Arctic permafrosts are rapidly thawing in response to climate change, stimulating microbial activity and release of additional greenhouse gases, such as methane. Recently, catechin amendment of thawed permafrost soil microcosms demonstrated >80% decrease in methane production over 35 days compared to unamended controls, with metagenome, metatranscriptome and metabolome analyses revealing a shift in prokaryotic carbon metabolism from a syntrophic network feeding methanogens to one dominated by catechin fermentation. Here we leverage these same data to investigate potential virus impacts on this microbial community metabolism shift. In total, 900 DNA virus operational taxonomic units (vOTUs) were identified as actively lytic based on transcribed structural and lysis genes. Of these, 41% were predicted to infect at least one of 56 transcriptionally active prokaryote genera representing 13 phyla. The most active vOTUs were predicted to infect key catechin-degrading genera including Clostridium and undescribed Bacillota genus JAGFXR01. Temporally, viral communities responded later than prokaryotes to catechin amendment, reflecting a virus production lag to targeting key microbial responders. An induced prophage represented by a vOTU predicted to infect JAGFXR01 dominated the catechin-amended viral community response. It reached abundances 20-156 times higher than its host, suggesting intense viral lysis that could release degraded catechin intermediates. Indeed, catechin intermediate gene expression in non-catechin-degraders were elevated in catechin-amended samples, suggesting JAGFXR01 lysis products were taken up by non-catechin-degraders as part of community carbon metabolism rewiring. Together, these results potentially place viruses at the heart of carbon rewiring that modulates ecosystem outputs of climate-critical thawing permafrosts.

Analyzing biological networks demands scalable annotation tools, yet existing methods fall short in clustering power, statistical flexibility, and broad data compatibility. We introduce RISK (Regional Inference of Significant Kinships), a next-generation tool that overcomes these challenges by integrating community detection algorithms, rigorous overrepresentation analysis, and a modular architecture that supports diverse network types. RISK identifies biologically coherent relationships within networks and generates publication-ready visualizations, as demonstrated by its ability to resolve compact functional modules in Saccharomyces cerevisiae protein–protein interaction and genetic interaction networks. Its application to a high-energy physics citation network reveals structured relationships among research subfields, highlighting its versatility beyond biological systems. As biological and interdisciplinary networks increase in size and complexity, RISK’s scalability and adaptability make it a powerful solution for modern network analysis.

Pooled CRISPR screens using single-cell RNA sequencing (scRNA-seq) have emerged as powerful tools to uncover gene function, map regulatory networks, and identify genetic interactions. However, the inherent sparsity of bacterial scRNA-seq data has posed a major challenge toward applying these approaches to bacteria. Here, we present mapSPLiT, a pooled bacterial CRISPR activation and interference (CRISPRa/i) screening platform that enables large-scale experiments that simultaneously link hundreds of perturbations to their corresponding single-cell transcriptomes. By targeting 52 known or putative transcription factors with 118 perturbations in a pooled experiment, we expanded the E. coli regulatory network map, determined the function of putative regulators, and identified emergent global phenotypes. By targeting combinations of transcription factors simultaneously, we uncovered genetic interactions and regulatory logic between them. Mapping regulatory networks for carbon utilization in P. putida revealed control points that could expand metabolic flexibility and improve biomanufacturing. Together, these results establish mapSPLiT as a generalizable platform for single‑cell functional genomics in bacteria.

Microbiomes, complex communities of microorganisms and their genetic material, hold immense potential for addressing global challenges in diverse sectors, including healthcare, agriculture, and bioproduction. Engineering these intricate ecosystems, however, necessitates a comprehensive understanding of the complex web of microbial interactions. The emergence of machine learning (ML) has revolutionized microbiome research, offering powerful tools to analyze massive data sets, uncover hidden patterns, and predict microbial behavior. ML algorithms have demonstrated remarkable success in identifying and characterizing microbial communities, predicting interactions between organisms and optimizing the design of microbial communities for specific functions. This Perspective examines the transformative applications of ML in the context of microbiome engineering, encompassing both microbiome data analysis and the targeted manipulation of microbial communities. These techniques employ a variety of strategies, including the manipulation of quorum sensing molecules, antimicrobial peptides, growth conditions, the introduction of probiotics, and the utilization of bacteriophages. By integrating ML with experimental approaches, researchers are pushing the boundaries of microbiome engineering, paving the way for novel applications in diverse fields. However, it is important to acknowledge the challenges that ML algorithms face, such as the limited availability of high-quality, large-scale data sets, the inherent complexity of biological systems, and the need for improved integration of experimental and computational methods. This perspective further discusses the future perspectives of the field, highlighting expected developments in data generation, algorithm development, and interdisciplinary collaboration. These advancements hold the key to unlocking the full potential of microbial communities for addressing pressing global challenges.

Heterotrophic bacteria supply cobalt-containing cobalamin (Cbl) to marine microalgae, which in return provide organic substrates. Such metabolic cross-feeding can regulate the species composition of prolific marine plankton and biofilms. Algal-produced glycine betaine (GB) can be catabolized by bacteria using a Cbl-dependent demethylase (MtgBCD). Yet, GB's impact on bacterial Cbl production during cross-feeding, and its control by cobalt scarcity remains elusive. Here, we demonstrate that Phaeobacter inhibens bacteria boost their Cbl production 25-fold when grown in monocultures with GB compared to glucose. During co-cultivation, P. inhibens satisfied the Cbl requirements of Gephyrocapsa huxleyi algae. Transcriptomic analysis of mono- and co-cultures revealed co-expression of mtgBCD and gbcAB genes encoding Cbl-dependent and -independent GB demethylases, respectively. Distinct growth defects of deletion mutants indicate that P. inhibens switches from Cbl-dependent to -independent GB demethylation under cobalt limitation, involving a Cbl riboswitch. Our findings suggest a positive feedback loop in which algal GB release stimulates the bacterial supply of Cbl. We predict the breakdown of this interaction under naturally occurring cobalt limitation which potentially contributes to the transient nature of algal blooms. Comparative genomics indicate that this mechanism is widespread in Rhodobacterales and other abundant marine Alphaproteobacteria, underscoring its pivotal role in global ocean productivity.

Precise, reversible control of transgene expression is essential for safe and durable gene therapy, yet current inducible systems remain difficult to translate in vivo due to large size, limited induction duration, and dependence on immunogenic regulators. Here we present the adaptamer (ADAR modulatable aptamer), a compact (<120 bp) RNA switch that couples an FDA approved small molecule responsive aptamer with endogenous ADAR mediated RNA editing to regulate expression. Ligand binding stabilizes a double stranded RNA structure that recruits ADAR to convert a stop codon into a sense codon, restoring downstream protein translation. This cleavage free, post transcriptional mechanism enables precise, small molecule dependent modulation without exogenous protein machinery. We demonstrate that adaptamers are highly functional across multiple cell lines, including human T cells. In mice, AAV delivered adaptamer controlled FGF21 expression induced metabolic remodeling, significantly increasing energy expenditure and reversing obesity. This minimal, programmable system offers a clinically compatible approach for tunable genetic medicines and safe deployment of pleiotropic and dose limited proteins.

The rising rate of drug-resistant fungal infections and the emergence of fungal pathogens with intrinsic resistance phenotypes are a growing concern. The close evolutionary distance between mammals and fungi complicates the design of new antifungals and increases the chances of toxic off-target effects. As such, antifungal drug development usually focuses on fungal-specific proteins when considering potential new targets. Ideal drug targets should mediate essential cell processes and be highly sensitive to inhibition. Targeted gene repression can serve as a model for drug-mediated inhibition and for determining the dosage-sensitivity profile of genes of interest. In the fungal pathogen Candida albicans, classical approaches for gene repression can be labor-intensive and limited to one genetic background due to low throughput. Here, we adapt pooled CRISPRi in C. albicans for the first time and exploit this technique for large-scale functional genomic analysis. Through pooled CRISPRi screening, we test the repression sensitivity of over a hundred essential genes conserved in fungi but absent in humans in a single flask, and successfully identify highly dosage-sensitive genes across multiple cell components and pathways. We further extended our analysis to ten diverse environmental conditions, allowing us to measure how the environment influences dosage-sensitivity profiles. Finally, we extend our experiments to two clinical drug-resistant C. albicans strain backgrounds and demonstrate that many of the fitness defects we observed are conserved in resistant clinical isolates. Together, our results highlight a set of genes that are highly dosage-sensitive across different genetic and environmental contexts, making them attractive targets for further investigation. By facilitating rapid, efficient large-scale functional genomics assays across diverse genetic backgrounds, CRISPRi pooled screening will open new frontiers in C. albicans biology.

Invertebrates, as the majority of macroscopic species on the Earth, are important resources for natural products. Chemical investigations of animals can date back to the early 20th century and have led to the discovery of thousands of compounds with diverse biological functions. These natural products can be structurally classified as terpenoids, polyketides, and alkaloids. Additionally, many compounds have been isolated from symbionts, leading to the widespread belief that animals lack the capability for secondary metabolism. Recent biochemical studies challenge this notion, revealing great potential for animal biosynthesis research. Animals possess larger genomes and more complex metabolic pathways, suggesting untapped biosynthetic potential. In contrast to microorganisms, studies on the biosynthesis of natural products in animals remain limited. Characterized genes represent only a small fraction of their vast genomes. The discovery of biosynthetic gene clusters suggests that the methods used to mine the biosynthetic genes of microorganisms may also be applicable to animals. The characterization of 4-vinylanisole in locusts demonstrates that the pathways lacking clear core biosynthesis enzymes still require multidisciplinary experimental approaches. In summary, further biosynthesis studies will expand methodological approaches and accelerate the characterization of remaining natural product pathways.

Intra-strain variation in model bacterial pathogens can compromise experimental reproducibility and obscure biological interpretations. Dickeya solani, a necrotrophic potato pathogen, is widely studied using the type strain IPO 2222. However, phenotypic discrepancies among laboratories led us to investigate the genetic integrity of this reference stock. We identified at least three distinct IPO 2222 variants co-existing in the original stock, differing solely by mutations in the gene encoding the small regulatory RNA (sRNA) ArcZ. These findings resolve conflicting reports of antimicrobial activity in this strain described by Brual et al. (PLOS Genetics 19, e1010725, 2023) and Matilla et al. (mBio, e02472-22, 2022). We demonstrate that these are adaptive mutations rapidly selected in planta during host infection. Crucially, rather than systematically inactivating the gene, these mutations modulate the cellular levels of processed ArcZ. This modulation can uncouple virulence from antimicrobial activity. These variants behave as social cheaters, exhibiting a frequency-dependent fitness advantage over the cooperative wild-type strain during co-infection. These findings provide evidence that remodeling of a pleiotropic sRNA drives the emergence of bacterial cheaters within a plant host. The speed at which these mutants sweep through the population underscores the intense selective pressure acting on regulatory networks during infection, identifying sRNA modulation as a pivotal mechanism for rapid short-term adaptation.

Microbial interactions shape the composition and stability of the human gut microbiome. Yet, the molecular mechanisms underlying these relationships remain poorly understood. Here, we systematically profiled 1,224 pairwise growth interactions among 36 representative gut bacterial strains using a spent medium assay. Most interactions were inhibitory and largely explained by resource depletion or medium acidification. We investigated further the molecular basis of a positive interaction, showing that Clostridium perfringens promotes the growth of Mediterraneibacter gnavus through extracellular vesicles. Additionally, we identified Veillonella parvula as a key species capable of modulating environmental pH, thereby enabling the growth of Parabacteroides merdae, a strain highly sensitive to acidic conditions. This pH-buffering effect was enhanced by guanine supplementation and persisted in multi-species communities containing different pH-lowering strains from diverse bacterial phyla. This demonstrates the ecological relevance of V. parvula, a species that, despite being commonly present in human gut microbiomes, is generally found at low levels. Overall, the comprehensive dataset and mechanistic insights presented here provide a foundation for rational strategies to manipulate microbiome composition and function.

Plant genome editing has undergone a transformative shift with the advent of advanced molecular tools, offering unprecedented levels of precision, flexibility and efficiency in modifying genetic material. While classical site-directed nucleases such as ZFNs, TALENs and CRISPR-Cas9 have revolutionized genome engineering by enabling targeted mutagenesis and gene knockouts, the landscape is now rapidly evolving with the emergence of novel systems that go beyond the conventional double strand break (DSB)-mediated approaches. Advanced and recent tools include LEAPER, SATI, RESTORE, RESCUE, ARCUT, SPARDA, helicase-based approaches like HACE and Type IV-A CRISPR system, and transposon-based techniques like TATSI and piggyBac. These tools unlock previously inaccessible avenues of genome and transcriptome modulation. Some of these technologies allow DSB-free editing of DNA, precise base substitutions and RNA editing without altering the genomic DNA, a significant advancement for regulatory approval and for species with complex genomes or limited regeneration capacity. While LEAPER, RESCUE and RESTORE are the new advents in the RNA editing tool, SATI allows DSB-free approach for DNA editing, ARCUT offers less off-target and cleaner DNA repairs and Type IV-A CRISPR system induces gene silencing rather than editing. The transposon-based approaches include TATSI, piggyBac and TnpB, and helicases are used in HACE and Type IV-A CRISPR system. The prokaryotic Argonaute protein is used in SPARDA tool as an endonuclease to edit DNA. The transient and reversible nature of RNA editing tools such as RESTORE and LEAPER introduces a new layer of epigenetics-like control in plant systems, which could be harnessed for tissue-specific and environmentally-responsive trait expression. Simultaneously, innovations like ARCUT and SPARDA utilize chemically-guided editing, minimizing reliance on biological nucleases and reducing off-target risks. Their modularity and programmability are enabling gene function studies, synthetic pathway designs and targeted trait stacking. These advances represent a novel synthesis of genome engineering and systems biology, positioning plant genome editing not just as a tool of modification but as a platform for designing adaptive and intelligent crops, tailored to future environmental and nutritional challenges. Although, many of these recent tools remain to be applied on plant systems, they are proven to be effective elsewhere and hold a great potential to be effective in creating climate-resilient crops.