Peripheral tolerance to insulin is encoded by mimicry in the microbiome
Rubio Garcia, A.; Paterou, A.; Lee, M.; Slawinski, H.; Wicker, L. S.; Todd, J. A.; Pekalski, M. L.
Show abstract
How organisms achieve sustained peripheral tolerance throughout their lifetime, a correct immune discrimination between self and non-self, remains poorly understood. Host-microbiome interactions carry fundamental information that facilitates this process. We hypothesize that commensal microbes are under evolutionary pressure to develop epitopes that, when presented along with other antigens from their own bacterial community, lead to an overall tolerogenic self classification by the host immune system. Hosts, which have co-evolved with commensals, may rely on mimotopes, bacterial epitopes that are indistinguishable from key self epitopes, as a homeostatic feedback mechanism to establish and maintain tolerance. Using a probabilistic sequence model of peptide mimicry, we show that the gut microbiome contains a set of genes that are likely to trigger identical immune responses to insulin B 9-25, a widely distributed self epitope across tissues and the primary autoantigen in type 1 diabetes. Similarities in the antigen receptor sequences determined from CD4 T cells reacting to insulin epitopes and mimotopes provide experimental evidence for mimicry. All predicted high posterior probability mimotopes belong to the transketolase superfamily, an enzyme that allows efficient harvest of commensal-derived sugar polymers and dietary fibre, an advantage during host colonisation. Microbial transketolase upregulation during infant weaning coincides in time with the peak in autoantibody development against insulin. Abundance changes in bacterial genera that carry these mimotopes have also been observed to precede disease diagnosis. Our findings suggest gut dysbiosis followed by immune response to insulin mimotopes as a primary cause of type 1 diabetes, and may contribute towards unraveling similar causal patterns in a wide variety of disorders.
Matching journals
The top 6 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Single cell genome sequencing of laboratory mouse microbiota improves taxonomic and functional resolution of this model microbial community 94%
- Long-read sequencing to interrogate strain-level variation among adherent-invasive Escherichia coli isolated from human intestinal tissue 93%
- Fecal Virome Transplantation (FVT) from healthy humans remodels the gut bacteriome and virome and reduces metabolic syndrome in mice 93%
Similar papers in this journal
- The host genotype actively shapes its microbiome across generations 94%
- PathoFact: A pipeline for the prediction of virulence factors and antimicrobial resistance genes in metagenomic data 94%
- Functional characterization of helminth-associated Clostridiales reveals covariates of Treg differentiation 93%
Similar papers in this journal
- SPARTA: Interpretable functional classification of microbiomes and detection of hidden cumulative effects. 93%
- Individualized network analysis reveals a link between the gut microbiome, diet intervention and Gestational Diabetes Mellitus 92%
- Providing new insights on the byphasic lifestyle of the predatory bacterium Bdellovibrio bacteriovorus through genome-scale metabolic modeling. 91%
Similar papers in this journal
- Transmembrane helices are an overlooked and evolutionarily conserved source of major histocompatibility complex class I and II epitopes 92%
- A detailed Molecular Network Map and Model of the NLRP3 Inflammasome 90%
- Significant and conflicting correlation of IL-9 with Prevotella and Bacteroides in human colorectal cancer 90%
Similar papers in this journal
- Inference of Causal Interaction Networks of Gut Microbiota Using Transfer Entropy 93%
- Precursor peptide-targeted mining of more than one hundred thousand genomes expands the lanthipeptide natural product family 92%
- Inferring directional relationships in microbial communities using signed Bayesian networks 92%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.