Search in the Abstract Database

Abstracts Search 2017

* = Presenting author

P774 Metagenomics and metabolomics of patients with inflammatory bowel disease and their unaffected relatives

Vancamelbeke M.*1, Sabino J.1, Deroover L.1, Vandermeulen G.1, Luypaerts A.1, Ferrante M.1,2, Falony G.3, Vieira-Silva S.3, Verbeke K.1, Raes J.3, Cleynen I.4, Vermeire S.1,2

1KU Leuven, Department of Clinical and Experimental Medicine, Translational Research Center for Gastrointestinal Disorders, Leuven, Belgium 2University Hospitals Leuven, Department of Gastroenterology and Hepatology, Leuven, Belgium 3Rega Institute for Medical Research, VIB, KU Leuven, Department of Microbiology and Immunology, Laboratory of Molecular Bacteriology, Leuven, Belgium 4KU Leuven, Department of Human Genetics, Laboratory for Complex Genetics, Leuven, Belgium

Background

Dysbiosis, intestinal barrier dysfunction and metabolic alterations of the gut microbiota have been implicated in the pathogenesis of inflammatory bowel disease. We studied the faecal microbiome and metabolome, as well as intestinal permeability of multiple-affected families with Crohn's disease (CD) or ulcerative colitis (UC) to investigate which factors are associated with disease.

Methods

Faecal and urine samples were obtained from 84 individuals of 19 families (37 CD, 11 UC and 36 unaffected first-degree relatives (FDR)). Faecal microbial profiling was done using 16S rDNA paired-end sequencing (Illumina MiSeq). Sequencing depth was downsized to 10,000 reads/sample. Taxonomic annotation was performed with the RDP classifier. Faecal volatile organic metabolites were measured using GC-MS. Metabolite data were relatively quantified to an internal standard, and subject-specific compounds were discarded. Metabolite profiles were clustered by PLS-DA (Unscrambler). Small intestinal permeability (IP) was measured using a 2-hour lactulose-mannitol urine test. Statistical analyses were conducted in R with multiple testing correction (Benjamini-Hochberg).

Results

Microbial richness and composition were significantly different in patients with CD compared to UC and FDR (p<0.05), whereas these comparisons were not significant for UC versus FDR. Vector fitting confirmed diagnosis as the main driver of the variability in microbial composition (p<0.001), followed by family ID (p=0.02). The genera discriminating CD and FDR included 16 known and new genera, such as Faecalibacterium, Ruminococcus and Gemmiger (corrected p<0.05). Analysis of the metabolites also showed separate clusters for CD and FDR, while samples from UC patients partially overlapped with both groups. In contrast to the microbiota results, family did not significantly drive the metabolic profiles. The chemical classes associated with FDR were short- and medium-chain fatty acids, while samples from CD patients were associated with esters. Comparison of individual metabolites identified eight compounds with significantly different levels for CD versus FDR (corrected p<0.05). Among these, acetic acid and butyric acid are known for their anti-inflammatory properties and beneficial effect on gut barrier function. A subset of CD patients (30%) had increased small IP values, but this trait was not associated with any of the individual metabolites, nor bacterial genera.

Conclusion

Significantly different metagenomic and metabolomic profiles were observed between CD patients and healthy individuals with a shared familial background. Faecalibacterium, Ruminococcus and Gemmiger genera, amongst others, drive the phenotype of CD, as do esters and lower levels of short-chain fatty acids.