Anthropogenic nitrogen (N) enrichment profoundly influences the growth of N-fixing plants, but the drivers underlying their highly variable biomass responses remain poorly quantified. Through a meta-analysis of 2,110 paired observations across 76 N-fixing plant species from 121 experiments, we demonstrate that N enrichment increases N-fixing plant biomass by 14.2% on average, with responses ranging from −99.4% to 1301.3%. Intraspecific variability (26.9%; i.e., variation among individuals within species) and plant phylogeny (21.1%) represent substantial yet frequently overlooked sources of uncertainty in biomass responses of N-fixing plants to N enrichment, even though environmental conditions (climatic and edaphic factors) and experimental design (N enrichment rate and duration) account for a significant portion of biomass response variation. Intraspecific variability, phylogeny, and experimental design collectively explain a larger proportion of variance in root than in shoot biomass responses, whereas environmental conditions show the opposite trend. These results provide quantitative evidence that phylogeny and intraspecific variability are indispensable for accurately predicting the response of N-fixing plants to global change. Nitrogen enrichment alters the growth of nitrogen-fixing plants, but responses vary widely. This meta-analysis shows that phylogeny and intraspecific variation are major contributors to that variability.

Protospacer adjacent motif (PAM) requirements constrain the targeting scope of CRISPR-Cas9 systems and limit genome editing applications that require precise nuclease positioning. Although broad-PAM and near-PAMless Cas9 variants have expanded the editable genome, robust targeting across certain PAM families remains difficult, particularly at NNAN PAMs. Here, we apply Sequence Display to generate multiplexed, large-scale sequence–activity datasets for Staphylococcus lugdunensis Cas9 (SlugCas9) across the NNAN PAM family, including NNAA, NNAT, NNAC and NNAG. These datasets quantitatively resolve how individual and combinatorial mutations reshape nuclease activity across distinct PAM contexts. By integrating Sequence Display-derived datasets with protein language models, we construct fine-grained activity landscapes and identify SlugCas9 variants with enhanced and broadened activity across NNAN PAMs. The resulting variants reprogram PAM recognition beyond the native G-compatible landscape of SlugCas9 and enable improved base editing at selected endogenous NNAN-constrained sites in cultured mammalian cells, including disease-relevant sites in APOC3 and PCSK9. Together, our results establish a strategy for reprogramming PAM preference in compact Cas9 nucleases and highlight Sequence Display as a general platform for evolving genome editors toward specific and difficult-to-access PAM spaces. Compact CRISPR-Cas9 enzymes are easier to deliver but recognize few targets. Here, authors use Sequence Display to generate large-scale SlugCas9 sequence–activity datasets and apply protein language models to engineer variants that recognize difficult NNAN PAMs and edit selected human genomic sites.

Despite its prominent industrial use for L-proline production, microbial fermentation by Corynebacterium glutamicum achieves a suboptimal yield of glucose conversion, far below the theoretical maximum of 0.63 g/g. This study aimed to develop an efficient engineered E. coli chassis to produce L-proline from glucose by addressing the rate-limiting enzymatic step and high glucose conversion. The γ-glutamyl kinase encoding gene (proB) was screened by directed evolution, and the catalytic efficiency of the ScProB** variant toward L-glutamate increased 2.75-fold. After channeling 2-oxoglutarate flux and supplying sufficient NADPH, L-proline production reached 86.27 ± 2.58 g/L, with a 0.33 ± 0.01 g/g glucose yield in 10-L fed-batch fermentation. Further increasing the flux through central carbon metabolism by implementing a non-PTS glucose uptake system, amplifying the anaplerotic reaction, and redirecting the flux away from the glyoxylate cycle, the resulting strain BP23 achieved 119.89 ± 5.66 g/L L-proline, with a productivity of 1.93 ± 0.03 g/L/h. The glucose conversion yield improved to 0.49 ± 0.01 g/g after 62 h of fed-batch fermentation, indicating that the E. coli strain is a promising workhorse for producing L-proline with high glucose conversion.

RNA is an increasingly attractive target for small-molecule therapeutics, yet the rational design and computational modelling of RNA-binding small molecules are constrained by the scarcity of experimentally measured binding affinities linked to 3D structures. Here we present ARIEL (Affinity Resource for Interactions of RNA and Ligands), an openly accessible, manually curated database that pairs experimentally determined dissociation constants (Kd) of RNA–ligand complexes with the corresponding experimental structures. We screened 6954 RNA-containing structures from the Protein Data Bank, extracted 2843 complexes with a small ligand, and manually curated the associated primary literature to retrieve the corresponding Kd values. This yielded a nonredundant dataset of 278 RNA–ligand complexes with experimental affinities, including 179 unique ligands bound to 171 different RNA targets, with Kd values spanning nine orders of magnitude (−log Kd 2.1–11.0). Each entry also provides the Kd measurement method and conditions, full literature provenance, ligand physicochemical descriptors, and binding-pocket annotations, with direct cross-links between structures, ligands and source publications. ARIEL will be maintained through regular, versioned releases; it is freely available at https://aocdweb.com/BioTools/ARIEL/.

Millions of synthesis procedures have been published, but their meta-analysis, important for identifying successful synthesis patterns, is difficult due to varied reporting structures. Using this information requires converting unstructured text into a precise sequence of actions, which requires understanding domain jargon, a challenge usually addressed by fine-tuning models on labeled data. We present a simple, training-free workflow for laboratory action extraction that works across fields, encoding domain knowledge through explicit action sets and using powerful, readily available Large Language Models, or LLMs. Applied to zeolite synthesis, a widely used class of porous materials, our approach outperforms existing open-source tools and delivers context-aware results with open, locally runnable LLMs. It also matches state-of-the-art, field-specific models in their own domains, highlighting the generalization capabilities of current LLMs. The methodology showcased,  along with the open algorithms made available, allow chemists to evaluate action sets, screen LLMs, and digitize laboratory procedures accurately with minimal effort. A domain-agnostic, training-free workflow leverages open-source large language models with explicit action sets and post-processing rules to extract laboratory procedures from unstructured text, demonstrated through zeolite synthesis.

BacMap 3.0 (https://bacmap.ca) is a comprehensive, freely accessible resource integrating 23 524 bacterial and archaeal species, 32 866 genomes, >100 million genes, 1756 primary metabolic pathways, and 45 284 metabolites within one interactive platform. Despite the extraordinary growth of bacterial and archaeal genome sequencing, relevant data remain distributed across numerous specialized databases, impeding comprehensive genome-to-function analysis. While existing microbial resources generally emphasize genome annotation, taxonomy, metabolic pathways, or phenotypes separately, BacMap 3.0 unifies these complementary information layers within an organism-centred framework. It resolves this fragmentation by consolidating genomic, functional, structural, metabolic, phenotypic, ecological, and phylogenetic information into richly detailed, organism-specific “BacCards”. Each BacCard combines an interactive genome map with extensive annotations covering bacterium-specific taxonomy, physiology, morphology, metabolism, protein structures, pathways, and metabolites. To facilitate data exploration, BacMap 3.0 supports taxonomic and phenotype-based browsing, chemical structure searches, interactive genome visualization, as well as gene, genome, 16S RNA, and protein BLAST searches. Building on earlier releases devoted primarily to genome visualization, BacMap 3.0 increases organismal coverage by >13-fold and transforms BacMap into an integrated bacterial knowledgebase for genome-to-function investigation across multiple biological scales. By bringing diverse biological information together in a single searchable environment, BacMap 3.0 offers a powerful platform for comparative genomics, microbiome research, infectious disease investigation, and biotechnology with a goal of accelerating the interpretation of microbial diversity, biology, and function.

Enhancers are cis-regulatory elements that control gene expression during development, cell differentiation, and disease. Genome-wide studies have mapped large numbers of candidate regulatory elements, but experimentally supported enhancer information remains dispersed across primary studies and experimental systems. Here, we present ENdb 2.0 (http://www.licpathway.net/ENdb/index.php), an updated, manually curated database of experimentally supported enhancers. Compared with ENdb 1.0, the major updates are (i) literature coverage through 2026 and five species: Homo sapiens, Mus musculus, Drosophila melanogaster, Danio rerio, and Rattus norvegicus; (ii) 4831 enhancer evidence records, including 2793 enhancer–TF–target-gene records involving 496 TFs and 1070 target genes; (iii) expanded curation of genomic coordinates, experimental evidence, enhancer type, regulatory state, biological condition, tissues, cell sources, cell types, and diseases, covering 169 tissues, 1255 cell sources, 415 cell types, and 226 diseases; and (iv) biological search, genome browsing, pathway enrichment analysis, agent-assisted query, and structured data download. ENdb 2.0 links experimentally supported enhancer regions with regulatory factors, target genes, biological conditions, and supporting experiments across physiological and disease states.

Positive-sense RNA viruses often use 3′ tRNA-like structures (TLSs) instead of poly(A) tails to capture the host translation machinery. While TLSs resemble canonical tRNAs and engage host factors, their ability to directly recruit ribosomes has remained unresolved. Here, we present cryo-electron microscopy snapshots of the histidine-accepting TLS (TLSHis) from tobacco mosaic virus bound to the tobacco 60S subunit, the 80S ribosome, and the 80S-tRNAiMet initiation complex. Across these states, TLSHis is consistently anchored to the L1 stalk of 60S, even under cycloheximide treatment, yet remains dynamic on the 40S subunit. Structural analysis shows that TLSHis transitions from the E-site to an adjacent Z-site to accommodate initiator tRNA. Functional assays show that TLSHis preferentially associates with ribosomal subunits over polysomes and can substitute for a poly(A) tail to promote robust cap-dependent translation. These results provide structural and mechanistic evidence that the viral tRNA mimic exploits the tRNA-exit channel to stabilize initiation-competent ribosomes and drive poly(A)-independent translation. Some plant RNA viruses use 3′ tRNA-like structures to promote translation without poly(A) tails. Here, the authors reveal how the tobacco mosaic virus tRNA-like structure directly engages host ribosomes to facilitate translation initiation.