Studies On SSR Marker Assisted Selection For Genetic Background In Wheat

Backcrossing is widely regarded as one of important methods in plant breeding. And marker assisted selection may play a considerable role in backcrossing breeding. In this study, a transgenic line MG349 with GmDREB1 gene, which has been cloned in soybean and proved to improve drought tolerance, was chosen as the donor parent. Jimai 22, the dominant wheat cultivar in Shandong province at present, was chosen as the recurrent parent. 360 BC₁F₂ plants from the cross between the two parents were detected for GmDREB1 gene by using the gene itself function maker, and the recovery of the genetic background of Jimai22 in the offspring was analyzed by 46 SSR markers, which have polymorphism between the parents and cover 21 linkage groups of wheat, aiming to explore the suitable number of markers and their distribution in the genome, cost-efficient selection methods in SSR marker assisted backcrossing (MAB). The main results are as follows:1. The analysis of recovery ratio of genetic background (RRGB) of the 360 BC₁F₂ plants showed that means of RRGB was 0.762, close to the theoretical value of 0.75, in which 48.1% plants ranged 0.70⁰.80, 32.3% plants over 0.80 and 1.7% plants over 0.90 in RRGB respectively. So it indicated that, in terms of speeding-up wheat backcrossing, selecting the plants having both the targeted gene or genes and increased RRGB by increasing the population size is possible. In other words, the required times of backcrossing could be reduced by increasing the backcrossing population.2. The effects of number of SSR markers on selection efficiency of RRGB under the both conditions of considering the distribution of the markers in the genome and randomly selection of the markers were analyzed. Under the condition of considering the marker distribution, 7 markers (one in each homoeologous group), 14 markers (two in each homoeologous group) and 21 markers (three in each homoeologous group) were chosen. Correlation coefficients between the different number of markers and total of 46 markers in RRGB were 0.506,0.645 and 0.773 respectively, and all reached the significant level (P<0.01) . Among the 36 plants with the highest RRGB value obtained by 46 markers, 34%, 43% and 51% of the plants were the same as those obtained by 7, 14 and 21 markers respectively. When only considering the number of markers, without considering their distribution in the genome, correlation coefficients between the different number of markers and total of 46 markers in RRGB were 0.432, 0.556 and 0.652 respectively, and all reached the significant level (P<0.01) . Among the 36 plants with the highest RRGB value obtained by 46 markers, 32%, 41% and 47% of the plants were the same as those obtained by 7, 14 and 21 markers respectively. So in considering the cost-efficient way in the background selection, we recommended that, 2³ markers in every homoeologous group were selected at the beginning to scan the initial population, and then select plants with higher RRGB by using more markers if the result is not satisfied.3. The effects of SSR markers on selection efficiency of RRGB from genome A, B, D and the 7 homoeologous groups were also studied. The result showed that there were slight differences among the correlation coefficients between each genome and total RRGB obtained by 46 markers. The RRGB of B genome was significant lower than A, D genome and total RRGB, and there was significant positive correlation between B and A(P<0.01), as well as B and D. Thinking that the total RRGB is determined by the three genomes, we speculate that it's a good choice to increase the number of markers in the B genome for a better selection efficiency. Certainly, the difference between A, B and D genome may be induced by special genetic background of the two parents, so its validation is needed to be verified further. We also found that there's no correlation among the 7 homoeologous group across the marker data. But they all have positive correlation with the total RRGB and the sequence of correlation coefficients were: homoeologous group 6 (0.545) > homoeologous group 5 (0.478) > homoeologous group 4 (0.427) > homoeologous group 7 (0.421) > homoeologous group 3 (0.406) > homoeologous group 1(0.379)> homoeologous group 2(0.271).The order is consistent with the number of markers in the homoeologous group. But homoeologous group 2, homoeologous group 3 and homoeologous group 7 with the same number of markers have different correlation coefficients with total RRGB. The correlation coefficients with total RRGB of homoeologous group 3 and homoeologous group 7 were almost as twice as that of homoeologous group 2. And RRGB of homoeologous group 2 was the only one which had significant positive correlation with total RRGB. So it indicated that with the same number of markers, increasing the number of markers in the homoeologous group 2 may increase the selection efficiency in RRGB. As the same as what mentioned in comparing the differences in correlation with total RRGB between genomes, the differences in RRGB between homoeologous groups may be induced by specific background of the two parents.