Abstract
Background
Irritable bowel syndrome (IBS) patients are subtyped by predominant bowel habit rather than pathophysiological mechanisms and this may underlie challenges in identifying more effective targets for designing new treatments. Metabolomics and microbial analysis can distinguish IBS patients from healthy controls but it is unknown if they can identify pathophysiological IBS subgroups.
Aims
To phenotype subgroups of IBS patients using metabolomics and microbial analysis and determine if these subgroups have different underlying pain signalling mechanisms.
Methods
Symptom history, stool and urine were collected from 30 diarrhea predominant (IBS-D) and 30 constipation predominant (IBS-C) IBS patients (Rome IV). Liquid Chromatography-Mass Spectrometry quantified 130 metabolites in stool and urine. The GA-map™ Dysbiosis Test targeting ≥300 bacteria on different taxonomic levels was used to identify stool microbial composition. Multivariate OPLS discriminatory analysis assessed metabolomics and microbial profiles. To assess potential effects on pain signalling, the effect of stool supernatant on dissociated dorsal root ganglia (DRG) neuron responses to capsaicin (10nM) was assessed using Ca2+ imaging.
Results
Within both IBS-D and IBS-C, combined stool/urine metabolomic profiles of patients with a dysbiosis-like (DL) IBS (onset following antibiotics, enteric infection, or travel) were distinct from patients with a non-DL IBS onset (IBS-D R2=0.7, Q2=0.5; IBS-C R2=0.5, Q2=0.4); fecal glutamic acid and urinary pyruvic acid were the main metabolites driving separation. However, microbial profiles of DL vs non-DL onset could only be discriminated in IBS-D (R2=0.8, Q2=0.4). In the patients with a DL IBS onset, stool metabolomic profiles of the 7 IBS-C discriminated from the 8 IBS-D patients (R2=0.9, Q2=0.8). Profile differences were not seen between IBS-C and IBS-D with a non-DL onset of IBS. In preliminary studies, incubation of DRG neurons with stool supernatant from 1 DL IBS-D and 1 non-DL IBS-D increased peak [Ca2+]i responses to capsaicin compared to incubation with media (DL: 5.5±0.9 vs 2.3±0.7; non-DL: 6.9±0.7 vs 3.9±0.4% ΔF/F). Similarly, the number of responsive neurons to capsaicin was increased after incubation with IBS stool supernatant vs media (DL: 27% vs 8%; non-DL 19% vs 12%).
Conclusions
Different metabolomic and bacterial profiles between DL and non-DL onset of IBS-D suggests a novel means to better phenotype clinically defined IBS subgroups. While initial results with stool supernatants from both a DL and non-DL IBS-D patient suggest increased pain signalling in DRG neurons, more studies are needed to determine if there are differences between these two subgroups as well as healthy controls.
Funding Agencies
CIHRSoutheastern Ontario Academic Medical Organization (SEAMO)
Comparison of the antibacterial activity of essential oils and extracts of medicinal and culinary herbs to investigate potential new treatments for irritable bowel syndrome
