The morphology of bulliform cells located on the upper epidermis of leaves is one of the most important cell structures affecting leaf shape. Although many mechanisms regulating the development of bulliform cells have been reported, the fine regulatory mechanisms governing this process have rarely been described. To identify novel components regulating rice leaf morphology, a mutant showing a constitutively rolling phenotype from the seedling stage to flowering, known as crm1‐D, was selected for further analysis. Anatomical analyses in crm1‐D were attributable to the size reduction of bulliform cells. The crm1‐D was controlled by a single dominant nuclear gene. Map‐based cloning revealed that Roc8, an HD zipper class IV family member, was responsible for the crm1‐D phenotype. Notably, the 50‐bp sequence in the 3′‐untranslated region (3′‐UTR) of the Roc8 gene represses Roc8 at the translational level. Moreover, the roc8 knockdown lines notably increased the size of bulliform cells. A series of assays revealed that Roc8 negatively regulates the size of bulliform cells. Unexpectedly, Roc8 was also observed to positively mediate lignin biosynthesis without incurring a production penalty. The above results show that Roc8 may have a practical application in cultivating materials with high photosynthetic efficiency and low lignin content.

Photorespiration, often considered as a wasteful process, is a key target for bioengineering to improve crop yields. Several photorespiratory bypasses have been designed to efficiently metabolize 2-phosphoglycolate and increase the CO2 concentration in chloroplasts, thereby reducing photorespiration. However, the suppression of primary nitrate assimilation remains an issue when photorespiration is inhibited. In this study, we designed a carbon and nitrogen metabolism–coupled photorespiratory bypass, termed the GCBG bypass, in rice (Oryza sativa) chloroplasts. Our results demonstrated efficient assembly and expression of the GCBG bypass in rice chloroplasts, which affected the levels of typical metabolites and their derivatives of natural photorespiration and enhanced the photosynthetic efficiency. Metabolomic analyses revealed that oxaloacetate, produced from glycolate in chloroplasts, positively impacted amino acid synthesis, energy metabolism, and sugar synthesis. The engineered GCBG plants showed an average yield increase of 19.0% (17.8% to 20.2%) compared with wild-type plants under natural growth conditions, alongside improved nitrogen uptake, which compensated for 44.1% of yield losses under nitrogen-limited conditions. In summary, the GCBG bypass substantially improved the photosynthetic efficiency, biomass, and yield in rice by integrating carbon and nitrogen metabolism. This study introduces a strategy for engineering high-yielding rice or other crops with improved photosynthetic efficiency and nitrogen uptake.

C₃ and C₄ grasses directly and indirectly provide the vast majority of calories to the human diet, yet our understanding of the molecular mechanisms driving photosynthetic productivity in grasses is largely unexplored. Ground meristem cells divide to form mesophyll or vascular initial cells early in leaf development in C3 and C4 grasses. Here we define a genetic circuit composed of SHR (SHORT ROOT), IDD (INDETERMINATE DOMAIN), and PIN (PIN-FORMED) family members that specifies vascular identify and ground cell proliferation in leaves of both C₃ and C₄ grasses. Ectopic expression and loss-of-function mutant studies of SHR paralogs in the C₃ plant Oryza sativa (rice) and the C₄ plant Setaria viridis (green millet) revealed the roles of these genes in both minor vein formation and ground cell differentiation. Genetic and in vitro studies further suggested that SHR regulates this process through its interactions with IDD12 and 13. We also revealed direct interactions of these IDD proteins with a putative regulatory element within the auxin transporter gene PIN5c. Collectively, these findings indicate that a SHR-IDD regulatory circuit mediates auxin transport by negatively regulating PIN expression to modulate minor vein patterning in the grasses.

Grain size and weight are crucial yield-related traits in rice (Oryza sativa). Although certain key genes associated with rice grain size and weight have been successfully cloned, the molecular mechanisms underlying grain size and weight regulation remain elusive. Here, we identified a molecular pathway regulating grain size and weight in rice involving the MPS ONE BINDER KINASE ACTIVATOR-LIKE 1A–SERINE/THREONINE-PROTEIN KINASE 38–CYCLIN C (OsMOB1A–OsSTK38–OsCycC) module. OsSTK38 is a nuclear Dbf2-related kinase that positively regulates grain size and weight by coordinating cell proliferation and expansion in the spikelet hull. OsMOB1A interacts with and enhances the autophosphorylation of OsSTK38. Specifically, the critical role of the OsSTK38 S322 site in its kinase activity is highlighted. Furthermore, OsCycC, a component of the Mediator complex, was identified as a substrate of OsSTK38, with enhancement by OsMOB1A. Notably, OsSTK38 phosphorylates the T33 site of OsCycC. The phosphorylation of OsCycC by OsSTK38 influenced its interaction with the transcription factor KNOTTED-LIKE HOMEOBOX OF ARABIDOPSIS THALIANA 7 (OsKNAT7). Genetic analysis confirmed that OsMOB1A, OsSTK38 and OsCycC function in a common pathway to regulate grain size and weight. Taken together, our findings revealed a connection between the Hippo signalling pathway and the Cyclin-Dependent Kinase (CDK) module in eukaryotes. Moreover, they provide insights into the molecular mechanisms linked to yield-related traits and propose innovative breeding strategies for high-yielding varieties.

Neural Radiance Fields (NeRF) have been widely adopted for reconstructing high quality 3D point clouds from 2D RGB images. However, the segmentation of these reconstructed 3D scenes is more essential for downstream tasks such as object counting, size estimation, and scene understanding. While segmentation on raw 3D point clouds using deep learning requires labor intensive and time-consuming manual annotation, directly training NeRF on binary masks also fails due to the absence of color and shading cues essential for geometry learning. We propose Invariant NeRF for Segmentation (InvNeRFSeg), a two step, zero change fine tuning strategy for 3D segmentation. We first train a standard NeRF on RGB images and then fine tune it using 2D segmentation masks without altering either the model architecture or loss function. This approach produces higher quality, cleaner segmented point clouds directly from the refined radiance field with minimal computational overhead or complexity. Field density analysis reveals consistent semantic refinement: densities of object regions increase while background densities are suppressed, ensuring clean and interpretable segmentations. We demonstrate InvNeRFSegs superior performance over both SA3D and FruitNeRF on both synthetic fruit and real world soybean datasets. This approach effectively extends 2D segmentation to high quality 3D segmentation.