Massive single-cell atlas reveals how diverse gut infections reprogram the intestinal lining

The distal small intestine, or ileum, is one of the busiest crossroads in mammalian biology. It absorbs vitamin B12, bile acids, fat-soluble vitamins, water and electrolytes, all while acting as a physical and immunological barrier against a constant stream of microbes. Yet despite its central role in nutrition and host defense, and its prominence in Crohn’s disease, the ileum has been surprisingly neglected by single-cell genomics compared with the colon. A new Resource published in Nature Immunology by Daniel Beiting of the University of Pennsylvania and colleagues, working as part of the Mucosal Immunology Studies Team consortium, aims to close that gap. The team has built GutPath, a publicly accessible atlas of more than 500,000 individual cells from the mouse ileum and the ileal-draining mesenteric lymph node, profiled both at steady state and during six phylogenetically distinct infections or colonizations.

The technical foundation of GutPath is CITE-seq, a method that simultaneously captures transcriptomes and roughly 100 surface proteins from the same single cells. The researchers profiled the intestinal epithelium, the underlying lamina propria and the mesenteric lymph node in mice infected with segmented filamentous bacteria, Yersinia pseudotuberculosis, the helminth Nippostrongylus brasiliensis, the fungus Candida albicans, the apicomplexan parasite Cryptosporidium parvum and murine norovirus strain CR6. After rigorous quality control, the atlas retained 505,956 high-quality cells, split between roughly 321,000 ileal cells and 185,000 lymph node cells, with individual cells averaging more than 6,000 transcripts. Annotation combined automated reference-based methods with extensive manual curation, resolving 91 distinct transcriptional cell states in the ileum and 49 in the lymph node, organized in multiple layers from broad lineages down to highly specific phenotypes.

A crucial validation step was to confirm that the atlas faithfully reproduces established immunological archetypes. Differential abundance analysis of cell neighborhoods showed that Nippostrongylus infection expanded tuft and goblet cell populations, the classic hyperplasia seen in helminth-driven type 2 immunity, while Yersinia drove a marked recruitment of neutrophils and monocytes, hallmarks of acute bacterial infection. Cryptosporidium and Yersinia both increased Ifng expression in natural killer and natural killer T cells and enriched the interferon-gamma signaling program, consistent with known interferon-dependent control of the parasite. Murine norovirus produced a more modest type I interferon signature, and both Candida colonization and segmented filamentous bacteria enriched T helper 17 differentiation programs in lymph node T cells, with Candida also elevating transforming growth factor-beta signaling and hypoxia-related gene sets. In other words, the atlas captures the canonical immunology of each model, giving researchers confidence in what it reports about the unexpected.

That confidence matters because the most striking findings emerged from the epithelial compartment rather than the immune one. When the team performed unbiased differential expression analysis, the magnitude of transcriptional change varied enormously across infections, with Nippostrongylus and Yersinia producing the most robust responses and murine norovirus and Candida the weakest. At stringent fold-change thresholds, intestinal epithelial cells, including enterocytes, stem cells, transit-amplifying cells and goblet cells, together with fibroblasts, emerged as major drivers of the innate response across all infections. Strikingly, the genes induced in stem cells, transit-amplifying cells and enterocytes showed remarkably little overlap between infections, indicating that the epithelium mounts largely pathogen-specific transcriptional programs rather than a generic alarm response.

To understand these programs, the researchers exploited the fact that their sample preparation captured the full developmental lifespan of the enterocyte, from Lgr5-positive stem cells in the crypts through transit-amplifying cells to mature absorptive cells at the villus tip. Pseudotime trajectory inference, validated against well-established marker genes, allowed them to map gene expression along this differentiation axis. Among 4,278 genes whose expression changes along the trajectory in naive mice, 82 encode solute carrier transporters, the transmembrane proteins on apical and basal enterocyte surfaces that move nutrients and ions. Twenty-nine of these transporters, 35 percent, were dysregulated in at least one infection. Some, such as the nucleoside transporter Slc28a2 and the taurine transporter Slc6a6, retained their developmental timing but were downregulated by Yersinia and Nippostrongylus, while others, including the phosphate symporter Slc20a1 and the mitochondrial glutathione transporter Slc25a39, showed infection-altered timing of expression, suggesting either spatially distinct niches along the villus or a broader shift in enterocyte state.

The helminth model revealed how deeply infection can reach into epithelial metabolism. Scoring metabolic tasks with the scCellfie framework showed that enterocytes from Nippostrongylus-infected mice had the most pronounced alterations of any condition, concentrated in lipid metabolism, with reduced synthesis of secondary bile acids and altered handling of fatty acids such as palmitate, linoleate and arachidonate. Cell-cell communication analysis predicted that these enterocytes were receiving interleukin-4 and interleukin-13 signals, the canonical type 2 cytokines, and were producing fibroblast growth factor family ligands, including Fgf15, a key regulator of bile acid and cholesterol metabolism. Mass spectrometry confirmed the biology: epithelial scrapings, but not luminal contents, showed elevated fatty acids in infected mice, and the accumulation correlated with worm burden, linking pathogen load directly to metabolic disruption of the tissue.

Yersinia pseudotuberculosis produced an even more distinctive epithelial phenotype. Clustering imbalance analysis showed that only Yersinia, among all six models, generated a unique enterocyte transcriptional state, which the authors named Yps enterocytes. Pseudotime modeling revealed that Yersinia elicited more temporally regulated differentially expressed genes than all other infections combined, and the 132 genes defining Yps enterocytes included acute-phase proteins such as Saa1 and Saa2, the antimicrobial effectors Nos2 and Reg3a, and a strong enrichment of immune defense ontology terms. Transcription factor motif analysis pointed to STAT3, a downstream effector of many cytokine cascades, and to SREBF1, a regulator of fatty acid synthesis, with the STAT3 signature being specific to the Yps population.

Spatial transcriptomics then connected this state to tissue pathology. Yersinia infection produces dense inflammatory foci called pyogranulomas, rich in bacteria, neutrophils and monocytes. Using the 10x Genomics Xenium platform with a 5,000-gene panel, the team profiled more than 680,000 spatially resolved cells and computationally unrolled intestinal Swiss rolls into a linear proximal-to-distal axis. Yps enterocyte gene expression peaked precisely at pyogranuloma locations, and cell-cell communication predictions identified an IL-22/STAT3/IL-18 signaling axis: IL-22 expression peaked at the pyogranuloma, STAT3-activated enterocytes produced IL-18, and IL-18 receptor signaling was predicted to reach innate lymphoid cells, natural killer cells and natural killer T cells within the inflammatory focus. Stereo-seq analysis showed that Yersinia transcripts remained confined to the base of the pyogranuloma and did not extend into the overlying enterocytes, and immunohistochemistry confirmed phospho-STAT1 activation at pyogranulomas and broader phospho-STAT3 activation across the tissue, supporting a model in which immune-derived cytokines, rather than direct bacterial contact, drive the epithelial phenotype.

Perhaps the most conceptually important observation is that some infection-induced programs ripple far beyond the sites of heaviest pathogen burden. Genes such as Nos2 and Hif1a were restricted to enterocytes immediately above pyogranulomas, but Saa1, Stat3 and the iron exporter Slc40a1 were elevated across the entire ileum. A parallel ripple appeared in the helminth model, where a proximal intestinal infection remodeled the distal ileum. The authors caution that their atlas captures a single early time point per infection and that fragile granulocytes may be underrepresented, but the resource itself, freely queryable through CELLxGENE at gutpath.org with all raw data and code available, offers the immunology community a reference for label transfer, hypothesis generation and the study of how focal infections and inflammatory lesions, including those of Crohn’s disease, broadcast their effects across the intestinal landscape.

Subject of Research: Single-cell transcriptomic profiling of intestinal epithelial and immune cell responses to diverse enteric infections in mice

Article Title: Diverse infections transcriptionally reprogram the intestinal epithelium and epithelial–immune cell interactions

Article References: Hart, A., Merolle, M., Howard, C., Haskins, B. E., Cohn, I. S., Bobba, S., Xiao, R., Yang, Y., Cadwell, K., Ma, J., Yano, H., Hou, X., Wallbank, B. A., Cutillo, D., Ivanov, I. I., Striepen, B., Shin, S., Brodsky, I. E., Artis, D., … Artis, D. (2026). Diverse infections transcriptionally reprogram the intestinal epithelium and epithelial–immune cell interactions. Nature Immunology. https://doi.org/10.1038/s41590-026-02665-6

Image Credits: AI Generated

DOI: 10.1038/s41590-026-02665-6

Keywords: GutPath, single-cell RNA sequencing, CITE-seq, intestinal epithelium, ileum, enterocytes, Yersinia pseudotuberculosis, Nippostrongylus brasiliensis, Cryptosporidium, mucosal immunology, spatial transcriptomics, cell-cell communication