| Maize is one of the most important crops in China. Heterosis plays an important role in maize production. In US, theoretic research on maize heterosis stared at the early stage of 19 century and maize hybrids were used in 1930's. In China, maize hybrid breeding and extension stared in 1950's. Study on the genetic basis of herterosis has been paying a lot of attention since the beginning of the last century. But until now, it is not fully understood because of analysis method and technology limitations. Recently, the rapid development of molecular markers technology provides a great opportunity to study heterosis. The purposes of this investigation are to explore the genetic basis of maize heterosis with molecular markers based on the field tests in two locations, Wuhan and Xiangfan, using F2:3 population from an elite maize combination in China, and to compare the genetic basis of heterosis between rice and maize based on molecular markers and sequence information from rice. The major results are as follows. 1. A genetic linkage map containing 150 SSR and 24 RFLP markers was constructed, which spanned a total of 2531.6 cM with an average interval of 14.5 cM. Compared with other published maize linkage maps in marker alignments and intervals, the linkage map established in this study was consistent with them. All the markers were analyzed for the genetic segregation distortion. Totally 14 SDRs (segregation distortion regions) were found among 9 chromosomes, 4 of them were near the known gametophyte gene of segregation distortion. The causes of segregation distortion and effect to map QTL with distorted molecular markers were discussed. 2. Maize plant height was used as a model trait to study complex quantitative traits. Totally 10 QTL affecting plant heights were mapped on 6 different chromosomes with the composite interval mapping method. At the same time, 23 significant digenetic interactions were simultaneously detected in both locations (p<0.01). These results demonstrated that epistatic interactions might play an equal importance role as the single-locus effects in determining maize plant heights. QTL affecting developmental behavior of plant height also were mapped based on the combining analysis of composite interval mapping method with the conditional analysis method. The results revealed that the genes controlling plant height had obvious dynamic characters. Based on these results we brought forward the principle to perform molecular marker assistant selection of quantitative traits. 3. By CIM (LOD≥2.5), 102 QTL were detected for 10 yield traits and 34 were common in both locations. The number of QTL for different single yield trait ranged between 7 and 14. The contributions to phenotypic variations for the single QTL varied from 2.8% to 22.1%. Additive effects, partial dominance effects, dominance effects and overdominance effects were all detected at QTL for different traits. More proportion overdominance effects were always observed in QTL for traits that had higher heterosis. 4. At P≤0.005 level, 1097 and 1352 significant digenic interactions were detected in Wuhan and Xiangfan for 10 yield traits using all possible loci pairs by two ANOVAs between the 174 co-dominace markers, respectively. At P ≤0.01 level, 174 significant digenic interactions were detected simultaneously for 10 yield traits in both two locations. The number of significant digenic interactions varied between 4 and 56 according to the different traits. All 4 possible digenic interaction types were observed. Totally the proportion of AA type was 43.8%, AD (DA) type was 45.8% and DD type was 10.4%. Each of the interactions accounted for only a small proportion of the phenotypic variation with the average of 4.0% for single interaction. Most interactions (71.7%) occurred between two loci both showing non-significant effects to traits. It clearly demonstrated that epistasis play an important role in the maize genetics basis of heterosis. 5. At the molecular marker level of whole genome, the relationship of heterosis with heterozygosity was analyzed. For most characters, there was no significant correlation between heterosis and heterozygosity. For some characters that had higher heterosis, such as kernel weight per ear, ear weight, and ear length, they were significant correlations, but correlation coefficients were below 0.3. Taking ear length, row number and plant height for examples, effects among digenetic interactivegenotypes were compared. The results showed that one of the two-locus homologues of digenic interaction was often the best genotype, at the same time, another two-locus homologues of digenic interaction was also the worst genotype. On the other hand, double heterozygote of digenic interaction was neither the best nor the worst genotype. So, heterozygosity was also the important genetic basis of maize heterosis. 6. Based on molecular marker linkage maps of rice and maize, the syntenic relationships of quantitative trait loci (QTL) were compared for agronomic traits between maize and rice. Many QTL controlling same or similar traits in maize and rice had corresponding locations and conserved specificity, which suggested a common origin and conserved functionality of the genes underlying the QTL. Occasionally, one QTL for a given character in rice corresponded with two QTL for the same character in maize. There were QTL clusters in some regions of different chromosomes, which might cause by linkage or pleiotropy. These results may be useful in identifying gene or QTL among grasses, and understanding the function of the genomes and the evolution of structured the organization for grass genomes. 7. At E<1e-5 level, 16,939 homologous sequences were found in rice database by blast search using sequences of 224 RFLP and SSR markers. And 7853 homologous sequences were found at E<1e-10 level. 896 of 2729 homologous sequences were integrated in the rice high density molecular map by comparing with the rice physical map and molecular map. A real comparative map of maize and rice was constructed with this strategy, which provided a powerful technique platform for further investigation on heterosis comparative genomics.
|
PDF Full Text Request