Abstract Background As a dietary approach to reducing inflammation in ulcerative colitis (UC), the 4-SURE diet was designed to correct pathogenic alterations of excessive protein fermentation and hydrogen sulphide (H2S) production in the distal colon. Specific dietary objectives were achieved (Day et al,J Nutr 2022;152:1690) but it is uncertain whether mechanistic objectives could be achieved with 8 weeks of diet. Therefore, we aimed to perform a deep functional analysis (microbial and metabolomic) of the faeces. Methods Faecal samples of 28 adults with mild-moderately active UC were collected at week 0 and 8 of diet intervention, subsampled and stored at−80 °C. Shotgun metagenomic sequencing was used to identify genes involved in H2S metabolism. Metagenomic reads were trimmed, reads mapping to the human genome removed. Gene identification was performed using Diamond v2.1.9 for alignment against a database of genes involved in global sulphur cycling. Gas-chromatography mass-spectrometry was used to characterise volatile organic compounds (VOCs) with specific analysis of products of protein fermentation. Microbiota capacity to produce H2S was assessed by anaerobic incubation of homogenates followed by spectrophotometric determination of total H2S production. Results Majority of microbiome members belonged to the Bacteroidota and Bacillota phyla, with no significant difference in bacterial diversity determined between time points (p=0.16). However, using a Random Forest Classifier, known H2S producers, Odoribacter and Peptostreptococcaceae, were identified as the most important taxa in discriminating between pre and post-diet samples, with abundances markedly lower after diet intervention. A shift in microbiota metagenomic profiles putatively involved in H2S metabolism was identified post-diet, with differences (p<0.05, Wilcoxon signed-rank with Benjamini-Hochberg correction) in 12 of 23 analysed genes involved in H2S cycling determined, including iscS, cysK (Fig), metC, luxS, asrC and sseA. 173 faecal VOCs were identified across all samples. Indole (a specific marker of protein fermentation) decreased from 0.42 [-0.25,0.61] at week 0 to -0.28 [-0.77,0.48] at week 8 (p=0.007, FDR correction; Fig). Faecal H2S reduced from median 2.61 (interquartile range 2–4.96)µmol to 1.41 (0.81-2.27)µmol of H2S/g stool (wet weight)(p=0.002)(Fig). Conclusion Deep functional analysis of the faeces demonstrated the micro-environmental objectives of the 4-SURE diet successfully reduced protein fermentation, changed microbial community sulphur-metabolic gene profile and reduced H2S production ex vivo. Applying functional analysis to a diet study is novel, highlighting exemplar framework for including biomarkers of pathogenic relevance in analysis of therapeutic diets.
Read full abstract