*4.3. qDTY1.1-Specific Contributions to Drought Tolerance*

Of the 16 N22-derived introgressions in DTY-IL, two overlapped with known *qDTY*s for which N22 was a known donor [21]. While the large introgressions on chromosome 1 contain the full *qDTY*1.1 region, a smaller introgression on chromosome 3 partially overlaps with *qDTY*3.2 (Figure 10). Working under the assumption that the genome-wide changes in transcriptomes between the drought-tolerant and drought susceptible genotypes in both flag-leaves and panicle, would at least partially trackback to either transcriptional or allelic di fferences of specific loci within the *qDTY* regions we took a closer look at the *qDTY*1.1 fine mapped region and *qDTY*3.2 overlap.

Based on the Nipponbare reference, the fine-mapped *qDTY*1.1 region encompasses 79 genes, of which six were di fferentially expressed between DTY-IL and Swarna under RDS in either tissue (Table S15). An auxin-responsive protein (LOC\_Os01g67030), specifically upregulated in DTY-IL panicle under RDS was considered as a likely causative candidate. Generally, auxin has been shown to negatively regulate drought adaptation in plants [63]. Notably, the DEEPER ROOTING 1 (DRO1) promoter was shown to contain auxin-responsive elements (AuxRes) and negatively regulated by auxin [102]. While two AuxRes were found in the LOC\_Os01g67030 promoter of Nipponbare and MH63v2 (-513 bp and -1606 bp) only one was found at position -1573 bp in N22v2. The absence of the proximal AuxRe motif, in addition to the presence of novel, putative drought-responsive elements (Figure S18) could explain the observed di fferential regulation of LOC\_Os01g67030 under RDS in DTY-IL.

LOC\_Os01g67030.1 contains a cytochrome b561 (Cyt\_b561) and a dopamine β-monooxygenase (DOMON) domain (Figure S19). Cyt\_b561 proteins are involved in the regeneration of ascorbate through transmembrane electron transport [103,104] and have previously been implicated in drought tolerance through redox homeostasis [105]. The functionally uncharacterized β sheet-rich DOMON domain has been implicated in sugar and heme recognition [106] and predicted to be involved in protein-protein interactions, putatively functioning in metabolic signaling, in redox reactions, or both. Interestingly, LOC\_Os01g67030 thus has the potential to link sugar signaling and ROS signaling, both of which have emerged as essential in the DTY-IL-specific drought response.

The 5 AA di fferences in the N22v2 prediction of LOC\_Os01g67030 sequence fall under the two conserved domains (Figure S19). Notably they include two proline conversions and a glycine deletion, with potential structural implications, particularly in the context of transmembrane domains and β sheets [103,104]. This could a ffect the ability of the Cyt\_b561 domain to mediate transmembrane transport and the DOMON domain to mediate protein-protein interactions or ligand binding.

Efficient ROS scavenging was identified as a key mechanism of RDS tolerance in both panicles and flag-leaves of DTY-IL. In the panicle, LOC\_Os01g67030 could directly contribute to ROS homeostasis and with ROS being increasingly implicated in stress signaling including modulation of gene expression [107,108], LOC\_Os01g67030 activity could be responsible for the some of the expression changes observed in P-M10 and P-M15. LOC\_Os01g67030, however, was not expressed in flag-leaves. It is possible that the ultimate positive effects of the N22 allele of LOC\_Os01g67030 on seed setting in DTY-IL panicles could increase sink strength in a way that it positively affects the source strength of flag-leaves, which could contribute to maintained photosynthetic rates. A similar sink on source effects has been demonstrated by manipulation of SnRK1 dependent metabolic signaling in maize under control and drought [109].

LOC\_Os03g03510 was found downregulated in DTY-IL. Both CIPK\_C domain [110] and SnRK3 domain [111,112] have been implicated in abiotic stress responses, including drought tolerance. In addition, SnRK3, like SnRK1 [62,109] has been demonstrated to function in metabolic signaling and source-sink relationships. In sinks, SnRK1 activity has detrimental effects on grain filling [109]. A 35 AA C-terminal extension in the N22 allele could have functional implications, which, in addition to its observed downregulation could reduce its efficiency in DTY-IL. The postulated effect would be altered downstream phosphorylation responses with potential transcriptional changes that reflect some of the differences seen between Swarna and DTY-IL Ultimately this could contribute to maintained sink strength of the panicle with putative effects on flag-leaf source metabolism.

In theory the postulated functions of both candidates could have synergistic effects that could explain a range of the observed DTY-IL specific drought responses. Gene validation studies expressing the N22 allele of LOC\_Os01g67030, LOC\_Os03g03510, or both, under control of their native promoters in drought susceptible *indica* background, respective knock-outs in DTY-IL, or both, are needed to confirm their postulated roles.
