Georgia C. Eizenga1*†
Yami Santamaria2
Aaron K. Jackson1†Huy Phan2Melissa H. Jia1†Quynh P. - H. Grunden1
Jeremy D. Edwards1†Ed Himelblau2†
Michael R. Schläppi2*†- 1Dale Bumpers National Rice Research Center, US Department of Agriculture, Agricultural Research Service (USDA-ARS), Stuttgart, AR, United States
- 2Department of Biological Sciences, Marquette University, Milwaukee, WI, United States
Introduction: Exploring natural genetic variation to facilitate breeding of improved rice seedling cold tolerance will allow the crop to be planted earlier in the growing season, taking advantage of spring rainfall and decreasing exposure to high summer nighttime temperatures, which reduce grain quality.
Methods: To uncover genomic regions in rice that manage cold stress tolerance response mechanisms in the cold-sensitive aus (AUS) and the relatively cold-tolerant tropical japonica (TRJ) subpopulations, and to identify cold tolerance genes, AUS and TRJ recombinant inbred line populations developed from crosses between cold-tolerant and cold-sensitive parents were used for quantitative trait locus (QTL) mapping of two traits: degree of membrane damage after 1 week of cold exposure, quantified as percent electrolyte leakage (EL), and percent low-temperature seedling survivability (LTSS) after 1 week of recovery growth.
Results and discussion: Thirteen subpopulation-specfic QTL were revealed: three EL and four LTSS QTL for AUS, and two EL and four LTSS QTL for TRJ, with no overlap between AUS and TRJ QTL. Only two AUS QTL overlapped with regions previously identified by our AUS × temperate japonica biparental mapping, further confirming the discovery of subpopulation-specific QTL. Based on high-impact genomic differences between the cold-tolerant and cold-sensitive parents, 35 cold tolerance candidate genes were identified—23 in AUS and 12 in TRJ—of which about 50% encode proteins involved in signal transduction and protein homeostasis processes. Although most QTL showed that alleles from cold-tolerant parents improved the two cold tolerance traits, alleles from cold-sensitive parents enhanced these traits at several other QTL. Therefore, alleles from both cold-tolerant and cold-sensitive parents can be used in breeding efforts to generate AUS and TRJ lines with better cold tolerance potential than their respective cold-tolerant parents.
Introduction
Cultivated Asian rice (Oryza sativa L.) is grown worldwide, in diverse environments and on every continent except Antarctica. Rice has two geographically, genetically, and phenotypically diverged varietal groups (VG), Indica and Japonica. The Indica VG is composed of indica and aus (AUS) cultivars, and the Japonica VG is composed of tropical (TRJ) and temperate (TEJ) japonica cultivars. It is generally accepted that TEJ rice cultivars, which predominate in temperate climates including northern China, Korea, and Japan, are more tolerant to cold and chilling temperatures than Indica cultivars, which grow in tropical and subtropical regions like southern China and South Asia (Li et al., 2022; Lu et al., 2014; Lv et al., 2015; Mackill and Lei, 1997; Shakiba et al., 2017).
Improving cold tolerance in rice at the seedling stage will increase the crop’s resilience to extreme temperatures, as lower temperatures often lead to poor germination and sterility, which reduce yield (Li et al., 2022). Enhanced tolerance in temperate regions, such as the US Mid-South, will improve seedling vigor, allowing rice to be planted a few weeks earlier without the risk of poor stand establishment and reducing production costs by taking advantage of spring rains. Planting rice earlier in the growing season allows the grain to mature sooner, reducing exposure to high nighttime temperatures during the hottest summer months. Such temperatures (around 30°C) increase grain fissuring (grain cracking), resulting in lower whole-kernel yield—commonly referred to as head rice yield—and reduced grain quality (Gann et al., 2021; Nagata et al., 2013).
Cold tolerance in rice is a multigenic and complex quantitative trait because, depending on temperature, humidity, and photoperiod, different cold stress tolerance response mechanisms are employed by the rice genome through various modules of coexpressed genes (reviewed by Li et al., 2022). As a result, individual genes generally have small effects and act in concert with other genes to protect rice plants from detrimental low-temperature effects. Since 2001, more than 30 studies using traditional biparental linkage mapping projects have identified over 80 germination/seedling stage and approximately 75 reproductive stage quantitative trait loci (QTL), including work from our labs (Schläppi et al., 2023). In addition, over 11 studies since 2015, including four from our labs (Shakiba et al., 2017; Schläppi et al., 2017; Shimoyama et al., 2020; Phan and Schläppi, 2021), using genome-wide association studies (GWAS), have identified at least 435 germination/seedling stage and 65 reproductive stage QTL (reviewed in Li et al., 2022). For breeding purposes, combining different QTL mapping approaches is beneficial to select chromosomal regions consistently uncovered and to identify candidate genes associated with cold tolerance.
To dissect cold tolerance in rice at the germination and seedling stages, Schläppi et al. (2017) developed five assays. One assay evaluated the germination stage, one assay evaluated both the germination and seedling stages, and three assays evaluated the seedling stage. Based on these assays, five cold tolerance indices were calculated and assessed using 202 O. sativa accessions from the Rice Mini-Core (RMC) collection. The RMC includes accessions representing the five major rice subpopulations, AUS, indica, TRJ, TEJ, and aromatic (Sweeney and McCouch, 2007). For young seedling stage indices, the Japonica VG and Indica VG clustered into cold-tolerant and cold-sensitive accessions, respectively. From this study, the percent low-temperature seedling survivability (LTSS) assay was determined to be the most reliable measure of seedling cold tolerance.
Further investigation of the rice cold tolerance at the seedling stage used a subset of 354 O. sativa accessions from the Rice Diversity Panel 1 (RDP1), including representatives of the five rice subpopulations as well as accessions that were admixtures of two or more subpopulations (Shimoyama et al., 2020). These accessions were evaluated for low-temperature tolerance at the seedling stage using three indices: LTSS, percent electrolyte leakage (EL) as a measure of plasma membrane integrity after cold exposure, and median lethal cold temperature (LT50), the temperature at which 50% of 2-week-old seedlings die. Overall, these indices confirmed that accessions in the Japonica VG, including both TEJ and TRJ, were cold tolerant; accessions in the Indica VG were cold sensitive; and aromatic accessions exhibited intermediate cold tolerance.
Mapping cold tolerance traits in both the RMC (Schläppi et al., 2017) and the RDP1 (Shimoyama et al., 2020) using GWAS identified 37 QTL in the RMC and 245 QTL in the RDP1 associated with tolerance. To refine these results and identify smaller genomic regions overlapping with the larger genomic regions detected by GWAS, Schläppi et al. (2023) developed two mapping populations from crosses between two cold-tolerant TEJ accessions and one cold-sensitive AUS accession. Mapping in these populations revealed 16 QTL that overlapped with at least one of the QTL identified by GWAS. The genomic regions delineated by these QTL contained at least 25 candidate genes with polymorphisms between the cold-tolerant TEJ parents and the cold-sensitive AUS parent.
The main objective of this study was to unravel the cold tolerance mechanisms within the AUS and TRJ rice subpopulations and identify subpopulation-specific mechanisms that could be used to improve low-temperature tolerance. To achieve this, we (1) used the RMC LTSS data (Schläppi et al., 2017) to select AUS and TRJ accessions that were either cold tolerant or cold sensitive for developing AUS and TRJ subpopulation-specific biparental recombinant inbred line (RIL) mapping populations; (2) screened the RILs for chilling tolerance and conducted QTL mapping within the populations; (3) identified overlaps with QTL from GWAS mapping; and (4) discovered candidate genes underlying subpopulation-specific chilling tolerance.
Materials and methods
Development of the AUS and TRJ subpopulation-specific RIL populations
To develop subpopulation-specific AUS and TRJ biparental RIL populations, the LTSS indices of the O. sativa accessions categorized as AUS and TRJ in the RMC (Schläppi et al., 2017) were examined to identify the two most tolerant and two most susceptible accessions in each subpopulation. Seeds of the eight selected RMC accessions were obtained from the Genetic Stocks-Oryza (GSOR) collection (https://www.ars.usda.gov/GSOR) at the USDA-ARS Dale Bumpers National Rice Research Center (Stuttgart, AR, USA). Crosses were made between the susceptible and tolerant accessions within each subpopulation, and two to eight F1 plants from the successful crosses were grown to produce F2 progeny seeds. Leaf tissue was collected from the F1 hybrids in spring 2018 to validate their parents using 10 simple sequence repeats (SSR) markers described in Ricebase (https://ricebase.org; Edwards et al., 2016), selected from the original 128 SSR markers, three InDel markers, and six SNP markers used to genotype the RMC (Li et al., 2010). Based on the confirmed parents of the F1 hybrids, the number of days to 50% heading of the parents, the seed of the parents not having red pericarp, the number of F2 seeds, plant morphology, and the parents having resequencing data (Wang et al., 2016), the NC 1/536 (GSOR 310381) × Phudugey (GSOR 310715) cross was selected for developing the AUS subpopulation-specific RIL population, and the Taichu Mochi 59 (GSOR 310080) × British Honduras Creole (GSOR 310144) cross for developing the TRJ subpopulation-specific RIL populations.
For the AUS population, NC 1/536, the cold-susceptible female parent, originated from Pakistan and had a mean LTSS at 10°C of 1.54% (Schläppi et al., 2017). Phudugey, the cold-tolerant male parent, originated from Bhutan and had a mean LTSS at 10°C of 37.42% (Schläppi et al., 2017). Four (RM234, RM228, RM224, and RM154) of the 10 SSR markers were used to confirm these F1 hybrids, and the genetic distance between the AUS parents is shown in Supplementary Figure S1. To develop the NC 1/536 × Phudugey RIL population, about 220 single F2 seeds were planted in the greenhouse near Stuttgart, AR, USA, to produce F3 seeds. Using the single-seed descent method, all F2 plants that produced F3 seeds were advanced to the F4, F5, and F6 generations. For genotyping, leaf tissue was collected from 168 individual F6 plants grown in the greenhouse (for three RILs missing tissue, tissue from F7 field plants was used for genotyping). The F7 seeds harvested from these plants were planted in the field near Stuttgart, AR, USA, on 4 May 2022 to produce adequate seed for phenotyping the population for cold tolerance at the seedling stage (11 RILs did not produce an F7 seed; thus, an F6 seed was planted). Data on agronomic traits were collected from 153 F6:7 plants (described below), and F7:8 seed was collected from a single F6:7 plant of each of the 152 RILs for conducting the cold tolerance phenotyping (Supplementary Table S1) (the aus_197 RIL did not produce an F8 seed for the cold tolerance phenotyping).
For the TRJ population, Taichu Mochi 59, the cold-susceptible female parent, originated from Taiwan and had a mean LTSS at 10°C of 39.76% (Schläppi et al., 2017). British Honduras Creole, the cold-tolerant male parent, originated from Belize and had a mean LTSS at 10°C of 92.14% (Schläppi et al., 2017). Eight (RM234, RM190, RM215, RM1339, RM224, RM154, RM232, and RM5) of the 10 SSR markers were used to confirm the F1 hybrids, and the genetic distance between these TRJ parents is shown in Supplementary Figure S1. To develop the Taichu Mochi 59 × British Honduras Creole TRJ RIL population, 200 F2 seeds were planted in the greenhouse to produce F3 seeds. Using the single-seed descent method, all F2 plants that produced F3 seeds were advanced to the F4, F5, and F6 generations. For genotyping, leaf tissue was collected from 144 individual F6:7 plants grown in the greenhouse, and the seeds harvested from 140 TRJ RILs were planted in the field to produce adequate seeds for phenotyping the population for cold tolerance at the seedling stage (Supplementary Table S2; two RILs did not produce seeds, so F5 seeds were planted in the field). Data on agronomic traits were collected from 122 F5:7 plants (described below). Note that the trj_35 RIL seed did not germinate in the field, so it was grown in the greenhouse; thus, no agronomic data were recorded. For cold tolerance phenotyping, F6:8 seeds were collected from a single plant of 123 TRJ RILs, and 90 of these RILs were used for the cold phenotyping.
Phenotyping of the AUS and TRJ RIL populations
Days to heading, plant height, and culm habit
The AUS and TRJ populations were grown under field conditions to phenotypically characterize the RILs for important agronomic traits, including days to 50% heading (HD), plant height (PTHT), and culm habit (CULMHAB), and to determine which candidate genes were potentially present in the populations. Field conditions also allowed the production of larger amounts of seed to phenotype the populations for seedling cold tolerance traits. Thus, seeds of 168 AUS RILs and 140 TRJ RILs were planted in the field near Stuttgart, AR, USA, on 4 May 2022. For each RIL, eight F6:7 seeds were planted in a row, with 30.5 cm between plants within the row and 198 cm between rows. The number of seeds that germinated per row ranged from zero to eight plants, with an average of 3.3 plants per AUS RIL row and 3.2 plants per TRJ RIL row. The RILs were planted, one replication each, in two bays, with the four parents serving as the outer rows per tier: NC 1/536 and Phudugey for the AUS RILs, and Taichu Mochi 59 and British Honduras Creole for the TRJ RILs. Emergence occurred on 19 May 2022, and the bays were fertilized and flooded on 14 June. Seeds were harvested in mid-September, between 12 and 20 September 2022.
The RILs and parents were characterized for three agronomic traits. HD was determined as the number of days from planting until about 50% of the panicles in a row reached anthesis. For PTHT and CULMHAB, each row of an individual RIL was observed, and based on uniformity, one representative plant was selected. Plant height was measured from the soil surface to the tip of the longest panicle at harvest. Culm habit was rated as erect (1), intermediate (3), open (5), spreading (7), or prostrate (9) after observing all plants in the row. Panicles from the phenotyped RIL were harvested, and the resulting seed was used for cold tolerance tests.
Supplementary Figure S2A presents images of the AUS RIL parents, NC 1/536 and Phudugey, and a selected AUS RIL. Most of the AUS RILs were lodged due to being tall plants and having weak tillers; therefore, images were not recorded. Supplementary Figure S2B includes images of the TRJ RIL parents, Taichu Mochi 59 and British Honduras Creole, and selected TRJ RILs.
Germination and standard seedling growth conditions
For cold phenotyping, seeds of the four parents and RILs were germinated in the dark for 2 days at 37°C in deionized water containing 0.1% bleach to prevent bacterial contamination. Germinating seeds were transferred into PCR strips, placed into pipette tip boxes, and grown hydroponically in deionized water for 10 days in a growth chamber under 12-h light (approximately 150 µE photon flux)/12-h dark cycles, with day/night temperatures of 28°C/25°C. On day 10, the water was replaced with one-fourth Murashige–Skoog basal salt liquid medium to provide nutrients. Each line was represented by up to eight plants per box in quadruplicate, for a maximum of 32 plants (four boxes of eight plants) per experiment. The four boxes were randomly arranged within the growth chamber. Each box contained 11 strips of RILs and one strip containing four seedlings of each parent as controls. For the AUS RILs, there were four NC 1/536 seedlings, the cold-sensitive control, and four Phudugey seedlings, the cold-tolerant control. For the TRJ RILs, the cold-tolerant parent, British Honduras Creole, and the relatively cold-sensitive parent, Taichu Mochi 59, were used as controls.
Chilling stress treatment
Four boxes containing 2-week-old seedlings at the two-leaf stage were placed at random positions within a growth chamber maintained at a constant 10°C ± 1°C and incubated for 7 days (12-h light/12-h dark cycles). The seedlings were watered every other day.
Electrolyte leakage
At the end of the 7-day 10°C stress period, the middle section of the second leaf from three individual seedlings per RIL or control per box was collected and cut into three equally sized segments. The pieces were washed in deionized water and transferred into three different screw-cap glass tubes filled with 5 ml of deionized water (conductivity ≤ 2 µS/cm), then shaken at 200 rpm for 60 min at room temperature to release cellular electrolytes from low-temperature-damaged tissues. Initial conductivity of the three replicates per box (a total of 12 replicates across the four randomly distributed boxes) was measured by taking 120 µl of the solution and adding it to the well of a handheld LAQUAtwin B-771 conductivity meter (Horiba Scientific, Kyoto, Japan). Leaf samples were boiled for 20 min after the initial reading to release total cellular electrolytes. Samples were shaken again at 200 rpm for 30 min after cooling to room temperature, and the final conductivity reading was taken. Percent EL for each sample was determined as [(initial conductivity reading)/(final conductivity reading)] × 100.
Low-temperature seedling survivability
At the end of the 7-day 10°C stress period, seedlings were returned to standard growth conditions for 1 week to recover, after which seedling survival was assessed visually. Seedlings that were mostly green and stiff were scored as alive, whereas seedlings that were mostly wilted and/or bleached and soft were scored as dead. The mean percent survivability was calculated as [(number of seedlings scored as alive)/(total number of stressed plants)] × 100.
Statistical analysis
Both the percent EL and percent LTSS were calculated using a linear mixed model (LMM) to obtain best linear unbiased predictions (BLUPs). The LMM used an augmented design with a fixed-effect variable (group) based on whether the line was a control or RIL. The line nested within the group variable, set (growth chamber and experiment date), and the box were assigned as random effects. In addition to percent LTSS (useful for direct comparison to previous studies), a more appropriate statistical method for survival analysis was employed. Specifically, a binomial generalized linear mixed model (GLMM) with a logit link function was used to predict the probability of survival under cold treatment on the log-odds scale. The same fixed and random variables were used for all traits in both the GLMM and LMM models.
All calculations were performed using JMP 18 (JMP 18.2.0, 2022), including the mean and standard error of the mean for each of the six traits across the AUS and TRJ RILs. Frequency distribution images for the AUS and TRJ RILs for each trait were generated in JMP. Correlations among the six traits were calculated using the multivariate option with the restricted maximum likelihood (REML) method, and the corresponding scatterplot matrix was also produced in JMP.
Genotyping and QTL mapping
Leaf tissue was collected from the parents and progeny of both F6:7 RIL populations and lyophilized. Genotyping was performed with the Cornell-IR LD Rice Array (C7 AIR) with 7,098 SNP markers (Morales et al., 2020) available as an Illumina Infinium array. Lyophilized leaf tissues were sent to Eurofins Diagnostics Inc. (www.eurofinsgenomics.eu/en/genotyping-gene-expression/service-platforms/illuminaarrayplatforms/) for DNA extraction and sequencing. Genotypes were filtered to remove markers or lines that were monomorphic, had 30% or greater missing data, had excessive heterozygous calls, or did not appear to map correctly. For each population, all available genotyped lines were used to create a genetic linkage map in MapDisto version 1.7.7 (Lorieux, 2012); lines missing all three traits from cold tolerance traits or agronomic traits were then removed for mapping purposes.
The AUS RIL population had 519 SNPs with 152 RILs used for mapping cold tolerance traits (percent electrolyte leakage, percent low temperature seedling survivability, and low temperature seedling survivability logit) and 153 progeny lines used to map agronomic traits (DH, PTHT, and CULMHAB). The TRJ RIL population had 1,042 SNPs, with 90 RILs used to map cold tolerance traits and 122 RILs used to map agronomic traits. Parents of the respective mapping populations were included in the QTL analysis. QTL mapping was performed in Qgene version 4.4.0 (Joehanes and Nelson, 2008) using single-trait multiple interval mapping, with the scan size set at 0.1 centimorgans (cM) for both populations. For each trait, a permutation analysis was performed in Qgene using the default setting of 1,000 permutations and alpha levels of 0.01, 0.05, and 0.10. For comparison, QTL mapping of both populations was also conducted in IciMapping software version 4.2 (Integrated Software for Building Genetic Linkage Maps and Mapping Quantitative Trait Loci; Meng et al., 2015) using default settings for Inclusive Composite Interval Mapping (ICIM). One thousand permutations were used to test each trait at a significance level of 0.05 to determine LOD thresholds.
Annotation of candidate genes
Candidate genes were manually identified within 1.5 Mb on each side of the reported QTL peak by searching a master file compiled from multiple databases and manuscripts, as described in Huggins et al. (2019). The file was updated in May 2024 and includes gene annotations from the Os-Nipponbare-Reference _IRGSP-1.0 assembly (https://rice.uga.edu/pub/data/Eukaryotic_Projects/o_sativa/annotation_dbs/; accessed 28 May 2024) (Kawahara et al., 2013; Ouyang et al., 2006), the Rice Annotation Project (RAP-Db, https://rapdb.dna.affrc.go.jp/download/irgsp1.html?version=2024-01-11; accessed 28 May 2024) (Sakai et al., 2013), candidate genes for biotic and abiotic stress identified from Cohen and Leach (2019), genes listed in Oryzabase (https://shigen.nig.ac.jp/rice/oryzabase/download/gene; accessed 28 May 2024) (Yamazaki et al., 2010), genes listed in RiceNavi (Wei et al., 2021), and cloned genes from FunRiceGenes (https://funricegenes.github.io/geneInfo.table.txt; accessed 28 May 2024) (Huang et al., 2022).
In silico analysis of candidate genes within mapped QTL
All loci associated with a particular QTL were retrieved from the Rice Annotation Project Database (https://rapdb.dna.affrc.go.jp; Sakai et al., 2013). Genomic variation within these loci was obtained from the RiceVarMap database (https://ricevarmap.ncpgr.cn; Zhao et al., 2015) by identifying polymorphic positions between cultivars (cultivar IDs W025 and W409 for the TRJ-derived QTL, cultivar IDs W005 and W303 for the AUS-derived QTL). Loci containing InDels, nonsense mutations, and/or likely splice site mutations were considered for further analysis, as these variants are likely to cause changes in protein function.
Expression data for these candidate genes were obtained from NCBI project IDs PRJNA610422 and PRJNA430015 using the Rice RNA-seq Database (https://plantrnadb.com/ricerna/; Yu et al., 2022). For PRJNA610422, 4°C-0 h samples and 4°C-12 h samples were compared. For PRJNA430015, “CTR control” samples and “CTR cold stress” samples were compared. A heteroscedastic t-test was performed to determine whether expression differed significantly between samples.
Results and discussion
Description of AUS and TRJ RIL populations
The NC 1/536 × Phudugey AUS RILs, Taichu Mochi 59 × British Honduras Creole TRJ RILs, and the four parents were phenotyped under field conditions for the agronomic traits, HD, PTHT, and CULMHAB (Supplementary Figure S2 shows images of the AUS parents and a representative RIL in Supplementary Figure S2A, and the TRJ parents with selected RILs in Supplementary Figure S2B). In the AUS population, Phudugey, the male parent, headed slightly later and was slightly taller than the female parent, NC 1/536. The 153 RILs exhibited transgressive variation, with progeny taking 93 to 128 days to reach 50% heading, and plant height ranged from 100 to 195 cm (Table 1; Supplementary Table S1; Supplementary Figure S3A). The mean RIL rating for culm habit was open (5.7), similar to both parents, but individual RILs ranged from erect (1) to spreading (7).
Table 1. Summary statistics (overall mean, SE, and range of progeny) of six traits measured in the NC 1/536 (GSOR 310381) × Phudugey (GSOR 310715) aus (AUS) RIL population included in the quantitative trait locus analysis (Tables 3, 4) and in the parents.
The parents of the TRJ population, Taichu Mochi 59 and British Honduras Creole, also differed for the three agronomic traits evaluated. Compared to Taichu Mochi 59, the male parent, British Honduras Creole, took about five fewer days to head, was taller, and had an open (4.4) culm habit compared with the erect (1.4) culm habit of Taichu Mochi 59. The 122 TRJ RILs exhibited transgressive variation for all three agronomic traits, as presented in Table 2, Supplementary Table S2, and Supplementary Figure S3B.
Table 2. Summary statistics (overall mean, SE, and range of progeny) of six traits measured in the Taichu Mochi 59 (GSOR 310080) × British Honduras Creole (GSOR 310144) tropical japonica (TRJ) RIL population included in the quantitative trait locus analysis (Tables 3, 4) and in the parents.
Both the AUS and TRJ RIL populations were evaluated for two low-temperature traits: EL, where more cold-tolerant RILs generally exhibit less leakage, and LTSS, where more cold-tolerant RILs have higher survivability. The LTSS values for the RILs were not normally distributed, so the logit function in the GLMM was used to transform the data, resulting in a distribution closer to normal, particularly for the TRJ RILs (Supplementary Figures S3A, B). Comparing the distribution of the 152 AUS RILs evaluated for cold tolerance (Table 1; Supplementary Figure S3A) with that of the 90 TRJ RILs evaluated for the same traits (Table 2; Supplementary Figure S3B) confirmed that the Japonica VG, which includes the TRJ subpopulation, is more cold tolerant than the Indica VG, which includes the AUS subpopulation (Liu et al., 2018). This is evident by the lower mean EL in the TRJ RILs (21.62%) compared with the AUS RILs (26.52%). Similarly, the mean LTSS (82.46%) and LTSS-logit value (2.66) for the TRJ RILs were higher than the mean LTSS (58.58%) and LTSS-logit (0.56) value for the AUS RILs.
For the AUS RILs, significant negative correlations were observed between heading date and both plant height (r = − 0.32; p ≤ 0.0001) and culm habit (r = − 0.22; p ≤ 0.01), indicating that taller plants took fewer days to head and were more upright (Supplementary Figure S4A). Plant height and LTSS BLUP and logit (r = 0.26; p = 0.01) had significant positive correlations, suggesting that taller RILs had higher LTSS in this population. As expected, EL was negatively correlated with LTSS logit and BLUP (r = − 0.28; p = 0.001), since EL decreases with increased LTSS. For the TRJ RILs, culm habit had significant negative correlations with heading date (r = − 0.48; p ≤ 0.001) and both LTSS values, logit (r = − 0.27; p = 0.01) and BLUP, as shown in Supplementary Figure S4B, indicating that more upright TRJ RILs took longer to head and were more cold tolerant. The significant correlation between plant height and LTSS observed in the AUS RILs was not found in the TRJ RILs.
Mapping potential validation of agronomic traits for AUS and TRJ RIL populations
The 153 NC 1/536 × Phudugey AUS RILs and their parents were genotyped with the 7,098 DNA markers included on the C7AIR (Morales et al., 2020). Of these, 519 markers were polymorphic between the parents, and 516 were used to construct a 416.50-centimorgan (cM) linkage map. The markers mapped to 292 different cM positions, with an average interval of 1.43 cM and a maximum distance of 14.09 cM between markers (Supplementary Table S3). Four regions exhibited significant segregation distortion based on χ2 values. Three regions on chr. 2 (32.6–35.2 cM), chr. 4 (4.0–4.9 cM), and chr. 6 (23.6–33.6 cM) were distorted in favor of the Phudugey alleles, while one region on chr. 9 (22.0–22.5 cM) was distorted in favor of NC 1/536 alleles.
QTL mapping was conducted for three agronomic traits in both the AUS and TRJ RIL populations as a “quality control” to assess whether known QTL for these traits and the associated genes could be uncovered (Table 3). QTL mapping within the AUS RILs revealed five QTL for agronomic traits: three HD QTL, one PTHT QTL, and one CULMHAB QTL (Table 3A; Figure 1A). For the three HD QTL—qHD3, qHD6, and qHD7—the additional days were attributed to the Phudugey parent. Of note, these three HD-QTL regions include four well-known genes for heading date, namely qHD3, encompassing the MADS BOX GENE 50 (MADS50) at 1.30 Mb on chr. 3 (Lee et al., 2004); qHD6, including both RICE FLOWERING-LOCUS T 1 (RFT1) at 2.93 Mb (Komiya et al., 2009) and HEADING DATE 3A (HD3A) at 2.94 Mb (Kojima et al.., 2002), both on chr. 6; and lastly, qHD7, which most likely includes HEADING DATE 2 (HD2) on chr. 7 at 29.62 Mb, first identified in the late 1990s (Yano et al., 1997; Yamamoto et al., 1998).
Table 3. Days to 50% heading (HD), plant height (PTHT), and culm habit (CULMHAB) quantitative trait loci (QTL) identified in the NC 1/536 × Phudugey aus (AUS) and Taichu Mochi 59 × British Honduras Creole tropical japonica (TRJ) recombinant inbred line (RIL) mapping populations.
Figure 1. Quantitative trait loci (QTL) maps of (A) AUS and (B) TRJ recombinant inbred lines (RILs) in Mb. Cold tolerance traits are shown in blue and agronomic traits in green. Three cold tolerance traits were mapped in both panels: percent electrolyte leakage (qEL), percent low temperature seedling survivability (qLTSS(percent)), and low temperature seedling survivability logit (qLTSS(logit)), which predicts the probability of survival under cold treatment on the log-odds scale. Three agronomic traits were also determined: days to 50% heading (qHD), plant height (qPTHT), and culm habit (qCULMHAB). In (A), the mapping was conducted using NC 1/536 (GSOR 310381) × Phudugey (GSOR 310715) RILS—152 for cold tolerance and 153 for agronomic traits—with 516 single-nucleotide polymorphism (SNP) markers; while in (B), the mapping was conducted using Taichu Mochi 59 (GSOR 310080) × British Honduras Creole (GSOR 310144) RILs—90 for cold tolerance and 122 RILs for agronomic traits—with 1,042 SNP markers. The 12 rice chromosomes are shown with SNP markers as gray horizontal lines and centromeres in black. (Maps created with MapChart 2.3 [Voorrips, 2002].).
A single PTHT QTL, qPH11, was detected on chr. 11 and was attributed to the Phudugey parent (Supplementary Table S3). One candidate gene, PYRUVATE KINASE 1 (PK1) at 2.24 Mb, was identified within this QTL region. Pyruvate kinase (PK) catalyzes the final step of glycolysis and influences the glycolytic pathway, and studies of OsPK1 mutants have demonstrated its role in plant morphological development, particularly affecting plant height (Zhang et al., 2012).
Similarly, a single QTL for culm habit, qCULMHAB9, was detected on chr. 9. One of the two His phosphotransfer proteins (HP), HISTIDINE CONTAINING PHOSPHOTRANSMITTER 2 (AHP2), which regulates cytokinin signaling, is located in this QTL region at 22.65 Mb. Cytokinin is an important phytohormone that regulates plant growth and development, and Sun et al. (2014) reported that plants deficient in cytokinin signaling were shorter due to reduced internodes, had fewer tillers, and reduced fertility, which may have affected culm habit.
The 123 Taichu Mochi 59 × British Honduras Creole TRJ RILs and their parents were genotyped using the 7,098 DNA markers included on the C7AIR (Morales et al., 2020). Of these markers, 1,045 were polymorphic between the parents, and 1,042 were used to construct a 569.2-cM linkage map. The markers were mapped to 512 different cM positions, with an average interval of 1.11 cM and a maximum of 12.10 cM between markers (Supplementary Table S4). Six regions exhibited significant segregation distortion based on χ2 values. Four regions on chr. 1 (35.5–44.6 cM), chr. 5 (3.7–23.7 cM), chr. 7 (26.1–31.0 cM), and chr. 8 (30.1–38.7 cM) were distorted in favor of the British Honduras Creole alleles, whereas two regions on chr. 9 (5.7–25.1 cM) and chr. 12 (1.8–9.0 cM) were distorted in favor of Taichu Mochi 59.
QTL mapping in the TRJ population with 122 RILs for agronomic traits revealed six QTL (Table 3; Figure 1B). One TRJ RIL (trj_35) did not have agronomic data because the line did not grow in the field; thus, it was only used for mapping the cold tolerance traits. Mapping days to HD revealed a single QTL, qHD3, which included the well-known MADS50 gene at 1.30 Mb on chr. 3 (Lee et al., 2004). This QTL was attributed to the Taichu Mochi 59 parent, which had delayed heading. Notably, MADS50 was also one of the four candidate genes identified with HD QTL mapping in the AUS RIL population.
Three PTHT QTL were found, with qPTHT1 and qPTHT5 attributed to the British Honduras Creole parent and qPTHT3 to Taichu Mochi 59. The qPTHT1 region (31.8–32.3 Mb) was proximal to the well-known “green revolution gene”, SEMI-DWARF 1 (SD1), on chr. 1 at 38.38 Mb, which produced semidwarf plants due to a mutation in the GIBBERELLIN 20-OXIDASE2 (GA20OX2) gene, resulting in dwarfism caused by a deficiency in the gibberellin (GA) plant growth hormones (Sasaki et al., 2002; Spielmeyer et al., 2002). One candidate gene in the qPH1 region, GIBBERELLIN 2-OXIDASE 3 (GA2OX3) at 31.80 Mb (Sakai et al., 2003), is involved in the regulation of GA. A second gene in this region, GRETCHENHAGEN 3-2 (OsGH3-2) at 32.46 Mb, encodes an indoleacetic acid-amido synthetase and interacts with the GA20OX1 and GA20OX2 genes to affect plant height by regulating GA biosynthesis and auxin metabolism (Kim et al., 2024). Also, in the qPTHT5 region, SEMI-DWARF 5 (SDF5) is located at 20.33 Mb and is involved in the GA biosynthesis pathway and modifies plant architecture, including plant height (Yachun et al., 2021). In the qPTHT3 region, the COP9 SIGNALOSOME COMPLEX SUBUNIT 1 (CSN1) was a positive regulator of plant height under far-red light, and exogenous GA3 treatment increased height (Li et al., 2025). Lastly, single mutations in the GDP DISSOCIATION INHIBITOR 1 (OsGDI1) in the qPTHT5 region resulted in dwarf plants as well as longer and thinner grains (Shad et al., 2023).
Culm habit (CULMHAB) QTL were discovered on chr. 3, qCULMHAB3, with the increased spreading attributed to British Honduras Creole, and chr. 8, qCULMHAB8, with the increase attributed to Taichu Mochi 59. Genes associated with plant architecture were identified in or near each of these QTL regions. TILLERING AND DWARF 1 (TAD1) is on chr. 3 at 1.33 Mb (Xu et al., 2012) mainly affects tiller number and plant height. The WEALTHY FARMER’S PANICLE (WFP) or IDEAL PLANT ARCHITECTURE 1 (IPA1) is on chr. 8 at 25.27 Mb (Miura et al., 2010; Jiao et al., 2010) and encodes SOUAMOSA PROMOTER BINDING PROTEIN-LIKE 14 (OsSPL14), which affects plant height, tiller number and angle, panicle branching, and the number of grains produced.
Cold tolerance trait mapping
QTL mapping was conducted for two low-temperature traits in both the AUS and TRJ RIL subpopulations. Percent EL was used to measure membrane damage right after 1 week of exposure to 10°C, and percent LTSS was used to measure seedling survival after 1 week of exposure to 10°C and 1 week of recovery at warm temperatures. This revealed three EL QTL (qEL) and four LTSS QTL (qLTSS) for the AUS subpopulation, and two qEL and four qLTSS for the TRJ subpopulation (Table 4). For qLTSS mapping in the AUS subpopulation, increased seedling survivability associated with three of the four LTSS QTL—qLTSS1–1, qLTSS6, and qLTSS11—was attributed to alleles from the relatively cold-tolerant Phudugey parent, while for qEL mapping, increased electrolyte leakage and thus increased membrane damage after cold exposure associated with two of the three EL QTL, qEL8 and qEL11 (overlapping with qLTSS11), was attributed to alleles from the cold-sensitive NC 1/536 parent. This showed that although the cold-tolerant parent contributed more alleles to enhance cold tolerance scores of the RIL population, the cold-sensitive parents also provided positive alleles at some loci. This agrees with the observation that the percent LTSS and percent EL range of the RIL population was wider than the range between the two AUS parents (see Table 1), thus explaining the observed “transgression” values when positive or negative alleles from each parent were combined in individual RILs. This is particularly noteworthy for the EL trait, where both parents had a mean of 25%, while the RILs had a range of 15%–54% (Table 1).
Table 4. Cold tolerance quantitative trait loci (QTL) for percent electrolyte leakage (EL, BLUP) and low-temperature survivability (LTSS, both a percent BLUP and logit) identified in the NC 1/536 × Phudugey aus (AUS) and Taichu Mochi 59 × British Honduras Creole tropical japonica (TRJ) recombinant inbred line (RIL) mapping populations.
For qLTSS mapping in the TRJ subpopulation, increased seedling survivability associated with three of the four LTSS QTL—qLTSS4, qLTSS7-2, and qLTSS10—was attributed to alleles from the very cold-tolerant British Honduras Creole parent, while for qEL mapping, increased electrolyte leakage and thus increased membrane damage after cold exposure associated with qEL4 and qEL11 was attributed to alleles from the more cold-sensitive Taichu Mochi 59 parent and the more tolerant British Honduras Creole parent, respectively (Table 4). This again showed, as for the AUS subpopulation, that both TRJ parents contributed positive or negative alleles to enhance or reduce cold tolerance scores, respectively, because the percent LTSS and percent EL range of the RIL population was wider than the range between the two TRJ parents (Table 2).
Identification of cold tolerance candidate genes
One rationale for the current study was to identify rice subpopulation-specific cold tolerance QTL and associated candidate genes that could be used to improve cold tolerance traits within the two rice subpopulations. Due to the lower density of SNPs within subpopulation RILs compared to RILs generated from crosses between parents from different subpopulations, the 13 cold tolerance trait QTL identified here covered larger genomic regions than those we mapped previously using TEJ × AUS biparental crosses (Schläppi et al., 2023). Interestingly, only two AUS QTL overlapped with our previously mapped QTL from TEJ × AUS biparental, qLTSS1–2(logit and percent) and qLTSS6(logit and percent), while none of the TRJ QTL did (Table 4; Schläppi et al., 2023). This showed that our approach indeed uncovered subpopulation-specific QTL and associated cold tolerance candidate genes. However, the larger QTL regions overlapped with several smaller QTL we previously identified using genome-wide association studies (Supplementary Table S5; Schläppi et al., 2017; Shimoyama et al., 2020).
To identify EL and LTSS trait candidate genes, we searched within the 13 QTL regions for mostly high-impact single-nucleotide polymorphisms (SNPs) or insertion/deletion (InDel) variants in genes leading to frameshift, premature termination, lost start codon, lost stop codon, or splice variants. Altogether, 23 and 12 genes with variants were identified for the AUS and TRJ RILs, respectively (Table 5). For the AUS QTL qEL8(BLUP) and qEL11(BLUP), alleles from the cold-sensitive parent increased EL and thus membrane damage after cold exposure. Consistently, high-impact variants were identified in alleles of the cold-sensitive parent for all seven candidate genes listed in Table 4. Specifically, biological functions for candidate genes encoding a peroxidase (OsPRX134), a receptor-like cytoplasmic kinase (OsRLCK352), and a potential splice regulator (OsU2AF65B) associated with qEL11(BLUP), and genes coding for an aldolase (Badh2), annexin (OsANN7), and F-box factor (OsFBDUF45) associated with qEL8(BLUP), were most likely disrupted in the cold-sensitive AUS parent. Among these candidate genes, OsANN7 might be of particular interest because it was recently shown that another annexin family member, OsANN5, is involved in cold stress tolerance at the seedling stage (Que et al., 2023). In contrast, for the AUS QTL qEL10(BLUP), alleles from the cold-tolerant parent increased EL, and consistently, a high-impact variant was identified in a wall-associated kinase (OsWAK22) of the cold-tolerant parent. Disruption of the biological function of this candidate gene might increase membrane damage after cold exposure, possibly because signal transduction from the cell wall across the plasma membrane to the cytoplasm might be partially compromised. Correct signaling by OsWAK22 might dampen immune responses during cold temperature exposure (Brusamarello-Santos et al., 2019).
Table 5. High-impact SNP and InDel variants in aus parents, comparing the cold-sensitive NC 1/536 with cold-tolerant Phudugey, and in tropical japonica parents, comparing the cold-sensitive Taichu Mochi 59 with cold-tolerant British Honduras Creole.
For AUS QTL qLTSS1–1(logit and percent), qLTSS6(logit and percent), and qLTSS11(logit and percent), alleles from the cold-tolerant parent increased the low-temperature seedling survival trait. Consistently, high-impact variants were identified in alleles of the cold-sensitive parent for most of these QTL candidate genes listed in Table 4, including qLTSS11(logit and percent), which overlaps with qEL11(BLUP) (see above). Thus, defects in the biological functions of at least one of qLTSS1–1(logit and percent)-associated candidate genes coding for IPA1 interacting protein 1 (IPI1), OsWAK6, a C2 calcium-dependent membrane targeting domain containing protein (OsC2DP4), or Laccase 1 (OsLAC1) might have negative effects on cold tolerance. Interestingly, the cold-tolerant AUS parent has a frameshift in an esterase-encoding gene (LOC_Os01g25360), which may have no consequence or could even positively affect cold tolerance. Moreover, in the cold-sensitive parent, defects in the biological functions of at least one of qLTSS6(logit and percent)-associated candidate genes—coding for a resistance protein (RLP32), MAP kinase kinase 1 (OsMKK1; described in Schläppi et al., 2017), a ubiquitin family gene (LOC_Os06g05760), an RNA-binding protein (RRM1–115), an esterase (LOC_Os06g06080; described in Schläppi et al., 2023), a U-box E3-ubiquitin ligase (OsPUB64), or a Gly-Asp-Ser-Leu (GDSL)-type esterase/lipase (OsGELP80)—might negatively affect cold tolerance. Of special interest is OsMKK1, which we previously flagged as a candidate gene in Schläppi et al. (2017) and which is downregulated by cold temperatures (Table 6), possibly to dampen cold stress signaling events. Of further interest are IPI1, LOC_Os06g05760, and OsPUB64, because, first, they belong to ubiquitin and E3-ubiquitin ligase family genes that might be involved in regulating protein homeostasis in response to cold temperature stress (Wang et al., 2017; Kim et al., 2021); second, IPI1 and OsPUB64 are upregulated by cold temperatures (Table 6); and third, we recently have shown that OsUBC7, an E2-ubiquitin conjugase-encoding gene, is involved in the early cold tolerance response of rice (Phan and Schläppi, 2025). Thus, IPI1 and OsPUB64 might belong to a similar biochemical pathway as previously identified rice cold tolerance genes, such as OsUBC7, OsHOS1, and OsSRFP1—two E3-ubiquitin ligase-encoding genes (Lourenço et al., 2013; Fang et al., 2015). Interestingly, the cold-tolerant AUS parent also contributed defective alleles of genes coding for a histone chaperone (OsNAPL2; described in Schläppi et al., 2023) and a transferase (LOC_Os06g05790; described in Schläppi et al., 2023), which either had no effect or a positive effect on cold tolerance. By contrast, for AUS QTL qLTSS1-2(logit and percent), alleles from the cold-sensitive parent increased low-temperature seedling survival. Consistently, the cold-tolerant parent had a high-impact variant in a gene coding for a heat-shock chaperone DnaJ (Hsp40), while the cold-sensitive parent had the (most likely) functional reference allele. However, the second candidate gene, coding for a calmodulin-binding transcription activator (OsCAMTA5), had a splice acceptor variant (vg0140402807) in the cold-sensitive parent, which either has no effect or might positively affect calcium signaling during cold stress, meriting further investigation.
Table 6. Cold temperature-regulated gene expression of QTL-associated candidate genes with high-impact variants between cold-tolerant and cold-sensitive rice accessions.
Taken together, the AUS subpopulation-specific QTL mapping for the EL and LTSS traits provided 23 candidate genes involved in phospholipid binding, lipase activity, calcium and kinase signaling, immune response, RNA binding, and ubiquitin-mediated protein turnover. Most of these candidate genes likely carried functional reference alleles from the cold-tolerant parent, correlating with enhanced cold tolerance potential. An additional interesting candidate gene is Badh2, which encodes a betaine aldehyde dehydrogenase that inhibits the biosynthesis of the major rice fragrance component 2-acetyl-1-pyrroline (Chen et al., 2008) and is upregulated by cold (Table 6). Whether variants of this gene for eating quality also influence cold tolerance merits further investigation; however, it is noteworthy that its paralog, Badh1, plays a positive role in rice cold stress tolerance (Tang et al., 2014).
For overlapping TRJ QTL qEL4(BLUP) and qLTSS4(logit), alleles from the relatively cold-sensitive parent increased EL and decreased LTSS, respectively. Only one candidate gene, coding for an intracellular Ras-group-related leucine-rich repeat (LRR) family protein (OsIRL1), was identified, which had a frameshift variant allele in the relatively cold-sensitive parent. This defect might negatively affect stress tolerance signaling during cold exposure (You et al., 2010) and merits further investigation. For TRJ QTL qEL11(BLUP), alleles from the cold-tolerant parent increased EL; however, no candidate gene with high-impact variants was identified. For TRJ QTL qLTSS7-2(logit) and qLTSS10(logit), alleles from the cold-tolerant parent increased LTSS. Consistently, almost all candidate genes identified in the QTL regions carried high-impact variants in alleles from the relatively cold-sensitive parent. Defects in genes coding for a trichome birefringence-like protein (OsTBL24), a magnesium chelatase (OsCHLH), a ubiquitin carboxyl-terminal hydrolase domain containing protein (OsUBP10), and a PHD-finger family protein (OsPHD45), associated with qLTSS7-2(logit), as well as genes coding for an AAA-type ATPase (LOC_Os10g22700), a hypothetical protein (LOC_Os10g22770), a Cf2/Cf5 disease resistance protein (LOC_Os10g22930), a triacylglycerol lipase (OsABH1), and a basic helix–loop–helix factor (OsbHLH045), associated with qLTSS10(logit), might have negative effects on cold tolerance. Only one gene associated with qLTSS7-2(logit), OsANN6, had a high-impact splice donor variant in the cold-tolerant parent, which either had no effect or positively affected cold tolerance. In contrast, for TRJ QTL qLTSS7-1(logit), alleles from the relatively cold-sensitive parent increased LTSS. Only one candidate gene coding for a hypothetical protein (LOC_Os07g43770) was identified, which had a high-impact variant in the cold-tolerant parent, indicating that the reference allele might have a positive effect on cold tolerance.
In addition to OsIRL1, additional TRJ genes of interest that might merit further investigation include the following: OsCHLH, for which transposon insertions are lethal due to an albino phenotype (Jung et al., 2003; Miyao et al., 2007), whereas the variant uncovered here (Table 5) is viable and downregulated by cold (Table 6); OsPHD45, which might regulate growth and development during cold stress and is upregulated by cold (Table 6); LOC_Os10g22930, previously identified as a candidate gene for low-temperature tolerance during germination (Wang et al., 2018); and the transcription factor encoding gene OsbHLH045, expressed in the root (Li et al., 2006), downregulated by cold (Table 6), and potentially regulating cold stress-responsive genes in this organ, as well as possibly playing a role in carotenoid biosynthesis (Palaniswamy et al., 2024). Finally, OsABH1 is another interesting candidate because this lipase affects cellular triglyceride content, possibly influences cell wall components, and is downregulated by cold, which might positively affect cold tolerance potential (Jin et al., 2024).
Conclusions
In this study, we used AUS and TRJ subpopulation-specific biparental mapping populations, derived from crosses between cold-tolerant and cold-sensitive parents, to map subpopulation-specific cold tolerance QTL. We first performed “quality control” QTL analyses to demonstrate that QTL for heading date, plant height, and culm habit contained known genes/QTL for all traits, such as MADS50, RFT1, HD2, PK1, GA2OX3, CSN1, SDF5, OsGDI1, TAD1, and WFP, thus validating that the populations were suitable for cold tolerance trait QTL discovery (Table 3). Only two of the 13 cold tolerance QTL identified here overlapped with the QTL we previously obtained through biparental mapping using different subpopulation parents, indicating that this approach identified subpopulation-specific QTL that could be used to improve the cold tolerance potential of each respective subpopulation. We identified 35 candidate genes for cold tolerance in the QTL regions—23 for the AUS subpopulation and 12 for the TRJ subpopulation—and 80% (28/35) had gene-disruptive nucleotide variants in the cold-sensitive parent (Table 5). Based on allelic contributions to the QTL effect, we infer that the relatively cold-sensitive AUS and TRJ parents also contain functional genes that could be introduced into the cold-tolerant parents to further enhance their cold tolerance potential. The 35 candidate genes could be assigned to modules that regulate different cellular processes. Specifically, 10 genes (29%) encoded receptor-like (OsRLCK), wall-associated (OsWAK), MAPK, or other kinases, as well as ROS or calcium signaling components that, together with the three (9%) transcription factor genes OsPHD45, OsbHLH045, and OsCAMTA5, might be involved in transducing cold stress tolerance responses. Among the other 25 genes, five (14%) coded for ubiquitin- and F-box-type proteins, such as OsUBP10, IPA1, OsPUB64, and OsFBDUF45, which, together with the two (9% of AUS candidates) chaperone genes OsNAPL2 and Hsp40, might regulate protein integrity and/or protein homeostasis as part of the cold stress tolerance response. Seven (20%) candidate genes coded for enzymes such as aldolases, esterases, laccases, lipases, and transferases that, together with the two (6%) RNA-binding genes OsU2AF65B and RRM1–115, might regulate cellular and RNA metabolic processes as part of specific cold stress tolerance response mechanisms. These candidate genes can be functionally analyzed in future studies using genomics, molecular genetics, and biochemical approaches, while the QTL regions containing them can be used for marker-assisted breeding of cold-tolerant rice cultivars.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.
Author contributions
GE: Conceptualization, Formal Analysis, Funding acquisition, Investigation, Resources, Supervision, Validation, Writing – original draft, Writing – review & editing. YS: Investigation, Writing – review & editing. AJ: Formal Analysis, Methodology, Software, Visualization, Writing – review & editing. HP: Investigation, Methodology, Writing – review & editing. MJ: Data curation, Methodology, Validation, Writing – review & editing. QG: Investigation, Writing – review & editing. JE: Formal Analysis, Methodology, Software, Validation, Writing – review & editing. EH: Formal Analysis, Visualization, Writing – review & editing. MS: Conceptualization, Formal Analysis, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Writing – original draft, Writing – review & editing.
Funding
The author(s) declared financial support was received for this work and/or its publication. This work was supported by USDA-AFRI-NIFA grant no. 2016-67013-24587 and by USDA-AFRI-NIFA grant no. 2023-67013-39534. Mention of a trademark or proprietary product does not constitute a guarantee or warranty of the product by the US Department of Agriculture and does not imply its approval to the exclusion of other products that also can be suitable. The USDA is an equal opportunity provider and employer. All experiments complied with the current laws of the United States, the country in which they were performed.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declare that no Generative AI was used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2025.1716845/full#supplementary-material
Supplementary Figure 1 | Genetic distance between the AUS and TRJ recombinant inbred line parents using SSR genotypes of RMC accessions.
Supplementary Figure 2 | (A) Images of NC 1/536 (GSOR 310381) and Phudugey (GSOR 310715) parents of the AUS recombinant inbred line (RIL) population. (B) Images of Taichu Mochi 59 (GSOR 310080) and British Honduras Creole (GSOR 310144), parents of the TRJ recombinant inbred line population. The eight TRJ RILs illustrate the diversity in the population.
Supplementary Figure 3 | (A) Frequency distributions of 152 NC 1/536 x Phudugey AUS RILs evaluated for three seedling cold tolerance traits (A) percentage electrolyte leakage (EL), (B) low temperature seedling survivability-logit (LTSS logit) and (C) percentage LTSS (LTSS) and 153 RILs evaluated for three agronomic traits (D) days to 50% heading, (E) plant height and (F) culm habit, in the field near Stuttgart, Arkansas, USA. The Tukey outlier box plot illustrates the distribution for the traits evaluated. The diamond in the box indicates the mean, the line in the box indicates the median, and the whisker indicates the SE. The red bracket defines the shortest half of the data (the densest region). (B) Frequency distributions of 90 Taichu Mochi 59 (TM59) x British Honduras Creole (BHC) TRJ RILs evaluated for three seedling cold tolerance traits (A) percentage electrolyte leakage (EL), (B) low temperature seedling survivability logit (LTSS logit) and (C) percentage LTSS (LTSS) and 122 RILs evaluated for three agronomic traits (D) days to 50% heading, (E) plant height and (F) culm habit, in the field near Stuttgart, Arkansas, USA. The Tukey outlier box plot illustrates the distribution for the traits evaluated. The diamond in the box indicates the mean, the line in the box indicates the median, and the whisker indicates the SE. The red bracket defines the shortest half of the data (the densest region).
Supplementary Figure 4 | (A) Correlation table and scatterplot matrix of the correlations between 153 NC 1/536 x Phudugey AUS RILs evaluated for three agronomic traits (A) days to 50% heading (HD), (B) plant height (PTHT) and (C) culm habit (CULMHAB) in the field near Stuttgart, Arkansas, USA and the 152 RILs evaluated for three seedling cold tolerance traits (D) percentage electrolyte leakage (EL), (E) percentage low temperature seedling survivability-blup (LTSS-blup) and (F) LTSS-logit (LTSS-logit). Correlations highlighted in blue are significant at p = 0.01 and in orange at p=<0.0001. Correlation coefficients were calculated using the multivariate option, restricted maximum likelihood (REML) method. (B) Correlation table and scatterplot matrix of the correlations between 122 Taichu Mochi 59 x British Honduras Creole TRJ RILs evaluated for three agronomic traits (A) days to 50% heading (HD), (B) plant height (PTHT) and (C) culm habit (CULMHAB) in the field near Stuttgart, Arkansas, USA and the 90 RILs evaluated for three seedling cold tolerance traits (D) percentage electrolyte leakage (EL), (E) percentage low temperature seedling survivability-blup (LTSS-blup) and F) LTSS-logit (LTSS-logit). Correlations highlighted in blue are significant at p = 0.01 and in orange at p=<0.0001. Correlation coefficients were calculated using the multivariate option, restricted maximum likelihood (REML) method.
Supplementary Table 1 | List of the days to 50% heading (HD), plant height (PTHT) and culm habit (CULMHAB) for each AUS recombinant inbred line (RIL) and means for the parents, NC 1/536 and Phudugey. For the cold tolerance traits, the “adjusted means” were calculated for percent electrolyte leakage (EL) and low-temperature seedling survivability (LTSS), both BLUP and logit values are listed for each AUS RIL and the parents.
Supplementary Table 2 | List of the days to 50% heading (HD), plant height (PTHT) and culm habit (CULMHAB) for each TRJ recombinant inbred line (RIL) and means for the parents, Taichu Mochi 59 (TM59) x British Honduras Creole (BHC). For the cold tolerance traits, the “adjusted means” were calculated for percent electrolyte leakage (EL) and low-temperature seedling survivability (LTSS), both BLUP and logit values are listed for each TRJ RIL and the parents.
Supplementary Table 3 | List of the 516 single nucleotide polymorphism (SNP) markers from the Cornell-IR LD Rice Array (C7AIR) (Morales et al., 2020) that were polymorphic between the NC 1/536 and Phudugey, the AUS parents.
Supplementary Table 4 | List of the 1,042 single nucleotide polymorphism (SNP) markers from the Cornell-IR LD Rice Array (C7AIR) (Morales et al., 2020) that were polymorphic between Taichu Mochi 59 and British Honduras Creole, the TRJ parents.
Supplementary Table 5 | Overlap between AUS or TRJ QTL intervals listed on the left side and GWAS-QTL intervals identified in the Rice Mini-core Collection (RMC), Rice Diversity Panel 1 (RDP1), and the backcross inbred line (BIL) population which are listed on the right side. The GWAS-QTL found in the complete panel are designated as “ALL” and those found in a subgroup are designated as the panel and the subgroup, Japonica VG (JAP), Indica VG (IND) or a subpopulation, tropical japonica (TRJ) or aus (AUS).
Supplementary Table 6 | Cold temperature regulation of AUS and TRJ RIL QTL associated candidate genes. The 22 candidate genes are highlighted in yellow and the expression data reported by Yu et al. (2022) is listed below the gene and highlighted in green. The significant difference was confirmed by t-tests highlighted in green.
References
Brusamarello-Santos, L. C. C., Alberton, D., Valdameri, G., Camilios-Neto, D., Covre, R., Lopes, K. D. P., et al. (2019). Modulation of defence and iron homeostasis genes in rice roots by the diazotrophic endophyte Herbaspirillum seropedicae. Sci. Rep. 9, 10573. doi: 10.1038/s41598-019-45866-w
Chen, S., Yang, Y., Shi, W., Ji, Q., He, F., Zhang, Z., et al. (2008). Badh2, encoding betaine aldehyde dehydrogenase, inhibits the biosynthesis of 2-acetyl-1-pyrroline, a major component in rice fragrance. Plant Cell 20, 1850–1861. doi: 10.1105/tpc.108.058917
Cohen, S. P. and Leach, J. E. (2019). Abiotic and biotic stresses induce a core transcriptome response in rice. Sci. Rep. 9, 6273. doi: 10.1038/s41598-019-42731-8
Edwards, J. D., Baldo, A. M., and Mueller, L. A. (2016). Ricebase: a breeding and genetics platform for rice, integrating individual molecular markers, pedigrees and whole-genome-based data. Database 2016, baw107. doi: 10.1093/database/baw107
Fang, H., Meng, Q., Xu, J., Tang, H., Tang, S., Zhang, H., et al. (2015). Knock-down of stress inducible OsSRFP1 encoding an E3 ubiquitin ligase with transcriptional activation activity confers abiotic stress tolerance through enhancing antioxidant protection in rice. Plant Mol. Biol. 87, 441–458. doi: 10.1007/s11103-015-0294-1
Gann, P. J., Esguerra, M., Counce, P. A., and Srivastava, V. (2021). Genotype-dependent and heat-induced grain chalkiness in rice correlates with the expression patterns of starch biosynthesis genes. Plant-Environment Interact. 2, 165–176. doi: 10.1002/pei3.10054
Huang, F., Jiang, Y., Chen, T., Li, H., Fu, M., Wang, Y., et al. (2022). New data and new features of the FunRiceGenes (functionally characterized rice genes) database: 2021 update. Rice 15, 23. doi: 10.1186/s12284-022-00569-1
Huggins, T. D., Chen, M.-H., Fjellstrom, R. G., Jackson, A. K., McClung, A. M., and Edwards, J. D. (2019). Association analysis of three diverse rice (Oryza sativa L.) germplasm collections for loci regulating grain quality traits. Plant Genome 12, 170085. doi: 10.3835/plantgenome2017.09.0085
Jahan, N., Zhang, Y., Lv, Y., Song, M., Zhao, C., Hu, H., et al. (2019). QTL analysis for rice salinity tolerance and fine mapping of a candidate locus qSL7 for shoot length under salt stress. Plant Growth Regul. 90, 307–319. doi: 10.1007/s10725-019-00566-3
Jiang, X., Yu, S., Huang, Y., Huang, J., Liu, S., Yang, D., et al. (2024). Identification of the RRM1 gene family in rice (Oryza sativa) and its response to rice blast. PeerJ 12, e17668. doi: 10.7717/peerj.17668
Jiao, Y., Wang, Y., Xue, D., Wang, J., Yan, M., Liu, G., et al. (2010). Regulation of OsSPL14 by OsmiR156 defines ideal plant architecture in rice. Nat. Genet. 42, 541–544. doi: 10.1038/ng.591
Jin, X., Chen, J., Khan, A., Chen, Z., Gao, R., Lu, Y., et al. (2024). Triacylglycerol lipase, OsSG34, plays an important role in grain shape and appearance quality in rice. Plant J. 117, 840–855. doi: 10.1111/tpj.16532
Joehanes, R. and Nelson, J. C. (2008). QGene 4.0, an extensible Java QTL-analysis platform. Bioinformatics 24, 2788–2789. doi: 10.1093/bioinformatics/btn523
Jung, K.-H., Hur, J., Ryu, C.-H., Choi, Y., Chung, Y.-Y., Miyao, A., et al. (2003). Characterization of a rice chlorophyll-deficient mutant using the T-DNA gene-trap system. Plant Cell Physiol. 44, 463–472. doi: 10.1093/pcp/pcg064
Kawahara, Y., de la Bastide, M., Hamilton, J. P., Kanamori, H., McCombie, W. R., Ouyang, S., et al. (2013). Improvement of the Oryza sativa Nipponbare reference genome using next generation sequence and optical map data. Rice 6, 4. doi: 10.1186/1939-8433-6-4
Kim, J. S., Chae, S., Jo, J. E., Kim, K. D., Song, S.-I., Park, S. H., et al. (2024). OsMYB14, an R2R3-MYB transcription factor, regulates plant height through the control of hormone metabolism in rice. Molecules Cells 47, 100093. doi: 10.1016/j.mocell.2024.100093
Kim, M.-S., Kang, K.-K., and Cho, Y.-G. (2021). Molecular and functional analysis of U-box E3 ubiquitin ligase gene family in rice (Oryza sativa). Int. J. Mol. Sci. 22, 12088. doi: 10.3390/ijms222112088
Kim, S.-M., Suh, J.-P., Lee, C.-K., Lee, J.-H., Kim, Y.-G., and Jena, K. K. (2014). QTL mapping and development of candidate gene-derived DNA markers associated with seedling cold tolerance in rice (Oryza sativa L.). Mol. Genet. Genomics 289, 333–343. doi: 10.1007/s00438-014-0813-9
Kojima, S., Takahashi, Y., Kobayashi, Y., Monna, L., Sasaki, T., Araki, T., et al. (2002). Hd3a, a rice ortholog of the Arabidopsis FT gene, promotes transition to flowering downstream of Hd1 under short-day conditions. Plant Cell Physiol. 43, 1096–1105. doi: 10.1093/pcp/pcf156
Komiya, R., Yokoi, S., and Shimamoto, K. (2009). A gene network for long-day flowering activates RFT1 encoding a mobile flowering signal in rice. Development 136, 3443–3450. doi: 10.1242/dev.040170
Lee, S., Kim, J., Han, J.-J., Han, M.-J., and An, G. (2004). Functional analyses of the flowering time gene OsMADS50, the putative SUPPRESSOR OF OVEREXPRESSION OF CO 1/AGAMOUS-LIKE 20 (SOC1/AGL20) ortholog in rice. Plant J. 38, 754–764. doi: 10.1111/j.1365-313X.2004.02082.x
Li, X., Duan, X., Jiang, H., Sun, Y., Tang, Y., Yuan, Z., et al. (2006). Genome-wide analysis of basic/helix-loop-helix transcription factor family in rice and arabidopsis. Plant Physiol. 141, 1167–1184. doi: 10.1104/pp.106.080580
Li, C., Wu, X., Wang, P., Wang, H., Wang, L., Sun, F., et al. (2025). Genome-wide association study of image-based trait reveals the genetic architecture of dark-induced leaf senescence in rice. J. Exp. Bot. 76, 331–345. doi: 10.1093/jxb/erae391
Li, X., Yan, W., Agrama, H., Hu, B., Jia, L., Jia, M., et al. (2010). Genotypic and phenotypic characterization of genetic differentiation and diversity in the USDA rice mini-core collection. Genetica 138, 1221–1230. doi: 10.1007/s10709-010-9521-5
Li, J., Zhang, Z., Chong, K., and Xu, Y. (2022). Chilling tolerance in rice: Past and present. J. Plant Physiol. 268, 153576. doi: 10.1016/j.jplph.2021.153576
Liu, Q., Luo, L., Wang, X., Shen, Z., and Zheng, L. (2017). Comprehensive analysis of rice laccase gene (OsLAC) family and ectopic expression of OsLAC10 enhances tolerance to copper stress in Arabidopsis. Int. J. Mol. Sci. 18, 209. doi: 10.3390/ijms18020209
Liu, C., Ou, S., Mao, B., Tang, J., Wang, W., Wang, H., et al. (2018). Early selection of bZIP73 facilitated adaptation of japonica rice to cold climates. Nature Communications 9, 3302. doi: 10.1038/s41467-018-05753-w"10.1038/s41467-018-05753-w
Liu, Y., Zeng, H., Shang, Y., Zhang, H., Jiao, T., Yin, L., et al. (2025). Functional characterization of OsCSN1 in the agronomic trait control of rice seedlings under far-red light. Int. J. Mol. Sci. 26, 522. doi: 10.3390/ijms26020522
Lorieux, M. (2012). MapDisto: fast and efficient computation of genetic linkage maps. Mol. Breed. 30, 1231–1235. doi: 10.1007/s11032-012-9706-y
Lourenço, T., Sapeta, H., Figueiredo, D. D., Rodrigues, M., Cordeiro, A., Abreu, I. A., et al. (2013). Isolation and characterization of rice (Oryza sativa L.) E3-ubiquitin ligase OsHOS1 gene in the modulation of cold stress response. Plant Mol. Biol. 83, 351–363. doi: 10.1007/s11103-013-0092-6
Lu, G., Wu, F.-Q., Wu, W., Wang, H.-J., Zheng, X.-M., Zhang, Y., et al. (2014). Rice LTG1 is involved in adaptive growth and fitness under low ambient temperature. Plant J. 78, 468–480. doi: 10.1111/tpj.12487
Lv, Y., Guo, Z., Li, X., Ye, H., Li, X., and Xiong, L. (2015). New insights into the genetic basis of natural chilling and cold shock tolerance in rice by genome-wide association analysis. Plant Cell Environ. 39, 556–570. doi: 10.1111/pce.12635
Mackill, D. J. and Lei, X. (1997). Genetic variation for traits related to temperate adaptation of rice cultivars. Crop Sci. 37, 1340–1346. doi: 10.2135/cropsci1997.0011183X003700040051x
Meng, L., Li, H., Zhang, L., and Wang, J. (2015). QTL IciMapping: Integrated software for genetic linkage map construction and quantitative trait locus mapping in biparental populations. Crop J. 3, 269–283. doi: 10.1016/j.cj.2015.01.001
Meng, X., Wang, Y., Tang, B., Zhou, J., Gu, Y., Shen, Q., et al. (2025). A comprehensive analysis of the alternative splicing co-factor U2AF65B gene family reveals its role in stress responses and root development. Int. J. Mol. Sci. 26, 3901. doi: 10.3390/ijms26083901
Miura, K., Ikeda, M., Matsubara, A., Song, X.-J., Ito, M., Asano, K., et al. (2010). OsSPL14 promotes panicle branching and higher grain productivity in rice. Nat. Genet. 42, 545–549. doi: 10.1038/ng.592
Miyao, A., Iwasaki, Y., Kitano, H., Itoh, J.-I., Maekawa, M., Murata, K., et al. (2007). A large-scale collection of phenotypic data describing an insertional mutant population to facilitate functional analysis of rice genes. Plant Mol. Biol. 63, 625–635. doi: 10.1007/s11103-006-9118-7
Morales, K. Y., Singh, N., Perez, F. A., Ignacio, J. C., Thapa, R., Arbelaez, J. D., et al. (2020). An improved 7K SNP array, the C7AIR, provides a wealth of validated SNP markers for rice breeding and genetics studies. PloS One 15, e0232479. doi: 10.1371/journal.pone.0232479
Nagata, K., Sasaki, R., and Ohdaira, Y. (2013). Cultivar differences in the grain cracking of rice under the high air temperature conditions during grain filling. Japanese J. Crop Sci. 82, 42–48. doi: 10.1626/jcs.82.42
Nischal, L., Mohsin, M., Khan, I., Kardam, H., Wadhwa, A., Abrol, Y. P., et al. (2012). Identification and comparative analysis of microRNAs associated with low-N tolerance in rice genotypes. PloS One 7, e50261. doi: 10.1371/journal.pone.0050261
Ouyang, S., Zhu, W., Hamilton, J., Lin, H., Campbell, M., Childs, K., et al. (2006). The TIGR Rice Genome Annotation Resource: improvements and new features. Nucleic Acids Research 35, D883–D887. doi: 10.1093/nar/gkl976
Palaniswamy, R., Kambale, R., Mohanavel, V., Rajagopalan, V. R., Manickam, S., and Muthurajan, R. (2024). Identifying molecular targets for modulating carotenoid accumulation in rice grains. Biochem. Biophysics Rep. 40, 101815. doi: 10.1016/j.bbrep.2024.101815
Panahabadi, R., Ahmadikhah, A., Farrokhi, N., and Bagheri, N. (2022). Genome-wide association study (GWAS) of germination and post-germination related seedling traits in rice. Euphytica 218, 112. doi: 10.1007/s10681-022-03069-x
Phan, H. and Schläppi, M. (2021). Low temperature antioxidant activity QTL associate with genomic regions involved in physiological cold stress tolerance responses in rice (Oryza sativa L.). Genes 12, 1700. doi: 10.3390/genes12111700
Phan, H. and Schläppi, M. (2025). The RAD6-like ubiquitin conjugase gene OsUBC7 has a positive role in the early cold stress tolerance response of rice. Genes 16, 66. doi: 10.3390/genes16010066
Que, Z., Lu, Q., Li, Q., and Shen, C. (2023). The rice annexin gene OsAnn5 is involved in cold stress tolerance at the seedling stage. Plant Direct 7, e539. doi: 10.1002/pld3.539
Sakai, H., Lee, S. S., Tanaka, T., Numa, H., Kim, J., Kawahara, Y., et al. (2013). Rice Annotation Project Database (RAP-DB): An integrative and interactive database for rice genomics. Plant Cell Physiol. 54, e6–e6. doi: 10.1093/pcp/pcs183
Sasaki, A., Ashikari, M., Ueguchi-Tanaka, M., Itoh, H., Nishimura, A., Swapan, D., et al. (2002). A mutant gibberellin-synthesis gene in rice. Nature 416, 701–702. doi: 10.1038/416701a
Sakai, M., Sakamoto, T., Saito, T., Matsuoka, M., Tanaka, H., and Kobayashi, M. (2003). Expression of novel rice gibberellin 2-oxidase gene is under homeostatic regulation by biologically active gibberellins. J. Plant Res. 116, 161–164. doi: 10.1007/s10265-003-0080-z
Schläppi, M. R., Jackson, A. K., Eizenga, G. C., Wang, A., Chu, C., Shi, Y., et al. (2017). Assessment of five chilling tolerance traits and GWAS mapping in rice using the USDA mini-core collection. Front. Plant Sci. 8. doi: 10.3389/fpls.2017.00957
Schläppi, M. R., Jessel, A. R., Jackson, A. K., Phan, H., Jia, M. H., Edwards, J. D., et al. (2023). Navigating rice seedling cold resilience: QTL mapping in two inbred line populations and the search for genes. Front. Plant Sci. 14. doi: 10.3389/fpls.2023.1303651
Shad, M. A., Wang, Y., Zhang, H., Zhai, S., Shalmani, A., and Li, Y. (2023). Genetic analysis of GEFs and GDIs in rice reveals the roles of OsGEF5, OsGDI1, and OsGEF3 in the regulation of grain size and plant height. Crop J. 11, 345–360. doi: 10.1016/j.cj.2022.07.018
Shakiba, E., Edwards, J. D., Jodari, F., Duke, S. E., Baldo, A. M., Korniliev, P., et al. (2017). Genetic architecture of cold tolerance in rice (Oryza sativa) determined through high resolution genome-wide analysis. PloS One 12, e0172133. doi: 10.1371/journal.pone.0172133
Shimoyama, N., Johnson, M., Beaumont, A., and Schläppi, M. (2020). Multiple cold tolerance trait phenotyping reveals shared quantitative trait loci in Oryza sativa. Rice 13, 57. doi: 10.1186/s12284-020-00414-3
Singh, A., Kanwar, P., Yadav, A. K., Mishra, M., Jha, S. K., Baranwal, V., et al. (2014). Genome-wide expressional and functional analysis of calcium transport elements during abiotic stress and development in rice. FEBS J. 281, 894–915. doi: 10.1111/febs.12656
Spielmeyer, W., Ellis, M. H., and Chandler, P. M. (2002). Semidwarf (sd-1), “green revolution” rice, contains a defective gibberellin 20-oxidase gene. Proc. Natl. Acad. Sci. 99, 9043–9048. doi: 10.1073/pnas.132266399
Sun, M., Jia, B., Yang, J., Cui, N., Zhu, Y., and Sun, X. (2017). Genome-wide identification of the PHD-finger family genes and their responses to environmental stresses in Oryza sativa L. Int. J. Mol. Sci. 18, 2005. doi: 10.3390/ijms18092005
Sun, W., Shi, J., Hong, J., Zhao, G., Wang, W., Zhang, D., et al. (2022). Natural variation and underlying genetic loci of γ-oryzanol in Asian cultivated rice seeds. Plant Genome 15, e20201. doi: 10.1002/tpg2.20201
Sun, L., Zhang, Q., Wu, J., Zhang, L., Jiao, X., Zhang, S., et al. (2014). Two rice authentic histidine phosphotransfer proteins, OsAHP1 and OsAHP2, mediate cytokinin signaling and stress responses in rice. Plant Physiol. 165, 335–345. doi: 10.1104/pp.113.232629
Sweeney, M. and McCouch, S. (2007). The complex history of the domestication of rice. Ann. Bot. 100, 951–957. doi: 10.1093/aob/mcm128
Tang, W., Sun, J., Liu, J., Liu, F., Yan, J., Gou, X., et al. (2014). RNAi-directed downregulation of betaine aldehyde dehydrogenase 1 (OsBADH1) results in decreased stress tolerance and increased oxidative markers without affecting glycine betaine biosynthesis in rice (Oryza sativa). Plant Mol. Biol. 86, 443–454. doi: 10.1007/s11103-014-0239-0
Tripathi, A. K., Singh, K., Pareek, A., and Singla-Pareek, S. L. (2015). Histone chaperones in Arabidopsis and rice: genome-wide identification, phylogeny, architecture and transcriptional regulation. BMC Plant Biol. 15, 42. doi: 10.1186/s12870-015-0414-8
Voorrips, R. E. (2002). MapChart: Software for the graphical presentation oflinkage maps and QTLs. J. Heredity 93, 77–78. doi: 10.1093/jhered/93.1.77
Wang, F., Jing, W., and Zhang, W. (2014). The mitogen-activated protein kinase cascade MKK1–MPK4 mediates salt signaling in rice. Plant Sci. 227, 181–189. doi: 10.1016/j.plantsci.2014.08.007
Wang, H., Lee, A.-R., Park, S.-Y., Jin, S.-H., Lee, J., Ham, T.-H., et al. (2018). Genome-wide association study reveals candidate genes related to low temperature tolerance in rice (Oryza sativa) during germination. 3 Biotech. 8, 235. doi: 10.1007/s13205-018-1252-9
Wang, H., Xu, X., Vieira, F. G., Xiao, Y., Li, Z., Wang, J., et al. (2016). The power of inbreeding: NGS-based GWAS of rice reveals convergent evolution during rice domestication. Mol. Plant 9, 975–985. doi: 10.1016/j.molp.2016.04.018
Wang, J., Yu, H., Xiong, G., Lu, Z., Jiao, Y., Meng, X., et al. (2017). Tissue-specific ubiquitination by IPA1 INTERACTING PROTEIN1 modulates IPA1 protein levels to regulate plant architecture in rice. Plant Cell 29, 697–707. doi: 10.1105/tpc.16.00879
Wei, X., Qiu, J., Yong, K., Fan, J., Zhang, Q., Hua, H., et al. (2021). A quantitative genomics map of rice provides genetic insights and guides breeding. Nat. Genet. 53, 243–253. doi: 10.1038/s41588-020-00769-9
Wu, J., Cao, X., Sun, X., Chen, Y., Zhang, P., Li, Y., et al. (2025). OsEL2 regulates rice cold tolerance by MAPK signaling pathway and ethylene signaling pathway. Int. J. Mol. Sci. 26, 1633. doi: 10.3390/ijms26041633
Xu, C., Wang, Y., Yu, Y., Duan, J., Liao, Z., Xiong, G., et al. (2012). Degradation of MONOCULM 1 by APC/CTAD1 regulates rice tillering. Nat. Commun. 3, 750. doi: 10.1038/ncomms1743
Yachun, Y., Juan, L., Hao, L., Zuntao, X., Ruiying, Q., Wenge, W., et al. (2021). SDF5 encoding P450 protein is required for internode elongation in rice. Rice Sci. 28, 313–316. doi: 10.1016/j.rsci.2021.05.001
Yamamoto, T., Kuboki, Y., Lin, S. Y., Sasaki, T., and Yano, M. (1998). Fine mapping of quantitative trait loci Hd-1, Hd-2 and Hd-3, controlling heading date of rice, as single Mendelian factors. Theor. Appl. Genet. 97, 37–44. doi: 10.1007/s001220050864
Yamazaki, Y., Sakaniwa, S., Tsuchiya, R., Nonomura, K.-I., and Kurata, N. (2010). Oryzabase: an integrated information resource for rice science. Breed. Sci. 60, 544–548. doi: 10.1270/jsbbs.60.544
Yano, M., Harushima, Y., Nagamura, Y., Kurata, N., Minobe, Y., and Sasaki, T. (1997). Identification of quantitative trait loci controlling heading date in rice using a high-density linkage map. Theor. Appl. Genet. 95, 1025–1032. doi: 10.1007/s001220050658
You, C., Dai, X., Li, X., Wang, L., Chen, G., Xiao, J., et al. (2010). Molecular characterization, expression pattern, and functional analysis of the OsIRL gene family encoding intracellular Ras-group-related LRR proteins in rice. Plant Mol. Biol. 74, 617–629. doi: 10.1007/s11103-010-9704-6
Yu, Y., Zhang, H., Long, Y., Shu, Y., and Zhai, J. (2022). Plant Public RNA-seq Database: a comprehensive online database for expression analysis of ~45 000 plant public RNA-Seq libraries. Plant Biotechnol. J. 20, 806–808. doi: 10.1111/pbi.13798
Zhang, X.-Y., Li, S., Opachaloemphan, C., Zhang, C.-X., He, S. Y., and Jiang, Y. (2025). Augment rice ANNEXIN expression to counter planthopper NlAnnexin-like5 as an anti-virulence strategy against a major crop pest. bioRxiv 2025, 8.18.670976. doi: 10.1101/2025.08.18.670976
Zhao, H., Yao, W., Ouyang, Y., Yang, W., Wang, G., Lian, X., et al. (2015). RiceVarMap: a comprehensive database of rice genomic variations. Nucleic Acids Research 43, D1018–D1022. doi: 10.1093/nar/gku894
Zhang, Y., Xiao, W., Luo, L., Pang, J., Rong, W., and He, C. (2012). Downregulation of OsPK1, a cytosolic pyruvate kinase, by T-DNA insertion causes dwarfism and panicle enclosure in rice. Planta 235, 25–38. doi: 10.1007/s00425-011-1471-3
Keywords: aus rice, chilling tolerance, heading date, Oryza sativa, rice, plant height, QTL mapping, tropical japonica rice
Citation: Eizenga GC, Santamaria Y, Jackson AK, Phan H, Jia MH, Grunden QP-H, Edwards JD, Himelblau E and Schläppi MR (2026) Divergent genetic architecture of cold stress tolerance in aus and tropical japonica rice. Front. Plant Sci. 16:1716845. doi: 10.3389/fpls.2025.1716845
Received: 01 October 2025; Accepted: 24 November 2025; Revised: 21 November 2025;
Published: 07 January 2026.
Edited by:
Francis Chuks Ogbonnaya, Grains Research and Development Corporation, AustraliaReviewed by:
Tian Qing Zheng, Chinese Academy of Agricultural Sciences, ChinaOlive Onyemaobi, Commonwealth Scientific and Industrial Research Organisation (CSIRO), Australia
Lv Yonggang, Yunnan Academy of Agricultural Sciences, China
Copyright © 2026 At least a portion of this work is authored by Georgia C. Eizenga, Aaron K. Jackson, Melissa H. Jia, Quynh P. - H. Grunden and Jeremy D. Edwards on behalf of the U.S. Government and as regards Dr. Eizenga, Mr. Jackson, Ms. Jia, Ms. Grunden, Dr. Edwards and the U.S. Government, is not subject to copyright protection in the United States. Foreign and other copyrights may apply. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Georgia C. Eizenga, Z2VvcmdpYS5laXplbmdhQGFycy51c2RhLmdvdg==; Michael R. Schläppi, bWljaGFlbC5zY2hsYXBwaUBtYXJxdWV0dGUuZWR1
†ORCID: Georgia C. Eizenga, orcid.org/0000-0003-0330-9225
Aaron K. Jackson, orcid.org/0000-0002-5540-9255
Melissa H. Jia, orcid.org/0000-0003-0484-0608
Jeremy D. Edwards, orcid.org/0000-0002-0237-0932
Ed Himelblau, orcid.org/0000-0002-8045-2415
Michael R. Schläppi, orcid.org/0000-0003-1657-3252