【作者】 王松凤；

【导师】 江玲；

【作者基本信息】 南京农业大学 ， 遗传学， 2007， 硕士

【摘要】 近年来，随着国民经济的迅速发展和人民生活水平的不断提高，优质高产已成为水稻育种最重要的目标之一。由于稻谷粒形相关性状及千粒重是典型的数量性状，利用传统的育种方法对其进行改良，效率较低，而MAS选择是一种比较有效的育种途径。要实现MAS在育种上的应用，需要完成的工作主要包括以下几个步骤：1)获得不同群体在不同环境中重演率高的主效QTLs；2)构建全基因组的单染色体片段置换系或者近等基因系；3)利用分离群体对目标QTL进行单基因分解，并进行精细定位与MAS评估；4)建立分子育种相关的表型与数据库；5)多个优良性状QTLs的转移、渗透与聚合。由于目前绝大部分研究仍停留在利用单一群体，在单一环境下对相关QTLs进行检测，缺乏环境稳定性分析，难以发掘对环境钝感的QTLs，这严重影响了其在优质高产改良育种中的应用。为了发掘利用表达稳定的QTLs，本研究利用多个籼粳交永久性群体，在多个环境条件下对控制稻米粒长、粒宽、粒形及千粒重的QTLs进行了检测、比较和分析。主要结论如下：1．利用IR36／热研2号RIL群体在2个环境中检测到的控制粒长、粒宽、粒形和千粒重的QTLs的数目分别为3、3、4和6个。在2个环境中都能稳定表达的QTLs包括：qGL-3、qGW-5、qGW-6、qLWR-5和qTGW-3。其中，qGL-3的平均LOD值为3.94，平均贡献率为17.60％；qGW-5和qGW-6的平均LOD值分别为4.35和3.16，平均贡献率分别为16.10％和7.85％；qLWR-5的平均LOD值为2.25，平均贡献率为9.36％；qTGW-3的平均LOD值为2.14，平均贡献率为8.30％。2．利用Nipponbare／Kasalath∥Nipponbare BIL群体在3个环境中检测到的控制粒长、粒宽、粒形和千粒重的QTLs的数目分别为11、9、8和11个。在3个环境中均被重复检测到的QTLs有：qGL-2、qGW-2-1、qGW-5、qLWR-5和qTGW-2，其平均LOD值分别为5.43、3.06、6.50、3.68和4.26，平均贡献率分别为17.60％、10.30％、25.67％、14.27％和14.17％。进一步，结合以Nipponbare为背景，Kasalath为插入片段的CSSL群体在3个环境中的表型值，可以发现插入片段包含qGL-2、qGW-2-1和qTGW-2-1三个位点的株系NK50在3个环境中的表型值均与背景亲本存在显著差异，主要表现为：粒长明显变长，粒宽明显变窄，千粒重明显降低。而插入片段包含qGW-5的四个株系(NK22、NK32、NK38和NK50)与背景亲本相比，粒宽均明显变窄。3．利用Koshihikari/Kasalath∥Koshihikari BIL群体在3个环境中检测到的控制粒长、粒宽和粒形的QTLs的数目分别为9、12和13个。可以稳定表达的QTLs包括qGL-5、qGL-6、qGL-7、qGW-2-2、qGW-3-1、qGW-5和qLWR-5，其平均LOD值分别为3.89、4.42、3.93、4.76、5.40、13.49和18.57，平均贡献率分别为8.81％、9.29％、9.76％、8.65％、8.77％、26.6％和39.18％。对以Koshihikari为背景，Kasalath为插入片段的CSSL群体在3个环境中的表型进行分析，发现插入片段包含qGL-5的株系KK12和另外2个插入片段包含qGL-6的株系(KK17和KK18)，与背景亲本相比，粒长均明显增长；而插入片段包含qLWR-5的株系KK12和KK14，与背景亲本相比，粒形均明显增大。4．利用Sasanishiki/Habataki∥Sasanishiki///Sasanishiki BIL群体在3个环境中检测到控制粒长、粒宽和粒形的QTLs的数目分别为6、5和8个。其中控制粒长的QTLs，qGL-12在2006年南京和连云港2个环境中检测到，其平均LOD值为2.46，平均贡献率为8.80％；控制粒宽的QTLs，qGW-3-2和qGW-12在2年南京的环境中均检测到，平均LOD值为分别为5.10和3.91，平均贡献率分别为16.45％和12.40％；控制粒形的QTL，qLWR-3在3个环境中均检测到，其平均LOD值为4.93，平均贡献率为21.40％。综上，在本研究检测到的众多控制粒形相关性状及千粒重的QTLs中，能够稳定表达的QTLs主要包括：qGL-3、qGL-2、qGL-5、qGL-6、qGL-7、qGW-2-1、qGW-2-2、qGW-3-1、qGW-5、qGW-6、qLWR-3、qLWR-5、qTGW-2和qTGW-3。我们还发现，其中的一些QTLs在染色体上存在成簇分布现象，如在第5染色体上存在一个控制粒宽及粒形的QTLs簇(IR36／热研2号RIL群体，Nipponbare/Kasalath∥Nipponbare BIL群体，Koshihikari/Kasalath/Koshihikari BIL群体，Sasanishiki/Habataki∥Sasanishiki///Sasanishiki BIL群体)；在第2染色体上存在一个既控制粒长、粒宽又影响千粒重的QTLs簇(Nipponbare/Kasalath BIL)。这些研究结果为稻米的优质高产MAS育种奠定基础。更多还原

【Abstract】 Improvement of grain quality and yield has become one of the most important goals inrice breeding programs. Grain shape related traits and 1000-grain weight have been shownto be quantitatively inherited. Therefore, the use of marker-assisted selection (MAS) mayprovide a more efficient approach for improving these quality traits than classical breedingalone. MAS can be carried out by following five steps: 1. mapping major non-environment-specific loci that can be repeatedly detected in different populations; 2. constructingchromosome substitution segment lines carrying a single donor segment or near isogeniclines covering the whole genome; 3. dissecting target QTLs into single Mendelian factorsand precisely mapping, evaluating the feasibility of MAS by QTL pyramiding; 4.constructing phenotypic database for breeding by molecular technology; 5. transferring orintrogressing, and combinating multiple desirable QTLs. However, most of studies on ricegrain quality and yield were conducted in a single environment or using a single population,the stability of the resultant QTLs could not be evaluated. This characteristic is critical fordetermining the utility of a QTL for MAS. In the present study, the stability of QTLsaffecting grain length, grain width, grain length width ratio and 1000-seed weight wereanalyzed using different populations across different environments. The main conclusionswere as follows:1. Sixteen main effect QTLs were resolved using IR36/Nekken2 RIL population in twoenvironments. The numbers of QTLs for grain length, grain width, grain length width ratioand 1000-grain weight were 3, 3, 3, and 6, respectively. The stable QTLs were qGL-3qGW-5, qGW-6, qLWR-5, and qTGW-3, whose averge PVEs were 17.60％and 16.10％,7.85％, 9.36％, and 8.30％, with average LOD scores of 3.94, 4.35, 3.16, 2.25 and 2.14,respectively.2. A total of thirty nine main effect QTLs were detected for grain length, grain width, grainlength width ratio and 1000-grain weight in the Nipponbare/Kasalath//Nipponbare BILpopulation across three environments, qGL-2, qGW-2-1, qGW-5, qLWR-5 and qTGW-2 were repeatedly detected in three environments. Their averge PVEs were 17.60％16.10％,7.85％, 9.36％, and 8.30％, with average LOD scores of 5.43, 3.06, 6.50, 3.68 and 4.26,respectively. Further stability analysis was conducted using CSSLs with Nipponbare as thegenetic background and Kasalath as the donor in three environments. The result showedthat significant difference of phenotypic values existed between the recurrent parent,Nipponbare, and the CSSL NK50 harboring qGL-2, qGW-2-1 and qTGW-2-1 alleles.Compared to Nipponbare, NK50 had longer, narrower, and lighter grains. The four CSSLsharboring qGW-5 allele, including NK22, NK32, NK38 and NK50, all had markedlydecreased grain width across three environments, indicating the effects of the allele weresignificant and stable.3. Thirty four main effect QTLs were detected in the Koshihikari/Kasalath//KoshihikariBIL population across three environments, including 9 QTLs controlling grain length, 12controlling grain width and 13 controlling grain length width ratio, qGL-5, qGL-6, qGL-7,qGW-2-2, qGW-3-1, qGW-5 and qLWR-5 were expressed stably in three environments.Their averge PVEs were 8.81％, 9.29％, 9.76％, 8.65％, 8.77％, 26.6％and 39.18％, withaverage LOD scores of 3.89, 4.42, 3.93, 4.76, 5.40, 13.49 and 18.57, respectively.Significant difference of phenotypic values was observed between the recurrent parent,Koshihikari, and the CSSL KK12 harboring qGL-5 allele across three environments. Thesame case was for the two CSSLs harboring qGL-6 allele, namely KK17 and KK18, as wellas the two CSSLs KK12 and KK14 harboring qLWR-5 allele.4. A total of six QTLs controlling rice grain length, five QTLs controlling rice grain width,eight QTLs controlling rice grain length width ratio were detected using theSasanishiki/Habataki//Sasanishiki///Sasanishiki BIL population in all three environments.QTL qGL-12 for grain length was detected only in two of the three environments, Nanjingand Lianyungang, in 2006, with an average PVE of 8.80％and an average LOD score of2.46. QTLs qGW-3-2 and qGW-12 were identified in two environments of Nanjing in twosuccessive years. Only qLWR-3 was repeatedly detected in all three environments Thesethree QTL had average PVEs of 16.45％, 12.40％and 21.40％, and average LOD scores of5.10, 3.91 and 4.93, respectively.In conclusion, a total of 14 QTLs for grain shape relevant traits and 1000-grain weight, including qGL-2, qGL-3, qGL-5, qGL-6, qGL-7, qGW-2-1, qGW-2-2, qGW-3-1, qGW-5,qGW-6, qLWR-3, qLWR-5, qTGW-2, qTGW-3, were detected in different mappingpopulation across different environments. Moreover, a QTL cluster for grain width andgrain shape was located on chromosome 5 based on different populations, and another QTLcluster controlling grain length, grain wide and 1000-grain weight was detected onchromosome 2 in Nipponbare/Kasalath//Nipponbare BILs in different environments. Theseresults will provide useful information for elite rice breeding by MAS.更多还原