摘 要

鹅掌楸为孑遗植物，目前仅存鹅掌楸（Liriodendron chinense(Hemsl.) Sarg.）和北美鹅掌楸（Liriodendron tulipifera Linn.）两个种。鹅掌楸目前仅存于中国陕西汉中汉水流域以南，福建武夷山脉以西，南岭山脉东段向西至贵州、云南境内；以及越南北部的老街地区也有自然分布；北美鹅掌楸分布于美国东部密西西比河以东，沿阿巴拉契山脉两侧，向北延伸至加拿大东南部五大湖区的安大略省多伦多地区。鹅掌楸属植物，不仅是研究东亚和北美地区植物区系演化等基础科学问题重要研究材料，也是研究和利用鹅掌楸属种间杂种优势、遗传图谱构建和开展分子育种的重要基因资源。国内已经报道过鹅掌楸遗传连锁图谱的构建，但是由于已有图谱上的锚定的分子标记数目相对较少，遗传图距也较大，制约了鹅掌楸QTL定位和分子辅助育种等后续研究。限制性位点关联DNA测序（RAD-seq）技术是一种新型的测序方法，能大幅降低基因组的杂合度，快速有效获得作图群体中成千上万个SNP标记，以构建高密度遗传连锁图谱。

本研究以北美鹅掌楸‘NK’为母本，以中国鹅掌楸‘LS’为父本，选取198株杂交F₁子个体为作图群体。选用限制性内切酶EcoRI对杂交鹅掌楸的2个杂交亲本和198个子代在内的200个个体基因组DNA进行酶切，构建了RAD文库并进行了RAD-seq测序。采用读长为91bp的双末端测序，2个亲本的平均测序深度为2×，198个子代的平均测序深度为0.8×，数据平均产出量为1.94Gb，共得到了约387.21Gb的数据。使用Stacks软件将每个样品的RAD-reads进行聚类，比对亲本RAD-tag，SNP识别与基因型分型等生物信息学分析后，共检出22,019个候选群体SNP位点。然后将得到的SNP位点进行卡方检验和缺失率检验。通过过滤，得到了4,223个SNP位点。将符合孟德尔遗传分离比的4,223个SNP标记对应的RAD-tag序列与鹅掌楸参考基因组序列进行比对，最终获得3,501个候选SNP标记。

结果表明，RAD-seq策略能够作为一个快速有效的手段来进行大规模分子标记的开发，用于构建鹅掌楸等林木的高密度遗传图谱。这将为进一步开展QTL定位、分子辅助育种等研究提供科学依据。

关键词：鹅掌楸；北美鹅掌楸； RAD-seq；SNP标记；遗传连锁图谱

Genomic SNP Markers Development based on the parents and Interspecies hybridization Population of Liriodendron

Abstract

Liriodendron, an endangered woody plant genus, has only two species, Liriodendron chinense (Hemsl.)Sarg. And Liriodendron tulipifera Linn, remained. Liriodendron chinense presently naturally scattered in the south of China and extend to the North of Vietnam. Liriodendron tulipifera is distributed in the southeast of United States of America, i.e. along with the Appalachian Mountains from the north of Florida to the Great Lakes areas in Canada. The germplasm of Liriodendron is the important material both in the basic research of molecular genetics and tree breeding, and the wood production. Although the genetic linkage maps construction, quantitative trait loci (QTL) allocating, and marker-assisted breeding are to be caught the attention, but limited valuable information available on Liriodendron. Restriction site-associated DNA sequencing (RAD-seq) technology, a new sequencing approach, can significantly reduce the genomic heterozygosity and to provide thousands of SNP markers in mapping population. It makes the possible to construct high-density genetic linkage maps on Liriodendron.

In this study, an F₁ mapping population was produced by the hybridization with L. tulipifera L(♀) and L. chinense (♂). RAD libraries were constructed using genomic DNA from both two parents and 198 individuals of the progeny population, and the genomic DNAs were digested with the restriction enzyme EcoRI. We performed RAD sequencing with paired end sequencing read length in 91bp, a total of 387.21Gb data obtained with average sequencing depth 2× in parents and 0.8× for he progeny, and the average production was 1.94Gb. A total number of 22,019 candidate SNP were identified through the analysis of clustering each sample’s RAD-reads, the comparison of the parent RAD-tag, SNP identification and genotyping by a software called Stacks. Then, the obtained SNP were tested by Chi-square test and the test of miss rate, there were 4,223 SNPs acquired after filtering previous SNPs. At last, a total number of 3,501 candidate SNP markers were obtained through mapping the 4,223 SNP markers corresponding with RAD-tag sequence that follow Mendel’s segregation ratio to the reference genome of L. chinense. The results demonstrates that RAD-seq is a fast and effective technique capable of developing a large number of SNPs for constructing genetic linkage maps on Liriodendron. The high-density linkage maps of L. tulipifera and L. chinense could be convenient for developing QTLs allocating, and molecular marker-assisted breeding.

keywords: Liriodendron tulipifera；Liriodendron chinense；RAD-seq；SNP maker；Genetic linkage map

目 录

第一章 前 言 1

1.1 鹅掌楸概述 1

1.2 DNA分子标记 3

1.2.1 DNA分子标记介绍 3

1.2.3 二代测序技术及RAD-seq 7

1.3 遗传连锁图谱的构建 15

1.3.1 遗传连锁图谱介绍 15

1.3.2 作图群体 16

1.3.3 作图软件 16

1.4 SNP标记开发的技术路线 17

1.5 本研究的目的与意义 18

第二章 SNP标记开发 19

2.1 实验材料 19

2.2.1 基因组DNA的提取 19

2.2.2 DNA质量检测 20

2.2.3 选择构建RAD文库的酶 20

2.2.4 RAD文库建立 21

2.2.5 RAD序列分析、SNP识别及基因型分型 21

2.3 结果与分析 24

2.3.1 选择合适的酶构建RAD文库 24

2.3.2 RAD的序列分析 26