The translocation of the short arm of rye chromosome 1 from the cultivar Petkus into the long arm of wheat chromosome 1B confers improved tolerance to several abiotic and biotic stresses. Although several genes for resistance to biotic stresses are no longer effective, the1RS.1BL translocation is still widely used because of its beneficial effects on grain yield and improved abiotic stress tolerance . We have previously shown that the presence of a short segment of wheat 1BS chromosome from cultivar Pavon in the distal region of the 1RS translocation was associated with reduced grain yield, biomass, and canopy water status relative to near-isogenic lines carrying the complete 1RS chromosome arm . Carbon isotope discrimination data showed that the lines with the complete 1RS chromosome arm achieve higher yields and better water status through increased access to water throughout the season, rather than through water conservation . A subsequent field study showed that the improved water status of the isogenic lines with the 1RS chromosome was associated with enhanced root density below 20 cm relative to the lines with the 1RSRW chromosome . Changes in root architecture in the field were correlated with drastic changes in root development in hydroponic growth systems, where the 1RSRW line showed a regulated arrest of the seminal root apical meristem ∼2 wk after germination. By the same time, the 1RSRW plants displayed altered gradients of reactive oxygen species in the root tips and emergence of lateral roots close to the RAM . In this study, we performed exome captures for 1RS, 1RSRW, ebb flow tray and its parental lines T-9 and 1B+40 . We show that, as a result of a distal inversion between 1RS and 1BS chromosome arms, T-9 and 1B+40 have duplicated 1BS and 1RS orthologous regions in opposite orientations and that a crossover between these chromosomes resulted in a duplicated 1RS region colinear to the inserted 1BS segment in 1RSRW.
Using these genetic stocks, we demonstrate that the dosage of the genes in the duplicated region plays an important role in the regulation of the seminal root growth. We also describe a radiation mutant with a deletion in the inserted 1BS segment and the adjacent 1RS region that restored long roots, confirming the importance of the dosage of the genes in this region on root development. Finally, we identified 38 genes within this deletion and used published RNA-sequencing data and annotation to discuss their potential as candidates for the genes regulating seminal root elongation in wheat.The genetic stocks including the 1RS and 1RSRW chromosome arms were initially generated in the cultivar Pavon 76 , a spring wheat developed at the International Maize and Wheat Improvement Center . The 1RS chromosome arm translocation in Pavon was introgressed from the CIMMYT cultivar Genaro, which, in turn, received the translocation from the cultivar Kavkaz . The donor of the 1RS arm in Kavkaz was the rye cultivar Petkus, one of the leading rye cultivars in the 20th century.To name the different chromosome constitutions we used two superscripts, with the first superscript indicating the proximal position and the second superscript the distal position. The 1RSRW chromosome arm was generated by homologous crossover in overlapping wheat segments of the primary 1BS–1RS recombinant T-9, which possessed a distal wheat 1BS segment, and 1B+40, which possessed a distal 1RS segment . The 1RSWR arm was generated by a crossover in overlapping wheat segments in primary 1BS–1RS recombinants T-38, which possessed a large distal wheat 1BS segment, and 1B+44, which possessed a long distal 1RS segment . The 1RSWW chromosome was generated by a crossover between 1RSRW and 1RSWR chromosomes and was designated as chromosome MA1 in Lukaszewski . The lines carrying the 1RSRW, 1RSWR, and 1RSWW chromosomes were previously back crossed into the CIMMYT common wheat cultivar Hahn, which has the 1RS.1BL translocation, with 1RS also originating from cultivar Kavkaz, the same as in Pavon-1RS.
The introgressions involved six marker-assisted back crosses, resulting in near-isogenic lines that were deposited in the National Small Grains Collection as accessions PI 672839 , PI 672838 , and PI 672837 . We have previously shown that the 1RSRW chromosome results in short roots in the Hahn background but not in the Pavon background. Therefore, to analyze the effects of different 1RS/1BS recombinant chromosomes on root length, we back crossed primary recombinants with varying lengths of wheat and rye segments—T-9, T-18, T-21, and 1B+40 —four times into Hahn. Line T-21 is identical to T-9 and line T-18 carries a large distal 1BS segment on its 1RS arm and was used as 1BS reference in the calculation of ratios for copy number determination. Line 1B+37, which carries a large distal 1RS segment on its 1BS arm , was used as 1RS reference in the exome capture comparisons but was not used in the hydroponic experiments.To dissect the chromosome region affecting root length, we irradiated 5,000 wheat F2 seeds from the cross between Hahn × Hahn-1RSWW with 300 Gy . This mutant population was established in 1RSWW before we knew which wheat segment was affecting root length. The objective of mutagenizing F2 plants rather than homozygous plants was to detect deletion mutants in the heterozygous plants of the first generation without having to wait for progeny tests. We extracted DNA from the 2,200 mutagenized plants that survived and used a dominant wheat marker and a dominant rye molecular marker to eliminate plants that were homozygous for the 1RS or 1BS segments. We identified 907 plants that were heterozygous for the proximal segment, of which, we expected the majority to also be heterozygous for the distal 1BS segment. We then screened the selected plants with multiple markers for the distal 1BS insertion and identified one mutant . From the progeny of this plant, we selected two sister homozygous plants, designated hereafter as 1RSWW-del-8 and 1RSWW-del-10. We then back crossed these two deletions independently to Hahn-1RSRW and to Hahn four times to reduce background mutations and to test the effect of the deletion on the root length in both backgrounds. Although the two lines carry the same deletion, independent back crosses increase chances of eliminating different background mutations, and they served as biological replicates in the root length experiments.We performed two exome capture experiments using different platforms. In the wheat exome capture using the assay developed by Arbor Bio-sciences, we included lines T-9, T- 18, T-21, 1B+37, 1B+40, and 1RSRW ethyl methanesulfonate mutant lines RW_M4_43_11 and RW_M4_47_12 . In the wheat exome capture using the assay developed by NimbleGen , we included lines 1RS, 1RSRW, and deletion lines 1RSWW-del-8 and 1RSWW-del-10. Based on the average similarity between the wheat and rye genes and the hybridization conditions used in the capture, we expect most of the rye genes to be captured with both wheat exome capture assays. The exome captures were sequenced using the Illumina platform and 150 bp paired-end reads at the University of California, Genome Center.
The sequencing reads were preprocessed to trim adapters with Trimmomatic v0.39 . Since the capture included both wheat and rye reads, we mapped the reads to a combined reference including wheat Chinese Spring RefSeq v1.0 chromosome 1B and the rye chromosome arm 1RSAK58 from the 1RS.1BL translocation in cultivar Aikang58 . To minimize off-target mapping, we mapped the reads at high-stringency with ‘bwa aln’ v0.7.16a-r1181 , allowing only perfectly mapped reads . Alignments were sorted by using samtools v1.7 , and duplicate reads were removed with Picard tools v2.7.1 . We normalized the number of mapped reads so that all lines have the same total number of reads mapped to the chromosome arm 1BL. We selected the 1BL arm as reference because 1RS/1BS recombinant lines differ in their short arm constitutions, but all share identical 1BL arms. We then calculated normalized read number ratios using a common reference line . We generated heat maps for these ratios and visually determined the borders of duplication, recombination, and deletion events. We then validated these borders using t tests of the ratios at both sides of the border . For these analyses we excluded genes with less than six reads in the accessions used as denominator for normalization. We report wheat gene coordinates using CS RefSeq v1.1 and rye gene coordinates using the 1RSAK58 genome as references , which is almost identical to our 1RS sequence. The other available genome reference for rye inbred line Lo7 is less similar to the 1RS sequences from Hahn 1RS.1BL translocation.Hydroponic experiments were performed in growth chambers at 22–23 ˚C with a photoperiod of 16 h light vs. 8 h dark . In all experiments, grains were imbibed at 4 ˚C for 4 d and then placed at room temperature. Once the coleoptiles emerged, seedlings were floated on a mesh to develop roots for 4 d. After removing the grain, seedlings were wrapped at the crown with foam and inserted in holes precut in a foam core board placed on top of the solution. The detail protocols and solutions are described in our previous paper . As in our previous study,flood and drain tray experiments in this study were performed in two different laboratories in Argentina and the United States using tanks of 0.35 and 13 L, respectively. As a result of the different conditions, final root lengths differed across experiments. However, differences among genotypes were consistent across experiments, and all statistical comparisons among genotypes were performed within experiment or using experiments as blocks. In experiments performed in 13-L tanks, we changed nutrient solution every 3 d and we included all genotypes in each tank. When necessary, we used multiple tanks as blocks. In experiments performed in 0.35- L tanks, we changed nutrient solution every 2 d, and a single genotype was included per pot, with multiple pots used as replications. To determine the effect of the 1RSWW-del-8 and 1RSWW-del-10 deletions on root development, we evaluated the segregating plants in the BC2F2 and BC4F2 generations to account for potential random effects of residual deletions in other chromosomes.To define the borders of the inserted 1BS region, we used the Arbor Biosciences exome capture to characterize the 1RSRW line and its two parental lines 1B+40 and T-9 .
We also included line T- 21 that appears to be identical to T-9 , line T-18 that has a distal 1BS segment longer than T-9/T-21 and was used as a wheat reference, and line 1B+37 that has a longer distal 1RS segment than 1B+40 and was used as a rye reference. We mapped the reads of each capture to a combined reference without allowing any SNP and then normalized the counts to a similar number of mapped reads per capture in the 1BL arm.The recent sequencing of the 1RS arm revealed the presence of a large inversion between the distal region of chromosome arms 1RS and 1BS , which suggests that lines with breakpoints within this region, such as T-9, T-21 and 1B+40, may be more complex than originally thought. The 1RSRW line was generated by a crossover of the primary recombinant lines T-9 and 1B+40 , and the previous results indicate that 1RSRW has retained the 1RS-1BS breakpoints of T-9 and 1B+40 . The 1RSRW chromosome arm also has the same strong telomeric C-band as 1RS and 1B+40, indicating that it has retained the complete 1RS segment present in 1B+40 . We initially assumed that the 1BS segment in 1RSRW replaced the orthologous rye genes and that the loss of these genes could be responsible for the shorter roots of Hahn- 1RSRW. However, the codominant marker THdw11 has both the 1RS and 1BS bands in T-9, 1B+40, and 1RSRW but not in T-18 or 1B+37 , suggesting a duplication rather than a replacement in the lines with distal crossover events. To investigate the extent of this duplication, we first identified 14 orthologous 1BS-1RS gene pairs including high-confidence wheat genes located within the 1BS insertion and rye 1RSAK58 genes that were at least 90% identical with an aligned region covering >90% of the gene . Surprisingly, all 14 ryeorthologues were present in the exome capture of T-9, 1B+40, and 1RSRW , which indicated that the complete rye region orthologous to the 1BS insertion was present in these lines. Since no 1RS gene was missing in the 1BS orthologous region, we rejected the hypothesis that lost rye genes were responsible for the differences in root length between Hahn- 1RS and Hahn-1RSRW isogenic lines.