The radix of Glehnia littoralis (beishashen) is used as an antitussive, mucolytic, antibacterial, antiphlogistic and immune response enhancer in Chinese medicine (CM) [1]. It is also used as a diaphoretic, antipyretic and analgesic in Japan [2]. Glehniae Radix is listed in the Japanese and Chinese Pharmacopoeia and is widely recognized as a nutritional and healthy food [3–6].

Due to the high market demand, overexploitation of Glehniae Radix already threatened the existence of this wild species. And it was protected now by China Plants Red Data Book [7]. However, the scarcity of this wild species has resulted in frequent fraudulent adulteration and substitution of G. littoralis with the species Chuanminshen violaceum Sheh et Shan (Chuanmingshen), Changium smyrnioides Wolff (Mingdangshen), Sphallerocarpus gracilis (Bess.) K.-Pol. (Miguoqin), Adenophora polyantha Nakai (Shishashen) and Silene tatarinowii Regel (Shishengyingzicao), because of their similar appearances [8, 9]. Due to their similar Chinese names, Adenophorae Radix is also easily confused with Glehniae Radix in clinical use [10]. The botanical origins of Adenophorae Radix are two species in the family Campanulaceae, Adenophora tetraphylla (Thunb.) Fisch (Lunyeshashen) and Adenophora stricta Miq. (Shashen).

The biologically active compounds of these adulterants are significantly distinct from those of Glehniae Radix. However these adulterants do not contain these biologically active compounds coumarins, coumarin glyeosides and polyacetylenes as Glehniae Radix [11]. The identification of Glehniae Radix and its adulterants has been based on morphological and microscopic observation [8, 9], while molecular identification has been rarely used [10, 12]. However, the assessment procedures are affected by environmental factors and often produce ambiguous results [13].

The DNA barcoding technique uses standard genomic regions to discriminate species [14–16], and this method provides consistent and reliable results regardless of the age, plant part or environmental factors of the sample [17]. Because of its speed and accuracy, DNA barcoding has gained attention in CM identification [12]. Although the Barcode of Life Plant Working Group (BOL) recommends the sequence combination rbcL + matK for barcoding [16], other genomic regions such as nuclear ITS (Internal Transcibed Spacer) may also be useful for medicinal material identification [17]. The ITS2 DNA sequence was suggested as a universal (medicinal) plant barcode [18–20]. The China Plant BOL Group (CBOL) has also suggested that ITS/ITS2 should be incorporated into the core barcode for seed plants [21]. The ITS2 region has been successfully applied to identify diverse medicinal plants and herbal materials [18, 22–26]. However, the ITS2 barcode was more likely to be affected by genetic anomalies, such as gene multiplication, pseudogenes and introgression [20].

For our study, we would like to identify commercial processed Glehniae Radix and its adulterants from a large pool of samples by the ITS2 sequences. This study aims to evaluate the suitability and feasibility of ITS2 barcode to accurately discriminate between Glehniae Radix and its adulterants, particularly the sequence divergences and differentiation powers of the ITS2 region.

Previous studies have successfully used 5S rRNA spacer domains to distinguish the original G. littoralis plant from its related medicinal species [10, 13] These studies extracted genomic DNA samples from leaves, but did not mention commercial Glehniae Radix samples. A study [31] explored the utility of ITS2 sequences for barcoding Glehniae Radix, their results displayed the similar trend that Glehnia Radix and its three adulterants could be identified by the ITS2 region. But the study included only four Glehniae Radix sequences and six sequences derived from three adulterant species. In this study, we used 46 and 59 ITS2 sequences from Glehniae Radix and its seven adulterants, respectively.

The processing procedures of Glehniae Radix are as following: remove the fibrous roots, stems and residual impurities, wash, slightly dry, rear blanched in boiling water, remove the skin, cut and dry. After processing, genomic DNA extracted from the roots of samples was usually severely degraded or contaminated by micro-organisms [32, 33]. Moreover, it was difficult to obtain high-quality DNA from samples enriched in polysaccharide polyphenols, fibre or other storage materials [17, 24]. In this study, we added 10 % PVP-30 powder when grinding the sample in liquid nitrogen to remove contaminants (polyphenols and polysaccharides). Subsequently, we added nuclear separation liquid to the plant powder 2–3 times, and extended water bath time to ensure DNA could be fully released into the extraction buffer. Additionally, root herbs often contain soil fungi; therefore, to avoid the influence of fungal contamination, one should clean the herb’s surface and isolate the interior material during the sampling procedure. We successfully extracted genomic DNA from 59 various herb samples, including 38 commercial crude drugs and 16 dried root samples.

Because of the DNA degradation of herbal medicines, DNA barcodes that are favourable for some plant species such as rbcL and matK cannot be used to identify commercial herbs [34]. In our study, the DNA isolated from the samples was severely degraded; however, 59 different versions of the ITS2 barcodes were readily retrievable due to the short length of the ITS2 in these samples (202–266 bp). Shorter fragments are easier to amplify from herbarium DNA [35], and the length of the ITS2 region might be sufficient to allow amplification without high-quality DNA [19, 24, 32]. Although multiple sequences were detected from a single individual due to its multicopy genes [36], intragenomic ITS2 variation typically occurred at only a very few, extremely variable, positions; thus, the ITS2 region can be treated as a single gene [37–39]. In our study, 46 sequences of the Glehniae Radix ITS2 barcode were obtained and were identical to each other. Xin et al. [40] also recognized that the application of the ITS2 barcode was not affected by the presence of multiple copies in Goji.

A successful barcode sequence requires both low intraspecific variation and high interspecific divergence. The 46 samples of G. littoralis analysed in this study, which were from original plants and commercial crude drugs, were representatives selected from extensive distribution areas, and their characteristics varied. The ITS2 sequences of G. littoralis within a given sample pool all belonged to the same haplotype, supporting Mizukami et al.’s assertion [41] that the genetic diversity among geographical strains of G. littoralis is narrow. The ITS2 intraspecific genetic distances in the medicinal Panax species were low [39]. Moreover, ITS2 displayed considerable interspecific divergence between Glehniae Radix and its adulterants. Thus, the ITS2 sequence as a barcode shows strong stability within species and high variability between species. Similar results have been found in Flos Lonicerae Japonicae (Jinyinhua) [23]. ITS2 sequences could be suitable for Glehniae Radix identification due to high conservation of the ITS2 regions derived from commercial crude drugs and original plant samples.

An NJ tree was constructed and compared the molecular morphological features of the ITS2 secondary sequences of Glehniae Radix and its adulterants. In the cluster dendrogram, each of the medicinal species was unambiguously distinguishable from each of the others. Furthermore, the secondary structure of ITS2 is considered a molecular morphological characteristic [30, 37]. This study demonstrated that ITS2 has a powerful identification capability, as the ITS2 sequences readily distinguished Glehniae Radix from its adulterants. DNA barcode discrimination technology has been applied to identify commercial plant products in teas [42] and Hypericum species [43], among others [26, 44].

DNA barcoding based on ITS2 distinguished commercial processed Glehniae Radix from common herbal adulterants.