Identification of Crocus sativus and its Adulterants from Chinese Markets by using DNA Barcoding Technique

Abstract

Background

Saffron (Crocus sativus L.) is a common but very expensive herbal medicine. As an important traditional medicine, it has an outstanding effect in treating irregular and painful menstruation. Recently, the over-demand tendency of saffron results in an unusual phenomenon in the medicinal markets. Adulterants and saffron-like substitutes are intentionally mixed into medicinal markets and pharmacies or online stores, affecting drug safety and food quality.

Objectives

Our study aimed to identify saffron from its adulterants via DNA barcoding.

Materials and Methods

Samples (13 saffron + 4 others containing Carthamus tinctorius or Chrysanthemum x morifolium) obtained from 12 different provinces of China. Through DNA barcoding, samples were compared using three candidate markers, trnH-psbA, rbcL-a and ITS2.

Results

trnH-psbA and rbcL-a were capable of distinguishing different accessions. ITS2 could identify samples even at intra-specific level. According to these three barcodes, four samples were identified saffron-like substitutes.

Conclusions

The adulterant rate in Chinese markets reaches as high as 33.33% that may cause health risks and further may reduce saffron efficacy once is being used as herbal remedy. In order to make a distinction between C. sativus with other genera as adulterants, DNA barcoding is suggested.

1. Background

Crocus sativus L., commonly known as saffron, or Zang-Hong-Hua in Chinese, is a perennial stemless herb of Iridaceae (1). Saffron originally grew in Iran, India, Spain, Greece. It has been also successfully cultivated in various places in China for many centuries (2). Since ancient times, the dried stigmas of C. sativus have been considered as the most precious and expensive medicine (3, 4). In folk, it is commonly used for its analgesic and/or sedative properties. In traditional Chinese medicine (TCM), it has been used as an anti-anginal (2). The other pharmacological activities of saffron, such as anti-cancer, anti-inflammatory, and anti-atherosclerotic activities have also been reported (5, 6). The major active constituents responsible for these biological activities are two natural carotenoids, crocin and crocetin (7). For example, crocetin, an antioxidant, has memory enhancing effects in aged mice (8) and crocin demonstrated to be useful in pharmacological alleviation of cognitive defects (9). Even saffron odor may be effective in treating menstrual distress (10). In addition, people use saffron to season their food for its flavor and taste.

However, due to the limited production and labor-intensive process of harvesting, it is sold at an extremely high price around $2,000/kg. Due to its high demand and price, adulterations are common with the crude drugs of Carthamus tinctorius, Chrysanthemum x morifolium, Zea mays and Nelumbo nucifera (11-14). These plants have fiber-type structure similar to C. sativus anthers that makes the differentiation rather impossible by just morphological investigation. However, the contents of the active compounds of these adulterants are distinct from that of saffron that obviously reduce the efficacy of C. sativus. Thus, for the benefits of consumers and the quality control of saffron, its correct identification is a must.

Although herbal medicine has a great sayings in pharmalogical sciences, it suffers from contaminations that may occur through the processing steps during collection, drying and grinding (15). To check the authenticity of a medicine, physical, chemical, and molecular marker techniques are amongst the suggested methods. Although physical methods such as microscopy are easy to operate, the amount of helpful data to check the purity of product is not enough. Chemical methods such as chromatography or spectroscopic analysis are time-consuming and cost-effective and may be strongly influenced by the experimental conditions. DNA barcoding on the other hand that takes advantage of using particular nuclear or organellar genome sequences as genetic markers seems a way forward in determination any impurities. This technique is rapid, sensitive, accurate and simple and in our case is efficient for species identification (16-18).

Recently, this technique has been widely applied to discriminate medicinal herbs of TCM, such as Codonopsis Radix, Paeoniae Radix Rubra, Clematis chinensis, using ITS2 (19-21). Sabia species and Radix astragali were correctly identified by the combined DNA barcodes (22, 23). Therefore, use of DNA barcodes in identification of medicinal herbs is recommended.

According to previous saffron identification studies, most materials came from wild or garden species (24, 25). The authentication of saffron from medicinal markets or drugstores was rarely tested (26). In this study, we employed three DNA barcodes, including trnH-psbA intergenic spacer (trnH-psbA), large subset of ribulose-bisphosphate carboxylase (rbcL-a) and nuclear internal transcribed spacer 2 (ITS2), to differentiate saffron from its adulterants by sequences diversity analysis.

2. Objectives

Our objectives were not only to investigate and detect the market substitution of saffron, but also to identify what the adulterants were as well as to remind people to be cautious to select.

3. Materials and Methods

3.1. Plant Materials

All 13 samples were called Zang-Hong-Hua in Chinese, which meant they had been used or sold as saffron (C. sativus). They were collected from 12 different provinces. Among them, only two samples were identified as C. sativus (S1) and Carthamus tinctorius (C1) respectively, based on their morphological characteristics. Other 11 samples unidentified, labeled saffron were bought from medicinal markets and drugstores during 2012-2013 (Table 1). These samples were from commercial products, except the one from Nanjing Botanical Garden Mem. Sun Yat-Zen. The familiar adulterants, such as Carthamus tinctorius (C2), Nelumbo nucifera (NN), Crysanthemum x morifolium (CM) and Zea mays (ZM), their relevant gene sequence numbers downloaded from Genbank were also listed in Table 1.

Table 1. Plant samples used in the study .

Latin name		Sample ID		Origin		Genbank No.
Crocus sativus (To be determined) (To be determined) (To be determined) (To be determined) (To be determined) (To be determined) (To be determined) (To be determined) (To be determined) (To be determined) (To be determined) Carthamus tinctorius Carthamus tinctorius Nelumbo nucifera Chrysanthemum x morifolium Zea mays		S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 C1 C2 NN CM ZM		Nanjing Botanical Garden Mem. Sun Yat-Sen, Jiangsu Shanghai Plantation, Shanghai Beijing Tongrentang Pharmacy, Beijing Lhasa Pharmacy, Xizang Bozhou medicinal market, Anhui Chengdu medicinal market, Sichuan Nanjing Pharmacy, Jiangsu Dalian, Liaoning Baoding, Hebei Yantan, Shandong Zhangzhou, Fujian Kunming, Yunnan Urumqi, Xinjiang ---- ---- ---- ----		KF88648, KF88658, KF886671 KF886649, KF886659, KF886672 KF886650, KF886660, KF886673 KF886651, KF886661, KF886674 KF886652, KF886662, KF886675 KF886653, KF886663, KF886676 KF886654, KF886664, KF886677 KF886665 KF886666, KF886678 KF886667, KF886679 KF886655, KF886668, KF886680 KF886656, KF886669, KF886681 KF886657, KF886670, KF886682 GQ435089, GU990409, GU724317 AB331309, JN407324, JF977132 GU575275, JF949972, FJ539128 GU575286, M16836, DQ68301

3.2. DNA Isolation, Amplification and Sequencing

Total genomic DNA was isolated from the dried stigma (20 mg each). The stigma of each sample was immersed in liquid nitrogen and crushed into a fine powder. Plant Genome DNA Kit (DP305, Beijing, China) was used for DNA extraction. Quality of the extracted DNA was determined using gel electrophoresis.

Amplifications of the three loci trnH-psbA, rbcL-a, and ITS2 were obtained. The primer pair names, primer sequences, and reaction conditions (27-30) used are listed in Table 2.

Table 2. Primers and reaction conditions used in this study .

Locus	Primer name	Primer sequence (5'-3')	Reaction conditions
trnH-psbA rbcL-a ITS2	PA TH R-F R-R ITS3F ITS2R	GTTATGCATGAACGTAATGCTC CGCGCATGGTGGATTCACAATCC ATGTCACCACAAACAGAGACT TCGCATGTACCTGCAG-TAGC ATGCGATACTTGGTGTGAAT GACGCTTCTCCAGACTACAAT	94ºC 5 min 94ºC 1 min, 55ºC 1 min, 72ºC 1.5 min, 35 cycles 72ºC 7 min 95ºC 2 min 94ºC 1min, 55ºC 30 s, 72ºC 1 min, 34 cycles 72ºC 7min 94ºC 5 min 94ºC 30 s, 56ºC 30 s, 72ºC 45 s, 40 cycles 72ºC 10 min

trnH-psbA: trnH-psbA intergenic spacer; rbcL-a: ribulose-1,4-bisphosphate carboxylase large subunit; ITS2: internal transcribed spacer 2.

Polymerase chain reaction (PCR) amplifications of the three candidate DNA barcode loci carried out in a T100 Thermal Cycler (Bio-Rad, United States) were performed in a 25 μL reaction mixture containing 2.5 μL of 10× PCR buffer, 2.0 μL of 25 mmol.L^-1 MgCl₂, 2.5 μL of 2.5 mmol.L^-1 dNTPs, 0.5 μL of each primer (10 mmol.L^-1), 0.5 U of Taq DNA polymerase, 1 μL of genomic DNA (~30 ng) and ddH₂O. PCR products were examined using 1% agarose gel electrophoresis in 1×TAE buffer at 100 Volt for about 30 min. Gel images were obtained using Gel documentation imaging system (Bio-Rad, United States). Purifying and sequencing were completed by Biosune Co., Ltd (Beijing, China).

3.3. Sequence Alignment and Analysis

Sequences were assembled and aligned by Clustal X software (Version 1.83) and adjusted manually in CodonCode Aligner (Version 4.2.3). The nucleotide sequences data of the partial trnH-psbA spacer, rbcL-a and ITS2 genes were submitted to Genbank. Genetic distance was computed using MEGA5.2 by Kimura two-parameter (K2P) model (31). Based on each locus, maximum parsimony trees were built with bootstrap testing of 1000 replicates. The DNA sequences were then deposited in Genbank.

4. Results

Three genes from nuclear and chloroplast were selected to discover genetic variations and to carry out analysis on the identification of 11 samples retailed as saffron in different markets. A total of 39 DNA sequences from three DNA barcodes (trnH-psbA, rbcL-a, and ITS2) were generated (Table 3). They were able to differentiate authentic saffron from the substituted ones with inter-specifically variable sites among 11 commodity samples from S2 to S12. The three DNA barcodes were successfully amplified in the order of trnH-psbA (13/13) > rbcL-a (11/13) > ITS2 (10/13). By combining a portion of trnH-psbA gene with ITS2 gene (Figure 1), it was demonstrated that both sequence variations can distinguish C. sativus from different sources at intra-species taxa. Based on the genetic analysis of rbcL-a region, C1 and C2 were separated into two different clades with 67% supporting rate and clustered C1, S11, S12 and CM (C. x morifolium) into one clade with bootstrap support of 98%.

Table 3. Properties of the three DNA barcoding regions of C. sativus and its adulterants. Numbers about the polymorphic sites are their positions in the multiple sequence alignment.

Property	Sample	trnH-psbA	rbcL-a	ITS2
Sequence length avg. (bp) All inter-specific distance (K2P) Selected variable sites	S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 C. tinctorius C. x morifolium N. nucifera Z. mays	431 1.04 105bp-112bp TCGGGAAA TCGGGAAA TCGGGAAA TCGGGAAA TCGGGAAA TCGGGAAA TCGGGAAA TCGGGAAA GTAGGATT GTAGGATT GTAGTATT TTAGTAT- GTAGTATT GTACTAT- ATTTTATT TTTGCGAT	655 0.58 347bp-354bp CCCCCTGC CCCCCTGC CCCCCTGC CCCCCTGC CCCCCTGC CCCCCTGC CCCCCTGC – – CCTCCTGC CCTACTGC CCTACTGC CCTACTGC CCTACTGC CCTCCTGC GTCCCTGT	240 1.27 127bp-134bp TCCGGCTT TCCGGCTT TCCGGCTT TCCGGCTT TCCCCCTC TCCGGCTT TCCGGCTT – – – CTAAAGAC TTAAAAAC CTAAAGAC TTAAAAAC GTGTGACC CCCGGCGC

Figure 1 .

Classification tree of combining trnH-psbA with ITS2 gene using the Maximum Parsimony (MP) method

Figure 2 .

Classification tree of partial trnH-psbA gene using the Maximum Parsimony (MP) method

Maximum parsimony tree constructed by either single locus or combined loci suggested that the samples were divided into two major clades (Figures 1 and 2). Both of the phylogenetic trees demonstrated that S2 to S8 were authentic saffron among market samples, while S9 (Hebei), S10 (Shandong), S11 (Fujian) and S12 (Yunnan) were not with 99% bootstrap support (BS). The accessions including S9, S10, S11, S12, C1, C2 and CM were divided into two sub-clusters in the trnH-psbA tree. As for the relationship among species, S11 and S12 were closer to C. tinctorius (99% BS) and C. x morifolium (97% BS), respectively through phylogenetic analysis. From Figure 1, S9 and S10 were inferred to be the same species (99% BS) and could belong to Carthamus. Thus, S9 (Hebei), S10 (Shandong), S11 (Fujian) and S12 (Yunnan) were proved not saffron, but its adulterants.

5. Discussion

The result showed that the trnH-psbA region could place an unidentified specimen into a family and even genus (32), while rbcL-a may be effective in identifying specimen at family level. As one of the most common regions used for phylogenetic analyses, the ITS2 region was selected as a barcode candidate that can distinguish accessions from different geographical origin at the species level (33, 34). But one of potentially negative factor for sequencing ITS2 is the presence of poly-G, poly-C, and poly-A repeats (35). Moreover, multi-locus approach may be a much more effective strategy for species identification. As C. sativus and its adulterants are completely in different families, each of the three loci is adequate when used at the family level. In addition, DNA integrity and purity are the major concerns in DNA barcodes. Poor DNA quality and quantity may lead to the unsuccessful amplification of DNA barcodes in some of the commodity samples. When materials are highly processed or stored for a long time, total DNA is highly degraded or contaminated (36) that leads to the DNA amplification and sequencing difficulties.

When presented with a completely unknown sample, it would be highly desirable to place it in a smaller group of taxa (i.e. within one genus) (37). DNA based methods maybe more useful in quickly and efficiently discriminate adulterated or substituted raw materials (38). By using DNA barcoding method, total DNA of species can be achieved from fresh and dried plant parts, processed herbal drugs, as well as tablets and capsules. For its easy operation, DNA barcodes will also be powerful tool for non-professional users. Nowadays, with the exponential growth of international market and increasing demand of high quality medicinal materials, the adulterated or substituted source materials have sprung up and could confuse the wholesalers, retailers and consumers (39). Thus, correct identification by DNA barcoding technique has become an essential task for the regulatory authorities and related industries in order to ensure drugs safety and fair trade. As a result, adverse health problems can be avoided and negative effects such as purchasing products mixed with adulterants or protected species marketed as unprotected species can be reduced (40, 41). Though many substitutions have been disclosed by researchers, still lots of adulterants have remained undiscovered in the medicine market.

Through our research, we found the adulterant rate of saffron from Chinese markets reached as high as 33.33% (excluding the sample we collected from the botanical garden). Saffron can be identified by a unique DNA barcode or a combination of multiple DNA barcodes. This technology is useful in providing a reliable and effective means for the differentiation of saffron from its substitutes or adulterants. With the adoption of barcoding as an authentication tool, perhaps it will be possible to discourage medicinal plants substitution in the marketplace. As a high-valued product, the official authentication and monitoring of saffron become very important in China for safety reasons. And DNA barcode tool is highly recommended.