海洋生态与环境科学重点实验室知识库

在全球气候变化的背景下，生物种群面临快速变化的环境，因此探讨生物复杂性状在环境因子变化的驱动下快速适应性进化的遗传机制，查明自然种群快速适应性进化的遗传变异的来源以及进化轨道，对于理解生物种群早期适应性进化的机制、评估物种的进化潜力以及生物资源的保护具有非常重要的科学意义。本文的研究对象刀鲚（Coilia nasus Temminck et Schlegel，1846）隶属于鲱形目（Cluperiformes）、鳀科（Engraulidae）、鲚属（Coilia），是我国江海洄游型经济鱼类之一。刀鲚存在两种不同的生活史形式。一种是洄游型刀鲚，广泛分布在西北太平洋的沿岸和河口地区，春季溯河洄游到淡水产卵。另一种是淡水定居型刀鲚，包括湖鲚（C. nasus taihuensis）和短颌鲚（C. brachygnathus），终生生活在长江中下游的附属湖泊里，例如太湖、巢湖、鄱阳湖和洞庭湖等，并且是这些湖泊生态系统的优势种。由于人类活动的影响，包括捕捞强度超过资源恢复能力、修建水坝隔断鱼群上溯产卵通道、水域环境污染等，刀鲚种群的数量锐减，并且出现小型化的现象。目前刀鲚被世界自然保护联盟公布的最新版《濒危物种红色名录》列为“濒危”物种。洄游型和定居型刀鲚生活在不同的环境中，两者在基因组上某些区域可能已经发生分化。厘清刀鲚种群不同的生态型之间的遗传关系，保护不断下降的长江洄游型刀鲚资源已经迫在眉睫。然而目前关于刀鲚分类学方面的研究仍然存在分歧，特别是关于短颌鲚的分类地位一直存在较大争议，定居型刀鲚对淡水的生态适应遗传机制仍未查明，极大地限制了其管理和保护工作。随着二代测序技术的不断完善，在全基因组层面获取成百上千上万个分子标记变成了现实。本研究采用简化基因组测序（RAD-seq）技术，从基因组层面精确解析刀鲚种群不同的生态型之间的遗传关系和动态演化历史，揭示淡水定居型刀鲚对淡水快速适应性进化的遗传机制，阐明洄游型刀鲚种群的遗传结构和本地适应性机制。相关研究结果将为我国濒危物种洄游型刀鲚的资源评估、有效管理与保护提供参考依据。主要研究结果如下：

    1、采用28,691个全基因组层面的单核苷酸多态性（SNP）位点和196条线粒体部分基因组序列（2,916 bp）研究了刀鲚种群三种生态型之间的遗传关系，推断洄游型刀鲚入侵淡水环境的历史。基于AMOVA的分析结果表明，洄游型刀鲚与湖鲚组群间的分化比组群内种群间的分化小，暗示两者之间的遗传分化比较小（FCT = 0.022）。洄游型刀鲚和湖鲚与短颌鲚组群间的分化比组群内种群间的分化大，暗示两者之间的遗传分化比较大（FCT = 0.648）。在196个刀鲚样本中，单倍型网络图将80个单倍型主要分为两个谱系。谱系I包含51个单倍型，主要由洄游型刀鲚和淡水定居型湖鲚种群的个体构成，同时包含8个淡水定居型短颌鲚个体。谱系II包含29个单倍型，主要由淡水定居型短颌鲚种群的个体构成，同时包含1个淡水定居型湖鲚种群的个体。两个主体单倍型之间存在5个核苷酸突变（c. 0.17%）。结合形态学的证据，确认了短颌鲚（Coilia brachygnathus）的物种有效性。根据贝叶斯因子物种界定分析（BFD*）结果，淡水定居型湖鲚和洄游型刀鲚遗传差异不显著。邻接树（NJ tree）、主成分判别分析（DAPC）和ADMIXTURE分析结果进一步证实了贝叶斯因子物种界定分析结果。DIYABC分析结果显示洄游型刀鲚向淡水环境入侵大约发生在10,000年前，与古地理事件——末次冰盛期是一致的。研究结果强烈暗示短颌鲚与另外两个类群（洄游型刀鲚和湖鲚）在遗传上是隔离的，因此短颌鲚是一个有效物种。

    2、利用种群基因组学方法，通过比较四组淡水定居型与洄游型种群刀鲚，揭示了其对淡水快速平行适应性进化的遗传机制。基于费希尔精确检验（Fisher’s exact test）和Pcadapt方法在四个种群对共检出1,147个离散位点，其中1101个（96%）离散位点分布在LG6和LG22染色体上，在种群对之间形成相似的基因岛。在淡水定居型种群中，大部分SNP位点的淡水等位基因频率近乎固定。但是，在洄游型种群中淡水等位基因频率很低，平均值为0.12（SD = 0.07）。在LG6和LG22染色体上存在两个大的（31.54 Mb和21.97 Mb）染色体倒置，表现出与淡水适应的平行关系。在不同种群对之间，两个染色体倒置区域的等位基因频率变化剧烈。基于LG6和LG22染色体上倒置区的所有SNP位点，洄游型和淡水定居型种群之间F_ST平均值为0.37，比基于中性SNP位点的F_ST（平均值0.03）高出一个数量级。高的遗传分化水平，暗示两个染色体倒置区域在洄游型和定居型两种生态型之间受到歧化选择。LG6和LG22两个染色体倒置共覆盖超过50 Mb，占整个基因组大约6%，包含1,800多个基因。两个染色体倒置内的基因富集分析在生物过程分类中分别筛选到72（LG6）和73（LG22）个显著富集的GO terms（p < 0.01）。基因富集分析表明显著富集的基因与代谢过程、免疫调节、生长、性成熟、渗透压调节等有关，这些基因很有可能是导致洄游型和定居型刀鲚之间在形态、生理和行为方面差异的主要原因。有益的固有遗传变异的存在，海洋和淡水栖息环境的剧烈差异以及大的有效群体对应的高效选择作用，可能导致基因组上快速的平行适应性进化。研究结果显示，染色体倒置可能在刀鲚快速平行生态适应性进化中发挥重要作用，固有遗传变异是适应性遗传变异的重要来源。

    3、基于简化基因组测序技术，对中国洄游型刀鲚4个地理种群共95个个体进行了种群基因组学研究。经过严格质量过滤后，得到167,905个高质量SNP位点。遗传结构分析结果显示4个洄游型种群可以分为3个遗传组群，即鸭绿江口和辽河种群为一组，长江口种群、富春江种群各自为一组。全基因组范围内洄游型种群之间的F_ST比较低（平均值0.06）但是显著。归群分析结果显示，长江口种群回归正确率最高达到100%，富春江种群次之为91.7%，鸭绿江口和辽河种群最低分别为79.2%和73.1%。离散位点SNP的基因注释结果发现多个与生长发育、能量代谢、免疫反应、跨膜运输、信号转导、渗透压调节相关的功能基因，表明这些基因可能参与了迁徙过程。因此，笔者建议对洄游型刀鲚种群进行原地保护，保护其洄游路线，以维持洄游型刀鲚的遗传多样性和进化潜力。

    综上，利用简化基因组测序，通过单核苷酸多态性分子标记和线粒体部分基因组分子标记，结合贝叶斯因子界定方法和形态学的证据，确认了短颌鲚（Coilia brachygnathus）的物种有效性。基于种群基因组学的技术手段，阐明了淡水定居型刀鲚对淡水快速平行适应的遗传机制，染色体倒置在淡水快速适应性进化过程中发挥了重要作用，固有遗传变异是适应性遗传变异的重要来源。通过基因组范围的SNP分子标记，阐明了洄游型刀鲚种群间存在显著的遗传分化，为保护不断下降的长江洄游型刀鲚资源提供合理评价和理论支撑。相关研究结果填补了刀鲚种群基因组学研究的空白，揭示了刀鲚适应淡水平行演化的遗传机制，并且为其他非模式生物的群体遗传学研究的开展提供了参考。

With the global climate change, locally adapted populations need to keep up with the pace of a rapidly varying environment. Therefore, it is necessary to investigate genetic architecture of rapid adaptation with complex traits controlled by a large number of environmental influences, and to illustrate the source of rapid adaptive genetic variants and evolutionary trajectory in wild populations. Moreover, this will provide a significant sight into understanding the initiative adaptive genetic mechanism of populations, assessment of evolutionary potential and conservation of biological resources. Japanese grenadier anchovy, Coilia nasus Temminck et Schlegel, 1846 (Cluperiformes: Engraulidae) is an economic and anadromous fish. Coilia nasus exists in two distinct life history forms. One is an anadromous form, widely distributed in coastal and estuarial regions of the Northwest Pacific, which migrates upstream into fresh water in the spring for breeding. The other is a landlocked freshwater-resident form including C. nasus taihuensis and C. brachygnathus spending their whole life in the affiliated lakes in middle and lower reaches of the Yangtze River in China, such as Taihu Lake, Chaohu Lake, Poyang Lake and Dongting Lake, and serving as the most dominant species in the lake ecosystem. Due to the impact of anthropogenic activities, such as overfishing, widespread artificial dams and environmental pollution, the quantity of C. nasus populations has declined dramatically, and the fish become mature at early age with small body size. C. nasus has been classified as Red List of Threatened Species published by International Union for Conservation of Nature. The anadromous form and landlocked freshwater-resident form of C. nasus may have become different in certain genomic regions since they live in different habitats. Therefore, the rational assessment and theoretical foundation for preventing the continuous descending population size are extremely urgent through illustrating the genetic relationship among three ecotypes of C. nasus populations. However, the researchers have not come to an agreement on the taxonomy of C. nasus complex at this moment, especially, the validity of C. brachygnathus. Moreover, the genetic architecture underlying rapid parallel adaptation of C. nasus to fresh water is still poorly understood, limiting the management and conservation of C. nasus. With the advent of next-generation sequencing (NGS) technology, hundreds of thousands of molecular markers across the whole genome of interest become realistic. In this study, we performed Reduced-Representation Sequencing (RRS), through single-nucleotide polymorphism (SNP) genotyping and mitochondrial partial genome assembly of wild C. nasus species complex, aiming to clarify accurate genetic relationship and demographic history among three ecotypes of C. nasus. The studies were designed to uncover genomic architecture of rapid adaptation to fresh water in resident C. nasus populations and to illustrate population genetic structure and local adaptation in anadromous C. nasus populations. In addition, the corresponding results can be a reference for stock assessment, effective management and conservation of endangered anadromous C. nasus. Main results are presented as follows:

    1. We studied genetic relationship and demographic history among three ecotypes of Japanese grenadier anchovy (C. nasus species complex) using 28,691 genome-wide SNPs and 196 mitochondrial partial genome sequences consisting of 2,916 bp, and tested the history of freshwater invasion of C. nasus. Based on the results of Analysis of Molecular Variance, there was more percentage variation among populations within groups than that of among groups, suggesting the shallow but significant genetic differtiation between anadromous C. nasus and freshwater resident C. nasus taihuensis populations (FCT = 0.022). On the contrary, there was more percentage variation among groups than that of among populations within groups, suggesting the high and significant genetic differtiation between anadromous C. nasus, freshwater resident C. nasus taihuensis populations and freshwater resident C. brachygnathus populations (FCT = 0.648). Among 196 C. nasus species complex individuals, we detected 80 haplotypes. The haplotype network analysis showed the haplotypes divided into two main clades, named clades I and II. Clade I comprised 51 haplotypes mainly in anadromous C. nasus and freshwater resident C. nasus taihuensis populations, but eight freshwater resident C. brachygnathus individuals were also found in this group. Clade II comprised 29 haplotypes mainly from freshwater resident C. brachygnathus populations, but only one individual from Taihu Lake population were found in this group. The two dominant haplotypes differed by five nucleotide substitutions (c. 0.17%). Combined with morphological evidence, the validity of C. brachygnathus was confirmed. However, the freshwater resident C. nasus taihuensis and the anadromous C. nasus were not different genetically based on Bayes factor species delimitation (BFD*). Neighbour-joining tree, DAPC and ADMIXTURE analyses all corroborated the results of BFD*. The independent freshwater invasion event occurred around 10,000 years ago according to DIYABC analysis. The divergence time between anadromous C. nasus and freshwater resident C. brachygnathus is consistent with the paleogeographic event – last glacial maximum. Our results strongly suggest that C. brachygnathus is genetically isolated from C. nasus and C. nasus taihuensis and thus represents a valid species.

    2. Using population genomic approaches, we investigated the genomic architecture that underlies rapid parallel adaptation of C. nasus to fresh water by comparing four freshwater-resident populations with their ancestral anadromous population. A total of 1,147 outlier SNPs detected by Fisher’s exact test (FET) and Pcadapt were shared among the four population pairs. Most (96%) of the 1,147 candidate outlier SNPs were preferentially distributed on chromosomes LG6 (263) and LG22 (838), clustering into similar genomic islands among population pairs. The freshwater allele (FWA) of most candidate SNPs was nearly fixed in the freshwater populations. In contrast, the frequency of FWA in the anadromous population was generally low with an average of 0.12 (SD = 0.07). Two putative large chromosome inversions on LG6 and LG22 (31.54 Mb and 21.97 Mb) were revealed, which were enriched for outlier loci and exhibited parallel association with freshwater adaptation. The frequency of the putative chromosome rearrangements on LG6 and LG22 displayed large shifts between the anadromous and freshwater-resident populations. The F_ST between the anadromous population and the four freshwater-resident populations calculated with all SNPs located in the chromosome inversion regions on LG6 and LG22 showed elevated differentiation with an average value of 0.37, which was an order of magnitude larger than the F_ST values (0.03) based on neutral SNPs. Drastic frequency shifts and elevated genetic differentiation were observed for the two chromosome inversions among populations, suggesting that both inversions would undergo divergent selection between anadromous and resident ecotypes. Combined, the two inversion regions in LGs 6 and 22 covered more than 50 Mb (~ 6% of the genome) and contained more than 1,800 genes. Gene Ontology (GO) enrichment analyses of genes within the two chromosome inversions in terms of biological process identified a total of 72 and 73 significantly enriched GO terms on LG6 and LG22, respectively (P < 0.01). Enrichment analysis of genes within chromosome inversions showed significant enrichment of genes involved in metabolic process, immunoregulation, growth, maturation, osmoregulation, and so forth, which probably underlay differences in morphology, physiology and behavior between the anadromous and freshwater-resident forms. The availability of beneficial standing genetic variation, large optimum shift between marine and freshwater habitats, and high efficiency of selection with large population size could lead to the observed rapid parallel adaptive genomic change. We propose that chromosomal inversions might have played an important role during the evolution of rapid parallel ecological divergence in the face of environmental heterogeneity in C. nasus and standing genetic variation is important source for adaptation.

    3. Using RAD-seq data from 95 individuals in four anadromous populations of C. nasus, we analysed population structure based on genome-wide SNPs. A total of 167,905 SNPs were retained after quality filtering. Results of population structure analysis suggested that the sampled populations should be divided into three genetic clusters, with Yalu River Estuary and Liaohe River populations forming into one group, Yangtze River Estuary and Fuchun River populations grouping into distinct clusters respectively. The genome-wide fixation indexes (F_ST) among the anadromous populations were generally shallow but significant (mean = 0.06). Self-Assignment analysis implemented in GENECLASS2 indicated that all individuals of Yangtze River Estuary population were assigned to the true home, while the assignment accuracy of Fuchun River population was 91.7%, Yalu River Estuary and Liaohe River populations had the lower rates of 79.2% and 73.1%, respectively. Gene annotations of outlier SNPs indicated some genes that were possibly involved in migration. These genes involved in growth, metabolic process, immunoregulation, transmembrane transport, maturation, signal transduction, osmoregulation, and so forth. We highlight in situ conservation for C. nasus across its entire distribution range, and protect the migrate route to maintain the genetic diversity and evolutionary potential of C. nasus.

    Overall, with Bayes factor species delimitation (BFD*) analysis and morphological evidence, these comprehensive data acquired through SNPs and mitochondrial partial genome confirmed the validity of C. brachygnathus. We investigated the genomic architecture that underlies rapid parallel adaptation of C. nasus to fresh water through population genomics. We propose chromosomal inversions might have played an important role during the evolution of rapid parallel ecological divergence in the face of environmental heterogeneity in C. nasus and standing genetic variation is important source for adaptation. There was significant genetic difertiation among four anadromous populations of C. nasus based on genome-wide SNPs, providing the rational assessment and theoretical foundation for protecting the Yangtze River anadromous C. nasus. Our results have filled the gap of genome research and revealed the genetic mechanism of adaptation to freshwater for C. nasus, which can be a good reference for studying other nonmodel organisms.

第1章 绪论... 1

1.1 种群适应环境变化的遗传学机制... 1

1.1.1 经典群体遗传学认为选择性清除是快速适应性进化的重要机制... 1

1.1.2 数量遗传学范畴内的多基因性状快速适应性进化机制... 2

1.1.3 基因组结构变异普遍存在并在生态与演化过程中发挥重要作用... 2

1.1.4 复杂性状快速适应性进化进程中遗传变异的来源是广受关注的热点科学问题... 3

1.2 基因组结构变异的形式及在生态演化中的作用... 4

1.2.1 染色体倒置的基本特征及其在生态演化中的作用... 5

1.2.2 染色体倒置的生态进化意义... 8

1.3 刀鲚研究进展... 13

1.3.1 刀鲚的分类地位及其生物学特征... 13

1.3.2 刀鲚的形态学研究进展... 16

1.3.3 刀鲚的分子遗传学研究进展... 17

1.4 本研究的目的与意义、技术路线... 19

1.4.1 研究的目的与意义... 19

1.4.2 研究技术路线... 20

第2章 刀鲚三种生态型之间的遗传关系... 23

2.1 引言... 23

2.2 材料与方法... 24

2.2.1 样品采集与RAD文库构建、测序... 24

2.2.2 RAD数据过滤与SNP提取、过滤... 27

2.2.3 种群遗传多样性和遗传结构... 29

2.2.4 线粒体基因组组装... 29

2.2.5 单倍型网络图分析... 29

2.2.6 物种界定... 30

2.2.7 种群动态演化历史... 31

2.3 实验结果... 32

2.3.1 种群遗传多样性和遗传结构... 32

2.3.2 线粒体基因组组装和单倍型网络图... 35

2.3.3 物种界定... 39

2.3.4 种群动态演化历史... 40

2.4 讨论... 41

2.4.1 短颌鲚物种有效性... 41

2.4.2 短颌鲚的动态演化历史... 42

2.5 本章小结... 43

第3章 刀鲚淡水定居种群快速适应性进化的遗传机制... 45

3.1 引言... 45

3.2 材料与方法... 46

3.2.1 样品采集与RAD文库构建、测序... 46

3.2.2 RAD数据过滤与SNP提取过滤... 48

3.2.3 种群遗传多样性和遗传结构... 48

3.2.4 种群动态演化历史... 48

3.2.5 筛选离散位点... 51

3.2.6 推断染色体倒置区域... 52

3.2.7 基因组注释... 52

3.2.8 基因功能富集... 53

3.3 研究结果... 54

3.3.1 种群遗传多样性和遗传结构... 54

3.3.2 种群动态演化历史... 58

3.3.3 离散位点筛选... 59

3.3.4 淡水等位基因频率变化... 60

3.3.5 与淡水适应平行演化相关的染色体倒置区域... 61

3.3.6 离散位点基因注释和倒置区域基因功能富集... 65

3.4 讨论... 78

3.4.1 染色体倒置在适应性分化中的重要作用... 78

3.4.2 影响基因组适应性进化速率的因素... 79

3.5 本章小结... 80

第4章 洄游型刀鲚的种群遗传结构... 81

4.1 引言... 81

4.2 材料与方法... 82

4.2.1 样品采集与RAD文库构建、测序... 82

4.2.2 RAD数据过滤与SNP提取、过滤... 82

4.2.3 种群遗传多样性和遗传结构... 83

4.2.4 遗传归群分析... 84

4.2.5 推断倒置区域... 84

4.2.6 筛选离散位点和基因注释... 84

4.3 实验结果... 84

4.3.1 种群遗传多样性和遗传结构... 84

4.3.2 归群分析... 87

4.3.3 与淡水适应相关的倒置区域... 89

4.3.4 离散位点筛选和基因注释... 92

4.4 讨论... 97

4.4.1 洄游型刀鲚种群遗传多样性... 97

4.4.2 洄游型刀鲚种群遗传结构... 98

4.4.3 洄游型刀鲚本地适应性... 99

4.4.4 洄游型刀鲚保护管理建议... 100

4.5 本章小结... 101

第5章 总结与展望... 103

5.1 研究总结... 103

5.2 主要创新点... 104

5.3 研究不足之处... 104

5.4 展望... 105

参考文献... 107

致 谢... 127

作者简历及攻读学位期间发表的学术论文与研究成果... 129