Forensically informative nucleotide sequencing (FINS) for the authentication of Chinese medicinal materials

DNA extraction refers to an invasive method that extracts DNA from tissues and cells via physical disruption and/or chemical fractionation. Cetyl trimethylammonium bromide (CTAB) and phenol/chloroform extraction [10] was employed by a number of commercial kits for DNA extraction. For example, DNA from highly processed Chinese medicinal materials, such as the mule skin extract Asini Corii Colla (Ejiao) [11].

DNA release, a non-invasive method that allows DNA to release from a sample into a solution without destruction, is particularly useful for obtaining DNA from important voucher specimens. DNA release is also used to investigate samples by analyzing the environmental DNA or the preservative, as demonstrated by recent studies of DNA detection from the water in which frogs (Rana catesbeiana) live and from worms (Hypopta agavis) preserved in 95% ethanol [12, 13]. The quantity and quality of the obtained DNA is a major concern with this method. While purification may be achieved by commercially available kits, the yield of DNA is quite minute and should be stored in safe conditions (freezing, cyanide and ethanol), certain chemicals that can damage DNA, such as ethyl acetate or formaldehyde, should be avoided [14].

Usually, only a small amount of DNA can be extracted or released from highly processed or improperly stored Chinese medicinal materials. Polymerase chain reaction (PCR) can produce a sufficient amount of a specific DNA fragment obtained from a tiny amount of DNA extract. The selection of a DNA region for amplification is one of the crucial factors for FINS because the resolution of FINS depends heavily on the variability and the number of informative sites of the DNA sequences of the tested samples and reference materials. As the evolutionary rates of different DNA regions vary, DNA regions with sufficient variability are essential for providing a high resolution result. Rapidly evolving regions among taxonomic groups can be used for the identification at the genus or species level. Slowly evolving regions among groups can be used to differentiate at the section or family level. An ideal DNA region for identifying Chinese medicinal materials should have high inter-specific variation but low intra-specific variation and have sufficient informative polymorphic sites to allow differential sequence alignment among the samples and the reference species. The evolutionary rate of the same DNA region may vary among animals, plants and fungi. For example, mitochondrial cytochrome c oxidase subunit 1 (COI) is suitable for the identification of specific animal species [15]; however, it is not suitable for most plants as few polymorphic sites are found across the 1.4 kb COI sequences [16], probably due to the slow mutation rate [17]. Thus, prior knowledge of the evolutionary rates of various DNA regions facilitates the selection of an appropriate DNA region. In the past few years, short DNA sequences for global barcoding of species have been proposed [15]. For example, the DNA barcodes for animals is COI and for fungi is ITS; the core DNA barcodes for plants are chloroplast large subunit of ribulose-bisphosphate carboxylase gene (rbcL) and chloroplast maturase K coding region (matK), while chloroplast trnH-psbA intergenic spacer (trnH-psbA) and nuclear internal transcribed spacer (ITS) are supplementary DNA barcodes for plants [15, 18, 19]. Recent studies suggested that ITS should be incorporated into the core DNA barcode for seed plants [20‐22]. These DNA barcodes have also been commonly applied to identify medicinal materials and should be considered as the primary DNA target region for FINS [23]. Chinese medicinal materials are often dried or processed, which may affect the quality and quantity of the extractable DNA. A shorter DNA region should be considered for samples with degraded DNA. The universal primers for PCR amplification of some commonly used regions in FINS are listed in Table 1.

DNA sequencing is the most direct approach to obtaining maximum genetic information of the amplified DNA regions. With significantly lowered costs and time, DNA sequencing is now routinely used to identify medicinal materials. The amplified and purified DNA fragments may be sequenced directly; however, molecular cloning may be applied in some cases. Cloning is required if (1) some DNA regions (e.g. ITS and 5S rRNA gene spacer) have non-homogenous multiple copies or secondary structures [24, 25]; (2) non-specific PCR amplification generates multiple amplicons of similar size; (3) there is simultaneous amplification of DNA from samples and fungal contaminants (e.g. due to improper storage) and (4) there is poly-A/T structure (e.g. in trnH-psbA) interfering with the DNA sequencing [26].

A sequence database is necessary because a successful application of FINS relies on the comparison of DNA sequences among the samples and reference species. Phylogenetic analyses of many taxa using various DNA regions have been performed, providing useful reference for FINS. Our group has recently constructed an online Medicinal Materials DNA Barcoding Database http://​www.​cuhk.​edu.​hk/​icm/​mmdbd.​htm, which contains over 20,000 DNA sequences of 1,300 medicinal species found in the Pharmacopoeia of the People's Republic of China and the United States Pharmacopoeia [27]. DNA sequences can also be found in the open access NCBI GenBank http://​www.​ncbi.​nlm.​nih.​gov/​genbank, EMBL Nucleotide Sequence Database http://​www.​ebi.​ac.​uk/​embl as well as the Chloroplast Genome Database http://​chloroplast.​cbio.​psu.​edu. With the vast amount of sequence data, it is possible to roughly identify any unknown sample even if the sequence of its source species is not yet available. However, the quality of publicly available DNA sequences could sometimes be incorrect or derived from wrongly identified species [6, 28, 29]. Generation of tailor-made reference sequences is essential if the concerned reference sequences do not exist and high resolution identification is required.

The original idea of FINS is to perform phylogenetic analysis of unknown samples together with the reference species to trace their source origin [4], which is different from molecular identification based solely on multiple sequence alignment and comparison of polymorphic sites. FINS emphasizes the use of phylogenetic analysis to identify species via phylograms [4]. In general, phylogenetic analysis carefully selects sequence alignment to find the informative homologous sites for subsequent analysis. Phylogenetic trees are then constructed using tree construction methods, such as maximum parsimony (MP), maximum likelihood (ML) and Bayesian analysis, to reflect the evolutionary history of the concerned taxa. Available computer programs for constructing multiple sequence alignment and phylogenetic analysis are Align-M, ClustalW, BioEdit, PAUP and MEGA [30].

To identify medicinal materials, FINS users would rather identify a sample than its phylogenetic relationship with the reference species. The topology of the phylogram is the major concern and the phylogenetic relationship among the reference species is less focused in FINS identification. It was suggested that DNA distance-based methods are preferred to phylogeny-based methods in the application of FINS for identification [31]. The DNA distance-based method provides similarities between the reference and the unknown species whereas the phylogeny-based method explores the evolutionary history of the species. The major difference between these two methods lies in the way that the DNA sequences are analyzed. Phylogenetic relationship analysis, such as maximum parsimony and maximum likelihood, uses a matrix of discrete phylogenetic informative characters or statistical models to infer the optimal phylogenetic trees of selected taxa. Distance-matrix methods, such as unweighed pair-group mean analysis (UPGMA) and neighbor-joining (NJ), calculate the genetic distance from multiple sequence alignments to determine the similarities among reference sequences. A cladogram is then constructed based on the pair-wise distance values to build up the relationship of similarity. The distance-matrix methods are simple to implement and do not invoke any evolutionary indications because similar looking species may not necessarily be phylogenetically related (i.e. convergent evolution).

Over 800 medicinal species are officially recorded in the Pharmacopoeia of the People's Republic of China [32]. Some of these Chinese medicinal materials are economically important and ecologically valuable, such as Dendrobii Caulis (Shihu), while some others are highly toxic, such as Aristolochiae Fructus (Madouling) and Radix Tripterygii Wilfordii (Leigongteng). FINS is one of the most definitive methods to ensure they are used safely and to protect consumers from adulteration. Over the years, FINS has been used to identify economically important materials and ecologically valuable species, as well as toxic and commonly used Chinese medicinal materials. Examples of the identification of these Chinese medicinal materials using FINS are given in Table 2.