Identification of plant materials containing ephedrine alkaloids based on DNA barcoding and TaqMan real-time PCR assay

An intentional or inadvertent mixing of plant materials containing ephedrine alkaloids, especially Ephedra, is illegal. In order to better detect plant materials containing ephedrine alkaloids in export and smuggling, DNA barcoding combined with a TaqMan real-time PCR-based assay were used in this study. We collected 201 samples from 18 species belonging to four genera distributed in three families to amplify two barcoding markers, internal transcribed spacer 2 (ITS2) and psbA-trnH. 175 ITS2 sequences and 136 psbA-trnH sequences were obtained. Alignments and the neighbor-joining tree indicated that ITS2 showed a better discrimination of species than psbA-trnH. In addition, based on the sequence comparison of the ITS2 region from 18 species, three sets of primer/probes were designed using the real-time PCR assay platform. A sensitivity test showed the above primer/probe sets were specific and sensitive (as little as 0.0050 ng/μL target DNA) to identify species containing ephedrine alkaloids. Hence, the combination of novel DNA barcoding and the TaqMan real-time PCR-based technique are promising tools for the identification of plant materials containing ephedrine alkaloids. It is beneficial to standardize market circulation and detection of plant materials containing ephedrine alkaloids.


Introduction
Beginning in 1995, the China Food and Drug Administration (CFDA) enacted several laws to regulate the use of ephedrine alkaloids. One of the most publically known restrictions is the policy for cold medications. More notably, the US Food and Drug Administration (FDA) Import Alert 54-13 regarding the dietary supplements and bulk dietary ingredients containing ephedrine alkaloids, which went into effect on December 14, 2012. Consequently, plant materials containing ephedrine alkaloids have triggered a wide concern in public health care safety around the world.
In China, plant species containing ephedrine alkaloids are generally found in the genus Ephedra, known as ma huang in Mandarin Chinese. Ephedra is the only genus in Ephedraceae, and it includes approximately 50 species around the world (Price 1996). There are ca. 14 Ephedra species in China, distributed mainly along shores, rocky beaches, and in sandy soils (Committee 1999). Apart from Ephedra (Caveney et al. 2001), some species also are known to contain ephedrine alkaloids, including Pinellia (Araceae), Typhonium (Araceae), and Sida (Malvaceae) (Caveney et al. 2001;Khatoon et al. 2005). Some species that do not contain ephedrine alkaloid and are easy to mix with ephedra herbs include Equisetum hyemale L., Equisetum ramosissimum Desf. et al.
Based upon a literature review, we identified 18 species containing ephedrine alkaloids distributed in China. Among them, Ephedra sinica Stapf, Ephedra intermedia Schrenk et C.A. Mey., Ephedra equisetina Bge., and Pinellia ternata (Thunb.) Breit. are described in the Chinese Pharmacopoeia, serving as a remedy for asthma, respiratory tract diseases, and vomiting in traditional Chinese and Indian medical Communicated by J. Van Huylenbroeck.
Yaqin Zheng and Han Gao have contributed equally to this work. systems (Committee 2015). Sida cordifolia Linn. is rarely used in China, but it is known as Bala in Ayurvedic medicine and widely used to treat illnesses like pulmonary tuberculosis, rheumatism, hematuria, urinary and heart diseases in India (Khatoon et al. 2005).
In addition to disease prevention, medicinal plants containing ephedrine alkaloids are widely used in health care products for fat loss, which are effective and popular, but often ephedrine alkaloids are added to these pills without proper labeling. Use in such a way may increase the risk of heart attack, stroke, and even death if misused (Haller and Benowitz 2002;Samenuk et al. 2002). Furthermore, medicinal plants containing ephedrine alkaloids have served as an important source of amphetamine in clandestine laboratories (Lee et al. 2006) and the lawless smuggling of medicinal plants containing ephedrine alkaloids continues to be a problem. Hence, it is urgent to develop a valid and sensitive method to effectively identify and detect plant materials containing ephedrine alkaloids to standardize the market circulation and ensure clinical safety.
Plant morphology and chemical analysis methods have long been used to identify plant materials containing ephedrine alkaloids (Sun et al. 2004;Kim et al. 2005;Li et al. 2008;Chatterjee et al. 2013). Nevertheless, if the diagnostic morphological features are not visible or overlooked, it is difficult to identify. For example, when the specimen is processed, it may no longer harbor its classical morphological features. Furthermore, chemical variability of the plant materials may obstruct the verification of the botanical traceability, as chemical composition varies with geographic condition and post-harvest processing (Figueredo et al. 2017).
Sequence-based technology has proven to be a powerful tool to authenticate species. DNA barcoding is diagnostic technology that uses a short, standard DNA fragments to distinguish species. It stands out for its advantages in generality and practicality without being affected by the morphological identification experience (Gu et al. 2013). In plants, the Consortium for the Barcode of Life (CBOL) recommends the combination of matK and rcbL as the core barcode. This combination has a high degree of species differentiation between genera and a low rate of species identification within genera (Hollingsworth et al. 2009). Recently, the internal transcribed spacer 2 (ITS2) has been shown to be useful in exploring systematic relatedness under genera and species (Goudsmit et al. 2003;Schultz and Wolf 2009). ITS2 sequence had an obvious advantage in the success rate of amplification and sequencing and the comparative similarity (Chen et al. 2010). And it was proposed to be a core barcode and psbA-trnH to be a complementary barcode for identifying medicinal plant species . Many previous studies have validated that the combination of ITS2 and psbA-trnH possesses much more recoverability, universal sequencing applicability, and species discrimination ability (Xiang et al. 2013;Han et al. 2016;Mishra et al. 2016;Carrero and Hoyos 2018;Ha et al. 2018).
Real-time PCR is a technique that can quantify the nucleic acid copies based on different fluorescences. TaqMan realtime PCR is characterized by its high specificity and sensitivity, and has been widely applied in molecular fields such as clinical microbiology, gene expression, tumor immunology, detection of minimal residual lesions, and polymorphism (Ohst et al. 2015;Xu et al. 2016;Kading et al. 2017). Alasaad developed a TaqMan real-time PCR-based assay to identify Fasciola spp (Alasaad et al. 2011). Wu distinguished toxic Aristolochia manshuriensis from Akebia quinata to prevent renal damage using TaqMan real-time PCRbased assay (Wu et al. 2015).
In the present study, ITS2 and psbA-trnH were amplified from 201 samples, spanning four genera and three families. The Ephedra group included 128 samples from 11 species. The Araceae group contained 53 samples from four species. The Sida group included 20 samples from three species. TaqMan real-time PCR assays were conducted to detect medical plant materials containing ephedrine alkaloids. The aims of this study were to: (1) study the diversity of two barcode regions in several species and genera and (2) develop an assay to discriminate the different groups (Ephedra vs Pinellia/Typhonium vs Sida).

Materials
Two-hundred and one sample belonging to Ephedra, Pinellia, Sida, and Typhonium were collected from Xinjiang, Inner Mongolia, Sichuan, Gansu, Beijing, Guangxi, Hainan, Yunnan, and Hubei Provinces (Table s1). All the samples were morphologically identified by Wei Sun at the Institute of Chinese Materia Medica, China Academy of Chinese Medical Sciences. The stem and root samples collected from the field were dried on silica gel. The majority of Ephedra collected in Xinjiang were dried naturally. All corresponding voucher samples were further deposited in the Institute of Chinese Materia Medica, China Academy of Chinese Medical Sciences, Beijing, China. Forty-five ITS2 sequences and 17 psbA-trnH sequences were downloaded from GenBank for comparative and complementary analyses (Table s2).

DNA extraction
Total genomic DNA was extracted using Plant Genomic DNA Kits (Tiangen Biotech Co., China) following the manufacturer's protocol with the following adjustments: repeated Page 3 of 10 143 chloroform:isoamylol (24:1, v/v) treatment and a final DNA precipitation using ice cold isopropanol to precipitate DNA (Moyo et al. 2008).

DNA amplification and sequencing using primers of ITS2 and psbA-trnH
The ITS2 region was amplified from genomic DNA using two pairs of primers, S2F/3R or P3/E4 (Table S3) (Chen et al. 2010;Li et al. 2011). The psbA-trnH sequences were amplified using the universal primers: the psbAF and trnHR and another pair of primer fwd PA and rev TH (Table S3) (Chen et al. 2010;Li et al. 2011). The PCR reaction mixture consisted of 2 μL (~ 35 ng) DNA, 12.5 μL of 2 × Taq PCR Master Mix (Aidlab Biotechnologies Co., China), 1.0 μL of 2.5 μmol·L −1 forward and reverse primers, respectively (synthesized by Sangon Biotech (Shanghai) Co., China), and distilled deionized water in a final volume of 25 μL. General PCR conditions were conducted as previously reported (Kress et al. 2005). The PCR products were examined with 1.0% agarose gel electrophoresis and were sequenced bidirectionally using a 3730XL sequencer (Applied Biosystems, USA).

Sequence analysis
A Codon Code Aligner V 3.7.1 (CodonCode Co., USA) was used to proofread and assemble the sequencing peak diagrams. The ITS2 region was obtained based on the Hidden Markov model (HMM) to remove the 5.8S and 28S rRNA sections at both ends of the sequence on the ITS2 database (http:// its2. bioap ps. bioze ntrum. uni-wuerz burg. de/) (Keller et al. 2009). The psbA-trnH region was obtained after trimming the psbA gene and trnH gene by referencing the submitted sequence noted clearly in GenBank. All of the ITS2 and psbA-trnH sequences were deposited in GenBank, omitting accessions with identical sequence information. The GenBank IDs are as follows: all psbA-trnH sequences range from KX779046-KX779070, the ITS2 sequences for Ephedra range from KX779071-KX779088, and the ITS2 sequences for Sida are KX779089-KZ779111. The ITS2 sequences for Pinellia and Typonium were KM236606, KM236618, KM236622, KM236659.
To evaluate the variability and the discriminating ability, MEGA 6.0 and Clustal W were applied. And the neighborjoining (NJ) tree analysis was used for specie identification.

Designed primers and probes
To develop the TaqMan Real-time PCR, we firstly designed three groups of primers and probes. We divided the studied plant species into three groups according to their homology. In addition to the 11 studied species, another four species distributed in China without acquisition of samples, E. saxatilis, E. gerardiana, E. likiangensis, and E. minuta, were downloaded from GenBank to design the primers and probe for the Ephedra group. Primers and probes were designed by Applied Biosystems 7500 real-time quantitative PCR software-primer Express 3.0 using default parameters. Three groups of probes were purified using HPLC, all of the 5' ends of the oligonucleotides were attached to FAM (6-carboxyfluorescein) reporter dye, and the 3' ends were labeled with a TAMARA (6-carboxy-tetramethyl-rhodamine) quencher dye. Both the primers and probes were synthesized by Invitrogen (Life Technologies, China; Table 1).
To guarantee the accuracy of TaqMan Real-time PCR, the specificity of the three pairs of primers was tested as follows. The genomic DNA was firstly examined by the amplification of ITS2 region using universal primers mentioned above. Once the quality of DNA was examined, the DNA was used as template to conduct regular PCR using the designed specific primers. In Ephedra group, all the 21 haplotypes from 141 Ephedra samples and one representative sample from Pinellia and Sida were amplified using the specific primers of Ephedra group. The amplification products were inspected with gel electrophoresis. The same assays were conducted in the Pinellia and Typhonium group, Sida group.

TaqMan real-time PCR assay
Amplification reactions contained 0.5 μL of 10 mol·L −1 primer, 0.5 μL of 10 mol·L −1 of TaqMan probe, 10 μL of 2 × Goldstar TaqMan Mixture with ROX (TaKaRa Biomedical Technology (Beijing) Co. China), 1 μL of DNA solution (replaced by water for negative controls), and nuclease free water in a final volume of 20 μL. Cycling conditions for the Table 1 The primer and probe of three sets According to the homology, the experimental species were divided into three groups. Three groups primers and probes were designed

Assessment of sensitivity of the TaqMan assay
To evaluate the specificity, all the 42 haplotypes covered 201 ephedrine alkaloids containing plant samples were used as template to develop the TaqMan Real-time PCR. The sensitivity of our assay was assessed using a tenfold dilution series (between 80 ng/μL and 8 ng/μL) and a twofold dilution series (between 8 ng/μL and 0.0050 ng/μL) of E. sinica gDNA. The detection limit was based on the final dilution at which the fluorescent signal of the TaqMan probes can still be exponentially amplified.

Efficiency of the PCR amplification
In the Ephedra group, the success rates of PCR amplification of ITS2 and psbA-trnH sequences were 88.4 and 92.2%, respectively. Considering Sida, the PCR amplification success rates for the two regions were both 45%. However, in the Pinellia and Typhonium group, the success rate of the PCR amplification for ITS2 was 100% using the primer pair P3/E4, and 62.5% using the pair S2F/3R. The primer pair PA/TH for the psbA-trnH region had 12.5 and 75% success amplification rates, respectively, for P. ternata and P. pedatisecta. Considering the low amplification rate for psbA-trnH in Pinellia and Typhonium, another primer pair, fwd/rev, was utilized. Nevertheless, the amplification rate success for psbA-trnH in P. ternata was 36.1, and only 30.8% of the PCR products were sequenced. No successful amplification was obtained for P. cordata and T. flagelliforme using the PA/TH and fwd/re primer pairs.

Alignment of ITS2 sequences and psbA-trnH sequences
The GC content was calculated and variable sites were analyzed, the characteristics of the ITS2 sequences and psbA-trnH sequences of the ephedrine alkaloids containing plants were summarized in Table 2 and Table 3, respectively. The length of the obtained ITS2 region from the 11 species of Ephedra in this study was 251 bp with 21 haplotypes ( Figure S1). The length of the ITS2 region for P. ternata, P. cordata, P. pedatisecta, and T. flagelliforme was 251, 250, 252, and 239 bp, respectively. For these four species, the aligned length of ITS2 was 255 bp, with 16 existing haplotypes and 71 variable sites in total. The length of ITS2 among Sida species was 230 bp. Only two haplotypes and two nucleotide variation sites were detected among Sida spp. For psbA-trnH sequences, the GC content among Ephedra species was approximately 38.0%. The sequence length in eight Ephedra spp. was 461 bp, and the sequence length was 473 bp in E. fedtschenkoae and E. monosperma. However, the length psbA-trnH in E. equisetina was 461 bp and 473 bp. From the multiple sequence alignments of psbA-trnH of Ephedra, we found that the differences consisted of insertions, deletions, and single nucleotide changes. Therein, only E. sinica showed a C/T transition (1/19) and E. monosperma showed A/C transversion (4/10). Notably, the single nucleotide was a "C" at site 300 in E. monosperma in the four psbA-trnH sequences obtained from individuals collected in Sichuan, while sequences from individuals collected in Xinjiang and Gansu read as "A". Compared with the corresponding psbA-trnH sequence downloaded from GenBank, deletions or insertion of 12 bases were found in E. intermedia, E. regeliana and E. fedtschenkoae, which was the base "TTG GAT TTC CTG ". In all the 33 psbA-trnH sequences of E. equisetina, 3 experimental sequences existed 12 bases deletion compared with other experimental sequences and those downloaded by GenBank.

Phylogenetic analysis
In this study, a phylogenetic tree based on Kimura 2 parameter (K2P) model was constructed for both ITS2 and psbA-trnH sequences. The NJ-tree based on ITS2 sequences demonstrated that the three groups clustered into three clades (Fig. 1). In the Ephedra clade, the haplotypes of E. intermedia, E. przewalskii, E. distachya dispersed to different group. The remaining nine Ephedra species clustered into one subclade. In the Pinellia and Typhonium clade, four species grouped into four subclades, illustrating monophyly for these genera. The three Sida species included in this study grouped into one clade.
The NJ tree based upon psbA-trnH sequences was constructed without P. cordata and T. flagelliforme, as no sequences were available on GenBank or obtained in our study (Fig. 2). The psbA-trnH NJ tree showed that Ephedra, Sida, Pinellia, grouped into three clades. Seventeen haplotypes of 11 Ephedra species grouped into one clade, with no subclades, and P. ternata and P. pedatisecta grouped into two branches. Four species of Sida clustered together.

The specificity of the designed primers
The ITS2 regions of all the 42 haplotypes of ephedrine alkaloids containing plant materials, including Ephedra, Pinellia, Typhonium and Sida, were successfully amplified firstly ( Figure S2). We demonstrated that the genomic DNA was successfully extracted and could be used as templates to evaluate the specificity of primers. All 21 haplotype samples from Ephedra generated PCR products of 166 bp, showing a band between 100 and 250 bp, while samples from Pinellia and Typhonium group, Sida group did not generate an amplicon (Fig. 3). A fragment of 168 bp was amplified using the specific primers of Pinellia and Typhonium group, while samples from Ephedra and Sida showed no bands (Fig. 4). In the Sida group, the same result was found (Fig. 5), which illustrated three pairs of primers of three groups were specific.

Rapid and sensitive detection of plant materials containing ephedrine alkaloids using TaqMan qPCR
Based on the results of the specific primers (Table 1), we found that these primers can specifically amplify the DNA of three groups. On the basis, we designed the probes (Table 1) and conducted the Real-time PCR assays. All 11 Ephedra species included in this study showed normal fluorescence with the primers and probe, but no amplifications were Fig. 1 The NJ tree of plant materials containing ephedrine alkaloids constructed with ITS2 sequences observed with the primers and probe from the Pinellia and Typhonium group or the Sida group. Four species of Pinellia and Typhonium showed normal fluorescence amplification only when the primers and probe from the Pinellia and Typhonium group was used. The same is true for the Sida group. Results from the Real-time PCR showed that all haplotypes generated normal fluorescence amplification curves with CT < 35. The results of a series TaqMan qPCR assays on diluted genomic DNA of E. sinica showed that fluorescence  amplification curves were exponential at 0.0050 ng/ μL and the fluorescence signal was no longer normal at 0.0025 ng/μL. This indicated that the minimum amount of genomic DNA of E. sinica detected with the TaqMan qPCR was about 0.0050 ng/μL ( Figure S3).

The comparison of the ITS2 and psbA-trnH barcodes used for identification of ephedrine alkaloid containing species
An ideal barcode should be easily amplified and sequenced (Taberlet et al. 2007). In the present study, 175 ITS2 sequences and 136 psbA-trnH sequences from 201 specimens of 18 species were obtained, indicating that the utilized ITS2 primers are more universally applicable and can be utilized to trace botanical origin from ephedrine alkaloid containing plant species. This can be attributed to several reasons: (i) the length of ITS2 is much smaller than psbA-trnH, varying between 230 and 252 bp, which makes it easier to amplify and sequence; (ii) The efficiency of PCR amplification and sequencing of ITS2 is higher than psbA-trnH; and (iii) High-quality sequencing peak diagrams of psbA-trnH of Sida were difficult to obtain because of the existence of poly A or T sequence.
The psbA-trnH region is more applicable to distinguish Ephedra, as the entire region is highly conserved, and E. monosperma, E. equisetina, and E. fedtschenkoae individuals had unique insertions, which could be used for identification purposes. The stable single change at site 300 in E. monosperma could provide geographical information because the psbA-trnH sequences from E. intermedia, E. regeliana, and E. fedtschenkoae did not match the psbA-trnH sequences obtained from GenBank. However, the 12 base insertion or deletion could be due to geographical variation (Lahaye et al. 2008) or pseudogene amplification. For E. sinica, E. glauca, E. lomatolepis, E. distachya, E. intermedia var. tibetica, and E. przewalskii, both ITS2 and psbA-trnH sequences proved inappropriate to differentiate among species. A more informative barcode is needed to distinguish these species.
In this study, we did not experiment with easily mixed species that did not contain ephedrine. Because compared to ephedrine-containing species, their genetic distance and sequence are quite different. Our results indicated that two universal primer pairs, PA/TH and fwd/rev, showed no successful amplification and sequencing in P. cordata and T. flagelliforme, which suggested that more applicable primers need to be designed. A previous study documented that specific primers based on the difference of psbA-trnH in Ephedra can identify Ephedra DNA in dietary supplements using PCR (Techen et al. 2006). However, in our work, psbA-trnH cannot be amplified and successfully sequenced in several species. In addition, some species also had no referenced psbA-trnH sequences in GenBank. ITS2 stood out for its universality and accessibility. Therefore, we selected ITS2 to design primers and probes. The TaqMan assay was specific and sensitive enough to detect as little as 0.0050 ng/μL of DNA in our samples. In addition, under normal PCR conditions, results showed no smeared or invisible bands when visualized using electrophoretic analysis at ~ 0.0050 ng/μL of DNA.

The significance of using DNA barcoding and TaqMan probe technology
Although the supervision and administration of medical plant materials containing ephedrine alkaloids are strict, cases of smuggling are still prevalent. The identification of medical plant materials containing ephedrine alkaloids is of great importance not only for ensuring safety but for controlling reasonable market circulation. The present work developed a rapid, specific, and highly sensitive method to detect plant materials containing ephedrine alkaloids. We tested 18 species in this study, covering the most common medical species containing ephedrine alkaloids in China, and more than 90% of the ephedrine alkaloid containing plant species listed on the FDA Poisonous Plant Database and European Food Safety Authority. Because of our large sample size, we had a high success rate in identifying existing ephedrine-containing species. If new ephedrine-contained species appear, the false positive or false negative results might occur due to the lack of database information. In case of false negative result, we need to further sequence the new species and determine whether these primers and probes are suitable for them by comparing the sequence information with the current ephedrine-containing species. For the false positive result, we could further detect the content of ephedrine by chemical method to ensured the accuracy of result.

Conclusion
In this paper, we evaluated the feasibility of the ITS2 and psbA-trnH region as candidate barcodes to discriminate the 201 ephedrine alkaloids containing samples. ITS2 proved advantageous over psbA-trnH. Further, based on DNA barcoding, we developed a specific and sensitive TaqMan realtime PCR assay (as little as 0.0050 ng/μL target DNA) to distinguish plant species containing ephedrine alkaloids. Our study demonstrated that DNA barcoding along with TaqMan technology can be used to detect plant materials containing ephedrine alkaloids rapidly and sensitively. This study could be an assistant method to detect plant materials containing ephedrine alkaloids. In particular, it may also have general guiding implications in the examination of dietary supplements and ephedrine alkaloid containing powdered materials.
Author contribution statement YZ and HG analyzed the data and wrote the manuscript; MS, YL and JF collected the materials and performed the experiments; XL and YZ designed and guided experiments. All the authors agreed on the contents of the paper and have no conflicting interests to declare.