Limits...
Urine metabolome profiling of immune-mediated inflammatory diseases

View Article: PubMed Central - PubMed

ABSTRACT

Background: Immune-mediated inflammatory diseases (IMIDs) are a group of complex and prevalent diseases where disease diagnostic and activity monitoring is highly challenging. The determination of the metabolite profiles of biological samples is becoming a powerful approach to identify new biomarkers of clinical utility. In order to identify new metabolite biomarkers of diagnosis and disease activity, we have performed the first large-scale profiling of the urine metabolome of the six most prevalent IMIDs: rheumatoid arthritis, psoriatic arthritis, psoriasis, systemic lupus erythematosus, Crohn’s disease, and ulcerative colitis.

Methods: Using nuclear magnetic resonance, we analyzed the urine metabolome in a discovery cohort of 1210 patients and 100 controls. Within each IMID, two patient subgroups were recruited representing extreme disease activity (very high vs. very low). Metabolite association analysis with disease diagnosis and disease activity was performed using multivariate linear regression in order to control for the effects of clinical, epidemiological, or technical variability. After multiple test correction, the most significant metabolite biomarkers were validated in an independent cohort of 1200 patients and 200 controls.

Results: In the discovery cohort, we identified 28 significant associations between urine metabolite levels and disease diagnosis and three significant metabolite associations with disease activity (PFDR < 0.05). Using the validation cohort, we validated 26 of the diagnostic associations and all three metabolite associations with disease activity (PFDR < 0.05). Combining all diagnostic biomarkers using multivariate classifiers we obtained a good disease prediction accuracy in all IMIDs and particularly high in inflammatory bowel diseases. Several of the associated metabolites were found to be commonly altered in multiple IMIDs, some of which can be considered as hub biomarkers. The analysis of the metabolic reactions connecting the IMID-associated metabolites showed an over-representation of citric acid cycle, phenylalanine, and glycine-serine metabolism pathways.

Conclusions: This study shows that urine is a source of biomarkers of clinical utility in IMIDs. We have found that IMIDs show similar metabolic changes, particularly between clinically similar diseases and we have found, for the first time, the presence of hub metabolites. These findings represent an important step in the development of more efficient and less invasive diagnostic and disease monitoring methods in IMIDs.

Electronic supplementary material: The online version of this article (doi:10.1186/s12916-016-0681-8) contains supplementary material, which is available to authorized users.

No MeSH data available.


Urine diagnostic biomarkers in IMID diseases. a Shows the distribution of the concentrations in logarithmic scale of the metabolites associated to multiple IMID diseases (i.e., hub metabolites). The concentrations have been previously normalized to the median concentration of the control cohort. b Shows the clustering graph of both diseases and metabolites according to their corresponding disease-metabolite associations
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC5016926&req=5

Fig2: Urine diagnostic biomarkers in IMID diseases. a Shows the distribution of the concentrations in logarithmic scale of the metabolites associated to multiple IMID diseases (i.e., hub metabolites). The concentrations have been previously normalized to the median concentration of the control cohort. b Shows the clustering graph of both diseases and metabolites according to their corresponding disease-metabolite associations

Mentions: Among the validated metabolites, six were found to be associated to three or more IMIDs (Fig. 2a). Since their patterns were very similar between diseases (i.e., significance of association and direction of change), they were considered as hub metabolites in IMIDs. From these, citrate showed the strongest hub properties, showing a significantly lower concentration in the urine of most IMIDs compared to controls (Fig. 2, PCD = 6.2 × 10–16, PSLE = 2.3 × 10–10, PPs = 2.9 × 10–8, PRA = 4.3 × 10–7, PPsA = 3.5 × 10–5). In UC, citrate levels were also lower than in controls both in the discovery and validation cohorts, although the difference was only significant at the nominal (P < 0.05) level.Fig. 2


Urine metabolome profiling of immune-mediated inflammatory diseases
Urine diagnostic biomarkers in IMID diseases. a Shows the distribution of the concentrations in logarithmic scale of the metabolites associated to multiple IMID diseases (i.e., hub metabolites). The concentrations have been previously normalized to the median concentration of the control cohort. b Shows the clustering graph of both diseases and metabolites according to their corresponding disease-metabolite associations
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC5016926&req=5

Fig2: Urine diagnostic biomarkers in IMID diseases. a Shows the distribution of the concentrations in logarithmic scale of the metabolites associated to multiple IMID diseases (i.e., hub metabolites). The concentrations have been previously normalized to the median concentration of the control cohort. b Shows the clustering graph of both diseases and metabolites according to their corresponding disease-metabolite associations
Mentions: Among the validated metabolites, six were found to be associated to three or more IMIDs (Fig. 2a). Since their patterns were very similar between diseases (i.e., significance of association and direction of change), they were considered as hub metabolites in IMIDs. From these, citrate showed the strongest hub properties, showing a significantly lower concentration in the urine of most IMIDs compared to controls (Fig. 2, PCD = 6.2 × 10–16, PSLE = 2.3 × 10–10, PPs = 2.9 × 10–8, PRA = 4.3 × 10–7, PPsA = 3.5 × 10–5). In UC, citrate levels were also lower than in controls both in the discovery and validation cohorts, although the difference was only significant at the nominal (P < 0.05) level.Fig. 2

View Article: PubMed Central - PubMed

ABSTRACT

Background: Immune-mediated inflammatory diseases (IMIDs) are a group of complex and prevalent diseases where disease diagnostic and activity monitoring is highly challenging. The determination of the metabolite profiles of biological samples is becoming a powerful approach to identify new biomarkers of clinical utility. In order to identify new metabolite biomarkers of diagnosis and disease activity, we have performed the first large-scale profiling of the urine metabolome of the six most prevalent IMIDs: rheumatoid arthritis, psoriatic arthritis, psoriasis, systemic lupus erythematosus, Crohn&rsquo;s disease, and ulcerative colitis.

Methods: Using nuclear magnetic resonance, we analyzed the urine metabolome in a discovery cohort of 1210 patients and 100 controls. Within each IMID, two patient subgroups were recruited representing extreme disease activity (very high vs. very low). Metabolite association analysis with disease diagnosis and disease activity was performed using multivariate linear regression in order to control for the effects of clinical, epidemiological, or technical variability. After multiple test correction, the most significant metabolite biomarkers were validated in an independent cohort of 1200 patients and 200 controls.

Results: In the discovery cohort, we identified 28 significant associations between urine metabolite levels and disease diagnosis and three significant metabolite associations with disease activity (PFDR&thinsp;&lt;&thinsp;0.05). Using the validation cohort, we validated 26 of the diagnostic associations and all three metabolite associations with disease activity (PFDR&thinsp;&lt;&thinsp;0.05). Combining all diagnostic biomarkers using multivariate classifiers we obtained a good disease prediction accuracy in all IMIDs and particularly high in inflammatory bowel diseases. Several of the associated metabolites were found to be commonly altered in multiple IMIDs, some of which can be considered as hub biomarkers. The analysis of the metabolic reactions connecting the IMID-associated metabolites showed an over-representation of citric acid cycle, phenylalanine, and glycine-serine metabolism pathways.

Conclusions: This study shows that urine is a source of biomarkers of clinical utility in IMIDs. We have found that IMIDs show similar metabolic changes, particularly between clinically similar diseases and we have found, for the first time, the presence of hub metabolites. These findings represent an important step in the development of more efficient and less invasive diagnostic and disease monitoring methods in IMIDs.

Electronic supplementary material: The online version of this article (doi:10.1186/s12916-016-0681-8) contains supplementary material, which is available to authorized users.

No MeSH data available.