#49 Uremic toxins and inflammation: Metabolic pathways affected in stage 5 kidney disease

Abstract

Abstract Background and Aims Chronic kidney disease (CKD) is a global health threat, affecting around 12% of the population. The uremic phenotype associated with CKD is considered a predictor for the high cardiovascular and overall mortality in these patients and inflammation is regarded as an essential contributor to it. We seek to discover new inflammation-related metabolites in the CKD population through the use of semitargeted metabolomics, in order to gain insight into the pathways involved and identify potential biomarkers. Method The study employed an observational prospective design to investigate metabolite impact on inflammation in 50 non-dialysis-dependent stage 5 CKD patients. Blood samples were collected before the start of the first haemodialysis session. A semitargeted metabolomic approach was followed, utilising liquid chromatography and high-resolution mass spectrometry with an in-house library of polar compounds. Univariate (Mann-Whitney test) and multivariate (logistic regression with LASSO regularization) analyses identified metabolomic variables associated with inflammation. Results Univariate analysis revealed that adenosine-5′-phosphosulfate (APS) (p-value (p) 0.0211), dimethyglycine (p 0.0015), pyruvate (p 0.0066), lactate (p 0.0156) and 2-ketobutyric acid (p 0.0458) were significantly increased with inflammation, with differences over 1,5-fold changes (inflammation = yes vs inflammation = no). Cholic acid (p 0.01655), homogentisic acid (p 0.0007), 2-phenylpropionic acid (p 0.0351) and 3-indoxylsulfate (p 0.0087) show similar patterns but in the opposite direction, as observed in the volcano plot in Fig. 1. Multivariate analysis revealed significant associations between metabolites and inflammation in CKD stage 5 patients. Specifically, odds ratios for N-Butyrylglycine (OR: 1.87), dimethylglycine (OR: 1.81), 2-Oxoisopentanoic acid (OR: 1.70), and pyruvate (OR: 1.71) were related to a higher probability of experiencing inflammation. Conversely, homogentisic acid (OR: 072), 2-Phenylpropionic acid (OR: 0.77), and 2-Methylglutaric acid (OR: 0.78) exhibited an OR<1, suggestive of decreased probability. This is graphically represented in Fig. 2. Conclusion In our semi-targeted metabolomics study on stage 5 CKD patients, we pinpointed various metabolites linked to inflammation. Notably, dimethylglycine, a uremic toxin, may contribute to vascular damage, while pyruvate's abnormal metabolism is a known feature in heart failure and chronic kidney disease. N-butyrilglycine and 2-oxoisopentanoic acid were also associated with inflammation, potentially tied to mitochondrial fatty acid beta oxidation and incomplete breakdown of branched-chain amino acids. APS, impacting bone development and intestinal permeability, showed a positive correlation with inflammation. Conversely, homogentisic acid and phenylpropionic acid were negatively correlated. While homogentisic acid's anti-inflammatory role isn't clear, phenylpropionic acid exhibits protective effects against hepatotoxicity by reducing immune response. 2-methylglutaric acid and cholic acid were inversely correlated with inflammation, suggesting a protective effect related to fatty acid metabolism and bile acid regulation. These identified metabolites enhance our understanding of CKD-associated inflammation and present strong candidates for future biomarker validation with potential for clinical translation.