Identification of constituent herbs in ginseng decoctions by DNA markers

Background DNA in herbal decoctions is usually fragmented by extensive boiling and is usually regarded as unsuitable for molecular authentication. This study aims to evaluate the feasibility for molecular authentication methods by multiplex polymerase chain reaction (PCR) and DNA sequencing for decoctions. Methods The DNA extraction procedure, sample pulverization and boiling time were examined in (1) single herb decoction with Panax ginseng (“ginseng”) or P. quinquefolius (“American ginseng”), (2) decoctions of two classical ginseng prescriptions of five herbal materials (Aconitum carmichaeli, Atractylodes macrocephala, P. ginseng, Glycyrrhiza uralensis and Zingiber officinale) and (3) a commercial ready-to-serve ginseng soup. Primers were designed from 26S-18S, ITS2 and trnH-psbA region, with DNA sequences obtained from GenBank. Multiplex PCR was also employed in ginseng or American ginseng decoctions to differentiate between these two herbs. Results All six herbal species tested in this study could be identified in decoctions. We had four main observations: (1) sample pulverization before boiling improved PCR detection; (2) prolonged boiling increased the DNA concentration but decreased the intactness of DNA fragments; (3) ginseng could be differentiated from American ginseng by multiplex PCR; (4) identification of individual herbs in multi-herb decoctions with prolonged boiling time of 180 min could be achieved. Conclusions DNA could be amplified from extensively boiled ginseng decoctions, multi-herb decoctions and commercial soup although DNA degradation was critical to successful PCR. Electronic supplementary material The online version of this article (doi:10.1186/s13020-015-0029-x) contains supplementary material, which is available to authorized users.


Background
Intentional or inadvertent adulteration of herbal products, including health food and dietary supplements, can cause adverse effects or even poisoning [1][2][3][4][5][6]. For instance, misidentification of the toxic vine Gelsemium elegans as Mussaenda pubescens (an ingredient of Chinese health drinks) caused two poisoning cases in Hong Kong [3]. Adulteration of the anti-inflammatory agent Stephania tetrandra by the toxic herb Aristolochia fangchi led to more than 100 cases of kidney failure and urothelial carcinoma in Belgium [4]. Adulterants have also been found in expensive herbs such as Panax quinquefolius [7], and materials from endangered species such as the horn of Saiga tatarica [8]. Therefore, reliable and effective methods of authentication are crucial for the quality control of herbal materials and their products.
Authentication of highly processed herbal materials is difficult. Examples include comparison of the morphological and microscopic characteristics of easily confused herbs and their dregs after boiling [9,10]. The approaches described could be useful, but the dregs in a decoction are usually discarded before consumption. Authentication is also possible by chemical or physical analyses such as high performance liquid chromatography [11], mass spectrometry [12] and nuclear magnetic resonance [13]. However, the chemical constituents in decoction may be altered by physiological conditions, harvesting period, storage and boiling conditions of the herbs involved. Such methods also make differentiation between related species quite difficult owing to their similar chemical profiles and demanding sophisticated instrumentation [14].
Since the 1990s, molecular methods have been used for authentication. There have been reports of using internal transcribed spacer 2 (ITS2) region to distinguish Lycium barbarum from its various types [15]; polymerase chain reaction (PCR) fragment length polymorphisms in the 3′ untranslated region of cDNA sequences to identify the root of Astragalus membranaceus [16]; DNA microarrays to authenticate ginseng drugs [17]; and PCR based method to characterize alfalfa and red clover [18]. Molecular methods and employment of DNA barcodes can produce accurate, reliable and sensitive results [19,20].
China was the first country to recognize molecular methods as a legal basis for authentication of crude herbal drugs, including Fritillaria unibracteata [21] and Zaocys dhumnades [22]. The future of molecular authentication is promising, even with DNA degradation during processing. For instance, molecular authentication of meats in processed food had been reported [23][24][25] and short microsatellite DNA fragments could be obtained from a highly processed medicinal product prepared from the skin of Equus asinus [26]. A recent report described using the ribosomal external transcribed spacer region to authenticate ginseng products such as tea and red ginseng extract [7]. However, how these products were processed and whether they contained ginseng residues was not mentioned.
Identification of herbal ingredients in decoctions is highly significant because this method will extend molecular authentication of herbal products. This method may also be useful for law enforcement authorities and may benefit consumers. Thus this study aims to evaluate the feasibility for molecular authentication methods by PCR and DNA sequencing for decoctions. We employed a single herb decoction with P. ginseng ("ginseng") or P. quinquefolius ("American ginseng"), decoctions of two classical ginseng prescriptions with multiple herbal materials, and a commercial ready-to-serve ginseng soup to investigate the availability of DNA for molecular authentication. Then, we identified the DNA in the decoctions by multiplex PCR and DNA sequencing.

Samples
Herbal materials were purchased in a herbal shop (Zisun Pharmaceutical Ltd., Guangzhou, China). All herbal samples were identified with their organoleptic characteristics by Dr. David Tai-Wai Lau, Curator of the Shui-Ying Hu Herbarium, School of Life Sciences, The Chinese University of Hong Kong according to the Pharmacopoeia of the People's Republic of China (2010 edition) [22] and confirmed by DNA sequencing. All authentic voucher specimens were deposited in the Institute of Chinese Medicine, The Chinese University of Hong Kong (Table 1).
Dust and soil on the surface of samples were removed by cleaning with water. These herbs were prepared for decoctions. A commercial ready-to-serve ginseng soup, Korean Ginseng Chicken Stew (Harim, Korea; product code: 8801492307002; voucher code: T3475), which claimed to contain P. ginseng on the label, was purchased from a local supermarket.

Preparation of decoctions
A single herb decoction was prepared from P. ginseng or P. quinquefolius. Each herb was divided into two equal portions. One portion was pulverized into fine powder by a grinder. Another portion was retained in its sliced form. Herb (10 g) was boiled in 200 mL double-distilled water. Volume of the decoction was maintained at 200 mL by adding water. Herbal decoction was collected after 30, 60 and 120 min of boiling.
Two multi-herb decoctions (i) ginseng and prepared aconite, and (ii) ginseng and ginger combination were prepared. (i) was prepared from Aconitum carmichaeli and P. ginseng according to the protocol from A Repertory of Traumatology. A. carmichaeli (9 g) and P. ginseng (12 g) were cut into small pieces and boiled in 200 mL double-distilled water [27] for 60 min. (ii) was prepared from Atractylodes macrocephala, P. ginseng, Glycyrrhiza uralensis and Zingiber officinale according to the protocol from Synopsis of Golden Chamber. Each herb (46.9 g) was boiled for 180 min in 1600 mL double-distilled water until 600 mL decoction remained [28].

Treatment of decoction samples
For each decoction sample collected at different times of boiling, large sediments were removed by MiniSpin centrifuge (Eppendorf, Hamburg, Germany). One portion of the supernatant was used for direct PCR and was regarded to be "crude DNA". Another portion was lyophilized, followed by DNA extraction using a QIAquick Nucleotide Removal kit (Qiagen, Hilden, Germany) according to manufacturer instructions for the single herb decoctions; modified cetyltrimethylammonium bromide (CTAB) DNA extraction (protocol 1) (Additional file 1) for the multi-herb decoctions; and modified CTAB DNA extraction (protocol 2) (Additional file 1) for the Korean Ginseng Chicken Stew. This portion was regarded to be "extracted DNA". Crude and extracted DNA were stored at −20°C immediately and PCR analyses were carried out within one month.

Sequence alignment and primer design
Forty five DNA sequences of herbs were downloaded from the GenBank, National Center for Biotechnology Information (NCBI). These reference sequences were from A. carmichaeli trnH-psbA intergenic spacer, A. macrocephala ITS2, G. uralensis ITS2, P. ginseng and P. quinquefolius 26S-18S intergenic spacer and Z. officinale ITS2 (Additional file 2). These multi-copy sequences are abundant in the herbs, available in GenBank, highly variable and can be used effectively for identification.
Sequence alignment was carried out using ClustalW2 (European Bioinformatics Institute, Hinxton, UK). Primers were designed with sequences shown in Table 2. Their melting temperatures, percentage of GC content, hairpin formation, as well as formation of self-dimers and hetero-dimers were determined by OligoAnalysis v3.1 (Integrated DNA Technologies, Coralville, IA, USA).

PCR and DNA sequencing
Different DNA loci were amplified by conventional PCR or multiplex PCR using the designed primers (

DNA concentration and statistical analyses
Extracted DNA was diluted so that its concentration was the same as that in the decoction. DNA concentration was measured using a Qubit 2.0 Fluorometer (Invitrogen, Carlsbad, CA, USA) and presented as the mean ± standard deviation. Differences between boiling times were assessed by one-way analysis of variance. Differences between sliced and pulverized samples at the same boiling time were analyzed by the two-sample t test. P < 0.05 was considered statistically significant.

PCR products and DNA concentration in single herb decoction
We investigated if a 191 bp DNA of the P. ginseng 26S-18S intergenic spacer region could be amplified from the single herb P. ginseng decoction prepared by boiling sliced or pulverized samples for 30, 60 and 120 min. With the concentration of the crude and extracted DNA adjusted to that in the original decoction, visible PCR products were found in crude and extracted DNA with samples boiled for 30 min. More PCR product was obtained in pulverized samples boiled for 60 min ( Figure 1A), and there was a significant increase in DNA released into the decoction in pulverized samples for each boiling time compared with sliced samples (30 min: P = 0.0280; 60 min: P = 0.0319; 120 min: P = 0.0479) ( Figure 1B). A visible PCR product was not observed upon boiling for 120 min, although it was a time at which the highest DNA concentration was measured.

Size of PCR product with boiling time
We investigated the relationship between boiling time and the size of amplifiable DNA. DNA fragments with sizes (in bp) of 88, 121, 191, 249 and 311 were amplified.
For pulverized samples boiled for 30 and 60 min, DNA fragments of size 88-311 bp were amplified (Figure 2A and B). However, only short PCR products with sizes ranging from 88-121 bp were amplified for samples boiled for 120 min ( Figure 2C).

Multiplex PCR for differentiation between ginseng and American ginseng
Multiplex PCR was developed to differentiate P. ginseng from P. quinquefolius. PCR master mixes involved one pair of conserved primers for P. ginseng and P. quinquefolius (Panax_F1 and Panax_R1), and a forward P. ginsengspecific primer (Pgi_SPF1) region for pairing with the same reverse conserved primer (Panax_R1) (Additional file 3). Three nucleotides at the 3′ end of the 21 nucleotide The concentration of extracted P. ginseng DNA was adjusted to make it similar to that in the original decoction and measured by Qubit 2.0 Fluorometer. Water was used as blank and the data represent mean ± standard deviation (n = 3). Difference between boiling time was analyzed by one-way analysis of variance. Difference between sliced and pulverized samples at the same boiling time was analyzed by two sample t test (*p < 0.05).
forward P. ginseng-specific primer were different from those of the DNA sequence of P. quinquefolius ( Figure 3A). As a result, the 191 bp amplicon was found in P. ginseng and P. quinquefolius decoction and the 121 bp amplicon was found only in the P. ginseng decoction ( Figure 3B and C). However, the 191 bp amplicon was not found in the decoction boiled for 120 min ( Figure 3B and C) due to DNA degradation ( Figure 2C).

Identification of herbs used in multi-herb decoctions and commercial soup
In a decoction involved two herbs (ginseng and prepared aconite), A. carmichaeli (148 bp) and P. ginseng (121 bp) could be amplified in extracted DNA with samples boiled for 30 and 60 min ( Figure 4A). In another decoction involved four herbs (ginseng and ginger combination), A. macrocephala (108 bp), G. uralensis (106 bp), P. ginseng (121 bp) and Z. officinale (106 bp) were amplified in extracted DNA with samples boiled for 180 min ( Figure 4B-E). We could also amplify the P. ginseng component in Korean Ginseng Chicken Stew after DNA was extracted ( Figure 5).

Discussion
In the present study, DNA released into the decoction by boiling could be amplified by PCR. BLAST analyses of these amplicons showed ≥99% coverage and identity between the amplified and reference sequences (Additional file 4). Pulverization of the sample increased the yield of DNA released into the decoction ( Figure 1B) due to the increase in the surface area of the sample. The highest DNA concentration was measured in the decoction boiled for 120 min but a PCR product was not observed ( Figure 1A). An increase in the duration of heat treatment can cause a higher degree of DNA fragmentation in meat samples

EF031236
T A T [29]. This phenomenon might also be the case in decoction, resulting in a less amplifiable DNA template that is intact. We confirmed this hypothesis by amplifying DNA of sizes ranging from 88-311 bp. Only short PCR products of ≤121 bp were observed with 120 min boiling ( Figure 2C). Therefore, the amplicon to be studied in the decoction should be short. Sometimes this requirement may mean that the ability to differentiate between species is sacrificed. In our case, because of limited sequences in GenBank, our primers designed for A. carmichaeli (primer pair 1 in Table 2) and G. uralensis (primer pair 3 in Table 2) were specific at the genus level. Other primers had specificity at the species level (Additional file 4). The PCR product was detected by amplifying crude DNA in the single herb decoction directly ( Figure 1A). The chemicals released from herbs might not be detrimental to PCR in certain cases. This phenomenon increased the efficiency of authentication because time-consuming DNA extraction procedures could be omitted.
In the market, it is common to find cultivated P. ginseng as an adulterant of P. quinquefolius. These two herbs have different pharmacological effects and chemical   compositions [30,31]. We investigated the possibility of differentiating P. ginseng and P. quinquefolius in decoctions by multiplex PCR. PCR master mixes involved one pair of conserved primers for P. ginseng and P. quinquefolius, and a forward P. ginseng-specific primer (Additional file 3). The 191 bp conserved amplicon was amplified and served as the positive control to ensure sufficient intact DNA fragments presented in the decoction. There were three nucleotide mismatches with the DNA sequence of P. quinquefolius at the 3′ end of the forward P. ginsengspecific primer sequence (i.e., nearly 15% mismatch) ( Figure 3A), which made the primer sufficiently specific for producing the 121 bp P. ginseng-specific PCR product ( Figure 3B). This approach may be explored for differentiating between genuine and adulterant herbs in a decoction and, if coupled to quantitative PCR, for finding the ratio of the two herbs. Herbal prescriptions usually consist of more than one herb. Two classical ginseng prescriptions (ginseng and prepared aconite, ginseng and ginger combination) and a commercial ginseng soup were tested to illustrate the possibility of identification of herbs by DNA markers in multi-herb decoctions. Results showed that DNA fragments of the relevant herbal ingredients were amplified (Figures 4 and 5).
Some of the herbs used in the prescriptions were dried and some were even processed (Table 1). In comparison with fresh materials, carrying out molecular authentication for dried and processed herbal materials is difficult because DNA is degraded or interacts with other chemicals. Several protocols have been developed to facilitate DNA extraction of different herbal samples: silica binding [32], salting-out precipitation [33], magnetic beads binding, such as ChargeSwitch gDNA Plant Kit (Invitrogen, Carlsbad, CA, USA) and anion exchange purification, such as Genomic Tip 20/G (Qiagen, Hilden, Germany) [34]. There are also reports to extract PCR amplifiable DNA from maize, soy bean [35], fish derived food products [36] and rice wine [37]. Our study shows the presence of PCR amplifiable DNA fragments of herbal materials in the aqueous solution after boiling. This phenomenon may be investigated further for other herbal decoctions.