Publishable summary

Investigation of Novel biomarkers and Definition of the role of the microbiome In Graves’ Orbitopathy

INDIGO [Grant No. 612116] is a Marie Sklodowska Curie funded Industry & Academia Partnerships & Pathways project comprising 3 universities and 2 SMEs. The project is coordinated by Cardiff University; other academic partners are University of Milan and University of Essen. The 2 SME partners are Cultech and PTP.

Graves’ orbitopathy (GO), also known as thyroid eye disease, is a complication of Graves’ disease (hyperthyroidism) and both have an autoimmune pathogenesis. GO treatment remains unsatisfactory and the majority of patients report long-term impairment of quality of life. To improve GO outcomes and thus reduce long-term illness and cost to society, research is needed to identify risk factors, develop a better understanding of the pathophysiology of the disease, devise approaches for early diagnosis during the pre-clinical stage of the disease, and create novel and safe interventions. Whilst progress has been made in understanding the end-stages of disease, i.e. the tissue remodelling responsible for GO signs and symptoms, little is known about the processes causing loss of tolerance to target antigens, mainly the thyrotropin receptor (TSHR). Although several mechanisms for triggering autoimmunity by microorganisms have been suggested e.g. molecular mimicry, more recently the gut microbiota have been implicated in maintaining the balance between inflammatory Th17 and non-inflammatory Treg in the gut-associated lymphoid tissue. Technological advances make it possible to sequence the microbial 16S rRNA genes, facilitating the description of bacterial communities, rapidly and cheaply. INDIGO will test the hypothesis that in GD patients bacteria inducing tolerance (Treg) are under-represented or those generating a pro-inflammatory cytokine milieu (Th17) are over-represented (please see attachments for figures illustrating dysbiosis and its effect on the immune system).

To explore this hypothesis INDIGO applies four complementary approaches:

– Analysis of the gut microbiome of patients with GD and GO, and comparison with that of healthy individuals, using 16S rRNA gene sequencing and next generation sequencing;

– Clinical trial of a probiotic formulation in GD patients to investigate whether the microbiome can be modified;

– Development of in vitro models comprising gut epithelium and immune cells plus micro-organisms from GD/GO patients followed by analysis of cytokines released;

– Modification of the gut microbiome of mice, using micro-organisms from GD/GO patients, to improve and extend an in vivo model of GO and be a prototype for similar studies in other autoimmune disorders.

In addition, INDIGO is using several methodologies to identify novel biomarkers to identify GD patients most likely to develop GO and hence inform the early selection of the most appropriate treatment regimen: high throughput analysis of microRNA (miRNA) in serum/plasma, proteomics analyses of tears and serum and a comparison of antibody responses in GD/GO patients and controls to microbial or food derived antigens to determine whether these environmental triggers are involved in GD or associated with GO progression.

The project began in May 2014 and comprises 9 work packages (WPs), 2 (WP1 & 9) are dedicated to management, training, dissemination and consolidation; 3 (WP2, 3 & 4) are focused on generation of samples from patients, in vitro and in vivo models for analysis in the other 3 (WP 6,7 & 8) whilst WP5 has produced probiotics and contrabiotics for use in WP2, 3 & 4 to assess their biological effects. The project will end in April 2018.

WP2 has applied in vitro gut models to investigate 2 questions; 1) Does the gut microbiota affect cytokine production likely to alter the Th17/Treg balance? 2) Is gut permeability altered by Graves’ disease? The first question used co-cultures of Caco2: Human Epithelial Colorectal Adenocarcinoma Cell line and THP1 monocytes supplemented by faecal water (FW) from GD/GO patients and compared with healthy controls. Cytokines produced were measured using ELISA and results to date indicate an increase in IL1 beta in GD/GO compared with healthy controls but this needs to be confirmed. To answer the second question we measured the trans-epithelial electrical resistance (TEER) on Caco2 cells cultured in the presence of thyroxine (to mimic hyperthyroidism), methimazole (treatment for GD) and FW (represents the gut metabonome and proteome, mainly from the microbiota). Data are still being analysed, but methimazole and thyroxine display dose dependent increase and decrease in TEER respectively indicating that hyperthyroidism increases gut permeability, independently of the microbiota, whilst standard treatment for GD corrects this change in permeability.

WP3 recruits GD/GO patients and healthy controls for the provision of samples for analysis in WP6, 7 & 8 (observational study) and also for the probiotic trial. Patients with recent onset GD receive standard block & replace treatment and are randomized to receive the probiotic LAB4 (2 species of lactobacilli and 2 of bifidobacteria) or placebo. Samples collected include faeces, blood, tears and nasal swabs at baseline, when euthyroidism has been achieved and when the patient remits or relapses for both the observation study and the probiotic trial. WP3 will be completed late in 2016.

WP4 has established an animal model of GD/GO in Essen in which female BALBc mice are immunized with an expression plasmid for the TSHR A subunit using in vivo electroporation. A proportion of immunized mice develop a GO-like disease and comparison with the same model induced in London indicates some differences e.g. hyperthyroidism induced in London but not in Essen (Berchner-Pfannschmidt et al Endocrinology 2016 157:1673). Furthermore there are differences in their gut microbiota (please see WP8). Experiments are underway in which the model has been repeated but including strategies to modify the gut microbiota by treating with probiotic, contrabiotic (see WP5) or antibiotic.

WP5 has produced LAB4 probiotic for use in WP2 and for administering to patients and mice. The WP has prepared large scale cultures from patients with severe GO, obtained from UK Essen, for incorporation into the contrabiotic to be used to modify the GD/GO animal model. WP5 also included extensive microbiological investigation of human and mouse faecal samples on more than 10 different agar types and 6 different dilutions. Significant differences were found in the human samples with lactobacilli and total anaerobes higher in GD than controls whilst staphylococci were significantly lower in GD than controls. There were significantly higher levels of yeast in the GO samples used for the contrabiotic (yeasts in 4/6 samples for contrabiotic versus in 7/36 total GD, GO, control and follow up samples, p=0.0321, two-tailed-value ).

WP6 is completing sample collection and applying high-throughput methods to perform LC/MS/MS proteomic profiling of patients’ blood serum and next generation sequencing (NGS) to analyse miRNA in patients’ blood plasma. In addition it will compare the metabonome of FW from GD/GO and healthy controls using LC/MS/MS proteomics  and NMR spectroscopy.

WP7 utilizes ELISA technology to investigate antibody responses of GD/GO patients and controls to bacterial and food antigens. Significant responses were obtained in GD patients to Yersinia enterocolitica but not to Salmonella and Escherichia species. A combined ELISA kit will be used to assess the immune response of sera against 40 different food antigens organized into 21 groups. In preliminary experiments cow’s milk and egg white were most frequently recognized by patients’ sera but numbers studied need to be increased for both bacterial and food antigens.

WP8 applies next generation sequencing of 16S rRNA genes to identify the bacterial species present in the gut microbiota of mouse and human samples. We compared the gut microbiota of immunized TSHR BALB/c mice established in Essen and London, described above, to observe whether the gut microbiota may have an impact on the GO preclinical mouse model in different laboratories. We conducted a comparison using the estimator of richness and diversity (also known as alpha-diversity) of the microbial communities. beta-diversity, or the change in the community composition of samples, is analysed via principal component analysis (PCA) and Non-Metric Dimensional Scaling (NMDS). We observed significant differences in the gut microbiome of animals both at alpha and beta-diversity, possibly explaining differences observed in the induced disease. The gut microbiota was also analysed in endpoint comparison among treatment groups within the Essen cohort, comprised of TSHR or beta-gal control immunized mice and naïve untreated mice as background control. No significant differences were observed in alpha-diversity but we observed a spatial shift of the TSHR immunized mice bacterial communities (Benjamini-Hochberg (BH) adjusted permutation test with 999 permutations p<0.01). In humans, preliminary microbiome analysis identified no significant differences between the control and GD microbiota composition. However, for 2 GD patients, who developed GO, there was a dramatic shift in their microbiota, with the common feature being a decrease in the genus Bacteroides (BH adjusted p<0.0001). This NGS approach confirmed WP2 findings using traditional microbiology techniques.

In conclusion, INDIGO will increase understanding of the mechanisms leading to loss of tolerance and an autoimmune response. The findings will be applicable not only to GO, which affects approximately 3 million people in Europe with an estimated socioeconomic burden of 6.4 billion euros per annum, but also to other autoimmune conditions.