Fingerprint profile of active components for Andrographis
ORIGINAL PAPER
Fingerprint pro?le of active components for Andrographis paniculata Nees by HPLC-DAD
Yue Cui ?Yuzhi Wang ?Xiangyuan Ouyang ?Yubo Han ?Hongbin Zhu ?Qingmei Chen
Received:1December 2008/Accepted:15June 2009/Published online:27June 2009óSpringer Science+Business Media,LLC 2009
Abstract Andrographis paniculata Nees (A.paniculata )has been used as herbal medicine for thousands of years in China.In this work,a quick extraction method with microwave-assisted treatment was studied for a complete extraction of active compounds from A.paniculata.Fur-thermore,the proposed ?ngerprint method,enhanced ?n-gerprint by HPLC-DAD,has the advantage of ef?ciency and accuracy.In comparison with common ?ngerprint at ?xed wavelength,enhanced ?ngerprint compiled additional spectral data and was hence more informative.It could ef?ciently identify,distinguish and assess A.paniculata.So it could be used to conduct the quality control of this tra-ditional Chinese medicine comprehensively.Keywords Food quality áAndrographis paniculata Nees áEnhanced ?ngerprint áHPLC-DAD áMicrowave-assisted extraction
Introduction
Andrographis paniculata Nees (A.paniculata ),one of the main traditional Chinese medicines (TCM),commonly known as ‘‘King of Bitters’’,belongs to the Acanthaceae family.It grows widely in the tropical area of South East Asia,India and China [1].This herb has been used tradi-tionally for several applications such as against fever,
dysentery,in?ammation sore throat and snakebite [2,3],antimicrobial [4],antimalarial [5,6].Furthermore,it is a promising new way for the treatment of several diseases caused by immune disorders such as HIV and AIDS [7].During the last decade,as the use of TCM becomes popular in the world,identi?cation and the quality control of it has attracted more and more attention.But the tradi-tional quality control technique of TCM encounters more and more challenges because one or two markers or pharmacological active compounds in the herbal medicine employed for evaluating the quality and authenticity of herbal medicine cannot give a complete picture of herbal product.Chromatographic ?ngerprint,a more signi?cant formulation for controlling the quality of herbal medicines and their products,has been accepted by many countries and organizations [8,9].This technique emphasizes the systemic characterization of compositions of samples and focuses on the identi?cation and assessment of the stability of components [8,10].
Conventional ?ngerprint contains only chromatographic information.However,in some cases,the limited infor-mation provided by conventional ?ngerprint might not be enough to reveal the quality characteristics of some extremely complex TCM products comprehensively.Therefore,it is necessary to develop a more effective ?n-gerprint.The proposed ?ngerprint,enhanced ?ngerprint,seems promising to ful?ll the https://www.wendangku.net/doc/50317306.html,pared with conventional ?ngerprint,enhanced ?ngerprint with addi-tional spectral information contained,has many more advantages.Firstly,it contains more peaks which can not be detected at a ?xed wavelength but have UV responses at other wavelengths;Secondly,in quality assessment,con-ventional ?ngerprint might give different chromatographic patterns at different wavelengths,whereas enhanced ?n-gerprint technique is more objective and credible [11].
Y.Cui áY.Wang (&)áX.Ouyang áY.Han áH.Zhu áQ.Chen
State Key Laboratory of Chemo/Biosensing and Chemometrics,College of Chemistry and Chemical Engineering,Hunan University,Changsha 410082,People’s Republic of China e-mail:wyzss@https://www.wendangku.net/doc/50317306.html,
Sens.&Instrumen.Food Qual.(2009)3:165–179DOI 10.1007/s11694-009-9082-4
By using the?ngerprint analysis,a particular herb can be identi?ed and distinguished from some closely related species.Due to the variation of the collecting time,sites, processing methods and so on,there are some differences in the content of active compounds in the same plant. Therefore,it is necessary to establish?ngerprint of A.paniculata for its quality control.
Conventionally,the sample preparation of herbal med-icines in?ngerprint studies was accomplished by heating, boiling,re?uxing or ultrasonic extraction[12–14].But all of them are time-consuming,laborious,involving laborious operation and bulk amount of organic solvents[15].The microwave-assisted extraction[16–21]has been exten-sively used for herbal medicine analysis;however,its application in the study on the?ngerprint of A.paniculata has been rarely reported.
Although there were different approaches to the analysis of active compounds in A.paniculata[22–26],to our knowledge,few?ngerprint analysis methodologies have been reported.Zhang et al.[27]performed the?ngerprint analysis of A.paniculata.He used re?uxing to extract active compounds and HPLC-UV at a?xed wavelength to establish?ngerprint.However,as a traditional extraction method,re?uxing could not extract active compounds completely and effectively.The objective of our study is to establish a simple and timesaving extraction method: microwave-assisted extraction and a more effective?n-gerprint which contains additional spectral information: enhanced?ngerprint.
In the present study,different sample preparation pro-cedures were compared,including microwave-assisted and marinated extraction,re?ux and ultrasonic extraction. Microwave-assisted extraction was utilized with HPLC-DAD for identi?cation and?ngerprint establishment of A.paniculata.At the same time principal component analysis,similarity analysis and hierarchical cluster anal-ysis were used to study the factor affecting the quality of A.paniculata.
Experimental
Instrumentation and reagents
HPLC analysis was carried out on a Agilent Series1100 liquid chromatograph,equipped with a vacuum degasser,a quaternary pump,an auto sampler and a diode array detection(DAD)system,connected to a reversed-phase column(Diamonsil C185l m150mm94.6mm i.d., Agilent,USA).Data collection was performed using Chem.Station software(Agilent).The water used for all the solutions and dilutions was prepared with a Millipore ZMQS50F01water puri?cation system(USA).Pressure Self-Control Microwave Decomposition System(Shang-hai,P.R.China)and Auto Science AS-2060B Ultrasonic Cleaner(Tianjin,P.R.China)were used for extraction. Ethanol(95%)is of AR grade and methanol(99.8%)is of HPLC grade(Ludu,P.R.China).
Materials
Raw herbs of A.paniculata were collected directly from various production sites in China.They are presented in https://www.wendangku.net/doc/50317306.html,mercial drugs,purchased in different drug-stores,are presented in Table2.Andrographolide(98%) and dehydroandrographolide(98%)were purchased from Chinese Medicine Research Institute(Nanjing,P.R. China).
Table1Raw herbs of A.paniculata
Sample No.Number
of samples
Location Collecting time Medical parts
of A.paniculata
15Different collecting time Huazhou,Guangdong2007.07.25Stem and leaf 25Huazhou,Guangdong2007.08.20Stem and leaf 35Huazhou,Guangdong2007.09.15Stem and leaf 45Huazhou,Guangdong2007.10.11Stem and leaf 55Different locations Yongkang,Guangdong2007.09.10Stem and leaf 65Yulin,Guangxi2007.08.29Stem and leaf 75Haikang,Hainan2007.09.09Stem and leaf 85Hebei2007.08.26Stem and leaf 95Linquan,Anhui2007.09.13Stem and leaf 105Haozhou,Anhui2007.09.15Stem and leaf 115Different medical parts Haozhou,Anhui2007.09.15Leaf
125Haozhou,Anhui2007.09.15Stem
166Y.Cui et al.
Sample preparation
Before extraction,A.paniculata was ground into powder with particle diameter of0.2–0.9mm by plant grinder. Samples were prepared by four procedures as follows: Marinated extraction
A1g aliquot of sample was added with30mL85%eth-anol and then placed into the sealed vessel,followed by the extraction for30h.The extracts were?ltered and con-centrated to dryness under vacuum.The dried extracts were dissolved in25mL of methanol,?ltered through0.45-l m millipore membrane and an aliquot of10l L of the?ltrate was used for HPLC analysis.
Re?ux extraction
A1g aliquot of sample was added with30mL85%eth-anol and then placed into the re?ux device,followed by the extraction for two times,2h each.The extracts were col-lected together,?ltered and concentrated to dryness under vacuum.Then the remaining steps were done as described in‘‘Marinated extraction’’.
Ultrasonic extraction
A1g aliquot of sample was added with30mL85%eth-anol and then placed into the sealed vessel,followed by ultrasonication for two times,0.5h each.The extracts were collected together,?ltered and concentrated to dryness under vacuum.Then the remaining steps were done as described in‘‘Marinated extraction’’.
Microwave-assisted extraction
A1g aliquot of sample was added with30mL70%eth-anol and then placed into the pressure self-control micro-wave decomposition system.The irradiation time was kept for4min,then the extraction solution was cooled down to ambient temperature.After that,the resulted solution was centrifugal separated,?ltered and concentrated to dryness under vacuum.Then the remaining steps were done as described in‘‘Marinated extraction’’.
Preparation of reference substance solutions
Andrographolide and dehydroandrographolide are the two typical medicinal components of A.paniculata,so both of them were considered as the reference substances in?n-gerprint.Stock solutions of andrographolide and dehydro-andrographolide were prepared by dissolving suitable amount of pure andrographolide and dehydroandrogra-pholide,respectively,in methanol.Reference substance solution was obtained by mixing andrographolide and de-hydroandrographolide stock solutions together.After that, the reference substance solution was directly injected into the HPLC.
HPLC analysis
A reversed-phase C18column(150mm94.6nm,i.d. 5l m)from Agilent was used with the mobile phase con-sisting of methanol(A)and water(B).Gradient elution was performed using the following linear gradient:0min—30%A;5min—45%A;35min—70%A;40min—80% A;45min—80%A.The column compartment was kept at the temperature of25°C.When the wavelengths were below230nm or above300nm,most components have weak or even no responses,so the DAD detection was performed in the range of230–300nm at2nm/step.The sample volume injected was20l L.
Enhanced?ngerprint generating
The information of?ngerprint extracted from DAD data is two-dimensional matrices.Principal Component Analysis (PCA),a chemometrics method,was used for decomposi-tion of two-dimensional matrices.According to the PCA,matrix X(n9m)was set to store the data.The retention time of?ngerprint is set as rows(n)and the wavelength as columns(m).X can be described as m conventional chromatographic?ngerprints at various wavelengths,and each?ngerprint is measured at n regular time intervals.It can be described as follows:
X?
x11x12áááx1m
x21x22áááx2m
.........
x n1x n2áááx n"m
2
66
64
3
77
75?TV0?t1t2...t n
?
v0
1
v0
2
...
v0
m
2
66
64
3
77
75
e1TBy PCA,X can be decomposed into two matrices:the score matrix T and the loading matrix V.According to Eq.1,the score matrix T can be described as:
Table2Commercial drugs
Sample No.Number of samples Drugstore
135Laobaixing drugstore in Anhui
145Kanglian drugstore in Zhengzhou
155Jinsha drugstore in Changsha
165Hongrentang drugstore in Jinan
Fingerprint pro?le of active components for Andrographis paniculata Nees167
T?XVe2TCommonly,an authentic sample(AUS)is required for reference.It is arranged that X0is the DAD matrix of AUS and t i0is the i th principal component(PC)of the score matrix.T i0can be described as follows:
t0i?x0v0ie3TTherefore the?rst principal component(PC1)can be described as follows:
t01?x0v01e4TConsidering that PC1represents the maximal variation of the data matrix,we simply used PC1to describe the systematic information as follows:
t01?x0v01?x01v11tx02v21tááátx0m v m1e5TIn Eq.5,x10–x m0is chromatographic?ngerprint detected at various wavelengths and the loading vector v11–v m1is the weight of each?ngerprint.Obviously,t10transforms the two dimensional information into one dimensional linear rela-tionship and it contains both the chromatogram and spec-trum information.
In PCA,singular value decomposition of original matrix,an algebra method,was applied to calculate the Covariance Matrix as follows:
S?
k10ááá0
0k2ááá0
.........
00ááák n
2
66
64
3
77
75e6T
k1–k n is the variance of every PC.With the assistance of
MATLAB7.0,the variance percentage of PC1of A.pan-
iculata is calculated.The value is98.87%,which means
PC1of A.paniculata can accounts for98.87%variance in
the original data and the most information of DAD data is
successfully extracted.Thus using only PC1to explain the
original matrix is suf?cient.
Because of the quality of PC1and AUS,v10derived from
AUS describes the common spectral characteristics for
samples on the same species.Therefore,according to
Eq.5,the?ngerprint of one sample is as follows
t1?xv01e7T
With the assistance of MATLAB7.0and professional
similarity evaluation system named‘‘Similarity Evaluation
System for Chromatographic Fingerprint of TCM’’,the
enhanced?ngerprint of A.paniculata was established
which is shown in Fig.1.
Data analysis
Data analysis was performed using the professional soft-
ware named‘‘Similarity Evaluation System for Chro-
matographic Fingerprint of TCM’’.It is mainly applied to
the similarity analysis of chromatographic patterns.The
software about the calculation of similarity was mainly
based on the peak area and retention time.Firstly,the
chromatogram data should be imported into the software.
Then the marker peaks were chosen by the user and all
peaks would be matched.Subsequently,the
mean
168Y.Cui et al.
chromatogram was produced,and the similarity value of all introduced chromatograms relative to that of mean chro-matogram would be calculated [15].The correlation coef-?cient is expressed by the following formula:
r ?P n
i ?1x i x 0
i
?????????????????????????????????P n i ?1x 2i P n i ?1x 02
i q e8Twhere r is the correlation coef?cient between ?ngerprint x
and the reference ?ngerprint x 0.x i and x 0i represent the i th elements of the i th peak of each ?ngerprint,which can be the response value of signal,area or height of the peak,etc.The value r of the correlation coef?cient is in the range 0\r B 1.The larger the value of r is,the higher the quality of target samples is.It is considered as the best,the better or the worst drug with a corresponding correlation coef?cient above 0.9,between 0.8and 0.9,or below 0.8,respectively.When r equals 1,it is identical.Results and discussion Selection of extraction method
Different extraction methods were compared (shown in Fig.2)to select an optimum sample preparation.From Fig.2,we can see the HPLC ?ngerprint marked as ‘‘D’’has more peaks than other ?ngerprints,which implies micro-wave-assisted extraction is a more effective method.At the same time it is timesaving.So it is more suitable for complete extraction of main components.Preliminary test of microwave-assisted extraction conditions
In microwave-assisted extraction,there are many factors affecting the extraction yield,such as the extraction time,solid–liquid ratio and the concentration of extraction
solvent.An orthogonal test design in the extraction mode was used to optimize the extraction conditions.The levels and factors of the orthogonal test are shown in Table 3and the results are presented in Table 4.From Table 4,we can see that the maximum extraction yield of the extract was 1.53%.However,we cannot select the best extraction conditions only based on these outcomes in Table 4.Therefore a further orthogonal analysis was warranted.Thus,the K and R values were calculated and listed in Table 4.K value is the average extracting ef?ciency of every factor at each level.R value is the range of K value.As seen in Table 4,according to the R values,we can ?nd that the in?uence to the mean extraction yields of the compounds decreases in the order:factor B [factor C [factor A.We also can see that K 2j has the highest K value in factor A,which means that compared with other levels of factor A,level two has the highest extracting ef?ciency,and is the most suitable level for extraction.Based on the same principle,level four of factor B and level two of factor C were chosen for extraction.In other words,the maximum yield of the active compound was obtained at the following conditions:30mL solvent per gram of sample,70%alcohol solution,4min of irradiation time,respectively.
Fingerprints of A.paniculata
The establishment of the reference ?ngerprint
In order to identify the main active components of A.paniculata,the HPLC pro?le of the standard compounds is shown in Fig.3.Andrographolide and dehydroandrog-rapholide could be identi?ed by comparison of their retention time in Fig.3and the HPLC chromatography of the sample.Their structures are as follows:(a):Androg-rapholide,(b):Dehydroandrographolide.
Fingerprint pro?le of active components for Andrographis paniculata Nees 169
170Y.Cui et al.
The sample,from Huazhou,P.R.China (No.4),extracted by microwave-assisted extraction,was detected by HPLC-DAD with the wavelength in the range of 230–300nm.It is worthy to demonstrate that sample No.4was the raw material like others and is reliable to show the common character of all samples .Therefore it is feasible to pick No.4to show the HPLC pro?le of samples.The chromatography is presented in Fig.4.
For the establishment of the reference ?ngerprint,all the raw materials of A.paniculata,presented in Table 1,were analyzed by HPLC.Then the HPLC pro?les were inputted into the similarity evaluation system software.The average ?ngerprint,that was computed by this software,was con-sidered as the reference ?ngerprint of samples which is shown in Fig.5.Based on the retention time,seven com-mon peaks were determined and peaks 4,5were identi?ed as andrographolide and dehydroandrographolide,respec-tively,by comparing their retention time in Fig.3.
Table 3Factors and levels of orthogonal experiment Level No.Factor A:ratio of material to solvent (g/mL)Factor B:time (min)Factor C:ethanol concentration (mL:mL)11:2016:421:3027:331:4038:24
1:50
4
9:1
Table 4Arrangement and results of L9(34)orthogonal experiment Test No.Factor A Factor B Factor C Extracting ef?ciency(%)11:2016:40.99721:2027:3 1.53031:2038:20.52241:2049:1 1.29551:3017:3 1.35161:3026:4 1.15571:3039:10.73481:3048:2 1.27591:4018:20.795101:4029:10.651111:4036:40.753121:4047:3 1.195131:5019:1 1.135141:5028:20.781151:5037:30.978161:5046:4 1.448
K 1j 1.086 1.070 1.088K 2j 1.129 1.029 1.264K 3j 0.8490.7470.843K 4j 1.086 1.3030.954R
0.280
0.556
0.421
Fig.3The HPLC pro?le of andrographolide and dehydroandrogra-pholide.(A)Andrographolide,(B)
dehydroandrographolide
Fig.4The HPLC pro?le of the sample (No.4)extracted by microwave-assisted
extraction
Fingerprint pro?le of active components for Andrographis paniculata Nees 171
The reference?ngerprint is in practice a chromato-graphic pattern of some common kinds of pharmacologi-cally active and chemically characteristic components in A. paniculata.This chromatographic pro?le should feature the fundamental attributions of‘‘integrity’’,in other words,‘‘sameness’’.It could demonstrate the‘‘sameness’’between various samples successfully.If the HPLC pro?le of one sample has the common peaks in the reference?ngerprint and the retention time of the peaks are consistent with the ones in the reference?ngerprint,the sample could be identi?ed as A.paniculata.Meanwhile,in similarity anal-ysis,the reference?ngerprint is considered as the standard. Because the reference?ngerprint can show the common character of A.paniculata,the sample could be considered as an authentic one if the similarity value of it is high. The?ngerprints of A.paniculata from different sources In order to indicate the differences from various treatments,?ve batches of samples which were listed in Tables1and2 were analyzed statistically.RSD values of seven common peaks(shown in Fig.5)in HPLC chromatograms among ?ve batches of samples were shown in Table5.From Table5,we can see RSD values were all less than1.15%for retention time and less than4.18%for peak area,which means the common peaks were in good correspondence in
Table5RSD value of seven common peaks in HPLC chromatograms among?ve batches of samples
Sample No.No.of samples Peak1a Peak2a Peak3a Peak4a Peak5a Peak6a Peak7a
15Area b 3.81 4.14 4.11 3.97 3.61 3.78 4.15
RT c0.830.790.860.740.760.810.91 25Area b 2.89 3.11 3.14 3.12 3.17 2.89 2.76
RT c0.98 1.140.940.91 1.05 1.12 1.07 35Area b 4.13 3.91 3.87 3.97 3.76 3.88 3.91
RT c0.810.830.780.910.880.910.83 45Area b 3.76 3.91 3.59 3.48 3.85 3.98 4.13
RT c0.860.910.910.830.880.860.92 55Area b 3.72 3.81 3.59 3.91 3.73 3.59 3.86
RT c0.560.510.610.630.690.590.63 65Area b 3.79 3.95 3.86 4.17 3.71 3.87 3.48
RT c0.850.940.910.930.860.89 1.01 75Area b 3.76 3.49 3.82 3.95 3.83 3.69 3.91
RT c0.890.860.92 1.140.970.820.78 85Area b 2.86 2.49 3.12 2.81 3.24 3.37 3.43
RT c0.480.670.620.590.390.540.61 95Area b 2.61 2.37 2.64 3.02 2.71 2.17 2.21
RT c0.970.860.880.950.910.790.81 105Area b 1.72 1.76 1.89 2.07 1.81 2.12 1.91
RT c0.870.780.810.790.690.710.69 115Area b 2.71 2.48 1.97 2.38 3.07 2.94 2.67
RT c0.570.670.610.630.590.710.73 125Area b 2.83 2.72 2.91 3.21 3.18 2.93 3.11
RT c0.380.890.710.960.790.850.92 135Area b 3.81 3.76 3.91 3.71 3.69 3.19 3.41
RT c0.620.590.670.640.610.570.54 145Area b 3.04 3.19 3.31 3.46 3.41 3.47 3.51
RT c0.790.690.730.640.680.750.72 155Area b 3.74 3.75 3.69 3.81 3.75 3.57 3.82
RT c0.510.490.570.390.470.380.42 165Area b 3.84 3.76 3.75 3.67 3.73 3.81 3.71
RT c0.590.510.490.460.480.370.53
a Peaks1–7are the peaks marked in Fig.5
b Area peak area(mAu min)
c RT retention time(min)
172Y.Cui et al.
?ve batches of every sample.Therefore the average HPLC chromatogram of ?ve batches which was generated by MATLAB 7.0can show the sameness of every sample.Meanwhile average HPLC chromatograms from different samples can show the differences among various treatments.On the basis of the demonstrations above,we use the average HPLC chromatogram among ?ve batches of every sample to do the HPLC ?ngerprints and the clustering analyses in the following study.
All the samples (showed in Table 1)were evaluated by their similarities,which came from the calculation on the correlation coef?cient or cosine value of vectorial angle of original data.The results are shown in Fig.6and Table 6.The bigger the similarity value is,the better the approach
degree between the sample and reference.Peaks (A)–(G)in Fig.6are the seven peaks (marked as peaks 1–7)showed in Fig.5.And the similarity values in Table 5are all above 0.8.So they could be identi?ed as A.paniculata.The ?ngerprints of A.paniculata collected in different time
Samples (Nos.1–4showed in Table 1)were collected in July,August,September and October at the same location (Huazhou,P.R.China).The ?ngerprints are presented in Fig.7.The results show that absorption intensity of some peaks is different,especially peak (A)(andrographolide)and peak (B)(dehydroandrographolide).Obviously,
peaks
Fig.6Enhanced HPLC ?ngerprints of A.paniculata (Nos.1–12).Peaks (A)–(G)are the seven common peaks showed in Fig.5(peaks 1–7)
Table 6Results of similarity analysis
a
The value is mean ±standard deviation
Sample No.Number of samples Correlation coef?cient a Cosine value
of vectorial angel a 150.9632±0.0170.9732±0.018250.9515±0.0290.9582±0.029350.9749±0.0170.9791±0.018450.9657±0.0310.9737±0.031550.9473±0.0260.9617±0.026650.9771±0.0280.9816±0.029750.9841±0.0320.9476±0.032850.8759±0.0290.8789±0.029950.8680±0.0360.8723±0.0371050.8737±0.0160.8869±0.0161150.8719±0.0170.8859±0.01712
5
0.8699±0.021
0.8723±0.021
Fingerprint pro?le of active components for Andrographis paniculata Nees 173
of Nos.2and 3have greater absorption intensity which implies that the concentration of principal components in Nos.2and 3is higher than that in other samples.So it is favorable to collect A.paniculata in August and September.The ?ngerprints of different medicinal parts of A.paniculata
It has been well known that the different medicinal parts of raw herbs usually are used for different curative purpose.It is therefore important to learn about any difference among various medicinal parts of A.paniculata.Samples (Nos.10–12showed in Table 1)with different effective parts of the plant were analyzed with the software and are presented in Fig.8.Although the chromatograms of the Sample Nos.10–12have the similar peaks,peaks of No.11have greater absorption intensity which means that the concentration of principal effective compounds in No.11is higher than that in other samples.Therefore the leaf was better for the medicinal part of A.paniculata than others
.
Fig.7Enhanced HPLC ?ngerprints of A.paniculata (Nos.1–4).(A)
Andrographolide,(B)
dehydroandrographolide
Fig.8Enhanced HPLC ?ngerprints of A.paniculata (Nos.10–12).(A)Andrographolide,(B)dehydroandrographolide
174Y.Cui et al.
The ?ngerprints of A.paniculata collected in different locations
Samples from various locations (No.3and Nos.5–10,showed in Table 1)were compared with the software and are presented in Fig.6and Table 5.Based on the similarity values,it is interesting to note that all the samples fell into two groups:Group A and Group B.The similarities of Groups A,B are 0.947–0.984and 0.868–0.879,respec-tively.Group A consisted of No.3and Nos.5–7was collected from south of China.Group B consisted of Nos.8–10was collected from north of China.The important ?nding is that the similarities of herbs are very relative to their collecting locations (Tables 1,5).Moreover,the effects brought from collecting locations are more visible than those from collecting time and medicinal parts of A.paniculata .So in order to control the stability of A.paniculata ,the collecting location should be ?xed.As a result,in order to control the quality of A.pan-iculata,attentions should be paid to collecting time,loca-tion and the medicinal parts.Especially,location is the main factor affecting the quality of A.paniculata.The clustering analyses of samples from different sources
Hierarchical cluster analysis based on peak characteristics from the average ?ngerprint pro?les among ?ve batches of every sample presented in Table 1was applied,and the Euclidean distance,which is a pattern similarity measure,was selected as measurement for hierarchical cluster analysis.The cluster analysis was operated with the assis-tance of MATLAB 7.0and the result is shown in Fig.9.In cluster analysis,samples can be classi?ed as one category
because of the short Euclidean distance,which means they are more similar to each other than to others.In Fig.9,comparing sample Nos.1–4,we can see that Nos.2and 3can be classi?ed as one category for the higher similarity.According to Table 1,Nos.2and 3were collected in August and September,therefore it is favorable to collect A.paniculata in August and September.
Meanwhile,using this method,we are able to classify all the samples into two broad categories,samples 1–7col-lected from south of China and samples 8–12collected from north of China.The result proves that in order to control the stability of A.paniculata ,the collecting location should be ?xed.
The ?ngerprints of A.paniculata purchased in different commercial drugs
Commercial drugs (Nos.13–16,showed in Table 2)pur-chased from different drugstores were analyzed.The results are presented in Fig.10and Table 7.The results in Fig.10indicated that they all have the seven common peaks showed in the reference ?ngerprint in Fig.5(peaks (A)–(G)in Fig.10corresponding to peaks 1–7in Fig.5).So they could be identi?ed as A.paniculata.However,the similarities of Nos.15and 16in Table 7are not very high,they are both below 0.80.If 0.80is set as an appropriate threshold,it is easy to ?nd that sample Nos.15and 16are unacceptable.Meanwhile,hierarchical cluster analysis was applied.The result is presented in Fig.11.Seen from Fig.11,Nos.13and 14are clustered together with the reference ?ngerprint ?rstly,which suggests that the quality of Nos.13and 14are more similar to the reference.Sample Nos.15and 16are different from other samples and the reference,which indicates they are unacceptable
samples
Fig.9The clustering analyses of chromatograms of samples from different sources (the samples are presented in Table 1)
Fingerprint pro?le of active components for Andrographis paniculata Nees 175
for quality control.As a result,the conclusion of similarity analysis and hierarchical cluster analysis are https://www.wendangku.net/doc/50317306.html,parison of enhanced ?ngerprint and conventional chromatographic ?ngerprint
The conventional chromatographic ?ngerprints of A.pan-iculata were obtained at wavelengths of 235or 250nm (shown in Fig.12)which were the characteristic UV
absorption wavelengths of the two typical medicinal components respectively (i.e.andrographolide and dehy-droandrographolide).From Fig.12,we can see that com-pared with conventional ?ngerprints,enhanced ?ngerprint contains more peaks,especially in the retention time of 15–25.Moreover,the signals in enhanced ?ngerprint are much stronger than those in conventional ?ngerprints.Meanwhile,the similarities of conventional ?ngerprint (250nm)and enhanced ?ngerprint are compared in Table 8.From Table 8,it can be easily seen that using enhanced ?ngerprint,all the samples are classi?ed into two groups with different quality grades.Numbers 1–7are classi?ed into Group A with the similarities of 0.9171–0.9749,and Group B is consists of Nos.8–12,whose similarity values are in the range of 0.8680–0.8759.Whereas all the samples from different sources could not be distinguished by conventional ?ngerprint.Thus,com-pared with enhanced ?ngerprint,the classi?cation ability and discrimination power of conventional chromatographic ?ngerprint is limited to perform further quality
control.
Fig.10Enhanced HPLC ?ngerprints of commercial drugs (Nos.13–16).Peaks
(A)–(G)are the seven common peaks showed in Fig.5(peaks 1–7)
Table 7Similarity values of commercial drugs purchased in differ-ent drugstores Sample No.Number of samples Correlation coef?cient a Cosine value of vectorial angel a 1350.9578±0.0120.9590±0.0131450.8701±0.0210.8825±0.0211550.6854±0.0270.6965±0.02916
5
0.7121±0.041
0.7201±0.042
a
The value is mean ±standard
deviation
Fig.11The clustering analyses of chromatograms of different commercial drugs (the samples are presented in Table 2).Re the reference ?ngerprint
176
Y.Cui et al.
Methodology study
Precision test of the apparatus
Apparatus precision was evaluated by the analysis of ?ve successive injections of the same sample solution.The correlation coef?cients between the reference HPLC ?n-gerprint and the sample HPLC ?ngerprint were:0.9973,0.9959,0.9984,0.9982and 0.9963,respectively.These data showed that the precision of the apparatus was satisfactory.
Stability test of the solution
Stability was evaluated by the analysis of ?ve injections of the same sample solution every 4h.The correlation coef-?cients were 0.9987,0.9991,0.9972,0.9978and 0.9983,
respectively.The data indicated that the stability of the solution was very nice.
Repeatability of analytical method
Repeatability was evaluated by the analysis of ?ve injec-tions of ?ve sample solutions prepared by the same method.The correlation coef?cients were 0.9979,0.9992,0.9995,0.9987and 0.9988,respectively,showing that the repeatability of the analytical method is very suitable.
Discussion
In the present study,an impersonal,valid and rapid extraction method and an ef?ciency ?ngerprint analysis method was developed and https://www.wendangku.net/doc/50317306.html,pared
with
Fig.12Comparison of
conventional HPLC ?ngerprint and enhanced HPLC ?ngerprint.(1)Andrographolide (2)dehydroandrographolide.(A)Enhanced HPLC ?ngerprint,(B)conventional HPLC ?ngerprint at 250nm,(C)conventional HPLC ?ngerprint at 235nm
Table 8Similarity values of A.paniculata
a
The value is mean ±standard deviation
Sample No.Number of samples Enhanced ?ngerprint a Conventional
?ngerprint (250nm)a 150.9632±0.0170.8817±0.011250.9515±0.0290.8707±0.012350.9749±0.0170.8924±0.021450.9657±0.0310.8768±0.032550.9473±0.0260.8307±0.027650.9771±0.0280.8473±0.031750.9841±0.0320.8776±0.038850.8759±0.0290.8687±0.011950.8680±0.0360.8883±0.0341050.8737±0.0160.8797±0.0181150.8719±0.0170.8679±0.02112
5
0.8699±0.021
0.8934±0.023
Fingerprint pro?le of active components for Andrographis paniculata Nees 177
traditional extraction,microwave-assisted extraction was suitable for the sample preparation of A.paniculata.It has the characters of speediness and no-pollution.Meanwhile, samples from different sources were analyzed.According to their similarities and hierarchical cluster analysis,those herbs were assorted into two groups.The taxonomy based on similarities had a fair consistency with the reference ?ngerprint in Fig.5.In a word,our results have demon-strated that the enhanced?ngerprint with the similarity analysis may be accepted in general quality standards of herbal medicines in the recent future.The important?nd-ing is that the most relevant factor on the quality of A.paniculata was the collecting location then the har-vesting time and the medicinal parts.In order to get the consistent raw materials of A.paniculata,the collecting location should be?xed and then the harvest time.At last, commercial drugs purchased from different drugstores were analyzed to show the feasibility and applicability of the extraction and analysis method.As a result,two com-mercial drugs were identi?ed as unquali?ed ones for their poor similarity.
Conclusion
In summary,the microwave-assisted extraction and enhanced?ngerprint can be established for the quality control of A.paniculata.This method is very simple with high precision,stability and repeatability.The HPLC enhanced?ngerprint,with additional spectral information compiled,can comprehensively and properly reveal the quality characteristics of TCM,and will become a very useful complementary technique for Chinese herbal quality control and a powerful support for the progress of Chinese herbal medical prescription.
Acknowledgements This work was?nancially supported by the National Natural Science Foundation of China(Nos.20875025, J0830415),the Hunan Provincial Natural Science Foundation (No.07JJ3018).
References
1.A.C.Kumoro,M.Hasan,Supercritical carbon dioxide extraction
of Andrographolide from Andrographis paniculata:effect of the solvent?ow rate,pressure and temperature.Chin.J.Chem.Eng 15,877–883(2007)
2.R.Wongkittipong,L.Prat,S.Damronglerd,C.Gourdon,Solid–
liquid extraction of andrographolide from plants—experimental study,kinetic reaction and model.Sep.Purif.Technol.40,147–154(2004)
3.W.Martz,Plants with a reputation with snakebite.Toxicon30,
1131–1142(1992)
4.P.K.Singha,S.Roy,S.Dey,Antimicrobial activity of Androg-
raphis paniculata.Fitoterapia74,692–694(2003)
5.N.N.N.A.Rahman,T.Furuta,S.Kojima,K.Takane,M.A.Mohd,
Antimalarial activity of extracts of Malaysian medicinal plants.
J.Ethnopharmacol.64,249–254(1999)
6.V.K.Dua,V.P.Ojha,R.Roy,B.C.Joshi,N.Valecha,C.U.Devi,
M.C.Bhatnagar,V.P.Sharma,S.K.Subbarao,Anti-malarial activity of some xanthones isolated from the roots of Androgra-phis paniculata.J.Ethnopharmacol.95,247–251(2004)
7.R.A.Kumar,K.Sridevi,N.V.Kumar,S.Nanduri,S.Rajagopal,
Anticancer and immunostimulatory compounds from Androgra-phis paniculata.J.Ethnopharmacol.92,291–295(2004)
8.S.Wang,H.-Q.Ma,Y.-J.Sun,C.-D.Qiao,S.-J.Shao,S.-X.
Jiang,Fingerprint quality control of Angelica sinensis(Oliv.) Diels by high-performance liquid chromatography coupled with discriminant analysis.Talanta72,434–436(2007)
9.W.Li,Y.Deng,R.Dai,Y.Yu,M.K.Saeed,L.Li,W.Meng,X.
Zhang,Chromatographic?ngerprint analysis of Cephalotaxus sinensis from various sources by high-performance liquid chro-matography—diodearray detection—electrospray ionization-tan-dem mass spectrometry.J.Pharm.Biomed.Anal.45,38–46 (2007)
10.P.Drasˇar,J.Moravcova,Recent advances in analysis of Chinese
medical plants and traditional medicines.J.Chromatogr.B812, 3–21(2004)
11.S.-K.Yan,W.-F.Xin,G.-A.Luo,Y.-M.Wang,Y.-Y.Cheng,An
approach to develop two-dimensional?ngerprint for the quality control of Qingkailing injection by high-performance liquid chromatography with diode array detection.J.Chromatogr.A 1090,90–97(2005)
12.M.Palma, C.G.Barroso,Ultrasound-assisted extraction and
determination of tartaric and malic acids from grapes and wine-making by-products.Anal.Chim.Acta458,119–130(2002) 13.L.Tong,Y.Wang,J.Xiong,Y.Cui,Y.Zhou,L.Yi,Selection
and?ngerprints of the control substances for plant drug Eu-commia ulmodies Oliver by HPLC and LC–MS.Talanta76,80–84(2008)
14.B.Zhang,R.Yang,C.-Z.Liu,Microwave-assisted extraction of
chlorogenic acid from?ower buds of Lonicera japonica Thunb.
Sep.Purif.Technol.62,480–483(2008)
15.J.Kang,L.Zhou,J.Sun,J.Han,D.-A.Guo,Chromatographic
?ngerprint analysis and characterization of furocoumarins in the roots of Angelica dahurica by HPLC/DAD/ESI-MSn technique.
J.Pharm.Biomed.Anal.47,778–785(2008)
16.A.V.Filgueiras,J.L.Capelo,https://www.wendangku.net/doc/50317306.html,villa,C.Bindicho,Talanta53,
433–436(2000)
17.Z.K.Guo,Q.H.Jin,G.Q.Fan,Y.P.Duan,C.Qin,M.J.Wen,
Anal.Chim.Acta436,41–46(2001)
18.E.D.Conte, C.Y.Shen,P.W.Perschbacher, https://www.wendangku.net/doc/50317306.html,ler,J.
Agric.Food.Chem.44,829–834(1996)
19.B.Kaufmann,P.Christen,J.L.Venuthey,Phytochem.Anal.12,
327–330(2001)
20.N.Hong,V.A.Yaylayan,G.S.Raghavan,J.R.Park,J.M.Blan-
ger,Nat.Prod.LETT15,197–200(2001)
21.W.H.Ho,S.J.Hsieh,Anal.Chim.Acta428,111–116(2001)
22.L.Chen,A.Yu,X.Zhuang,K.Zhang,X.Wang,L.Ding,H.
Zhang,Determination of andrographolide and dehydroandrogra-pholide in rabbit plasma by on-line solid phase extraction of high-performance liquid chromatography.Talanta74,146–152(2007) 23.L.Chen,H.Jin,L.Ding,H.Zhang et al.,On-line coupling of
dynamic microwave-assisted extraction with high-performance liquid chromatography for determination of andrographolide and dehydroandrographolide in Andrographis paniculata Nees.J.
Chromatogr.A1140,71–77(2007)
24.Q.Dua,G.Jerz,P.Winterhalter,Separation of andrographolide
and neoandrographolide from the leaves of Andrographis pan-iculata using high-speed counter-current chromatography.J.
Chromatogr.A984,147–151(2003)
178Y.Cui et al.
25.M.K.Bhan,A.K.Dhar,S.Khan,https://www.wendangku.net/doc/50317306.html,ttoo,K.K.Gupta,D.K.
Choudhary,Screening and optimization of Andrographis pan-iculata(Burm.f.)Nees for total andrographolide content,yield and its components.Sci.Hortic.107,386–391(2006)
26.K.Muntha Reddy,V.B.Mopuru Reddy, D.Gunasekar,M.
Madugula Murthy,C.Caux,B.Bodo,A?avone and an unusual
23-carbon terpenoid from Andrographis paniculata.Phyto-chemistry62,1271–1275(2003)
27.Z.Zhang,H.Dong,J.Yu,The?ngerprints of Andrographis
paniculata by HPLC/UV/MS.Chin.J.Nat.Med3,373–377 (2005).(In Chinese)
Fingerprint pro?le of active components for Andrographis paniculata Nees179