Integrative genomic approaches to unravel genomic regions and candidate genes associated with flag leaf photosynthesis at the reproductive stage in rice.

Photosynthesis is a complex polygenic trait that directly influences crop yield and remains a challenging physiological trait to dissect genetically. This study is novel in its large-scale evaluation of photosynthetic performance across 181 diverse rice accessions, representing six subpopulations, under controlled conditions. By integrating photosynthetic measurements from the flag leaf at the booting stage with genome-wide association studies (GWASs) using 3.7 million high-quality single nucleotide polymorphism (SNP) markers, candidate genes were identified for photosynthesis. A total of 18 putative quantitative trait nucleotides (QTNs) [logarithmic odds (LOD) ≥ 8.0] were identified across the MLMM, FarmCPU, and BLINK models and are dispersed on all chromosomes except 5 and 10, including two QTNs (1-2050328 and 12-11380740) that were commonly identified in two models. Following QTN identification, gene mining analysis revealed 1,091 structural and regulatory genes in flanking regions. Subsequently, fine mapping using the gene haplotype analysis suggested 43 genes, including signaling/regulatory genes (receptor like kinases, F-box, RING, and auxin-responsive genes), gene regulators (histones, NAC/NAM, and PPR), membrane trafficking/transport (Exo70 and ADP-ribosylation factor), stress and defense components (heat shock protein, thionin, and MAC/perforin), and 19 uncharacterized proteins, of which seven genes were further selected as candidate gene using ortholog and regulatory analyses. The candidate genes may associate with photosynthesis, including carotenoid isomerase and various kinases, along with genes involved in stomatal regulation (OsABA4), sugar transport (sugar transporter 14 and UDP-glucose transporter), and leaf development (auxin-responsive genes), collectively contributing to efficient photosynthate production and assimilate translocation. Furthermore, the integration of genomic prediction analyses across GWAS, ridge regression best linear unbiased prediction (RRB), and bootstrap trees (BTS) models identified 44 common SNPs corresponding to these candidate regions, thereby enhancing the accuracy of genomic breeding value estimation across the population. To further explore the phenotypic response of photosynthesis based on subpopulation, five diverse rice accessions were selected for a detailed study of light response (A/Q), carbon dioxide response (A/Ci), and non-photochemical quenching (NPQ), enabling the assessment of photosynthetic variation across diverse genetic backgrounds. This study identified significant putative genomic regions, candidate genes, and a set of SNP markers associated with high photosynthesis in rice accessions, providing valuable resources for plant breeding and genomics-assisted breeding to enhance rice yield.