Molecular mapping and characterization of QTLs for grain quality traits in a RIL population of US rice under high nighttime temperature stress

doi:10.21203/rs.3.rs-1832853/v1

Download PDF

Article

Molecular mapping and characterization of QTLs for grain quality traits in a RIL population of US rice under high nighttime temperature stress

https://doi.org/10.21203/rs.3.rs-1832853/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 25 Mar, 2023

Read the published version in Scientific Reports →

You are reading this latest preprint version

Elevated nighttime temperatures resulting from climate change significantly impact the rice crop worldwide. The rice (Oryza sativa L.) plant is highly sensitive to high nighttime temperature (HNT) during grain-filling (reproductive stage). HNT stress negatively affects grain quality traits and has a major impact on the value of the harvested rice crop. In addition, along with grain dimensions determining rice grain market classes, the grain appearance and quality traits determine the rice grain market value. During the last few years, there has been a major concern for rice growers and the rice industry over the prevalence of rice grains opacity and the reduction of grain dimensions affected by HNT stress. Hence, the improvement of heat-stress tolerance to maintain grain quality of the rice crop under HNT stress will bolster future rice value in the market. In this study, 185 F₁₁-recombinant inbred lines (RILs) derived from two US rice cultivars, Cypress (HNT-tolerant) and LaGrue (HNT-sensitive) were screened for the grain quality traits grain length (GL), grain width (GW), and percent chalkiness (%chalk) under control and HNT stress conditions and evaluated to identify the genomic regions associated with the grain quality traits. In total, there were 15 QTLs identified; 6 QTLs represented under control condition explaining 3.33–8.27% of the phenotypic variation, with additive effects ranging from − 0.99 to 0.0267 on six chromosomes and 9 QTLs represented under HNT stress elucidating 6.39 to 51.53% of the phenotypic variation, with additive effects ranging from − 8.8 to 0.028 on nine chromosomes for GL, GW, and % chalk. These 15 QTLs were further characterized and scanned for natural genetic variation in a japonica diversity panel (JDP) to identify candidate genes for GL, GW, and %chalk. We found 6,160 high impact single nucleotide polymorphisms (SNPs) characterized as such depending on their type, region, functional class, position, and proximity to the gene and/or gene features, and 149 differentially expressed genes (DEGs) in the 51Mbp genomic region comprising of the 15 QTLs. Out of which, 11 potential candidate genes showed high impact SNP associations. Therefore, the analysis of the mapped QTLs and their genetic dissection in the US grown japonica rice genotypes at genomic and transcriptomic levels provide deep insights into genetic variation beneficial to rice breeders and geneticists for understanding the mechanisms related to grain quality under heat stress in rice.

Rice (Oryza sativa L.), a major staple food, provides 35–60% of the dietary calorie intake for an estimated three billion people in the world^1–3 and plays a vital role in food security and sustainability in Asia⁴, Americas, and Africa⁵. To meet the demands of the increasing world’s population, rice breeders and geneticists have been working on developing high yielding and improved grain quality rice varieties⁶. In recent years, rice growers and consumers have been paying attention to grain quality, along with grain yield in rice ever since climate change was recognized as a serious threat to the stability and quality of the rice crop. Based on recent statistics of Intergovernmental Panel on Climate Change (IPCC), the mean temperature of the global air surface has been increased from 1.0 to 3.7^oC in the recent years, which would potentially increase the frequency and magnitude of heat stress events^{7, 8}. Under these critical conditions, climate change due to global warming has increased both the nighttime temperature and daytime temperature, however greater warming has been observed at night than during the day in rice growing regions worldwide^6,9. High nighttime temperature (HNT) is one of the detrimental factors deteriorating grain quality year-by-year and declining the grain yield of the rice crop⁶. Several studies have shown that a rise of 1^oC in nighttime temperature reduced rice grain yield by 10% and grain quality⁶. Rice is highly sensitive to HNT stress at various phenological growth stages; however, grain filling at the reproductive stage is the most sensitive, leading to poor grain quality under greenhouse and field conditions^10–14. In particular, HNT stress during grain filing stage increases respiration and oxidative stress in cells causing reduced flux of carbohydrates from the flag leaf to developing seeds and directly affects the grain quality¹⁵. A rise in nighttime temperatures has been shown to have a negative impact on grain quality and grain yield in the major rice growing areas of Asia⁶ and the United States¹⁶. Thus, rice breeders and geneticists need to focus on grain quality characteristics in order to improve overall grain quality under HNT stress using novel genomics-assisted techniques.

Rice grain quality is a complex trait with various components, including grain appearance, milling, and eating, cooking, and nutritional qualities¹⁷. The appearance quality of rice grain is primarily determined by grain dimensions, a combination of grain length (GL), grain width (GW), and grain thickness (GT)^18,19, endosperm opacity (chalkiness), chalky grain percentage, chalky area, and chalky degree (percentage chalkiness-% chalk) ^{20, 21}. The chalky grains have opaque spots either in the endosperm ranging in size, on the dorsal side of the grain (white belly) or in the center (white corner)²². Moreover, several studies have reported that HNT stress affects physio-chemical attributes such as increased chalk formation²³, reduced amylose content resulting smaller grain size in rice crop^{24, 25}, variation in flour protein ingredients²⁶, altered the enzyme activities^{11, 27}, and poor starch packing resulting increased chalkiness in developing grain²⁸. When rice grains are exposed to HNT stress (≥ 27 ⁰C/80.6 ⁰F at nighttime), 90.2% grains of japonica rice cultivar “Koshikihari” during 3–35 days after flowering showed milky-white and white-back²⁹. The HNT stress causes impairment of starch biosynthesis and triggers inhomogeneous starch deposition resulting in loosely packed irregular and smaller starch granules in rice grains^29–32. Recently, Shi et al³³ reported that HNT stress declined the head rice yield and grain width, altered the amylose content while it increased the chalkiness in several rice genotypes. Amylose content is lower under HNT stress (16.1%) compared to control nighttime temperature conditions (19.1%), suggesting that lower activity of amylose synthesis might be involved in chalk formation²⁹. The grain length, grain width, and chalkiness in rice grains determine overall grain quality and price. The genetic basis of these grain quality traits is poorly understood under heat stress conditions and the important goal lies in improving these complex correlated traits using advanced genetic tools/techniques on multi-omics datasets such as genome, transcriptome, metabolome, and proteome for genomics-assisted breeding.

Grain quality traits (GL, GW and % chalk) are polygenic quantitative traits³⁴. Recent revolutionary genomic advances in QTL mapping in bi-parental recombinant inbred lines (RILs) and genome-wide association studies (GWAS) on diverse populations³⁵ can resolve genetic architecture of complex traits. Up until now, more than 400 QTLs for grain size have been mapped across 12 chromosomes of rice of which 109, 95, and 107 QTLs were associated with GL, GW, and 100-grain weight, respectively³⁶. A few of these QTLs have been fine-mapped³⁷ and a total of 14 major genes are shown to be significantly involved in controlling grain size and shape including four independent genes (GW2, GW8, GS5, and GW5/qSW5) for GW, eight genes (GS2/GL2, GL3.1/qGL3, GL4, GLW7, GS3, OsLG3, OsLG3b/qLGY3, and TGW6) for GL, and two pleiotropic genes (GW6a and GL7/GW7) controlling both GL and GW that also have larger impact on grain weight^38–40. In last several years, many QTLs for grain appearance and eating quality have been identified in different mapping populations (RILs, DHs, CSSLs, F₂, and BC₂F₂) ⁴¹. Over the past two decades, several QTLs have been identified and are associated with grain dimension, across the chromosomes of rice⁴⁰. More than one hundred QTLs have been related to the chalkiness trait, including 30 QTLs for percent grain with white core (PGWC), 26 QTLs for degree of endosperm chalkiness (DEC: %chalk), 12 QTLs for area of endosperm chalkiness (AEC), 11 QTLs for white-backed kernel, and 3 QTLs for basal white (BW), all distributed across12 chromosomes and are catalogued in the Gramene database^34,40,42. A larger number of the QTLs associated with chalkiness were detected on chromosomes 5, 6 and 8 in different genetic backgrounds and environments⁴³. The QTL clusters were found on the hotspot regions of these three chromosomes^44–59. Even though HNT stress significantly deteriorates rice grain quality, a limited number of studies have been conducted to dissect the HNT stress tolerance in rice, and as a result very few QTLs/loci related to chalkiness under heat stress have been identified worldwide^60–64. However, until now, no study has been reported on QTLs mapping for GL and GW under HNT stress in bi-parental populations in rice. Despite the increasing knowledge of QTL mapping, metabolomics⁶⁵, and transcriptomic^{66, 67} responses to HNT stress in rice, the understanding and dissection of genetic basis of grain quality traits and HNT stress tolerance in rice is still limited. Hence, based on our best knowledge and recent survey of the literature, in the United States, this is the first study to identify genetic loci and putative candidate genes controlling grain quality traits and heat stress tolerance mechanisms in tropical japonica rice under HNT stress. In this study, we investigated a recombinant inbred line (RIL) population of two US rice cultivars “Cypress” and “LaGrue” for grain quality traits under HNT stress. The study was carried out with the objectives: to screen and evaluate the RIL populations with both parents for grain quality traits (GL, GW, and % chalk) and their correlations under HNT, identify the QTLs with SNP markers by using Inclusive Composite Interval mapping (IM-Add and IM-EPI) method, validate the QTL-linked markers with previously reported QTLs, and characterize the QTL regions to identify the candidate genes for grain quality traits with HNT stress tolerance using genomic and transcriptome analysis. This novel study of QTL mapping and characterization of these QTLs at genome and transcriptome levels will be useful to provide insights on allele mining, identification of potential candidate genes related to grain quality and heat stress tolerance mechanisms in US growing japonica rice cultivars, which would be beneficial to rice breeders and geneticists for developing good grain quality and heat tolerant rice cultivars in the US.

Plant material and growth conditions. The MY2 population of 185 F₅-derived bulk RILs (F₁₂) developed from a cross between two released U.S. long grain rice varieties, “Cypress and LaGrue”, was developed as part of the RiceCap project ⁶⁸. Cypress is tolerant to heat stress and good grain quality (low chalky) under HNT stress while LaGrue is highly sensitive to heat stress and poor grain quality (high chalky) under HNT stress^11,69. The MY2 RIL population was originally obtained from the USDA ARS Dale Bumpers National Rice Research Center, Stuttgart, AR, United States and each line was purified using single plant (F₁₀) lines and further re-derived from single panicle in the 2016 field trial at H. Rouse Cafey Rice Research Station, Louisiana State University (LSU) Agricultural Center, Rayne, LA 70578, USA. The permissions from the USDA and LSU Agricultural Center, LA, USA were obtained to collect and use the MY2 RIL population for this study.

A sample of ~ 30 seeds from each RIL along with both parents was germinated in single plastic pots, of size 15 cm × 15 cm, filled with a mixture of the SunGro professional potting mix (Sun Gro Horticulture Distribution, Agawam, MA, United States) and field soil, and grown in the greenhouse at the Harry R. Rosen Alternative Pest Control Center at the University of Arkansas, Fayetteville, AR, United States⁷⁰. At 10 days after germination, equal-sized seedlings of each RIL and both parents were transplanted in 3-gallon-plastic pots filled with a mixture of potting mix and field soil. The RILs and parents were then grown in the greenhouse until booting stage (R2). For greenhouse conditions, the temperature was set to 30 ± 1°C (86 ± 1°F) during the day and 22.2 ± 1°C (72 ± 1°F) at night⁷¹. The light was set to a light/dark 13/11-h cycle with maximum photo-synthetically active radiant (800-1,000 µmolPAR m^− 2 s^− 1) light and 60–65% relative humidity (RH) for the growth of the rice plants. The experimental design was a completely randomized design (CRD) with three replications (each replication is one plant in the pot). The plants were checked for water every day and fertilized with the Peter Professional soluble fertilizer (Allentown, PA, United States) containing chelated iron once a week for full vegetative growth. Plant protection methods were applied to prevent insect pests and diseases, which followed the Rosen Center (University of Arkansas, Fayetteville, AR, USA) greenhouse standard procedures.

Phenotypes and high nighttime temperature (HNT) stress treatment. At booting stage (R2), as described by Counce et al⁷², three major panicles per plant were tagged in each RIL and both parents. The plants with tagged panicles were then moved to the greenhouse with the HNT stress treatment that was maintained at a day/night temperature of 30°C (86°F)/28°C (82.4°F) for 10 h (20:00–6:00) while the control treatment was set at a day/night temperature of 30°C (86°F)/22.2°C (72°F) until harvest maturity (approximately 18–20% grain moisture content). The HOBO data loggers/sensors (Onset HOBO® data logger, Cape Cod, MA, United States) were installed in both greenhouses (Control and HNT stress treatments) for the continuous monitoring and recording of the day and nighttime temperatures until physiological maturity (Fig. 1). The data logger system was operated by the HOBOware® Pro software/app with compatible devices. At harvest maturity, all the panicles (under Control and HNT stress treatments) of each RIL and both parents were collected separately in individual brown bags, air-dried (12–14% grain moisture content), and used for the phenotyping of grain quality traits such as GL, GW, and %chalk. For grain quality estimation, rough rice of each plant of the RIL and both parents was de-hulled using a manually operated dehuller (Rice Husker TR120, China). The GL, GW, and %chalk were estimated in 100-brown/dehulled rice grain samples twice in each replicate of the RIL and parents, using an image analysis system WinSeedle™ Pro 2020a (Regent Instruments Inc., Sainte-Foy, Quebec, Canada) and expressed as percent of affected grains in the projected area.

Statistical analysis and phenotypic evaluations. The statistical analysis of the phenotype data was performed using R statistical packages v4.1.0 and John’s Macintosh Project (JMP) Genomics Pro version 12.0 for descriptive statistics. The test for normal distribution and homogeneity of variances was done using the Shapiro–Wilk test and Brown–Forsythe test, respectively. An analysis of variance (ANOVA) was carried out with a statistical model that included the effects of RIL, treatment (control and HNT stress), and the interaction between RIL and treatment. The Tukey’s Honest Significant Difference (HSD) test was used to compare the means of treatments among all the RILs for significant effects (Tukey’s HSD, p < 0.05) using the hsd function in R packages and JMP version 12, as Tukey’s HSD can determine slight differences between the means. Pearson’s correlation coefficient was carried out between grain quality traits under control and HNT stress conditions using the R package ‘corrplot’. The broad-sense heritability (H²) was estimated to describe how the environment affected GL, GW and %chalk in the RIL population using the lmer4 function in R v4.1.0^{73, 74}. The confint function⁷³ was used to compute the standard errors of the variance estimates provided by lmer, and these were then proliferated to use the 95% confidence intervals for the H².

Genotyping and construction of genetic map. The MY2 RIL population with both parents was genotyped with 1K-Agriplex platform on single plant re-derived from single panicle in the field trial. In addition, allele-specific primer (KASP) data was generated to genotype the RIL population to be more informative genetically. For genotyping of the MY2 RIL population, DNA was extracted from the young leaf tissue of each individual plant of each RIL using a modified CTAB method ⁷⁵, and KASP assays were designed by LGC genomics⁷⁶. The SNPs, monomorphic, heterozygous, and more than 30% missing data, were removed and a set of 1178 (1042 SNPs + 136 KASP) was used for further genetic analysis: QTL analysis and construction of genetic and physical linkage maps. The genetic map was constructed using 1178 SNP markers with MAP functionality of QTL IciMapping v4.2 software⁷⁷. Map has three basic steps: grouping, ordering, and rippling. Grouping of binned markers was performed using logarithms of the odds (LOD) threshold value, resulting in assembling 12 linkage groups representing the 12 chromosomes of rice. The RECORD (Recombination Counting and ORDering) algorithm was performed to order 1178 SNP markers within the linkage groups over 12 chromosomes. Recombination frequencies between markers were converted into centiMorgan (cM) by using the Kosambi mapping function. The physical map was drawn with 1178 SNP markers with their physical positions (extracted from IRGSPv1.0: http://rice.uga.edu/) using MapChart 2.0 software⁷⁸.

Quantitative trait loci (QTL) analysis. QTL analysis was performed using IciMapping v4.2 software⁷⁷ with the BIP functionality (https://www.isbreeding.net/) in the MY2 RIL population for grain quality traits. The significant LOD threshold to declare the significant QTLs was calculated using 1000 permutations and Type I error rate (P < 0.05). To detect the putative additive and epistatic QTLs, the Inclusive Composite Interval Mapping (ICIM-ADD) and Epistatic QTL (ICIM-EPI) functions were performed, respectively as mapping methods. The mapping parameters were 1.0 cM scanning steps and a probability of 0.01 in stepwise regression for additive QTL identification. For the detection of epistasis, scanning steps of 5.0 cM, a probability of 0.0001 in stepwise regression, and a LOD threshold of 4.0 or more were used to declare significant epistasis. QTLs explaining ≥ 10.0% phenotypic variance with ≥ 2.5 LOD were declared significant. The MapChart v2.0 software⁷⁸ was used for graphical representation of QTLs on the rice chromosomes.

Allele mining for the QTL regions in a japonica diversity panel. The raw FASTQ reads from the 83 genotypes (Supplementary Table S3) comprising the japonica diversity panel (JDP) of the USDA mini-core rice collection (URMC) were mapped to the reference rice genome cv. Nipponbare (IRGSP 1.0), using the Burrows-Wheeler Aligner (BWA), and the bam files were used to call SNPs using GATK⁷⁹. Approximately 6.1 million high quality SNPs were retained (less than 2% missing rate and more than 5% minor allele frequency) and annotated. Of these, the SNPs present in 15 QTLs (51 Mb genomic region across nine chromosomes) were selected and analyzed. This SNP dataset was further filtered to retain the SNPs that were polymorphic between both parents “Cypress and LaGrue”. The chromosome wide distribution of the various genomic features including QTLs and SNPs was visualized in circos⁸⁰.

Characterization of the QTL regions and Identification of candidate genes and their functions using transcriptome and genomic analyses. We identified candidate genes involved in GL, GW and % chalk under HNT by integrating QTLs and the expression QTL (eQTL) analyses of both parents “Cypress and LaGrue”. RNA-Seq data from caryopsis tissues at R6 stage (milky endosperm) under control and HNT stress conditions with replicates were used for differential gene expression analyses using the DESeq2 R package⁸¹. The differentially expressed (DE) genes were investigated using the GxE model on the ratio Cypress-HNT vs. Cypress-Control / Lagrue-HNT vs. Lagrue-Control to understand the effect of HNT stress in a genotype-dependent manner⁸². Genes with threshold value of log₂ fold change ≥ 1 and padj < 0.05 were assigned as DE and used for subsequent analyses. All the experiments on rice plants were performed in accordance with relevant guidelines and regulations.

Phenotypic variation and trait correlation. To evaluate the effects of HNT stress compared to control conditions (Fig. 1), the MY2 RIL populations with both parents, “Cypress and LaGrue” were analyzed and showed a wide range of phenotypic variation in gain quality traits such as GL, GW, and %chalk. Frequency distribution and histograms of all the grain quality traits indicate distribution among the RIL populations under control and HNT stress conditions. Values of each grain quality trait among the RIL populations extended beyond both parents show transgressive variation under control and HNT stress conditions (Fig. 2A-C, Table-1), indicating that alleles with positive and negative effects were derived from both parents. A wide range of phenotypic variation for all the grain quality traits under control and HNT stress conditions are summarized in Table 1 and indicates, cumulative action of many genes, quantitative inheritance. The extended variation, explained by the percent genetic variation (PGV), was observed in the RIL populations, showing 29.51%, 32.84%, and 215.04% genetic variation in GL, GW, and % chalk under HNT stress, respectively, while 27.41%, 28.85%, and 122.28% genetic variation in GL, GW, and % chalk under control treatment, respectively (Table 1). The parents, Cypress and LaGrue, were significantly different for GL, GW, and % chalk under HNT stress as LaGrue shows higher reduction in GL, GW and increase in % chalk, showing HNT stress sensitivity compared to Cypress that shows HNT stress tolerance under HNT stress conditions (Fig. 3A-C). High broad sense heritability (H²), ranging from 0.587 to 0.897, was observed for GL, GW, and %chalk in the RIL populations under control and HNT stress treatments (Table 1). Under HNT stress, the RIL populations show higher H² (0.702–0.897) compared to control condition, ranging from 0.587 to 0.657, for GL, GW, and %chalk.

Table 1

Descriptive statistics of phenotypic data of grain quality traits of 185 recombinant inbred lines (RILs) with both parents (Cypress and LaGrue) evaluated under control and high nighttime temperature (HNT) stress conditions. ^aTreat: treatment conditions (Control and HNT stress), ^bGL (mm): grain length in millimeter (mm), ^cGW (mm): grain width in millimeter (mm), and ^d%chalk: percent chalkiness.
Descriptive Statistics	Genotype	Treat^a	GL (mm)^b	GW (mm)^c	%Chalk^d
Mean	Cypress	Control	6.49	2.01	3.14
	Cypress	HNT	6.37	1.9	9.36
	LaGrue	Control	6.28	2.07	5.41
	LaGrue	HNT	5.92	1.94	32.09
	RILs	Control	6.64	2.08	5.16
	RILs	HNT	6.33	2.01	21.34
Range	RILs	Control	1.82	0.6	6.3
Range	RILs	HNT	1.86	0.702	45.89
Standard Deviation	Cypress	Control	0.082	0.023	0.82
	Cypress	HNT	0.057	0.077	1.09
	LaGrue	Control	0.276	0.015	0.84
	LaGrue	HNT	0.071	0.018	3.472
	RILs	Control	0.152	0.046	0.84
	RILs	HNT	0.128	0.048	2.17
Variance	Cypress	Control	0.006	0.0005	0.67
	Cypress	HNT	0.003	0.005	1.19
	LaGrue	Control	0.076	0.0002	0.7
	LaGrue	HNT	0.005	0.0003	1.02
	RILs	Control	0.041	0.003	1.05
	RILs	HNT	0.031	0.004	1.25
Coefficient of Variation (%)	Cypress	Control	1.27	0.77	26.25
	Cypress	HNT	0.9	0.95	11.68
	LaGrue	Control	4.39	1.13	15.54
	LaGrue	HNT	1.21	3.96	10.81
	RILs	Control	2.3	2.25	16.26
	RILs	HNT	2.03	2.42	10.19
Percent Genetic Variation (PGV)	RILs	Control	27.41	28.85	122.28
Percent Genetic Variation (PGV)	RILs	HNT	29.51	32.84	215.04
Skewness	RILs	Control	0.215	-0.077	-0.227
Skewness	RILs	HNT	0.153	-0.522	0.9545
Kurtosis	RILs	Control	0.338	0.533	-0.6841
Kurtosis	RILs	HNT	-0.094	0.285	1.023
W-test	RILs	Control	0.9876	0.9915	0.9657
W-test	RILs	HNT	0.9851	0.977	0.9389
Broad Sense Heritability (H²)	RILs	Control	0.657	0.651	0.587
Broad Sense Heritability (H²)	RILs	HNT	0.762	0.702	0.897

The trait correlation analysis was performed to determine the relationship between %chalk, GL, and GW in the RIL populations under control and HNT stress conditions (Fig. 3D). The %chalk shows positive correlation with GL (r = 0.1) and GW (r = 0.05) under control condition while it exhibits significant positive correlation with GL (r = 0.17, p < 0.05) and negative correlation with GW (r=-0.19, p < 0.01) under HNT stress condition. The correlation of GL with GW under control condition is positive correlation (r = 0.08) while under HNTstress, it shows significant positive correlation (r = 0.25, p < 0.001).

Genetic Map Construction. The genetic map was constructed using 1178 polymorphic SNP markers containing a total length of 1897.87 cM with an average maker density of 1.6 markers per cM in the rice genome. The marker information and marker position (in cM) on each chromosome are shown in Supplementary Fig. S1. The number of SNPs per chromosome ranged from 54 SNPs on chromosome 10 to 157 SNPs on chromosome1. More than 80% SNP markers show a distance of 3.0cM or less between adjacent markers in the map. Several gaps more than 20 cM between markers were shown on different chromosomes with the largest gaps where recombination is less frequent. To fill these gaps, allele mining in the QTL regions would be helpful to make strong resolution on the map for further genetic analyses. The physical position of each SNP marker across the rice genome based on their genomic position is shown in Fig. 4A.

QTL mapping and QTL x QTL Interaction. A total of 15 additive QTLs associated with grain quality traits (GL, GW, and % chalk) showing major and minor effects were identified in 185-RIL populations under control and HNT stress conditions, using 1178 SNP markers and BIP functionality of CIM mapping (Table 2, Fig. 4A, Supplementary Fig. S2-3). Out of these 15 additive QTLs, a total of 6 QTLs, with additive effects ranging from − 0.99 to 0.0267 and individually explained by 3.33 to 8.27% of the phenotypic variation, were identified for GL, GW and % chalk, under control condition (Fig. 4A, Supplementary Fig.S2A-C), on six chromosomes (chr1, chr4, chr5, chr6, chr7, and chr9) while under HNT stress condition, 9 QTLs, with additive effects ranging from − 8.8 to 0.028 and individually explained by 6.39 to 51.53% of the phenotypic variation, were detected for GL, GW, and % chalk on six chromosomes (chr1, chr2, chr3, chr6, chr7, and chr10) in rice genome (Table 2, Fig. 4A, Supplementary Fig. 3A-C).The additive effect is explained as one-half of the difference between the average effects of parental alleles (Cypress and LaGrue). In the results, a negative additive effect indicates that the favorable allele was contributed by Cypress; whereas a positive additive effect suggests that LaGrue added the favorable allele. A QTL was considered as major effect when the phenotypic variation explained in percentage (PVE %) was more than 20%. The details information of all the QTLs under control and HNT stress conditions are shown in Table 2. On physical map, the average distance between left and right markers flanking QTLs was less than 1.0 Mb and had a range of 157,670 bp to 19,092,637 bp (Fig. 4A, Supplementary Fig.S2-3).

Table 2

Quantitative trait loci (QTLs) associated with grain quality traits (GL, GW, and % chalk) under control and high nighttime temperature (HNT) stress conditions in the MY2 RIL population derived from the cross of two U.S. rice cultivars (Cypress and LaGrue) using ICIM mapping. ^aTrt: treatment conditions (Control and HNT stress), ^bQTL: quantitative trait locus, ^cChr: chromosome in the rice genome, ^dPos(cM): genetic position in centiMorgan (cM) on the linkage map, ^eLOD: logarithm of the odds peak/score was set ≥ 2.5 as threshold for detecting significant QTL in the mapping, ^ePVE(%): total phenotypic variance explained by the QTL in percentage (%), ^gAdd: additive effect values indicate the direction of favorable allele from the parents, ^hFav.allele: Favorable allele showing the negative additive effect value that means the direction of favorable allele from Cypress, while showing positive additive effect value pointing the direction of the favorable allele from LaGrue.
Treat^a	QTL^b	Chr^c	Pos (cM)^d	Left Flaking Marker		Right Flaking Marker		LOD^e	PVE (%)^f	Add^g	Fav. Allele^h
Treat^a	QTL^b	Chr^c	Pos (cM)^d	Marker	Pos (cM)	Marker	Pos (cM)	LOD^e	PVE (%)^f	Add^g	Fav. Allele^h
Control	qGL-C-1	1	113	SNP0129_129	110	SNP0130_130	114	4.78	8.274	-0.11	Cypress
	qGL-C-4	4	102	SNP1179_1179	101.5	SNP386_4	102.5	3.51	5.469	-0.082	Cypress
	qGL-C-7	7	73	SNP405_7	67.5	SNP827_7	87.5	2.61	4.841	-0.077	Cypress
	qGW-C-5	5	19	SNP399_5	17.5	SNP0511_511	21.5	3.21	7.634	0.0267	LaGrue
	qGW-C-6	6	152	SNP0638_638	149.5	SNP333_6	154	3.25	7.441	0.0264	LaGrue
	q%Chalk-C-9	9	55	SNP1170_1170	33.5	SNP0829_829	68.5	3.32	3.325	-0.996	Cypress
HNT	qGL-HNT-1	1	107	SNP0118_118	101.5	SNP291_1	107.5	4.85	10.453	-0.12	Cypress
	qGL-HNT-2	2	104	SNP0212_212	97	SNP506_2	105.5	2.93	6.389	-0.081	Cypress
	qGW-HNT-1.1	1	86	SNP0108_108	83.5	SNP0109_109	89.5	3.79	9.2892	-0.033	Cypress
	qGW-HNT-1.2	1	88	SNP0109_109	87.73	SNP0112_112	89.18	3.56	8.0977	-0.031	Cypress
	qGW-HNT-3	3	135	SNP310_3	132.5	SNP0350_350	139.5	2.83	6.4213	0.028	LaGrue
	q%Chalk-HNT-1	1	12	SNP1203_1203	6.5	SNP99_1	16.5	4.33	49.874	-7.273	Cypress
	q%Chalk-HNT-6	6	98	SNP0603_603	73.29	SNP0640_640	107.5	4.86	51.47	-7.294	Cypress
	q%Chalk-HNT-7	7	12	SNP1163_1163	1.09	SNP0672_672	17.5	4.42	51.527	-7.259	Cypress
	q%Chalk-HNT-10	10	70	SNP0923_923	69.5	SNP212_10	72.5	2.61	28.775	-8.785	Cypress

Under control condition, three QTLs for GL, qGL-C-1, qGL-C-4, and qGL-C-7, were detected with LOD scores of 4.78, 3.51, and 2.61 and showed PVE (%) of 8.27, 5.469, and 4.84%, respectively. The size of qGL-C-1, qGL-C-4, and qGL-C-7 QTLs was 4.0cM, 1.0cM, and 20.0cM within flanking SNP makers, SNP0129_129 (38,652,270bp) & SNP0130-130 (39,278,883bp), SNP1179_1179 (29,114,708bp) & 29,299,845bp), and SNP405_7 (2,036,558bp) & SNP827_7 (2,721,936bp) on chromosomes 1, 4, and 7, respectively (Table 2, Fig. 4A, Supplementary Fig. 2A). For GW, two QTLs, qGW-C-5 and qGW-C-6, were identified with LOD scores of 3.21 and 3.25, and exhibited PVE (%) of 7.63% and 7.44%, respectively. The size of these QTLs, qGW-C-5 and qGW-C-6, was 4.0 cM and 4.5cM within flanking SNP markers, SNP399_5 (3,709,209 bp) & SNP0511_511 (4,129,551bp) and SNP0638_638 (28,004,251bp) & SNP333_6 (28,688,881bp) on chromosomes 5 and 6, respectively (Table 2, Fig. 4A, Supplementary Fig. 2B). For % chalk, the QTL, q%chalk-C-9, was detected with LOD score of 3.32 and showed PVE (%) of 3.33%. The size of the QTL was 35cM within flanking SNP markers, SNP1170_1170 (4,052,650bp) and SNP0829_829 (6,667, 280bp) on chromosome 9 (Table 2, Fig. 4A, Supplementary Fig. 2C). Under HNT stress conditions, for GL, two QTLs, qGL-HNT-1 and qGL-HNT-2, were identified with LOD scores of 4.85 and 2.93 and showed PVE (%) of 10.45% and 6.40%, respectively. These QTLs were sized of 6.0cM (1,936,656bp) and 8.5 cM (613,935bp) within flanking SNP markers, SNP0118_118 (36,107,132bp) & SNP291_1 (38,043,788bp) and SNP0212_212 (11,707,132bp) & SNP506_2 (12,321,173bp) on chromosomes 1 and 2 respectively (Table 2, Fig. 4A, Supplementary Fig. 3A). For GW, three QTLs, qGW-HNT-1.1, qGW-HNT-1.2, and qGW-HNT-3, were identified with LOD scores of 3.79, 3.56, and 2.83, and showed PVE (%) of 9.28%, 8.1%, and 6.5%, respectively. These three QTLs were sized of 6.0cM (402,983bp), 1.45cM (157,670bp), and 7.0cM (214,274bp) within flanking SNP markers, SNP108_108(34,274,861bp)&SNP109_109(34,677,844bp),SNP109_109(34,677,844b) &SNP0112_112(34,835,514bp),andSNP310_3(22,796,525bp)&SNP350_350(23,010,799bp) on chromosomes 1, 1, and 2, respectively (Table 2, Fig. 4A, Supplementary Fig. 3B). Four QTLs for %chalk, q%chalk-HNT-1, q%chalk-HNT-6, q%chalk-HNT-7, and q%chalk-HNT-10, were detected with LOD scores of 4.33, 4.86, 4.42, and 2.61 and explained phenotypic variation (%) of 49.87%, 51.47%, 51.53%, and 28.77%, respectively. The size of these four QTLs was 10.0cM (7,919,913bp), 34.2cM (19,092,637bp), 16.4cm (11,701,288bp) within flanking SNP markers, SNP1203_1203 (30,498,826bp) &SNP99_1 (38,418,739bp), SNP1163_1163(4,001,958bp) & SNP0672_672(15,703,246bp), and SNP0923_923(17,908,351bp) & SNP212-10(21,817,967bp) on chromosomes 1, 6, 7 and 10, respectively (Table 2, Fig. 4A, Supplementary Fig. 3C).

The polygenic behavior of grain quality traits (GL, GW, and %chalk) may identify a large number of QTLs with smaller effects and epistasis between distinguish loci in the genome. Several QTLxQTL interactions were detected between different loci for each grain quality trait under individual control and HNT stress conditions (Supplementary Table S1, Fig. 4B). To declare a significant epistatic interaction, QTL x QTL interaction with a threshold LOD > 4.0 was used ⁸³. Under control conditions, loci associated with %chalk on chr1 (30.49mb-38.49mb), chr2 (17.mb-18.97mb), chr4 (27.70mb-29.29mb), and chr5 (22.63mb-23.97mb) showed epistatic interactions with loci on chr3 (2.87mb-3.39mb) & chr4(30.15mb-33.15mb), chr11(16.41mb-17.36mb), and chr9 (4.06mb- 6.67mb), respectively exhibiting PVE range 0.51–2.37% with epistatic effect range − 8.23 to 5.34 (Supplementary Table S1, Fig. 4B). Loci associated with GL on chr3 (5.89mb-27.14mb) and chr7 (2.73mb-4.74mb) interacted with the loci (17.63mb-21.87mb and 6.27mb-9.21mb) on chr8 showing PVE 2.79% and 2.68% and epistatic effect 0.134 and 0.109, respectively. For GW, the epistatic interaction was observed between loci on chr9 (17.74mb-18.82mb) and the loci on chr10 (11.93mb-13.43mb) with PVE 1.85% and epistatic effect 0.036 (Supplementary Table S1, Fig. 4B). Under HNT stress conditions, for GL and GW, no interaction was detected. For % chalk, locus on chr1 (30.50mb-38.42mb) showed significant epistatic interaction with loci on chr2 (21.33mb-22.03mb), chr3 (7.35mb-31.57mb), chr4 (16.29mb-16.33mb), chr5 (6.61mb-14.64mb), chr6 (0.36mb-9.98mb), chr7 (18.63mb-25.95mb), chr8 (22.62mb-24.46mb), chr10 (13.72mb-16.83mb), chr11 (17.59mb-27.83mb), and chr12 (22.75mb-25.68mb) showing PVE range from 0.62–2.37% and epistatic effect range − 6.06 to 1.31. In addition, loci on chromosomes 2, 4, 5, 6, and 7 showed significant interactions with loci on several chromosomes showing PVE range 0.50–2.79% and epistatic effect range − 8.23 to 5.39 (Supplementary Table S1). The details information about QTLxQTL interaction is shown in Supplementary Table S1.

Co-localization of QTLs with previously reported QTLs. There has been no bi-parental mapping for grain quality traits in tropical japonica background under HNT stress reported. In order to validate the identified QTLs in this study, we investigated the co-localization of all the detected QTLs related to GL, GW, and % chalk under control and HNT stress conditions with 15 previously reported studies worldwide in japonica rice background^{84,46,21, 85–91,62,55,51,92,93}. The 15 independent publicly available mapping studies, published between 2001 to 2021, reported 29 major effect QTLs related to GL, GW, ratio of grain length and width (RLW), and chalkiness related components such as percent chalkiness (PC), degree of chalky endoderm (DEC), chalk rate (CR), percent grain white chalk (PGWC), white kernel (WK), white back (WB), back chalk (BCHK), percent white back (PWB), percent grain chalkiness (PGC), and chalky kernel (CK), were downloaded and used for extraction of the genome position of the QTLs for co-localization analysis (Table 3). The QTLs, qGL-C-1 under control conditions and qGL-HNT-1 under HNT stress on chr1 related to GL, were co-localized with previously reported major effect QTL GL1-4⁸⁴ in the genomic region between 36,107,132 to 39,278,883 bp in the genome. Under control conditions, the QTLs, qGW-C-5 on chr5 and qGW-C-6 on chr6 for GW, were co-localized with three previously reported major effect-QTLs (gGL5-3, qGW-5, and qGW5)^84,46,21,85 in the genomic region (3,709,209bp-4,129,551bp) on chr5 and two major QTLs (qGW6-7 and qRLW-6)^84,85 in the genomic region (28,004,251bp-28,688,881bp) on chr6, respectively. The QTL, q%chalk-C-9 for % chalk under control conditions, was coincided in the genomic region (4,052,650bp-6,667,280bp) with three well-known and major effect-QTLs, qPC9.1, qDEC-9, and qCR9, on chr9^{86, 46, 87}.

Table 3

Previously reported quantitative trait loci (QTLs) co-localized with the genomic regions of 15 QTLs related to grain quality traits (GL, GW, and % chalk) under control and high nighttime temperature (HNT) stress. ^aTrt: treatment conditions (Control and HNT stress), ^bQTL: 15 quantitative trait loci, ^cChr: chromosome in the rice genome.
Treat^a	QTL^b	Chr^c	Left Flaking Marker	Right Flaking Marker	QTL Interval (bp)	Co-localization with Pre Reported QTLs	Reference
Control	qGL-C-1	1	SNP0129_129	SNP0130_130	38,652,270 − 39,278,883	qGL1-4	⁸⁴
	qGL-C-4	4	SNP1179_1179	SNP386_4	29,114,708 − 29,299,845	-	⁻
	qGL-C-7	7	SNP405_7	SNP827_7	2,036,558-2,721,936	-	-
	qGW-C-5	5	SNP399_5	SNP0511_511	3,709,209-4,129,551	qGL5-3, qGW-5, qGW5	^{84, 46, 21, 85}
	qGW-C-6	6	SNP0638_638	SNP333_6	28,004,251 − 28,688,881	qGW6-7, qRLW-6	^{84, 85}
	q%Chalk-C-9	9	SNP1170_1170	SNP0829_829	4,052650-6,667,280	qPC9.1, qDEC-9, qCR9,	^{86, 46, 87}
HNT	qGL-HNT-1	1	SNP0118_118	SNP291_1	36,107,132 − 38,043,788	qGL1-4	⁸⁴
	qGL-HNT-2	2	SNP0212_212	SNP506_2	11,707,238 − 12,321,173	-	-
	qGW-HNT-1.1	1	SNP0108_108	SNP0109_109	34,274,861 − 34,677,844	q%Chalk-HNT-1	-
	qGW-HNT-1.2	1	SNP0109_109	SNP0112_112	34,677,844 − 34,835,514	q%Chalk-HNT-1	-
	qGW-HNT-3	3	SNP310_3	SNP0350_350	22,79,6525-23,010,799	qGL3, qGW3	⁸⁸
	q%Chalk-HNT-1	1	SNP1203_1203	SNP99_1	30,498,826 − 38,418,739	qGL-HNT-1,qGW-HNT-1.1, qGW-HNT-1.2, qPC1.1, qPGWC-1, qDEC-1b, qWK1-2, qBCHK1-3	^{86, 46, 89, 90}
	q%Chalk-HNT-6	6	SNP0603_603	SNP0640_640	10,049,864 − 29,142,501	qGW-C-6, qGW6-7, qPC6.1, qRLW-6, qGL-6, gw6, qWB6, qPGWC6.1, qPGWC6.2, qCR6	^{84, 86, 85, 88, 91, 62, 55}
	q%Chalk-HNT-7	7	SNP1163_1163	SNP0672_672	4,001,958 − 15,703,246	qPWB7.1	⁹²
	q%Chalk-HNT-10	10	SNP0923_923	SNP212_10	17,908,351 − 21,817,967	qPGC-10, qCR10, qCK10, qCK10h, qCD10h	^{87,51, 93}

Under HNT stress, the QTL, qGW-HNT-3 on chr3 for GW, was coincided with two previously reported major QTLs, qGL3 and qGW3 in the genomic region (22,796,525bp-23,010,799bp) on chr3 in rice genome⁸⁸. For % chalk under HNT stress, the QTLs, q%Chalk-HNT-1 on chr1, q%Chalk-HNT-6 on chr6, q%Chalk-HNT-7 on chr7, and q%Chalk-HNT-10 on chr10, were co-localized with prior reported 5 QTLs (qPC1.1, qPGWC-1, qDEC-1b, qWK1-2, and qBCHK1-3) in the genomic region (30,498,826bp -38,418,739bp) on chr1^86,47,89,90, 9 QTLs (qGW6-7, qPC6.1, qRLW-6, qGL-6, gw6, qWB6, qPGWC6.1, qPGWC6.2, qCR6) in the genomic region (10,049,864bp-29,142,501bp) on chr6^{84,86,85,88,91,62,55}, 1 QTL (qPWB7.1) in the genomic region (4,001,958bp-15,703,246bp) on chr7⁹², and 5 QTLs (qPGC-10 ,qCR10, qCK10, qCK10h, qCD10h) in the genomic region (17,908,351bp-21,817,967bp) on chr10^87,51,93,, respectively. The graphic representation of the identified QTLs co-localized with previously reported 29 QTLs in the rice genome is shown in Circos (Fig. 5) and Supplementary Fig. S4.

Allele mining for the 15 QTL regions in the japonica diversity panel. A set of 83 genotypes of the JDP was used to mine for polymorphic alleles in the 51 Mbp genomic sequence that contains the 15 QTLs using IRGSP 1.0 as reference genome. A total of 868,823 polymorphic alleles, excluding in/dels and other structural variants, were identified of which 131,629 alleles (15% of total alleles) differentiated both parents, “Cypress and LaGrue” from each other. Based on the functional class of the SNP effects, 55%, 42%, and 3% of these SNPs were classified as missense, silent, and nonsense, respectively. The missense: silent (Mi:Si) ratio for these variants in the genome is 1.3 and the transitions: transversions (Ts:Tv) ratio is 2.6.

According to the SNP annotations, 6,160 out of 868,823 SNPs (0.7% of total SNPs) were classified as high impact SNPs (HISs) depending on their type, region, functional class, position, and proximity to the gene and/or gene features. 865 HISs representing 14% of these HISs differentiate the parents, Cypress and LaGrue, further.

The distribution of features found in the QTL regions such as, source markers used to identify the QTLs, genes in QTLs, DEGs, and SNPs associated with each of these features were visualized across the chromosomes (Fig. 5). A summary of the SNPs identified by allele mining in a set of 83 genotypes of the JDP is presented in the Supplementary Table S2. Assuming the SNPs are uniformly distributed across the cumulative QTL length of 51 Mbp, there is one SNP present every 59 bp for all 83 genotypes of the JDP, and one SNP every 389 bp that is polymorphic between the two parents. Similarly, there is one high impact SNP (HIS) every 8.3 Kbp and one Cypress-LaGrue differentiating HIS every 59.1 Kbp.

Genomic and transcriptomic characterization of the QTL regions and identification of candidate genes for grain quality under HNT stress. A genomic scan of the 51Mbp region spanning the 15 QTLs was performed. Using the IRGSP 1.0 genome feature file (gff3) as a reference, we identified a total of 7,117 genes in this region, of which 495 genes were shared between the overlapping QTL regions (qGW-C-6 & q%Chalk-HNT-6 on chromosome 6, and qGL-HNT-1, qGW-HNT-1, & qChalk-HNT-1 on chromosome 1). The QTL-wise distribution and order of these genes were also determined.

To understand the effect of HNT stress in a genotype-dependent manner, differential gene expression analysis was performed on the genotype x environment (GxE) interaction of HNT stress treated and control replicates of the Cypress and LaGrue genotypes as described earlier. We aimed to identify candidate genes involved in GL, GW, and % chalk under HNT stress, by integrating QTL and expression QTL (eQTL) analyses of both parents. Of the 7,117 genes, 149 genes were either up- or down-regulated in response to HNT stress at log₂fold change ≥ 1 and padj < 0.05 as shown (Fig. 6). The genes showing significant differences for gene expression were assigned to their respective QTLs based on their genomic locations/co-ordinates. Of these 149 differentially expressed genes (DEGs), a subset of 35 genes (six for GL, one for GW, and 28 for % chalk) was identified based on the known function in starch biochemical pathways, stress-responsive pathways, and gene ontology (GO) based gene functions. The expression profiles and annotations of these genes are also shown (Supplementary Fig. S5). The subset comprises genes such as the Beta-galactosidase 4 (LOC_Os01g65460) which is downregulated in Cypress under HNT stress but highly expressed in LaGrue under HNT stress. Other examples of GO based genes included two loci with unknown gene functions, LOC_Os01g55160 and LOC_Os01g64330, that have been reported to be upregulated under endoplasmic reticulum (ER) stress during rice endosperm development and could cause aggregation of abnormal PB bodies that ultimately affect the grain quality⁹⁴.

To identify the potential candidate genes, we followed a two-tier approach based on differential gene expression and allele mining to find SNPs that overlap with the 149 DEGs showing significant changes in expression under control and HNT conditions. We found 5,507 SNPs overlapped with the 149 DEGs ± 10Kbp (14 SNPs or 0.25% of these were HISs that differentiate between the Cypress and LaGrue genotypes). This approach identified 11 candidate loci of interest (DEGs at log₂FC ≥ 1 and padj < 0.05) and high impact SNP associations with these loci based on the position, predictive cause, and effect of the SNP (Table 4).

Table 4

Candidate genes identified using a two-tier approach of overlapping differentially expressed genes (*log2FC > = 1 and padj < 0.05*) and allele mining in the 51Mbp region spanning the 15 QTLs. Transposon derived protein coding genes are highlighted. ^aHISs: high impact single nucleotide polymorphisms (SNPs).
Gene	Type	No. of HISs^a	Gene Annotation
LOC_Os06g25760	protein_coding	15	retrotransposon protein, putative, Ty3-gypsy subclass, expressed
LOC_Os06g42710	protein_coding	3	transposon protein, putative, Mutator sub-class, expressed
LOC_Os01g60050	protein_coding	2	HEAT, putative, expressed
LOC_Os01g64330	protein_coding	2	retrotransposon protein, putative, unclassified, expressed
LOC_Os07g16330	protein_coding	2	expressed protein
LOC_Os01g53560	protein_coding	1	phosphoesterase, putative, expressed
LOC_Os01g54300	protein_coding	1	OsMan02 - Endo-Beta-Mannanase, expressed
LOC_Os06g19980	protein_coding	1	MYB family transcription factor, putative, expressed
LOC_Os06g45360	protein_coding	1	peptidase, M24 family protein, putative, expressed
LOC_Os07g12650	protein_coding	1	ribosomal protein L7Ae, putative, expressed
LOC_Os07g16690	protein_coding	1	retrotransposon protein, putative, Ty3-gypsy subclass, expressed

HNT stress is one of the detrimental factors that mainly effects grain quality traits leading to poor grain quality in rice grain. Enhancing HNT stress tolerance with minimizing poor grain quality in US rice cultivars is one of the evolving targets in rice breeding. It has been determined that HNT stress (≥ 27^oC) during grain filing days in japonica rice triggers impairment of starch biosynthesis resulting into the increased chalkiness²⁹. The increase in chalkiness alters the amylose content declining grain shape and size in rice³³. To improve the grain quality under HNT stress, it is beneficial to identify favorable alleles/loci linked with candidate genes related to grain quality traits using phenomics to genomics-assisted approaches. Thus, these functional alleles or loci would be useful for functional SNPs-based marker-assisted selection in the rice-breeding program. Here, we have carried out phenotyping of the MY2 RIL population with both parents “Cypress and LaGrue” for grain quality traits under HNT stress and control conditions (Fig. 1), which allowed us to map and identify the genomic loci linked with grain quality traits determining HNT stress response.

The broad range of phenotypic variation for grain quality traits (GL, GW, and % chalk) under HNT stress and control conditions in the MY2-RIL population indicates that GL, GW, and % chalk have quantitative inheritance that would be useful for QTL mapping analysis. We found that the parent Cypress showed least reduction in GL, GW, and smaller increase in % chalk while other parent LaGrue exhibited significantly higher reduction in GL, GW, and greater increase in % chalk under HNT stress compared to control conditions. The significant reduction in GL and GW is related to decline in carbohydrate availability and caused by high nighttime respiration or reduced carbon assimilates supply under HNT stress in rice⁹⁵. Thus, the activity of several enzymes involved in biosynthesis and deposition of starch in the endosperm, negatively impacts grain dimensions and grain appearance in rice⁹⁵ and other cereal crops⁹⁶. The investigation of genetic variation derived from the both parents in the MY2 RIL populations, under control and HNT stress conditions, showed the highest genetic variation in % chalk followed by GW and GL (Fig. 2A-C, Table-1). Our results suggest that there was transgressive variation in the RIL populations exhibiting larger phenotypic values of each quality trait compared to both parents under prolonged HNT stress. This indicates that positive and negative alleles or loci with epistasis or additive effects, derived from both parents, can function together to substantially affect the GL, GW, and % chalk⁹⁰. Several studies have shown that grain appearance is primarily determined by grain shape & size, a combination of grain length (GL), grain width (GW), and grain thickness (GT)^{18, 9}. During grain filling stage (R6 stage), air temperature in the environment plays an important role in generating unexpected variations in rice grain quality ¹¹. High broad sense heritability and relationship between grain traits under control and HNT stress conditions indicate strong trait stability in different environments and strong genetic control directing the selection of genotypes-based GL, GW, and % chalk for HNT stress tolerance, which can be used in rice breeding.

The physical appearance traits such as grain length, width, and chalkiness components determining grain quality in rice are important to consumers. Although grain appearance traits are heritable and we can easily make grains in grain size and shape though conventional breeding approaches, significant improvement in grain architecture such as chalkiness is difficult to achieve⁹⁷. The use of genomics-assisted breeding and the completion of rice genome sequencing have extremely facilitated the discovery and mapping of QTLs, and genes for rice grain appearance quality and dimensions determining grain quality traits. The grain quality traits including grain length, width, ratio of length & width, and chalkiness are the polygenic quantitative traits³⁴. To date, more than 200 QTLs for GL and GW³⁶, and more than 100 QTLs for chalkiness traits have identified in indica and japonica background of rice^34,40,42,43. However, until now, this is first study focusing on QTLs mapping and their characterization for GL and GW in the My2 RIL population derived two tropical japonica US rice cultivars “Cypress and LaGrue” under HNT stress. In the RIL populations, the additive effect is defined as 50% of the variation between the average effects of alleles coming from both parents. The additive effects suggest that the favorable alleles with negative effect was contributed by Cypress and favorable alleles with positive effect was come from LaGrue (Table 2). A QTL with more than 20% PVE was considered as a major effect QTL. Thus, among 9 additive effect QTLs for GL, GW, and %chalk under HNT stress, 4 QTLs related to % chalk are considered to be major effect-QTLs showing more than 20% phenotypic variance and these QTLs were mapped on chr1, ch6, ch7, and chr10. The higher phenotypic variation explained for most of the studied %chalk trait revealed major genes/QTLs to be responsible. To analyze the QTLxQTL interaction, surprisingly, there was no QTLXQTL interaction for GL and GW under HNT stress. However, for % chalk, locus on chr1 showed significant epistatic interaction with loci on chr2, chr4, chr5, ch6, chr7, chr8, chr10, chr11, and ch12 with PVE range from 0.62–2.37% and epistatic effect range from 0.50–2.79%. The epistatic interactions for % chalk in the genome suggest the genomic regions on chr1 play an important role in controlling % chalk. The presence of genetic effects of additive and additive epistasis plays a significant role in controlling complex traits such chalkiness and grain dimension traits.

Based on the physical positions of the SNP makers (IRGSP 1.0), identified QTLs related to GL, GW, and % chalk under control and HNT stress were investigated for co-localization with the previously reported QTLs for GL, GW, and chalkiness mapped in japonica background. The study maps grain quality traits in tropical japonica RIL populations under HNT stress reported specially in the US. To validate the strength of the identified 15 QTLs for GL, GW, and % chalk under control and HNT stress conditions in the present study, 29 major effects QTLs related to grain quality traits (GL, GW, ratio of length and width, chalkiness components) reported in 15 independent publicly available studies in last two decades were analyzed^{84,46,21, 85–91,62,55,51,92,93}. Among 15 QTLs, 12 QTLs were coincided with the genomic regions of these previously reported 29 QTLs (Table 2, Fig. 5, and Supplementary Fig. S4). Interestingly, qGL-C-1 for GL under control and qGL-HNT-1 under HNT stress conditions were highly co-localized with the genomic region of a major effect QTL GL1-4 related to grain length⁸⁴. In addition, qGW-C-5 and qGW-C-6 related to GW under control conditions were overlapped with the genomic regions of two QTLs, qGW-5, and qGW5, on chr5^86,87,88 and two major effects QTLs, qGW6-7, and qRLW-6, on chr6^84,85 for grain width, respectively. Under control conditions, q%chalk-C-9 was coincided with the genomic regions of three well-studied major effects QTLs, qPC9.1, qDEC-9, and qCR9, for chalkiness^86,46,87. qGW-HNT-3 for GW under HNT stress was mapped in the genomic region of well-characterized QTL, qGW3 for grain width in rice⁸⁸. For % chalk, q%Chalk-HNT-1 on chr1, q%Chalk-HNT-6 on chr6, and q%Chalk-HNT-7 on ch7, and q%Chalk-HNT-10 on chr10 were highly co-localized with the genomic regions of 5 QTLs (qPC1.1, qPGWC-1, qDEC-1b, qWK1-2, qBCHK1-3) on chr1^86,46,89,90, 9 QTLs (qGW6-7, qPC6, qRLW-6, qGL-6, gw6, qWB6, qPGWC6.1, qPGWC6.2, qCR6, 1) on chr6^{84,86,85,88,91,62,55}, and 1 QTL (qPWB7.1) on chr 7⁹², and 5 QTLs (qPGC-10 ,qCR10, qCK10, qCK10h, qCD10h) on chr10^{87, 51, 93}. These 12 QTLs for GL, GW, and % chalk in rice genome showed the strongest evidence of co-localization with other well-characterized and major effects previously reported QTLs, which suggests that these QTLs are consistently stable, and their genomic regions bear useful gene targets to understand the HNT stress tolerance mechanism. This information can be implemented in rice breeding using SNP-based marker-assisted selection and introgression of the co-localized genomic regions with these QTLs of grain quality traits into elite US rice cultivars in order to develop new rice cultivars. In addition, a follow-up search for the favorable alleles we found in other germplasms might enable the selection of the trait phenotypes for other novel elite alleles.

To characterize the 51Mbp of genomic region comprising15 QTLs for grain quality traits (GL, GW, and % chalk) in the present study, a set of 83 rice genotypes of the JDP was used to discover new favorable alleles using existing natural genetic variation. Among 868,823 polymorphic alleles, 131,269 alleles representing 15% of the total alleles are newly discovered alleles showing novel genomic variation between both parents, Cypress and LaGrue. Based on the SNP type, region, functional class, position, proximity to the gene or gene features, 6,160 newly discovered SNPs/alleles (out of 868,823 total) are high impact SNPs (HISs), of which 865 HISs representing 14% of the total HISs have shown novel genetic variation between both parents. The HISs identified in this study are possible candidates for their use as functional markers to differentiate the two US cultivars, and thus find useful in genomics-assisted breeding and introgression studies. The SNPs that are unable to differentiate between Cypress and LaGrue are also useful in the sense that they indicate the presence of natural genetic variation for GL, GW, and % chalk in the japonica genotypes. The detailed summary of these SNPs with the genomic region of 15 QTLs showing highly dense distribution of SNPs/alleles in 83 rice genotypes of the JDP is shown in Fig. 5 and Supplementary Table S2. These observations emphasize the power of this analysis as there was a 2,600-fold increase in SNP density over the initial set of markers that were used to identify the QTLs thus providing higher marker saturation (initial marker density = one marker every 155 Kb). Upon validation, these SNPs will provide a valuable resource for the future genomics-assisted breeding efforts for the traits of interest such as grain yield and grain quality.

A total of 7,117 genes were identified in the 51Mbp region comprising the 15 QTLs using IRGSP 1.0 genomic feature file (GFF) as a reference. Out of these, we identified 495 genes there either up- or down-regulated in response to HNT stress and showed significant variation for gene expression under control and HNT stress conditions in a genotype-dependent manner (Fig. 6). Narrowing down further, a subset of 35 genes associated with GL, GW, and % chalk under HNT stress were selected based on their known functions in starch biochemical pathways, stress-responsive pathways, and gene ontology (GO). (Supplementary Fig. S5). Two genes, LOC_Os01g55160 and LOC_Os_01g64330 with unknown functions, identified in this study, have previously been shown to be upregulated under ER stress during endosperm development in rice, which could be responsible for aggregation of abnormal protein bodies overall affecting the rice gran quality⁹⁴. Moreover, plant beta-galactosidases (B-gals) have been well characterized and are involved in degradation of structural pectin and xyloglucans in plant cell walls⁹⁸. The observed expression pattern of the Beta-galactosidase 4 gene (LOC_Os01g65460) in this study, which is downregulated in Cypress HNT under stress but highly expressed in LaGrue under HNT stress, therefore alludes to starch degradation in LaGrue under HNT stress as compared to Cypress.

Other genes identified in this study are thought to have potentially relevant roles such as in starch accumulation, endosperm formation, development, stress response, grain size, grain quality, chalkiness etc., some of which are described below in detail ^99–104. Five potential candidate genes identified in the genomic region of QTL, qChalk7HNT, belong to α-amylase/trypsin inhibitors upregulated in Cypress under HNT stress conditions. In addition, upregulation of OsDGAT1-2, LOC_Os06g36800, in LaGrue under HNT stress could be a cause of increased chalkiness because of accumulation of triacylglycerol (TAG) and lipid bodies, thereby hindering proper starch accumulation under HNT stress⁹⁹. The cyclin protein, LOC_Os01g59120, is 7X upregulated in Cypress under HNT stress vs LaGrue under HNT stress and plays a critical role in endosperm formation and maintenance of embryo development in rice¹⁰⁰. The upregulation of LOC_Os06g38790, an orthologue gene (ZmEBE-1) of maize, in rice may play a role under HNT stress conditions in Cypress under HNT stress vs LaGrue under HNT stress in early development of specialized domains of the endosperm¹⁰⁰. OsHAP3//nuclear factor-YB (NF-YB)/CCAAT binding factor-A (CBF-A) in rice regulates chloroplast plastid biogenesis¹⁰². OsLMS, LOC_Os01g63820, shares amino-acid similarity to the Arabidopsis FIERY2/CPL1 gene, known to control many plant stress responses and development¹⁰³. FERONIA-like receptor (FLR1) family members’ effect grain size & quality and stress tolerance and FLR2 and FLR8 negatively regulate grain size¹⁰⁴. Downregulation of FLR2 gene in Cypress under HNT stress suggests maintaining grain filling compared to LaGrue under HNT stress condition. The cumulative slight perturbation in the expression of multiple genes at multiple loci under HNT conditions could cause a QTL for a particular trait to be additive and epistatic. Similarly, the upregulation of a gene in non-photosynthetic tissue (endosperm) in Cypress under HNT stress vs LaGrue under HNT stress could have a potential role in amyloplastic formation.

To find the candidate genes for grain quality in US japonica rice, we followed a two-tier approach using differential gene expression and newly discovered alleles/SNPs overlapped with the 149 DEGs exhibiting significant changes in the expression under control and HNT stress identified 11 candidate genes with high impact SNP associations (Table 4). Of particular interest is the HEAT protein (LOC_Os01g60050), a microtubule organization protein that is essential for cortical microtubule organization in plant cells and is speculated to play an important role during cellurization¹⁰⁵. This gene is stress responsive and is 12x upregulated in Cypress (HNT tolerant parent) as compared to LaGrue (HNT sensitive parent) under HNT conditions. It is also worth noting that four out of 11 loci are transposon derived protein-coding genes that are marred with HISs and show differential gene expression under HNT conditions. For instance, the loci LOC_Os06g25760- a retrotransposon belonging to the Ty3-gypsy subclass contains 15 HISs. Such rapid rate of mutations accumulating SNPs at a higher rate in transposons as compared to genes is a meaningful observation from an evolutionary perspective of genome structure and function. Additionally, due to their presence in the transcriptome, these transposons are presumably active in the genome and are differentially expressed in a stress responsive manner. This makes them candidates of choice to study transposon-mediated regulation of genes involved in HNT stress response.

The power of allele mining and using the top-down approach to integrate the genomic and transcriptomic datasets has its takeaway. Of the possible 6.1 million SNPs distributed across the 12 chromosomes or 868,823 SNPs distributed across the 15 QTLs on nine chromosomes, we narrowed down to a subset of 30 High Impact SNPs for further testing (Table 4). Additionally, from a total of ~ 48K genes in the rice genome, we shortlisted a potential set of 11 protein coding genes including transposons for validation and characterization (Table 4). These genes are important because i. They reside in the QTL regions of interest, ii. They are differentially expressed between the parental genotypes “Cypress” and “LaGrue” in a stress responsive manner, and iii. They harbor the HISs and could be functionally relevant. Besides increasing the marker saturation of the initial set of 1178 SNP markers by 2,600-fold, we identified a set of polymorphic markers that distinguish between two parents, Cypress and LaGrue, that can be put to future use for introgression or other breeding strategies.

This is the first integrated study to screen the MY2 RIL population (185 lines) derived from two US rice cultivars, “Cypress and LaGrue”, for grain quality traits such as GL, GW, and % chalk at booting stage (R2: reproductive stage) under HNT stress, map & characterize the QTLs regions by mining for novel alleles and identify the candidate genes for grain quality under HNT stress. Phenotypic screens indicated that the MY2 RIL population with both parents (US rice cultivars) showed a wide range of variation for GL, GW, and % chalk under HNT stress compared to control conditions and the RIL population exhibited strong trait correlation and high broad sense heritability for grain quality traits under HNT stress. Thus, the phenotyping insights suggest that the MY2 RIL population exhibited transgressive segregation that would be useful for QTLs mapping and their characterization to find the candidate genes for GL, GW, and %chalk under HNT stress. Co-localization of 15 QTLs with the previously reported 29 major effects-QTLs for GL, GW, and % chalk validates the robustness of the favorable alleles linked with the QTLs and strengthens the QTL analysis in this study. Further characterization of the QTLs through allele mining and identification of genes that showed genotype dependent differential expression under HNT stress and control conditions led us to identify 11 candidate genes with high impact SNP associations based on the position, function, and predictive cause/effect of the SNPs. Therefore, all the findings from this study are noteworthy and the approach used demonstrates the power of integration of phenotypic, genomic, and transcriptomic data for gaining insights into genetic variations to understand the mechanisms responsible for heat stress tolerance that contribute to the grain quality in rice. These insights will be useful for developing allele-specific markers for genotyping US (japonica background) rice germplasms/elite breeding lines, optimizing genomic predictions for breeding program, establishing SNPs-assisted breeding strategies to breed superior varieties with improved grain quality and yield.

Data availability

Data generated in this study are available in the main tables, figures, and supplementary tables and figures.

Acknowledgements

We thank the USDA ARS Dale Bumpers National Rice Research Center, Stuttgart, AR and H. Rouse Cafey Rice Research Station, Louisiana State University Agricultural Center, LA, United States for providing the seeds of the MY2 RIL populations, Ms. Jessica Kivet for the maintenance of the well-controlled greenhouses, and Mr. Mohmmad Ibrahim Kazim, Mr. Mohit Shrestha, and Ms. Hale Wade for their assistance in greenhouse and phenotypic evaluation.

Author Contributions

AK and AP conceived the idea and designed the study. AK, JT, YD, and CR performed phenotypic screening in the greenhouses. AK, JT, and NG performed data curation and formal bioinformatics analysis. AK performed the phenotypic and genotypic data analysis with AF. AK drafted the manuscript. AF provided the RIL population, and SNP markers for the study. AP acquired the funding and supervised the research. All authors reviewed and approved the manuscript.

Funding

This study was supported by the National Science Foundation NSF-EPSCoR RII Track-2 award (#1826836): “Systems genetics studies on rice genomes for analysis of grain yield and quality under heat stress”. We also acknowledge the funding from the Arkansas Rice Research and Promotion Board for supporting and promoting an early work (2014-2019) and initiating this study.

Competing interests

The Authors declare no competing interests.

Fageria N. K. Yield physiology of rice. J. Plant Nutrit.30 843–879 (2007).
FAO. FAOSTAT Database. Rome: FAO (2009).
GRISP (ed.). Rice almanac, 4th Edn. Los Baòos: International Rice Research Institute (2013).
Wassmann, R. et al. Regional vulnerability of climate change impacts on Asian rice production and scope for adaptation. Advances in Agronomy vol. 102 91-133 (Academic Press, 2009).
Muthayya, S., Sugimoto, J. D., Montgomery, S. & Maberly, G. F. An overview of global rice production, supply, trade, and consumption: Global rice production, consumption, and trade. Ann. N. Y. Acad. Sci.1324, 7-14 (2014).
Peng, S., Huang, J., Sheehy, J.E., Laza, R.C., Visperas, R.M., Zhong, X., Centeno, G.S, Khush, G.S., Cassman, K.G. Rice yields decline with higher night temperature from global warming. Proc. Natl. Acad. Sc. USA101(27): 9971–9975 (2004). https:// doi.org/10.1073/pnas.0403720101
Intergovernmental Panel on Climate Change. Climate Change 2014 Mitigation of Climate Change: Working Group III Contribution to the Fifh Assessment Report of the Intergovernmental Panel on Climate Change. (Cambridge University Press, 2014). doi:https:// doi.org/10.1017/CBO9781107415416.
Intergovernmental Panel on Climate Change. IPCC. Global Warming of 1.5°C. An IPCC Special Report on the impacts of global warming of 1.5°C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty (2018).
Elagib, N. A. Trends in intra- and inter-annual temperature variabilities across Sudan. Ambio39 413–429 (2010). 10.1007/s13280-010-0042-3
Counce, P. A., Bryant, R. J., Bergman, C. J., Bautista, R. C., Wang, Y. J., Siebenmorgen, T. J., et al. Rice milling quality, grain dimensions, and starch branching as affected by high night temperatures. Cereal Chem.82:645–648 (2005). 10.1094/CC-82-0645
Cooper, N. T. W., Siebenmorgen, T. J., Counce, P. A. Effects of nighttime temperature during kernel development on rice physicochemical properties. Cereal Chem.85 276–282 (2008). 10.1094/cchem-85-3-0276
Jagadish, S. V. K., Murty, M. V. R., Quick, W. P. Rice responses to rising temperatures-challenges, perspectives, and future directions. Plant Cell Environ.381686–1698 (2015). 10.1111/pce.12430
Kumar, A., Thomas, J., Yingling, S., Dwiningsih, Y., Ramegowda, V., Gaspar, J., et al. Screening of diverse rice cultivars for heat tolerance and grain quality under high nighttime temperature. Fayetteville, AR: University of Arkansas, 61-67 (2017).
Kumar, A., Gupta, C., Thomas, J., & Pereira, A. Genetic dissection of grain yield component traits under high nighttime temperature stress in a rice diversity panel. Frontiers in plant science, 12, 712167 (2021). https://doi.org/10.3389/fpls.2021.712167
Park, J.R., Kim, E.G., Jang, Y.H. et al. Screening and identification of genes affecting grain quality and spikelet fertility during high-temperature treatment in grain filling stage of rice. BMC Plant Biol21, 263 (2021). https://doi.org/10.1186/s12870-021-03056-9
Lyman, N. B., Jagadish, K. S. V., Nalley, L. L., Dixon, B. L. & Siebenmorgen, T. Neglecting rice milling yield and quality underestimates economic losses from high-temperature stress. PLoS ONE 8, e72157 (2013).
Zheng, X., Wu, J.G., Lou, X.Y., Xu, H.M., Shi, C.H. The QTL analysis on maternal and endosperm genome and their environmental interactions for characters of cooking quality in rice (Oryza sativa L.). Theor Appl Genet, 116:335–342 (2008).
Tan, Y. F., Xing, Y. Z., Li, J. X., Yu, S. B., Xu, C. G., and Zhang, Q. F. Genetic bases of appearance quality of rice grains in Shanyou 63, an elite rice hybrid. Theor. Appl. Genet.101, 823–829 (2000). doi: 10.1007/s001220051549
Xing, Y., Tan, Y., Hua, J., Sun, X., Xu, C., and Zhang, Q. Characterization of the main effects, epistatic effects and their environmental interactions of QTLs on the genetic basis of yield traits in rice. Theor. Appl. Genet. 105, 248–257 (2002). doi: 10.1007/s00122-002-0952-y
Zhang, Q.F. Strategies for developing green super rice. Proc Natl Acad Sci USA, 104:16402–16409 (2007).
Wan, X.Y., Weng, J.F, Zhai, H.Q., Wang, J.K., Lei, C.L., Liu, X.L., Guo, T., Jiang, L., Su, N., Wan, J.M. Quantitative trait loci (QTL) analysis for rice grain width and fine mapping of an identified QTL allele gw-5 in a recombination hotspot region on Chromosome 5. Genetics, 179:2239-2252 (2008).
Khush, G.S., Paule, C.M., de la Cruz, N.M. Rice grain quality evaluation and improvement at IRRI. In. Proceedings of the Workshop on Chemical Aspects of Rice Grain Quality. pp. 21.31 (1979). (International Rice Research Institute: Los Baños, Philippines)
Ishimaru, T., Horigane, A.K., Ida, M., Iwasawa, N., San-oH, Y.A., Nakazono, M. et al. Formation of grain chalkiness and changes in water distribution in developing rice caryopses grown under high-temperature stress. Journal of Cereal Science, 50:166–174 (2009).
Yamakawa, H., Hirose, T., Kuroda, M., Yamaguchi, T. Comprehensive expression profiling of rice grain filling-related genes under high temperature using DNA microarray. Plant Physio, 2007; 144(1): 258–277 (2007). PMID: 17384160
Perrotta, C., Treglia, A.S., Mita, G., Giangrande, E., Rampino, P., Ronga, G. et al. Analysis of mRNAs from ripening wheat seeds: the effect of high temperature. Journal of Cereal Science, 27: 127–132 (1997).
Wardlaw, I.F., Blumenthal, C., Larroque, O., Wrigley, C.W. Contrasting effects of chronic heat stress and heat shock on kernel weight and flour quality in wheat. Functional Plant Biology, 29: 25-34 (2002).
Fitzgerald, M.A., Resurreccion, A.P. Maintaining the yield of edible rice in a warming world. Functional Plant Biology, 36: 1037–1045 (2009).
Ashida, K., Iida, S., Yasui, T. Morphological, physical, and chemical properties of grain and flour from chalky rice mutants. Cereal Chemistry, 86: 225–231 (2009).
Ishimaru, T., Parween, S., Saito,Y., Shigemitsu,T., Yamakawa, H., Nakazono, M., Masumura, T., Nishizawa, N.K., Kondo, M., Sreenivasulu, N. Laser Microdissection-Based Tissue-Specific Transcriptome Analysis Reveals a Novel Regulatory Network of Genes Involved in Heat-Induced Grain Chalk in Rice Endosperm, Plant and Cell Physiology, 60 (3): 626–642 (2019). https://doi.org/10.1093/pcp/pcy233
Shi, W., Yin, X., Struik, P.C., Soils, C., Xie, F., Schmidt, R.C., Huang, M., Zou, Y., Ye, C., Jagadish, S.V.K. High day and nighttime temperatures affect grain growth dynamics in contrasting rice genotypes. J. Exp. Bot.68, 5233-5245 (2017).
Kaneko, K., Sasaki, M., Kuribayashi, M., Suzuki, H., Sasuga, Y., Shirya, T., Inomata, T., Itoh, K., Baslam, M., Mitsui, T. Proteomic and glycomic characterization of rice chalky grains produced under moderate and high temperature conditions in field system. Rice, 9, 26 (2016).
Zhang, H., Xu, H., Feng, M., Zhu, Y. Suppression of OsMADS7 in rice endosperm stabilizes amylose content under high temperature stress. Plant Biotech. J., 16, 18-26 (2018).
Shi, W., Yin, X., Struik, P.C., Xie, F., Schmidt, R.C., Jagadish, S.V.K. Grain yield and quality responses of tropical hybrid rice to high nighttime temperature. Field Crops Research, 190: 18-25 (2016).
Sreenivasulu, N., Butardo, V.M., Mishra G., Cuevas, R.P., Anacleto, R., Kavi -Kishor, P.B. Designing climate-resilient rice with ideal grain quality suited for hightemperature stress. J Exp Bot66(7):1737–1748 (2015). https://doi.org/10.1093/jxb/ eru544
Gong, J., Miao, J., Zhao, Y., Zhao,Q., Feng, Q.,Zhan, Q.,Cheng, B., Xia, J., Huang, X.,Yang, S.,Han, B. Dissecting the Genetic Basis of Grain Shape and Chalkiness Traits in Hybrid Rice Using Multiple Collaborative Populations, Molecular Plant,10(10):1353-1356 (2017). https://doi.org/10.1016/j.molp.2017.07.014.
Huang, R., Jiang, L., Zheng, J., Wang, T., Wang, H., Huang, Y., et al. Genetic bases of rice grain shape: so many genes, so little known. Trends Plant Sci. 18, 218–226 (2013). doi: 10.1016/j.tplants.2012.11.001
Ponce, K., Zhang, Y, Guo, L., Leng, Y. and Ye, G. Genome-wide association study of grain size traits in indica rice multiparent advanced generation intercross (MAGIC) population. Front. Plant Sci. 11:395 (2020). doi: 10.3389/fpls.2020.00395
Dong, Q., Zhang, Z.H., Wang L.L. et al. Dissection and fine mapping of two QTL for grain size linked in a 460-kb region on chromosome 1 of rice. Rice, 11 (1): pp 44 (2018).
Li, N. and Li, Y. Signaling pathways of seed size control in plants, Current Opinion in Plant Biology, 33:23–32 (2016).
Bazrkar-Khatibani, L., Fakheri, B.A., Hosseini-Chaleshtori, M., Mahender, A., Mahdinejad, N., Ali, J.Genetic mapping and validation of quantitative traitl loci (QTL) for the grain appearance and quality traits in rice (Oryza sativa L.) by using recombinant inbred line (RIL) population. International Journal of Genomics, v2019, Article ID 3160275, 13 pages (2019). https://doi.org/10.1155/2019/3160275
Mahender, A., Anandan,A., Pradhan,S.K. and E. Pandit, E. Rice grain nutritional traits and their enhancement using relevant genes and QTLs through advanced approaches. Springerplus, 5(1): 2086 (2016).
ZeOFu, L.I., JianOMin, W.A.N., JiaOFa, X. I. A. and HuOQu Z. H. A. I. Mapping quantitative trait loci underlying appearance quality of rice grains (Oryza sativa L.). Rice Science, 30, 3 (2003).
Yang, W., Liang, J., Hao, Q. et al. Fine mapping of two grain chalkiness QTLs sensitive to high temperature in rice. Rice14, 33 (2021). https://doi.org/10.1186/s12284-021-00476-x
He, P., Li, S.G., Qian, Q., Ma, Y.Q., L,i J.Z., Wang, W.M., Chen, Y., Zhu, L.H.Genetic analysis of rice grain quality. Theor. Appl. Genet. 98 (3-4): 502–508 (1999). https://doi. org/10.1007/s001220051098
Tan, Y.F., Xing, Y.Z., Li, J.X., Yu, S.B., Xu, C.G., Zhang, Q. Genetic bases of appearance quality of rice grains in Shanyou 63, an elite rice hybrid. Theor. Appl. Genet.101(5-6):823–829 (2000). https://doi.org/10.1007/s001220051549
Wan, X.Y., Wan JM, Weng JF, Jiang L, Bi JC, Wang CM, Zhai HQ (2005) Stability of QTLs for rice grain dimension and endosperm chalkiness characteristics across eight environments. Theor Appl Genet 110(7):1334-1346. https://doi. org/10.1007/s00122-005-1976-x
Hao, W., Zhu M, Gao J, Sun S, Lin H (2009) Identification of quantitative trait loci for rice quality in a population of chromosome segment substitution lines. J IntegrPlant Biol 51(5):500–512. https://doi.org/10.1111/j.1744-7909.2009. 00822.x
Chen, H., Zhao Z, Jiang L, Wan X, Liu L, Wu X, Wan J (2011) Molecular genetic analysis on percentage of grains with chalkiness in rice (Oryza sativa L.). Afr J Biotechnol 10:6891–6903.
Guo, T., Liu, X., Wan, X., Weng, J., Liu, S., Liu, X., Chen, M., Li, J., Su, N., Wu, F., Cheng, Z., Guo, X., Lei, C., Wang, J., Jiang, L., Wan, J. Identification of a stable quantitative trait locus for percentage grains with white chalkiness in rice (Oryza sativa). J Integr Plant Biol 53(8):598–607 (2011). https://doi.org/10.1111/j.1744-7909.2011.01 041.x
Liu, X., Wan, X., Ma, X., Wan, J. Dissecting the genetic basis for the effect of rice chalkiness, amylose content, protein content, and rapid viscosity analyzer profile characteristics on the eating quality of cooked rice using the chromosome segment substitution line population across eight environments. Genome 54:64–80 (2011).
Liu, X., Wang, Y., Wang, S.W. QTL analysis of percentage of grains with chalkiness in japonica rice (Oryza sativa). Genet Mol Res 11(1):717-724 (2012). https://doi.org/10.4238/2012.March.22.1.
Li, Y., Fan, C., Xing, Y., Yun, P., Luo, L., Yan, B., Peng, B., Xie, W., Wang, G., Li, X., Xiao, J., Xu, C., He, Y. Chalk5 encodes a vacuolar H+ -translocating pyrophosphatase influencing grain chalkiness in rice. Nat Genet 46(4):398-404 (2014). https://doi. org/10.1038/ng.2923
Peng, B., Wang, L., Fan, C., Jiang, G., Luo, L., Li, Y., He, Y. Comparative mapping of chalkiness components in rice using five populations across two environments. BMC Genet 15(1):49 (2014). https://doi.org/10.1186/1471-2156-15-49.
Zhao, X., Zhou, L., Ponce, K., Ye, G. The usefulness of known genes/QTLs for grain quality traits in an indica population of diverse breeding lines tested using association analysis. Rice 8(1):29 (2015). https://doi.org/10.1186/s12284-015- 0064-3.
Chen, L., Gao, W., Chen, S., Wang, L., Zou, J., Liu, Y., Wang, H., Chen, Z., Guo, T. High-resolution QTL mapping for grain appearance traits and co-localization of chalkiness-associated differentially expressed candidate genes in rice. Rice 9(1):48 (2016). https://doi.org/10.1186/s12284-016-0121-6
Wang, X., Pang, Y., Wang, C., Chen, K., Zhu, Y., Shen, C., Ali, J., Xu, J., Li, Z. New candidate genes affecting rice grain appearance and milling quality detected by genome-wide and gene-based association analyses. Front Plant Sci 7:1998 (2017).
Zhu, A., Zhang ,Y., Zhang, Z., Wang, B., Xue, P., Cao, Y., Chen, Y., Li, Z., Liu, Q., Cheng, S., Cao, L. Genetic dissection of qPCG1 for a quantitative trait locus for percentage of chalky grain in rice (Oryza sativa L.). Front Plant Sci 9:1173 (2018).
Misra, G., Anacleto, R., Badoni, S., Butardo, V., Molina, L., Graner, A., Demont, M., Morell, M.K., Sreenivasulu, N. Dissecting the genome-wide genetic variants of milling and appearance quality traits in rice. J Exp Bot 70(19):5115–5130 (2019). https://doi.org/10.1093/jxb/erz256
Misra, G., Badoni, S., Parween, S., Singh, R.K., Leung, H., Ladejobi, O., Mott, R., Sreenivasulu, N. Genome-wide association coupled gene to gene interaction studies unveil novel epistatic targets among major effect loci impacting rice grain chalkiness. Plant Biotechnol J. (2020). https://doi.org/10.1111/ pbi.13516
Nevame, A.Y.M., Emon, R.M., Malek, M.A., Hasan, M.M., Alam, M.A., Muharam, F.M., Aslani, F., Rafii, M.Y., Ismail, M.R. Relationship between high temperature and formation of chalkiness and their effects on quality of rice. Biomed Res Int 1-18 (2018).
Kobayashi, A., Genliang, B., Shenghai, Y., Tomita, K. Detection of quantitative trait loci for white-back and basal-white kernels under high temperature stress in japonica rice varieties. Breed Sci 57(2):107–116 (2007). https://doi.org/10.12 70/jsbbs.57.107
Wada, T., Miyahara, K., Sonoda, J., Tsukaguchi, T., Miyazaki, M., Tsubone, M., Ando, T., Ebana, K., Yamamoto, T., Iwasawa, N., Umemoto, T., Kondo, M., Yano, M. Detection of QTLs for white-back and basal-white grains caused by high temperature during ripening period in japonica rice. Breed Sci 65(3):216-225 (2015). https://doi.org/10.1270/jsbbs.65.216
Miyahara, K., Wada, T., Sonoda, J., Tsukaguchi, T., Miyazaki, M., Tsubone, M., Yamaguchi, O., Ishibashi, M., Iwasawa, N., Umemoto, T., Kondo, M. Detection and validation of QTLs for milky-white grains caused by high temperature during the ripening period in Japonica rice. Breed Sci 67(4):333–339 (2017). https://doi. org/10.1270/jsbbs.16203
Bheemanahalli, R., Knight, M., Quinones, C. et al. Genome-wide association study and gene network analyses reveal potential candidate genes for high night temperature tolerance in rice. Sci Rep11, 6747 (2021). https://doi.org/10.1038/s41598-021-85921-z
Schaarschmidt, S. et al. Season afects yield and metabolic profles of rice (Oryza sativa) under high night temperature stress in the feld. Int. J. Mol. Sci.21, 3187 (2020).
Liao, J.-L. et al. Transcriptome changes in rice (Oryza sativa L.) in response to high night temperature stress at the early milky stage. BMC Genomics16, 18 (2015).
Desai, J. S. et al. Warm nights disrupt global transcriptional rhythms in field-grown rice panicles. https://doi.org/10.1101/702183 (2019).
Oard, J.H., Moldenhauer, K., Fjellstrom, B., Nelson, J.C., Scheffle,r B et al. Registration of the MY2 ‘Cypress’/‘LaGrue’ rice recombinant inbred line mapping population. J. Plant Regist4(3):261 (2010). https://doi.org/10.3198/jpr2009.11.0668crmp
Lanning, S.B., Siebenmorgen, T.J., Ambardekar, A.A., Counce, P.A., Bryant, R.J. Effects of nighttime air temperature during kernel development of field- grown rice on physiochemical and functional properties. Cereal Chem. 89(3):168–175 (2012) https://onlinelibrary.wiley.com/doi/epdf/10.1094/CCHEM-12-11-0146
Kumar, A. Development and Characterization of Rice Genotypes for Water Use Efficiency and Drought Resistance. Ph. D. thesis. Fayetteville, AR: University of Arkansas (2017).
Ghadirnezhad, R., Fallah, A. Temperature effect on yield and yield components of different rice cultivars in flowering stage. Int. J. Agron.20: 846707 (2014). 10.1155/2014/846707
Counce, P. A., Siebenmorgan, T. J., Ambardekar, A. A. Rice reproductive stage thermal time and calendar day intervals for six US cultivars in the Grand Prairie, Arkansas, over 4 years. Ann. App. Biol.003:4746 (2015).
Bates, D. M. lme4: Mixed-effects modeling with R. Draft. Geneva: R Core Team (2010).
Bates, D., Maechler, M., Bolker, B. M., Walker S. Fitting linear mixed-effects models using lme4. Geneva: R Core Team (2015).
Murray, M., Thompson, W. Rapid isolation of high molecular weight plant DNA. Nucleic Acids Res.8, 4321-4326 (1980).
LGC, Group. SNPline genotyping automation (2016).https://www.lgcgroup.com/products/genotyping instruments/snpline/#.XFCv91xKhaQ.
Meng, L., Li, H., Zhang, L., Wang, J. QTL IciMapping: integrated software for genetic linkage map construction and quantitative trait locus mapping in bi-parental populations. Crop J 3:269-283 (2015).
Voorrips, R.E. MapChart: software for graphical presentation of linkage maps and QTLs. J. Hered. 93: 77-78 (2002).
Li, H. and Durbin, R. Fast and accurate short read alignment with Burrows-Wheeler Transform. Bioinformatics, 25:1754-60 (2009).
Krzywinski, M., Schein, J.,Birol, I., Connors, J., Gascoyne, R., Horsman, D., Jones, S. and Marra, M. Circos: an Information Aesthetic for Comparative Genomics. Genome Res 19:1639-1645 (2009).
Love, M. I., Huber, W., and Anders, S. Moderated Estimation of Fold Change and Dispersion for RNA-Seq Data with DESeq2. Genome Biol. 15, 550. doi:10.1186/s13059-014-0550-8 (2014).
Gill, N. and Dhillon, B. RNA-seq data analysis for differential expression. Methods in Molecular Biology, Springer Nature doi: 10.1007/978-1-0716-1795-3_4 (2022).
Cao, Z., Guo, Y., Yang, Q., He, Y., Fetouh, MI., Warner, RM., Deng, Z. Genome-wide identification of quantitative trait loci for important plant and flower traits in petunia using a high-density linkage map and an interspecific recombinant inbred population derived from Petunia integrifoliaand P. axillaris. Hort Res6:27 (2019).
Nagata, K., Ando, T., Nonoue, Y., Mizubayashi, T., Kitazawa, N., Shomura, A., Matsubara, K., Ono, N., Mizobuchi, R., Shibaya, T., Ogiso-Tanaka, E., Hori, K., Yano, M., & Fukuoka, S. Advanced backcross QTL analysis reveals complicated genetic control of rice grain shape in a japonica × indica cross. Breeding science, 65(4): 308-318 (2015). https://doi.org/10.1270/jsbbs.65.308
Zhang, T., Wang, S., Sun, S. et al. Analysis of QTL for grain size in a rice chromosome segment substitution line Z1392 with long grains and fine mapping of qGL-6. Rice 13, 40 (2020). https://doi.org/10.1186/s12284-020-00399-z
Dwiningsih, Y., Kumar, A., Thomas, J., Ruiz, C., Alkahtani, J., Al-hashimi, A., Pereira, A. Identification of genomic regions controlling chalkiness and grain characteristics in a recombinant inbred line rice population based on high-throughput SNP markers. Genes 12, 1690 (2021). https://doi.org/ 10.3390/genes12111690
Yun, Y.T., Chung, C.T., Lee, Y.J. et al. QTL mapping of grain quality traits using introgression lines carrying Oryza rufipogon chromosome segments in japonica rice. Rice 9, 62 (2016). https://doi.org/10.1186/s12284-016-0135-0
Li, Z.F., Wan, J.M., Xia, J.F., and Zhai, H.Q. Mapping quantitative trait loci underlying appearance quality of rice grains (Oryza sativa L.). Yi Chuan Xue Bao 30, 251-259 (2003).
Tabata, M., Hirabayashi, H., Takeuchi, Y., Ando, I., Iida, Y., and Ohsawa, R. Mapping of quantitative trait loci for the occurrence of white-back kennels associated with high temperature during the ripening period of rice (Oryza Sativa L.). Breeding Science 57:47-52 (2007).
Edwards, J.D., Jackson, A.K. and McClung, A. Genetic architecture of grain chalk in rice and interactions with a low phytic acid locus. Field Crops Research 205:116-123 (2017). https://doi.org/10.1016/j.fcr.2017.01.015
Xing, Y.Z., Tan, Y.F., Xu, C.G., Hua, J.P., and Sun, X.L. Mapping quantitative trait loci for grain appearance traits of rice using a recombinant inbred line population. Acta Bot Sinica 43, 840-845 (2001).
Qin, Y., Kim, S. and Sohn, J. Genetic analysis and QTL mapping for grain chalkiness characteristics of brown rice (Oryza sativa L.). Genes & Genomics 31 (2): 155-164 (2009).
Dai, L., Wang, L., Leng, Y., Yang, Y., Huang, L., Chen, L., Wang, Y., Ren, D., Hu, J., Zhang, G., Zhu, L., guo, L., Qian, Q. and Zeng, D. Quantitative trait loci mapping for appearance quality in short-grain rice. Crop Sci. 56:1484-1492 (2016). doi: 10.2135/cropsci2015.07.0404
Oono, Y., Wakasa, Y., Hirose, S., Yang, L., Sakuta, C., Takaiwa, F. Analysis of ER stress in developing rice endosperm accumulating β-amyloid peptide. Plant Biotechnology Journal 8: 691–718 92010). doi: 10.1111/j.1467-7652.2010.00502.x
Bahuguna, R. N., Solis, C. A., Shi, W. & Jagadish, K. S. V. Post-flowering night respiration and altered sink activity account for high night temperature-induced grain yield and quality loss in rice (Oryza sativa L.). Physiol. Plant. 159, 59–73 (2017).
Impa, S. M. et al. High night temperature induced changes in grain starch metabolism alters starch, protein, and lipid accumulation in winter wheat. Plant Cell Environ. 43, 431-447 (2020).
Nan, J., Feng, X., Wang, C., Zhang, X., Wang, R., Liu, J., Yuan, Q., Jiang G. and Lin, S. Improving rice grain length through updating the GS3 locus of an elite variety Kongyu 131. Rice 11:21 (2018). doi: 10.1186/s12284-018-0217-2
Tanthanuch, W., Chantarangsee, M., Maneesan, J., Ketudat-Cairns, J. Genomic and expression analysis of glycosyl hydrolase family 35 genes from rice (Oryza sativa L.). BMC Plant Biol. 30 (8):84 (2009). doi: 10.1186/1471-2229-8-84.
Bhunia RK, Sinha K, Chawla K, Randhawa V, Sharma TR. Functional characterization of two type-1 diacylglycerol acyltransferase (DGAT1) genes from rice (Oryza sativa) embryo restoring the triacylglycerol accumulation in yeast. Plant Mol Biol.105(3):247-262 (2021). doi: 10.1007/s11103-020-01085-w
Guo, J., Wang, F., Song, J., Sun, W., Zhang, X.S. The expression of Orysa;CycB1;1 is essential for endosperm formation and causes embryo enlargement in rice. Planta 231(2):293-303 (2010). http://doi.org/10.1007/s00425-009-1051-y
Magnard, JL., Lehouque, G., Massonneau, A. et al. ZmEBE genes show a novel, continuous expression pattern in the central cell before fertilization and in specific domains of the resulting endosperm after fertilization. Plant Mol. Biol. 53, 821-836 (2003). https://doi.org/10.1023/B:PLAN.0000023672.37089.00
Miyoshi, K., Ito, Y., Serizawa, A., Kurata, N. OsHAP3 genes regulate chloroplast biogenesis in rice. Plant J. 36(4):532-40 (2003). doi: 10.1046/j.1365-313x.2003.01897.x
Xiong, L., Lee, H., Ishitani, M., Tanaka, Y., Stevenson, B., Kiowa, H., Bressan, A.R., Paul M. Hasegawa, P.M., and Zhu, J. Repression of stress-responsive genes by FIERY2, a novel transcriptional regulator in Arabidopsis. PNAS 99(16): 10899–10904 (2002). https://www.pnas.org/doi/pdf/10.1073/pnas.162111599
Wang, L., Wang, D., Yang, Z. et al. Roles of FERONIA-like receptor genes in regulating grain size and quality in rice. Sci. China Life Sci. 64, 294-310 (2021). https://doi.org/10.1007/s11427-020-1780-x
Xing MQ, Zhang YJ, Zhou SR, et al. Global Analysis Reveals the Crucial Roles of DNA Methylation during Rice Seed Development. Plant Physiol.168 (4):1417-1432 (2015). doi:10.1104/pp.15.00414

No competing interests reported.

SuppTablesFigures.pdf

Download PDF

Journal Publication

published 25 Mar, 2023

Read the published version in Scientific Reports →

Editorial decision: Major revision
11 Sep, 2022
Reviews received at journal
19 Aug, 2022
Reviews received at journal
12 Aug, 2022
Reviewers agreed at journal
05 Aug, 2022
Reviewers agreed at journal
01 Aug, 2022
Reviewers invited by journal
01 Aug, 2022
Editor assigned by journal
22 Jul, 2022
Editor invited by journal
18 Jul, 2022
Submission checks completed at journal
18 Jul, 2022
First submitted to journal
06 Jul, 2022

You are reading this latest preprint version

Molecular mapping and characterization of QTLs for grain quality traits in a RIL population of US rice under high nighttime temperature stress

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Materials And Methods

Results

Discussion

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1