AbstractPlant germ cells are specified de novo from somatic tissues, yet how early anther progenitors initiate divergent germline and somatic fates remains unclear. Here, we generated a high-resolution single-cell transcriptomic atlas of Arabidopsis anther development, profiling 66,864 cells from early lineage specification through pollen maturation. We resolved major anther cell types and identified an early L2-derived intermediate population, termed the archesporial-derived state (Ar-d), marked by REM22, ER/ERL1/2, SPL/NZZ, and BAM1. Although classical anatomical studies define archesporial cells, primary sporogenous cells, primary parietal cells, and secondary parietal derivatives based on developmental position and inferred lineage relationships, our analysis showed that these early L2-derived intermediates remain globally similar, forming a shared progenitor-like transcriptional state before lineage stabilization. Trajectory reconstruction, RNA velocity, and module analyses revealed that germline and somatic trajectories first emerge through restricted branch-biased transcriptional programs, which are later amplified into robust lineage-specific identities. Mutant single-cell analyses defined two regulatory steps: SPL/NZZ establishes early branch-associated transcriptional programs within the Ar-d state, whereas EMS1 stabilizes the somatic/tapetal trajectory and restricts inappropriate enrichment of germline-associated states. Lineage-resolved analyses further defined maturation programs in germline, tapetum, middle layer, and endothecium cells, and identified LBD transcription factors required for normal pollen development. Together, our study reveals a cryptic mode of anther fate priming in which shared early L2-derived intermediates progressively transform limited branch-biased transcriptional programs into distinct reproductive and somatic cell identities through SPL-dependent priming and EMS1-dependent somatic/tapetal stabilization.Read more

AbstractPlants encounter diverse pathogens and have evolved a two-layered innate immune system to detect pathogen molecules and activate defense mechanisms that restrict infection. Most cloned plant Resistance (R) genes encode NLR immune receptors. NLR genes are often found in clusters of paralogs with sequence and copy number variation; whether these NLR clusters evolve in response to single or multiple pathogens has been unclear. We report here the isolation of a Phytophthora capsici resistance gene, Rpc2, along with a novel P. infestans resistance gene, Rpi-amr5, from two Solanum americanum accessions. These orthologous genes reside in the Rpi-amr1 cluster, which has previously been associated with resistance to P. infestans. By screening RXLR effector libraries of P. infestans and P. capsici, we identified multiple effectors recognised by both NLRs. Our findings highlight the complexity of NLR clusters and evolution driven by interactions with multiple pathogens. This work will underpin efforts to elevate resistance against Phytophthora pathogens and enhances our understanding of NLR evolution.Read more

AbstractThe endoplasmic reticulum (ER) is the largest intracellular membrane-bound organelle in eukaryotic cells, comprising an interconnected network of tubules and sheets. Lunapark (LNP) functions as an E3 ubiquitin ligase that targets RHD3, a dynamin-like GTPase essential for homotypic fusion of ER tubules, for proteasomal degradation, thereby modulating ER tubule stability. However, the regulatory mechanism governing LNP activity remains largely unknown. Here, we report the characterization of the rice oskish mutant, which exhibits aberrant aggregation of ER tubules and defective ER exit of seed storage proteins. OsKish is an ER-localized small protein that physically interacts with both OsLNPs and RHD3-like (RHD3L) protein. OsKish stabilizes OsLNPs by suppressing their auto-ubiquitination; concomitantly, OsKish attenuates the membrane fusion activity of RHD3L possibly by inhibiting its oligomerization. Altogether, our studies propose an OsKish–OsLNPs–RHD3L framework for fine-tuning homotypic ER tubule fusion in rice, providing mechanistic insights into the maintenance of ER architecture in plants.Read more

AbstractLand plant chloroplasts are characterized by a prominent morphological feature: the stacking of thylakoid membranes into grana, i.e. defining a continuous membranous structure surrounded by the stroma and folded into appressed grana and non-appressed domains, enclosing the thylakoid lumen.Photosystem II and its antenna system are concentrated, whereas Photosystem I and ATPase are located in the grana end membranes and, in the stroma, membranes connecting grana. Thylakoid stacking promotes the physical separation of Photosystems I and II, thereby ensuring a balanced flow of excitation energy between them. Through a genetic dissection of the light-harvesting complex subfamily Lhcb, we showed that removal of the antenna system serving Photosystem II completely abolished grana formation, whereas deletion of specific Lhcb subgroups reduced stacking by up to 40%. Lhcb5 alone was sufficient to partially restore stacking, whereas exclusive expression of Lhcb2 resulted in grana with an unusually large size, which disassembled upon illumination. Time-resolved fluorescence measurements revealed that the extent of excitation energy spillover from PSII to Photosystem I correlated with the level of Lhcb depletion, rather than with granal architecture. These findings provide genetic evidence for a functional link between Photosystem II antenna composition, lateral heterogeneity and excitation energy partition between photosystems.Read more

AbstractProximity labeling (PL) proteomics, primarily powered by engineered biotin ligases such as BioID and TurboID, has emerged as a transformative approach for mapping protein association networks and subcellular proteomes in living cells. By covalently biotinylating proteins within a nanometer-scale radius of a bait protein, PL captures transient, weak, and spatially restricted associations that often escape conventional affinity-based methods. However, applying PL in plants introduces distinctive challenges, including rigid cell walls that can limit biotin penetration, tissue-specific variation in substrate accessibility, enzyme temperature sensitivity, and the difficulty of removing excess free biotin after labeling. Here, we outline best practices for designing and implementing biotin ligase-based PL experiments in plant systems, drawing on experience across multiple species. We discuss key considerations, including enzyme selection, expression system design, fusion protein validation, biotin delivery strategies, protein extraction and enrichment, and mass spectrometry-based analysis. We also highlight emerging quantitative and conditional PL approaches that enable dynamic comparison of proxiomes across developmental stages, environmental conditions, and genetic backgrounds. Throughout, we emphasize the importance of rigorous controls, careful terminology, and orthogonal validation to ensure biologically meaningful interpretation of PL datasets. These guidelines aim to standardize experimental design and interpretation, facilitating reproducible and biologically meaningful PL studies in plant systems.Read more

Read more