupole mass analyzergreatly improves the duty cycle, which significantly enhances thesignal intensity of full-scan spectra for both MS and product ionsin a quadrupole instrument.

The daughter ions obtained from PIC were compared withthose of authentic triptolide and could be used to unambigu-ously identify the peak in biological samples (Fig. 2C). In order toaid the identification/confirmation of the peak, the ScanWave DSmode was used to collect the PIC spectrum, which typicallyshowed a four-fold signal enhancement when compared with aconventional product ion spectrum (Fig. 2D), which could beattributed to the more efficient duty cycle achieved. The scanrange for the PIC was from m/z 50 to 410.

In spite of the narrow UPLC peaks, the high data acquisitionrate of the Xevo TQ-MS allowed for accurate determination of theMRM transitions. Confirmation was achieved by PIC, which is trig-gered after the peak top is detected where the definition of thepeak itself is not affected. As such, both quantitative and qualita-tive data can be acquired simultaneously.

Method validation

Linearity, lower limits of detection and quantification. Thelinear calibration curves were plotted by internal standardmethod based on linear regression analysis of the integrated

853

Simultaneous determination and confirmation of triptolide in zebrafish embryos

Biomed. Chromatogr. 2011; 25: 851–857 wileyonlinelibrary.com/journal/bmcCopyright © 2010 John Wiley & Sons, Ltd.

peak areas (y) vs concentrations (x, ng/mL) in the concentrationranges from 0.115 to 360 ng/mL. The typical regression equationobtained by least squared regression was y = 5.997 ¥ 10-2x + 9.907¥ 10-4, where y is the peak area ratios of analyte to IS, and x is theconcentration of analyte. The correlation coefficients (r2) wereover 0.9986, and the observed deviation was within �15% for allcalibration concentrations.

The LLOD and LLOQ were 0.02 and 0.06 ng/mL respectively,which represent nearly 10-fold sensitivity enhancement com-pared with the previously published LC-MS methods (King et al.,2003; Chang et al., 2005; Qi et al., 2009).

Precision and accuracy. QC samples at three concentrationlevels (4, 20, 60 ng/mL) were analyzed in six replicates for deter-mining the accuracy and precision of the method by intra-dayand inter-day variability. From the results obtained (Table 2), theintra-day and the inter-day precisions were less than 6.79 and7.88%, and the intra-day and inter-day accuracies were within�4.75 and �2.31 respectively. All data were within the accept-able range and indicated that the present method has a satisfac-tory accuracy, precision and reproducibility.

Recovery. Under the given set of operating conditions, therecovery of triptolide from the zebrafish embryos samplesranged from 87.6 to 105.3% for the three QC concentration levels.The recovery was consistent over its calibration range, indicatingthat the extraction efficiency assay is independent on the con-centrations in the studied ranges. The recovery of IS was 89.1 �6.2% and was constant throughout the study.

Matrix effect. The matrix effect was in the range from 83.3% to102.5% (Table 2), indicating the ion suppression or enhancementin signal ranged from -16.7 to 2.5, with RSD less than 12.8% for thethree QC levels, indicating that the matrix effect on the ionizationof analyte is not serious under these experimental conditions.

Determination of triptolide. The optimized UPLC-TQ-MSmethod was subsequently applied to the quantitative analysis ofthe zebrafish embryos and their culture medium samples treatedwith two concentrations of triptolide for three time points. Theresults were shown in Table 3 and Fig. 3. As expected, the amountof triptolide absorbed by the zebrafish embryos appeared to bedose- and time-dependent. This was in accordance with our

Figure 2. (A) Chromatogram for IS with MRM 409 > 345; (B) chromatogram for triptolide with MRM 361 > 105; (C) product ion confirmation (PIC) forScanWave DS spectrum from MRM peak of triptolide at Rt 3.06 min; (D) a regular DS of triptolide.

854

Y. Zhou et al.

Biomed. Chromatogr. 2011; 25: 851–857wileyonlinelibrary.com/journal/bmc Copyright © 2010 John Wiley & Sons, Ltd.

Figure 3. TIC chromatograms of (A) triptolide absorbed in 20 embryos samples treated with 1.25 mM triptolide for 36 h; (B) triptolide remained inculture medium of 1.25 mM triptolide treated embryos for 36 h; (C) triptolide absorbed in 20 embryos samples treated with 0.625 mM triptolide for 36 h;(D) triptolide remained in culture medium of 0.625 mM triptolide treated embryos for 36 h.

855

Simultaneous determination and confirmation of triptolide in zebrafish embryos

Biomed. Chromatogr. 2011; 25: 851–857 wileyonlinelibrary.com/journal/bmcCopyright © 2010 John Wiley & Sons, Ltd.

previous report that triptolide inhibits zebrafish angiogenic geneexpression in a dose- and time-dependent manner (He et al.,2009). In zebrafish embryos treated with 0.625 mM triptolide, theamount of triptolide absorbed by the embryos was only 0.53%following treatment for 12 h and reached a plateau at 1.2% after24 h treatment. In zebrafish embryos treated with1.25 mM trip-tolide, 0.50% triptolide was absorbed by the embryos after 12 htreatment and increased to 12.5% after 48 h treatment. Theamount of triptolide absorbed by 1.25 mM triptolide treatedembryos was higher than that for 0.625 mM-treated ones. It ispostulated that the amount of triptolide absorption wouldincrease with the concentration of triptolide used in treatment.

ConclusionsA new approach to simultaneously perform quantitative andqualitative analysis of trace amounts of triptolide absorbed byzebrafish embryos on a TQ-MS was developed. This was achievedby combining the MRM acquisition and the PIC acquisition withinthe same UHPLC-MS/MS run. The method has proved to be rapid,sensitive and selective. As such, the trace level tripolide could beunambiguously determined. This strategy also has the potentialto be utilized as the ideal platform for simultaneously acquiringuseful qualitative and quantitative information from complexmatrices in the field of metabonomics.

AcknowledgementsThis research was supported by the Hong Kong Jockey ClubCharities Trust Fund. The authors are grateful to Dr Karl Lo ofWaters China Limited.

ReferencesChang XL, Chen HB, Zhao XZ, Gao ZH, Xu HB and Yang XL. High-

performance liquid chromatography determination of triptolide invitro permeation studies. Analytica Chimica Acta 2005; 534: 215–221.

Chatterjee D and Gerlai R. High precision liquid chromatography analysisof dopaminergic and serotoninergic responses to acute alcohol expo-sure in zebrafish. Behavior and Brain Research 2009; 200: 208–213.

Goldsmith P and Harris WA. The zebrafish as a tool for understanding thebiology of visual disorders. Seminars in Cell and Development Biology2003; 14: 11–18.

He MF, Liu L, Ge W, Shaw PC, Jiang RW, Wu LW and But PPH. Antiangio-genic activity of Tripterygium wilfordii and its terpenoids. Journal ofEthnopharmacology 2009; 121: 61–68.

He MF, Huang YH, Wu LW, Ge W, Shaw PC and But PPH. Triptolide func-tions as a potent angiogenesis inhibitor. International Journal ofCancer 2010; 126: 226–278.

Isaacson CW, Usenko CY, Tanguay RL and Field JA. Quantification offullerenes by LC/ESI-MS and its application to in vivo toxicity assays.Analytical Chemistry 2007; 79: 9091–9097.

Jin MC, Chen XH and OuYang XK. Simultaneous determination of trip-tolide, tripdiolide and tripterine in human urine by high-performanceliquid chromatography coupled with ion trap atmospheric-pressurechemical ionization mass spectrometry. Biomedical Chromatography2009; 23: 289–294.

King RC, Gundersdorf R and Fernandez-Metzler CL. Collection of selectedreaction monitoring and full scan data on a time scale suitable fortarget compoundquantitative analysis by liquid chromatography/tandem mass spectrometry. Rapid Communications in Mass Spectrom-etry 2003; 17: 2413–2422.

Kiviharju TM, Lecane PS, Sellers RG and Peehl DM. Antiproliferative andproapoptotic activities of triptolide (PG490), a natural product enter-ing clinical trials, on primary cultures of human prostatic epithelialcells. Clinical Cancer Research 2002; 8: 2666–2674.

Langheinrich U. Zebrafish: a new model on the pharmaceutical catwalk.Bioessays 2003; 25: 904–912.

Table 2. Intra-day and inter-day precision, recovery and matrix effect of triptolide in zebrafish samples

Concentrationadded (ng/mL)

Intra-day (n = 6) Inter-day (n = 6) Recovery(n = 5)

(mean �SD, %)

Matrix effect(%, n = 3)Measured

concentration(mean �

SD, ng/mL)

Precision(RSD%)

Accuracy(RE%a)

Measuredconcentration

(mean �SD, ng/mL)

Precision(RSD%)

Accuracy(RE%a)

4.0 4.19 � 0.28 6.79 4.75 3.93 � 0.31 7.88 -1.75 87.6 � 3.5 83.3%20.0 20.41 � 1.19 5.83 2.05 20.39 � 1.60 7.87 1.95 105.3 � 5.4 91.5%60.0 61.21 � 1.23 2.01 2.02 61.39 � 1.85 3.02 2.31 94.1 � 4.8 102.5%

RE is expressed as [(mean measured concentration)/(spiked concentration)] ¥ 100.

Table 3. Quantitative analysis of triptolide in zebrafish embryos and their aqueous-exposure solution samples

No. Samples Concentration(ng/mL)

RSD (n = 6, %)

1 0.625 (mM) Exposure Solution 12 h 22.9 4.272 Exposure Solution 24 h 21.6 11.33 Exposure Solution 48 h 18.7 12.64 Embryo 12 h 1.20 9.175 Embryo 24 h 1.89 6.026 Embryo 48 h 2.75 7.247 1.25 (mM) Exposure Solution 12 h 431.7 5.848 Exposure Solution 24 h 380.8 4.929 Exposure Solution 48 h 368.3 3.14

10 Embryo 12 h 2.27 2.6111 Embryo 24 h 33.6 4.9412 Embryo 48 h 56.5 8.09

856

Y. Zhou et al.

Biomed. Chromatogr. 2011; 25: 851–857wileyonlinelibrary.com/journal/bmc Copyright © 2010 John Wiley & Sons, Ltd.

Matuszewski BK, Constanzer ML and Chavez-Eng CM. Strategies for theassessment of matrix effect in quantitative bioanalytical methodsbased on HPLC-MS/MS. Anal. Chem. 2003; 75: 3019.

Parng C. In vivo zebrafish assays for toxicity testing. Current Opinion inDrug Discovery and Development 2005; 8: 100–106.

Qi XY, Wu B, Cheng YY and Qu HB. Simultaneous characterization ofpyrrolizidine alkaloids and N-oxides in Gynura segetum by liquidchromatography/ion trap mass spectrometry. Rapid Communicationsin Mass Spectrometry 2009; 23: 291–302.

Rose J, Holbech H, Lindholst C, Norum U, Povlsen A, Korsgaard B andBjerregaard P. Vitellogenin induction by 17beta-estradiol and17alpha-ethinylestradiol in male zebrafish (Danio rerio). ComparativeBiochemistry and Physiology C: Toxicology and Pharmacology 2002;131: 531–539.

Song XR, Yang GL, Zhao JX, Sun HW and Li ZW. Separation of triptolideand triptonide in Tripterygium wilfordii Hook f. by capillary

electrophoresis. Journal of Hebei University (Natural Science Edition)2001; 21: 183–187.

Tao X, Younger J, Fan FZ, Wang B and Lipsky PE. Benefits of an extract ofTripterygium wilfordii Hook F in patients with rheumatoid arthritis: adouble-blind, placebo-controlled study. Arthritis and Rheumatism2002; 46: 1735–1743.

Xia ZL, Huang SQ, Chen JY and Deng FX. Studies on the chemical con-stituents of the stem and leaves of Tripterygium wilfordii Hook. ZhongYao Tong Bao 1988; 13(10): 36–37.

Yao J, Zhang L, Zhao X, Hu L and Jiang Z. Simultaneous determination oftriptolide, wilforlide A and triptonide in human plasma by high-performance liquid chromatography-electrospray ionization massspectrometry. Biological and Pharmaceutical Bulletin 2006; 29: 1483–1486.

857

Simultaneous determination and confirmation of triptolide in zebrafish embryos

Biomed. Chromatogr. 2011; 25: 851–857 wileyonlinelibrary.com/journal/bmcCopyright © 2010 John Wiley & Sons, Ltd.