Home > Gastroenterology > ECCO 2020 > Big Data Analysis > Multi-omics help describe CD phenotypes

Multi-omics help describe CD phenotypes

Presented by
Dr B. Verstockt, University Hospital Leuven, Belgium
ECCO 2020
By using integrated signature profiles generated from multiple 'omic' datasets, molecular mechanisms were identified which could potentially describe Crohn’s disease (CD) phenotypes, like the occurrence of perianal disease [1].

Dr Bram Verstockt (University Hospital Leuven, Belgium) presented the results of this multi-omic data integration with network analysis. He explained that the molecular mechanisms which orchestrate the heterogeneity of CD are poorly understood.

In a discovery cohort of 98 CD patients, 576 unique proteins were measured in blood. All patients were also genotyped. The 2 resulting datasets were then integrated using an innovative algorithm called Multi-Omics Factor Analysis (MOFA). From this analysis, 5 so-called latent factors (LFs: representative variables capturing the sources of variation in omic datasets) associated with at least 1 clinical phenotype were identified. Clustering patients along LFs achieved meaningful separation of clinical phenotypes like perianal penetrating disease.

The top-ranking proteins associated with perianal disease were involved in inflammatory pathways and autophagy, or were already known to be implicated in CD. Dr Verstockt: “Many of the genes encoding these proteins are significantly dysregulated in draining perianal fistula, as compared to paired rectal mucosa, in independent patients. The majority of the differentially expressed serum proteins (77.8%) could be validated in an independent cohort of 88 newly diagnosed CD patients”. Seven mutations mapped to transcription factors (SMAD3, BACH2) and post-translational regulators (such as IFNGR2, IL10, IL2RA, SLC2A4RG, and ZMIZ1) could potentially regulate perianal disease pathophysiology and could thus be considered novel drug targets. Dr Verstockt added that network analysis can also highlight potential repurposing of already existing drugs in other diseases.

In another study, an integrated multi-omics data analysis revealed microbiome-driven proteolysis as a contributing factor to the severity of ulcerative colitis (UC) disease activity. Certain members of the microbiome, such as Bacteroides vulgatus, may contribute to exacerbating disease activity in UC through protease activity. In vivo and in vitro experiments provide evidence that bacterial protease inhibition may be a novel therapeutic approach in UC [2].

A third multi-omics analysis revealed specific bio-geographical and functional characteristics in inflammatory bowel disease (IBD) intestinal mucosa. The results provide a unique and comprehensive cell map of IBD in a location-specific context, potentially shedding light upon unexplained clinical phenomena [3]. For example, an upregulated IL-6 pathway in Treg cells was recognised as sigmoid-specific in UC patients, which may help explain colonic perforations associated with anti-IL-6R treatment. The results may allow one to tailor therapies to affected areas to improve treatment efficacy.

      1. Sudhakar P, et al. ECCO-IBD 2020, OP15.

      2. Mills R, et al. ECCO-IBD 2020, OP31.

      3. Maimon N, et al. ECCO-IBD 2020, OP33.

Posted on