The centerpiece of the study is an in vivo CRISPR interference, or CRISPRi, screening platform. Unlike traditional knockouts, which permanently delete genes and can be impractical at genome-wide scale, CRISPRi uses a catalytically dead Cas9 protein, dCas9, guided by hundreds of thousands of small RNA molecules to reversibly silence transcription of individual genes. Each bacterial cell in a massive pooled library carries a guide RNA targeting a specific gene, so the abundance of each variant after a period in the host serves as a direct readout of how important the silenced gene was for survival. If a gene is essential in the gut, cells whose guide suppresses it will dwindle in number; if it is dispensable, those cells persist. Gene fitness, measured by deep sequencing of guide RNA abundance with unique molecular identifiers to reduce noise, thus becomes a high-resolution functional reporter of adaptation.

To make the experiment tractable, the researchers colonized mice with OligoMM12, a defined minimal microbial community of twelve mouse gut strains. This gnotobiotic approach, building on earlier genome-guided designs of minimal microbiotas, provides a reproducible yet realistic ecological backdrop in which E. coli must compete with other residents for nutrients and space, without the overwhelming complexity of a conventional microbiome. The team introduced genome-wide CRISPRi libraries into three very different E. coli strains: MG1655, the classic laboratory K-12 derivative; CFT073, a uropathogenic strain; and LF82, an adherent-invasive E. coli, or AIEC, associated with Crohn’s disease. As the libraries circulated through the mice, sequencing revealed which genes each strain depended on under specific conditions.

The first major finding concerns diet. When mice were fed standard chow, a high-fiber diet, or a high-fat Western-style diet, the metabolic landscape facing E. coli changed dramatically, and the essential gene profile shifted with it. Genes required for growth on one diet were dispensable or even deleterious on another, confirming at functional-genomic resolution that what an animal eats directly rewrites the nutritional rules of the gut. The screen also exposed cross-feeding interactions, identifying genes whose loss was tolerated only because other members of the defined community supplied the missing metabolites. In this way the platform does more than list essential genes; it sketches the metabolic web that connects E. coli to its microbial neighbors, showing that the bacterium’s survival strategy is woven into the ecology of the whole community.

Comparing the three strains side by side revealed that there is no single genetic playbook for intestinal colonization. Although the core essential genomes overlapped, reflecting shared housekeeping functions, each strain displayed a distinct signature of requirements. The analysis highlighted divergent strategies in motility, stress response and respiration. MG1655, CFT073 and LF82 each relied on different subsets of genes to cope with bile, acidity, oxygen tension and immune pressures, consistent with the idea that genetic diversity across the E. coli species translates into fundamentally different ecological tactics. The results extend earlier transposon-insertion sequencing studies in enterohemorrhagic E. coli, K1 strains and other pathogens, but with the added precision and tunability that CRISPRi offers, since partial silencing can reveal dose-sensitive phenotypes that complete knockouts miss.

One of the most medically relevant strands of the study concerns AIEC LF82, a strain strongly enriched in the ileal mucosa of Crohn’s disease patients. When the researchers induced intestinal inflammation with dextran sulfate sodium, they found that LF82 preferentially colonized the small intestine in the inflammatory environment, and its genetic dependencies shifted accordingly. Genes tied to iron acquisition became more important, echoing the known biology of lipocalin 2 and neutrophil gelatinase-associated lipocalin, host proteins that sequester iron-bearing siderophores during inflammation. Respiratory genes also rose in prominence, in line with evidence that inflammation disrupts intestinal oxygen balance and creates new terminal electron acceptors that facultative anaerobes can exploit. The inflammatory gut, in other words, is a different nutritional world, and the screen captured that transformation in genetic terms.

Perhaps the most striking discovery involves a mobile genetic element. LF82 carries the Gally prophage, and the CRISPRi data revealed that this integrated virus plays an important role in modulating the strain’s fitness in the inflamed gut. Follow-up analyses quantified phage particle release and showed that the prophage’s influence on fitness differed between inflamed and healthy conditions, adding to growing evidence that prophages are not merely parasitic passengers but functional modulators of host physiology. Previous work has shown that phage integration can rewire the respiratory strategy of E. coli, and the new findings place the Gally element squarely within the colonization toolkit of a Crohn’s-associated pathobiont. The observation that phage production itself appears to be regulated, and blocked under some host-like conditions, underscores how finely tuned these relationships are.

Technically, the study sets a new benchmark for rigor. Gene-level calls were derived using MAGeCK with robust ranking aggregation and Benjamini-Hochberg false-discovery-rate correction across individual mice as biological replicates, and guide activity predictions were used to improve on-target performance, drawing on earlier CRISPR-dCas9 screening work in E. coli. Raw CRISPRi and 16S sequencing data are available through the Sequence Read Archive under accession PRJNA1293909, and processed log2 fold-change values, statistical results, analysis scripts and notebooks are shared through public repositories, allowing other groups to reproduce and extend the work. The authors acknowledge support from the European Research Council, the Agence Nationale de la Recherche, the Fondation pour la Recherche Médicale and the Swiss National Science Foundation’s NCCR AntiResist program.

The broader significance of the work extends well beyond E. coli. By demonstrating that gene fitness can serve as a quantitative sensor of the gut environment, the platform offers a way to probe intestinal conditions directly through the organisms that live there, effectively turning bacteria into living microsensors of diet, inflammation and microbial competition. For microbiome science, where correlational studies abound, this is a shift toward mechanism: every essential gene identified is a hypothesis about a nutrient, a stress or a host defense that can be tested. For medicine, the strain-specific genetic requirements of Crohn’s-associated AIEC and the role of the Gally prophage suggest new angles for interventions that disarm pathobionts without indiscriminately wiping out beneficial residents. As researchers refine gnotobiotic models and CRISPRi delivery, in vivo functional genomics of this kind could become a routine tool, mapping not just which microbes are present in the gut but precisely which molecular functions make them winners or losers in one of the most densely populated ecosystems in the mammalian body.

Subject of Research: In vivo CRISPRi functional genomics of Escherichia coli adaptation to the mouse gut

Article Title: In vivo CRISPRi screens reveal Escherichia coli functional adaptations in the mouse gut

Article References: Maire, A., Ortelli, M., Tkacz, E., Dehio, C., Chassaing, B., Sokol, H., Rolhion, N., & Bikard, D. (2026). In vivo CRISPRi screens reveal Escherichia coli functional adaptations in the mouse gut. Nature Microbiology. https://doi.org/10.1038/s41564-026-02471-8

Image Credits: AI Generated

DOI: 10.1038/s41564-026-02471-8

Keywords: E. coli, CRISPRi, gut microbiome, Nature Microbiology, adherent-invasive E. coli, Crohn's disease, Gally prophage, gene essentiality, mouse model, OligoMM12, functional genomics, intestinal colonization