鬼针草属（Bidens L.）植物为菊科（Asteraceae）一年或多年生草本，属内有5种为药用植物，其中，金盏银盘Bidens biternata为三九感冒灵的原料药材之一。鬼针草属药材在广东和广西等地方中药材标准均有收载，然而，地方标准中对其基原植物的规定存在差异，且存在学名和中文名不对应、与《中国植物志》记载不一致等问题，严重阻碍了鬼针草属药材的规范使用。鬼针草属植物性状特征相似、化学成分相近，传统鉴定方法较为困难，因此，亟需开发一个准确有效的方法来鉴定鬼针草属药用植物。叶绿体作为植物特有的细胞器，其结构高度保守，携带许多具有重要遗传价值的基因信息，可用于特异性条形码筛选、超级条形码鉴定和系统进化等研究。高等植物的线粒体基因组结构虽然复杂，但其基因组成较为保守，可以用于物种的遗传进化等研究。

本研究分析了通用条形码鉴定鬼针草属药用植物的可行性，利用Illumina高通量测序技术对6种鬼针草属药用植物的叶绿体和线粒体基因组进行测序、注释及比较分析，以探讨同属近缘物种的叶绿体和线粒体基因组结构变化；基于叶绿体基因组筛选了高变区作为特异性条形码对6种鬼针草属药用植物进行鉴定；利用细胞器基因组构建系统进化树，以期为鬼针草属药用植物的物种鉴定、用药安全、规范使用以及种质资源保护奠定基础。本研究的主要结果如下：

1.鬼针草属药用植物叶绿体基因组解析及其鉴定

1）利用通用条形码ITS/ITS2和psbA-trnH序列对6种鬼针草属药用植物（金盏银盘B. biternata、婆婆针B. bipinnata、鬼针草B. pilosa var. pilosa、白花鬼针草B. pilosa var. radiata、小花鬼针草B. parviflora和狼杷草B. tripartita）共21份样本进行扩增测序，K2P遗传距离结果显示ITS/ITS2和psbA-trnH序列在鬼针草属药用植物内不满足种间遗传距离远大于种内遗传距离的条件，NJ树结果显示鬼针草属药用植物存在多个物种聚为一支的情况，因此，ITS/ITS2和psbA-trnH序列不能用于这6种鬼针草属药用植物的鉴定研究。

（2）通过高通量测序获得的这6种鬼针草属药用植物共12个样本的叶绿体基因组均为典型的四分体结构，包括一个大单拷贝（LSC）区、一个小单拷贝（SSC）区和两个反向重复区（IR）区，序列长度为150,489~151,635 bp，总GC含量均为37.5%。均含有130个基因，包括85个蛋白编码基因、37个tRNA基因和8个rRNA基因，有17个基因在IR区重复。鉴定到60~71个简单重复序列（SSRs），同一物种不同样本鉴定到的SSRs不完全相同，且SSRs具有A/T碱基偏好性。密码子使用分析显示大多数氨基酸密码子都具有偏好性。

（3）叶绿体全基因组比对分析表明，6种鬼针草属药用植物叶绿体基因组相对保守，且非编码区的序列变异程度高于编码区。同源性分析显示金盏银盘和婆婆针叶绿体基因组序列的同源性很高，鬼针草和白花鬼针草叶绿体基因组序列的同源性很高。

（4）利用mVISTA和DnaSP分析结果筛选出7个叶绿体基因组高变区作为特异性条形码，用6种鬼针草属药用植物共31份样品对7个高变区的鉴定能力进行验证，成功扩增出其中4个特异性条形码（ndhA，accD，clpP和ccsA-ndhD）。3个（ndhA，accD和ccsA-ndhD）特异性条形码仅能鉴定出狼杷草和小花鬼针草，1个（clpP）特异性条形码建树聚类结果不清晰。因此，基于叶绿体基因组筛选的特异性条形码不适用于这6种鬼针草属药用植物的鉴定研究。基于叶绿体基因组共有蛋白编码基因序列构建的系统发育树结果示鬼针草属物种聚为一支，但婆婆针Y18102样本未与其它3个样本聚在一起。基于叶绿体全基因组序列构建的ML系统进化树显示，6种鬼针草属药用植物聚为一大分支，其中金盏银盘3份样本聚为一支，婆婆针4份样本聚为一支，且金盏银盘和婆婆针互为姐妹分支；鬼针草和其变种白花鬼针草聚为一支，此聚类结果支持《Flora of China》将鬼针草和白花鬼针草合并为同一个种；小花鬼针草和狼杷草各自单独聚为一支。系统发育树的聚类结果证明，叶绿体全基因组序列可以作为超级条形码鉴定6种鬼针草属药用植物。

2.鬼针草属药用植物的线粒体基因组解析

本研究首次解析了6种鬼针草属药用植物的线粒体基因组，其大小在183 kb（鬼针草）和216 kb（狼杷草）之间，6个物种的线粒体基因组GC含量均为45%左右，注释的CDS数目相近。除金盏银盘和婆婆针2个物种的线粒体基因组SSR和长重复序列完全一致外，其余4个物种SSR和长重复的数目与类型均不同。利用菊科已公布的植物线粒体基因组和本研究获得的线粒体基因组构建系统进化树，结果显示金盏银盘和婆婆针聚为一支，鬼针草和白花鬼针草聚为一支，6种鬼针草属药用植物并未聚为一大支，说明利用线粒体全基因组序列鉴定鬼针草属药用植物不可行。基于线粒体基因组19个共有蛋白编码基因构建的系统发育树显示6种鬼针草属药用植物聚为一支，说明共有蛋白编码基因序列具有系统进化分析的潜力，但仍需要更多数据支持和验证。高等植物线粒体基因组数据匮乏，本研究丰富了高等植物线粒体基因组数据库，为菊科药用植物的线粒体基因组分析提供数据支持。

本研究通过以上分析发现通用条形码、基于叶绿体基因组的特异性条形码、线粒体基因组均不能鉴定6种鬼针草属药用植物，叶绿体全基因组可以作为超级条形码鉴定6种鬼针草属药用植物，为鬼针草属药用植物的分子鉴定、用药安全和规范使用以及种质资源保护提供数据支持。

Plants of Bidens L. are annuals or perennials of Asteraceae. A total of ten species in the genus, and five of them are medical plants recorded in Guangdong, Guangxi, Shandong and other local standards of Chinese medicinal materials. Bidens biternate of this genus is one of the medicinal materials of 999 Ganmaoling. However, there are some problems in records about these medicinal plants of these local standards, for example, the scientific name is not consistent with their Chinese name and the records in flora of China. Bidens plants have similar morphologies and chemical components, so it is difficult to identify them by traditional identification methods. These problems seriously hinder the development of Bidens medicinal plants. Therefore, it is an urgent need to develop an accurate and effective method to identify the medicinal plants of Bidens. The chloroplast genomes are highly conserved in gene order and content, and they are thus ideal research models for the study of specific barcode and super barcode, species identification and evolution. The structure of mitochondrial genome in higher plants is more complex than chloroplast genome, but ITS/ITS2 gene composition is relatively conserved, which can be used for the study of genetic evolution of plants.

In this study, DNA barcodes were used to identified Bidens medicinal plants. Illumina sequencing was used to sequence the chloroplast and mitochondrial genomes of six Bidens medicinal plants and the complete organelles genomes were assembled, annotated and analyzed. Specific barcodes were screened and ML phylogenetic tree were constructed based on the chloroplast genome to identified these medicinal plants. The results obtained in this study could provide valuable genetic resource for species identification, resource conservation and phylogenetic relationship analysis of the Bidens medicinal plants. The main results of this study are as follows:

1. Comparative and identification analysis of Bidens medicinal plants by chloroplast genomes

(1) ITS/ITS2 and psbA-trnH sequences were used to identified six Bidens medicinal plants (B. biternate, B. bipinnata, B. pilosa var. pilosa, B. pilosa var. radiata, B. parviflora and B. tripartita). A total of 21 samples of these six species were amplified and sequenced by sanger sequencing. The K2P genetic distance of intraspecific and interspecific of ITS/ITS2 and psbA-trnH sequences within these species were similar to each other. NJ phylogenetic tree showed that different species of Bidens clustered in one branch. Therefore, ITS/ITS2 and psbA-trnH sequences could not identify the six medicinal plants of Bidens.

(2) The complete chloroplast genomes of 12 samples of B. biternate, B. bipinnata, B. pilosa var. pilosa, B. pilosa var. radiata, B. parviflora and B. tripartite were typical circular structure containing a pair of inverted repeats (IRs) separated by the large single copy (LSC) and the small single copy (SSC) regions. Chloroplast genomes ranged from 150,489 bp to 151,635 bp in length. The GC contents in the whole chloroplast genomes is 37.5%, which was highly conserved in Bidens species. A total of 130 genes, including 85 protein-coding genes, 37 tRNAs and eight rRNAs were annotated, 17 of which were duplicated in the IRs. The number of SSRs ranged from 60 to 71 in the 12 chloroplast genomes, which was not exactly the same in one species. The codon usage frequency was estimated and most of the amino acid codons show A/T base preferences. The relative synonymous codon usage (RSCU) value increased with the increase in the quantity of codons which coded for a specific amino acid.

(3) The comparative analysis showed that the chloroplast genomes of six Bidens species were highly conserved, and the coding-gene sequences were more conserved than the noncoding sequences. Co-linear and homology analysis showed that the sequences of B. biternate and B. bipinnata are highly homologous, and B. pilosa var. Pilosa and B. pilosa var. radiata are highly homologous.

(4) Seven specific barcodes were screened base on the comparison of 12 complete chloroplast genomes, and four specific barcodes (ndhA, accD, clpP and ccsA-ndhD) were successfully amplified in these six species with 31 samples. Three specific barcodes (ndhA, accD and ccsA-ndhD) could identify B. parviflora and B. tripartite, and one specific barcodes (clpP) could not identify any species. Therefore, the specific barcodes were not suitable for the identification study of the six Bidens species. The result of ML phylogenetic tree constructed by the complete chloroplast genomes showed that six Bidens species clustered in a clade. Three B. biternate samples clustered in one branch, and four B. bipinnata clustered in one branch. B. pilosa var. pilosa and B. pilosa var. radiata clustered in one branch, which supported the result of Flora of China that the two species merged into one species. Therefore, the complete chloroplast genome could be used as a super barcode to identify six Bidens species.

2. Mitochondrial genome analysis of Bidens medicinal plants

In this study, the mitochondrial genomes of six Bidens species were analyzed for the first time. The length of complete mitochondrial genomes was between 183 kb and 216 kb. The GC contents in the whole mitochondrial genomes was approximately 45% and the number of coding sequence was similar among these six species. Apart from B. biternate and B. bipinnata, the number and type of SSRs and long repeats detected in mitochondrial genomes were different in the other four species. ML phylogenetic tree was constructed based on the published mitochondrial genomes of Asteraceae plants and the mitochondrial genomes obtained in this study. The results showed that B. biternate and B. bipinnata clustered in one branch, and B. pilosa var. pilosa and B. pilosa var. radiata clustered in one branch. Six Bidens species were not clustered in a clade. So, it is not feasible to identify Bidens medicinal plants by using the mitochondrial genome sequence. This study enriched the mitochondrial genome database of higher plants and was benefit for mitochondrial genome analysis of Asteraceae medicinal plants.

The results of this study showed that the two DNA barcodes (ITS/ITS2 and psbA-trnH), special barcodes based on complete chloroplast genome and mitochondrial genome could not be used to identified these six Bidens species. The complete chloroplast genome could be used as super barcode to identified these six Bidens species. The results could enrich the organelle genome database of Asteraceae medicinal plants and provide foundation for molecular identification of Asteraceae plants.