中国农业科学 ›› 2018, Vol. 51 ›› Issue (5): 821-834.doi: 10.3864/j.issn.0578-1752.2018.05.002

• 作物遗传育种·种质资源·分子遗传学 • 上一篇    下一篇

玉米株型相关性状的全基因组关联分析

刘坤1,张雪海1,孙高阳1,闫鹏帅1,郭海平1,陈思远2,薛亚东1,郭战勇1,谢惠玲1,汤继华1,李卫华1

 
  

  1. 1河南农业大学农学院/省部共建小麦玉米作物学国家重点实验室,郑州450002;2郑州市第一中学,郑州 450002
  • 收稿日期:2017-09-26 出版日期:2018-03-01 发布日期:2018-03-01
  • 通讯作者: 李卫华,Tel:0371-63558122;E-mail:liwh416@163.com
  • 作者简介:刘坤,E-mail:lkun2017@163.com。张雪海,E-mail:xuehai85@126.com。刘坤和张雪海为同等贡献作者。
  • 基金资助:
    国家自然科学基金(U1604231)、河南省重大科技专项(161100110500)、河南省科技攻关项目(152102110060)

Genome-Wide Association Studies of Plant Type Traits in Maize

LIU Kun1, ZHANG XueHai1, SUN GaoYang1, YAN PengShuai1, GUO HaiPing1, CHEN SiYuan2, XUE YaDong1, GUO ZhanYong1, XIE HuiLing1, TANG JiHua1, LI WeiHua1   

  1. 1College of Agronomy, HenanAgriculturalUniversity/Key Laboratory of Wheat and Maize Crops Science, Zhengzhou 450002; 2No.1 Middle School of Zhengzhou, Zhengzhou 450002
  • Received:2017-09-26 Online:2018-03-01 Published:2018-03-01

摘要: 【目的】玉米株型性状与植株产量、光合效率、抗倒性等密切相关,是理想株型设计育种的基础,通过对玉米多个株型相关性状进行全基因组关联分析,构建玉米理想株型,为抗倒伏、耐密性的玉米新品种选育提供理论基础。【方法】以284份温带、亚热带和热带材料组成的关联群体为研究对象,在郑州与浚县2个环境下对玉米总叶片数(LN)、穗上叶片数(LNAN)、穗上叶片数与总叶片数比值(LNAN/LN)、株高(PH)、穗位高(EH)、穗位高与株高的比值(EH/PH)等6个玉米株型相关性状进行调查,借助覆盖玉米全基因组约56万个SNP标记,进行全基因组关联分析,以期为玉米新品种的选育提供理论依据。【结果】2个环境下,6个性状均表现为正态分布;大多数性状之间均存在高度正相关或负相关;方差分析表明这6个株型相关性状的基因型与环境以及基因型×环境的互作均达到显著水平。在选择最优关联分析模型时,发现Q模型假阳性较高,Q+K模型对假阳性的控制过于严格,而K模型的表现最好;在2个环境下,全基因组关联分析(K模型)共检测到56个显著SNP-性状关联(P≤3.99E-6),这些SNPs共涉及5个性状的17个位点,每个位点解释的表型变异从7.97%—10.56%不等。同时发现有4个位点能够在2个环境中同时被检测到,表明这4个位点受环境影响较小,在不同环境下可以稳定存在。通过对显著关联的SNP上下游各50 kb范围内候选基因进行搜索,共发现80个候选基因,其中42个具有功能注释。例如,与株高和穗位高显著相关的基因GRMZM2G161293编码一个具有乙酰葡糖转移酶活性的蛋白,该酶催化UDP-N-乙酰氨基葡萄糖生成糖过程中的N-乙酰葡糖氨基残基的转移,可能通过影响玉米籽粒可溶性糖含量进而影响产量,推测其为最可能的候选基因。【结论】K模型对假阳性的控制效果最好,基于K模型的GWAS结果,一共检测到17个株型性状相关的位点。

关键词: 玉米(Zea maysGWAS), 株型性状, 理想株型 L.), 全基因组关联分析(

Abstract: 【Objective】 Plant morphological traits are the basis of ideotype-based maize breeding which are closely related to photosynthetic efficiency, lodging resistance and grain yields. Genome-wide association studies (GWAS) of 558 629 SNPs with genome-wide coverage was used to elucidate the genetic basis of six plant morphological traits, including total number of leaves (LN), leaf number above ear (LNAN), the ratio of LNAN to LN (LNAN/LN), plantheight (PH), earheight (EH) and the ratio of EH to PH (EH/PH), which could provide theoretical basis in enhancing ideotype-based maize breeding and facilitating the genetic improvement of new maize varieties with high plant density and lodging resistance. 【Method】 In this study, a representative panel of 284 inbred lines planted in Zhengzhou and Xunxian, including temperate, subtropical and tropical materials, were used for association mapping.【Result】 All traits measured in the two locations exhibited an approximately normal distribution. Highly positive or negative correlations between paired traits were observed. Variance analysis of these traits indicated that significant variations were observed across environment, genotype and the genotype × environment interaction. When test with the optimal GWAS model, we found that Q model showed high type I errors (false positive), while Q+K model were too strict in reducing false positive. K model is the best in reducing false positive. Totally, 56 significant SNP-trait associations involving in 17 loci were identified for five traits (P≤3.99E-6), each locus can explain phenotypic variation ranging from 7.97% to 10.56%. Moreover, four loci were detected in both environments, indicating that these 4 loci were less affected by environment effects and could be stable in different environments. All potential candidate genes and their annotations within 100 kb (50 kb upstream and downstream of the lead SNP) of the loci were identified, in total, 80 candidate genes were found, including 42 genes that have functional annotation. For example,the gene GRMZM2G161293 encoding a protein that has acetylgluco-saminyl transferase is associated with plant height and ear height. It catalyzes the transfer of the amino group from N-acetyl glucosamine to glucose, which may improve yield by influencing the content of soluble sugars in maize kernels.【Conclusion】 The results indicated that K model having the best result in reducing the type I errors (false positive). Based on K model, a total of 17 loci associated with plant morphological traits were identified.

Key words: maize (Zea mays L.), genome-wide association study, plant type traits, ideotype