henhaobio
Sign in

bioRxivImmunologyPreprint

Gut microbial interaction networks control autoimmunity to neuroretina

Amy Zhang*, Reiko HoraiiD, Yingyos JittayasothorniD, Jonathan H. BadgeriD, Zhichao Wu, Guangpu ShiiD, Akriti GuptaiD, Samyuktha Arunkumar, Caitlin E. MurphyiD, Vijayaraj NagarajaniD, John A. McCullochiD, Shilpa KodatiiD, Hatice Nida SeniD, Jung Wha LeeiD, Jonathan P. JacobsiD, Xiaoyan XuiD, Mary Joseph MattapalliliD, Zixuan PengiD, Biying XuiD, Robert J. Palmer Jr., Nadim MajdalaniiD, Kenya HondaiD, Colm Ó’hUigíniD, Rachel R CaspiiD

PDFDOICode
In-depth full-text digest available · Jump to digest ↓

Abstract

The gut microbiome influences the development of immune-mediated inflammatory diseases, including autoimmune uveitis, a sight-threatening ocular inflammation driven by retina-specific T cells1. Using a model of spontaneous autoimmune uveitis (sEAU) we showed that gut commensals provide immune stimuli that trigger disease2. Here we report that uveitis-promoting microbes are present in human gut flora and that colonization of germ-free (GF) mice with commensals from healthy human donors was sufficient to provoke disease. Severity of sEAU correlated with expansion of Akkermansia and contraction of short-chain fatty acid (SCFA)-producing Firmicutes, accompanied by decreased SCFA levels and a dominant gut Th1 effector response. Mechanistic gain-of-function experiments, enriching GF sEAU mice with A. muciniphila, reproduced these microbiome, metabolite and immune phenotype shifts, and exacerbated disease. We propose that Akkermansia promotes autoimmunity by outcompeting SCFA producers and enhancing Th1-type responses. An inverse correlation between Akkermansia (Verrucomicrobia) and Firmicutes was also present in fecal microbiota from patients with uveitis; related Verrucomicrobia/Firmicutes shifts were observed in multiple sclerosis and Crohn's disease datasets. These findings reveal a stereotypic gut microbial interaction network that regulates systemic immune balance, and may represent an ecologically conserved mechanism through which the gut microbiome modulates autoimmune and inflammatory diseases.

In-depth summary

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 4 (22 Oct 2026) · CC0 1.0

Akkermansia expansion reduces short-chain fatty acid-producing Firmicutes in human-flora uveitis mice, raising Th1 responses and disease; a Verrucomicrobia/Firmicutes ratio is elevated in uveitis, MS and Crohn's patients.

The paper asks how human gut communities affect spontaneous autoimmune uveitis, using germ-free mice with a retina-specific TCR colonised by healthy-donor flora. It contrasts a network view with earlier studies, which mostly looked at single taxa, and reports that Akkermansia effects depend on community context.

Key results

  1. ShownHealthy-donor human flora supports spontaneous uveitis in germ-free mice, with lower scores than mice carrying mouse flora. Colonic CD4 T cells showed fewer Th1 and Th17 cells and more regulatory T cells than in SPF mice.

    Histology scores differed across groups (Kruskal-Wallis p = 0.0002; U, V, W n = 28, 38, 18; SPF n = 22; GF n = 10). Immune profiling: n = 15 GF, 19 U, 11 V, 11 W, 27 SPF. · 1. Human gut commensals support development of…; Figure legends

  2. SuggestedIn human-flora mice, severe uveitis tracked with Akkermansia, IFN-γ-producing (Th1) CD4 T cells and low SCFA levels. Akkermansia was inversely correlated with SCFA-producing Firmicutes, and a similar negative link appeared in uveitis patient networks.

    Pearson r = 0.5 for Akkermansia and r = -0.3 for certain Firmicutes vs histology (FDR < 0.05, n = 105 mice). Akkermansia vs Firmicutes regression used n = 234 samples. Patients: n = 8 vs 12 controls (SCFA levels: 6 patients). · 2. HuFl sEAU tracks with a dominant Th1…; 3. SCFA signatures are enriched in low disease…; 4. Akkermansia negatively correlates with…; Figure legends

  3. ShownAdding A. muciniphila to a low-Akkermansia human flora worsened uveitis, lowered Firmicutes and serum SCFAs; serum and/or gut IFN-γ was higher in Akkermansia-enriched mice (minimal-flora and HuFl models). An E. coli spike did not reproduce the disease or Firmicutes effect.

    Histology n = 11 Akk vs 7 PBS (2 of 3 experiments); serum SCFAs n = 13 vs 10; serum IFN-γ n = 18 vs 14 (five experiments combined, two of them in the minimal-flora model). In a minimal-diversity model, Akkermansia and Firmicutes had r-squared > 0.99 (p < 0.0001). · 5. Akkermansia enrichment increases uveitis and…; Figure legends

  4. SuggestedAkkermansia did not activate retina-specific T cells in vitro and shared no sequence with the cognate epitope. Colonic scRNA-seq suggested a stronger Th1 programme and more myeloid-to-T-cell signalling in Akkermansia-high mice.

    scRNA-seq used 6 mice (3 per group); about 15% of colonic CD4 T cells expressed the R161H TCR. UMAP and regulon panels show a representative mouse per group. In vitro gut-extract assay: n = 4 (HuFl+AKK) and 3 (HuFl+PBS). · 5. Akkermansia enrichment increases uveitis and…; Single cell RNAseq data analyses; Figure legends

  5. SuggestedThe Verrucomicrobia/Firmicutes (V/F) ratio rose with disease in mice and was higher in uveitis, multiple sclerosis (MS) and Crohn's disease patients vs controls, but not in ulcerative colitis.

    Mann-Whitney U tests, one-tailed for uveitis and MS; Kruskal-Wallis for IBD. No effect sizes or group sizes are given for MS and IBD in the text. · 6. The inverse relationship between…; Figure legends

Methods, data and reproducibility

Germ-free R161H TCR-transgenic mice were colonised with healthy human donor stool, then given Akkermansia or E. coli. Uveitis scoring was combined with 16S and shallow shotgun metagenomics, SCFA metabolomics, flow cytometry and colonic scRNA-seq, plus analysis of patient and public datasets.

  • Gnotobiotic R161H mouse lines carrying flora from three healthy human donors (U, V, W); fecal 16S profiling · mouse · n = U 50, V 59, W 64 fecal samples; histology n = 28, 38, 18 · PRJNA1451962
  • Healthy human donor stool (Keio University) used to colonise germ-free mice · human · n = 3 donors
  • Akkermansia reconstitution experiments in germ-free R161H mice (minimal flora and HuFl-base models) · mouse · n = histology n = 11 vs 7; serum SCFAs n = 13 vs 10
  • Colonic CD4 T cell scRNA-seq from Akk-rich vs control mice · mouse · n = 6 mice (3 per group)
  • NEI treatment-naive autoimmune uveitis patients vs healthy controls (16S and shotgun stool profiling) · human · n = 8 patients, 12 controls
  • Public stool microbiome datasets for inflammatory bowel disease (IBD) and MS (iMSMS) · human · PRJNA1198911
Methods & tools
YAMS16 v1.7, DADA2, SILVA 138.1, phyloseq v1.42.0, MaAsLin2 v1.12.0, NetCoMi, JAMS v1.9.9, HUMAnN v3.9, Cell Ranger v9.0.1, Scanpy v1.11.0, CellChat, SCENIC
Software & environment
Software versions are given in the Methods (R v4.4, Python v3.10, Scanpy v1.11.0, Seurat v5.1.0, SingleR v2.8.0); no containers or requirement files are mentioned.
Reproducibility gaps
No custom code was written; the listed URL is a third-party 16S pipeline. Sequencing data are under BioProject PRJNA1451962; no URL is printed.
Limitations (5)
  • Species- and strain-level resolution of the Akkermansia-Firmicutes relationship is hard to achieve, and how Akkermansia antagonises SCFA producers was not resolved. · stated by the authors
  • Disease develops asynchronously in individual mice and germ-free R161H mice breed slowly, which limits the scale and breadth of experiments. · stated by the authors
  • The model cannot reveal antigenic targets in human uveitis; the authors chose not to identify the HuFl microbial mimic antigen, citing lack of translational relevance. · stated by the authors
  • Human validation rests on small cohorts (8 uveitis patients vs 12 controls for SCFA and network analyses), with one-tailed tests for uveitis and MS V/F comparisons; the uveitis and MS cohorts included patients on treatment. · noted in this summary
  • The scRNA-seq mechanism data come from only 3 mice per group and are partly correlational (CellChat and SCENIC are inferences). · noted in this summary

Read this if you study gut microbiome community interactions in autoimmunity and want gnotobiotic evidence linking Akkermansia, SCFA-producing Firmicutes and Th1 responses.

Citations
1

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISVaella-Alar '26Ma '21Yamada '19Shirahama '24Alexander '23Fasano '26
63% semantic matchImmunology
Maricielo Vaella-Alarcon, Miki Kume, Silvia R. Davila-Paico+6 more·bioRxiv·Oct 19, 2026