We also observed a significant increase in the transcripts of the JA-regulated genes PLETHORA 1 , PLT3 and PLT5 in 1RW and UBI::OPRIII-R5 relative to 1RS at 16 DAG but not at 6 DAG . These results support a role of the OPRIII genes on the JA-Ile signaling pathway, which is known to be involved in the regulation of root architecture. To pharmacologically test if the differences in JA-Ile were responsible for the reduced seminal root growth of the 1RW lines, we used Ibuprofen , a known inhibitor of the JA biosynthetic pathway. The addition of 5 μM-IBU in the hydroponic culture from 6 DAG resulted in similar seminal root elongation in 1RS and 1RW by the end of the experiment at 16 DAG . The addition of 5 μM-IBU also accelerated seminal root elongation in 1RW when added from 8 or 10 DAG, but not from 12 DAG , suggesting that IBU is no longer effective when the JA-Ile signaling cascade is already induced. The addition of 5 μM-IBU also eliminated the differences in the distribution of ROS between 1RW and 1RS . These results indicate that differences in both root length and ROS distribution between 1RS and 1RW are likely mediated by changes in JA-Ile. We also analyzed the effect of combined levels of 5 μM-IBU and 40 μM-MeJA applied from 6 DAG on 1RS and 1RW root length. Plants treated only with IBU showed similar results to those in the previous experiment,large pot with drainage whereas those treated only with MeJA showed inhibited root growth . The MeJA treatment alone reduced but not eliminated the differences between 1RS and 1RW genotypes.

At 20 DAG, the differences between genotypes were significant in the MeJA treatment but not significant in the combined MeJA-IBU treatment . A three-way ANOVA showed significant differences for the threemain factors and the three two-way interactions . The significant genotype x MeJA and genotype x IBU interactions are consistent with the hypothesis that the root length differences between 1RS and 1RW are associated with changes in the JA bio-synthetic / signaling pathway. These results are also consistent with the observed differences in endogenous JA/JA-Ile levels in the roots between genotypes and the demonstrated role of the OPRIII genes on the differences in root length in the transgenic plants.A comparison of the transcriptomes of the distal 1 cm of the seminal roots between 1RS and both 1RW and UBI::OPRIII-R5 revealed a large number of differentially expressed genes at both 6 DAG and 16 DAG . The large number of DEGs indicate major developmental changes in the distal region of the seminal roots. Two lines of evidence indicate similar changes in the transcriptomes of 1RW and UBI::OPRIII-R5 relative to 1RS . First, we observed a large number of shared DEGs at 6 DAG and 16 DAG . In addition, we detected a highly significant regression between the log2 expression ratios in 1RW/1RS and in UBI::OPRIII-R5/ 1RS for all genes with more than two-fold difference in expression . These results indicate similar transcriptome changes in 1RW and UBI::OPRIII-R5 relative to 1RS, supporting the hypothesis that the observed changes in 1RW are mainly driven by the increased dosage and expression of the OPRIII genes. To characterize the main pathways affected by the changes in OPRIII expression, we carried out an additional Quant-Seq transcriptome analysis comparing 1RW and mut-OPRIII-B1 . We performed KEGG analyses using the closest rice homologs of the DEGs between the 1RW/1RS and UBI::OPRIII-R5/1RS comparisons at 6 and 16 DAG and for RW/mut-OPRIII-B1 comparisons at 6 and 20 DAG .

These analyses showed three pathways that were significantly enriched in all three comparisons in any combination across the two time points . The P values for significant enrichment for each pathway were calculated using a 2 × 2 contingency table comparing the number of hits over the total in the selected list and the population . The significant enrichment in the alpha-linolenic acid and glutathione pathways are likely associated with the observed changes in JA and ROS in the roots, respectively. An additional pathway significantly enriched at both time points was the phenylpropanoid biosynthetic pathway, which is critical for the establishment of root water barriers. In summary, the three root transcriptome datasets presented in this study represent a valuable resource for functional studies of root genes in wheat.Altered levels of the phytohormone JA-Ile were shown in different studies to affect root architecture and responses to drought in different plant species. However, natural variation in JA- biosynthesis and signaling, as well as its incidence in root architecture of major crops remains largely unknown. This work shows that differences in OPRIII gene dosage underlie the variation between 1RS and 1RW in root architecture, in their ability to access water from deep soil layers, and in their yield potential under limited water condition. Knock-outs of single members of this multigene family were sufficient to generate changes in seminal root length in 1RW, suggesting that this is a sensitive regulatory point of the JA pathway. We show here that the understudied genes from the monocot specific OPRIII subfamily encode cytoplasmic and nuclear 12- OXOPHYTODIENOATE REDUCTASE enzymes that regulate a critical step in the synthesis of JA-Ile. Using transcriptome and pharmacological studies, we demonstrate that the effects of the OPRIII dosage or expression on root development are mediated by changes in JA-Ile and the downstream JA-signaling pathway in the distal region of the seminal roots.

The addition of the JA-biosynthesis inhibitor IBU not only eliminated the differences between 1RS and 1RW in root length, but also in ROS distribution . Our previous results indicate that the JA and ROS pathways may be interconnected, a hypothesis also supported by a previous study in Arabidopsis showing that changes in ROS distribution play an important role in root stem cell maintenance. There is also evidence that ROS, and in particular glutathione, are important components of the root growth regulatory pathway in maize affecting both the meristematic and elongation zones. This connection is supported by our root tip transcriptome studies. Our KEGG analyses comparing the 1RS, 1RW, OPRIII mutants and transgenic plants over expressing OPRIII showed consistent effects on the glutathione pathway, which is known to contribute to the control of redox homeostasis. Another consistently enriched pathway in the KEGG analyses was the phenylpropanoid biosynthetic pathway, which is critical for the establishment of root water barriers. Autonomous production of phenylpropanoids is required for the establishment of the endodermal Casparian strip as well as for adherence of the suber in matrix to the cell wall of the endodermis. Additional support for the role of the OPRIII genes in the regulation of this pathway comes from the increased expression of the rice homologs Os06G0215500and Os06G0215600in the endo and exodermis of the rice roots. The rice root spatial heat map for OPRIII genes Os06G0215500, Os06G0215600, Os06G0215900 and Os06G0216300 also revealed increased expression of the OPRIII genes in the apical region, including the quiescent center . This information, combined with the expression of the wheat OPRIII genes in the distal part of the seminal roots at different time points and the developmentally regulated arrest of the RAM in 1RW and UBI::OPRIII transgenic plants, suggest a role of the OPRIII genes in the developmental regulation of the RAM in cereal plants. In Arabidopsis, PLT genes have been shown to be dose-dependent master regulators of root development,drainage collection pot and in rice they are expressed in the root stem cell niche and in the nascent lateral roots. Therefore, it is possible that the observed increase in the expression of the wheat PLT1, PLT3, and PLT5 in the lines with increased OPRIII dosage or expression contributed to the arrest of the root meristem and/or to the different distribution of the lateral roots. Since PLT protein gradients are critical for their roles in root development, it would be interesting to investigate the effect of changes in OPRIII dosage on the spatial distribution of wheat PLT proteins along the roots. In addition to the basic biological questions that can be investigated by the manipulation of the OPRIII genes, this study also points to potential practical applications. One example is the increased root length in the 1RW CRISPR-Cas9 mutants, which provides a path to restore the good performance of the 1RS lines under water stress to the 1WW line with improved bread making quality. However, the seminal roots of the single gene mutants were still shorter than the original 1RS line , suggesting that the OPRIII gene dosage may need to be fine-tuned to restore root growth and grain yield potential to the 1RS levels. Finally, the extensive variation detected in the number of functional OPRIII genes in the available sequenced wheat genomes suggests that natural variation in these genes may have contributed to the adaptation of wheat to different soil environments.

The identification of OPRIII gene dosage as a sensitive point in the regulation of the JAbio-synthetic pathway provides a target to engineer root architecture in wheat and possibly other cereal crops.An isochromosome 1RS was developed by centric mis-division of the 1RS.1BL translocation in wheat. This translocation originated from the Kavkaz source of the translocation, via cv. Genaro. After self-pollination, a diisosomic 1RS line of Pavon 76 was isolated. For this study, the diiso 1RS of Pavon 76 was crossed, and back crossed four times to Hahn-1RS with cytological selection for the presence of diiso 1RS in each generation. The BC4F1 was self-pollinated and individual plants with diiso 1RS as well as telo 1RS were isolated, grown and selfpollinated. As Hahn-1RS has the 1RS.1BL translocation, the homozygous diiso 1RS addition line has six doses of the rye chromosome 1RS. A cytogenetic study of the progeny of a diiso 1RS plant showed that this extra diiso chromosome is not stable and is lost in approximately 31% of the progeny . To determine the 1RS copy number variation in 23 progenies of a diiso 1RS plant, we performed six independent DNA extractions for each of the 23 plants, and from 8 Hahn-1RS and 8 Hahn- 1RW additional plants as controls. We then used qRT-PCR primers qrt1RS5-F and -R to determine the dosage of the OPRIII-R5 gene located in the 1RS chromosome arm. We used the CO2 gene as an endogenous control for a single copy gene. For each of the 23 progenies and 16 controls, we determined root length in hydroponic tanks as described below, and then calculated a regression between the 1RS CNV and root length using SAS v9.4. To select plants with more than two 1RS chromosomes, we performed t tests between the CNV in each recombinant and the 1RS control . Root length in plants with more than two 1RS arms were compared with plants carrying the 1RS or 1RW controls. Raw data and statistical analyses for these experiments are available in source data of Fig. 1.To overcome the redundancy of OPRIII homologs and paralogs, we first amplified OPRIII-B3 and OPRIII-B1 from 1RW with primers on the UTR region . We then cloned the PCR products into T vector and sequenced them to confirm the presence of the complete coding region. We next added the attB site to the coding region using primer OPRIII-B-attB-F combined with either OPRIII-B3-attB-R or OPRIII-B2- attB-R . We were not able to clone OPRIII-B2 and OPRIII-A2 from the reverse transcription products, so we cloned each exon from genomic DNA using the primers described in source data of Supplementary Fig. 3. The capillary voltage was 1 kV in positive ion mode and 2 kV in negative ion mode. A quality control sample, generated by pooling equal aliquots of each sample, was analyzed periodically to monitor system stability and performance. Samples were analyzed in random order. Skyline-Daily software was used to detect and integrate peak areas. Exported peak areas was used to calculate linear regression of analytical standards used for quantification. JA was normalized to isotopically labeled internal standard. The corresponding linear regression equation was used for quantification for each analyte, which was then adjusted for precise weight of leaf tissue for each sample . For JA-Ile and OPDA, for which there are not internal standards for, peak areas were used for final statistical analyses.