Genetic maps of the A. tauschii D-genome have previously been constructed using segregating populations in diploid backgrounds (Gill et al. 1991; Boyko et al. 1999; Ter Steege et al. 2005; Luo et al. 2009). In contrast, our map of the D-genome of A. tauschii was constructed in a hexaploid background. Although the DH population used in this study was small—only 39 lines, 96% of 440 polymorphic DNA markers were mapped onto the final A. tauschii D-genome map. This high genetic map construction efficiency may be a consequence of the unique genetic structure of the hexaploid DH population used in this study, in which only D-genome chromosomes between A. tauschii accessions AS66 and AS87 were involved in genetic recombination under a background of non-recombinant A- and B-genomes from the T. turgidum line PI377655 (Luo et al. 2012). Interference due to A- and B-genome polymorphism was thus avoided (Poole et al. 2007; Barker and Edwards 2009; Allen et al. 2011). When 50 shared markers were compared between this map and a consensus map constructed by Somers et al. (2004), 37 exhibited consistent orders. The orders of the remaining 13 differed, however, perhaps as a result of small structural rearrangements (such as translocations, deletions, and inversions) and/or because of the small number of DH lines used.
To evaluate the usefulness of the hexaploid wheat DH population for gene identification in A. tauschii, we analyzed the morphological trait of glaucousness. Glaucousness in A. tauschii is controlled by a dominant gene, W2, located on chromosome arm 2DS. This phenotype is inhibited by the epistatic influence of the dominant inhibitor gene W2
found on the distal region of 2DS (Watanabe et al. 2005; Liu et al. 2007). In our study, the non-glaucous trait was also mapped to 2DS. This result suggests that the diploidization-hexaploid DH population has value as a tool for mapping qualitative trait genes.
Segregation distortion is a common phenomenon in plants and can be influenced by various factors affecting the fertility of either gametes or zygotes (Lyttle 1991). In a previous study on 54 F2 diploid plants derived from two A. tauschii accessions (Faris et al. 1998), 57 (29%) out of 194 RFLP markers were significantly distorted (P < 0.05) from expected segregation ratios, with segregation distortion regions (SDRs) detected on chromosomes 1D, 3D, 4D, 5D, and 7D. In the present study, 48 (11.4%) out of 422 markers showed distorted segregation (P < 0.05), and seven SDRs were detected on chromosomes 1D, 2D, 3D, 6D, and 7D. The longest SDR was QSd.scau-7D3, with a length of 16.46 cM including the centromere, and favoring the A. tauschii parent AS66 (Additional file 1: Table S1). There may be an important locus associated with this SDR, as a similar SDR has also been detected on a homoeologous chromosome, 7E, in another species (Cai et al. 2011). Out of seven SDRs, six skewed in favor of A. tauschii AS66, the maternal parent in the cross with A. tauschii AS87. This is consistent with the observations of Faris et al. (1998) that loci affecting gametophyte competition in male gametes via nucleo-cytoplasmic interactions may play a role in SDR production. Allelic variation associated with the production of wide-hybrid plants may have also contributed to the segregation distortion observed in our study, as these hexaploid DH lines were derived from the wide hybridization of T. turgidum PI377655 with diploid F1 hybrids of A. tauschii AS66 × AS87. For example, allelic variations in loci that control crossability between different species can affect seed-setting of interspecific crosses (Tixier et al. 1998). Allelic variations related to hybrid seed germination and plant vigor may also affect the production of DH lines (Ter Steege et al. 2005). During the production of the DH lines used in our study, only 16.5% (71/430) of F1 hybrid seeds germinated; from these seeds, 39 vigorous haploid plants possessing ABD genomes were obtained (Luo et al. 2012).
We demonstrated in this study that diploidization-hexaploid DH population can be used to generate genetic maps. However, the limitation of a small DH population with only 39 lines should be pointed out. The present study identified 12 linkage groups which were significantly larger than the number of the 7 haploid chromosomes of D genome. The small population could result in more linkage groups and could be one of factors for segregation distortion and inconsistent marker order. To generate a better genetic map, a larger size of mapping population is needed.