DNA Barcoding for the Identification and Authentication of Animal Species in Traditional Medicine

Animal-based traditional medicine not only plays a significant role in therapeutic practices worldwide but also provides a potential compound library for drug discovery. However, persistent hunting and illegal trade markedly threaten numerous medicinal animal species, and increasing demand further provokes the emergence of various adulterants. As the conventional methods are difficult and time-consuming to detect processed products or identify animal species with similar morphology, developing novel authentication methods for animal-based traditional medicine represents an urgent need. During the last decade, DNA barcoding offers an accurate and efficient strategy that can identify existing species and discover unknown species via analysis of sequence variation in a standardized region of DNA. Recent studies have shown that DNA barcoding as well as minibarcoding and metabarcoding is capable of identifying animal species and discriminating the authentics from the adulterants in various types of traditional medicines, including raw materials, processed products, and complex preparations. These techniques can also be used to detect the unlabelled and threatened animal species in traditional medicine. Here, we review the recent progress of DNA barcoding for the identification and authentication of animal species used in traditional medicine, which provides a reference for quality control and trade supervision of animal-based traditional medicine.


Introduction
Traditional medicine (TM) has been widely used for the prevention and treatment of various common ailments and complicated illnesses in human history, and the use of animals as medicine, which is known as zootherapy, has long been an essential part of traditional therapeutic practices, such as Traditional Chinese Medicine (TCM), Kampo medicine, Ayurvedic medicine, and American folk medicine [1][2][3][4][5][6]. Besides the traditional usages, modern clinical studies have demonstrated that animal-based TMs possess a number of pharmacological effects, including anti-inflammatory, antitumor, anti-infective, anticonvulsant, analgesic, and immunomodulatory activities [6][7][8][9]. Therefore, zootherapy continues to serve as an important complementary and alternative therapy in modern societies. For example, over 1,500 animal species have medicinal benefits according to the historical records in China, and approximately 77 kinds of medicinal animals and 50 kinds of medicinal materials derived from animal sources have been included in the Chinese Pharmacopoeia 2015 Edition [1,10]. In Brazil and Latin America, at least 354 and 584 animal species have been reported to be used in TM, respectively [4,5]. It is estimated that animal-derived TM has been increasingly used in many countries and its annual global trade accounts for billions of dollars [2]. Moreover, recent studies have revealed that animal-based TM contains a variety of bioactive compounds, which provides a valuable chemical library for drug discovery [11], indicating the potential of medicinal animals in developing modern pharmaceuticals.
During recent centuries, a wide range of wild animal species is known to be threatened with extinction; for 2 Evidence-Based Complementary and Alternative Medicine example, according to the latest statistics of the International Union for Conservation of Nature (IUCN) Red List of Threatened Species, the number of endangered and critically endangered animal species has reached 7597 and 5101, respectively [12]. Such rapid loss of biodiversity is closely associated with animal habitat loss and human overexploitation. For instance, despite some parts such as horns and secretions, almost all types of tissues and organs used in TM require animal sacrifice, which inflicts substantial stress to medicinal animal resources [2,13,14]. To solve the species crisis, the trade of threatened animals has been regulated under international legislation including the Convention on International Trade in Endangered Species (CITES), and the therapeutic application of such species has been forbidden in many countries. However, some countries such as those in East and Southeast Asia still utilize threatened animals for medicinal purposes [2,15], and the illegal trade of medicinal animals further impairs biodiversity [16,17]. Moreover, a variety of adulterants and counterfeits has emerged in the market for greater interests, which brings challenges to the safety of animal-based TM. Therefore, identification and authentication of animal species is of vital importance for TM trade supervision and quality control.
It is known that animal-based TMs are mainly derived from animal tissues, organs, or metabolites, and these parts are usually processed into diverse preparations such as slice, powder, and tablet, resulting in difficulties for morphologybased species identification. In addition, several commercial technologies used for species identification, including chromatography and immunoassay, are relatively costly and expertise-dependent, which is not suitable for bulk detection of TM preparations and forensic specimens [18,19]. Therefore, developing an accurate and efficient identification method represents an urgent need for animal-based TM. In recent decade, advance in molecular techniques has promoted the application of simple and precise DNA analysis in taxonomic field. Among the prevailing genome-based approaches, DNA barcoding provides a robust strategy to identify existing species and discover unknown species via comparative analysis of sequence variation [19][20][21]. Extensive studies have shown that DNA barcoding is capable of identifying a wide range of animal species, including mammals, birds, reptiles, amphibians, fishes, and crustaceans [21][22][23][24]. Moreover, many international organizations such as the Consortium for the Barcode of Life (CBOL) and the Barcode of Life Data System (BOLD) have been established to promote DNA barcoding as a global standard for species identification. In addition to taxonomy, DNA barcoding has recently been applied in various fields, including medicine and food science, forensics, and conservation biology [19,20,[25][26][27]. In the field of TM, DNA barcoding has been widely used to authenticate herbal sources [19,20]. However, studies on identifying medicinal animal species including the threatened species and discriminating the authentic TMs from the adulterants are relatively less addressed. Therefore, we tend to review the general process and current progress of DNA barcoding in analyzing animal-based TM and then discuss its limitations and potential strategies.

Animal Identification by DNA Barcoding
2.1. DNA Barcoding and Minibarcoding. The accurate species identification by DNA barcoding relies on a suitable DNA barcode, which refers to a standardized sequence (usually less than 1,000 bp) of the genome [21]. The barcode should have universality so that it can be easily amplified from diverse species, and it should contain few insertions or deletions to facilitate sequence alignment. Also, its mutation rate must be sufficient to generate a barcoding gap, which means the maximum intraspecific variation is less than the minimum interspecific distance [21]. For animal identification, the most broadly used barcode marker is mitochondrial cytochrome c oxidase subunit I (COI), which is highly conserved across species employing oxidative phosphorylation for metabolism [22]. Numerous studies have shown that COI-based DNA barcoding can delimit diverse animal species, indicating the high rates of sequence change at species level and constraints on intraspecific divergence in COI sequence [21][22][23][24]. For analyzing fresh and well-preserved animal tissues, a full-length barcode such as a 658-bp region of COI gene is recommended as its PCR amplification and sequencing can be feasible. However, animal tissues are usually processed before used as TM materials, and some processing approaches, such as sundrying, stir-frying, and boiling, can cause DNA degradation, leading to difficulties for PCR amplification of full-length barcodes [19]. Since the amplification feasibility improves with reduced sequence length, a minibarcoding method, which utilizes a shorter region within the standard barcode, has received increasing attention [28]. Recent studies have shown that an array of minibarcodes can be effectively amplified and sequenced from various processed products including TMs and foods and provide sufficient sequence information necessary for species identification [26,29,30].

DNA Metabarcoding.
It is known that many TM preparations are composed of multiple materials derived from diverse animal and/or plant species. In such mixtures, multiple barcode sequences will be coamplified and the conventional sequencing will generate multiple or overlaying sequencing peaks, resulting in ambiguity or false sequence information [31]. Therefore, DNA metabarcoding has been proposed to identify multiple species within complex samples using next-generation sequencing (NGS) technology [32,33]. NGS can rapidly yield millions of DNA reads and obtain all representative sequences presented in mixtures, which facilitates a high-throughput multitaxa identification. This technology can also be used for sequencing minibarcodes (approximately 50-400 bp), which makes it suitable for analyzing processed samples with degraded DNA [19]. For example, Arulandhu et al. [34] have shown that a multilocus DNA metabarcoding, including a set of full-length barcodes and minibarcodes combined with Illumina MiSeq amplicon sequencing, can reproducibly identify animal and plant species present in mixtures with both low and high complexity at 1% dry weight content, demonstrating its accuracy and sensitivity in analyzing complex samples. Due to these advantages, DNA metabarcoding has been widely applied in biodiversity research, ecological management, and community analysis [33,35]. Recently, it has also been used to identify animal species in medicinal preparations, foods, and forensic specimens [34,[36][37][38]. For instance, Coghlan et al. [36] used mitochondrial 16S ribosomal RNA (16S rRNA) barcode and NGS technology to detect species in 15 complex TCMs presented in the form of powders, capsules, and tablets and revealed that some of them contained CITES-listed animal species and unlabeled species, demonstrating the capability of DNA metabarcoding in analyzing TM formulae.

General Process of Animal Identification by DNA Barcoding
Although diverse animal tissues and animal-based TMs are used in different studies, the main procedures of DNA barcoding are similar. After extraction of DNA from the tested samples, appropriate DNA barcodes must be selected and then amplified via the polymerase chain reaction (PCR). The amplified regions can be sequenced by conventional methods or NGS according to the sample complexity and then matched to existing sequences from reference database or voucher specimens. Further comparative analysis can be performed to detect the intraspecific and interspecific sequence divergences (Figure 1).

DNA Extraction.
Extraction of DNA with high yield and quality is a crucial premise for DNA barcoding of animal species. For some types of animal tissues that contain a small amount of DNA, appropriate sampling and sufficient homogenization are important before DNA extraction. The extraction protocol usually involves several critical steps, including tissue lysis, removal of impurities from DNA, and DNA precipitation [39]. There are many classical methods for extracting out DNA from animal tissues, such as Sodium Dodecyl Sulfate (SDS) extraction, guanidinium thiocyanatephenol-chloroform extraction, silica matrix-based purification, and magnetic bead-based purification [40]. In many laboratories, these conventional methods have been modified to improve DNA extraction efficiency; for example, Ivanova et al. [41] developed an inexpensive and automationfriendly animal DNA extraction protocol, which employed SDS and proteinase K for tissue lysis followed by silicabased purification using glass fiber filtration plates. In addition, most of these protocols have been developed into commercial kits, which provide standard DNA extraction means for different studies. As different types of animal tissues have distinct characteristics, it is necessary to select appropriate extraction methods or commercial kits (Table 1).

4
Evidence-Based Complementary and Alternative Medicine  [38,42]. For instance, COI barcoding region has been found to offer insufficient and unreliable discrimination for some species in the class Gastropod and Anthozoa [43,44]. In fact, several other genes, such as mitochondrial cytochrome b (Cytb), 16S rRNA, 12S ribosomal RNA (12S rRNA), and nuclear ribosomal internal transcribed spacer (ITS), have also been used for barcoding of animal species [42,45,46]. For example, the ITS2 region can achieve an identification rate of 91.7% at the species level among 12,221 kinds of animals recorded in GenBank and differentiate some animal species such as the Argasidae that can be not identified by COI barcode [45]. On the other hand, minibarcodes usually exhibit higher success rate of amplification than full-length barcodes in highly processed samples. However, since the success of taxonomic differentiation is positively correlated with the barcode length, the minibarcode length is usually kept above 100 bp. For example, an approximately 250-bp region of 16S rRNA can be successfully amplified from various medicinal preparations and food products and further provides correct identification of animal species [34,36,47]. In addition, as a single DNA barcode may generate insufficient or false identification of hybrid species as well as animal species with high diversity [48], multilocus barcodes can be used to improve the identification accuracy and sensitivity. The DNA barcoding guideline for molecular identification of TCM, which is included in the Appendix of Chinese Pharmacopoeia, has proposed a comprehensive animal identification using both COI and ITS2 barcodes [10].

PCR Amplification of Barcode Region.
The amplification efficiency of the barcode region is closely associated with the primer pairs; for example, appropriate barcode primers should be versatile across a wide range of animal species and have high affinity to DNA templates and a balanced melting temperature [49]. Ivanova et al. [50] have developed universal primer sets for amplifying COI barcode. When the universal primers are not applicable for certain taxa or specimens, it is necessary to redesign primers, such as those for minibarcodes [19,47]. Many molecular biology software and related websites, including Primer Premier, Oligo, and Whitehead, can be used to design and evaluate primers. In addition to the primers, PCR reaction system also contains other necessities including heat-stable DNA polymerases, dNTP mixtures, and DNA templates. The reaction parameters, such as the temperature and time of melting and annealing processes, play an important role in the selective amplification of target templates, and these parameters should be optimized according to the specific circumstances [51].

Sequencing of Amplified Region.
As a widely used DNA sequencing method, Sanger dideoxy sequencing is capable of generating sequencing reads up to 1,000 bp [52]. However, Sanger sequencing is low throughput, which makes it suitable for DNA barcoding of single species at a small scale [31,53]. In contrast, NGS technology can parallelly sequencing multiple DNA fragments in a single reaction [38]. The 454 pyrosequencing is the first commercially available NGS and has been used to analyze various types of mixtures, including environmental specimens, food products, and medicinal preparations [36,38,54,55] . These databases contain a variety of sequences assigned to corresponding taxa, which is useful for comparative analysis of sequence variations. For example, GenBank, a commonly used database in barcoding studies, has included more than 1 Terabase sequence data with relatively broad taxon coverage [115]. Another database BOLD has already collected more than 2 million COI sequences from about 170,000 species, and INSDC has recorded extensive Cytb and ITS sequence information [38].
In addition, MMDBD focuses on the barcode information of medicinal plants and animals (over 1,700 species) listed in the Chinese Pharmacopoeia, American Herbal Pharmacopoeia, and other related references. Interestingly, MMDBD also includes sequence information of common adulterants and substitutes [116].
3.6. Sequence Analysis. Several comparative methods, including similarity-, distance-, and tree-based approaches, have been widely used to analyze sequence variations [38,117]. The Basic Local Alignment Search Tool (BLAST) is a similarity-based algorithm, which matches the query sequence to those in reference databases and then provides a similarity score according to the portion of the query aligned to the reference [118]. For distance-based analysis, Kimura 2parameter (K2P) model can be used to calculate the intraspecific and interspecific genetic distances among sequences, and then the barcoding gap can be used for species delimitation [117,119]. In addition, tree-based methods are often used to establish the phylogenetic relationships, which assign query sequences to species on the basis of their membership of clusters in a barcode tree. The closest relative animal species will appear in a cluster, while distinct species should form discreet clusters. Several hierarchical clustering algorithms, including neighbor joining (NJ), maximum likelihood (ML), maximum parsimony (MP), and Bayesian inference (BI), have been used to establish phylogenetic tree, and a combination of these algorithms can provide more reliable identification as compared with a single algorithm [22,117,120,121]. A number of commercial tools, such as MEGA, PHYLIP, and PAUP, can be used for tree construction and visualization.

Authentication of Animal-Based Traditional Medicine by DNA Barcoding
It is estimated that the animals used in TM are mainly from several phyla, including Chordata, Arthropoda, Echinodermata, Annelida, Mollusca, and Coelenterata. The authentication of genetic composition is important for quality control of animal-based TM and trade regulation of threatened medicinal animals. Recently, DNA barcoding as well as minibarcoding and metabarcoding has been successfully applied to identify single ( Table 2) and multiple species (Table 3) in various types of animal-based TMs and discriminate the authentics from the adulterants [34,36,.

Identification of Single Species
The phylum Chordata used in TM mainly includes mammals, reptiles, fishes, and amphibians. Among these vertebrates, many kinds of wild mammals as well as their organs and tissues, such as horns, scales, muscles, and gallbladders, have long been an important source of TM materials in many countries, especially those in Asia and Africa [2,122]. For example, the horns from a migratory ungulate Saiga tatarica have been used in TCM for thousands of years, but the wild population of S. tatarica has rapidly declined in recent decades due to persistent hunting [57,58]. CITES has listed this antelope in Appendix II since 1995, and the market trade of Saiga horns has been rigorously monitored. However, the horns from other species such as Capra hircus and Procapra gutturosa have been sold as Saiga horns in the market [57]. It is difficult to discriminate these horns as they share similar appearance, especially when they are processed in slices or powders. To distinguish the authentics from the adulterants, Chen et al. [57] recovered a 644 bp region of COI gene from well-preserved horns with specific primers and a 349 bp fragment from degraded horn samples with nested primers. Further analysis using K2P model and NJ tree method revealed that the mean intraspecific genetic distances of both full-length barcode and minibarcode are far less than the mean interspecific distances, and S. tatarica and its adulterant species can form independent clades in phylogenetic tree. Another famous horn-based TCM is Pilos antler, which is used for many illnesses including impotence, arthritis, and anemia [123]. It is officially derived from velvet antlers of Cervus nippon and Cervus elaphus according to the Chinese Pharmacopoeia [10], while the antlers of other deer such as Rangifer tarandus and Dama dama are often sold as the genuine products [62]. Recently, several studies have shown that using a set of DNA barcodes, including COI, Cytb, and 16S rRNA, can identify the biological origins of various animal horn samples and discriminate Pilos antler from its adulterants with high sensitivity [58,62,64]. Together, these results demonstrate that DNA barcoding and minibarcoding are capable of authenticating horn-based TM. Moreover, barcoding techniques have been used to analyze TM derived from other mammalian tissues, such as pangolin scales, deer musk, bear bile, and donkey-hide gelatin [36,[65][66][67][68][69][70]. For instance, pangolin scales, a rare TM used for many conditions including asthma and rheumatism, are derived from Manis pentadactyla according to the Chinese Pharmacopoeia [10]. However, the scales of other pangolin species often cause market confusion due to the scarcity of effective detection methods, and the illegal pangolin trade has escalated globally in recent years despite the legislation that all pangolin species have been listed in CITES Appendix II since 1994.
To identify several batches of confiscated pangolin remains in Philippines and Hong Kong, several studies used COI and Cytb barcodes and found that barcoding approach can accurately assign unknown scales to specific species and distinguish different pangolin species [66,67], indicating that DNA barcoding is a useful strategy for customs to suppress illegal trade of threatened mammals. It is reported that many reptiles, including snakes, geckos, and turtles, play an important role in traditional folk medicine worldwide, and these reptile-based TMs have various therapeutic benefits, such as anti-inflammatory, sedative, and analgesic effects [124]. To identify the medicinal reptiles that exhibit similar morphology interspecifically, DNA barcoding provides a simple but reliable strategy ( Table 2). For example, Zaocys dhumnades and Bungarus multicinctus are two important snakes used in TCM, and their dried bodies without the viscera have been widely applied to treat several disorders including rheumatoid arthritis, stroke, and convulsion [73][74][75][76][77]. However, many other snake species are marked as Z. dhumnades and B. multicinctus in the market, and the accurate identification of these snake-based TMs highly relies on professional experience. Interestingly, several recent studies have used a panel of full-length barcodes and minibarcodes including COI, 12S rRNA, 16S rRNA, and Cytb to analyze various snake specimens and related TMs collected from the wild and markets in China. The results showed that each sample can be identified as specific snake species, and Z. dhumnades and B. multicinctus as well as their adulterants can be clearly distinguished at the species level [73][74][75][76][77], indicating the efficacy of DNA barcoding and minibarcoding in authentication of snake-based TM. Another example is the dried body of Gekko gecko, which is traditionally used for relieving coughing and asthma [72]. Using 150-bp and 648-bp fragments of COI sequence, G. gecko specimens and their adulterants such as G. japonicas and Calotes versicolor have been found to exhibit significant barcoding gaps [72], further demonstrating the capacity of barcoding techniques in analyzing reptile-based TM.
As an indispensable group of vertebrates in aquatic animals, fish provides an enormous resource for humans as foods and medicines. Modern research has revealed that many fish species, especially marine fish, contain various substances with nutritional and pharmacological benefits [125]. For example, shark fins are not only the main constituent of the delicacy shark fin soup in Asian cuisines but also the precious TCM materials used for arthritis treatment and immune enhancement [97,126]. Although twelve shark species have been listed in Appendix II of CITES, overfishing and illegal trade continue to induce a rapid decline of shark populations. In response to this issue, DNA barcoding has recently been adopted to authenticate shark species in the market [94][95][96][97][98]. For processed shark products, degraded Evidence-Based Complementary and Alternative Medicine 7 Janjua et al. [69] Peppin et al. [70] 8 Evidence-Based Complementary and Alternative Medicine  10 Evidence-Based Complementary and Alternative Medicine  Evidence-Based Complementary and Alternative Medicine genomic DNA may lead to unsuccessful amplification of full-length barcodes. Therefore, several studies used shorter fragments of COI (less than 200 bp) and showed that these minibarcodes can accurately identify CITES-listed shark species from processed fins and even fin soup and skin-care products [97,98], demonstrating the capability of DNA minibarcoding in analyzing processed animal tissues. Another example is Hippocampus spp., which has a unique appearance in the family Syngnathidae and is a famous TM material used to treat impotence, asthma, and insomnia [84,85]. As a mass of seahorses have been harvested and traded annually, all seahorse species have been included in Appendix II of CITES since 2004. To investigate the usage of seahorses in Chinese TM market, several recent studies employed DNA barcoding to analyze dried seashore specimens and revealed that both COI and Cytb barcodes can efficiently authenticate seahorse species and identify endangered species [84,85]. Together, these studies demonstrate the availability of DNA barcoding in trade supervision of threatened fish species used in TM.
Amphibians are a group of ectothermic vertebrates characterized by their ability to exploit both aquatic and terrestrial habitats. Some of them, such as toads, frogs, and salamanders, are traditionally used to treat a number of ailments in many countries, including China, Japan, South Korea, and Spanish [127,128]. For example, Rana temporaria and Bufo gargarizans are two common amphibians used in TM. The dried body and oil derived from R. temporaria are traditionally used to relieve cough and asthma, while the toad cake and skin of B. gargarizans have been used for treating heart diseases, skin ailments, and other systemic illnesses [10,129]. However, other frog and toad species are often counterfeited as the authentics in the market. Recently, several studies have shown that distinct frogs and toads including R. temporaria and B. gargarizans can be clearly distinguished using DNA barcoding approach based on COI and 16S rRNA sequences [99,100], which provides a reference for authentication of amphibian-based TM.

Other Phyla-Derived Traditional Medicine.
A variety of animal species in the phylum Arthropoda have been reported to be used in traditional therapeutic practices worldwide [130]. For example, the dried larvae of Chrysomya megacephala are traditionally used to treat malnutrition and skin and soft tissue wounds [101]. To discriminate this blowfly from other Diptera, a recent study used a full-length COI-based DNA barcoding and found that C. megacephala exhibited a complete nonoverlapping barcode divergence with other flies including the closest relative species C. pinguis [101], demonstrating the ability of DNA barcoding in authentication of insect-based TM. Moreover, barcoding technique has been applied to identify other medicinal animals in the phylum Arthropoda, such as scorpions and crabs [73,106]. Interestingly, some animals have symbiotic associations with fungi, and they can be used together as TM materials. For the authentication of fungal species in such TMs, it is noteworthy that COI barcode cannot provide sufficient discrimination, while ITS barcode exhibits significant interspecific divergence [131]. For instance, Cordyceps sinensis, a well-documented tonic TCM widely used in Asia, is a combination of an entomopathogenic fungus and the larvae of the Hepialidae such as Hepialus armoricanus [104]. A number of close relative species such as C. gunnii and C. cicadae are commonly sold as C. sinensis, which seriously reduces the therapeutic efficacy or even leads to poisoning. To distinguish C. sinensis from counterfeit species, several recent studies have designed species-specific primers to amplify certain segments of ITS sequence and found significant barcoding gaps among these Cordyceps species [102][103][104]. Therefore, a combination of COI and ITS barcodes can be feasible to detect TM containing both insects and fungi. In addition to Arthropoda, a number of species in other phyla, including Echinodermata, Annelida, Mollusca, and Coelenterata, also provide important medicinal resources for traditional therapeutics [132][133][134][135]. Although DNA barcoding is relatively less addressed in TMs derived from these phyla, limited evidence has shown that it is still capable of identifying several species, including sea cucumbers, earthworms, oysters, and corals [108][109][110][111][112][113]. Together, these studies demonstrate that DNA barcoding and minibarcoding are accurate and efficient approaches to authenticate TM derived from a wide range of animal species.

Identification of Multiple Species.
Recent concerns about the safety and legality of TM have prompted more rigorous surveillance. Interestingly, several studies have shown that DNA metabarcoding is capable of authenticating labeled species and detecting undeclared taxa in animal-based TM formulae (Table 3). For example, using 16S rRNA barcode combined with Roche GS Junior sequencing, half of the TCM preparations legally purchased in South Australia were found to contain DNA from undeclared animal or plant taxa [114]. Another recent study has developed a multilocus metabarcoding approach that employs 12 DNA barcode markers and Illumina MiSeq amplicon sequencing and revealed that Ma pak leung sea-dog hard capsules and Cobra performance enhancer hard capsules, both of which are used to treat sexual weakness, contain DNA from nondeclared taxa such as Bos taurus and Homo sapiens instead of labeled species [34]. Together, these studies suggest that metabarcoding can provide a pharmacovigilance measure for pre-and postmarket auditing of TM. In addition, metabarcoding has recently been used to identify threatened animal species in a variety of complex samples including TM preparations (Table 3). For instance, Coghlan et al. [114] have shown that 16S rRNAbased DNA metabarcoding can detect DNA from endangered animals Panthera uncia and possibly Panthera tigris in a TCM sample used to treat arthritis and pain. Moreover, similar technique was used to audit the genetic composition of some TCM samples seized by Australian Customs and Border Protection Service, and the results showed that Saiga antelope horn powder contained DNA from the known CITES-listed species Saiga tatarica and other species including goat and sheep, and Chu pak hou tsao san powder contained DNA from Ursus thibetanus, which is recorded in both CITES Appendix I and IUCN Red List [36]. These results suggest that DNA metabarcoding is useful for custom authority to analyze forensic specimens. Interestingly, DNA metabarcoding has also been used to identify the diet composition from animal Evidence-Based Complementary and Alternative Medicine 13 secretions, such as the floral species of honey [136], further demonstrating its ability in biodiversity analysis.

Limitation and Prospect
Although DNA barcoding is an effective complement to conventional identification methods, it still has a few shortcomings [42,137]; for example, some animal tissues such as horns, shells, and scales contain relatively small amount of DNA, resulting in difficulties for template amplification. Moreover, a false identification may be generated due to the contamination in DNA extraction and PCR reaction processes. Thus it is important to sample the tissues containing more cells and make sure all procedures are standardized. For example, to obtain a high yield and quality of DNA from animal horns, 75% ethanol can be used for sterilization, and middle layer between the bone core and outer sheath can be collected and then ground into powder in liquid nitrogen [57]. On the other hand, some TM preparations such as the extracts are highly processed, inducing degradation of DNA into very small fragments or even complete removal of DNA. In such cases, it is preferred to perform DNA barcoding analysis before the raw materials are processed. Another concern for complex samples with degraded DNA is that the amplification success of a barcode region may be different for distinct species due to varied gene copy number [20,31]. It is thus necessary to employ multilocus barcoding and minibarcoding approaches and design novel primers for certain taxa.
The feasibility and accuracy of DNA barcoding are closely associated with both reference sequence data and taxonomically confirmed specimens. Although GenBank has included extensive COI sequences in a wide range of animal species, the information of some species used in TM is still lacking, and the sequence inventory of other barcodes such as 16S rRNA and Cytb also needs to be extended [38,42]. Moreover, it is important to improve the availability of professionally authentic vouchers in public DNA databases as the misidentified species will generate incorrect sequence information [138,139]. In addition, although MMDBD includes thousands of TM materials, the listed animal species and their adulterants are still insufficient, and the animal species used in other TM systems are also lacking. Therefore, more related databases should be established to provide sufficient bases for DNA barcoding analysis of animal-derived TM.
The quality of animal-based TM is known to be related to multiple factors, such as the specified tissues, the effective substances and even the growth stage. DNA-based analysis is feasible for genetic authenticity but unable to evaluate the pharmacological effects of TMs [19]; for example, DNA barcoding cannot distinguish different tissues from the same animal or determine the content of bioactive substances. It also fails to detect animal growth stages while some TM materials require animals at specific stages. Moreover, DNA barcoding is not feasible to identify the adulterants that do not contain DNA. To overcome these restrictions, it is necessary to combine genetic techniques with conventional approaches such as chromatography and metabolomics. Interestingly, a recent study used multidisciplinary techniques, including NGS, high performance liquid chromatography, and mass spectrometer, and showed that some TCM samples not only contained unlabelled animal species but also had undeclared pharmaceutical agents and excess heavy metals [114], indicating the potential of comprehensive analysis system in evaluating TM quality and reducing market fraud.

Conclusion
DNA barcoding offers a reliable and efficient strategy for the identification of authentic animal species and their adulterants in TM. It is noteworthy that the success of DNA barcoding is related to many factors, such as high quality of DNA and appropriate barcodes. For processed animal tissues with degraded DNA, minibarcodes usually exhibit higher success rate in species identification as compared with fulllength barcodes. For complex mixtures such as TM formulae, metabarcoding provides a feasible approach to simultaneously detect multiple animal ingredients. With a global accumulation of open access reference sequences, DNA barcoding gradually becomes an authoritative approach in TM authentication. Despite these advantages, DNA barcoding still has several inherent limitations, such as inability to identify medicinal parts or determine compounds with pharmacological activities. Therefore, establishing a comprehensive identification system including barcoding and other techniques will provide more information for quality assessment and trade monitor of animal-based TM.