Mucosal and plasma metabolomes in new-onset paediatric inflammatory bowel disease: correlations with disease characteristics and plasma inflammation protein markers

BACKGROUND AND AIMS

To advance the understanding of inflammatory bowel disease (IBD) pathophysiology, we compared the mucosal and plasma metabolomes between new-onset paediatric IBD patients and symptomatic non-IBD controls, and correlated plasma inflammation markers and disease characteristics with the altered metabolites.

METHODS

Paired colonic and ileal biopsies and plasma from 67 treatment-naïve children with incident Crohn's disease (CD; n=47), ulcerative colitis (UC; n=9), and non-IBD controls (n=11) were analysed using ultra-performance liquid chromatography-mass spectrometry (UPLC-MS/MS). Inflammatory plasma proteins (n=92) were assessed.

RESULTS

The metabolomes in inflamed mucosal biopsies differed between IBD patients and controls. In CD, mucosal levels of several lysophospholipids (lysophosphatidylcholines, lysophosphatidyletanolamines, lysophosphatidylinositols, and lysophosphatidylserines) were decreased, correlating with various plasma metabolites, including amino acid analogues and N-acetylated compounds. In both CD and UC, mucosal sphingolipids, including ceramide (d18:2/24:1, d18:1/24:2), lactosyl-N-palmitoyl-sphingosine (d18:1/16:0), behenoyl sphingomyelin (d18:1/22:0), lignoceroyl sphingomyelin (d18:1/24:0), and/or sphingomyelin (d18:1/24:1, d18:2/24:0) were increased, correlating with sphingolipids, bile acids, and/or N-acetylated metabolites in plasma. Among proteins associated with CD, interleukin-24 correlated with plasma metabolites, including lactosyl-N-palmitoyl sphingosine (d18:1/16:0) and phosphatidyletanolamine (18:1/18:1), haemoglobin, and faecal calprotectin. In UC, interleukin-24, interleukin-17A, and C-C motif chemokine 11 correlated with several plasma metabolites, including N-acetyltryptophan, tryptophan, glycerate, and threonate, and with the paediatric ulcerative colitis activity index, C-reactive protein, and faecal-calprotectin.

CONCLUSIONS

Mucosal perturbations of lysophospholipids and sphingolipids characterised the metabolome in new-onset paediatric IBD and correlated with plasma metabolites. By integrating plasma metabolomics data with inflammatory proteins and clinical data, we identified clinical and inflammatory markers associated with metabolomic signatures for IBD.

Integration of molecular profiles in a longitudinal wellness profiling cohort

An important aspect of precision medicine is to probe the stability in molecular profiles among healthy individuals over time. Here, we sample a longitudinal wellness cohort with 100 healthy individuals and analyze blood molecular profiles including proteomics, transcriptomics, lipidomics, metabolomics, autoantibodies and immune cell profiling, complemented with gut microbiota composition and routine clinical chemistry. Overall, our results show high variation between individuals across different molecular readouts, while the intra-individual baseline variation is low. The analyses show that each individual has a unique and stable plasma protein profile throughout the study period and that many individuals also show distinct profiles with regards to the other omics datasets, with strong underlying connections between the blood proteome and the clinical chemistry parameters. In conclusion, the results support an individual-based definition of health and show that comprehensive omics profiling in a longitudinal manner is a path forward for precision medicine.

An important aspect of precision medicine is to probe the stability in molecular profiles among healthy individuals over time. Here, the authors sample a longitudinal wellness cohort and analyse blood molecular profiles as well as gut microbiota composition.