JingLi Fan, Qianwen Xue, Jiayi Li, YunQi Liu, Fantao Liu, Zuchun Li, JingXuan Wang, YanMin Xu, Hengrui Zhang, ZiNing Zhu, Min Zhang, ChunHuan Chen, TingDong Li, XinLun Liu, Xue Shi, ChangYou Wang, Jixin Zhao, WanQuan Ji, Pingchuan Deng
Wheat grain development determines yield and quality but involves complex spatial organization across genetically distinct maternal and filial tissues. Dissecting the regulatory architecture of this composite organ requires approaches that resolve both cellular identity and spatial context. Here, we integrate single-nucleus RNA sequencing with spatial transcriptomics to construct a high-resolution spatiotemporal atlas of wheat grain development at the milk stage (25 days after flowering, 25 DAF). This integrated framework resolves 13 transcriptionally and spatially distinct cell populations spanning maternal tissues, endosperm, aleurone layer, and embryo. By combining cell-resolved and in situ expression profiles, we uncover pervasive subgenome-biased transcription, with pronounced B subgenome dominance in grain tissues. Spatially informed regulatory and hormone signaling analyses reveal stage-specific maternal-filial communication and coordinated auxin, abscisic acid, and gibberellin crosstalk. Co-expression network integration further identifies NAC transcription factors as central regulators linking storage metabolism, programmed cell death, and grain maturation. Our study demonstrates the power of integrating single-nucleus and spatial transcriptomics to resolve regulatory complexity in polyploid crop organs and provides a foundational resource for dissecting wheat grain development and improving yield and quality.