P018 Biologic functions between non-affected and affected margins of ileum samples from Crohn’s disease patients who will develop post-surgical recurrence.
V. Loren*1, 2, A. Garcia-Jaraquemada1, A. Aterido3, F. Kamberovic1, A. Julià3, Y. Zabana4, A. Alonso3, M. Mañosa2, 5, S. Marsal3, E. Domènech2, 5, E. Cabré2, 5, J. Manyé2, 6
1Health Sciences Research Institute of the ‘Germans Trias i Pujol’ Foundation (IGTP), Inflammatory Bowel Disease Unit, Badalona, Spain, 2Centro de Investigación Biomédica en Red (CIBER), Madrid, Spain, 3Institute for research in Biomedicine, Rheumatology Research Group, Barcelona, Spain, 4Mútua de Terrassa Hospital, Terrassa, Spain, 5Germans Trias i Pujol Hospital, Inflammatory Bowel Disease Unit, Badalona, Spain, 6Health Sciences Research Institute of the ‘Germans Trias i Pujol’ Foundation (IGTP), Inflammatory Bowel Disease Unit, Badalona, Spain
Transcriptome profile from ileocolectomy performed on non-affected margin of the ileum of the Crohn’s disease (CD) patients showed changes in NOTCH signalling that could be implicated in early post-surgical recurrence (PSR) (Zabana, 2015). Comparing transcriptome profile from inflamed vs healthy margin could reveal new knowledge about the CD recurrence.
Objectives: To identify specific biologic functions between non-affected and affected margins of surgical samples from CD patients which will develop recurrence >18 months of the surgery.
Material and methods: Transcriptomic profile study (CodeLink© microarrays, USA) amongst non-affected and affected margins of ileocolestomy from CD patients with PSR (n = 5) or not PSR (n = 10) were performed; 10 controls were also included in this study and compared with different intestinal margins. Raw data from microarray analysis underwent quality control, normalisation, and clustering. Next analysis of gene differentiation was done with the following statistical criteria: FDR < 0.05 or FDR < 0.01 and abs(log[FoldChange]) > 2. The most relevant genes were grouped in co-expression modules using WCGNA (selection power = 13, r2 > 0.92). GeneSet databases were used to perform enrichment analysis to screen biological and molecular functions in each module. The modules linked to recurrent or not recurrent CD were identified based on high correlation (abs[r2] > 0.75; p < 0.05).
After quality-control procedures, 53 samples and 20 902 gene expression probes mapping 1 gene were identified and included in further analysis. No single gene expression probe was identified when recurrent and non-recurrent patients within the inflamed or non-affected margins were compared. However, 222 probes (69 with FDR < 0.01) were identified amongst non-affected and inflamed margins from CD recurrent patients, and 342 probes (95 with FDR < 0.01) showed different expression non-recurrent patients. These individual probes were grouped in 5 modules; 3 of them involved specifically in recurrent patients. Modules related to oxygen levels and protein amidation had a positive correlation between PSR and non-PSR (abs[r2] >v0.96 and >v0.83; p < 0.0003, respectively), whereas centriole–centriole cohesion showed a negative correlation (abs[r2] = -1; p = 10-9).
PSR in CD show highlighted biological functions related to hypoxia and protein amidation, whereas centriole–centriole adhesion displays an antagonistic function compared with non-recurrent patients.