P737 Host-microbiome interactions in primary sclerosing cholangitis
J. Sabino*1, S. Vieira-Silva2, 3, I. Cleynen4, K. Machiels1, M. Joossens2, 3, 5, G. Falony2, 3, J. Wang2, 3, V. Ballet1, M. Ferrante1, G. Van Assche1, S. Van Der Merwe6, J. Raes2, 3, S. Vermeire1
1KU Leuven, Department of Clinical and Experimental Medicine, Translational Research Centre for Gastrointestinal Disorders (TARGID), Leuven, Belgium, 2REGA institute, KU Leuven, Department of Microbiology and Immunology, Leuven, Belgium, 3Vlaams Instituut voor Biotechnologie (VIB), Centre for the Biology of Disease, Leuven, Belgium, 4KU Leuven, Human Genetics, Leuven, Belgium, 5VUB, Department of Microbiology, Brussels, Belgium, 6University Hospitals of Leuven, Department of Hepatology, Leuven, Belgium
Primary sclerosing cholangitis (PSC) is a cholestatic liver disease, frequently associated with inflammatory bowel disease (IBD). The pathogenesis of PSC remains greatly unknown but 16 risk loci have been associated with PSC, and some of these overlap with known risk loci for IBD. As we recently showed that dysbiosis is a feature of patients with PSC, we assessed the influence of genetic risk factors for PSC or IBD on the intestinal microbiota.
Blood and faecal samples from 52 PSC patients were collected, and 41 patients had concomitant IBD (17 Crohn’s disease [CD] and 24 ulcerative colitis [UC]). A genetic risk score (GRS) was calculated for each patient, taking into account the risk allele frequency and odds ratio of each single nucleotide polymorphism (SNP). The GRS were calculated using 14 risk SNPs for PSC, 197 risk SNPs for CD and 191 risk SNPs for UC as generated using Immunochip. Second, a GRS only containing those SNPs implicated in microbiota-interaction was calculated for CD and UC. Further, 16S rDNA paired-end sequencing targeting the V4 hypervariable region was performed using the Illumina MiSeq sequencer. Sequencing depth was downsized to 10 000 reads/sample. The Ribosomal Database Project classifier was used for taxonomic assignment. Statistical analyses were performed with R, using parametric and non-parametric tests, with multiple testing correction (FDR). Correlation between genera abundances and GRS was performed with Spearman correlation.
Although microbial richness did not differ between patients with a high (n = 26) or low (n = 26) PSC GRS, and no differences were observed in the abundance of specific phyla, genera, or operational taxonomic units (OTU), the overall microbiota composition was significantly different between the 2 groups (Adonis test on Bray-Curtis dissimilarity, p-value 0.002). Interesting to note, the abundances of 9 genera were significantly different between the 2 groups before multiple testing correction, including Prevotella, Dorea, Methanosfera, and Paraprevotella. No significant association was observed between CD or UC GRS quartiles and microbiota composition, richness, or taxa abundances in this cohort.
The genetic risk for PSC seems to influence the overall gut microbiota composition in patients with PSC, however no specific significant differences in microbial diversity or differences of taxa abundances were observed between patients with high and low GRS for PSC. The cumulative effect of multiple small differences at genus level may explain the overall significant difference in community composition. Larger sample size might be needed to evidence these differences. Contrarily to PSC GRS, the CD or UC GRS do not seem to influence intestinal microbiota in this population.