Qualitative and quantitative analysis of phenolic compounds in the leaves of aquilaria sinensis using liquid chromatography-mass spectrometry

Page 1

Research Article

Received: 28 April 2012, Revised: 16 November 2012, Accepted: 22 November 2012 Published online in Wiley Online Library: 10 March 2013

(wileyonlinelibrary.com) DOI 10.1002/pca.2416

Qualitative and Quantitative Analysis ofPhenolic Compounds in the Leaves of Aquilariasinensis Using Liquid Chromatography–MassSpectrometryQian Yu,a,c† Jin Qi,a,b† Hai-Xiang Yu,a Lin-Lin Chen,a Jun-Ping Kou,a,b

Shou-Jin Liuc and Bo-Yang Yua,b*

ABSTRACT:Introduction – The leaves of Aquilaria sinensis traditionally have been used in China for a century. Phenolic compounds wereinvestigated to be the major active compounds in them.Objective – To establish a method that will simultaneously determine the phenolic compounds in Aquilaria sinensis leavesand identify their characteristic fragmentation patterns, and to make a comparison of A. sinensis leaf samples from differentareas of China.Methods – High-performance liquid chromatography (HPLC) coupled with mass spectrometry (MS) and photodiode array de-tection (DAD) were used for the qualitative and quantitative analysis.Results – Twenty-one compounds, including xanthones, benzophenones and flavones, were identified or tentativelycharacterised. The fragmentation patterns of xanthones and benzophenones were also described. Also, eight componentsin the herbal samples from different regions were determined by HPLC-DAD.Conclusion – The method developed is suitable for the qualitative and quantitative analysis of phenolic constituents in theleaves of A. sinensis. Copyright © 2013 John Wiley & Sons, Ltd.

Keywords: HPLC–ESI/MSn; phenolic compounds; Thymelaeaceae; Aquilaria sinensis

* Correspondence to: Bo-Yang Yu, Department of Complex Prescription ofTraditional Chinese Medicine, China Pharmaceutical University, Nanjing211198, China. Email: [email protected]

† Co-first authors.

a Department of Complex Prescription of Traditional Chinese Medicine,China Pharmaceutical University, Nanjing 211198, China

b State Key Laboratory of Natural Medicines, China Pharmaceutical Univer-sity, Nanjing 210009, China

c Anhui Province Key Laboratory of Modern Chinese Medicine, Anhui Univer-sity of Traditional Chinese Medicine, Hefei 230038, China

34

IntroductionThe genus Aquilaria (Thymelaeaceae) is widely distributed in Asia,and mainly cultivated in China’s Guangdong and Guangxi pro-vinces, and the Taiwan district. Aquilaria sinensis (Lour.) Gilg is themajor source of agarwood, one of the most highly valuable forestproducts currently traded internationally (Adzu et al., 2003). Leavesof A. sinensis can be prepared as tea, which is consumed on a dailybasis as a healthy drink in southern China. Moreover, leaves of A.sinensis are used traditionally in China for treatments of trauma-related illness such as fractures and bruising (Zhou et al., 2008).We have previously reported that A. sinensis leaf extract has notableanalgesic and anti-inflammatory activities (Zhou et al., 2008). Thechemical constituents of the leaves of A. sinensis have been identi-fied as flavones, xanthones and benzophenones. Around 30 pheno-lic constituents of these three types of compounds have beenisolated to date (Lu and Wang, 2008a; Qi et al., 2009; Feng et al.,2011), which may contribute to the efficacy of this leaf (Kou et al.,2008, 2010; Feng et al., 2011). Mangiferin, themajor active xanthonein the plant, exhibits anti-diabetic, anti-HIV, anti-cancer, immuno-modulatory and anti-inflammatory properties (Yoshimi et al.,2001). Iriflophenone 2-O-a-L-rhamnopyranoside, another benzo-phenone compound present in the plant, has been found to be ef-fective in treating allergies (Kou et al., 2008). In recent years, theleaves of A. sinensis have drawn increasing attention as potential aanti-inflammatory and laxative effects have also been described(Kou et al., 2008, 2010; Yu et al., 2011). The ai of this study was to

Phytochem. Anal. 2013, 24, 349–356 Copyright © 2013 John

optimise quantitative and qualitative methods based on HPLC–electrospray ionisation (ESI)/MSn for the evaluation of A. sinensisleaves. Additionally, eight phenolic compounds in the A. sinensissamples collected from different regions have been quantified.

Experimental

Samples, chemicals and reagents

The leaves of A. sinensiswere collected from natural growth sites located inthe provinces of Guangdong and Guangxi in China. Professor Bo-Yang Yuidentified the leaves according to their morphology. Voucher specimens(No. Yu-201003) have been deposited in the Department of Complex

Wiley & Sons, Ltd.

9

Page 2

Q. YU et al.

350

Prescription of Traditional Chinese Medicine, China PharmaceuticalUniversity. Eight standard compounds: mangiferin, iriflophenone 2-O-a-L-rhamnopyranoside, 7,40-di-O-methylapigenin 5-O-xylosylglucoside,hydroxygenkwanin, genkwanin, 7,40-di-O-methylluteolin, 7,30 ,40-tri-O-methylluteolin and 7,40-di-O-methylapigenin (Fig. 1) were isolated andpurified in our laboratory the Department of Complex Prescription of Tra-ditional Chinese Medicine, China Pharmaceutical University, and theirstructures were determined by chemical and spectroscopic methods(UV, IR, NMR, MS) and compared with data from the literature. The purityof each compoundwas determined to bemore than 95% using HPLC–UV.Acetonitrile was HPLC grade from Merck (Darmstadt, Germany); distilledwater was further purified by a Milli-Q system (Millipore, Milford, MA,USA), and all other chemicals were analytical grade.

Sample preparation

The leaves of A. sinensis were dried at 60�C until a constant weight wasreached. The dried material was ground to 30 mesh. Weighed samples(2.5 g) were placed in a flask, and refluxed with 100 mL ethanol:water(1:1, v/v) for 3 h. The extraction solutions were filtered, evaporated undervacuum at 100�C and diluted with methanol to 100 mL. The sampleswere filtered through a 0.45 mm membrane filter (Millipore), and a 10 mLaliquot was injected into the HPLC system.

HPLC conditions for qualitative and quantitative analysis

The HPLC fingerprinting analysis and quantitative determination wereperformed on an Agilent 1100 series HPLC system consisting of a binary

Figure 1. Chemical structures of phenolic constituents in the leaves of Aquilathe HPLC of the leaves of A. sinensis in Fig 3 and Table 1. *, Chemical structu

Copyright © 2013 Johnwileyonlinelibrary.com/journal/pca

pump, autosampler, thermostatically controlled column compartmentand a DAD. A ZORBAX Stablebond Analytical SB-C18 column (250 mm� 4.6 mm, 5 mm; Agilent Technologies, Santa Clara, CA, USA) was used.The mobile phase was composed of solvent A (0.5% formic acid aque-ous, v/v) and solvent B (acetonitrile). The linear gradient program usinga flow rate of 1.0 mL/min was performed as follows: 0–10 min, 10.0–18.0% B; 10–25 min, 18.0–22.0% B; 25–30 min, 22.0–30.0% B; 30–40min, isocratic 30.0% B; 40–45 min, 30.0–45.0% B; 45–60 min, 45.0–80.0% B. The injection volume, column temperature and UV wavelengthwere set at 10 mL, 25�C and 360 nm, respectively. The UV spectra of thepeaks were recorded from 200 to 400 nm.

HPLC–DAD–ESI/MSn conditions for qualitative analysis

The HPLC conditions for HPLC–MS and HPLC–UV analysis were identical.An Agilent 1100 LC/MSD Trap mass spectrometer was connected to theHPLC–DAD via an ESI ion source as an interface (Agilent Technologies,Palo Alto, CA, USA). The ESI/MS spectra were acquired in negative ionisa-tion modes recorded over a mass range of m/z 200–600. The capillaryvoltage was set at 3.5 kV, and the fragmentation amplitude was set at1.0 V. The nebuliser/drying gas was N2: 9.0 L/min; temperature 350�C;pressure 40 psi; flow rate 0.3 mL/min. The capillary exit voltage was setat �128.5 V for the negative ion mode. Data acquisition was performedby using Chemstation software (Agilent Technologies, Palo Alto, CA,USA). Data acquisition was performed by using a Chemstation software(Agilent Technologies, USA). Data acquisition was performed by usingChemstation software (Agilent Technologies).

ria sinensis. Compounds are numbered according to the peak numbers inres of reference standard compounds.

Phytochem. Anal. 2013, 24, 349–356Wiley & Sons, Ltd.

Page 3

Table

1.Re

tentiontim

e(tR)an

dMSda

tafortheph

enolicconstitue

ntstentativelyiden

tified

intheleaves

ofA.sinensisextract

Com

poun

dt R

(min)

MS2/M

S3Iden

tificatio

n

12.4

MS2[275

]:25

7(2),2

31(1),12

9(10

)9H

-Xan

then

-9-one

,1,2,3,6,7-pen

tahy

droxy

26.2

MS2[583

]:56

5(25

),52

3(5),4

93(43),4

63(92),4

03(43),3

43(62)

MS3[373

]:35

5(70

),32

7(77

),31

3(10

0),3

01(95),2

72(52)

Neo

man

giferin

37.8

MS2[569

]:55

1(6),4

79(10),4

49(73),3

89(31),3

59(66),3

29(100

)MS3[329

]:20

9(10

0),2

01(58),1

67(31),1

35(15)

Iriflop

heno

ne-3,5-C-b-diglucoside

48.4

MS2[407

]:38

9(3),3

17(12),2

87(100

)MS3[287

]:25

9(24

),19

3(10

0),1

67(58),1

25(37)

Iriflop

heno

ne3-C-b-glucoside

58.3

MS2[407

]:38

9(1),3

47(4),31

7(10

0)Aqu

ilarix

anthon

e6

10.6

MS2[421

]:40

3(17

),33

1(73

),30

1(10

0)MS3[301

]:27

3(56

),25

9(17

),25

7(17

),24

5(10

),22

9(1),1

89(5)

Man

giferin

711

.5MS2[421

]:40

3(7),3

61(1),33

1(75

),30

1(10

0),2

59(1)MS3[301

]:30

2(77

),27

1(63

),25

9(10

0)Isom

angiferin

814

.8MS2[391

]:24

5(10

0)MS3[245

]:20

1(95

),15

1(10

0),1

25(24),8

3(12

)Iriflop

heno

ne-2-O-a-L-rha

mno

pyrano

side

915

.6MS2[435

]:41

7(1),3

75(3),34

5(33

),31

5(10

0)MS3[315

]:30

0(19

),27

2(10

0),2

73(53)

Hom

oman

giferin

1018

.7MS2[245

]:22

7(1),2

01(1),19

7(7),1

51(100

),10

7(7)

Iriflop

heno

ne11

–13

16.8

20.0

20.4

MS2[433

]:39

1(18

),24

6(8),2

45(100

)MS3[245

]:20

1(23

),17

3(12

),15

1(10

0),

125(6),1

07(12)

Iriflop

heno

ne,[2-(2-O-acty-

Α-L-Rh

amno

pyrano

syl)oxy]

Iriflop

heno

ne,[2-(3-O-acty-Α-L-rham

nopy

rano

syl)oxy]

Iriflop

heno

ne,[2-(4-O-acty-Α-L-rham

nopy

rano

syl)oxy]

1425

.9MS2[477

]:45

9(1),4

33(1),34

1(1),3

01(100

)MS3[301

]:30

1(10

0),2

55(2),20

1(7),1

65(5)

Hyp

olaetin

-5-O-b-D-glucuron

opyran

oside

1531

.9MS2[491

]:47

3(1),4

59(4),42

9(2),4

01(3),38

5(1),3

15(100

)MS3[315

]:30

1(18

),30

0(10

0),2

71(12)

Aqu

ilarisin

1636

.2MS2[591

]:45

9(1),3

39(2),28

3(10

0),2

55(1),23

9(1)

7,40-Di-O

-methy

lapige

nin-5-O-xylosylglucoside

1749

.4MS2[299

]:28

4(10

0),2

56(11),2

55(7),21

0(3),1

87(5),15

1(5)

Hyd

roxyge

nkwan

in18

52.7

MS2[283

]:26

9(26

),26

8(10

0),2

56(5),25

5(1)

Gen

kwan

in19

53.4

MS2[313

]:29

9(16

),29

8(10

0),2

95(1)MS3[298

]:28

4(10

0),2

71(50),2

55(40),1

51(1)

7,40-Di-O

-methy

lluteolin

2056

.8MS2[327

]:29

7(5),2

65(100

),24

7(1),2

27(1),151

(1)

7,30,4

0 -Tri-O-m

ethy

lluteolin

2159

.2MS2[297

]:29

7(1),2

79(24),2

77(100

),26

9(9),2

67(32),2

55(18)

7,40-Di-O

-methy

lapige

nin

Fragment Features of Xanthones and Benzophenones in A. Sinensis

Phytochem. Anal. 2013, 24, 349–356 Copyright © 2013 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/pca

351

Page 4

Q. YU et al.

352

Preparation of standard stock solutions

A stock solution containing eight reference phenolic constituents wasprepared and diluted to appropriate concentrations for the plotting ofcalibration curves, established for quantitative and qualitative analysis(mg/mL): mangiferin (319), iriflophenone 2-O-a-L-rhamnopyranoside(141), 7,40-di-O-methylapigenin 5-O-xylosylglucoside (19.6), hydroxygen-kwanin (2.52), genkwanin (8.32), 7,40-di-O-methylluteolin (1.87), 7,30 ,40-tri-O-methylluteolin (0.820) and 7,40-di-O-methylapigenin (1.98). Thestock solution, containing different concentrations of the eight analytes,was injected in triplicate. Calibration curves of the peak areas at 360 nmversus concentration were prepared for each analyte.

Evaluation of fingerprints

The HPLC fingerprints were matched automatically using the softwareSimilarity Evaluation System for Chromatographic Fingerprint of TraditionalChinese Medicine developed by the Chinese Pharmacopoeia Committee(Version 2004 A). The reference fingerprint was formed by the system usingtheMedianmethod from the general comparison of the chromatograms ofthe 11 leaves of A. sinensis extracts, and the similarity values between thereference fingerprint and the chromatograms of the extracts of A. sinensisleaves were calculated using this software.

Results and discussion

ESI/MSn fragmentation patterns of the standard samples

The phenolic constituents previously isolated from the leaves ofA. sinensis consisted of flavones, a xanthone and a benzophenone.According to a previous report, higher intensities of phenolicconstituents were obtained using the negative ion mode com-pared with the positive ion mode (Fabre et al., 2001). Therefore,analysis was conducted in the negative ion mode. The collision-induced dissociation (CID) spectra of the [M� H]� ions were stud-ied by MS2 and MS3. Typical MSn data for each type of phenolicconstituents are shown in Table 1.

As with previous reports (Ye et al., 2007; Wang et al., 2008b) anumber of common features for MS fragmentation of flavones

Figure 2. Proposed MS fragment pathway for the [M � H]� ions o

Copyright © 2013 Johnwileyonlinelibrary.com/journal/pca

and flavoneO-glycosides have been observed in this study. Accord-ing to Ye et al. (2007), the MS fragmentation behaviour of flavonesO-glycosides was characterised by the loss of the sugar moiety. Ad-ditionally, fragment ions from aglycones were generated by the lossof the following groups: CH3 (15 Da), H2O (18 Da), CO (28 Da), CH2O(30 Da) and CO2 (44 Da), and a series of Retro–Diels–Alder (RDA)fragmentation reactions were also observed, which were the com-mon features of the flavonoids during the fragmentation mechan-isms. Moreover, according to Wang et al. (2008a, 2008b), the MSfragmentation behaviour of the flavone C-glycosides is charac-terised by the loss of 60, 90, or 120 Da from the sugar moietybecause the glycosidic bonds of the flavone C-glycosides weretoo stable to be cleaved on CID. Although there were no flavoneC-glycosides identified or tentatively characterised, the related MSfragmentation feature of the flavone C-glycosides helped identifyother phenolic constituent C-glycosides in this study.

Xanthones are heterocyclic compounds with a dibenzo-g-pyrone skeleton. Compound 6 was taken as an example toillustrate the fragmentation pathways of a xanthone (Fig. 2A).In the MS spectrum compound 6 produced the [M � H] � ionat m/z 421. The MS2 spectrum gave the predominant fragmentions at m/z 331 and 301 owing to the neutral loss of C3H6O3

(90 Da) and C4H8O4 (120 Da), respectively. Unexpectedly, thisfragmentation pattern was consistent with the fragmentationcharacteristics of the flavone C-glycosides, indicating that thefragmentation character of the flavone C-glycosides may beexpanded to other phenol C-glycoside compounds (Schieberet al., 2003). Furthermore, the low intensity product ion atm/z 259 appeared in the MS2 spectrum due to the loss of theglucose residue from the [M � H]� ion. The MS3 spectrum ofthe precursor ion at m/z 301 gave the product ion at m/z 273,271 and 257 resulting from the loss of CO (28 Da), CH2O (30Da) and CO2 (44 Da), respectively. The RDA fragmentation reac-tion was also observed in the MS/MS spectrum indicated bythe literature (Qin et al., 2008). Therefore, compound 6 wasidentified as mangiferin by comparing its retention time andmass fragmentation pattern with those of the standard.

f (A) mangiferin and (B) iriflophene 2-O-a–L-rhamnopyranoside.

Phytochem. Anal. 2013, 24, 349–356Wiley & Sons, Ltd.

Page 5

Figure 3. LC–MS chromatograms of the extract of the leaves of Aquilaria sinensis. (A) LC–UV chromatogram monitored at 360 nm. (B) LC–ESI/MS totalion current (TIC).

Fragment Features of Xanthones and Benzophenones in A. Sinensis

Phytochem. Anal. 2013, 24, 349–356 Copyright © 2013 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/pca

353

Page 6

Table 2. Calibration curves, LOD and LOQ for eight analytes

Analyte Calibration curve Correlation coefficient (r2) Linear range (mg/mL) LOD (ng/mL) LOQ (ng/mL)

6 y = 1 � 106x + 26.57 0.9997 0.3190–6.380 16.615 39.8758 y = 84238x + 0.81 0.9998 0.1420–2.830 12.150 42.31516 y = 247587x � 0.07 0.9997 0.0196–0.392 3.840 12.61217 y = 1 � 106x � 2.60 0.9990 0.0196–0.392 2.168 5.94018 y = 1 � 106x � 2.15 0.9992 0.0083–0.166 1.664 4.32019 y = 1 � 106x � 2.54 0.9997 0.0187–0.374 2.025 6.82420 y = 3 � 106x � 1.32 0.9994 0.0008–0.016 0.212 0.64221 y = 2 � 106x � 1.05 0.9994 0.0020–0.040 0.396 1.228

Table 3. Precision, repeatability and stability for eight

Q. YU et al.

354

Benzophenones are a group of compounds that becomeprenylated and/or further cyclised producing numerousstructurally unique compounds. Compound 8 was used toanalyse the fragmentation pathway of benzophenones. TheESI/MS of this compound produced a [M � H]� ion at m/z 391(Fig. 2B). Its MS2 spectrum produced a base peak at m/z 245due to the loss of C6H10O4 (146 Da), which complied withthe fragmentation behaviour of flavone O-glycosides. TheMS3 analysis of the precursor ion at m/z 245 suggested theloss of a CO (28 Da) and a B ring (94 Da), and resulted in thepredominant product ion at m/z 217 and 151, respectively. Alow intensity product ion at m/z 107 was observed in the MS3

spectrum of the precursor ion at m/z 151 due to the loss ofCO2 (44 Da). Taken together, compound 8 was identified asiriflophenone 2-O-a-L-rhamnopyranoside by comparing itsretention time and mass fragmentation pattern with thoseof the standards. The proposed fragmentation pathway isgiven in Fig. 2B.

analytes

Analyte Precision RSD (%) RepeatabilityRSD (%)

StabilityRSD (%)

Intraday Interday

6 0.837 2.407 1.110 0.7638 1.192 2.736 0.409 1.37816 0.585 3.209 1.765 1.52617 0.923 2.758 1.461 1.44518 0.878 2.806 1.290 1.35119 0.825 2.835 0.836 0.83020 1.102 2.955 1.072 0.99921 1.789 2.461 1.716 0.626

Table 4. Recovery of eight analytes determined by standardaddition method (n = 5)

Analytes Originalmean (mg)

Spikedmean (mg)

Foundmean (mg)

Recoverymean (%)

RSD(%)

6 21.33 21.33 42.91 101.18 2.828 10.59 10.59 21.41 102.16 1.5216 1.04 1.04 2.09 100.77 4.3117 0.041 0.041 2.09 103.02 4.5318 0.53 0.53 1.07 103.62 2.2019 0.36 0.36 0.74 103.88 2.0720 0.034 0.034 0.069 102.94 1.9521 0.10 0.10 0.21 102.91 2.71

Identification of phenolic constituents in the leaves ofA. sinensis

In total, 21 phenolic compounds including xanthones (1, 2, 5, 6, 7and 9), benzophenones (3, 4, 8 and 10–13) and flavones (14–21)were identified or tentatively characterised based on theirretention behaviours, UV and MS spectra with those of authenticcompounds or data from te literature. These compounds containthree new (11, 12 and 13) and three known compounds (2, 7and 9), which had not been previously isolated from the leavesof A. sinensis. The related LC–ESI/MSn data are presented in Table 1and chemical structures are shown in Fig. 1. The UV chromato-grams at 360 nm and the MS TIC chromatograms of the A. sinensisextract are presented in Fig. 3.

Compound 11 showed a [M� H]� ion atm/z 433. The [M� H]�

ion yielded a prominent fragmentation ion in theMS/MS spectrumat m/z 245 due to the loss of C8H12O5 (188 Da). The MS2 and MS3

spectra were consistent with that of 8, which characteristicallycomplied with benzophenones with an additional acetyl groupin the sugar moiety. Additionally, compounds 12 and 13 had thesame mass as 11 but different HPLC retention times, suggestingacetyl groups in different locations in the sugar moiety,which were not identified previously. Therefore, compounds11–13 were preliminarily identified as new structures, namelyiriflophenone [2-(2-O-acty-a-L-rhamnopyranosyl) oxy], iriflophe-none [2-(3-O-acty-a-L-rhamnopyranosyl) oxy] and iriflophenone[2-(4-O-acty-a-L-rhamnopyranosyl) oxy] (Fig. 1).

Compounds 2, 7 and 9 were tentatively identified as neoman-giferin (Lu et al., 2008b), isomangiferin and homomangiferin

Copyright © 2013 Johnwileyonlinelibrary.com/journal/pca

(Wu et al., 2010), respectively, which were initially identified inthis plant.

Compounds 14 and 15 have common fragmentation thatoccurred by the loss of the following groups: CH3 (15 Da), H2O(18 Da), CO (28 Da), CO2 (44 Da), as well as RDA fragmentationreactions, which were the common features of flavonoids. Inaddition, they gave the [M� H]� ion atm/z 477 and 491, respec-tively. Their MS/MS spectrum yielded ions at m/z 301 and 315due to the loss of C6H8O6 (176 Da), respectively, suggestingthe presence of a glucanic acid group in the two compounds.By comparing our data with the literature, compounds 14 and15 were identified as hypolaetin-5-O-b-D-glucuronopyranosideand aquilarisin (Feng et al., 2011), respectively.

Phytochem. Anal. 2013, 24, 349–356Wiley & Sons, Ltd.

Page 7

Table 5. Content of phenolic constituents in the leaves of A. sinensis collected from 11 different regions (n = 3)

Sample

Analyte Content (mg/g) Total(mg/g)

Total(%)

6 8 16 17 18 19 20 21

Dongguan-1 13.420 8.330 0.641 0.018 0.366 0.221 0.020 0.050 23.066 2.307Dongguan-2 39.599 29.142 1.186 0.026 0.370 0.204 0.024 0.025 70.576 7.058Dongguan-3 32.937 27.887 1.225 0.019 0.619 0.222 0.033 0.051 62.993 6.299Dongguan-4 41.577 37.535 1.096 0.073 1.253 0.735 0.127 0.092 82.488 8.249Dongguan-5 22.963 10.120 0.654 0.018 0.203 0.153 0.023 0.019 34.153 3.415Baidian-1 26.947 16.726 1.287 0.037 0.735 0.444 0.041 0.101 46.318 4.632Baidian-2 26.599 19.855 1.383 0.076 1.289 0.603 0.039 0.137 49.981 4.998Baidian-3 30.362 17.589 1.707 0.047 0.915 0.494 0.040 0.135 51.289 5.129Guangxi-1 28.621 20.774 1.192 0.040 0.369 0.354 0.029 0.022 51.401 5.140Guangxi-2 16.351 6.717 0.929 0.031 0.069 0.073 0.008 0.013 24.191 2.419Guangxi-3 11.046 10.401 0.833 0.014 0.039 0.048 0.007 0.007 22.395 2.240

Fragment Features of Xanthones and Benzophenones in A. Sinensis

Based on the fragmentation characteristics, part of the frag-mentation behaviours gave a regular pattern in the phenoliccompounds. The xanthone and benzophenone constituentsin the leaves of A. sinensis have similar mass fragmentationfeatures to flavones such as flavone O-glycosides and C-glycosides. Therefore, these fragmentation behaviours couldbe extended to other phenolic compounds, to help identifythem by LC–MS in complex systems such as herbs andherbal prescriptions.

Method validation

Calibration curves were constructed by plotting the peak area(y) versus content (x; mg) by analysing a set of standardsolutions. All the calibration curves showed a good linearitywith correlation coefficients (r2) no less than 0.999. The limitsof detection (LOD) and quantification (LOQ) were determinedat a signal-to-noise ratio (S:N) of approximately 3 and 10,respectively (Table 2).

Precision, repeatability and stability of the method were alsovalidated for each analyte. The intra- and interday precisionswere determined by analysing known concentrations of theeight analytes in the six replicates during a single day and byduplicating the experiments on three successive days. In orderto confirm the repeatability, five different working solutionswere prepared and analysed from the same sample (Baidian-1,Guangdong Province). Stability of the sample solution wastested at room temperature. The sample solution was analysedat 0, 2, 4, 6, 12 and 24 h to assess stability. Relative standarddeviations (RSD) of intra- and interday precision, repeatabilityand stability were less than 1.8%, 3.2%, 1.8% and 1.5%, respec-tively (Table 3), indicating good precision, repeatability andsample stability.

Recovery was performed by the addition of a known amountof each individual standard into a certain amount of A. sinensisleaves (1.25 g). The mixture was extracted and analysedfollowing the procedure described above, and the analyticalmethod developed was shown to be accurate with recoveriesof 100.8–103.9% (RSD < 4.5%, Table 4).

35

HPLC fingerprint and quantitative analysis

The relationship within a set of chromatographic fingerprintscan be analysed through comparison of similarity between the

Phytochem. Anal. 2013, 24, 349–356 Copyright © 2013 John

objects and the reference fingerprint. The similarity values of11 samples were 0.998, 0.997, 0.998, 0.999, 0.999, 0.998, 0.983,0.999, 0.994, 0.996 and 0.998.The quantitative method developed was used to analyse

cultivated or wild grown A. sinensis leaves in China and theresults are summarised in Table 5. The results show that thedifferences may be related to different environmentalconditions causing changes in the compositions of the phenoliccompounds. The total amount of the eight phenolic compoundsin the leaves of A. sinensis collected from two main producingareas: Guangdong (dongguan, baidian) and Guangxi werecompared with each other, and for medical purposes the plantsfrom Guangdong were considered to be of higher quality. Also,mangiferin was found to be the most abundant compound, witha content reaching 4.2%. The leaves of A. sinensis have a highcontent of mangiferin (Nong et al., 2005; Deng et al., 2006;Chen, 2007) and could be turned into an important source ofthis compound in the future.

Acknowledgements

This research was funded by the National Natural ScienceFoundation of China (NSFC No. 30973965, 30901956), SelectedScience and Technology Activities Foundation for the ReturnedOverseas Scholars of Nanjing, and the Priority AcademicProgram Development of Jiangsu Higher Education Institutions.

REFERENCESAdzu B, Amos S, Kapu SD, Gamaniel KS. 2003. Anti-inflammatory and

anti-nociceptive effects of Sphaeranthus senegalensis. J Ethnopharmacol84: 169–173.

Chen YR. 2007. HPLC determination of mangiferin content in commercialRhizoma Anemarrhenae. Chin Pharm Anal 27: 1478–1480.

Deng JG, Feng X, Wang Q, Tan JP, Ye Y. 2006. Comparative study ofmangiferin content from different origin and different varieties ofmango leaves. Chin Trad Patent Medi 28: 1755–1756.

Fabre N, Rustan I, Hoffmann ED, Quetin-Leclercq J. 2001. Determinationof flavone, flavonol, and flavanone aglycones by negative ion liquidchromatography electrospray ion trap mass spectrometry. J Am SocMass Spectrom 12: 707–715.

Feng J, Yang XW, Wang RF. 2011. Bio-assay guided isolation andidentification of a-glucosidase inhibitors from the leaves of Aquilariasinensis. Phytochemistry 72: 242–247.

Kou JP, Yu BY, Qi J, Inagaki N, Zhu DN. 2008. Medical Use of 2-O-a-L-rhamnosyl-4,6,40-trihydroxy benzophenone. Patent No. CN101439040, Application No. CN 200810244795.

Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/pca

5

Page 8

Q. YU et al.

356

Kou JP, Yu BY, Qi J, Inagaki N, Zhu DN. 2010. Medical application of2-O-a-L-rhamnose-4,6,40-trihydroxybenzophenone. Patent No. CN101816668 Application No. CN 20101143336.

Lu JJ, Qi J, Zhu DN, Yu BY. 2008a. Antioxidant activity and structure–activity relationship of the flavones from the leaves of Aquilariasinensis. Chin J Nat Med 6: 456–460.

Lu XR, Qin MJ, Xu DR. 2008b. Identification of main chemical compo-nents of tongguanwan prescription by HPLC–ESI–MS/MS. Chin J NatMed 6: 283–291.

Nong CZ, He WS, Fleming D, Pan LL, Huang HY. 2005. Capillary electro-phoresis analysis of mangiferin extracted from Mangifera indica L.bark and Mangifera persiciformis C.Y. Wu et T. L. Ming leaves. J Chro-matogr B 826: 226–231.

Qi J, Lu JJ, Liu JH, Yu BY. 2009. Flavonoid and a rare benzophenoneglycoside from the leaves of Aquilaria sinensis. Chem Pharm Bull57: 134–137.

Qin JP, Deng JG, Feng YQ, Feng X. 2008. Applications of organic massspectrometry in structural identification of an impurity compoundin prepared mangiferin extracted from Mangifera indica L. leaves. JChin Mass Spectr Soc 29: 218–225.

Schieber A, Berardini N, Carle R. 2003. Identification of flavonol and xan-thone glycosides from mango (Mangifera indica L. Cv. ‘Tommy Atkins’)peels by high-performance liquid chromatography–electrospray ioniza-tion mass spectrometry. J Agric Food Chem 51: 5006–5011.

Copyright © 2013 Johnwileyonlinelibrary.com/journal/pca

Wang HG, Zhou MH, Lu JJ, Yu BY. 2008a. Antitumor constituents from theleaves of Aquilaria sinensis (Lour. ) Gilg. Chem Ind Forest Prod 28: 1–5.

Wang Y, Yang L, He YQ, Wang CH, Welbeck EW, Bligh SW, Wang ZT.2008b. Characterization of fifty-one flavonoids in a Chinese herbalprescription Longdan Xiegan decoction by high-performance liquidchromatography coupled to electrospray ionization tandem massspectrometry and photodiode array detection. Rapid Commun MassSp 22: 1767–1778.

Wu ZT, Wei G, Lian GY, Yu B. 2010. Synthesis of mangiferin, isomangi-ferin, and homomangiferin. J Org Chem 75: 5725–5728.

Ye M, Liu SH, Jiang ZL, Lee YS, Tilton R, Cheng YC. 2007. Liquid chro-matography/mass spectrometry analysis of PHY906, a Chinesemedicine formulation for cancer therapy. Rapid Commun MassSp 21: 3593–3607.

Yoshimi N, Matsunaga K, Katayama M, Yamada Y, Kuno T, Qiao Z, Hara A,Yamahara J, Mori H. 2001. The inhibitory effects of mangiferin, anaturally occurring glucosylxanthone, in bowel carcinogenesis ofmale F344 rats. Cancer Lett 163: 163–170.

Yu BY, Kou JP, Zhou MH, Zhu DN, Yuan ST. 2011. One kind of the leavesof Aquilaria sinensis extract and its pharmaceutical and health careuse. Patent No. CN 101011499, Application No. CN 200710019969.

Zhou MH, Wang HG, Suolangjiba, Kou JP, Yu BY. 2008. Antinociceptiveand anti-inflammatory activities of Aquilaria sinensis (Lour.) Gilg,leaves extract. J Ethnopharmacol 117: 345–350.

Phytochem. Anal. 2013, 24, 349–356Wiley & Sons, Ltd.