多倍化（Polyploidization）或者全基因组复制（Whole-genome duplication, WGD）是植物进化的一个重要过程，能够增加生物体的复杂性和进化的可塑性。同源多倍体由于多倍化过程不受杂交的干扰，是探究多倍化进化效应的理想材料。然而，由于同源多倍体在形态上与二倍体差异较小，不易发现；且亚基因组之间非常相似，较难开展相关的遗传分析。因此，相对于异源多倍体物种，同源多倍体物种的进化研究还较少。青钱柳（Cyclocarya paliurus）隶属核桃科（Juglandaceae）青钱柳属，自然群体中同时存在二倍体和同源四倍体个体，是探究同源多倍体进化过程和机制的理想材料。本研究以青钱柳为研究材料，探究同源多倍体物种多倍化后基因组发生的结构和功能变化以及同源多倍体青钱柳的物种形成和适应性进化过程。

首先，本研究对二倍体和同源四倍体青钱柳，进行了基因组组装，获得了一个拆分等位基因的染色体水平的同源四倍体基因组和一个嵌合的染色体水平的二倍体基因组。四倍体的基因组大小为2.36 Gb，包含64条染色体，二倍体基因组大小为601 Mb，包含16条染色体；四倍体青钱柳约在120万年前经历了第三次WGD。同源四倍体青钱柳的四套同源染色体在长度、基因数量和SNP数量等方面没有显著差异；以二倍体基因组为参考，发现同源四倍体基因组在加倍后发生结构变异的区域占整个基因组的8.8%，而与二倍体共线性的区域占全基因组的81.3%；两个基因组共线性区域的遗传距离（D_XY）和直系同源基因的同义突变（K_s）都很小；进一步发现同源四倍体基因组上有69.18%的基因具有4个等位基因，30.82%的基因具有1-3个等位基因（即等位基因丢失）。这些具有等位基因丢失的基因经常位于结构变异区域的内部或周围、或者与基因组重复区域重叠，而且这些基因与其他基因的互作较少。同源四倍体青钱柳有102个基因的拷贝数量大于4，且表达量显著高于二倍体，GO富集分析表明这些基因参与青钱柳的逆境应答过程。

其次，对来自21个种群的118个个体进行全基因组重测序（四倍体个体的平均测序深度为60×，二倍体个体的平均测序深度为90×）。基于118个个体的重测序数据，鉴定出106个四倍体和12个二倍体，并且发现同源四倍体遗传多样性显著高于二倍体；同时种群聚类、主成分分析和最大似然树的结果均显示，同源四倍体个体组成一个群体，二倍体个体组成一个群体，表明同源四倍体青钱柳可能是单次起源；进一步推断二倍体和同源四倍体的分化时间约为57万年，两者之间存在微弱的基因流；同源四倍体基因组上约有30%的区域表现出二体遗传模式。四倍体中一些受正选择的与生物刺激应答相关，与减数分裂相关的基因也受到正选择，说明即使是保守的减数分裂过程，多倍化后也发生了进化。叶绿体系统发育的结果显示，单次起源的四倍体青钱柳在扩散的过程中，捕获了当地二倍体个体的叶绿体，导致四倍体的叶绿体单倍型与地理距离近的二倍体的叶绿体单倍型聚类在一起。

青钱柳属在核桃族（青钱柳属、枫杨属和核桃属）的系统发育地位一直存有争议，部分原因在于人们错误地将青钱柳四倍体当作二倍体来处理。本研究在探究青钱柳多倍体进化的基础上，利用7个青钱柳的近缘物种的全基因组数据，进行核桃族类群的系统发育分析。结果发现，无论是种群聚类、主成分分析，还是物种树推断，都显示青钱柳属与枫杨属的关系最近，但叶绿体基因组数据显示枫杨属与核桃属关系最近，表现出系统发育关系的核质不一致模式。进一步发现，核质不相容和不完全谱系分选不足以解释这种核质不一致模式，三个属之间存在广泛的基因流可能是主要因素。

本研究利用全基因组数据揭示了多倍化后四倍体青钱柳基因组发生的结构和功能上的变化，以及对群体进化历史的影响，同时厘清了青钱柳属与近缘类群的系统发育关系，为多倍化如何促进物种多样性和基因组进化提供了一定的依据和新的思路。

Polyploidization or whole-genome duplication (WGD) is a crucial process in plant evolution, as it can increase organismal complexity and evolutionary potential. Autopolyploidy is a valuable model for studying the evolutionary effects of polyploidization without hybridization. However, autopolyploids are morphologically similar to diploids and difficult to discover, and their subgenomes are highly similar, making related genetic analyses challenging. Thus, compared to allopolyploid species, fewer studies have explored the evolution of autopolyploid species. Cyclocarya paliurus (Juglandaceae) has natural populations containing both diploid and autotetraploid individuals, making it an ideal species to study the evolution and mechanisms of polyploidization. In this study, we used C. paliurus to investigate the structural and functional changes in the genome following polyploidization, as well as the speciation and adaptive evolution of autotetraploid C. paliurus.

First, this study conducted genome assembly of diploid and autotetraploid C. paliurus, resulting in the generation of an allele-aware, chromosome-level autotetraploid genome and a chimeric, chromosome-level diploid genome. The two genomes, consisting of 64 and 16 chromosomes respectively, had a size of 2.36 Gb and 601 Mb, and the autotetraploid C. paliurus underwent its third WGD about 1.2 million years ago. Notably, there were no significant differences observed between the four homologous chromosomes of autotetraploid C. paliurus regarding their length, gene number, and SNP number. Using the diploid genome as a reference, the study found that 8.8% of the autotetraploid genome underwent structural variation following polyploidization, while 81.3% of the genome exhibited collinearity with the diploid genome. The genetic distance (D_XY) of collinear regions and the synonymous mutation (K_s) of orthologous genes were also found to be small of two genomes. Furthermore, 69.18% of the genes in the autotetraploid genome had four alleles, whereas 30.82% of the genes had 1-3 alleles (i.e., allele loss). Genes with allelic loss were found to occur more often in proximity to or within structural variations and exhibited a marked overlap with transposable elements, and such genes showed a reduced tendency to interact with other genes. We also found 102 genes with more than four copies in the autotetraploid genome, and their expression levels were significantly higher than their diploid counterparts. The GO enrichment analysis showed that these genes were involved in stress response processes in autotetraploid C. paliurus.

Next, we performed whole-genome resequencing on 118 individuals from 21 populations, with an average sequencing depth of 60× for tetraploid individuals and 90× for diploid individuals. Based on the resequencing data, 106 tetraploid and 12 diploid individuals were identified. Autotetraploids showed significantly higher genetic diversity than diploids. Population clustering, principal component analysis, and maximum likelihood tree results all showed that the autotetraploids formed a distinct cluster and the diploids formed another, indicating that the autotetraploid C. paliurus has a single origin. Moreover, we observed a recent divergence (~0.57 Mya) of autotetraploids from diploids, with little interploidy gene flow, and 30% of the autotetraploid genome exhibited disomic inheritance. Eight genes under positive selection were detected for related to biotic stress response and eight genes related to meiosis, indicating that even the conserved meiosis process had evolved after polyploidization. Chloroplast phylogenetics revealed that during the expansion of the single-origin tetraploid C. paliurus, it captured chloroplasts from local diploids, resulting in the clustering of chloroplast haplotypes of tetraploids with those of diploids geographically close to them.

The phylogenetic position of the Cyclocarya genus within Juglandinae (which comprises Cyclocarya, Pterocarya, and Juglans) has long been a matter of debate, partly due to confusion between tetraploid and diploid C. paliurusindividuals. To shed light on this issue, we conducted a phylogenetic analysis of the Juglandinae using whole-genome data from seven closely related species of C. paliurus, based on the findings of C. paliurus evolution. Our analysis showed that Cyclocarya is most closely related to Pterocarya, as supported by population clustering, principal component analysis, and species trees based on various nuclear data types. However, species tree inference based on chloroplast genome data suggested that Pterocarya is most closely related to Juglans, indicating cytonuclear discordance within the three genera. Moreover, we found that neither cytonuclear incompatibility nor incomplete lineage sorting could account for this cytonuclear discordance, and suggested that extensive gene flow among the three genera might be the main factor contributing to the observed pattern.

This study utilized whole-genome data to reveal the structural and functional changes that occurred in the autotetraploid C. paliurus genome after polyploidization, as well as the population evolutionary history of autotetraploid C. paliurus. Additionally, we clarified the phylogenetic relationships between Cyclocarya and its closely related taxa. These findings provide new insights and a basis for understanding how polyploidization promotes species diversity and genome evolution.