The proteome of ulcerative colitis - Functional analyses of the active disease and the remission state in comparison with healthy controls

Abstract

This work characterizes the proteome of human colon mucosa in ulcerative colitis (UC).

We developed an optimized sample preparation method of colon mucosa biopsies for bottom-up proteomics and applied it to characterize the proteomes of active UC, remission from UC, and healthy controls. Abundances of proteins related to the immune system and to protein processing in the endoplasmic reticulum are increased in active UC compared to healthy controls. Lower abundant are metallothioneins, fibrillary collagens, bile acid transport proteins, carbonic anhydrases, and proteins related to nutrient, energy, and xenobiotic metabolism.

In general, the remission state seems to be a blend of healthy and diseased state. We characterized the remission state based on the proteins that were significantly different abundant in remission compared to active UC and/or healthy controls. A small fraction of these proteins (associated functions: hormones, vitamins, lipoproteins, muscle) is higher abundant in remission than in both active UC and healthy controls. Most proteins (associated functions: immune system, protein processing, collagen) show similar abundances in remission as in healthy controls. About one fourth of the remission abundances (associated functions: nutrient and energy metabolism, PPAR signaling) was between those in active UC and healthy controls and significantly different from both. Approximately one eighth of the proteins was at similar levels as in active UC (associated functions: immunoglobulins, metallothioneins, prostaglandin metabolism). Protein abundances that are not at equal levels as in healthy controls may contribute to relapses and symptoms in remission.

Our findings have clinical implications. Several functions apart from the inflammation could be readily addressable with medication. The abundances of the routinely used biomarkers calprotectin and lactotransferrin are representative for only a small minority of differently abundant proteins. An additional assessment of more representative proteins may be useful. We furthermore present a model for the prediction of the 1-year-outcome that could bring benefit for clinical decision-making.