P030 Gut-host metabolites and microbiota interactions in a murine model of checkpoint inhibitor induced colitis

Abstract

Abstract Background The advent of Immune Checkpoint Inhibitors (CPIs) has revolutionised cancer treatment by bolstering anti-tumour immune responses. However, the clinical success of CPIs is hampered by immune-related adverse events (irAEs), with colitis posing a common and formidable challenge. Despite this, the precise mechanisms triggering CPI-induced colitis remain elusive. Recognising the influence of gut microbiota on CPI therapy response and colitis development, this study seeks to unravel the intricate details of these mechanisms to facilitate the development of targeted therapies. Methods Utilising a murine model of CPI-induced colitis, we combined CPI therapy with faecal microbial transplant (FMT) of colitogenic bacteria. FMT induced colitogenic bacteria in wild-type mice, followed by weekly injections of anti-CTLA4 and anti-PD1 for 3 weeks. Metataxonomic data, obtained through 16S rRNA gene amplicon sequencing, and metabolomic profiles, captured by 1H-NMR spectroscopy, were analysed. Alpha and beta diversity, as well as differential expression of taxa, were assessed using 16S data. Metabolomics data underwent multivariate and univariate analyses to identify metabolic profiles and differential metabolites. Pathway analysis was conducted with MetaboAnalyst. Correlation analyses were performed between bacterial taxa at the family level and metabolite concentrations. Results After one week of FMT and CPI treatment, distinct profiles in faecal metabolome and microbiome were observed through Principal Component Analysis (PCA) and Principal Coordinates Analysis (PCoA) for beta-diversity, respectively. Altered metabolites included increased trimethylamine, methylamine, pyruvate, and cytidine, along with reduced tryptophan and proline. Changes in bacterial composition were noted, with increased Bacteroides, Faecalibaculum, Mucispirillum, and Roseburia, and decreased Muribaculum. Metabolic differences persisted after 3 weeks of treatment, revealing additionally decreased acetate, propionate and tyramine. Pathway analysis highlighted profound alterations in tyrosine, alanine, aspartate and glutamate metabolism, pyrimidine metabolism and glycolysis. Desulfovibrionaceae was positively correlated with dimethylamine, glycine and valine. Bacteroidaceae was positively correlated with propionate and butyrate. Conclusion This study provides novel mechanistic insights into the metabolite and microbiota-metabolism pathways associated with CPI-induced colitis. The alterations in short-chain fatty acids (SCFA), amino acids, and the SCFA-producing bacteria and colitogenic bacteria shed light on the metabolic changes underlying CPI colitis. Targeting these microbiomes and metabolites could offer promising avenues for enhancing clinical outcomes during CPI therapy.