Enzymes drive cellular metabolism, yet predicting catalytic properties from amino acid sequences remains challenging. Existing protein language models (PLMs) provide powerful general-purpose representations but are often inefficient for high-throughput screening and insufficiently adapted to enzyme-specific tasks. Here, we propose EnzGFM, an enzyme-specific PLM based on a Mamba-Transformer hybrid architecture with hierarchical pre-training to capture enzyme-specific patterns. Across enzyme property prediction benchmarks, EnzGFM consistently outperforms Transformer-based PLMs with 2–5-fold acceleration, achieving relative improvements of 16.67% in kinetic parameter prediction, 15.69% in enzyme-reaction mapping, 13.19% in EC number classification, and 20.04% in mutation effect assessment. Building on EnzGFM, we develop EnzGFM-Agent, an enzyme-focused agentic pipeline. Experimental validation further suggests that EnzGFM-Agent can enrich beneficial variants within small candidate pools. Together, these results demonstrate that EnzGFM captures enzyme-specific sequence-function patterns, while EnzGFM-Agent translates these predictions into experimentally actionable candidates and can help reduce wet-lab screening burden for practical enzyme engineering. Enzyme function prediction from amino acid sequences remains a central challenge in computational biology, despite recent advances in protein language models. This manuscript introduces EnzGFM, an enzyme-specific hybrid model that improves both accuracy and efficiency across multiple prediction tasks and, together with the EnzGFM-Agent pipeline, demonstrates the ability to identify experimentally validated beneficial variants while reducing screening effort.

The development of advanced genome engineering tools is crucial for optimizing metabolic pathways in Saccharomyces cerevisiae and achieving efficient biomanufacturing. This study proposes an enhancing multiplex genome editing strategy in S. cerevisiae by employing Escherichia coli-derived single-stranded annealing proteins (SSAPs) combined with S. cerevisiae-derived homologous recombinases (Rad51 and Rad52). The strategy utilizes an SSAP-Rad-Linearized CRISPR (SRLC) platform, which supports efficient simultaneous editing of multiple genomic loci without constructing complex multi-gRNA expression vectors. Co-overexpressing Rad51/Rad52 and E. coli SSAP proteins significantly enhances homologous recombination (HR), allowing precise multi-locus genome editing mediated by short homologous arms. Furthermore, SRLC employs a linearized CRISPR-Cas system to stimulate homologous recombination and enable counter-selection in S. cerevisiae, thereby improving precise multiplex genome editing efficiency. We applied SRLC to engineer the malonyl-CoA metabolic pathway in S. cerevisiae. Through a single round of editing and screening, we constructed a chassis strain with 9 targets simultaneously modification and achieved a 9.6-fold increase in intracellular malonyl-CoA. Using this chassis, 3-hydroxypropionic acid production increased 4.5-fold relative to wild-type S. cerevisiae. This platform offers a robust and scalable tool for S. cerevisiae manipulation and a practical pathway-engineering strategy for building for malonyl-CoA-derived factories.

All organisms defend against viral infections through active immunity mechanisms that clear the virus or population-level mechanisms that cause regulated cell death. Recent research increasingly shows that, across all domains of life, these two types of defence systems are coupled such that regulated cell death mechanisms safeguard or back up active immunity.

Cells have evolved to defend against perturbations by maintaining their intrinsic homeostasis to survive. However, no intrinsic pathway exists for them to expel new-to-biology synthetic nanostructures. Here we establish crosstalk between supramolecular transformations and genetic responses, achieving programmable influx–efflux cycles of nano-assemblies in living bacterial cells to restore redox and energy homeostasis. Specifically, a model photosensitizer–peptide conjugate undergoes multiple redox cycles between methionine and methionine sulfoxide (MetO), resulting in reversible morphological transformations between nanofibers (NFs) and nanoparticles (NPs). Upon irradiation, the oxidized peptide NPs are internalized into bacteria. To counteract the perturbations caused by internalized NPs, engineered bacteria activate the expression of MetO reductases in response to photo-oxidative stress. The internalized NPs are intracellularly enzymatically reduced such that they are expelled as reduced NFs, setting the stage for subsequent cycles. The concept presented here paves the way for the interlinked network between dynamic supramolecular assemblies and cellular regulatory behaviors. Cells have evolved defense pathways to maintain homeostasis against external perturbations, but not against new-to-nature nanomaterials, which can irreversibly accumulate and induce cell stress. Now, programmable peptide nanoassemblies have been designed that can morphologically transform in influx–efflux cycles in response to restoring redox homeostasis in bacterial cells.

Plant roots are hotspots for interactions with soil microbes, where a characteristic bacterial community structure is formed. Plant specialized metabolites often play pivotal roles in this assembly process. However, the molecular basis underlying root microbiota responses to these bioactive compounds, and how such metabolic interactions shape the assembly of host-specific root microbiota, remain largely unknown. Nicotine is a toxic alkaloid predominantly produced by the genus Nicotiana, and the genus Arthrobacter is known as one of the nicotine-degrading bacteria in the tobacco root microbiota. In this study, we used the tobacco–Arthrobacter interaction system as a model and integrated comparative genomics and experimental genetic manipulation assays to uncover the role of bacterial catabolism capacity for host specialized metabolites in shaping host-specific root microbiota. Nicotine catabolism genes are uniquely found in the Arthrobacter strains derived from nicotine-containing environments, and this restricted gene distribution is driven by a plasmid-mediated horizontal gene transfer. To assess the ecological consequences of this genomic adaptation in Arthrobacter fitness in tobacco roots, we characterized the nicotine utilization ability of Arthrobacter and conducted adaptation assays under in planta conditions using genetically manipulated Arthrobacter strains and tobacco mutants impaired in nicotine catabolism and biosynthesis, respectively. Nicotine improves Arthrobacter colonization of tobacco roots through a catabolism-dependent mechanism. Bacterial community analysis using a synthetic community approach further demonstrated that this metabolic adaptation enhances Arthrobacter fitness within tobacco root microbiota. Our findings illustrated that bacterial catabolic capacity toward host-derived plant specialized metabolites is key for successful root colonization. This metabolic adaptation is driven by plasmid-mediated horizontal gene transfer and ultimately shapes the structure of the root microbiota community.