Application of Partial Internal Transcribed Spacer Sequences for the Discrimination of Artemisia capillaris from Other Artemisia Species

Several Artemisia species are used as herbal medicines including the dried aerial parts of Artemisia capillaris, which are used as Artemisiae Capillaris Herba (known as “Injinho” in Korean medicinal terminology and “Yin Chen Hao” in Chinese). In this study, we developed tools for distinguishing between A. capillaris and 11 other Artemisia species that grow and/or are cultured in China, Japan, and Korea. Based on partial nucleotide sequences in the internal transcribed spacer (ITS) that differ between the species, we designed primers to amplify a DNA marker for A. capillaris. In addition, to detect other Artemisia species that are contaminants of A. capillaris, we designed primers to amplify DNA markers of A. japonica, A. annua, A. apiacea, and A. anomala. Moreover, based on random amplified polymorphic DNA analysis, we confirmed that primers developed in a previous study could be used to identify Artemisia species that are sources of Artemisiae Argyi Folium and Artemisiae Iwayomogii Herba. By using these primers, we found that multiplex polymerase chain reaction (PCR) was a reliable tool to distinguish between A. capillaris and other Artemisia species and to identify other Artemisia species as contaminants of A. capillaris in a single PCR.


Introduction
The genus Artemisia belongs to the Asteraceae family and is composed of 500 species that are mainly found in Asia, Europe, and North America [1,2]. Over 350 species in the genus Artemisia are grown in Asia, including China, Korea, and Japan [1]. Several Artemisia species have long been used for the treatment of disease in modern and traditional medicine [2,3]. For example, the dried aerial parts of A. capillaris are used as Artemisiae Capillaris Herba ("Injinho" in Korean medicinal terminology and "Yin Chen Hao" in Chinese) [4], which controls fever [2], protects the liver [5], and inhibits inflammatory responses [6]. However, the dried leaves of A. capillaris are often mistaken for those of A. japonica. Moreover, young A. capillaris leaves that are harvested in early spring are similar to those of A. argyi and A. princeps [7], which are sources of Artemisiae Argyi Folium ("Aeyup" in Korean and "Ai Ye" in Chinese) that is used for the treatment of pain, vomiting, and bleeding in the uterus [8].
Because of the morphological similarities among the dried and/or sliced shoots and leaves of Artemisia species, some are traded as other species in traditional herbal medicine markets [5,7]. To resolve this problem, various molecular biology techniques that are based on plant genetic information, such as gene nucleotide sequences (rbcL, matK, or a combination of both), have been used for plant identification and authentication, including medicinal plants [3]. Other gene sequences have been used to discriminate specific medicinal plants from an adulterant or substitute, for example, the trnL-F intergenic spacer for Coptis spp., matK for Rheum spp., and psbA-trnH for Phyllanthus spp. [9][10][11]. Internal transcribed spacer (ITS) sequences are effective discriminatory tools, and the ITS2 region in particular can be used as a universal DNA barcode for identifying plants

Plant Materials.
The fleshy aerial parts, including the leaves, of Artemisia species that grow and/or are cultivated in China, Korea, and Japan were collected ( Table 1). The samples were dried at room temperature, frozen, and stored at −80 ∘ C. The authenticity of the samples was verified by the Korea Institute of Oriental Medicine (KIOM) and the Department of Herbology, Wonkwang University. The voucher samples were deposited in the KIOM and the Department of Herbology.

Preparation of Genomic DNA.
Genomic DNA from each sample was extracted in accordance with the instruction manual for the NucleoSpin® Plant II (Macherey-Nagel, Duren, Germany). To improve DNA quality, phenolic compounds and polysaccharides were removed using 10% cetyltrimethylammonium bromide and 0.7 M NaCl. After the purity and amount of the prepared genomic DNA were determined using a NanoDrop™ DN-1000 spectrophotometer (Thermo Scientific, Wilmington, DE, USA), the DNA was diluted to 10 ng/ L and stored.

PCR Amplification
2.3.1. Amplification of ITS. A PCR for the amplification of the ITS, including the 5.8S rRNA coding region, was conducted using a T-personal cycler (Biometra, Goettingen, Germany) according to the protocol by White et al. [21]. In brief, 1.2 pmol of ITS1 (5 -TCCGTAGGTGAACCTGCGG-3 ) and ITS4 (5 -TCCTCCGCTTATTGATATGC-3 ) primers, 1 U Taq polymerase (ABgene, Epsom, UK), and 20 ng of genomic DNA extracted from each sample were used for the PCR amplification. During the 35-cycle PCR process, predenaturation was conducted for 5 min at 95 ∘ C and denaturation for 30 s at 95 ∘ C. The annealing process was conducted for 30 s at 52 ∘ C and the extension process for 1 min at 72 ∘ C. A final reaction step was conducted for 7 min at 72 ∘ C. The amplified products were separated on 1.2% agarose gel and revealed by staining with ethidium bromide (Sigma-Aldrich, St. Louis, MO, USA). The amplified PCR products were analyzed using MyImage (Seoulin Biotechnology, Seoul, Korea) and purified using a LaboPass™ Gel Kit (Cosmo Genetech, Seoul, Korea).

Amplification of DNA and SCAR Markers.
In brief, 1.2 pmol of primers, 1 U Taq polymerase (ABgene), and 50 ng of genomic DNA extracted from each Artemisia species were used for the PCR amplification. During the 23-cycle PCR process, predenaturation was conducted for 5 min at 95 ∘ C and denaturation for 30 s at 95 ∘ C. In general, the annealing process was conducted for 30 s at 53.5 ∘ C for the amplification of the DNA markers. However, to amplify the DNA markers for A. capillaris, A. japonica, A. apiacea, A. annua, and A. anomala, this process was conducted for 15-30 s at 54-58 ∘ C. The extension process was conducted for 20 s (except for A. apiacea, which had 30 s) at 72 ∘ C, and a final reaction step was conducted for 5 min at 72 ∘ C. To amplify an internal standard for the evaluation of the PCR conduct, a primer set (AYF/AYR) was used to amplify a 94 bp sequence. The amplified products were separated on 1.2% agarose gel and revealed by staining with ethidium bromide (Sigma-Aldrich). The amplified PCR products were then analyzed using MyImage (Seoulin Biotechnology).

Multiplex PCR.
For the multiplex PCR amplification, 0.07 pmol of the primers Fb and R7; 0.14 pmol of the primers AYF and AYR; 0.7 pmol of the primer Aam F3; 1.7 pmol of the primers AC F4, ACJ R3, and Aap R2; 3.4 pmol of the primers 2F1, 2F3, AJ F1, AC R3, Aap F1, AA F3, and Aa R4; 1x PrimeSTAR® Max DNA Polymerase (Takara Bio Inc., Kusatsu, Japan); and 20 ng of genomic DNA extracted from each Artemisia species were used. During the 30-cycle PCR process, predenaturation was conducted for 10 min at 95 ∘ C  and denaturation for 10 s at 95 ∘ C. The annealing process was conducted for 5 s at 56.5 ∘ C and the extension process for 10 s at 72 ∘ C. A final reaction step was conducted for 7 min at 72 ∘ C.
The amplified products were separated on 2% agarose gel and revealed by staining with ethidium bromide (Sigma-Aldrich).
In order to amplify an internal standard for the evaluation of the PCR, the AYF/AYR primer set was used to amplify a 94 bp sequence. The amplified PCR products were then analyzed using MyImage (Seoulin Biotechnology).

Nucleotide Sequencing of the PCR Products.
The nucleotide sequences of the PCR products were directly determined using the primers ITS1 and ITS4 by Macrogen (Seoul, Korea). In other cases, the PCR products resolved by agarose electrophoresis were cloned using a pGEM®-T Easy Vector System I (Promega, Madison, WI, USA). The nucleotide sequences of the subcloned PCR products were determined by Macrogen.

Alignment of the DNA Sequences and Construction of a
Dendrogram. The DNA sequences were manually edited and aligned by ClustalW multiple sequence alignment in BioEdit v7.0.9 (http://www.mbio.ncsu.edu/BioEdit/bioedit.html). A dendrogram was constructed using the neighbor-joining method [22] in the MEGA6 program [23] with 1000 bootstrap iterations. Evolutionary distances were computed using the maximum composite likelihood method [24] in MEGA6.

Determination and Analysis of ITS Sequences.
The 726-731 bp nucleotide sequences of the ITS, including the 5.8S region, were determined in 65 samples of 12 Artemisia species ( Table 1). Parts of the ITS sequences of each Artemisia species are presented in Figure 1 and were deposited in GenBank (accession numbers KT965653-KT965672). As shown in Figure 1, in the intraspecific samples of five Artemisia species (A. argyi, A. capillaris, A. iwayomogi, A. apiacea, and A. japonica), 4-9 bp differences in the ITS1 and ITS2 sequences were detected. In the case of A. japonica (sample numbers 55, 58, and 59), there were 8 bp differences in the ITS2 region and a 1 bp difference in the ITS1 region. These differences resulted mainly from substitutions (mostly base transitions) and a deletion. In A. apiacea (sample numbers 45 and 46), two base deletions in ITS1 and two substitutions in ITS2 were detected in sample number 45.
To determine whether each Artemisia species could be identified by interspecific ITS sequence differences, we constructed a dendrogram based on the ITS sequences. As outgroups, we used GenBank sequences of Aster yomena (accession number HQ154048.1) and Chrysanthemum coronarium (accession number EF577292.1) in the family Asteraceae, in which Artemisia is included ( Figure 2). As shown in Figure 2, each Artemisia species was classified into a separate group on the dendrogram. All of the A. japonica samples that exhibited excessive intraspecific ITS sequence variation were sorted into a group. Fortunately, the A. capillaris samples were separate from the A. japonica samples on the dendrogram. In addition, both A. annua and A. apiacea, which are sources of Artemisia Annuae Herba, were classified into the same cluster on the dendrogram ( Figure 2). Interestingly, A. argyi, A. princeps, A. montana, A. lavandulaefolia, and A. asiatica, which are sources of Artemisiae Argyi Folium, were classified into only one cluster.

Discrimination of A. capillaris from Other Artemisia Species by Differences in ITS Sequences.
Based on the results shown in Figure 2, we could discriminate A. capillaris from other Artemisia species, at least from the 11 Artemisia species used in this study, by differences in the ITS sequences. It was difficult to discriminate A. capillaris from A. japonica, which was close to A. capillaris on the dendrogram and exhibited significant variation in its ITS sequence. Most of the variation in the ITS sequences among the intraspecific A. japonica samples was found in the ITS2 region ( Figure 1); therefore, we excluded the ITS2 region when designing primers to amplify specific DNA markers for A. japonica. As shown in Figures 1  and 3, we designed the primer set AC F4/ACJ R3 in order to amplify a 189 bp PCR product in the ITS1 region that only appeared in A. capillaris samples (Figures 3 and 4(a)). Subsequently, we designed the AJ F1/AC R3 primer set in order to amplify a 176 bp PCR product in ITS2 that only appeared in A. japonica samples (Figures 3 and 4(b)).   25  26  28  33  35  37  40  45  46  50  55  58  59  61   490  500  510  520  530  540  550  560  570  580  590  600   1  4  9  13   19  18   25  26  28  33  35  37  40  45  46  50  55  58 Table 1). Bold arrows indicate the primers used to amplify DNA markers of the Artemisia species, and colored boxes represent nucleotide sequences as well as the positions of the ITS in the primers. Black boxes with an asterisk indicate variations in the nucleotides within species.
Based on these results, we suggest that two primer sets (AJ F1/AC R3 and AC F4/ACJ R3) could be used to discriminate A. capillaris not only from A. japonica but also from other Artemisia species.

Discrimination of Artemisia Species That Are Sources of
Artemisiae Annuae Herba and Artemisiae Anomalae Herba by Differences in ITS Sequences. We developed DNA markers in order to detect contamination of A. capillaris by other Artemisia species. As shown in Figure 2, A. annua and A. apiacea, which are sources of Artemisiae Annuae Herba, were close together on the dendrogram in a similar manner as A. capillaris and A. japonica. Therefore, we attempted to find region(s) in ITS1 and ITS2 to discriminate both A. annua and A. apiacea from other Artemisia species. As shown in Figures 1 and 3, we designed the AA F3/Aa R4 primer set in order to amplify a 543 bp PCR product in both A. annua and A. apiacea simultaneously as a common DNA marker. Subsequently, we designed primers to amplify a specific DNA marker to discriminate A. annua from A. apiacea. Based on Evidence- Based Complementary and Alternative Medicine   7   31  32  30  29  27  26  28  25  97 59  60   55  56  57   58  99  54  94  64  65  63  62  61  35  37  33  34  36  38  39  43  44  42  41  40  45  46  47  48  49  53  54  52  51  50  23  24  22  21  20  19  1  2  3  16  17  15  14   the differences found in the ITS1 and ITS2 sequences, we designed the Aap F1/Aap R2 primer set in order to amplify a 594 (in sample number 45, which had a 2 bp deletion) or 596 bp (in sample number 46) PCR product that only appeared in A. apiacea samples (Figures 1 and 3). Based on amplifications of the one or two PCR products expected on the gel (Figure 5(a)), we confirmed that the AA F3/Aa R4 and Aap F1/Aap R2 primer sets could discriminate not only A. annua from A. apiacea but also these two species from other Artemisia species. In the case of A. anomala, we designed an Aam F3/Aa R4 primer set in order to amplify a 492 bp PCR product in A. anomala samples (Figures 1 and 3) and confirmed that the expected 492 bp single band of the PCR product only appeared in A. anomala samples ( Figure 5(b)).

Detection of Contamination by Other Artemisia Species
Using Multiplex PCR. As shown in Figures 1 and 2, differences in the ITS sequences could discriminate five Artemisia species-A. asiatica, A. montana, A. lavandulaefolia, A. argyi, and A. princeps-that are sources of Artemisiae Argyi Folium and A. iwayomogi that is a source of Artemisiae Iwayomogii Herba from the six other Artemisia species. However, designing primers in order to amplify DNA markers for these species based on differences in the ITS sequences was difficult. Therefore, we tested the usability of the Fb/R7 and 2F1/2F3 primer sets in order to amplify SCAR markers that were developed in previous studies with six Artemisia species [5,7]. We confirmed that the Fb/R7 primer set amplified a 254 bp SCAR marker in samples of not only A. princeps and A. argyi but also A. asiatica, A. lavandulaefolia, and A. montana (data not shown). Furthermore, we confirmed that the 2F1/2F3 primer set amplified a 364 or 365 bp SCAR marker only in A. iwayomogi, and that this marker was not amplified in any other species, including A. asiatica, A. montana, A. lavandulaefolia, A. annua, A. apiacea, or A. anomala (data not shown). Therefore, these two RAPD-based primer sets could detect contamination by these Artemisia species in addition to the six other Artemisia species. Using the multiplex PCR method, we tested the reliability of these two primer sets and those developed based on the ITS sequences to discriminate A. capillaris from other Artemisia species and to detect contamination by other Artemisia species. For the multiplex PCR process, we randomly selected one sample from each Artemisia species listed in Table 1. As shown in Figure 6, these primer sets functioned reliably, not only to discriminate A. capillaris from other Artemisia species, but also to simultaneously detect contamination by other Artemisia species in a single PCR process. Finally, by mixing genomic DNA isolated from different Artemisia species at varying content ratios, we tested the reliability of this PCR method to detect contamination of other Artemisia species, such as A. japonica, A. princeps, and A. iwayomogi, which are mostly found in Korea and are easily misused. As shown in Figure 7, the multiplex PCR detected two Artemisia species that had been mixed at ratios of 9 : 1 and 19 : 1. Furthermore, the multiplex PCR detected a mixture of three Artemisia species (A. capillaris with A. japonica and A. princeps or A. capillaris with A. japonica and A. iwayomogi) at ratios of 8 : 1 : 1 and 18 : 1 : 1 (Figure 7). Therefore, we suggest that the multiplex PCR method is an accurate tool to discriminate A. capillaris from other Artemisia species and