接第二篇

3. Discussion

　　The identification of authentic species has become a crucial issue for the quality control of food and medicinal sources, because most of these materials are collected from wild habitats or are cultivated on farms [34]. Since the accurate identification of fungal species is very difficult, depending on the conventional method used, more reliable and objective methods are required in order to discriminate between strains at the species level [3]. In this study, we developed a simple PCR assay method, and confirmed that the comparative analysis of nrDNA-ITS sequences is one of the most reliable tools for overcoming these difficulties. To verify the accurate determination of species in the samples listed in Table 1, we carried out BLAST searches, compared entire nrDNA-ITS sequences, and constructed phylogenetic trees. From the BLAST searches, we confirmed that four fungal samples had been misidentified (Table 1). We also further confirmed the species identification results by an analysis of sequences and phylogenetic trees (Figures S1 and S2). A phylogenetic tree of Cordyceps and five related fungal species was constructed from 31 nrDNA-ITS sequences, including 18 nrDNA-ITS sequences that were obtained in the current study and 13 retrieved from the NCBI GenBank, based on the neighbor-joining method (Table 1, and Materials and Methods). The neighbor-joining analysis classified the 31 nrDNA-ITS sequences into six distinct clades with over 70% bootstrap values(C. militaris, C. pruinosa, I. cicadae, I. tenuipes, O. sinensis, and B. bassiana), and four genus cluster groups(Cordyceps, Isaria, Ophiocordyceps, and Beauveria) (Figure S1). These observations strongly confirmed that the identification results were correctly verified, and that the nrDNA-ITS sequences could be used to distinguish Cordyceps and its five related taxa to the genus level, with the species showing a phylogenetic relationship similar to relationship reported earlier [35]. Further, the phylogenetic analysis confirmed that these six fungal species, which are used as important food and medicinal ingredients, may indeed be distinguished based on nrDNA-ITS sequence divergence.

　　During the development of species-specific SCAR markers, the specificity of the primer sequence is the most important, because nucleotide substitutions and/or indels play a crucial role for primer specificity [36]. To identify the optimal species-specific SCAR primers, the entire nrDNA-ITS sequences were analyzed, comparing the positions of nucleotide substitutions and indels between the species. While species-specific nucleotide substitutions and indels suitable for SCAR primer design were most numerous in the ITS1 and ITS2 regions of C. militaris, C. pruinosa, and O. sinensis, only several suitable positions were identified for I. cicadae and I. tenuipes (Figure S1). Therefore, both types of ITS regions were considered during the identification of candidate SCAR primers to distinguish Cordyceps and its related fungal species. Species-specific primers were designed based on these candidate regions,and their specificities for their respective target templates and species were verified using 16 samples,listed in Table 1. This led to the identification of two species-specific SCAR primer sets for each species, to increase the discriminability and stability of the molecular authentication method. Using these primers as a starting point, two SCAR markers were developed for each species, which yielded different sized PCR amplicons, with the expected sizes being obtained only during the analysis of the target species (Figure 1 and Table 3).

　　In addition, to verify the specificity of SCAR primers, the sequences of a total of 55 different species belonging to Ophiocorcyceps and closely related species—which were comprised of six species that were obtained in the current study, and 49 species that had been published in previous reports—were downloaded from GenBank, and the respective primer regions were compared with ClustalW(Figure S4) [3,30–32]. As a result, all primers had enough species-specific nucleotide substitutions in both the forward and reverse primer regions to provide primer specificity, with the exception of several species. The sequences of C. roseostromata in the CM F3 and CM R1 regions, and the sequences of I. japonica in the IT F4 and IT R3 also had one or two species-specific nucleotide substitutions. Thus, all SCAR primers that were developed in the current study included at least two species-specific nucleotide substitutions in one of the forward or reverse primer regions. These results also strongly supported the ability of the SCAR markers developed in this study to differentiate the five fungal species, and to discriminate between each precise species.

　　The detection of adulterants or contaminants is very important for the safety and quality control of food and medicinal materials [15]. In a previous report, an O. sinensis-specific SCAR marker was used to yield a PCR product from 8 ng of O. sinensis template gDNA in a conventional PCR assay; i.e., the sensitivity of that assay was inferior to the sensitivity of the current assay. However, the detection limits of the other species were not determined in order to check the quantity of adulteration and/or contamination [3]. In this study, we determined the detection limits for each species based on both conventional and real-time assays. Hence, the SCAR markers developed in the current study could be used to detect less than 0.1% (at least 0.0001%) contamination in a conventional PCR assay. Consequently, the assay might be employed for the purity assessment of herbal medicines and diverse food ingredients related to Cordyceps. These results indicated that a real-time PCR assay with SCAR markers was able to distinguish between the five fungal species with high sensitivity. Moreover, the assay may constitute a very efficient tool, not only for the identification of authentic C. militaris, I. tenuipes, and O. sinensis species and their closely related adulterants at the species level, but also for the detection of less than 0.01% contamination with adulterants or other species.