Genetius

The authors present a cryogenic volume electron microscopy (cryo‑vEM) workflow that enables nanometer‑scale, three‑dimensional imaging of intact plant protoplasts without chemical fixation or staining. Using sorghum stem protoplasts, the method combines optimized protoplast isolation, plunge‑freezing vitrification, automated cryo‑FIB‑SEM, and machine‑learning‑based 3D segmentation to resolve organelles and membrane contacts in their native hydrated state, allowing quantitative analysis of organelle volumes and surfaces.

Using a large-scale sorghum mutant library, the study identified a cysteine substitution in the Zinc Sensor Motif (ZSM) of the sorghum bZIP19/23 homolog, which led to a three-fold increase in seed Zn concentration without adverse developmental effects. This demonstrates that targeted modification of the F‑bZIP ZSM can be an effective strategy for Zn biofortification in cereal crops.

The study employed high‑resolution micro‑CT and machine‑learning analysis to compare three‑dimensional leaf anatomy of wild‑type sorghum and a transgenic line expressing a synthetic Epidermal Patterning Factor (EPFsyn) that reduces stomatal density. Despite the lower stomatal density, overall leaf internal CO₂ conductance (gias) was unchanged because EPFsyn plants displayed larger sub‑stomatal crypts and altered mesophyll airspace, indicating a compensatory anatomical response. These results illuminate the coordination between surface stomatal patterning and internal mesophyll structure in a C4 species, informing water‑use efficiency engineering.

The study presents an optimized Agrobacterium-mediated transformation toolkit for Sorghum bicolor that achieves up to 95.7% editing efficiency using CRISPR/Cas9 targeting the SbPDS gene, and demonstrates comparable performance with a PAM‑broadened SpRY variant. This platform enables multiplex genome editing and is positioned for integration of advanced tools such as prime and base editors to accelerate sorghum breeding.

The study analyzed the population-genomic structure of 1,806 sorghum (Sorghum bicolor) landraces across Africa and Eurasia, revealing that spatial distance, human travel time, and linguistic affiliations together explain up to 27% of genetic variation. Regional patterns showed contrasting degrees of haplotype sharing, indicating differing evolutionary dynamics driven by geography, culture, and climate.

The study examined how two sorghum genotypes respond to increasing soil salinity, focusing on water status, sodium accumulation, and the role of leaf prickles and microhairs. Results indicated genotype‑specific mechanisms, including internal water redistribution at 200 mM NaCl and silicon–sodium exchange at 300 mM NaCl, with silicon accumulating in prickles but not in microhairs, potentially mediated by NIP aquaporins.

The study used single-nucleus RNA sequencing to profile transcriptional responses to drought in mature leaves of Sorghum bicolor, identifying major leaf cell types and their differential gene expression. Drought induced transcriptomic changes that outweighed cell-type differences, indicating a conserved stress response across mesophyll, bundle sheath, epidermal, vascular, and stomatal cells. This convergence enabled the identification of candidate regulators for drought-responsive genes, offering targets for improving water use efficiency in bioenergy crops.

The authors created a high‑throughput protoplast transformation and fluorescent‑protein tagging system for Sorghum bicolor and used it to map the subcellular locations of 234 enzymes across 184 pathways. Their data reveal both conserved and sorghum‑specific localization patterns, including a cytosolic chorismate biosynthesis enzyme and chloroplast puncta for several Calvin‑Benson cycle enzymes, providing the first extensive enzyme‑localization atlas for this C4 crop.

Using CRISPR‑Cas9 to delete CCD8 and LGS1 genes in sorghum, the authors show altered strigolactone‑related metabolites, reduced carbon assimilation, changes in root architecture, and delayed mycorrhizal colonization, with variable effects on Striga resistance suggesting trade‑offs and background‑dependent modifiers. The work highlights the broader physiological roles of strigolactones in sorghum growth and development.

The study employed UAV‐derived imagery and deep learning models (YOLO and Faster R-CNN) to detect sorghum panicles, achieving high precision (mAP@0.50 up to 0.98) and strong correlation with ground truth. Extracted phenotypic traits were then used in regression models (SVR, DTR, RFR) to predict sorghum yield, demonstrating improved prediction performance with YOLO‑based feature extraction.