全 文 ：2006年 9月
September 2006
色 谱
Chinese Journa l of Chrom atography
Vo.l 24 No. 5
499 ～ 502
R eceived date:2006-02-21
Corresponding author:HUANG Xuesong, ma le, PhD , professor, T el:(020)88152236, E-m a il:thxs1957@ yahoo. com. cn.
Simultaneous Determ ination of Trilobolide-6-O -Isobutyrates
A and B inWedelia trilobata by Gas Chromatography
HUANG Xuesong, OU Shiyi, TANG Shuze, FU Liang, WU Jianzhong
(Departm ent of Food Science and Technology, College of Science and Engineering ,
Jinan University, Guangzhou 510632 , China)
Abstract:A sim p le, sens itive, and specific gas chrom atograph ic(GC)m ethodwas developed to determ ine the
m ain bioactive sesqu iterpene lactones, trilobo lide-6-O-isobu tyrates A and B (TBO-A and TBO-B), inW edelia
triloba ta, a usefu l fo lk herb. A comm ercially availab le HP-5 cap illary co lumn (30 m ×0. 25 mm .i d. ×0. 25
μm) was ut ilized for the d irectdeterm ination o fTBO-A and TBO-B inW . trilobata. Calib ration curves were ob-
tained by sp iking au then tic com pounds and the in terna l standard ( feru lic acid) in toW . trilobata samp les before
extraction. Ex traction was carried ou t by reflux ing the dried herb (0. 5 g) fo r 1 h inm ethano l(25mL). A ll ca l-
ibration curves showed good linear regressions(r2 >0. 992) w ith in test ranges. The assayw as reproducib le and
accurate w ith the overall in traday and in terday re lative standard deviations and accuracies of less than 10% and
m ore than 90%, respect ively. The developed GC method was successfu lly u tilized to analyze the TBO-A and
TBO-B in aerial parts and flowers ofW. triloba ta , indicat ing that it was su itable for the quality con tro l o f th is
common ly used herb and re lated trad it ional Ch inesem ed icines.
Key words:gas chrom atography (GC); trilobo lide-6-O-isobutyrate A; trilobolide-6-O-isobu tyrate B;eudes-
m anolide
CLC number:O658　Document code:A　Article IC:1000-8713(2006)05-0499-04　Column:Articles
气相色谱法同时测定蟛蜞菊中的蟛蜞菊倍半萜内酯 A和 B
黄雪松 , 　欧仕益 , 　唐书泽 , 　傅　亮 , 　吴建中
(暨南大学理工学院食品科学与工程系 , 广东 广州 510632)
摘要:蟛蜞菊具有抗肿瘤 、抗病毒作用 ,其主要活性成分是倍半萜内酯类物质。以分离纯化所得的结晶倍半萜内酯
A与 B为参考标准 , 采用 HP毛细管色谱柱(30 m ×0. 25 mm .i d. × 0.25μm), 程序控制升温 ,氢火焰离子化检测
器检测 , 测定了蟛蜞菊的地上部分(茎叶和花)中倍半萜内酯 A与 B的含量。测定结果表明:蟛蜞菊茎叶中的倍半
萜内酯 A和 B的含量分别为(239±6.4)μg /g和(156±15)μg /g;花中倍半萜内酯 A和 B的含量分别为(233 ±
6. 5) μg /g和(173±16)μg /g。该法可用于蟛蜞菊原料及其药品的质量控制。
关键词:气相色谱法;倍半萜内酯 A;倍半萜内酯 B;蟛蜞菊
　　W edelia is a large genus of the Compositae
fam ily. Many species of the genus are used in fo lk
m edicine to prevent or cure a variety of diseases,
such as swe lling, headaches, fevers, hepatitis
[ 1] , in fections ofdigestion and pathologies of the
respiratory tract [ 2 - 3] , because they contain
bioactive components, such as ent-kaurenic acids
[ 4] , lactones, flavones, diterpenes and wedelo-
lactones [ 5 -7] .
　　W edelia trilobata belongs to the genus of
Compositae fam ily. It is a very noxious weed
found in certain places and its contro l is d ifficu lt
[ 8] . However, it is w idely used as a decorative
ground cover in gardens, on either sides of a
path, on the edges of a ponds and other public
places. Furthermore, it is also used as a folk
medicine in som e countries like, China, India,
and Singapore. It is reported that the n-hexane
and ethyl acetate extracts show antibacterial ac-
tiv ity against bacillu s subtilis, mycobacterium
sm egmatis , staphylococcus aureus , and other
bacteria, but not against yeast and fungi [ 9] .
Thyagarajan et a l. [ 10] have isolated five com-
pounds from Eclipta alba andW. trilobata and
found that they have antifunga l activity against
n ine species.
　　 In recent years, a num ber of eudesmanolide
sesqu iterpenes or sesqu iterpene lactones (F ig. 1-
7b) have been isolated from this genus and from
W . trilobata [ 4 - 6, 11 - 12] . Few studies about
色 谱 第 24卷
the bioactivities of these sesquiterpene lactones
have been reported although the bioactivities of
other sesquiterpene lactones have been rev iew ed.
As part of the effort to search for antitumor a-
gents from natural sources, the chem ical compo-
nents ofW. trilobata have been investigated, and
three compounds, that is, epimers, trilobolide-6-
O-isobutyrate A (TBO-A), trilobolide-6-O-isobu-
tyrate B (TBO-B), and 5a shown in F ig. 1, were
isolated from the aerial part and the flower ofW .
trilobata [ 13] . Their chem ical structures (F ig.
1) have been reported previously[ 4 - 5, 11] , but
no determ ination method has been g iven. In th is
article, a simultaneous determ ination method of
TBO-A and TBO-B inW . trilobata by gas chroma-
tography (GC) is reported.
F ig. 1　Sesquiterpene lactones inW edelia trilobata
1　Experim ental
1. 1　Chem icals andmaterials
　　Ferulic acid and organic solvents of analytical
reagent gradewere all purchased from Sigma (St.
LouisMO , USA). TBO-A and TBO-B were isola-
ted in the laboratory fromW. trilobata which w as
planted in the campus of the J inan Univers ity,
Guangzhou, China. The purities of the TBO-A
and TBO-B were higher than 95% and 93%, re-
spectively, analyzed by high performance liquid
chromatography (HPLC ), and their identities
were confirmed by X-ray analysis and m ass spec-
trometric analys is referred to in the published arti-
cles [ 4 - 5, 11, 13] .
1. 2　Apparatus and chromatographic condi-
tions
　　GC analysis was performed using HP6890 se-
ries gas chromatograph ( Hew lett-Packard )
equipped w ith a flame ionization detector (F ID).
Data were recorded and analyzed byGC Chemsta-
tion Rev. A. 06. 03 [ 509] in tegrator. An HP-5
capillary column (30 m ×0. 25 mm .i d. ×0. 25
μm)was utilized. Helium was used as the carrier
gas at a flow-rate of 2mL /m in. The oven temper-
ature was adjusted as fo llow s:in itial temperature
60 ℃ w ith 25 ℃ /m in rise to 215 ℃ and held at
215 ℃ for 10m in, and then 30℃ /m in rise to 270
℃ and held at270℃ for 15m in. Both the in jector
and detector temperatures were set at 250℃. An
aliquot (2 μL) of samplew as in jected w ith a split
ratio of 50∶1.
1. 3　Calibration curves
　　MeOH stock solutions contain ing TBO-A and
TBO-B were prepared and then diluted to appro-
priate concentration ranges for the construction of
calibration curves. Each calibration curve was
perform ed in trip licate w ith five different concen-
trations. The concentration of the internal stand-
ard, feru lic acid, was 10 mg /L. Calibration
curvesw ere plotted by spik ing authentic analytes
and the interna l standard into M eOH (25 mL)
contain ing a powder of the aerial parts or flowers
ofW. trilobata (0. 50 g), before reflux. The re-
sultantm ixtures were then extracted as described
in Section 1. 6. The extract (2 μL) of theM eOH
solution was directly analyzed by GC. For the
control sample, the extract ofW . trilobata spiked
w ith the internal standard only w as prepared and
analyzed in the sam e manner. Calibration curves
were derived by plotting concentrations of each
analyte as a function of difference of peak area ra-
tio (peak area ratiospiked - peak area ratiocontrol)
betw een spiked and control extracts.
1. 4　Accuracy and precision
　　The measurements of intraday and interday
re lative standard dev iations (RSD) were utilized
to determ ine the accuracy and precision of the de-
veloped assay. Known quantities ofTBO-A , TBO-
B, and the internal standard were added to the
MeOH solution (25 mL) contain ingW . trilobata
(0. 5 g) before reflux. Control samples spiked
w ith only the internal standard were also pre-
pared. The resultant samples were extracted and
analyzed as described in Section 1. 6. The differ-
ence of peak area ratio betw een the testing and
control samples for each analyte was calculated,
and the quantity of each analytewas subsequently
obtained from the corresponding calibration
500
第 5期 HUANG Xuesong, et a l:Simultaneous Determ ination of Trilobolide-6-O-Isobu tyrate A and B inW edelia trilobata by Gas Chrom atography
curve. Each sample was analyzed in trip licate to
determ ine the intraday RSD. The RSD was taken
as a measure of precis ion and the percentage
difference between the amount determ ined and
the amount spiked was cons idered as a m easure
of accuracy. The interday reproducib ility was ex-
am ined on three separate days.
1. 5　Detection lim its
　　A liquots of TBO-A and TBO-B were spiked
into the MeOH solution (25 mL) contain ingW .
trilobata (0. 5 g) to obtain a concentration range
of 0. 1 - 1. 0 mg /L. The resu ltant m ixtures were
extracted and analyzed in the sam em anner as de-
scribed in Section 1. 6. The detection lim it for
each analytew as determ inedwhen the ratio of the
peak height of the analyte to noise was greater
than five.
1. 6　Analysis of the TBO-A and TBO-B in the
aerial parts and flowers ofW. trilobata
　　To the ground dried aerial parts and flow ers
ofW. trilobata ( 0. 4 - 0. 6 g, adjusted according
to the TBO-A and TBO-B contents), 25 mL of
M eOH and 250 μL of the internal standard solu-
tion (1 g /L) were added. The m ixture w as re-
fluxed for 2 h and then centrifuged at 4 000 r /m in
for 10 m in. The supernatant (2 mL) was trans-
ferred into a vial and directly subjected to GC a-
nalysis. T he contents of the analytes w ere deter-
m ined from the corresponding calibration curves.
2　Resu lts and d iscussion
2. 1　Chrom atogram of TBO-A and TBO-B
　　F ig. 1 show s the chem ica l structures ofTBO-
A and TBO-B. The aim of the present study w as
to develop a direct GC analyticalmethod for ses-
quiterpene lactones. F ig. 2 show s thatTBO-A and
TBO-B were resolved we ll w ith base-line separa-
tions on theHP-5 column (30 m ×0. 25 mm .i d.
×0. 25 μm).
　　As shown in F ig. 2, TBO-A and TBO-B were
isolated completely, and their peaks were sharp
and symmetrica.l It also show s that the TBO-A
and TBO-B isolated from the materials m ight be
used as standards in the determ ination of TBO-A
and TBO-B in theW. trilobata.
　　The retention time of the internal standard,
feru lic acid, was (6. 209±0. 003)m in (n =15).
The relative retention times of internal standard to
TBO-A and TBO-B were 5. 13±0. 05 (n =15) and
4. 39±0. 03 (n =15), respectively.
F ig. 2　Chromatogram of TBO-A and TBO-B
2. 2　Accuracy and precision
　　The two calibration curves for TBO-A and
TBO-B showed good linear regressions (Table
1). Furthermore, the results demonstrated that
the developed directGC ana lyticalm ethod was re-
producible w ith good accuracy. The overa ll intra-
day and interday RSDs were less than 10% for all
analytes, and the overall intraday and interday ac-
curacies for the determ ination ofTBO-A and TBO-
B tested were all greater than 90%(Table 2). In
addition, the method provided better sensitivity
for TBO-A , w ith a detection lim it of 0. 3 μg /g.
A lthough TBO-B also has satisfactory accuracy
and precision, its detection lim it (40μg /g ) was
130 times higher than that of TBO-A. This m ight
be because TBO-B wasm ore unstable thanTBO-A
and was liable to degradation in the chromato-
graphic environm ent ( for instance, the column
temperature may be as high as 270 ℃). In any
case, TBO-B m ight be determ ined simultaneous ly
w ith TBO-A if the samplewasw ell prepared.
　　B lank controls are generally unavailable for
the study ofherbalmaterials, and thus calibration
curves are normally constructed w ithout us ing the
internal standard method. Therefore, reproduc-
ib ility and extraction yield becom e critical for
quantification of the principal components in
herbs. To solve th is problem, an internal stand-
ard method w as also developed. Ferulic acid was
selected as the internal standard, which was not
detected in the material sample. In brief, both
the internal standard, namely ferulic acid, and
the ana lytes tested were spiked into the samples
before extraction, although for the control, only
the internal standard was added separate ly to s im-
ilar herbal samples before reflux. Calibration
curvesw ere then constructed as a function of the
concentration of ana lyte versus the differences in
501
色 谱 第 24卷
peak area ratio between spiked and nonspiked
(control)W. trilobata extracts. The results of
the calibration curve and internal standard meth-
ods can be seen in Table 3, and their results are
almost the sim ilar. It m ight be concluded from
the s tandard deviation that the internal standard
method w as more precise than the calibration
method.
Table 1　Calibration curves for TBO-A and TBO-B determ inations
Com ponent Retention tim e /m in Ca libration curve1) r2 Linear range /(μg /g) D etection lim it /(μg /g)
TBO-A 24. 83 y =0. 995x +0. 954 0.997 0. 6 - 9. 6 0. 3
TBO-B 21. 49 y =76. 79x - 47. 03 0.992 60 - 640 40
　　 1) y:d ifference of peak area ratio (peak area ra tiosp iked - peak area ratiocontrol);x:concentration of ana ly te (μg /g).
Tab le 2　 Intraday and in terday RSDs for the assay of TBO-A and TBO-B (n =3)
Com ponent
Added /
(μg /g)
Intraday variabi lity
de tected (m ean±SD) /(μg /g) RSD /%1) accuracy /%2)
Interday variability
de tected (m ean±SD) /(μg /g) RSD /% accuracy /%
TBO-A 0. 6 0. 55±0.021 3. 8 91. 7 0. 56±0. 023 4.1 93.3
TBO-A 1. 2 1. 16±0.064 5. 5 96. 7 1. 17±0. 062 5.3 97.5
TBO-A 2. 4 2. 31±0.22 9. 5 92. 1 2. 35±0. 18 7.6 95.0
TBO-B 80 73. 7±3.60 4. 9 96. 3 72. 2±3. 75 5.2 90.3
TBO-B 160 144. 5±13. 87 9. 6 90. 3 145. 9±11. 96 8.2 91.2
TBO-B 320 298. 2±29. 10 9. 6 93. 2 302. 4±9. 68 3.2 94.5
　 1) RSD (relative standard dev ia tion)(%)=SD /mean×100%. 2) Accuracy (%)=(1 - |m ean concentra tion m easured - concentra-
tion sp iked|/concen tra tion spiked)×100%.
Tab le 3　Contents of TBO-A and TBO-B inWedelia trilobata (n=3)
　Sam ple De tection m ethod Content of TBO-A(m ean±SD) /(μg /g)
Content of TBO-B
(mean±SD) /(μg /g)
Tota l content of TBO-A and
TBO-B (m ean±SD) /(μg /g)
Ae ria l pa rts cal ibra tion curve 239±6.4 156±15 393±12
interna l 235±4.8 155±13 389±10
F lowe r par ts cal ibra tion curve 233±6.5 173±16 406±12
interna l 231±5.8 172±13 　403±9. 8
2. 3　Conten ts of TBO-A andTBO-B in theW.
trilobata
　　As shown in Table 3, the contents ofTBO-A
and TBO-B in theW. trilobata were from 389 to
406μg /g. The contents ofTBO-A and TBO-B be-
tween the aerial parts and flower show ed no s ig-
nificant difference. How ever, for TBO-A , the
contents of both the aeria l parts and flowers were
higher than that of TBO-B. This difference m ight
be attributed to the difference in stability between
TBO-A and TBO-B, which meant that TBO-A
m ight bemore stable than TBO-B.
3　Conclusions
　　To date, the presently deve loped direct GC
m ethod for the analysis of TBO-A and TBO-B is
an effective quantification method for determ ining
the m ain active components in W. trilobata ,
which is commonly known as a folk m edicine.
This GC analytica l assay is sensitive, accurate,
and reproducible. It prov ides a suitable quality
control method forW . trilobata samples and can
be readily utilized for determ ining themajor active
ingredients present inW. trilobata.
References:
[ 1] 　Lin S C, Lin C C, Lin Y H , Shyuu S J. Am erican Journal of
Chinese Medicine, 1994, 22:155
[ 2] 　Roque N F, G ianne llaT L, G iesbrechtA M , BarbosaR deC
S B C. Rev La tinoam Qu im , 1987, 18:110
[ 3] 　M iles D H , Ch ittawong V, Payne A M , H ed in P A, Kokpol
U. J Agric Food Chem , 1990, 38:1 591
[ 4] 　Farag S F, E lemary N A , N iw a M. Chem Pharm Bulle tin,
1996, 44:661
[ 5] 　Boh lmann F, Gerke T, Jakupov ic J, Borthakur N , K ing R
M. Phy tochem istry, 1984, 23:1 673
[ 6] 　Boh lmann F, AngW , K ing R M , Robinson H. Phy tochem is-
try, 1984, 23:2 069
[ 7] 　Block L C, Santos A R S, de SouzaM M , Sche idt C, Yunes
R A, Santos M A, De lle Monache F, Cech ime l V. J E thno-
pharm acol, 1998, 61:85
[ 8] 　Tham an R R. W edelia tri lobata:Da isy Invader of theP acific
Islands. [ 2006-01-20] . http:/ /www. issg. org /database /
species /reference_ files /sphtr i /Wedtri_Thaman. pdf
[ 9] 　T adde iA, Rosas-Romero A J. Phytomed icine, 1999, 6(2):133
[ 10] 　Thyagarajan R K , K rishnaswam y N R. G eobios, 1999, 26:
91
[ 11] 　Ragasa C Y, PadolinaW G, Bowden B F, L i S X, Tapiolas
D M , Coll J C. J N at Prod, 1994, 56:386
[ 12] 　Ferreira D T, Levora to A R, F aria T D J, D e Corva lho M
G , Braz-F ilho R. Na t P rod Lett, 1994, 4(1):1
[ 13] 　Huang X S, JiangRW , OoiV E C. ActaCryst, 2003, E59:771
502