Development Of DNA Barcodes For Citrus And Its Related Genera And Their Application In Species Identification

China is one of the most important origin centers of the genus Citrus and is rich in citrus germplasm resources in the world. The genus Citrus belongs to the subfamily Aurantioideae of Rutaceae family. The genus Citrus L. was established by Carl Linnaeus in 1753. Since then, scholars both in China and abroad have carried out a series of studies on the taxonomy of Citrus using morphology, cytology, and molecular data. In the current literature several classification systems for the genus Citrus have been suggested, of which the well-recognized are Swingle’s 16 species system and Tanaka’s 159 species system. In addition, Chinese scholar Tseng has also suggested a system composed of 30 species. Up to now, however, because of hybridization, apomixes and asexual variation, the number of species, inter-specific relationship and classification of the genus Citrus L. are still unresolved scientific questions.DNA barcoding is a technique for identifying and characterizing species of organisms using a short DNA region, which was proposed by Canadian zoologist Paul Hebert in 2003. DNA barcodes compared with the traditional morphological data can rapidly identify species based on DNA molecular evolution principles, and have many advantages in species identification, such as high accuracy, high efficiency, and not influenced by environmental factors and development stage of materials. Since the birth of DNA barcoding technology, the mitochondrial cytochrome c oxidase subunitⅠ(COⅠ) fragment as the barcoding sequence has been successfully used to identify species, find new species and cryptic species in insects, birds, fish and other animals. Unfortunately, CO I gene has a very slow evolutionary rate in plants, which makes it unsuitable for DNA barcoding in plants. Until now, plant barcoding is still at the stage of searching for a suitable locus and kinds of candidates. Many researchers propose a single barcoding region or combinations of regions from the chloroplast genome. At present, a well-characterized plant locus that meets all necessary criteria is lacking.In the present study, the 59 biological types of 6 genera belonged to the true citrus fruit plants were selected as the experimental materials. The chloroplast encoded genes (matK, rpoB, rpoC1, rbcL) and chloroplast intergenic sequences (trnH-psbA, trnG-trnS, psbH-petB, trnL-trnF), as well as nuclear ITS and its first and second space sequence (ITS 1, ITS2) were analysed as the candidate plant barcoding regions. The MEGA 5.05 software was used to calculate the basic composition, the frequency of DNA mutation and the ratio of transition/transversion. Additionally, SPSS17.0 software was used to analyse the sequence divergence through wilcoxon test. In our study, the distribution of intra-specific and inter-specific genetic distance was compared to identify the species of Citrus Swingle, Tanaka systems and the species of the Citrus and its related genera using the single fragment and the combined fragments respectively. The results of this study are as follows:1. Using Poncirus polyandra, P.trifoliata var. monstrosa, P.trifoliata as outgroups, the NJ trees of the Swingle system and Tanaka system with a single fragment were constructed to compare the identification rate. The results showed that the average identification ratio of nuclear ITS and its spacer sequence (ITS1, ITS2) was the highest, followed by the chloroplast intergenic sequence (trhH-psbA, trnG-trnS, psbH-petB, trnL-trnF), and the chloroplast encoded genes (matK, rpoB, rpoC1, rbcL) were third. In Swingle system, the strongest capability of identification was ITS 1 with a ratio of 100%; psbH-petB and ITS were second, and the rate was 93.8%; the rpoC1 was lowest, only 12.5%. In Tanaka system, the identification rate of psbH-petB and ITS were 100%, and the rate of rpoC1 was 14.8%.2. Using Murraya paniculata, Atalantia buxifolia, and Severinia buxifolia as outgroups, the NJ trees of the Citrus and its related genera with each fragments were constructed to compare the identification rate. The results showed that the identification rates of ITS, psbH-petB and matK were 59.3%,55.9% and 50.8% respectively, and the remaining fragments were below 50%. ITS was the highest, rpoC1 was lowest with only 23.7%. So because of the highest identification rate in their alike type, the nuclear gene ITS, chloroplast intergenic sequence psbH-petB, chloroplast encode gene matK were the important fragments in the DNA barcode of the Citrus and its related genera.3. Following the proposal of the third international barcode conference and the combined schemes of the predecessors, as well as the the identification results of the single fragment, matK+rbcL, matK+trnH-psbA, matK+ITS, trnH-psbA+ITS, trnG-trnS+ITS2, matK+ITS2, matK+psbH-petB+trnG-trnS, matK+ITS+trnG-trnS, matK+rpoB+rpoC1 fragments were selected together as the objects in our research. Using Poncirus polyandra, P.trifoliata var. monstrosa, P.trifoliata as outgroups, the phylogenies NJ trees of the Swingle system and Tanaka system with the combined fragments were constructed to compare the identification rate. The results showed that, in Swingle system, the rate of trnH-psbA+ITS was highest, reached 100%. The matK+trnH-psbA、matK+ITS and trnG-trnS+ITS2 were second with 93.8%, and matK+rpoB+rpoC1 was lowest with 81.3%. In Tanaka system, the rate of matK+ITS、trnG-trnS+ITS2 and matK+ITS+trnG-trnS were highest, reached 96.3%. However, the rate of trnH-psbA+ITS and matK+trnH-psbA was 74.1% and 51.9% respectively. So the rate of matK+ITS and trnG-trnS+ITS2 were higher than the other combined fragments to identify the speciese of Swingle system and Tanaka system.4. Using Murraya paniculata, Atalantia buxifolia, and Severinia buxifolia as outgroups, the NJ trees of 59 species of the Citrus and its related genera with combined fragments were constructed to compare the identification rate. The results showed that:the rate of matK+ITS+trnG-trnS was the highest, reached 83.1%, and matK+ITS was second with 81.4%. There was little difference of the identification rate between matK+ITS and marK+ITS+trnG-trnS, but there wasn’t a fragment of trnG-trnS in matK+ITS, which made it more efficient in the identification because of its shorter sequence length.In conclusion, in the identification of Swingle and Tanaka system of the Citrus and its related genera, ITS、psbH-petB as single fragment had the highest identification rate, and ITS、psbH-petB、matK were important fragments in the research of Citrus DNA barcode, which had a higher identification rate in the same category. Compared the combined fragments with the single fragments, the identification rate of combined fragments were greatly improved. Compared matK+ITS with other combined fragments, matK+ITS was more suitable for the research of Citrus and its related genera of plant, which had the characteristics of high identification rate and shorter sequence length. From what we have studied above, we can draw a conclusion that the DNA Barcode technique can be used to identify the species of Citrus and its related genera as a new species-identification method.