Background: In Graves' disease (GD) and orbitopathy (GO) autoimmune responses to the thyrotropin receptor (TSHR) cause hyperthyroidism and orbital tissue remodelling respectively. Previously we have demonstrated that the gut microbiota contributes to the heterogeneity in a TSHR-induced murine model of GD/GO: we also observed potential disease-associated microbial taxonomies. The present study aimed to modify the gut microbiota to provide additional insights into the complex host-microbiome interplay that underpins TSHR-induced GD/GO. Methods: Female BALB/c mice were treated with TSHR or empty expression plasmids alone or in combination with probiotics (lactobacilli & bifidobacteria) or faecal material transplant from severe GO patients (hFMT) or long-term treatment with vancomycin. Metataxonomics (16S rRNA gene sequencing) was performed on the intestinal contents of all mice, followed by prediction of gene functions of identified bacteria. RandomForest was used to select operational taxonomic units (OTUs) capable of discriminating among treatments and immunisations. Induced disease was assessed by measurement of serum TSHR-autoantibodies (TRAB) & T4, orbital histology and T cell-subset analysis. Results: TRAB were induced in all TSHR treated mice and accompanied by significant hyperthyroidism in the TSHR-alone and TSHR-probiotic groups. GO-like orbital histology was also apparent in these two groups in contrast to TSHR-hFMT and TSHR-vancomycin treated mice in which these features were reduced or absent respectively. We observed a significant reduction in richness and diversity (alpha-diversity) of the gut microbiota and an increase in the phylum Proteobacteria in the vancomycin treated mice which also showed a unique microbiota (Out-Of-Bag error 0%), previously described as associated with long-term antibiotic treatment. There was little variability in the microbiota of the plasmid-only, probiotic and hFMT, treated mice (OOB error rate 20%). We described at least 15 OTUs capable of discriminating between the probiotic, hFMT and plasmid-only groups; e.g the Lachnospiraceae UCG-008 family was more abundant in the plasmid-only compared to probiotic and hFMT mice. Between the two immunisation groups (TSHR or empty plasmid), the most discriminating OTU belongs to the genus Bacteroides, which we previously observed to be reduced in TSHR-immunized mice developing signs of GO, and also during the GD to GO transition in patients. Conclusions: Treatment with probiotics, hFMT or antibiotics all acted to modify the gut microbiota with the most extreme effects being observed with vancomycin. The lack of induced disease in the antibiotic group indicates a role for the gut microbiota in TSHR-induced disease but whether it is due to the absence of specific OTUs or to immune effectors is currently under investigation. The current data confirm our previous results from observational studies on the gut microbiota in both GD/GO patients and GO murine models. The SourceTracker Bayesian model will be used to calculate the direction and the success of the invasion of the hFMT, using the metataxonomic profile of the human donors' faeces and the plasmid-only mice as predictive sources. Further analyses will test the predicted gene functions of the microbiota identified and metataxonomics data will be correlated with the different disease phenotypes observed.
Unravelling the functional role of the gut microbiota in a murine model of Graves' Orbitopathy
F Biscarini;
2017
Abstract
Background: In Graves' disease (GD) and orbitopathy (GO) autoimmune responses to the thyrotropin receptor (TSHR) cause hyperthyroidism and orbital tissue remodelling respectively. Previously we have demonstrated that the gut microbiota contributes to the heterogeneity in a TSHR-induced murine model of GD/GO: we also observed potential disease-associated microbial taxonomies. The present study aimed to modify the gut microbiota to provide additional insights into the complex host-microbiome interplay that underpins TSHR-induced GD/GO. Methods: Female BALB/c mice were treated with TSHR or empty expression plasmids alone or in combination with probiotics (lactobacilli & bifidobacteria) or faecal material transplant from severe GO patients (hFMT) or long-term treatment with vancomycin. Metataxonomics (16S rRNA gene sequencing) was performed on the intestinal contents of all mice, followed by prediction of gene functions of identified bacteria. RandomForest was used to select operational taxonomic units (OTUs) capable of discriminating among treatments and immunisations. Induced disease was assessed by measurement of serum TSHR-autoantibodies (TRAB) & T4, orbital histology and T cell-subset analysis. Results: TRAB were induced in all TSHR treated mice and accompanied by significant hyperthyroidism in the TSHR-alone and TSHR-probiotic groups. GO-like orbital histology was also apparent in these two groups in contrast to TSHR-hFMT and TSHR-vancomycin treated mice in which these features were reduced or absent respectively. We observed a significant reduction in richness and diversity (alpha-diversity) of the gut microbiota and an increase in the phylum Proteobacteria in the vancomycin treated mice which also showed a unique microbiota (Out-Of-Bag error 0%), previously described as associated with long-term antibiotic treatment. There was little variability in the microbiota of the plasmid-only, probiotic and hFMT, treated mice (OOB error rate 20%). We described at least 15 OTUs capable of discriminating between the probiotic, hFMT and plasmid-only groups; e.g the Lachnospiraceae UCG-008 family was more abundant in the plasmid-only compared to probiotic and hFMT mice. Between the two immunisation groups (TSHR or empty plasmid), the most discriminating OTU belongs to the genus Bacteroides, which we previously observed to be reduced in TSHR-immunized mice developing signs of GO, and also during the GD to GO transition in patients. Conclusions: Treatment with probiotics, hFMT or antibiotics all acted to modify the gut microbiota with the most extreme effects being observed with vancomycin. The lack of induced disease in the antibiotic group indicates a role for the gut microbiota in TSHR-induced disease but whether it is due to the absence of specific OTUs or to immune effectors is currently under investigation. The current data confirm our previous results from observational studies on the gut microbiota in both GD/GO patients and GO murine models. The SourceTracker Bayesian model will be used to calculate the direction and the success of the invasion of the hFMT, using the metataxonomic profile of the human donors' faeces and the plasmid-only mice as predictive sources. Further analyses will test the predicted gene functions of the microbiota identified and metataxonomics data will be correlated with the different disease phenotypes observed.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
