Ｒeceived ：January 20，2014Accepted ：May 7，2014
Foundation item ：Major Progress of National New Drug Ｒearch （2012ZX09103201-038）
*Corresponding author Tel ：86-21-34204806；E-mail ：xbli@sjtu．edu．cn
天然产物研究与开发Nat Prod Ｒes Dev 2015，
27：126-133文章编号：1001-
6880（2015）1-0126-08土木香内酯与异土木香内酯的提取与纯化工艺
张乐，王梦月，李晓波
*
上海交通大学药学院中药基因组学实验室，上海200240
摘
要：本文报道了一种从土木香药材中提取以及纯化倍半萜内酯的方法。乙醇回流提取的方法采用正交试
验设计方法［L 9（34）］，考察了4个因素对产率与内酯浓度的影响，最终确定了乙醇回流提取的最佳方法：药材加10倍量的95%乙醇，
加热回流提取1次，2h 。而硅胶柱层析的最佳纯化方法为：以100 200目的硅胶，1ʒ3的上样量，径高比为1ʒ2。其洗脱剂成分为石油醚ʒ丙酮（100ʒ3，v /v ），洗脱流速为35mL /min 。采用该方法，异土木香内酯与土木香内酯的提取率大于70%，其纯度大于90%。关键词：土木香内酯；异土木香内酯；正交试验；硅胶柱层析；土木香中图分类号：Ｒ932
文献标识码：A
DOI ：10．16333/j．1001-6880．2015．01．026
手术后吃什么营养品好
Extraction and Purification of Alantolactone and
Isoalantolcatone from Ｒoot of Inula helenium
ZHANG Le ，WANG Meng-yue ，LI Xiao-bo *
School of Pharmacy ，Shanghai Jiao Tong University ，Shanghai 200240，China
Abstract ：The extraction and purification method of squiterpene lactones from the root of Inula helenium L．was opti-mized．The tests of orthogonal array design ［L 9（34）］were performed to analyze the effect of single factors ，namely etha-nol concentration ，extraction duration and times of extraction ，on the extraction yield of isoalantolcatone and alantolac-tone．As the result ，the optimal extraction conditions were found as follows ：the dried roots of I．helenium were extracted by reflux with 95%ethanol in solid /liquid ratio of 1ʒ10（w /v ）for 2hrs．The optimized purification process of ethanol extract was achieved by a pressurized silica gel （100-200mesh ）column chromatography with diameter to height ratio of 1ʒ2．The ratio of mixed sample to silica gel was 1ʒ3（w /w ），the eluent compod of petroleum ether （30-60ħ）and acetone （100ʒ3，v /v ），flowed at the rate of 35mL /min．The extraction yield of the lactones under the optimal method was above 70%，and the purity was above 90%．
Key words ：Isoalantolcatone ；alantolactone ；orthogonal array design ；silica gel column chromatography ；Inula helenium L．
Introduction
Inula helenium L．（Compositae ）is a perennial plant ，distributed in many areas of China ，Euro
pe and North America．The root of I．helenium is rich in eudesmane-type squiterpene lactones ，such as isoalantolcatone and alantolactone ［1］．In recent years ，extensive investi-gations on active constituents of I．helenium roots proved that squiterpene lactones were responsible for various biological activities ，such as antibacterial ，anti-fungal ，hepatoprotective and antitumor activities ［2-4］．
Therefore ，it is necessary to find an efficient method for extraction and purification of squiterpene lactones．The current study was mainly focud on the ultrasonic-assisted extraction of phenolic and microwave-assisted
extraction of flavones ［5，6］，but the purity was not repor-
ted．In this study ，we report an improved extraction method for squiterpene lactones by ethanol-reflux and purification by silica gel column chromatography．The extraction yield of the lactones obtained by the optimal method was above 70%，and the purity was above 90%．
Materials and methods
Plant material and reagents
Dried roots of I．helenium were purchad from the
herbal market in Sichuan Province，China．The plant material was ground and kept in a cool and dark place．Alantolactone and isoalantolcatone standards were in-ternally isolated and characterized by spectral methods （IＲ，NMＲand MS）［7，8］．The purity of each reference substance was above98%as determined by UPLC．All the solvents ud in the experiments were of analytical grade．
Instruments
UPLC（Agilent1290），Ultraviolet detector（G4212-60008），Ｒotovapor，Electric heating blast drying cham-ber，Electric balance，etc．
Standard preparation
Each reference substance was dissolved in methanol to the concentration of0．115，0．230，0．345，0．460，0.575and0．690mg/mL of isoalantolactone，and 0.114，0．228，0．342，0．456，0．570，and0．684mg/ mL of alantolactone．The samples of reference sub-stance were injected in triplicate（injection volume1μL）．The retention times（ＲTs）of isoalantolactone and al
天赐石麟antolactone were2．864min and3．039min，re-spectively．Linear regression equations of isoalantolac-tone and alantolactone were y=3736．8x+21．674（Ｒ2=0．9997）and y=4176x+25．284（Ｒ2=0．9996），respectively．
Sample preparation
The sample extracts of I．helenium roots were dissolved in methanol for UPLC analysis．1μL solution of each sample extract was injected into UPLC system．Contents of isoalantolactone and alantolactone in the examined samples were calculated by linear regression equations and expresd as milligram per milliliter solutions （mg/mL）．
Each experiment was performed in triplicate．All the data values were expresd as means with standard de-viation．
Quantitative determination of isoalantolactone and alantolactone
The determination of isoalantolactone and alantolactone content was carried out on an Agilent Zorbax Eclip Plus C
18
column（2．1ˑ50mm，1．8μm）with a mo-bile pha consisting of40%acetonitrile and60%wa-ter，at a flow rate of0．6mL/min，and the detection wavelength was t at210nm．The column oven tem-
perature was t at35ħ．The injection volume was1μL．Quantitative data were obtained from the peak are-as of each sample using an external standard method．Ethanol-reflux extraction
To optimize the ethanol-reflux extraction conditions，4 parameters were lected and investigated，namely：ratio of solid to liquid（A），solvent concentration（B），ex-traction duration（C）and times of extraction（D）．The extraction process was further optimized using the or-thogonal array design shown in Table1．Dried roots of I．helenium（20g）and ethanol were put into flask for ethanol-reflux extraction with designed ratio of plant material to ethanol（solid/liquid，w/v）and various conditions of experiments．The extract was filtered，and solvent was removed under vacuum．The obtained etha-nol extract was dried and concentrated under vacuum．All experiments were performed in duplicate．
The yield（X
1
）of extracts were calculated bad on the weight of obtained extract and plant material，while
the extraction yield（X
3
drink的过去分词
）was defined as the ratio of the weight of lactones in extracts to the total weight of lactones in the plant material．The weight of lactones in
extracts was calculated bad on the purity（X
2
）of the extracts，the ratio of the weight of lactones to the ex-tracts．
The method of weighed score（Y）comparison was ud to evaluate the effects of every factor on the yield and purity，and Y was calculated using the following expres-
sion：Y=0．8X
2
/6．55-0．2X
1
/66．37，where6．55was the average purity of the extracts in group9（Table
2），which was the best in purity（X
2
）of the extracts，
while the66．37was the average of the yield（X
1
）of the group4，which was the best in yield．A single factor analysis of variance（ANOVA）was adopted to investi-gate the effect of each factor in the process of extrac-tion．
Purification by silica gel column chromatography The extracts obtained from the I．helenium roots by an optimized ethanol-reflux process were further purified by silica gel column chromatography．T
he chromato-graphic column was packed by dry column-packing．The mixed sample was compod of the ethanol-extracts
721
Vol.27ZHANG Le，et al：Extraction and Purification of Alantolactone and Isoalantolcatone fromＲoot of Inula helenium
Table1Levels and factors of orthogonal design
Levels
Factors
（A）Ｒatio of solid
to liquid（g/mL）
（B）EtOH
concentration（%）
（C）Extraction
duration（h）
（D）Extraction
times（n）
186511
2107522
312952．53
and chromatographic silica gel（1ʒ1．5，w/w），and the
following factors were investigated in the process of
purification：granularity of chromatographic silica gel，
loading amount，ratio of diameter to height，ingredients
of eluent and flow rate．The purified ethanol-extract was
dried，concentrated under vacuum and further dissolved
in methanol for UPLC analysis．
The method of weighted score（Y）comparison was
ud to evaluate every effect of factors on the yield and
purity as well，according to extraction yield of lactones
（X
机械电子工程专业3），purity（X
2
）and the amount of eluent（X
4
）
：Y
=0．2X
3+0．4X
2
-0．4X
4纯菜
，where the amount of eluent
（X
4
）（mL/g）was defined as the ratio of the amount of eluent ud for purifying ethanol extracts to the weight of plant material for producing the ethanol ex-tract．Each experiment was performed in triplicate．All the data values were expresd as means with standard deviation．
The experimental designs of every factor were as fol-lows：
Granularity of chromatographic silica gel
Three types（60-100mesh，100-200mesh and200-300 mesh）of granularities of chromatographic silica gel were tested．The ratio of diameter to height of column chromatography was1ʒ5，and the ratio of mixed sample to silica gel was1ʒ3（w/w）．The eluent of column chromatography was compod of petroleum ether（30-60ħ）and acetone（100ʒ3，v/v），at a flow rate of5．5mL/min．
Loading quantity
The experiment was performed using silica gel（100-200mesh）column chromatography by eluent compod of petroleum ether（30-60ħ）and acetone（100ʒ3，v/ v），at a flow rate of5．5mL/min．Different loading quantities of mixed sample were tested．The ratios of mixed sample to silica gel were1ʒ6，1ʒ4，1ʒ3，1ʒ2，1ʒ
1（w/w），respectively．
Ｒatio of diameter to height
The experiment was performed by silica gel（100-200 mesh）column chromatography．Different ratios of di-ameter to height were studiedʒ1ʒ0．5，1ʒ2，1ʒ5，1ʒ16．The ratio of mixed sample to silica gel was1ʒ3（w/w）and the eluent was compod of petroleum ether（30-60ħ）and acetone100ʒ3（v/v）．The eluent flowed at the rate of5．5mL/min．
Ingredient of eluent
The experiment was performed by silica gel（100-200 mesh）column chromatography who diameter to height was1ʒ5．The ratio of mixed sample to silica gel was1ʒ3（w/w）．Different ingredients of eluent were studiedʒpetroleum ether（30-60ħ）/acetone100ʒ2，100ʒ3，100ʒ4，100ʒ5（v/v）．The eluent flowed at the rate of5．5mL/min．
Flow rate
The experiment was performed by silica gel（100-200 mesh）column chromatography who ratio of diameter to height was1ʒ5．The ratio of mixed sample to silica gel was1ʒ3（w/w）and the eluent was compod of petroleum ether（30-60ħ）and acetone100ʒ3（v/ v）．Three different flow rates were studiedʒ5．5，9．0，35mL/min．The low flow rate between5．0and9．0 mL/min was adjusted by the switch on the column chromatography，while the high flow rate of35mL/min was ach
ieved by pressurized column chromatography．Each experiment was performed in triplicate．All the data values were expresd as means with standard de-viation．
Ｒesults and Discussion
Extraction by ethanol-reflux
821Nat ProdＲes Dev Vol.27
Orthogonal array design
The results of each experimental group，yield and purity of extract，and data analysis were shown in Table2．Ac-cording to theＲvalue，the ethanol concentration was found to be the most important factor，the order of the effects of all factors was B＞D＞A＞C．This was also demonstrated by F and P value of ANOVA（Table3）．The ethanol concentration（factor B）and reflux times （factor D）were markedly correlative with yield of isoalantolactone and alantolactone．The optimal method of ethanol-reflux extraction of squiterpene lactones was determined asʒthe ratio of solid to liquid was1ʒ10（w/v，g/mL）with95%ethanol by reflux extraction for2hours，once（A
2
B
3
C
2
D
1会员活动
）．
Table2Experimental results and data analysis
FactorsＲesults（n=2）Yi n（n=1，2）
No．A B C D Yield（%）
个人收条模板X11
Purity（%）
X21
Yield（%）
X12
Purity（%）
X22
Yi1Yi2Yi=∑Yij
1111147．233．5243．773．4428．7628．8357．59 2122253．023．0549．912．9621．2721．1142．39 3133333．734．6930．034．5147．1246．0393．15 4212369．112．9963．632．7515．6914．4130．11 5223141．873．8835．783．6634．7733．9268．69 6231225．335．1822．765．0055．6354．21109．84 7313269．653．0062．872．9315．6516．8432．49 8321344．982．8340．082．7221．0121．1442．15 9332120．036．6118．996．4974．7073．54148．24 K1193．13120．19209．58274．52K=624．66 K2208．64153．23220．74184．73W=28148．96 K3222．89351．24194．34165．41P=21677．75 k132．1920．0334．9345．75
k234．7725．5436．7930．79
k337．1558．5432．3927．57
Ｒj4．9638．514．4018．18
Uj21751．6126882．3821736．2822807．75
Qj73．875204．6358．531130．00
Note：K=∑Yi，W=∑Yi2，P=1/18ˑK2，Ｒj=Max（Ki）-Min（Ki），Uj=1/6∑（Ki2），Qj=Ui-P
Table3Variance analysis of experiments results（ANOVA）
Source Corrected model Intercept Sum of squares
Qj
Degree of freedom
f
Mean
square
F value
Significant
level
A73．875204．6336．943．93
B5204．6322602．32276．91＊＊C58．53229．273．11
D1130．002565．0060．12＊＊Error4．1892．09
城市手抄报Total6471．2117
F0．01（2，9）=8．02；F0．05（2，9）=4．26
Note：Q T=W-P，Qe=Q T-Q A-Q B-Q C-Q D，Fj=（Qj/f j）/（Qe/f e）921
Vol.27ZHANG Le，et al：Extraction and Purification of Alantolactone and Isoalantolcatone fromＲoot of Inula helenium
Verification experiment
Toverify the above the results ，three experiments of eth-anol-reflux extraction were performed using the optimal method ：ratio of solid to liquid was 1ʒ10（w /w ，g /mL ）；ethanol concentration was 95%；extraction time
was 2hours ，
once．The result was shown in Table 4．The extraction yields of 3groups were above 75%and their purities were around 7%．As a result ，the opti-mized method was stable and reliable．
Table 4
Ｒesults of verification experiment
No．Plant material （g ）
Yield （%）Content （%）
Isoalantolactone
Alantolactone
Total lactones
Extraction yield （%）
1653．024．864．792．377．1675．162650．024．104．482．698．1782．383
650．8
24．21
4．94
2．20
7．94
81．09
Purification by silica gel column chromatography Effects of granularity of chromatographic silica gel Three types of granularities of chromatographic silica gel 60-100mesh ，100-200mesh and 200-300mesh were tested for comparing their effects on extraction yield （X 3），purity （X 2）and amount of eluent （X 4），and the results were showed in Fig 1．The experimental
results showed that the increasing of granularity of chromatographic silica gel would result in lower extrac-tion yield ，
in a little higher purity ，and remarkably higher amount of eluent．According to the comparative equation （Y =0．2X 3+0．4X 2-0．4X 4）．The optimal granularity of chromatographic silica gel was 100-200mesh （Fig．1）
．
Fig .1Effect of granularity of chromatographic silica gel on extraction yield （a ），purity （b ），amount of eluent （c ）and com-prehensive score （d ）of isoalantolactone and alantolactone （n =3）
Effects of loading quantity
A granularity of chromatographic silica gel 100-200mesh was ud to test the effect of loading quantity ，by eluent compod of petroleum ether （30-60ħ）and acetone （100ʒ3，v /v ）at a flow rate of 5．5mL /min．The effects of different loading quantities of mixed sam-ple on the extraction yield （X 3），purity （X 2）and a-mount of eluent （X 4）were shown in Fig．2．The results
demonstrated that with the increasing of loading quanti-ty ，the extraction yield would get lower．In addition ，the increasing of loading quantity would result in a little higher purity ，
and remarkably higher amount of eluent．According to the comparative equation （Y =0.2X 3+0．4X 2-0．4X 4），the optimal method was achieved when the ratio of mixed sample to silica gel was 1ʒ3（w /w ，g /mL ）．
031Nat Prod Ｒes Dev Vol.27
