QTL Mapping using Intervarietal Substitution Lines in Rapeseed (Brassica napus L.) (PDF)

Intervarietal substitution lines (ISLs) having one or a few defined segments of a donor genome in the common genetic background of a recurrent parent can be used to search the genome for donor alleles affecting traits. A complementary set of substitution lines represents ideally the whole donor genome divided into a limited number of distinct segments, each carried by a different line. ISLs were suggested as an alternative to a segregating population for QTL mapping. An ISL population can be used to overcome the limitations of a segregating population in the accuracy of QTL localization. With overlapping donor segments in different substitution lines and their respective phenotypic values, QTL positions can be narrowed down to a few centimorgan (cM), allowing a high precision of QTL localization.
The main objectives of this study were:
- To develop set of intervarietal substitution lines (ISL) from the cross of ‘Express’ x resynthesized line ‘RS239’;
- To map QTL for some agronomically important traits in two sets of ISL populations developed from the crosses of ‘Mansholt’ x ‘Samourai’ and ‘Express’ x resynthesized line ‘RS239’;
- To compare QTL results mapped in the ISL population with QTL mapped in an earlier generation (F1DH population) developed from the cross ‘Mansholt’ x ‘Samourai’;
For this, a set of ISLs was developed from a cross between the spring type resynthesized rapeseed line ‘RS239’ and the winter rapeseed variety ‘Express’ (‘ExRS239’). A second set, developed from a cross between doubled haploid lines of the two winter rapeseed varieties ‘Mansholt’ and ‘Samourai (‘MxS’), was available from earlier work. The ISLs were developed through five backcross generations. Using marker assisted selection with AFLP markers, lines carrying a complementary set of donor segments were selected.
Three hundred fifty ISLs developed from the cross ‘ExRS239’ and 270 ISLs from the cross ‘MxS’, were grown in the field at five locations in the year 2009, respectively 2006 and 2007, in Germany and agronomically important traits were evaluated. Molecular marker analysis using AFLP markers was conducted to characterize the segments of the donor parents carried by the ISLs. QTL were mapped by comparing the segments of the donor genome and their effects in the ISLs.
The donor segments in the ISLs generated from the cross of ‘ExRS239’ cover, depending on the method of estimation, a minimum of 950.9 cM or a maximum of 1587 cM, corresponding to 52.8% and 88.1% of the genetic map of this cross. This coverage is larger than that in the set of ISL generated from the cross ‘MxS’, where the donor segments cover a minimum of 515.5 cM or a maximum of 1095.3 cM, corresponding to 37.9% and 80.5% of the respective genetic map.
Forty four QTL of agronomically important traits in rapeseed were mapped using the ISLs of the cross ‘ExRS239’ and 24 QTL were mapped in the ISLs of the cross ‘MxS’. The higher number of QTL mapped in the ISLs from the cross ‘ExRS239’ is due to the larger genetic distance between the parents resulting in a larger number of functionally different alleles segregating in the progeny of this cross. Another reason is the better coverage of donor genome in this set of ISL.
Some QTL valuable for breeding purposes and to improve rapeseed quality were identified in the ISL from the cross of ‘ExRS239’. Six QTL were detected to have positive additive effects for seed protein content. Two of these QTL, the QTL on linkage groups N07 and N12, controlled protein content independently from QTL for oil content. The two QTL showed additive effects of 2.5% for protein content. Four QTL with donor alleles decreasing sinapate esters content were mapped on linkage groups N06, N11, N13, and N15. The additive effects of the donor alleles sum up to 2.19 mg/kg in sinapate esters reduction. Three QTL with donor alleles responsible for decreasing plant height at maturation were mapped on linkage groups N13, N18a, and N19 with a sum of additive effects of -50.4 cm.
The comparison of QTL of an early mapping generation, an F1DH population, and the advanced generation, the ISLs, from the cross of ‘MxS’ confirmed some QTL mapped in the F1DH population in the ISLs: two QTL on linkage groups N09 and N12 for seed glucosinolate content; two genes controlling seed erucic acid content on linkage groups N8 and N13; and two QTL for phytosterol, sinapate esters and oil content on linkage groups N08 and N13 corresponding to the two erucic acid genes. Some of the QTL identified in the F1DH population could not be detected in the ISL population because they had been mapped in regions that were not covered by the donor segments.
Several new QTL were identified in the ISLs that remained undetected earlier in the F1DH population. On the other hand, two QTL for protein content on linkage groups N01 and N03, and a QTL for plant height on linkage group N16 mapped in the F1DH population have to be considered as false positives. These QTL could not be detected in the ISL population, although ISLs carrying donor segments covering the regions where the QTL had been mapped were available.
It was observed that the additive effects of the QTL detected in the ISL population were higher than that in the F1DH population. Using ISLs, the QTL interval could also be narrowed down compared to the interval mapping in the F1DH population. The QTL were mapped in maximal intervals ranging from 1.8 to 30.2 cM in the ISL of the cross ‘ExRS239’ and from 1.2 to 24.8 cM in the ISL of the cross ‘MxS’. About 70% of the QTL were mapped in maximal intervals of less than 15 cM in both of the ISL populations. This was considerably smaller than the confidence intervals usually estimated when using segregating populations.
The QTL for phytosterol and sinapate esters were co-localized with the two erucic acid genes. A pleiotropic effect between erucic acid and phytosterol content was identified. On the other hand, close linkage rather than pleiotropism was suggested as an explanation for the correlation between erucic acid and sinapate esters content.
Further fine mapping of QTL can be done by backcrossing ISLs to the recurrent parent to reduce size and number of introgressed donor segments. After fine mapping it will also be possible to better distinguish between pleiotropy and close linkage for QTL of correlated traits.