Torreya is the surviving relict plant of the Cenozoic Era, belonging to the Gymnospermae Taxaceae. According to the Chinese flora (FRPS), there are 7 species of Torreya of which 4 species from China (Torreya fargesi, Torreya jackii, Torreya grandis and Torreya yunnanensis), 2 species from North America (Torreya taxifolia and Torreya californica), 1 species from Japan (Torreya nucifera) (Zheng et al., 1978). Most of the Torreya is a rare species endemic to China with complex morphological characteristics. The main classification bases are leaf length, leaf apex length, and endosperm morphology (Shen, 2011; Teng et al., 2017). It is difficult to accurately classify Torreya because of different growth environment and morphological variation caused by extensive introduction and cultivation. In recent years, some scholars still proposed to classify T. fargesi and T.yunnanensis as one species, and the systematic relationship between them needs to be further clarified. In practice, the difficulty in identification of Torreya has a great impact on research and application. For example, The seeds of T.grandis "Xiangfeizi" are famous high-grade nuts that arae known as "the king of dried fruits"(Li et al., 2014). T.grandis cv.'Merrillii' is an excellent cultivated variety of wild-type T.grandis. In recent years, the T.grandis cv.'Merrillii' cultivar has been vigorously developed in various places, and the supply of cedar tortoise seedlings is in short supply. However, due to the similar morphological characteristics of Torreya, other seedlings are often mixed with false T.grandis cv.'Merrillii' seedlings in the market, which it is difficult to identify effectively. T.grandis has a slow growth and long fruition cycle, so the fake seedlings will cause serious economic losses to the growers. Therefore, accurate classification and identification of plants for Torreya not only helps to clarify phylogenetic relationship, but also has important significance for the reasonable utilization of resources of Torreya.

At present, domestic and foreign researches on Torreya mainly focus on germplasm resources investigation, community characteristics and biological protection (Yu et al., 2014; Bo et al., 2016, Gao et al., 2019), but there are few studies on classification and identification. The morphological characteristics of Torreya are not obvious, especially in the long seedling stage before flowering and fruiting, which makes traditional morphological identification very difficult. DNA barcodes with the advantages of speed and accuracy have become an important tool for plant classification and identification which use a short and standard DNA fragment in the genome to identify species (Chen et al., 2010; Zhu et al., 2018; Li et al., 2020). ITS, ITS2, matK, and psbA-trnH and so on are commonly used DNA barcodes. China Plant BOL Group sampled 6 286 individuals representing 1 757 species in 141 genera of 75 families for data analysis. ITS showed the highest discriminatory power, and a combination of ITS and any plastid DNA marker was able to discriminate 69.9%~79.1% of species. In cases where it was difficult to amplify and directly sequence ITS in its entirety, just using ITS2 was a useful backup. China Plant BOL Group proposed ITS/ITS2 should be incorporated into the core barcode for seed plants (Li et al., 2011). Chen S.L research group evaluated the identification ability of ITS2 in a total of 6 600 medicinal plants in 753 genera, and found that the identification efficiency of ITS2 at the species level was 92.7%. Chen S.L research group recommended it as a standard barcode for medicinal plants (Miu et al., 2011). Tan et al. selected ITS sequence as the DNA barcode candidate sequence for Celios cristata, and successfully identified 6 common hybrids (Tan et al., 2020). Wang et al. selected ITS sequence to quickly identify succulents of Sedum and accurately identified 10 species of Sedum (Wang et al., 2020). Zhou et al. used the ITS2 sequence to successfully identify 25 species of “Xueteng”, and developed a medicinal material labeling system (Zhou et al. 2019).

In this study, nrDNA ITS sequences were used to identify and analyze the relationship between Torreya. Four different analysis methods (BLAST alignment, Kimura 2-parameter (K2P) genetic distance, SNP site analysis, and neighbor- Joining (NJ) tree) were used to identify Torreya. It is hoped to find an identification method for accurately identifying the T.grandis and T.grandis cv.'Merrillii'. At the same time, this study used the maximum likelihood (ML) to explore the genetic relationship of Torreya. The aim is to establish a rapid and accurate molecular identification method for Torreya, and provides a basis for the phylogenetic relationship of Torreya.

1 Results and Analysis

1.1 PCR amplification of ITS sequences

In this study, we used the DNA of the sample leaf as a template for PCR amplification, and the electrophoresis results showed that there were clear bands at the location of 1 000 ~ 2 000 bp (Figure 1). The PCR amplification effect was good which only a single and clear band was obtained. Then, the recovered and purified bands were sent to the sequencing company for sequencing.

Figure 1 PCR amplification of the sample ITS sequence Note: M: DL 2000 bp Marker; No. 1~8 is T. grandis samples, No. 9~12 is T.grandis cv.‘Merrillii’ samples; No. 13 is negative control

1.2 The ITS sequence features

In this study, the success rate of ITS sequence amplification and sequencing were 100%. After proofreading, the complete sequence was transmitted to NCBI database to obtain the ID. According to the preliminary analysis (Table 1), the length of ITS sequence in Torreya ranged from 1095 bp to 1105 bp, and the content of base C+G ranged from 59.9% to 61.5%, which was higher than A+T content. There were 51 variation sites and 1 051 conserved sites in ITS sequence.

Table 1 ITS sequence length and GC content of the sample

We used BLAST comparison method to evaluate the identification ability of ITS sequence in Torreya. This method was able to effectively identify 6 species of T. grandis, T. grandis cv.‘Merrillii’, T. jackii, T. nucifera, T. californica and T. taxifolia, but could not identify T. yunnanensis and T. fargesii.

1.3 The K2P genetic distance of ITS sequences

In this study, the genetic distance of Torreya was analyzed based on the K2P model, and the results showed that (Table 2): The average intraspecific genetice distance of Torreya was 0.006 (0~0.007 1), with the greatest genetic variation within T. taxifolia. The average interspecific genetice distance of Torreya was 0.015 0 (0~0.0302), among which the distance between T. yunnanensis and T. fargesii was the smallest, and the distance between T. grandis and T. taxifolia was the largest. By analyzing the K2P genetic distance of ITS sequence, we found that the minimum interspecific genetic distance was greater than the maximum intraspecific genetic in Torreya. which except T. yunnanensis and T. fargesii, T. grandis cv.‘Merrillii’ and T. nucifera. K2P genetic distance could effectively identify 4 species of T. grandis, T. jackii, T. californica and T. taxifolia.

Table 2 The genetic distance of ITS sequences in Torreya Note: The minimum genetic distance between species is below the diagonal, and the average genetic distance is above the diagonal

The barcoding gap showed (Figure 2) that there was no obvious barcoding gap on ITS sequence of Torreya, but the intraspecies and intraspecies genetic variation did not completely overlap. In the intraspecies genetic distance, 87.50% was less than 0.004 and 12.50% was between 0.004 and 0.008. The interspecific genetic distance was 75.00% greater than 0.008.

Figure 2 The barcoding of ITS sequences in Torreya

1.4 The SNPs site screening of ITS sequences

In this study, specific SNPs were screened by comparing ITS sequence variation sites in Torreya. The results showed that T. grandis cv.‘Merrillii’ had no SNP sites; T. grandis had three SNPs sites; T. nucifera had a SNP sites; T. jackii had two SNPs sites; T. yunnanensis and T. fargesii had no SNP. Moreover, we found that the T. californica and T. taxifolia in North America had 16 locis which were clearly different from other species, the T. californica had 4 SNPs sites, and the T. taxifolia had 7 SNPs sites (Figure 3). Further analysis revealed that there were T-C mutations at the bases 302, 610, and 1 044 of ITS sequences of T. grandis cv.‘Merrillii’ and T. grandis, which was the effective identification sites. SNP loci analysis could effectively identify 5 species of T. grandis, T. nucifera, T. californica and T. taxifolia.

Figure 3 Analysis of mutation sites in Torreya Note: Red is the SNP site of T. grandis, purple is the SNP of T. nucifera, orange is the SNP of T. jackii, blue is the SNP of T.californica, and green is the SNP of T.taxifolia. The black frame is the difference site between T. grandis cv.‘Merrillii’ and T. grandis

1.5 The Cluster analysis of ITS sequences

In this study, the NJ phylogenetic tree of Torreya was constructed based on ITS sequence, and the results showed that Torreya had good aggregation degrees of various species, and most of them was monophyletic (Figure 4A). Except T. yunnanensis and T. fargesii, other species of Torreya had obvious clustering without species crossing. Moreover, each species was clustered into a monophyletic group which had a high support rate. T. wallichiana and A. argotaenia, as two outgroups, were independent of Torreya and separately grouped into a single branch. NJ trees could effectively identify 6 species of T. grandis, T. grandis cv.‘Merrillii’, T. jackii, T. nucifera, T. californica and T. taxifolia.

Figure 4 Constructing clustering tree based on ITS sequence Note: A: Neighbor-Joining tree B: Maximum likelihood tree

The ML tree was used to analyze the genetic relationship of Torreya, and the results showed that the clustering situation of ML trees was similar to that of NJ tree (Figure 4B). Except T. yunnanensis and T. fargesii croosed to form a single cluster , other species of Torreya had obvious clustering with a high support rate. This result indicated that T. yunnanensis and T. fargesii were closely genetic relationship. In addition, T. grandis cv.‘Merrillii’ formed a parallel branch with T. nucifera, T. jacki formed a parallel branch with T. grandis, and T. californica formed a parallel branch with T. taxifolia, which showed a close genetic relationship between the two.

2 Discussion

Traditional morphological identification mainly relies on the phenotypic characteristics of plants (Yang et al., 2017), while DNA barcode identification relies on the polymorphism and difference of standard gene fragments to achieve rapid and accurate identification of species (Gogoi et al., 2018; Mosa et al., 2018). Plant ITS sequences include ITS1 and ITS2 sequences. In the preliminary test, the ITS2 fragment was amplified, and it was found that the ITS2 sequence of the sample was highly similar and the identification efficiency was low. Gao et al. (2017) also chosed ITS sequence as the core barcode and successfully identified peach kernel and almond due to the high similarity of ITS2 sequence. Therefore, The ITS sequences with 100% amplification rate and good sequencing quality were selected as the core identification sequence in this study. The results showed that all species of Torreya could be identified except T. yunnanensis and T. fargesii. If this single sequence could not fully identify the plant, the chloroplast gene could be used as an auxiliary barcode in combination for identification (Zhang et al., 2019; Chi et al., 2020; Intharuksa et al., 2020). The results showed that BLAST comparison, K2P genetic distance, SNP site screening and NJ trees could identify 6, 4, 5 and 6 species respectively, among which BLAST comparison and NJ trees had the highest identification efficiency. Yu et al. (2016) used RAPD method to identify Torreya, the method had poor stability and repeatability, and the reaction conditions needed to be optimized. But ITS sequence did not need to design multiple pairs of primers and optimized the reaction conditions. It has obvious advantages.

In this study, it was found that the T. yunnanensis and T. fargesii were very closely related. Not only was the K2P genetic distance between species small, but also the same evolutionary branch in the ML tree intersects with each other. In the FRPS, the T. yunnanensis and T. fargesii were divided into two different species. Silba et al. (1995) suggested that the two be merged under the T. grandis as a variant. However, Kang et al. (1995) suggested incorporating T. yunnanensis into T. fargesii. The results of this study supported the incorporation of T. yunnanensis into T. fargesii as variants, but not supported the incorporation of both as variants. Ma et al. (2014) and Zhou et al. (2015) analyzed the trnL_trnF and psbA-trnH genes of T. fargesii that reached the same conclusion. Later, in Flora of China (FOC), the T. yunnanensis was incorporated into the T. fargesii as a variant according to the same morphological characteristics such as the longitudinal groove of the leaf and the width of the midrib (Wu., 1999). The analysis results of the clustering tree in this study were basically consistent with the classification results of the traditional taxonomy in FOC, which could provide reference for the systematic classification of Torreya. At the same time, the study founded that the tree species located in North America had more mutation sites than Asian tree species. It not only was the mutation type more complicated, but the two were located in the two branches of the ML tree, which showed they were far from each other. It may be that different geographical growth environments gradually affect genetic characteristics. Geographic isolation had led to a reduction in gene exchange within the population, and genetic relationship had gradually become farther away (Guo et al., 2020). Similar geographical differentiation characteristics were also found in Bletilla Striata (Ren et al., 2019) and Oryza rufipogon Griff (Huang et al., 2019), the samples with closer geographical distribution had closer genetic relationships.

T.grandis cv.'Merrillii' is an excellent variety that has been cultivated and bred for a long time. In practice, T.grandis is often confused with T.grandis cv.'Merrillii' in the market, and it is difficult to rely only on morphological identification. This study found that the SNPs site and the NJ tree could effectively identify both. There were three bases serve as unique identification sites, and they were grouped into a monophyletic group without species crossover in the NJ tree. Therefore, in this study, nrDNA ITS fragments were used to successfully identify 6 species of Torreya, and the molecular identification between the T.grandis and T.grandis cv.'Merrillii' was carried out. It not only provides accurate identification of the plants of Torreya, but also helps to enrich the DNA barcode database of Torreya, At the same time, it can reduce the confusing use of T.grandis cv.'Merrillii' seedlings in the market, and rationally use Torreya resources to provide technical support.

3 Materials and Methods

3.1 Test materials and Reagents

The test samples were collected from Zhuji, Zhejiang (120°34 'E, 29°61' N), Shaoxing (120°44 'E, 30°07' N) and Yunyang County, Chongqing (108°67 'E, 30°95' N). They were identified as T. grandis cv.‘Merrillii’ and T. grandis by Liu Zhengyu, researcher of Chongqing Institute of Medicinal Plant Cultivation. Young and healthy plant leaves were taken, cleaned, quickly dried and preserved by silica gel. In this study, a total of 48 ITS sequences of plants for Torreya and outgroups were used for analysis, including 40 of Torreya (12 new ones) and 8 of Taxus Wallichiana and Ametotaxus argotaenia.

The plant genomic DNA Rapid Extraction Kit (centrifugal column) was purchased from Beijing Junord Biotechnology Co., LTD.I-5™ 2×High-Fidelity Master Mix was purchased from Beijing Qingke Biotechnology Co., LTD. DL2000 DNA Marker was purchased from Beijing Zhuang Meng International Biotechnology Co., LTD.ITS primer synthesis was completed by Beijing Qingke Biotechnology Co., LTD. Other experimental reagents were pure analytical reagents made in China.

3.2 Plant DNA extraction

The plant leaves were ground 1 800 times /min for 2 min with a ball mill, and 20 mg of sample fine powder was weighed. The sample DNA was extracted with the plant genomic DNA rapid extraction kit according to the instructions. The DNA purity and quality were detected by an ultraviolet-spectrophotometer and then placed at -20℃.

3.3 PCR amplification and sequencing

The reaction system included positive and negative primers (10 mol/L) 0.6μL, 2×Taq Mix 12.5μL, ddH₂O 9.8μL, template DNA 1μL. The amplification reaction procedure included predenaturation at 94℃ for 5 min; then 40 cycles containing denaturation at 94℃ for 30s, annealing at 53℃ for 30s, and elongation at 72℃ for 70 s. ITS primers (forward primers 5'-CCTTATCATTTAGAGGAAGG-3', reverse primers 5'- TCCTCCGCTT ATTGATATGC -3 '), PCR amplification products were detected by 1% agar-gel electrophoresis and sent to Beijing Qingke Biology Co., Ltd. for sequencing.

3.4 Data analysis

Sequencing results were edited and spliced by DNAMAN software. The software MEGA7.0.14 was used to analyze the sequence base variation sites, and Kimura 2-parameter model was used to analyze the genetic distance. The adjacent NJ trees were constructed for identification and analysis of Torreae, and the maximum likelihood ML trees were constructed for phylogenetic analysis. Then the Bootstrap value was set as 1000, and the support rate was more than 50%.

Authors’ contributions

ZW was the experimental designer and executor of this research. LY and LSE completed the analysis of experimental results and the writing of the first draft of the manuscript; BJ and XR participated in data collection and experimental design; RFM was the initiator and responsible person of the project, guiding experimental design, data analysis, manuscript writing and modification. All authors read and approved the final manuscript.

Acknowledgements

This study was funded by the Basic Scientific Research Project of Chongqing (19KF10-2012).

References

Bo J., Jian H.W., Jin S., and Xian R.Z., 2016, Characters and species diversity of Torreya fargesii Franch. community in the jinfo mountains, Journal of Landscape Research, 8(2): 71-74

Chen S.I., Yao H., Han J.P., Liu C., Song J.Y., Shi L.C., Zhu Y.J., Ma X.Y., Gao T., Pang X.H., Luo K., Li Y.L., Xi W., Jia X.C., Lin Y.L., and Leon C., 2010, Validation of the ITS2 region as a novel DNA barcode for identifying medicinal plant species, PLoS ONE, 5(1): e8613

https://doi.org/10.1371/journal.pone.0008613

Chi C.C., Jaakko H., and Harald S., 2020, Exploring phylogeny of the microsoroid ferns (Polypodiaceae) based on six plastid DNA markers, Mol. Phylogenet. Evol., 143: 106665-106672

https://doi.org/10.1016/j.ympev.2019.106665

Gao L., Lu Z.P., Le J.X., Zhang W., Wu J.J., Tian Q.L., and Liu Y., 2019, Research progress of conservation biology on Torreya fargesii, a rare and endangered plant endemic to China, Chongqing Shifan Daxue Xuebao (Journal of Chongqing Normal University (Natural Science), 36(2): 130-135

Gao L.H., Yin Y., Li J., Ma Y.M., Li W.B., Li B.W., and Zhang X.N., 2017, Identification of persicae semen and armeniacae semen amarum by PCR amplification of specific alleles based on ITS sequences, Shijie Zhongyiyao (World Chinese Medicine), 12(9): 2190-2194

Gogoi B., and Bhau B.S., 2018, DNA barcoding of the genus Nepenthes (Pitcher plant): a preliminary assessment towards its identification, BMC Plant Biology, 18(1): 153-160

https://doi.org/10.1186/s12870-018-1375-5

Guo L.P., Zhou L.Y., Kang C.Z., Wang H.Y., Zhang W.J., Wang S., Wang R.S., Wang X., Han B.X., Zhou T., and Huang L.Q., 2020, Strategies for medicinal plants to adapt to environmental stress and "simulative habitat cultivation" of Dao-di herbs, Zhongguo Zhongyao Zazhi (China Journal of Chinese Materia Medica), 45(9): 1969-1974

Huang J., Gao L.J., Gao J., Qing D.J., Zhu C.L., and Zhou Y.W., 2019, Analysis of genetic diversity on populations of common wild rice (Oryza rufipogon Griff.) from 6 countries in Asia, Southwest China Journal of Agricultural Sciences, 32(2): 232-240

Intharuksa A., Sasaki Y., Ando H., Charoensup W., Suksathan R., Kertsawang K., Sirisa-Ard P., and Mikage M., 2020, The combination of ITS2 and psbA-trnH region is powerful DNA barcode markers for authentication of medicinal Terminalia plants from Thailand, J. Nat. Med., 74(4): 282-293

https://doi.org/10.1007/s11418-019-01365-w

Kang N., and Tang Z.X., 1995, Studies on the taxonomy of the genus Torreyas, Zhiwu Yanjiu (Bulletin of Botanical Research), 15(3): 349-362

Li D.Z., Gao L.M., Li H.T., Wang H., Ge X.J, Liu J.Q., Chen Z.D., Zhong S.L., Chen S.L., Yang J.B., Fu C.X., Zeng C.Z., Yan H.F., Zhu Y.J., Sun Y.S., Chen S.Y., Zho L., Wang K., and Duan G.W., 2011, From the cover: Comparative analysis of a large dataset indicates that internal transcribed spacer (ITS) should be incorporated into the core barcode for seed plants, Proc. Natl. Acad Sci. USA, 108(49): 19641-19646

https://doi.org/10.1073/pnas.1104551108

Li R.F., 2014, The effects of trace elements of Torreya grandis ‘Merrillii’on production and quality, Thesis for M.S., Zhejiang Agriculture and Forestry University, Supervisor: Ye Z.Q., pp.5-8 

Li Z.Z., Zheng S.H., Shang X.G., Deng T.W., Wang J.J., and Zhao R.H., 2020, Molecular identification in genus of Glycyrrhiza based on DNA Barcoding, Zhongguo Xiandai Zhongyao (Modern Chinese Medicine), 22(2): 207-212

Ma J., Chen F.B., Zhou X.R., Wang J.H., and Song X.B., 2014, Sequence analysis and comparison of trnL-trnF of geographic populations of Torreya fargesii (Taxaceae) and its closely related species, Jiyinzuxue Yu Yingyong Shenwuxue (Genomics and Applied Biology), 33(5): 1053-1058

Miu Z.P., Zhang J.M., Li J.H., Hong X., and Pan T., 2018, The complete chloroplast genome sequence of an conifer plant Torreya grandis (Pinales, Taxaceae), Mitochondrial DNA Part B, 8(3): 1152-1153

https://doi.org/10.1080/23802359.2018.1522975

Mosa K.A., Gairola S., and Jamdade R., 2019, The promise of molecular and genomic techniques for biodiversity research and DNA barcoding of the Arabian Peninsula Flora, Front. Plant Sci., 9: 1929-1948

https://doi.org/10.3389/fpls.2018.01929

Ren F.M., Bai J., Liu Y., Cao H.Q., Wang Q.G., and Li S.Q., 2019, Study on genetic diversity of Bletilla striata from different localities based on nrDNA ITS, Shijie Kexue Jishu: Zhongyiyao Xiandaihue (Modernization of Traditional Chinese Medicine and Materia Medica-World Science and Technology), 21(6): 1189-1195

Shen D.F., 2011, Collection and evaluation of Torreya grandis germplasm resources and determination of core germplasm, Thesis for M.S., Zhejiang Agriculture and Forestry University, Supervisor: Zeng Y.R., pp. 8-10

Silba J., 1990, A supplement to the international census of the coniferae, II, Phytologia, 68(1): 7-78

Tan X.N., Wu W.R., Lai H.L., Zhang D.Y., An X., Huang Y.H., and Chen Z.Z., 2020, Identification of celosia cristata based on systematic identification of traditional Chinese medicine, Zhongyaocai (Chinese herbal medicine), (2): 296-302

Teng T.Y., Pan C.C., Liu J.L., Chen Z.I., Xie W.Y., and Jin X.F., 2017, Leaf variation of Torreya plants in Zhejiang and its taxonomic implications, Hangzhou Shifan Daxue (Journal of Hangzhou Normal University), 16(1): 9-12

Wang J.H., Zhou X.R., Shang J., Jiang B., and Yang L.P., 2014, Niche characteristics of dominant woody plant populations in a Torreya fargesii shrub community in the Jinfo Mountains, Shengtaixue Zazhi (Journal of Ecology), 33(5): 1135-1141

Wang Z.M., Jiang Y., Hu J.X., and Liao F., 2020, Analysis of succulent plants of Sedum based on ITS barcode , Zhiwu Jianyi (Plant Quarantine), 34(2): 38-41

Yang B.H., and Li C.P., 2017, The full-length mitochondrial rDNA sequence of Tyrophagus longior (Astigmata:Acaridae) and phylogeny of astigmatid mites based on mitochondrial rDNA sequence, J. Stored Prod Res., 71: 5-13

https://doi.org/10.1016/j.jspr.2016.12.004

Yu H.P., Zhao L., Cheng X.J., Wang Y.F., and Xu W.J., 2016, Identifying the genus of Torreya by artificial drawing of variety identification method, Zhejiang Linye Keji (Zhejiang Forestry Science and Technology), 36(3): 19-23

Zhang Z.H., Yang C., Tong S.S., Deng J.B., Gao G., Zhang L., and Yang R.W., 2019, Identification and analysis of DNA barcodes of Hemsleya, Fenzi Zhiwu Yuzhong (Molecular Plant Breeding), 17(9): 2899-2905

Zhou H., Ma S.J., Song J.Y., Lin Y.I., Wu Z.J., Han Z.Z., and Yao H., 2019, QR code labeling system for Xueteng-related herbs based on DNA barcode, Chinese Herbal Medicines, 11(1): 52-59

https://doi.org/10.1016/j.chmed.2018.09.006

Zhou X.R., Shang J., Chen F.b., Song X.B., and Jiang B., 2015, psbA-trnH sequence analysis of Torreya fargesii and its closely related species, Xibu Linye Kexue (Western Forestry Science), 44(1): 16-21

Zhu R.W.., Li Y.C., Zhong D.L., Zhang J.Q., 2018, Author correction: Establishment of the most comprehensive ITS2 barcode database to date of the traditional medicinal plant Rhodiola (Crassulaceae), Sci. Rep., 8(1): 6741

https://doi.org/10.1038/s41598-018-24988-7