Natural variations at the Stay-Green gene promoter control lifespan and yield in rice cultivars

Increased grain yield will be critical to meet the growing demand for food, and could be achieved by delaying crop senescence. Here, via quantitative trait locus (QTL) mapping, we uncover the genetic basis underlying distinct life cycles and senescence patterns of two rice subspecies, indica and japonica. Promoter variations in the Stay-Green (OsSGR) gene encoding the chlorophyll-degrading Mg++-dechelatase were found to trigger higher and earlier induction of OsSGR in indica, which accelerated senescence of indica rice cultivars. The indica-type promoter is present in a progenitor subspecies O. nivara and thus was acquired early during the evolution of rapid cycling trait in rice subspecies. Japonica OsSGR alleles introgressed into indica-type cultivars in Korean rice fields lead to delayed senescence, with increased grain yield and enhanced photosynthetic competence. Taken together, these data establish that naturally occurring OsSGR promoter and related lifespan variations can be exploited in breeding programs to augment rice yield.

Starting with an interest on the difference in the speed of leaf senescence after heading between indica rice (early senescence in general) and japonica rice (late senescence in general), authors cloned a causal gene for the different senescence pattern by mapping-based approach. The gene encodes a reported chlorophyll-degrading protein OsSGR that has been functionally characterized. After confirming the genetic effect of OsSGR gene on leaf and panicle senescence, this work further explored the natural variation of this gene and found the variations in the promoter, rather than coding sequence, determine the difference in senescence between indica and japonica, and revealed that the indica and japonica alleles are originated from nivara and rufipogon rice, respectively , which agrees with evolution rout of the two subspecies of cultivated rice. The work also demonstrated the potential value of this gene in increasing yield by introducing the japonica allele for late senescence into india cultivars. Although the OsSGR has been reported for its function in delaying senescence and post-flowering leaf stay-green has also been recognized for its effect in increasing grain filling and yield, this work provided solid genetic evidence and explored natural variation of the gene together with the origination of the indica and japonica allele, and proposed a evolutionary mechanism of the senescence and death for the r-selection life history trait. Therefore, this work provided very limited new knowledge on the function of the OsSGR protein in senescence but some useful information on the variation, evolutionary history and potential breeding value of this gene. However, for the potential value in breeding (high yield), authors may need consider how to balance the degree of stay-green traits because over stay-green, or greed for green after heading is not desirable and even bad for rice production. In addition, according to my expertise, variations in leaf and panicle senescence also exist within india or japonica rice, and breeders have selected optimal life span including leaf senescence in a specific rice-growing region. Wise strategy is needed to explore this gene in breeding. In general, the data in this work is of high quality. However, I noticed the transgenic rice overexpressing the OsSGR genes, activation lines and the KO mutant were not shown for the grain-filling rate and yield. Although the NILs in indica background with japonica allele showed significant increased in grain-filling rate and yield, it cannot be excluded that the effect is contributed by other unknown genes since no data to show how 'near' the NILs are. It could be more conclusive if grain-filling rate and yield data is collected for the OE and mutant.
Reviewer #2 (Remarks to the Author): Leaf senescence is the final stage of leaf development and the enlargement of this period is related to grain yield in rice. Between indica and japonica, two species, there is difference in this period and the indica rice has a rapid life cycle (early senescence). In this report, Shin D and his colleague indicated that natural variation in OsSGR promoter region might determine the lifespan variations between two species and the modify in this region could enhance grain yield. Unfortunately, I regret to inform that this manuscript cannot be considered for publication in Nature Comm. because it does not meet the requirements of the journal at least in this version.
Major point 1. First, ten or more genes have already been identified as the genes that cause stay-green (e.g., Leng et al., Int. J. Mol. Sci. 2017). Therefore, it is difficult to think that the lifespan variation between japonica and indica species can be determined by only one gene examined in this paper. The authors should clarify if there is a similar variation at least in several major genes. 2. It also relates to the relationship between the lifespan and the yield, which is another main theme of this paper. The idea that flag leaves, which are the main source organs of rice, could maintain photosynthetic ability for a long period and continue substance production leads to an increase in yield. Meanwhile, it was reported that OsSGR (senescence-inducible chloroplast staygreen protein I), the gene examined in this paper maintained the green color of leaves but did not affect photosynthetic ability and yield (Jiang H et al., Plant J. 2007). Authors should analyze photosynthetic ability of flag leaves or a canopy through growth stage and add the data of growth analysis (NAR) to clarify factors of yield characteristics.
Minor point 1. The authors should add the information about stay-green of rice or Arabidopsis into first paragraph and explain the relation between lifespan and senescence. 2. pp.4 line 88, please add accession number of OsSGR. 3. pp.8, Authors should discussion about the reason why the specific promoter region that causes short lifespan has been selected in indica but not in japonica through the domestication.
Reviewer #3 (Remarks to the Author): In this manuscript it has been convincingly demonstrated that yield of an important crop plant, i.e. rice, can be enhanced by delaying the senescence process. This is a result of high importance with regard to the increasing demand in food supply. The study has been performed with two agronomically important rice varieties having contrasting senescence processes. The molecular reason underlying the fast senescence feature of indica rice in comparison to japonica rice has been elucidated by QTL analysis and a broad repertoire of further sophisticated molecular analyses. The study shows that variations in the promoter of the so-called STAYGREEN gene encoding an enzyme of chlorophyll catabolism are responsible for high expression of the gene and the fast senescence phenotype of the rice variety indica. By investigating the STAYGREEN sequences of wild rice species the researchers authors could track the evolution of the sequences responsible for the fast senescence of indica rice and demonstrate that introgression of STAYGREEN alleles from japonica into indica cultivars increased yield. The results here presented have been obtained by a multitude of long-standing elaborate studies including field trials. A considerable part of the data is contained in the supplement. All data is presented sufficiently detailed and transparent to allow for an understanding of this very impressive work. Appropriate statistical tests have been done and are well documented. The conclusions are robust, valid and reliable. Indeed an increase in yield of about 10% is quite substantial. If the space would allow it, a comment on the different staygreen characters might be appropriate. Not in all conditions a staygreen character is advantageous, as detailed in the review of Gregersen et al. (2013, PMB). Specific comments: Line 25-26 -The introduction about the importance of senescence for crop yield should reference the highly cited paper on senescence and crop productivity of Gregersen et al. (2013, Plant Molecular Biology). Line 43: r-selection has to be explained to make the paper accessible for non-experts Line 60 and later in the text: The numbering of leaves should be changed. The flag leaves is certainly not the first leaf, but actually the last leaf. Accordingly, the "second" leaf is not the second leaf, but the first leaf below the flag leaf. The first leaf is the primary foliage leaf! The flag leaf is the leaf with the highest number, which depends on the species and variety. Line 88: For non-expert readers it might be misunderstanding that a gene called STAYGREEN accelerates senescence. The authors should clearly indicate that the gene has been detected to be impaired in a staygreen mutant and hence has been named accordingly. Last but not least the style of writing should be changed. Please avoid the use of the first person and write the complete manuscript in the third person. The first person is only appropriate for beliefs and convictions and is usually used in autobiographies. The description of scientific results rather requires impersonal reporting using third person pronouns instead of "we", because the results hopefully exist without the reflections of the writing persons.
Reviewer #4 (Remarks to the Author): The manuscript "Natural variations at the Stay Green gene promoter control lifespan and yield in rice cultivars" describes the identification of OsSGR as the gene responsible for the difference in lifespan between the two major rice subspecies, japonica and indica. Analysis on sequences of OsSGR in different rice accessions suggested that sequence variations in the promoter region are critical in the evolution of the indica subspecies as a rapid cycling rice in which higher expression of the OsSGR gene triggers earlier senescence of leaves and panicles. Results from this study could potentially bring about a better understanding of rice evolution in light of plant senescence and the authors have proposed potential application of this locus in rice breeding to increase yield.
One critical question needs to be addressed, however, is the discrepancy between the increase in yield in rice plants with reduced OsSGR expression in this study, and the results from earlier studies indicating that sgr is a nonfunctional stay-green locus (Park et al., 2007, Plant Cell;Sato et al., 2007, PNAS). For nonfunctional stay-green mutants, the plants retain high chlorophyll contents in senescent leaves but their photosynthetic competence decreases normally during senescence. This type of mutations is often associated with loss of function in genes involved in chlorophyll degradation, which is believed to be at relatively downstream part of the senescence program. The fact that OsSGR encodes for a chlorophyll-degrading Mg2+-dechelatase, is consistent with sgr being a nonfunctional stay-green mutant. In this study, however, the authors showed that when OsSGR expression was enhanced by introgression of the japonica OsSGR allele into indica type cultivars, the plants exhibited delayed senescence and increased yield.
I would suggest the author to address this question by measuring photosynthesis competence and examining senescence marker gene expression in the mutants, transgenic plants, and introgression lines to show that indeed enhanced expression of OsSGR is able to delay the senescence program -not just chlorophyll degradation. If that's the case, the authors also need to explain why in this case, the whole senescence process can be delayed by simply blocking a step in chlorophyll degradation.

Reviewers' comments:
Reviewer #1 (Remarks to the Author): Starting with an interest on the difference in the speed of leaf senescence after heading between indica rice (early senescence in general) and japonica rice (late senescence in general), authors cloned a causal gene for the different senescence pattern by mapping-based approach. The gene encodes a reported chlorophyll-degrading protein OsSGR that has been functionally characterized. After confirming the genetic effect of OsSGR gene on leaf and panicle senescence, this work further explored the natural variation of this gene and found the variations in the promoter, rather than coding sequence, determine the difference in senescence between indica and japonica, and revealed that the indica and japonica alleles are originated from nivara and rufipogon rice, respectively, which agrees with evolution rout of the two subspecies of cultivated rice. The work also demonstrated the potential value of this gene in increasing yield by introducing the japonica allele for late senescence into india cultivars. Although the OsSGR has been reported for its function in delaying senescence and post-flowering leaf stay-green has also been recognized for its effect in increasing grain filling and yield, this work provided solid genetic evidence and explored natural variation of the gene together with the origination of the indica and japonica allele, and proposed an evolutionary mechanism of the senescence and death for the r-selection life history trait. Therefore, this work provided very limited new knowledge on the function of the OsSGR protein in senescence but some useful information on the variation, evolutionary history and potential breeding value of this gene. However, for the potential value in breeding (high yield), authors may need consider how to balance the degree of stay-green traits because over stay-green, or greed for green after heading is not desirable and even bad for rice production.
-Response: We agree with reviewer's concern. Stay-green refers to the heritable delay of leaf senescence, which has been identified as an important component for the improving crop yield through extending of photosynthesis period in various crop species. For example, progressive increases in grain yield of maize were positively correlated with stay-green traits. However, most of previous results for stay-green characters in rice were studied by loss-of-function mutants and/or overexpression lines of specific genes, leading to over stay-green together with undesirable crop productivity, as the reviewer pointed out. For example, ossgr mutants exhibiting delayed leaf senescence did not show an advantage in grain yield.
The key point in our report is that the japonica OsSGR allele does not lead to over-stay green but shows difference in its induction kinetics and induction level compared to the indica allele. Thus, the japonica allele exhibits a capacity to balance the degree of stay-green traits when introduced into indica 2 varieties, as we have shown in this report that NILs-OsSGR in the indica varieties harboring a japonica OsSGR allele leads to delay of senescence in a japonica fashion with balanced stay-green in indica, not like stay-green knock out mutants. As a results, NILs showed the improved productivity along with the enhanced grain filling, which was useful for breeding purpose. We added this argument in Discussion.
In addition, according to my expertise, variations in leaf and panicle senescence also exist within indica or japonica rice, and breeders have selected optimal life span including leaf senescence in a specific ricegrowing region. Wise strategy is needed to explore this gene in breeding. In general, the data in this work is of high quality. However, I noticed the transgenic rice overexpressing the OsSGR genes, activation lines and the KO mutant were not shown for the grain-filling rate and yield.
Although the NILs in indica background with japonica allele showed significant increase in grain-filling rate and yield, it cannot be excluded that the effect is contributed by other unknown genes since no data to show how 'near' the NILs are. It could be more conclusive if grain-filling rate and yield data is collected for the OE and mutant. Leaf senescence is the final stage of leaf development and the enlargement of this period is related to grain yield in rice. Between indica and japonica, two species, there is difference in this period and the indica rice has a rapid life cycle (early senescence). In this report, Shin D and his colleague indicated that natural variation in OsSGR promoter region might determine the lifespan variations between two species 3 and the modify in this region could enhance grain yield. Unfortunately, I regret to inform that this manuscript cannot be considered for publication in Nature Comm. because it does not meet the requirements of the journal at least in this version.
Major point 1. First, ten or more genes have already been identified as the genes that cause stay-green (e.g., Leng et al., Int. J. Mol. Sci. 2017). Therefore, it is difficult to think that the lifespan variation between japonica and indica species can be determined by only one gene examined in this paper. The authors should clarify if there is a similar variation at least in several major genes.
-Response: We understand the reviewer's concern. As reviewer has pointed, leaf senescence is controlled by a complex set of genes with sophisticated processes. We agree with the reviewer that there will be other genes that lead to the different senescence between indica and japonica varieties. We also described in this report that there are several QTL loci that contribute to the different senescence between indica and japonica varieties (Fig. 2a, b). In this report, we focused on one of those loci that significantly contribute to the different senescence phenotype between indica and japonica varieties. We added the point of the reviewer in Introduction with the reference (page 3, line 63-66). We like to note that we are trying to narrow down for another leaf senescence QTLs as a separate project.
2. It also relates to the relationship between the lifespan and the yield, which is another main theme of this paper. The idea that flag leaves, which are the main source organs of rice, could maintain photosynthetic ability for a long period and continue substance production leads to an increase in yield. Meanwhile, it was reported that OsSGR (senescence-inducible chloroplast stay-green protein I), the gene examined in this paper maintained the green color of leaves but did not affect photosynthetic ability and yield (Jiang H et al., Plant J. 2007). Authors should analyze photosynthetic ability of flag leaves or a canopy through growth stage and add the data of growth analysis (NAR) to clarify factors of yield characteristics.
-Response: We thank for the reviewer's suggestion. In the revised manuscript, we added the data, as the reviewer requested. Following the reviewer's comment, we analyzed the photosynthetic ability of flag leaves of four NILs grown in field to evaluate the effect on grain yield. Three NILs of the indica background harboring japonica OsSGR allele displayed extended the photosynthetic capability with higher chlorophyll contents (Fig. 4g, i, and k, and Supplementary Fig. 17b), leading to the improved grain-filling rate and grain yield. Therefore, we inferred that japonica OsSGR allele in indica background 4 could increase the grain productivity by maintaining a higher chlorophyll content and photosynthetic capability of flag leave during grain filling period. On the contrary, NILs of the japonica background harboring indica OsSGR allele displayed decreased the photosynthetic capability with less chlorophyll contents, resulting in poor grain filling rate and productivity, as compared with parental japonica (Fig. 4g, i, k).
Minor point 1. The authors should add the information about stay-green of rice or Arabidopsis into first paragraph and explain the relation between lifespan and senescence.
-Response: As suggested by reviewer, we added information about stay-green in the text (pp 3, line 55-63).
-Response: As suggested by the reviewer, we included the accession number of three candidate genes (page 5, line 119-120).
3. pp.8, Authors should discuss about the reason why the specific promoter region that causes short lifespan has been selected in indica but not in japonica through the domestication.
-Response: We thank for this comment. According to the reviewer's comment, we added the following point in the Discussion part (page12, line 259-277). The following is our idea. Optimal rice lifespan of varieties has been considered and selected for maximum productivity and/or economic income according to their cultivation climate and cropping system. In the single cropping region including most of japonica cultivation area, grain filling rate is important agricultural trait for maximum crop production. Therefore, the lower transcript levels of OsSGR in japonica could extend photosynthesis capacity during grain filling period for improved productivity. On the contrary, most of indica cultivation region has double or triple cropping system. Therefore, cultivation period with maximum production should be considered. For this reason, promoter of OsSGR in indica was naturally selected for the faster and higher expression level of OsSGR. However, natural selection on coding region of OsSGR was not occurred during domestication because ossgr mutants exhibiting impaired chlorophyll breakdown did not affect productivity. We Line 43: r-selection has to be explained to make the paper accessible for non-experts -Response: We thanks for reviewer's comment. We revise the text as reviewer suggested (page 2, 3, line

43-47)
Line 60 and later in the text: The numbering of leaves should be changed. The flag leaves is certainly not 6 the first leaf, but actually the last leaf. Accordingly, the "second" leaf is not the second leaf, but the first leaf below the flag leaf. The first leaf is the primary foliage leaf! The flag leaf is the leaf with the highest number, which depends on the species and variety.
-Response: We thank for the comments. According to the reviewer's comment, we revised the numbering of leaves through the manuscript; first leaf into flag leaf and second leaf into second uppermost leaf.
Line 88: For non-expert readers it might be misunderstanding that a gene called STAYGREEN accelerates senescence. The authors should clearly indicate that the gene has been detected to be impaired in a staygreen mutant and hence has been named accordingly.
-Response: We totally agree with the reviewer's concern. We included the meaning of the 'stay-green' term, which used for mutations in chlorophyll breakdown (page 3, line 58-61).
Last but not least the style of writing should be changed. Please avoid the use of the first person and write the complete manuscript in the third person. The first person is only appropriate for beliefs and convictions and is usually used in autobiographies. The description of scientific results rather requires impersonal reporting using third person pronouns instead of "we", because the results hopefully exist without the reflections of the writing persons.
-Response: We agree with the reviewer's concern. We revised the text through the manuscript according to the reviewer's suggestion.
Reviewer #4 (Remarks to the Author): The manuscript "Natural variations at the Stay Green gene promoter control lifespan and yield in rice cultivars" describes the identification of OsSGR as the gene responsible for the difference in lifespan between the two major rice subspecies, japonica and indica. Analysis on sequences of OsSGR in different rice accessions suggested that sequence variations in the promoter region are critical in the evolution of the indica subspecies as a rapid cycling rice in which higher expression of the OsSGR gene triggers earlier senescence of leaves and panicles. Results from this study could potentially bring about a better understanding of rice evolution in light of plant senescence and the authors have proposed potential application of this locus in rice breeding to increase yield.
One critical question needs to be addressed, however, is the discrepancy between the increase in yield in rice plants with reduced OsSGR expression in this study, and the results from earlier studies indicating that sgr is a nonfunctional stay-green locus (Park et al., 2007, Plant Cell;Sato et al., 2007, PNAS). For 7 nonfunctional stay-green mutants, the plants retain high chlorophyll contents in senescent leaves but their photosynthetic competence decreases normally during senescence. This type of mutations is often associated with loss of function in genes involved in chlorophyll degradation, which is believed to be at relatively downstream part of the senescence program. The fact that OsSGR encodes for a chlorophylldegrading Mg 2+ -dechelatase, is consistent with sgr being a nonfunctional stay-green mutant. In this study, however, the authors showed that when OsSGR expression was enhanced by introgression of the japonica OsSGR allele into indica type cultivars, the plants exhibited delayed senescence and increased yield. I would suggest the author to address this question by measuring photosynthesis competence and examining senescence marker gene expression in the mutants, transgenic plants, and introgression lines to show that indeed enhanced expression of OsSGR is able to delay the senescence program -not just chlorophyll degradation. If that's the case, the authors also need to explain why in this case, the whole senescence process can be delayed by simply blocking a step in chlorophyll degradation.
-Response: We understand the reviewer's concern. As a response to this and another reviewer, we now included the photosynthetic competence of flag leaves from four NILs grown in field. As described in detail below, our results clearly show that timing and level of OsSGR expression can control the whole senescence program in rice during grain filling period by regulating the photosynthetic capability not just the chlorophyll content.
The reason is as follows. Delay of chlorophyll loss may not be directly related to functional delay of senescence as well known . However, it is clear and well known that late degradation of chlorophyll leads to later functional senescence, as reported in many cases (Yoo et  The word 'Stay green' for the OsSGR gene can be easily misunderstood, as the knock out mutants shows 'stay green' phenotype (page 3, line 59-61), but the actual function of the OsSGR gene is chlorophyll degradation and the concomitant leaf senescence. We added this point in the text (page 5, 6, line 120-122). 8 Below, we describe the photosynthetic capacity data in detail. According to our analysis, three indica NILs having japonica OsSGR allele maintained their photosynthetic capability longer than their parental varieties (Fig. 4k and Supplementary Fig. 17b), and japonica NIL harboring indica OsSGR allele displayed decreased photosynthetic capability (Fig. 4k). As a senescence marker gene, we analyzed the expression of OsNAP (LOC_Os03g21060) in rice plants. Activation lines showing early senescence accumulated higher levels of OsNAP transcripts in flag leaves ( Supplementary Fig. 5d), delayed leaf senescence mutants including ossgr, CRISPR/Cas9-genome editing and RNAi-suppressed lines displayed the decreased expression of OsNAP, confirming their leaf senescence phenotype ( Supplementary Fig. 6d,   Supplementary Fig. 7c and Supplementary Fig. 8e). The expression of OsNAP in indica (IR72)-NIL harboring japonica OsSGR allele was lower, as compared to IR72, and its expression was higher in the japonica-NIL having indica OsSGR allele compared to parental japonica variety ( Supplementary Fig.   17a).

Reviewers' comments:
Reviewer #2 (Remarks to the Author): In this report, Shin D and his colleague indicated that natural variation in OsSGR promoter region might determine the lifespan variations between two species and the modify in this region could enhance grain yield. In previous review, I pointed out that authors should analyze photosynthetic ability of flag leaves or a canopy through growth stage and add the data of growth analysis (NAR) to clarify factors of yield characteristics. Because it is most important and valuable point of this report to clarify which the lower decrease in chr. content by their allele can maintain the photosynthetic ability and then increase in yield or not. However, in present version, they added the Fv/Fm (chlorophyll fluorescence) data in Figure 4k and Supplementary Fig. 17b. Their data could not indicate higher photosynthetic ability. For example, between IR72 and NIL the larger difference in chl. content did not reflect photosynthetic ability after 7 weeks after heading (Fig. 4k). Additionally, Fv/Fm value does not necessary match the photosynthetic ability (Harley et al. Plant Physiol. 1992). Unfortunately, I regret to inform that this manuscript cannot be considered for publication without adequate data and thorough discussion. Minor point 1. Authors should add the statistical analysis in supple. Fig. 3 and Table 4.
Reviewer #3 (Remarks to the Author): I accept the replies to my comments on the first version of this manuscript. However, I'm disappointed that the style of writing has not been changed throughout the text. I hence repeat: Please avoid the use of the first person (I, we) and write the complete manuscript in the third person (e.g. it has been shown...). The first person is only appropriate for beliefs and convictions and is usually used in autobiographies. The description of scientific results rather requires impersonal reporting using third person pronouns instead of "we", because the results hopefully exist without the reflections of the writing persons.
In the abstract in lines 23, 25, 31 and later in line 67 etc. the personal style has been used.
Reviewer #4 (Remarks to the Author): The authors have nicely addressed my concern about OsSGR being functional in regulating the senescence program (not just chlorophyll degradation) by showing that the indica NILs having japonica OsSGR allele maintained their photosynthetic capability longer than their parental varieties, and japonica NIL harboring indica OsSGR allele displayed decreased photosynthetic capability. They also analyzed the expression of the senescence marker gene OsNAP in rice plants. Activation lines showing early senescence accumulated higher levels of OsNAP transcripts in flag leaves while delayed leaf senescence mutants including ossgr, CRISPR/Cas9-genome editing and RNAi-suppressed lines displayed the decreased expression of OsNAP. The expression of OsNAP in indica (IR72)-NIL harboring japonica OsSGR allele was lower, as compared to IR72, and its expression was higher in the japonica-NIL having indica OsSGR allele compared to parental japonica variety. My second concern, about the discrepancy between OsSGR being a functional senescence regulator in this study and the results from earlier studies indicating that sgr is a nonfunctional stay-green locus (Park et al., 2007, Plant Cell;Sato et al., 2007, PNAS), is however not addressed. In Line 252-258, the authors stated that "The japonica OsSGR allele does not lead to an over stay-green but shows difference in its induction kinetics and induction level compared to the indica allele (Fig 3a). The japonica allele appears to exhibit a capacity to balance the degree of staygreen trait, photosynthetic competence, nutrient remobilization, when introduced into indica varieties, to lead to increased grain filling and productivity during the senescence period. Thus, NILs-OsSGR in the indica background harboring a japonica OsSGR allele exhibit the balanced staygreen with a delay of senescence, but unlike the stay-green knock out mutants that exhibit unbalanced and non-functional stay-green phenotype". This statement is not convincing to me because based on the new data presented in Supplementary Figure 6 and Supplementary Figure 7, knockout mutants ossgr generated via CRISPR/Cas9 genome editing did show significant decrease in expression of the senescence marker gene OsNAP, implying that OsSGR is a regulator of senescence and loss-of-function in OsSGR did case delayed leaf senescence and theoretically increased photosynthetic capability and yield in this study. I suggest the authors to address this question seriously because sgr being a nonfunctional staygreen locus has been reported in several articles in multiple plant species.

Reviewers' comments:
Reviewer #2 (Remarks to the Author): In this report, Shin D and his colleague indicated that natural variation in OsSGR promoter region might determine the lifespan variations between two species and the modify in this region could enhance grain yield. In previous review, I pointed out that authors should analyze photosynthetic ability of flag leaves or a canopy through growth stage and add the data of growth analysis (NAR) to clarify factors of yield characteristics. Because it is most important and valuable point of this report to clarify which the lower decrease in chr. content by their allele can maintain the photosynthetic ability and then increase in yield or not. However, in present version, they added the Fv/Fm (chlorophyll fluorescence) data in Figure 4k and Supplementary Fig. 17b. Their data could not indicate higher photosynthetic ability. For example, between IR72 and NIL the larger difference in chl. content did not reflect photosynthetic ability after 7 weeks after heading (Fig. 4k) -Response: We thank Reviewer #2 for these critical suggestions. In the revised manuscript, we added the requested data. Following the reviewer's comment, we measured the net CO 2 assimilation rate of flag leaves to evaluate the effect of OsSGR on photosynthetic ability in IR72-NIL and JN-NIL grown in the field (Fig. 4k, Supplementary Fig. 17h). The net CO 2 assimilation rate decreased continuously after flowering in the flag leaves (Fig. 4k). However, the extent of the decrease in the net CO 2 assimilation rate in IR72-NIL was significantly less than that of IR72 (Fig. 4k). At 7 weeks after heading, the net CO 2 assimilation rate and Fv/Fm ratio of IR72-NIL were still higher than those of IR72, which is consistent with the higher levels of chlorophyll in the flag leaves (Fig. 4g, i, k and Supplementary Fig. 17b). IR72-NIL of the indica background harboring the japonica OsSGR allele exhibited extended photosynthesis with higher chlorophyll contents (Fig. 4g, i, and k, and Supplementary Fig. 17b, c), resulting in enhanced grain filling rate and therefore grain yield (Fig. 4l, m and Supplementary Table. 4b). Therefore, indica NILs harboring japonica OsSGR allele displayed extended photosynthetic abilities and therefore increased productivity, similar to functional stay-green genotype. However, OsSGR is a gene involved in chlorophyll degradation. The japonica alleles of OsSGR are progenitors of the indica alleles. The indica alleles lose chlorophyll earlier than japonica alleles due to natural variations in the promoter that activate the expression OsSGR earlier during the senescence period in our field condition in a temperate zone. Thus, the indica allele exhibits earlier senescence and accelerates the loss of photosynthetic capacity. In contrast, the japonica allele leads to `2 typical slower senescence pattern observed for the japonica cultivars grown in temperate zones. Thus, when japonica alleles are introduced into indica cultivars, the NILs undergo the japonica type change of photosynthetic capacity during senescence period, which is slower than the parental indica cultivars.
Conversely, JN-NIL of japonica harboring the indica OsSGR allele displayed shortened photosynthetic activity with decreased chlorophyll levels, leading to the decreased grain filling rate and yield, as compared with parental japonica (Fig. 4l, m, Supplementary Table 4b and Supplementary Fig. 17e, h, i).
Along with analyzing the photosynthetic ability of NILs grown in the field, we also measured the relative growth rate (RGR) of NILs, as requested. The relative growth rates (RGR) quantifies plant growth speed and is considered to be a reliable standard for estimating plant productivity 1, 2 . It is calculated as the dry mass increment per aboveground biomass at a given time point. RGRs of IR72-NIL are shown in Supplementary Fig. 17d. The RGR of all the rice plants we examined declined continuously during the grain filling stage. However, IR72-NIL harboring japonica OsSGR allele maintained higher RGR, especially between 5 -7 weeks after flowering, than the parental IR72 ( Supplementary Fig. 17d), indicating the contribution of the japonica OsSGR allele in higher biomass productivity. On the contrary, JN-NIL harboring indica OsSGR allele showed a relatively faster decline in RGR than its parental cultivar, Junam ( Supplementary Fig. 17j).
To further test the yield potential of NILs, large-scale field trials of NILs were performed in the paddy field in 2019. The figure below shows NILs grown in the field at late grain filling stage (Response Figure Fig.1 a-c). When indica-NILs harboring japonica OsSGR allele and their parental indica varieties were cultivated in the field, parental varieties were lodged by early senescence, whereas indica-NILs remained standing in the field due to the delayed senescence induced by japonica OsSGR allele (Response Figure 1a-c). Lodging-resistance is important for rice breeding because this is a favorable trait for improving crop productivity 3 . Furthermore, we experienced typhoons three times during the grain-filling period in this year. JN-NIL harboring indica OsSGR allele exhibited earlier senescence than its parental japonica (Response Figure 1d). After harvest, we analyzed the total grain yields of NILs (kg/10a). Compared to their parental cultivars, the grain yields of the IR72-NIL, Milyang21-NIL, and Milyang23-NIL were 850, 794, and 896 kg/10a, which corresponds to an increase of 5.1, 9.9, and 4.8 %, respectively in our experimental field condition (Response Figure 2). Previously, we showed that the grain yields per Minor point 1. Authors should add the statistical analysis in supple. Fig. 3 and Table 4.
-Response: We performed the statistical analysis and revised the text in supplementary Figure 3 and Table 4. the personal style has been used.
-Response: We revised the text throughout the manuscript according to the reviewer's suggestion. -Response: We thank for this encouraging comments.
My second concern, about the discrepancy between OsSGR being a functional senescence regulator in this study and the results from earlier studies indicating that sgr is a nonfunctional stay-green locus (Park et al., 2007, Plant Cell;Sato et al., 2007, PNAS), is however not addressed. In Line 252-258, the authors stated that "The japonica OsSGR allele does not lead to an over stay-green but shows difference in its induction kinetics and induction level compared to the indica allele (Fig 3a). The japonica allele appears to exhibit a capacity to balance the degree of stay-green trait, photosynthetic competence, nutrient remobilization, when introduced into indica varieties, to lead to increased grain filling and productivity during the senescence period. Thus, NILs-OsSGR in the indica background harboring a japonica OsSGR allele exhibit the balanced stay-green with a delay of senescence, but unlike the stay-green knock out mutants that exhibit unbalanced and non-functional stay-green phenotype". This statement is not convincing to me because based on the new data presented in Supplementary Figure 6 and Supplementary Figure 7, knockout mutants ossgr generated via CRISPR/Cas9 genome editing did show significant decrease in expression of the senescence marker gene OsNAP, implying that OsSGR is a regulator of senescence and loss-of-function in OsSGR did case delayed leaf senescence and theoretically increased photosynthetic capability and yield in this study. I suggest the authors to address this question seriously because sgr being a nonfunctional staygreen locus has been reported in several articles in multiple plant species.
-Response: We understand the reviewer's concern. `6 OsSGR is a gene involved in chlorophyll degradation. The japonica alleles of OsSGR are progenitors of the indica alleles. The indica alleles lose chlorophyll earlier than japonica alleles due to natural variations in the promoter that activate the expression OsSGR earlier during the senescence period in our field condition in a temperate zone. Thus, the indica allele exhibits earlier senescence and accelerates the loss of photosynthetic capacity. In contrast, the japonica allele leads to typical slower senescence pattern observed for the japonica cultivars grown in temperate zones. Thus, when japonica alleles are introduced into indica cultivars, the NILs undergo the japonica type change of photosynthetic capacity during senescence period, which is slower than the parental indica cultivars.
To analyze the biochemical activity of ossgr allele, we conducted chlorophyll degradation activity assays in vitro and in planta to analyze the biochemical activity of the ossgr allele ( Supplementary Fig. 11a, b). The results indicated that the ossgr allele did not have any chlorophyll breakdown activity, as compared to japonica and indica alleles. Therefore, ossgr mutants maintained high chlorophyll levels in senescent leaves, but mutant leaves did not maintain their photosynthetic activities longer than the WT leaves 4, 5 , leading to no yield advantage ( Supplementary Fig. 6i-m).
Knockout mutants ossgr in indica generated via CRISPR/Cas9 genome editing have no chlorophyll breakdown activity, just like ossgr mutants in japonica. Although chlorophyll breakdown in ossgr mutants is slower than that in WT leaves, previous reports have shown that ossgr mutants did not maintain photosynthesis competence, leading to a non-functional stay-green phenotype 13,24 . OsNAP, a senescence marker gene, is increased in both WT and ossgr mutants, indicating that ossgr mutant plants undergo senescence despite of their stay-green phenotype. Thus, ossgr is a non-functional staygreen gene. However, the expression of OsNAP in ossgr mutants was not induced to WT levels. We suggest that the OsSGR gene itself or the higher level of chlorophyll in ossgr mutants may affect the expression of OsNAP. On the contrary, indica-NILs containing japonica OsSGR alleles maintain chlorophyll breakdown activity, which is lower than their parental indica cultivars, due to decreased expression of OsSGR during senescence. Thus, these NILs underwent slower chlorophyll loss and displayed prolonged photosynthetic abilities with slow senescence phenotypes similar to japonica cultivars, leading to a yield advantage. We revised the text in the 'Discussion' (page11~12, line 268-

285).
In this report, Shin D and his colleague indicated two important point, that natural variation in OsSGR promoter region might determine the lifespan variations between indica and japonica rice and the modify in this region could delay the senescence and enhance grain yield. In previous review, I pointed out that authors should analyze the photosynthetic ability of flag leaves or a canopy through growth stage and add the data of growth analysis (NAR) to clarify factors of yield characteristics. In new version, authors added the new data about net photosynthetic ratio (Fig.  4k) and RGR data (supplemental Fig.17). I am disappointed that this report is not suitable for Nature Communication, the top scientific journal because another group published the similar results about the delayed leaf senescence caused higher yield (Ramkumar et  Major point 1. Ramkumar MK and his colleague reported with a novel stay-green mutant the delay of leaf senescence could improve the photosynthetic ability and yield. A stay-green locus they identified (SGM-3; on chr.9) was overlapped with QTL mapped by Shin et al.. And Ramkumar's group analyzed function the photosynthetic ability, chlorophyll contents and yield under normal or drought condition. Additionally, they indicated that OsSGR was a strong candidate gene of this locus and its different expression determined the allele. 2. In new version, authors added the data of net photosynthetic ratio between IR72 and NIL ( Fig.4k) but the ratio of both plants was too low, for example, that of NIL was 8 or less and 5 or less in control at 5 weeks after heading. Generally, in same phase the photosynthetic ratio of rice flag leaf is more than 15 (μmol CO2/m2/s) in many reports (e.g., Jiang et al., Jpn. J. Crop Sci: 139-145, 1988). Authors should check the data. 3. Another important point of this paper is that the promoter polymorphisms in OsSGR affected the period of senescence. Authors should analyze the motif in the promoter region. 4. Authors indicated that variations in OsSGR controlled life span. However, heading dates were same among two japonica and indica rice plants (Supplementary Fig.1). These results showed these variations did not affect the life span but senescence period after heading.
Reviewer #4 (Remarks to the Author): The authors stated in the text "OsNAP, a senescence marker gene, is increased in both WT and ossgr mutants, indicating that ossgr mutant plants undergo senescence despite of their stay-green phenotype. Thus, ossgr is a non-functional staygreen gene. However, the expression of OsNAP in ossgr mutants was not induced to WT levels" (Line 277-280).
In Supplementary Figure 6 d, the OSNAP levels, although increased during senescence, are significantly lower than that of WT, which means that senescence is significantly delayed. This does not support the conclusion "ossgr is a non-functional staygreen gene". This above mentioned statement does not explain why the ossgr mutants showed "no yield advantage ( Supplementary  Fig. 6i-m) ", and why the NILs (for which the authors did not provide OsNAP expression data) "underwent slower chlorophyll loss and displayed prolonged photosynthetic abilities with slow senescence phenotypes similar to japonica cultivars, leading to a yield advantage".
Since in the case of the OsSGR gene, chlorophyll content can no longer represent the progress of senescence, the authors should provide OsNAP expression and Fv/Fm data for all the critical lines to claim "senescence" phenotypes. A logical explanation of why senescence in both ossgr and NILs plants are delayed but only the NILs showed a yield advantage is lacking.
In this report, Shin D and his colleague indicated two important point, that natural variation in OsSGR

Responses.
We appreciate the reviewer for bring up this reference paper. The paper mentioned by the reviewer is a report on three stay-green mutants 1 . The authors reported that one of the mutants, SGM-3, can improve harvest index and drought tolerance. Reviewer pointed out that the locus of SGM-3 mutant generated by EMS mutagenesis was overlapped with QTL (chr.9) mapped in our study and therefore OsSGR is a strong candidate gene of SGM-3 mutant. We carefully examined the article. We understand the paper is somewhat sophisticated but there is a discrepancy between the argument by the reviewer and the description in the paper.
The reviewer mentioned that "A stay-green locus they identified (SGM-3; on chr.9) was overlapped with QTL mapped by Shin et al. and Ramkumar's group analyzed function the photosynthetic ability, chlorophyll contents and yield under normal or drought condition. Additionally, they indicated that OsSGR was a strong candidate gene of this locus and its different expression determined the allele.
However, the paper mentioned that, at the end of the Abstract, "Analysis of the earlier reported Quantitative Trait Loci (QTL) regions in SGM-3 revealed negligible variations from WT, suggesting it to be a novel SG mutant'. The paper also mentioned that in the Discussion part "suggesting that SGM-3 could be a novel and functional SG mutant (Page 12, line 43)".
Thus, the authors of the paper have emphasized that the SGM-3 mutation is a novel one. The novelty of SGM-3 mutants in the paper was argued based on the following data.
There were no mutations on OsSGR gene in SGM-3, which was described in the article (page 11, line 8-10). The results suggested that SGM-3 is not caused by the OsSGR mutation.
They analyzed the QTL region on chromosome 9 of the functional SG mutant, SNU-SG1 6 , but results showed no significant changes from WT. (Page 12, line 40-42). Therefore, they suggested that SGM-3 could be the novel functional SG mutant and will map the causal gene for future works.
The authors mentioned that "Thus, SGM-1 and SGM-2 were inferred as cosmetic SG mutants while SGM-3 as a functional SG mutant. (Page 11, line 28-29). As ossgr mutant is a non-functional, SGM-3 is different from OsSGR.
In addition, the expression of OsSGR in SGM-3 was higher than that of WT ( Figure 3; Figure 7), although SGM-3 showed the delayed senescence ( Figure 1). The authors mentioned that "SGM-3 also had the highest upregulation for NOL, SGR and PAO" (Page 10, line [7][8]. Higher expression of OsSGR should lead to earlier loss of greenness but this was not the case in SGM-3, indicating the difference between SGM-3 and ossgr mutants. Yet, we appreciate the value of the reference and added the following sentence in the discussion part of the revised manuscript (Page 13, line 313-316).
In agreement with our observation, a recent report showed that an extended photosynthetic competence during senescence stage leads to increased harvest index in SGM-3 mutant of upland rice variety Nagina 22; SGM-3 was suggested to be a "novel and functional stay green mutant" 31 .
2. In new version, authors added the data of net photosynthetic ratio between IR72 and NIL ( Fig.4k) but 3 the ratio of both plants was too low, for example, that of NIL was 8 or less and 5 or less in control at 5 weeks after heading. Generally, in same phase the photosynthetic ratio of rice flag leaf is more than 15 (μmol CO2/m2/s) in many reports (e.g., Jiang et al., Jpn. J. Crop Sci: 139-145, 1988). Authors should check the data.

Response.
We thank the reviewer for raising this concern. According to the reviewer's suggestion, we checked the article by Jiang et al. (1988) 7 . They used different unit from ours as below.
They used the unit of mgCO 2 /dm 2 /hr. We used the unit of μmolCO 2 m -2 S -1 as below. Thus, a direct numerical comparison may not be appropriate. Besides, the text in the paper was in Japanese, which made it difficult for us to understand the details of the instrument and methods such as the exact timing (e.g., days after heading) of the three time points during ripening stage.
We understand that the exact numerical values of photosynthetic rates may be variable depending on rice varieties and growth conditions. Yet, the comparison of our data with the previous data as mentioned above shows that our data is in a comparable range. We did not include this response in the revised manuscript, as our values were not peculiar.
3. Another important point of this paper is that the promoter polymorphisms in OsSGR affected the period of senescence. Authors should analyze the motif in the promoter region.

Response.
We understand the reviewer's point. We added the following paragraph in the Discussion part of the revised manuscript (Page 11, line 256-267) 11 .
When we analyzed the binding motifs for transcriptional factors in the 2-kb OsSGR promoter region using NEW PLACE (A Database of Plant Cis-acting Regulatory DNA element; https://www.dna.affrc.go.jp/PLACE/?action=newplace), we found that the japonica and indica promoters contained same numbers (thirteen) of WRKY biding motifs and no NAC binding motif. However, there is a distinctive difference in the Dof binding motifs between the two promoters. There are fourteen Dof 5 protein binding motif (consensus AAAG sequences or its reversibly complementary sequence, CTTT 30 ) in the japonica promoter. On the other hand, in the indica promoter, there is a new Dof protein binding motif formed by insertion of AAAAGCTC (position -1, 377; Supplementary Fig. 9). This region with the new Dof binding motif is tightly associated with low levels of chlorophyll contents. We anticipate that this new Dof binding motif and the respective Dof transcription factor may together lead to the indica type phenotype of chlorophyll loss. Elucidating the molecular mechanisms regulating the early and higher induction of OsSGR through evolution of the promoter region should be a key future effort.

Authors indicated that variations in
OsSGR controlled life span. However, heading dates were same among two japonica and indica rice plants ( Supplementary Fig.1). These results showed these variations did not affect the life span but senescence period after heading.

Response.
We appreciate the reviewer's point. We added the following paragraph in Discussion part in response to the reviewer's point. We like to note that heading dates of cultivars do not necessarily reflect their respective lifespans, as cultivars show differences in senescence process after heading with the concomitant differences in their lifespans (Page 13/14, line 317-326) 12 .
Lifespan indicates the maximal life expectancy from the seed to seed and the length of time that plants live or expected to live 32 . In this regard, the lifespan of rice plants is the duration time from seed germination to the panicle senescence and death associated with grain maturation. In analyzing senescence processes of rice plants in our experiments, we chose two japonica and indica rice cultivars, which showed the same heading date ( Supplementary Fig. 1) to avoid an influence of the differences in reproductive timing on senescence processes and the related lifespan. Despite of the same heading dates, the panicles of the two indica cultivars showed earlier senescence than those of the two japonica cultivars (Fig. 1h), as quantified by colorimetric assays (Fig. 1i). Thus, the lifespans of these rice cultivars were largely related to senescence process of the panicles after heading, which is controlled by the differential expression levels and kinetics of OsSGR (Fig. 4c, f, h, j).

2) Responses to Reviewer #4's comments.
Reviewer #4's comments. 6 The authors stated in the text "OsNAP, a senescence marker gene, is increased in both WT and ossgr mutants, indicating that ossgr mutant plants undergo senescence despite of their stay-green phenotype.
Thus, ossgr is a non-functional staygreen gene. However, the expression of OsNAP in ossgr mutants was not induced to WT levels" (Line 277-280).
In Supplementary Figure 6d, the OSNAP levels, although increased during senescence, are significantly lower than that of WT, which means that senescence is significantly delayed. This does not support the conclusion "ossgr is a non-functional staygreen gene". This above mentioned statement does not explain why the ossgr mutants showed "no yield advantage ( Supplementary Fig. 6i-m)", and why the NILs (for which the authors did not provide OsNAP expression data) "underwent slower chlorophyll loss and displayed prolonged photosynthetic abilities with slow senescence phenotypes similar to japonica cultivars, leading to a yield advantage".
Since in the case of the OsSGR gene, chlorophyll content can no longer represent the progress of senescence, the authors should provide OsNAP expression and Fv/Fm data for all the critical lines to claim "senescence" phenotypes. A logical explanation of why senescence in both ossgr and NILs plants are delayed but only the NILs showed a yield advantage is lacking.

Response:
We fully understand the reviewer's concern. We tried to make the confusion clearer in a few places in Result and Discussion.
The first point of the reviewer' concerns is regarding the expression of OsNAP in the OsSGR mutant plants. The nature of OsSGR as a non-functional stay-green gene has been well described before 2 . It encodes the enzyme Mg ++ -dechelatase in chlorophyll degradation pathway and is not a regulatory gene 4 .
However, it was known that the knock-out mutation in this enzyme leads to some degree of delayed senescence upon a prolonged growth as shown below in c 2 . In this measurement, the Fv/Fm value was higher in 30 days after heading in the mutant than in wild type. 7 The data on the above figure is consistent with our observation that expression of OsNAP in ossgr mutant plants is not induced to the wild type level during senescence ( Supplementary Figures 6d and 7c), as pointed out by the reviewer as below.
"In Supplementary Figure 6 d, the OSNAP levels, although increased during senescence, are significantly lower than that of WT, which means that senescence is significantly delayed".
We do not know the mechanism underlying the partial delay of some senescence symptoms such as Fv/Fm and expression of OsNAP in the ossgr mutant after a prolonged growth. Perhaps, high level of the remaining chlorophyll may have some effect on the partial delay of senescence. However, in the above data, it is also noted that the net photosynthesis (Pn) is not maintained even after 30 days after heading, showing the characteristics of non-functional stay-green mutants 2 .
We added the following paragraph to respond to the reviewer's point (Page 12, line 276-292). In addition, we made some minor changes in Result section (in red in Page 6) to respond to the reviewer's concern in the expression of OsNAP 2, 3, 4, 5 .
The nature of OsSGR as a non-functional stay-green gene has been well characterized previously 13,15 . OsSGR encodes the enzyme Mg ++ -dechelatase in the chlorophyll degradation pathway and is not a regulatory gene 19 . We conducted chlorophyll degradation activity assays in vitro and in planta to analyze the biochemical activity of the proteins derived from various ossgr alleles ( Supplementary Fig.   11a, b). The results showed that the protein from the ossgr knockout mutant allele showed a negligible 8 enzyme activity toward chlorophyll degradation. Accordingly, ossgr mutant plants maintain a high level of chlorophyll content in senescent leaves but without comparable maintenance of net photosynthesis 15 , which is a characteristic feature of non-functional stay-green mutants. However, the ossgr knockout mutation leads to delay in some aspects of functional senescence upon a prolonged growth; Fv/Fm value was higher in 30 days after heading in the mutant plants than in wild type plants 15 15 We also added the paragraph below to respond another point of the reviewer which is "A logical explanation of why senescence in both ossgr and NILs plants are delayed but only the NILs showed a yield advantage is lacking" (Page 12/13, line 293-316) 1 .
Here, we showed that rice grain yield can be increased by replacing the indica allele of OsSGR with the japonica allele. As OsSGR is a non-functional stay-green gene in terms of net photosynthesis, it seems paradoxical to increase rice yield via OsSGR. It was the difference in the promoter region of the indica and japonica alleles of the OsSGR gene that led to yield increase in the NILs we generated. Unlike the ossgr knockout mutation with no enzyme activity of OsSGR, the proteins from japonica and indica alleles showed comparable enzyme activity (Supplementary Figure 11). However, the promoter regions of the two alleles of OsSGR are diverged (Supplementary Figure 9). Compared to the promoter of the japonica allele, the promoter of the indica allele led to earlier and higher induction of OsSGR (Fig. 3a) and to earlier loss of chlorophyll ( Fig. 1) with concomitant reduction of photosynthesis (Fig. 4k) and grain yield (Fig. 4l). The senescence response in rice plants with the indica allele is similar to that in the activation tagging lines (Supplementary Figure 5) with the japonica alleles, where OsSGR expression is increased at a later stage with lower level of chlorophyll and reduced grain yield (Supplementary Figure   5). In generating the NIL lines with increased yield, we replaced the indica allele with the japonica allele, so that the induction of OsSGR is slower than the parental lines with the indica allele (Fig. 4j). This leads to slower loss of chlorophyll (Fig. 4g. i), higher photosynthesis (Fig. 4k), and increased yield (Fig. 4l) compared to those in their parental lines. Thus, the senescence response of rice plants with the japonica allele with slower induction of OsSGR became comparable to a functional stay-green phenotype, unlike the ossgr knock-out mutant which showed non-functional stay-green phenotype with no yield advantage.
As the heading date of these lines are same, our results show that a senescence period with extended photosynthetic competence can lead to higher productivity. In agreement with this notion, a recent report showed that an extended photosynthetic competence during senescence stage leads to increased harvest index in SGM-3 mutant of upland rice variety Nagina 22; SGM-3 was suggested to be a "novel and functional stay green mutant" 31 .
We also like to note that we measured the expression of OsNAP and Fv/Fm values of IR72, IR72-NIL, JN and JN-NIL during senescence stage to follow the senescence state in these lines, which were already included in the manuscript (Supplementary Figure 17).