| Tree peony(Paeonia sect. Moutan), is Chinese traditional famous flower which has great ornamental, medicinal and economic value. Although conventional breeding protocols have greatly contributed to an increase in tree peony germplasm, but these resources can not be sufficient to meet the peony industrial needs. Especially, the breeding and improvement progress has been hindered by high heterozygosity, long breeding period, and the natural quantitative for horticultural traits of interest. Marker-assisted selection (MAS), which is used in fast and accurate identification preponderant genotype through traits linked molecular markers to shorten the breeding period and improve breeding efficiency, is one of the development direction of peony breeding. Therefore, on the basis of large-scale development of microsatellite markers using transcriptome sequences, we first carried out the association analysis, based on a large population of 462 unrelated cultivated Paeonia rockii individuals, then to validate the significant associations in a linkage mapping population, and finally to reveal excellent allelic loci. The main results in this study as follows:1. Development of SSR markers is the foundation of this research work. A total of 4373 EST-SSRs were identified from 3787 Unigenes with an average of one SSR per 9.24 kb. Di-nucleotide repeats (46.26%) were the most abundant type followed by tri-, hexa-, tetra-and penta-nucleotide repeats. Moreover,788 primer pairs were synthesised, out of which 373 primer pairs (47.3%) produced successful amplifications, and 149 primer pairs (39.9%) were polymorphic in both parents of one mapping population. The 30 polymorphic EST-SSR markers were randomly selected to assess polymorphism further in 36 tree peony accessions, in which the polymorphism information content values ranged from 0.36 to 0.85 with a mean of 0.69, thereby indicating a high level of informativeness. Transferability studies indicated that 95.3% EST-SSR markers were transferable to nine other Paeonia species, thereby expanding their utility. In addition, as a demonstration of utility, a genetic relationship study was performed in 56 accessions. Cluster analysis using the neighbour-joining tree method demonstrated that major clusters corresponded to known pedigree trees. Thus, these newly developed markers provide an important genetic resource for MAS breeding in tree peony.2. Selection and evaluation of population is the first step in association analysis. We screened out one population with 462 seedlings from the main cultivation area of P. rockii. Totally 35 traits were investigated, on which descriptive statistics analysis and principal component analysis (PCA) were conducted. The results showed that the variation of 29 numerical traits was extremely significant (P<0.01) and the variation coefficient was 9.52%-112.10%, which the average genetic variation of fruit traits was the largest, followed by leaf traits and floral traits. The population displayed continuous distribution of phenotypes and normal segregation. PCA indicated that phenotypic variation was uniformly distributed within this population without conspicuous structure Six qualitative traits had abundant phenotypic variation, representing phenotypic variations of P. rockii. The above results show that this population is well suited for the association analysis.3. An understanding of genetics diversity and population structure is important for developing association analysis for tree peony breeding. Using 40 EST-SSR markers, a total of 183 alleles were detected, and the average number of alleles per locus was 4.5 with a range of 2-10 in the 462 P. rockii samples, indicating a moderate level of polymorphism in the association population. Population structure was assessed using three different methods, and the results showed that the association population was divided into three subgroups (K=3).4. One hundred and thirty-eight SSRs with minor allele frequency (MAF)>1% were used for the estimation of linkage disequilibrium (LD) among the whole panel. The results showed the overall level of LD between SSRs was low and a larger number of SSRs were in linkage equilibrium (r2<0.1; P<0.001) across the genome regions. For floral traits, using mixture linear models,46 significant associations were identified, involving the 11 traits with 37 SSRs. These loci explained a small proportion of the phenotypic variance, ranging from 2.68% to 23.97%(mean,5.50%). For leaf traits, 25 significant associations were identified, involving the 6 traits with 21 SSRs, which proportion of phenotypic variance was 2.19%-7.15%(mean,5.38%). For fruit traits,64 significant associations were identified, involving the 7 traits with 28 SSRs, which proportion of phenotypic variance was 2.21%-29.62%(mean,7.56%).5. A total of 13 significant associations were detected in association population and linkage population, of which related significant associations with floral, leaf and fruit traits were 5,1 and 7, respectively. On the basis of the genotypic effects for each marker tested in both populations,9 associations with the same effect patterns were identified (including 4 with positive effect and 5 with negative effect). We obtained some markers associated with petal width, petal shape, petal color, seed dry weight, seed number and seed fresh weight. Subsequently, the excellent allele controlling the highest and lowest phenotypic values were detected.In a word, on the basis of large-scale development of microsatellite markers using transcriptome sequences, we carried out the association analysis in P. rockii for the first time, and the results of this study will serve as an important foundation for gene mapping and MAS in tree peony. |
PDF Full Text Request