A systems biology approach identifies ANP32E and other biomarkers of glucocorticoid refractoriness in ulcerative colitis [microarray]

BACKGROUND: Glucocorticoid (GC) refractoriness is a common and unpredictable phenomenon in ulcerative colitis (UC). Although it has been demonstrated that NFkB is involved in its mechanism of action (MoA), there are no conclusive studies on the molecular functions involved. The current study describes in depth the MoA related to GC failure, by integrating transcriptomic data from UC patients, and updated molecular data on UC and GC. METHODS: miRNA and mRNA expression from rectal biopsies of active UC patients, obtained before and on the 3rd day of GC treatment, were evaluated by sequencing and microarrays, respectively. The differential results obtained were integrated into the mathematical models generated by Systems Biology. RESULTS: This computational approach identified 64 proteins, among which 18 stand out, either by being associated to the MoA or by providing a major capacity to classify the patients according to GC response. The biological functions of these proteins have been linked with inflammation, machinery GC-induced transcription and angiogenesis. All the selected proteins but one —Acidic leucine-rich Nuclear Phosphoprotein 32 family member E (ANP32E) — had previously been related to UC and/or GC-induced biological actions. ANP32E has been linked to the exchange of H2A to H2A.z histone, which promotes CG receptor complex-activated transcription. Western blot and immunofluorescence assays confirmed the predictive results obtained by Systems Biology. CONCLUSIONS: A comprehensive MoA of GC refractoriness has been described, highlighting the key role of GC-induced transcription in this phenomenon. Several proteins have been identified as putative predictors of GC refractoriness. Intestinal tissue from individuals with or without active ulcerative colitis (AUC) was collected and, those with AUC were characterized for their capacity to response to glucocorticoids treatment. microRNAs and mRNAs profiles were studied for each sample by microRNAseq and gene expression microarray, respectively.