论文部分内容阅读
Abstract Identification of kiwifruit germplasm materials is the basis of protecting and utilizing these resources. However， identifying Actinidia arguta varieties based on morphology is difficult， especially for non-specialists. In this study， we collected 180 specimens comprising 60 varieties of three species （A. arguta， A. kolomikta， and A. polygama） from the Northeast of China. The emphasis of our study was on the feasibility of identifying A. arguta varieties. Here， we used common analysis methods （genetic distance， phylogenetic analysis and the DNA barcoding gap） and SNPs analysis to evaluate the discriminatory power of different DNA barcoding markers. The results revealed that common methods were insufficient to identify A. arguta varieties but SNPs analysis based on DNA barcoden was the potential method for identifying A. arguta varieties. Besides， Our study agree that DNA barcoding could be used to analysis the evaluation genetic relationship of the Actinidia plants.
Key words Actinidia arguta； DNA barcoding； SNP ； Identification
Actinidia arguta[A. arguta （Sieb. et Zucc.） Planch. Ex Miq.]， A. polygama[A. polygama （Sieb .et Zucc.） Maxim.]and A. kolomikta[A. kolomikta （Rupr. & Maxim.） Maxim.]are native to eastern-most Siberia， China， Korea， Japan， Sakhalin and the Kuril Islands[1-2]. As a result of geographical and climatic adaptation， these plants have good cold resistance， which makes them excellent gene donors for kiwifruit breeding.
A. arguta， also called hardy kiwifruit or mini kiwi， has been more and more popular in the market and has received the attention of researchers and local farmers[3-4]. A. arguta is the third kiwifruit species commercially cultivated followed the A. deliciosa and A. chinensis. Compared with A. deliciosa and A. chinensis， A. arguta is smaller （4-20 g） and can be eaten whole without peeling. Studies suggest that A. arguta can withstand temperatures down to 30 ℃ and have the resistance to bacterial canker disease of kiwifruit （Pseudomonas syringae pv actinidiae）[4]. Besides， the root of hardy kiwifruit has been served as a traditional herbal medicine mainly used for gastric cancer， esophageal cancer and other gastrointestinal cancer in China. New cultivars of kiwifruit， such as "kuilv"， "fenglv" and "jialv"， have been developed by breeding programs in northeastern China， and planting area has been extended year by year. However， the dummy variety has appeared in the nursery stock market duo to commercial profit. The identification of varieties is beneficial to the preservation， development and utilization of these plants. However， studies concerning the identification of these cold resistant kiwifruit are relatively rare. Numerous analytical methods that rely on molecular markers have been developed for research into the genetic diversity of genus Actinidia， such as simple sequence repeat （SSR） markers[4-5] and random amplified polymorphic DNA（RAPD） markers[6]. However， these methods are of limited use in variety identification.
DNA barcoding is a new technique for molecular identification proposed in recent years that uses a standard short genetic region that is universally present in the target lineages and has sufficient sequence variation to discriminate among species[7-10]. DNA barcoding， which was put forward for the first time in 2003 by Canadian anthropologist Paul Hebert， has attracted significant attention around the world. This method was introduced into the study of plants in approximately 2005[7， 9]. The Consortium for the Barcode of Life （CBOL） Plant Working Group identified and recommended the use of chloroplast genes rbcL and matK in 2009[11-12]. However， this recommendation was based on the study of only a relatively small number of species in which multiple individuals were sampled from multiple congeneric species.
Single nucleotide polymorphism（SNP） is a single nucleotide variation in a specific genetic location[13]. SNP is one of the most abundant and stable genetic polymorphisms and is applicable to resolve differences among closely related species[14-15]. In this study， SNP typing method was evaluated using ITS2 markers to identify hardy kiwifruit varieties.
Materials and Methods
Taxon sampling
We collected 180 samples comprising three individuals each from 60 varieties of three species from Northeastern China. These species include A. arguta （including A. arguta var. purpurea， 51 varieties total）， A. kolomikta （6 varieties） and A. polygama （3 varieties）. All the specimens were collected from northeastern China and preserved in silica gel. The varieties and test numbers are shown in Appendix1.
DNA extraction， PCR amplification and sequencing
Genomic DNA was extracted from silica gel-dried leaf material with liquid nitrogen using the Plant Genomic DNA Kit （TIANGEN BIOTECH， Beijing， China）， following the supplier’s protocols. Five regions （ITS， ITS2， matK， rbcL， trnH-psbA） were amplified and sequenced. Each of the five regions used one universal pair of primers for sequence amplification. The volume of the PCR reaction was 30 μl， including： 2×PCR Reagent 15 μl （0.1 U Taq plus polymerase/μl， 500 μM dNTP each， 20 mM Tris-HCl （pH=8.3）， 100 mM KCl， 3 mM MgCl2） produced by Tiangen Biotech （Beijing， China） Co.， Ltd； DNA 2 μl； forward primer （2.5 μM） 1.5 μl， Reverse primer（2.5 μM）1.5 μl； ddH2O 10 μl. The reaction conditions were optimized by gradient PCR and are shown in Table 1. The sequencing reaction were performed by Sangon Biotech （Shanghai， China） Co.， Ltd using the same primers as used for PCR. All samples were sequenced via direct sequencing of PCR products. Sequence reads were assembled using CodonCode Aligner 6.0.2 （CodonCode， Centerville， Massachusetts， U.S.A.）.
Data analysis
Candidate DNA barcodes were aligned by ClustalX using default parameters[21]. Poor quality base calls at the 5’ and 3’ ends of the sequenced PCR products were removed. Kimura-2-Parameter （K2P） distances， conserved sites， variable sites， parsimony-information sites and GC content were computed with MEGA 6.0[22]. "Intra-varieties distances"， "inter-specific distances"， and "intra-specific & inter-variety distances" were calculated using a K2P distance matrix[23-24]. Wilcoxon signed rank tests were performed by IBM SPSS Statistics 21 to compare and screen the genetic diversity of different markers. The distribution of genetic distance was shown using the DNA barcoding gap[18， 23]. Variety identification and phylogenetic relationships among species were analyzed by genetic distance and tree-based method[25]. SNP sites detection were performed using MEGA 6.0.
Results
Amplification， sequencing success and marker features
Amplification efficiency， sequencing success， variable sites and other information can be seen in Table 2.
The amplification efficiency and sequencing quality are important for the evaluation of candidate barcode sequences. Amplification efficiency of each candidate barcode and the tested combinations was 100%. The sequencing rates of different loci were obviously different. The sequencing rate of the ITS region was only 27.78%， which was too low to do further analysis， so the ITS region was rejected as a barcode for identifying wild kiwifruit resources in northeastern China. The remaining candidate barcodes were subjected to further analysis due to their high sequencing rates. Sequence lengths of the four candidate barcodes ranged from 424 to 832 bp. The nuclear gene ITS2 region’s variable sites among three species totaled 39 bp which was much more than the variable sites in the chloroplast genes. The trnH-psbA region’s variable sites were fewest， with only 4 bp. Therefore， the ITS2 region might have more advantages for identifying species with close relationships. The barcoding combinations had more informative sites， which might contribute to the analysis of genetic diversity.
Evaluation of identifying power based on the barcoding gap The barcoding gap was estimated by comparing the intra- and inter-varietal divergences of each candidate locus[12]. In an ideal situation， the genetic variation of a DNA barcode should demonstrate separate， non-overlapping distributions between intra- and inter-varietal samples[16].
Agricultural Biotechnology2018
Our result （Fig. 1） demonstrated that the distributions of all the markers in this paper failed to exhibit distinct gaps at variety level， which means it was difficult for this analysis method to identify these closely related varieties.
However， there was an obvious gap between inter- and intra-specific variations， meaning that species level identification could be completed.
Evaluation of identifying power based on a tree-based method
In this study， for phylogenetic reconstruction we used the Neighbor-Joining （NJ） methods implemented in MEGA version 6.0. Values at each node were assessed by 1 000 bootstrap replicates. For identifying the varieties we chosen the ITS2 region and the combination ITS2+rbcL to construct the NJ tree due to their better performance in inter-varietal genetic distance compared with others.
If the same varietal samples could be grouped into a variety-specific cluster using phylogenetic analysis based on a candidate barcode， it would indicate that this barcode could successfully identify the varieties. ITS2 could not detect variety-specific clusters using phylogenetic analysis （Fig. 2）， which indicated that the intraspecific variation of ITS2 region was too poor to identify varieties. Similarly， the combination of ITS2+rbcL could not complete the identification at the varietal level （Fig. 3）. However， the 3 species were grouped into species-specific clusters successfully， which means the 3 kiwifruit species could be identified based on phylogenetic analysis.
Evaluation of identifying power based on wilcoxon signed rank tests
Usually， wilcoxon signed rank tests were used to analyze the variability of sequences[26-28]. Wilcoxon signed rank tests on combined data showed that ITS2 was the most variable barcode at theinterspecific level， followed by the combination ITS2+rbcL （Table 3）. At the intraspecific level， Wilcoxon signed rank tests showed that rbcL and trnH-psbA had the lowest level of divergence， whereas the highest was provided by ITS2 （Table 4）.
Analysis of SNPs
SNP analysis is an effective method for detecting genetic diversity in plants particularly in very closely related species[29]. All the ITS2 sequences here from the A. arguta varieties were examined for SNPs at the variety level （Fig. 4）. Three differential site at position 135， 220， 243 bp were found， which could identify some cultivars. One polymorphism site was found at 416 bp of matK sequences. rbcL and trnH-psbA have been proved to be an important markers in addressing phylogenetic relationships at various taxonomic levels[19， 30-31]. However， these markers cannot provide even one polymorphism to distinguish individuals of A. arguta. At position 135 bp， "Fenglv"， an A. arguta cultivar， carrys C but "A11-10-2" and "B9-2-1" carry A. Therefore， "Fenglv" could be confirmed uniquely in these three varieties. "A11-10-2" and "B9-2-1" also can be distinguished based on the diversity at the position 220 or 243 bp.
Discussion
The identification of Actinidia based on the common method
In our study， we sampled 3 species and each of them could be identified by phylogenetic trees based on DNA barcodes （Fig. 2； Fig. 3）， which means that interspecific identification may be completed successfully through DNA barcoding. For proving this hypothesis， we constructed phylogenetic tree （Fig. 5） based on matK sequence to evaluate the identification power of Actinidia plants （some sequences from this experiment， others collected from GenBank）.
Previous studies have constructed phylogenetic trees of Actinidia using the sequences of ITS and suggested that DNA barcoding could be used in classification and phylogenetic analysis[32]. Li et al.[33] sequenced ITS and matK gene of 23 Actinidia species. They argued neither ITS nor matK phylogeny supported monophyly of any of the four sect of the genus （Maculatae， Strigosae， Stellatae and Leiocarpae）， which conclusion consistented with our study. Besides， we thought DNA barcoding could reveal the regulation and genetic relationship of Actinidia plants， which could be used to select appropriate parents in the kiwifruit crossbreeding.
The identification of A. arguta based SNPs-Barcode
ITS2 was found to have the highest intra- and inter-specific divergence. However， we could not distinguish 51 A. arguta （including A. arguta var. purpurea） only based on the common method. Clement and Donoghue[34] found very low levels of discrimination among closely related species of Viburnum and surmised that the efficacy of barcoding in plants has often been overestimated because of the lack of comparisons among closely related species. However， we found polymorphic sites could be used effectively by applying the SNPs analysis.
SNPs-Barcode， the divergence of signal polymorphism sites in the DNA barcoding sequences， is a accurate and rapid tools compared with previous molecular methods to identify varieties[14， 35]. We selected 3 A. arguta varieties and constructed MCID （Manual cultivar identification diagram） map （Fig.6）， a new strategy employed in identification of cultivar[36]， based on SNP sites of ITS2 sequences. Obviously， 51 A. arguta varieties tested in this paper will be uniquely identified if there are sufficient DNA barcoding sequences featured with stable and abundant polymorphic sites. These findings may allow us to precisely identify the varieties of A. arguta. In conclusion， DNA barcoding could be used in classification and phylogenetic analysis of Actinidia. None of these barcoding regions or combinations in this study could identify these A. arguta varieties based on common method. However， SNPs-Barcode might be a precise and rapid tool to identify closely related A. arguta varieties.
References
[1] LIM TK. Edible Medicinal and Non-Medicinal Plants， Volume 1， Fruits[M]. Springer Netherlands， 2012：5-11.
[2] SOOYEON L， SEUNG HYUN H， JEONGYUN K， et al. Inhibition of hardy kiwifruit （Actinidia aruguta） ripening by 1-methylcyclopropene during cold storage and anticancer properties of the fruit extract[J]. Food Chemistry， 2016， 190： 150-157.
[3] LINDHORST AC， STEINHAUS M. Aroma-active compounds in the fruit of the hardy kiwi （Actinidia arguta ） cultivars Ananasnaya， Bojnice， and Dumbarton Oaks： differences to common kiwifruit （Actinidia deliciosa ‘Hayward’）[J]. European Food Research and Technology， 2016， 242（6）： 967-975.
[4] LAI JJ， LI ZZ， MAN YP， et al. Genetic diversity of five wild Actinidia arguta populations native to China as revealed by SSR markers[J]. Scientia Horticulturae， 2015， 191： 101-107.
[5] KWON SJ， GIAN L， YONGBUM K， et al. Development of 34 new microsatellite markers from Actinidia arguta： intra- and interspecies genetic analysis[J]. Plant Breeding & Biotechnology， 2013， 1（2）： 137-147.
[6] DAI H， LIU C， SUN X， et al. Genetic diversity of Actinidia arguta in northern-east China as revealed by RAPD polymorphism[J]. Acta Horticulturae， 2011， 913（913）： 169-174.
[7] KRESS WJ， WURDACK KJ， ZIMMER EA， et al. Use of DNA barcodes to identify flowering plants[J]. Proceedings of the National Academy of Sciences of the United States of America， 2005， 102（23）： 8369.
[8] STAATS M， ARULANDHU AJ， GRAVENDEEL B， et al. Advances in DNA metabarcoding for food and wildlife forensic species identification[J]. Analytical and Bioanalytical Chemistry， 2016， 408（17）： 4615-4630.
[9] KRESS WJ， ERICKSON DL. A two-locus global DNA barcode for land plants： the coding rbcL gene complements the non-coding trnH-psbA spacer region[J]. Plos One， 2007， 2（6）： e508.
[10] WANG FH， LU JM， WEN J， et al. Applying DNA Barcodes to Identify Closely Related Species of Ferns： A Case Study of the Chinese Adiantum （Pteridaceae）[J]. Plos One， 2016， 11（9）： e0160611.
[11] GROUP CPW. A DNA Barcode for Land Plants[J].Proceedings of the National Academy of Sciences of the United States of America， 2009， 106（31）： 12794. [12] BOLSON M， SMIDT EC， BROTTO ML， et al. ITS and trnH-psbA as Efficient DNA Barcodes to Identify Threatened Commercial Woody Angiosperms from Southern Brazilian Atlantic Rainforests[J]. Plos One， 2015， 10（12）： e0143049.
[13] HUANG Q， DUAN Z， YANG J， et al. SNP typing for germplasm identification of Amomum villosum Lour. Based on DNA barcoding markers[J]. Plos One， 2014， 9（12）： e114940.
[14] GERMANO J， KLEIN AS. Species-specific nuclear and chloroplast single nucleotide polymorphisms to distinguish Picea glauca， P. mariana and P. rubens[J]. Theoretical & Applied Genetics， 1999， 99（1-2）： 37-49.
[15] YAMAMOTO T， NAGASAKI H， YONEMARU JI， et al. Fine definition of the pedigree haplotypes of closely related rice cultivars by means of genome-wide discovery of single-nucleotide polymorphisms[J]. Bmc Genomics， 2010， 11（1）：267.
[16] CHEN S， YAO H， HAN J， et al. Validation of the ITS2 region as a novel DNA barcode for identifying medicinal plant species[J]. Plos One， 2010， 5（1）： e8613.
[17] HUI Y， SONG J， CHANG L， et al. Use of ITS2 Region as the Universal DNA Barcode for Plants and Animals[J]. Plos One， 2010， 5（10）.
[18] LAHAYE R， BANK MVD， BOGARIN D， et al. DNA barcoding the floras of biodiversity hotspots[J]. Proceedings of the National Academy of Sciences， 2008， 105（8）： 2923.
[19] FAY MF， CHASE MW. Plastid rbcL Sequence Data Indicate a Close Affinity between Diegodendron and Bixa[J]. Taxon， 1998， 47（1）：43-50.
[20] SANG T， CRAWFORD D， STUESSY T. Chloroplast DNA phylogeny， reticulate evolution， and biogeography of Paeonia （Paeoniaceae）[J]. American Journal of Botany， 1997， 84（8）： 1120.
[21] TANG Z， BLACQUIERE G， LEUS G. Clustal W and Clustal X version 2.0[J]. Bioinformatics， 2007， 23（21）：2947-2948.
[22] TAMURA K， STECHER G， PETERSON D， et al. MEGA6： Molecular Evolutionary Genetics Analysis version 6.0[J]. Molecular Biology & Evolution， 2013， 30（12）： 2725.
[23] MEYER CP， PAULAY G. DNA barcoding： error rates based on comprehensive sampling[J]. Plos Biology， 2005， 3（12）： 2229-2238.
[24] MEIER R， ZHANG G， ALI F. The use of mean instead of smallest interspecific distances exaggerates the size of the "barcoding gap" and leads to misidentification[J]. Systematic biology， 2008， 57（5）：809-813.
[25] ROSS HA， MURUGAN S， LI WL. Testing the reliability of genetic methods of species identification via simulation[J]. Systematic Biology， 2008， 57（2）：216.
[26] POON LLM， SONG T， ROSENFELD R， et al. Quantifying influenza virus diversity and transmission in humans[J]. Nature Genetics， 2016， 48（2）：195-200. [27] PINO-BODAS R， MARTíN MP， BURGAZ AR， et al. Species delimitation in Cladonia （Ascomycota）： a challenge to the DNA barcoding philosophy[J]. Molecular Ecology Resources， 2013， 13（6）：1058.
[28] LAHAYE R， BANK MVD， BOGARIN D， et al. From the Cover： DNA barcoding the floras of biodiversity hotspots[J]. Proceedings of the National Academy of Sciences of the United States of America， 2008， 105（8）： 2923.
[29] CHEN X， LIAO B， SONG J， et al. A fast SNP identification and analysis of intraspecific variation in the medicinal Panax species based on DNA barcoding[J]. Gene， 2013， 530（1）：39-43.
[30] KHASIM SM， RAMUDU J. Deoxyribonucleic Acid Barcoding of Some Indian Coelogyninae （Orchidaceae）[J]. Proceedings of the National Academy of Sciences India， 2016， 1-8.
[31] GERE J， YESSOUFOU K， DARU BH， et al. Incorporating trnH-psbA to the core DNA barcodes improves significantly species discrimination within southern African Combretaceae[J]. Zookeys， 2013， 365（365）： 129-147.
[32] LFSTRAND SD， SCHNENBERGER J. Molecular phylogenetics and floral evolution in the sarracenioid clade （Actinidiaceae， Roridulaceae and Sarraceniaceae） of Ericales[J]. Taxon， 2015， 64（6）： 1209-1224.
[33] LI J， HUANG H， TAO S. Molecular Phylogeny and Infrageneric Classification of Actinidia （Actinidiaceae）[J]. Systematic Botany， 2002， 27（2）：408-415.
[34] CLEMENT WL， DONOGHUE MJ. Barcoding success as a function of phylogenetic relatedness in Viburnum， a clade of woody angiosperms[J]. BMC Evolutionary Biology， 2012， 12（1）： 73.
[35] COLL F， MCNERNEY R， GLYNN JR， et al. A robust SNP barcode for typing Mycobacterium tuberculosis complex strains[J]. Nature Communications， 2014， 5： 4812.
[36] LIN J， WANG XC， CHANG YH， et al. Development of a novel and efficient strategy for practical identification of Pyrus spp （Rosaceae） cultivars using RAPD fingerprints[J]. Genetics & Molecular Research， 2011， 10（2）：932-942.
Key words Actinidia arguta； DNA barcoding； SNP ； Identification
Actinidia arguta[A. arguta （Sieb. et Zucc.） Planch. Ex Miq.]， A. polygama[A. polygama （Sieb .et Zucc.） Maxim.]and A. kolomikta[A. kolomikta （Rupr. & Maxim.） Maxim.]are native to eastern-most Siberia， China， Korea， Japan， Sakhalin and the Kuril Islands[1-2]. As a result of geographical and climatic adaptation， these plants have good cold resistance， which makes them excellent gene donors for kiwifruit breeding.
A. arguta， also called hardy kiwifruit or mini kiwi， has been more and more popular in the market and has received the attention of researchers and local farmers[3-4]. A. arguta is the third kiwifruit species commercially cultivated followed the A. deliciosa and A. chinensis. Compared with A. deliciosa and A. chinensis， A. arguta is smaller （4-20 g） and can be eaten whole without peeling. Studies suggest that A. arguta can withstand temperatures down to 30 ℃ and have the resistance to bacterial canker disease of kiwifruit （Pseudomonas syringae pv actinidiae）[4]. Besides， the root of hardy kiwifruit has been served as a traditional herbal medicine mainly used for gastric cancer， esophageal cancer and other gastrointestinal cancer in China. New cultivars of kiwifruit， such as "kuilv"， "fenglv" and "jialv"， have been developed by breeding programs in northeastern China， and planting area has been extended year by year. However， the dummy variety has appeared in the nursery stock market duo to commercial profit. The identification of varieties is beneficial to the preservation， development and utilization of these plants. However， studies concerning the identification of these cold resistant kiwifruit are relatively rare. Numerous analytical methods that rely on molecular markers have been developed for research into the genetic diversity of genus Actinidia， such as simple sequence repeat （SSR） markers[4-5] and random amplified polymorphic DNA（RAPD） markers[6]. However， these methods are of limited use in variety identification.
DNA barcoding is a new technique for molecular identification proposed in recent years that uses a standard short genetic region that is universally present in the target lineages and has sufficient sequence variation to discriminate among species[7-10]. DNA barcoding， which was put forward for the first time in 2003 by Canadian anthropologist Paul Hebert， has attracted significant attention around the world. This method was introduced into the study of plants in approximately 2005[7， 9]. The Consortium for the Barcode of Life （CBOL） Plant Working Group identified and recommended the use of chloroplast genes rbcL and matK in 2009[11-12]. However， this recommendation was based on the study of only a relatively small number of species in which multiple individuals were sampled from multiple congeneric species.
Single nucleotide polymorphism（SNP） is a single nucleotide variation in a specific genetic location[13]. SNP is one of the most abundant and stable genetic polymorphisms and is applicable to resolve differences among closely related species[14-15]. In this study， SNP typing method was evaluated using ITS2 markers to identify hardy kiwifruit varieties.
Materials and Methods
Taxon sampling
We collected 180 samples comprising three individuals each from 60 varieties of three species from Northeastern China. These species include A. arguta （including A. arguta var. purpurea， 51 varieties total）， A. kolomikta （6 varieties） and A. polygama （3 varieties）. All the specimens were collected from northeastern China and preserved in silica gel. The varieties and test numbers are shown in Appendix1.
DNA extraction， PCR amplification and sequencing
Genomic DNA was extracted from silica gel-dried leaf material with liquid nitrogen using the Plant Genomic DNA Kit （TIANGEN BIOTECH， Beijing， China）， following the supplier’s protocols. Five regions （ITS， ITS2， matK， rbcL， trnH-psbA） were amplified and sequenced. Each of the five regions used one universal pair of primers for sequence amplification. The volume of the PCR reaction was 30 μl， including： 2×PCR Reagent 15 μl （0.1 U Taq plus polymerase/μl， 500 μM dNTP each， 20 mM Tris-HCl （pH=8.3）， 100 mM KCl， 3 mM MgCl2） produced by Tiangen Biotech （Beijing， China） Co.， Ltd； DNA 2 μl； forward primer （2.5 μM） 1.5 μl， Reverse primer（2.5 μM）1.5 μl； ddH2O 10 μl. The reaction conditions were optimized by gradient PCR and are shown in Table 1. The sequencing reaction were performed by Sangon Biotech （Shanghai， China） Co.， Ltd using the same primers as used for PCR. All samples were sequenced via direct sequencing of PCR products. Sequence reads were assembled using CodonCode Aligner 6.0.2 （CodonCode， Centerville， Massachusetts， U.S.A.）.
Data analysis
Candidate DNA barcodes were aligned by ClustalX using default parameters[21]. Poor quality base calls at the 5’ and 3’ ends of the sequenced PCR products were removed. Kimura-2-Parameter （K2P） distances， conserved sites， variable sites， parsimony-information sites and GC content were computed with MEGA 6.0[22]. "Intra-varieties distances"， "inter-specific distances"， and "intra-specific & inter-variety distances" were calculated using a K2P distance matrix[23-24]. Wilcoxon signed rank tests were performed by IBM SPSS Statistics 21 to compare and screen the genetic diversity of different markers. The distribution of genetic distance was shown using the DNA barcoding gap[18， 23]. Variety identification and phylogenetic relationships among species were analyzed by genetic distance and tree-based method[25]. SNP sites detection were performed using MEGA 6.0.
Results
Amplification， sequencing success and marker features
Amplification efficiency， sequencing success， variable sites and other information can be seen in Table 2.
The amplification efficiency and sequencing quality are important for the evaluation of candidate barcode sequences. Amplification efficiency of each candidate barcode and the tested combinations was 100%. The sequencing rates of different loci were obviously different. The sequencing rate of the ITS region was only 27.78%， which was too low to do further analysis， so the ITS region was rejected as a barcode for identifying wild kiwifruit resources in northeastern China. The remaining candidate barcodes were subjected to further analysis due to their high sequencing rates. Sequence lengths of the four candidate barcodes ranged from 424 to 832 bp. The nuclear gene ITS2 region’s variable sites among three species totaled 39 bp which was much more than the variable sites in the chloroplast genes. The trnH-psbA region’s variable sites were fewest， with only 4 bp. Therefore， the ITS2 region might have more advantages for identifying species with close relationships. The barcoding combinations had more informative sites， which might contribute to the analysis of genetic diversity.
Evaluation of identifying power based on the barcoding gap The barcoding gap was estimated by comparing the intra- and inter-varietal divergences of each candidate locus[12]. In an ideal situation， the genetic variation of a DNA barcode should demonstrate separate， non-overlapping distributions between intra- and inter-varietal samples[16].
Agricultural Biotechnology2018
Our result （Fig. 1） demonstrated that the distributions of all the markers in this paper failed to exhibit distinct gaps at variety level， which means it was difficult for this analysis method to identify these closely related varieties.
However， there was an obvious gap between inter- and intra-specific variations， meaning that species level identification could be completed.
Evaluation of identifying power based on a tree-based method
In this study， for phylogenetic reconstruction we used the Neighbor-Joining （NJ） methods implemented in MEGA version 6.0. Values at each node were assessed by 1 000 bootstrap replicates. For identifying the varieties we chosen the ITS2 region and the combination ITS2+rbcL to construct the NJ tree due to their better performance in inter-varietal genetic distance compared with others.
If the same varietal samples could be grouped into a variety-specific cluster using phylogenetic analysis based on a candidate barcode， it would indicate that this barcode could successfully identify the varieties. ITS2 could not detect variety-specific clusters using phylogenetic analysis （Fig. 2）， which indicated that the intraspecific variation of ITS2 region was too poor to identify varieties. Similarly， the combination of ITS2+rbcL could not complete the identification at the varietal level （Fig. 3）. However， the 3 species were grouped into species-specific clusters successfully， which means the 3 kiwifruit species could be identified based on phylogenetic analysis.
Evaluation of identifying power based on wilcoxon signed rank tests
Usually， wilcoxon signed rank tests were used to analyze the variability of sequences[26-28]. Wilcoxon signed rank tests on combined data showed that ITS2 was the most variable barcode at theinterspecific level， followed by the combination ITS2+rbcL （Table 3）. At the intraspecific level， Wilcoxon signed rank tests showed that rbcL and trnH-psbA had the lowest level of divergence， whereas the highest was provided by ITS2 （Table 4）.
Analysis of SNPs
SNP analysis is an effective method for detecting genetic diversity in plants particularly in very closely related species[29]. All the ITS2 sequences here from the A. arguta varieties were examined for SNPs at the variety level （Fig. 4）. Three differential site at position 135， 220， 243 bp were found， which could identify some cultivars. One polymorphism site was found at 416 bp of matK sequences. rbcL and trnH-psbA have been proved to be an important markers in addressing phylogenetic relationships at various taxonomic levels[19， 30-31]. However， these markers cannot provide even one polymorphism to distinguish individuals of A. arguta. At position 135 bp， "Fenglv"， an A. arguta cultivar， carrys C but "A11-10-2" and "B9-2-1" carry A. Therefore， "Fenglv" could be confirmed uniquely in these three varieties. "A11-10-2" and "B9-2-1" also can be distinguished based on the diversity at the position 220 or 243 bp.
Discussion
The identification of Actinidia based on the common method
In our study， we sampled 3 species and each of them could be identified by phylogenetic trees based on DNA barcodes （Fig. 2； Fig. 3）， which means that interspecific identification may be completed successfully through DNA barcoding. For proving this hypothesis， we constructed phylogenetic tree （Fig. 5） based on matK sequence to evaluate the identification power of Actinidia plants （some sequences from this experiment， others collected from GenBank）.
Previous studies have constructed phylogenetic trees of Actinidia using the sequences of ITS and suggested that DNA barcoding could be used in classification and phylogenetic analysis[32]. Li et al.[33] sequenced ITS and matK gene of 23 Actinidia species. They argued neither ITS nor matK phylogeny supported monophyly of any of the four sect of the genus （Maculatae， Strigosae， Stellatae and Leiocarpae）， which conclusion consistented with our study. Besides， we thought DNA barcoding could reveal the regulation and genetic relationship of Actinidia plants， which could be used to select appropriate parents in the kiwifruit crossbreeding.
The identification of A. arguta based SNPs-Barcode
ITS2 was found to have the highest intra- and inter-specific divergence. However， we could not distinguish 51 A. arguta （including A. arguta var. purpurea） only based on the common method. Clement and Donoghue[34] found very low levels of discrimination among closely related species of Viburnum and surmised that the efficacy of barcoding in plants has often been overestimated because of the lack of comparisons among closely related species. However， we found polymorphic sites could be used effectively by applying the SNPs analysis.
SNPs-Barcode， the divergence of signal polymorphism sites in the DNA barcoding sequences， is a accurate and rapid tools compared with previous molecular methods to identify varieties[14， 35]. We selected 3 A. arguta varieties and constructed MCID （Manual cultivar identification diagram） map （Fig.6）， a new strategy employed in identification of cultivar[36]， based on SNP sites of ITS2 sequences. Obviously， 51 A. arguta varieties tested in this paper will be uniquely identified if there are sufficient DNA barcoding sequences featured with stable and abundant polymorphic sites. These findings may allow us to precisely identify the varieties of A. arguta. In conclusion， DNA barcoding could be used in classification and phylogenetic analysis of Actinidia. None of these barcoding regions or combinations in this study could identify these A. arguta varieties based on common method. However， SNPs-Barcode might be a precise and rapid tool to identify closely related A. arguta varieties.
References
[1] LIM TK. Edible Medicinal and Non-Medicinal Plants， Volume 1， Fruits[M]. Springer Netherlands， 2012：5-11.
[2] SOOYEON L， SEUNG HYUN H， JEONGYUN K， et al. Inhibition of hardy kiwifruit （Actinidia aruguta） ripening by 1-methylcyclopropene during cold storage and anticancer properties of the fruit extract[J]. Food Chemistry， 2016， 190： 150-157.
[3] LINDHORST AC， STEINHAUS M. Aroma-active compounds in the fruit of the hardy kiwi （Actinidia arguta ） cultivars Ananasnaya， Bojnice， and Dumbarton Oaks： differences to common kiwifruit （Actinidia deliciosa ‘Hayward’）[J]. European Food Research and Technology， 2016， 242（6）： 967-975.
[4] LAI JJ， LI ZZ， MAN YP， et al. Genetic diversity of five wild Actinidia arguta populations native to China as revealed by SSR markers[J]. Scientia Horticulturae， 2015， 191： 101-107.
[5] KWON SJ， GIAN L， YONGBUM K， et al. Development of 34 new microsatellite markers from Actinidia arguta： intra- and interspecies genetic analysis[J]. Plant Breeding & Biotechnology， 2013， 1（2）： 137-147.
[6] DAI H， LIU C， SUN X， et al. Genetic diversity of Actinidia arguta in northern-east China as revealed by RAPD polymorphism[J]. Acta Horticulturae， 2011， 913（913）： 169-174.
[7] KRESS WJ， WURDACK KJ， ZIMMER EA， et al. Use of DNA barcodes to identify flowering plants[J]. Proceedings of the National Academy of Sciences of the United States of America， 2005， 102（23）： 8369.
[8] STAATS M， ARULANDHU AJ， GRAVENDEEL B， et al. Advances in DNA metabarcoding for food and wildlife forensic species identification[J]. Analytical and Bioanalytical Chemistry， 2016， 408（17）： 4615-4630.
[9] KRESS WJ， ERICKSON DL. A two-locus global DNA barcode for land plants： the coding rbcL gene complements the non-coding trnH-psbA spacer region[J]. Plos One， 2007， 2（6）： e508.
[10] WANG FH， LU JM， WEN J， et al. Applying DNA Barcodes to Identify Closely Related Species of Ferns： A Case Study of the Chinese Adiantum （Pteridaceae）[J]. Plos One， 2016， 11（9）： e0160611.
[11] GROUP CPW. A DNA Barcode for Land Plants[J].Proceedings of the National Academy of Sciences of the United States of America， 2009， 106（31）： 12794. [12] BOLSON M， SMIDT EC， BROTTO ML， et al. ITS and trnH-psbA as Efficient DNA Barcodes to Identify Threatened Commercial Woody Angiosperms from Southern Brazilian Atlantic Rainforests[J]. Plos One， 2015， 10（12）： e0143049.
[13] HUANG Q， DUAN Z， YANG J， et al. SNP typing for germplasm identification of Amomum villosum Lour. Based on DNA barcoding markers[J]. Plos One， 2014， 9（12）： e114940.
[14] GERMANO J， KLEIN AS. Species-specific nuclear and chloroplast single nucleotide polymorphisms to distinguish Picea glauca， P. mariana and P. rubens[J]. Theoretical & Applied Genetics， 1999， 99（1-2）： 37-49.
[15] YAMAMOTO T， NAGASAKI H， YONEMARU JI， et al. Fine definition of the pedigree haplotypes of closely related rice cultivars by means of genome-wide discovery of single-nucleotide polymorphisms[J]. Bmc Genomics， 2010， 11（1）：267.
[16] CHEN S， YAO H， HAN J， et al. Validation of the ITS2 region as a novel DNA barcode for identifying medicinal plant species[J]. Plos One， 2010， 5（1）： e8613.
[17] HUI Y， SONG J， CHANG L， et al. Use of ITS2 Region as the Universal DNA Barcode for Plants and Animals[J]. Plos One， 2010， 5（10）.
[18] LAHAYE R， BANK MVD， BOGARIN D， et al. DNA barcoding the floras of biodiversity hotspots[J]. Proceedings of the National Academy of Sciences， 2008， 105（8）： 2923.
[19] FAY MF， CHASE MW. Plastid rbcL Sequence Data Indicate a Close Affinity between Diegodendron and Bixa[J]. Taxon， 1998， 47（1）：43-50.
[20] SANG T， CRAWFORD D， STUESSY T. Chloroplast DNA phylogeny， reticulate evolution， and biogeography of Paeonia （Paeoniaceae）[J]. American Journal of Botany， 1997， 84（8）： 1120.
[21] TANG Z， BLACQUIERE G， LEUS G. Clustal W and Clustal X version 2.0[J]. Bioinformatics， 2007， 23（21）：2947-2948.
[22] TAMURA K， STECHER G， PETERSON D， et al. MEGA6： Molecular Evolutionary Genetics Analysis version 6.0[J]. Molecular Biology & Evolution， 2013， 30（12）： 2725.
[23] MEYER CP， PAULAY G. DNA barcoding： error rates based on comprehensive sampling[J]. Plos Biology， 2005， 3（12）： 2229-2238.
[24] MEIER R， ZHANG G， ALI F. The use of mean instead of smallest interspecific distances exaggerates the size of the "barcoding gap" and leads to misidentification[J]. Systematic biology， 2008， 57（5）：809-813.
[25] ROSS HA， MURUGAN S， LI WL. Testing the reliability of genetic methods of species identification via simulation[J]. Systematic Biology， 2008， 57（2）：216.
[26] POON LLM， SONG T， ROSENFELD R， et al. Quantifying influenza virus diversity and transmission in humans[J]. Nature Genetics， 2016， 48（2）：195-200. [27] PINO-BODAS R， MARTíN MP， BURGAZ AR， et al. Species delimitation in Cladonia （Ascomycota）： a challenge to the DNA barcoding philosophy[J]. Molecular Ecology Resources， 2013， 13（6）：1058.
[28] LAHAYE R， BANK MVD， BOGARIN D， et al. From the Cover： DNA barcoding the floras of biodiversity hotspots[J]. Proceedings of the National Academy of Sciences of the United States of America， 2008， 105（8）： 2923.
[29] CHEN X， LIAO B， SONG J， et al. A fast SNP identification and analysis of intraspecific variation in the medicinal Panax species based on DNA barcoding[J]. Gene， 2013， 530（1）：39-43.
[30] KHASIM SM， RAMUDU J. Deoxyribonucleic Acid Barcoding of Some Indian Coelogyninae （Orchidaceae）[J]. Proceedings of the National Academy of Sciences India， 2016， 1-8.
[31] GERE J， YESSOUFOU K， DARU BH， et al. Incorporating trnH-psbA to the core DNA barcodes improves significantly species discrimination within southern African Combretaceae[J]. Zookeys， 2013， 365（365）： 129-147.
[32] LFSTRAND SD， SCHNENBERGER J. Molecular phylogenetics and floral evolution in the sarracenioid clade （Actinidiaceae， Roridulaceae and Sarraceniaceae） of Ericales[J]. Taxon， 2015， 64（6）： 1209-1224.
[33] LI J， HUANG H， TAO S. Molecular Phylogeny and Infrageneric Classification of Actinidia （Actinidiaceae）[J]. Systematic Botany， 2002， 27（2）：408-415.
[34] CLEMENT WL， DONOGHUE MJ. Barcoding success as a function of phylogenetic relatedness in Viburnum， a clade of woody angiosperms[J]. BMC Evolutionary Biology， 2012， 12（1）： 73.
[35] COLL F， MCNERNEY R， GLYNN JR， et al. A robust SNP barcode for typing Mycobacterium tuberculosis complex strains[J]. Nature Communications， 2014， 5： 4812.
[36] LIN J， WANG XC， CHANG YH， et al. Development of a novel and efficient strategy for practical identification of Pyrus spp （Rosaceae） cultivars using RAPD fingerprints[J]. Genetics & Molecular Research， 2011， 10（2）：932-942.