Biological and Pharmaceutical Bulletin
Online ISSN : 1347-5215
Print ISSN : 0918-6158
ISSN-L : 0918-6158
Regular Articles
Comparative Analysis of DNA Barcoding and HPLC Fingerprint to Trace Species of Phellodendri Cortex, an Important Traditional Chinese Medicine from Multiple Sources
Zhipeng ZhangYang ZhangZhao Zhang Hui YaoHaitao LiuBen’gang ZhangYonghong Liao
Author information
JOURNAL FREE ACCESS FULL-TEXT HTML

2016 Volume 39 Issue 8 Pages 1325-1330

Details
Abstract

Phellodendri Cortex is derived from the dried barks of Phellodendron genus species, has been extensively used in traditional Chinese medicine. The cortex is divided into two odorless crude drugs Guanhuangbo and Huangbo. Historically, it has been difficult to distinguish their identities due to a lack of identification methods. This study was executed to confirm the identity and to ensure the species traceability of Phellodendri Cortex. In the current study, analysis is based on the internal transcribed spacer (ITS) and psbA–trnH intergenic spacer (psbA–trnH) barcodes and HPLC fingerprint was carried out to guarantee the species traceability of Guanhuangbo and Huangbo. DNA barcoding data successfully identified the three plants of the Phellodendron genus species by ITS+psbA–trnH, with the ability to distinguish the species origin of Huangbo. Moreover, the psbA–trnH data distinguished Guanhuangbo and Huangbo except to trace species. The HPLC fingerprint data showed that Guanhuangbo was clearly different from Huangbo, but there was no difference between the two origins of Huangbo. Additionally, the result of hierarchical clustering analysis, based on chlorogenic acid, phellodendrine, magnoflorine, jatrorrhizine, palmatine and berberine, was consistent with the HPLC fingerprint analysis. These results show that DNA barcoding and HPLC fingerprint can discriminate Guanhuangbo and Huangbo. However, DNA barcoding is more powerful than HPLC fingerprint for species traceability in the identification of related species that are genetically similar. DNA barcoding is a useful scientific tool to accurately confirm the identities of medicinal materials from multiple sources.

Phellodendri Cortex is derived from the dried barks of Phellodendron amurense RUPR., Phellodendron chinense SCHNEID. and the variant Phellodendron chinense var. glabriusculum SCHNEID. (Rutaceae).1) It is commonly used in traditional oriental medicine, and it has demonstrated a wide range of biological and pharmacological activities—antimicrobial, anti-inflammatory and anticancer—in numerous studies.2,3) In the Chinese Pharmacopoeia (2000 version), P. amurense and P. chinense are listed as the original plants of Phellodendri Cortex.4) However, P. amurense is listed as an endangered species and is strictly protected by law because this resource has decreased dramatically.5) Additionally, there are differences between P. amurense and P. chinense regarding plant morphology and chemical composition. Thus, to control the quality of the herb and to protect the wild resource of P. amurense, Phellodendri Cortex is officially listed in the Chinese Pharmacopoeia (2005 version) as two different crude drugs, Guanhuangbo and Huangbo.6) Guanhuangbo is derived from the P. amurense bark, while Huangbo is derived from the P. chinense bark. Since the wild resource of P. chinense is insufficient, the cultivated herb of P. chinense var. glabriusculum is also utilized in the production of crude drugs. However, the three cortexes of P. amurense, P. chinense and P. chinense var. glabriusculum are odorless, and there has been difficulty in differentiating their identities due to a lack of identification methods.

In the Chinese Pharmacopoeia, the quality control of Phellodendri Cortex includes a confirmation of identity and an evaluation of the content of chemical markers. The identity of Guanhuangbo is confirmed by the presence of obacunone using the thin-layer chromatography (TLC) method, and its quantity is determined by measuring the content of berberine chloride and palmatine chloride using the HPLC method; Huangbo is identified by the presence of phellodendrine, and its quantity is determined by measuring the content of berberine chloride and phellodendrine chloride.7) However, the two crude herbs contain both obacunone and phellodendrine, making the TLC method insufficient. Traditional methods for quality control of Phellodendri Cortex include TLC, high performance capillary electrophoresis (HPCE), HPLC and random amplified polymorphic DNA (RAPD).811) These previous studies have successfully determined the content of chlorogenic acid, phellodendrine, magnoflorine, jatrorrhizine, palmatine and berberine for quality evaluation of Guanhuangbo using the HPLC method, and they have confirmed the polymorphisms of Guanhuangbo and Huangbo in 12 primers using the RAPD method.11,12) However, these methods are still inadequate for distinguishing the identity and for tracing the species of Phellodendri Cortex.

DNA barcoding is an emerging technology for modern species identification and has drawn increasing attention.13) In 2009, the Consortium for the Barcode of Life recommended the ribulose bisphosphate carboxylase large chain (rbcL) and maturase K (matK) genes as core DNA barcoding of the seed plant and the psbA–trnH intergenic spacer (psbA–trnH) and internal transcribed spacer (ITS) genes as supplementary DNA barcoding.14) In 2011, the Chinese Plant Barcode of Life Group proposed ITS/ITS2 as the core DNA barcoding for the seed plant.15) DNA barcoding has since been successfully applied towards the traceability of many herb medicines.1622) Additional candidate DNA barcoding genes in seed plants, which are used for identification of Chinese medicines and to confirm the identities of the original plants, are currently providing a basis for identifying Phellodendri Cortex.

For the quality control of crude drugs and the protection of the endangered species in the wild, it is highly desirable to confirm the identities of crude drugs and to ensure the species traceability of Phellodendri Cortex. In this study, we collected 32 samples including P. amurense, P. chinense and P. chinense var. glabriusculum. from 15 areas. We performed experiments to differentiate the crude herbs using ITS-PCR and psbA–trnH-PCR and to quantify 6 chemical markers using the HPLC/photodiode array detector (PAD) method, with the intent to establish a method that enables the species traceability of Phellodendri Cortex.

MATERIALS AND METHODS

Sampling of Plant Materials

Thirty-two samples were collected from the Liaoning, Jilin, Heilongjiang, Hebei, Beijing, Sichuan, Chongqing and Hubei provinces in China to ensure a sufficient representation of the herbs. Among the samples, thirteen were P. amurense (11 cortex and 2 leaf samples), six were P. chinense (5 cortex and 1 leaf samples) and nine were P. chinense var. glabriusculum (6 cortex and 3 leaf samples). Additionally, 4 leaf samples of Tetradium ruticarpum with a closer genetic relationship to the Phellodendron genus were collected as the outgroup for the molecular analysis (Table 1). The botanical identities of the stems were confirmed by Professor Zhao Zhang. All corresponding voucher samples were deposited in the Herbarium of the Institute of Medicinal Plant Development.

Table 1. Plant Materials Examined in This Study
CodesSamplesScientific nameFamilyLocalitiesGenBank accession No.
ITSpsbA–trnH
S1CortexP. amurenseRutaceaeLiaoning, ChinaKT961044KT961071
S2CortexP. amurenseRutaceaeJilin, ChinaKT961045KT961073
S3CortexP. amurenseRutaceaeJilin, ChinaKT961039KT961072
S4CortexP. amurenseRutaceaeJilin, ChinaKT961046KT961074
S5CortexP. amurenseRutaceaeJilin, ChinaKT961047KT961075
S6CortexP. amurenseRutaceaeHeilongjiang, ChinaKT961048KT961076
S7CortexP. amurenseRutaceaeHeilongjiang, ChinaKT961049KT961077
S8CortexP. amurenseRutaceaeHeilongjiang, ChinaKT961050KT961078
S9CortexP. amurenseRutaceaeHeilongjiang, ChinaKT961051KT961079
S10CortexP. amurenseRutaceaeBeijing, ChinaKT961054KT961082
S11CortexP. amurenseRutaceaeBeijing, ChinaKT961055KT961083
L1LeafP. amurenseRutaceaeHebei, ChinaKT961052KT961080
L2LeafP. amurenseRutaceaeHebei, ChinaKT961053KT961081
S12CortexP. chinenseRutaceaeChongqing, ChinaKT961061KT961089
S13CortexP. chinenseRutaceaeChongqing, ChinaKT961062KT961088
S14CortexP. chinenseRutaceaeHubei, ChinaKT961058KT961085
S15CortexP. chinenseRutaceaeHubei, ChinaKT961059KT961087
S16CortexP. chinenseRutaceaeChongqing, ChinaKT961060KT961086
L3LeafP. chinenseRutaceaeBeijing, ChinaKT961056KT961084
S17CortexP. chinense var. glabriusculumRutaceaeSichuan, ChinaKT961063KT961093
S18CortexP. chinense var. glabriusculumRutaceaeSichuan, ChinaKT961064KT961094
S19CortexP. chinense var. glabriusculumRutaceaeSichuan, ChinaKT961066KT961092
S20CortexP. chinense var. glabriusculumRutaceaeSichuan, ChinaKT961067KT961095
S21CortexP. chinense var. glabriusculumRutaceaeSichuan, ChinaKT961068KT961096
S22CortexP. chinense var. glabriusculumRutaceaeSichuan, ChinaKT961070KT961097
L4LeafP. chinense var. glabriusculumRutaceaeBeijing, ChinaKT961057KT961091
L5LeafP. chinense var. glabriusculumRutaceaeBeijing, ChinaKT961065KT961090
L6LeafP. chinense var. glabriusculumRutaceaeSichuan, ChinaKT961069KT961098
L7LeafTetradium ruticarpumRutaceaeChongqing, ChinaKT961040KT961099
L8LeafTetradium ruticarpumRutaceaeChongqing, ChinaKT961041KT961100
L9LeafTetradium ruticarpumRutaceaeChongqing, ChinaKT961042KT961101
L10LeafTetradium ruticarpumRutaceaeChongqing, ChinaKT961043KT961102

DNA Barcoding

All samples taken from the dried cortex or leaf (40 mg) were rubbed for two minutes at a frequency of 30 r/s. Total genomic DNA was extracted using the Plant Genomic DNA Kit (Tiangen Biotech Co., China) according to the manufacturer’s instructions. The extracted genomic DNA was amplified by polymerase chain reaction (PCR), using the ITS (ITS5F, 5′-GGA AGT AAA AGT CGT AAC AAG G-3′ and ITS4R, 5′-TCC TCC GCT TAT TGA TAT GC-3′) and psbA–trnH (fwdPA, 5′-GTT ATG CAT GAA CGT AAT GCT-3′ and revTH, 5′-CGC GCA TGG TGG ATT CAC AAT CC-3′) primers. PCR reaction mixtures contained 2 µL DNA template, 8.5 µL ddH2O, 12.5 µL 2×Taq PCR Master Mix (Beijing TransGen Biotech Co., China), 1 µL of each primer (2.5 µM), in a final volume of 25 µL. The PCR conditions of ITS were 35 cycles of denaturation at 94°C for 1 min, annealing at 50°C for 1 min, and extension at 72°C for 1.5 min (+3 s/cycle). The PCR conditions for psbA–trnH were 30 cycles of denaturation at 94°C for 1 min, annealing at 55°C for 1 min, and extension at 72°C for 1.5 min.14,2325) PCR products were separated and detected by 1.5% agarose gelelectrophoresis. Purified products were sequenced in both directions in a 3730XL sequencer (Applied Biosystems, U.S.A.).

HPLC Methods

Samples (0.5000 g) taken from dried cortex (sieved 65-mesh) were suspended in 50 mL 1% hydrochloric acid in methanol in a capped conical flask. They were then accurately weighed and extracted with three forty-minute ultrasonication steps. After cooling, weighing, and adding extraction solvent to compensate for the decrease in weight, the sample was mixed. The sample solution was filtered through a 0.22 µm membrane filter prior to the HPLC analysis. HPLC chromatographic conditions were conducted as described by Zhang et al.11)

Statistical Analysis

Sequences were assembled and aligned by the CodonCode Aligner 3.7.1 (CodonCode Co., U.S.A.). The ITS2 region was obtained using the HMMer annotation method based on the Hidden Markov model (HMM) to remove the 5.8S and 28S sections at both ends of the sequences.26) The inter/intra-specific genetic distances and the bootstrap Neighbor–Joining (NJ) tree were conducted by MEGA (4.0 Version), according to Kimura’s 2-parameter method with bootstrap testing of 1000 replicates. Hierarchical clustering analysis (HCA) was applied to demonstrate the variability of the relative peak areas of six bioactive compounds in 22 cortex samples of Phellodendri Cortex using SPSS (Version 19.0) and Unscrambler (Version 10.0). Principal component analysis (PCA), based on six compounds, was carried out using Unscrambler (Version 10.0).

RESULTS

Sequence Inter/Intra-Specific Variations Analysis

The base sequence of the ITS2 analysis showed 100% homology between all samples. The genetic distance of the ITS, psbA–trnH and ITS+psbA–trnH sequences were showed in Table 2, which was based on the K2P model calculated by MEGA5.1 Beta2. The average interspecific distance between Guanhuangbo and Huangbo was higher than their average intraspecific distance, except when the ITS sequence was used. This meant that the abilities of psbA–trnH and ITS+psbA–trnH to identify Guanhuangbo and Huangbo were superior to ITS. Moreover, the average interspecific distance of P. chinense and P. chinense var. glabriusculum was higher than their average intraspecific distance, except when the psbA–trnH sequence was used. This indicated that the ability of ITS and ITS+psbA–trnH to identify the origin of Huangbo was better than psbA–trnH. In summary, ITS+psbA–trnH was the superior sequence for tracing and identifying Phellodendri Cortex.

Table 2. The Genetic Distance of Phellodendron Genus Samples
Crude drugsScientific nameK2P genetic distancesRange of genetic distances (mean)
ITSpsbA–trnHITS+psbA–trnH
GuanhuangboP. amurenseIntraspecific distances0–0.0016 (0.0002)a)0–0.0047 (0.0018)0–0.0029 (0.0009)
HuangboIntraspecific distances0–0.0065 (0.0036)0 (0)0–0.0038 (0.0021)
P. chinenseIntraspecific distances0–0.0032 (0.0019)0 (0)0–0.0019 (0.0011)
P.chinense var. glabriusculumIntraspecific distances0–0.0048 (0.0030)0 (0)0–0.0029 (0.0014)
Interspecific distances between Guanhuangbo and Huangbo0.0016–0.0049 (0.0025)0.0141–0.0188 (0.0152)0.0067–0.0086 (0.0077)
Interspecific distances between P. chinense and P. chinense var. glabriusculum0.0032–0.0065 (0.0048)0 (0)0.0019–0.0038 (0.0029)

a) Average distance.

NJ Tree Analysis

An NJ tree illustrates the relationships between species and facilitates a characterization of their clustering. In this study, an NJ tree of DNA barcodes was built based on the K2P model. The NJ tree built on the ITS sequence demonstrated that P. amurense, P. chinense and P. chinense var. glabriusculum were short of monophyly. In the psbA–trnH sequence based NJ tree, Phellodendron genus species clustered into two clades: one was P. amurense and the other was P. chinense and P. chinense var. glabriusculum (Fig. 1). There was significant monophyly in the NJ tree built on the ITS+psbA–trnH sequence, where Phellodendron genus species separately clustered into three clades (Fig. 2). Thus, the NJ tree clearly distinguished between P. amurense, P. chinense and P. chinense var. glabriusculum as well as Guanhuangbo and Huangbo.

Fig. 1. The NJ Tree of the Phellodendron Genus Species with the psbA–trnH Sequence

The bootstrap scores (1000 replicates) were shown for each branch.

Fig. 2. The NJ Tree of the Phellodendron Genus Species with the ITS+psbA–trnH Sequence

The bootstrap scores (1000 replicates) were shown for each branch.

HPLC Fingerprints and PCA

A total of 22 cortex samples were examined, and representative HPLC fingerprints were showed in Fig. 3. The chromatographic data showed that the HPLC fingerprints of P. amurense had clear differences from those of P. chinense or P. chinense var. glabriusculum, whereas the fingerprints of the latter two species were almost identical. The chemical marker content in the cortex of P. amurense was distinct from that of P. chinense and P. chinense var. glabriusculum. The level of magnoflorine and palmatine chloride in the P. amurense samples was markedly higher with an apparently lower level of berberine, compared to P. chinense or P. chinense var. glabriusculum. By comparing the similarities of the HPLC chromatographic fingerprints together with the relative contents of magnoflorine, palmatine chloride and berberine chloride, Guanhuangbo and Huangbo were effectively differentiated. The result of the PCA’s loading plot revealed that magnoflorine, palmatine chloride and berberine chloride were powerful towards distinguishing between the two crude herbs (Fig. 4C). However, because of the HPLC fingerprints for P. chinense and P. chinense var. glabriusculum were similar, the HPLC fingerprints could not be used to discriminate the two species.

Fig. 3. HPLC Chromatographic Fingerprint Analysis of P. amurense (Upper), P. chinense, and P. chinense var. glabriusculum (Lower)

(1) Chlorogenic acid, (2) phellodendrine, (3) magnoflorine, (4) jatrorrhizine chloride, (5) palmatine chloride, (6) berberine chloride.

Fig. 4. HPLC Chemometrics Analysis of 22 Samples of Phellodendri Cortex

Dendrogram for hierarchical clustering by SPSS (A) and the Unscrambler (B); principle component loading plot by Unscrambler (C).

HCA

To validate the results of the the HPLC fingerprint analysis and to further elucidate the resemblance relationship among the samples, HCA was applied using the SPSS 19.0 software and the Unscrambler X 10.0 software. The relative peak areas of the six peaks that corresponded to chlorogenic acid, phellodendrine, magnoflorine, jatrorrhizine, palmatine and berberine were generated using the similarity analysis. The results of HCA showed that all 22 cortex samples were clearly divided into two clusters at level 1 (Figs. 4A, B). Group I was formed by all cortex samples of P. amurense, where as Group II consisted of all cortex samples of P. chinense and S17 (P. chinense var. glabriusculum) with all P. chinense samples forming one shorter cluster at level 2, indicating that P. chinense and P. chinense var. glabriusculum were similar in chemical composition. The results were consistent with the HPLC fingerprint analysis. Therefore, HCA was also helpful towards differentiating Guanhuangbo and Huangbo but insufficient for identifying the original plants of Huangbo.

DISCUSSION

Phellodendri Cortex is a typical example of the complex, phylogenetic origin of plants. P. amurense is the original plant of Guanhuangbo and is distributed in northeast China, where as P. chinense and P. chinense var. glabriusculum are the original plants of Huangbo and are distributed in southwest China.1) However, the crude materials of Guanhuangbo and Huangbo cannot be identified by cortex appearance, by organization structure or by their main chemical components.

DNA barcoding demonstrated an ability to universally and accurately discriminate the species in our study. The genetic distance and NJ tree analyses were assessed using DNA barcoding technology based on its ability to differentiate the Phellodendron genus species. The data supported the psbA–trnH barcode for discriminating Guanhuangbo and Huangbo (Fig. 1) and the ITS+psbA–trnH barcode for discriminating P. amurense, P. chinense and P. chinense var. glabriusculum (Fig. 2). DNA barcoding successfully traced the species of Phellodendri Cortex.

Additionally, we utilized the HPLC fingerprint and HCA methods to identify three cortexes. The data showed clear differences in the HPLC fingerprints (Fig. 3) and in the HCA (Fig. 4), allowing the differentiation of Guanhuangbo and Huangbo. However, the cortexes of P. chinense and P. chinense var. glabriusculum could not be distinguished using HPLC fingerprint and HCA methods.

In summary, we have established new chemical and molecular analysis methods for discriminating Phellodendri Cortex. Our results demonstrate that DNA barcoding overcomes the limitations of the HPLC fingerprint for differentiating close genetic relationships and similar chemical-composition species to guarantee an accurate and scientific confirmation of herbal identities in medicinal materials from multiple sources.

Acknowledgment

The authors are grateful for the financial support provided by the National Natural Science Foundation of China (No. 81473305).

Conflict of Interest

The authors declare no conflict of interest.

REFERENCES
 
© 2016 The Pharmaceutical Society of Japan
feedback
Top