Characterization of different fat depots in NAFLD using inflammation-associated proteome, lipidome and metabolome

Abstract

Non-alcoholic fatty liver disease (NAFLD) is recognized as a liver manifestation of metabolic syndrome, accompanied with excessive fat accumulation in the liver and other vital organs. Ectopic fat accumulation was previously associated with negative effects at the systemic and local level in the human body. Thus, we aimed to identify and assess the predictive capability of novel potential metabolic biomarkers for ectopic fat depots in non-diabetic men with NAFLD, using the inflammation-associated proteome, lipidome and metabolome. Myocardial and hepatic triglycerides were measured with magnetic spectroscopy while function of left ventricle, pericardial and epicardial fat, subcutaneous and visceral adipose tissue were measured with magnetic resonance imaging. Measured ectopic fat depots were profiled and predicted using a Random Forest algorithm, and by estimating the Area Under the Receiver Operating Characteristic curves. We have identified distinct metabolic signatures of fat depots in the liver (TAG50:1, glutamate, diSM18:0 and CE20:3), pericardium (N-palmitoyl-sphinganine, HGF, diSM18:0, glutamate, and TNFSF14), epicardium (sphingomyelin, CE20:3, PC38:3 and TNFSF14), and myocardium (CE20:3, LAPTGF-β1, glutamate and glucose). Our analyses highlighted non-invasive biomarkers that accurately predict ectopic fat depots, and reflect their distinct metabolic signatures in subjects with NAFLD.

Figure 1

Heatmap visualization based on Spearman correlation between variables of interest (column wise) and clinical parameters (row wise). Colour key indicates strength of a relationship: blue colour – negative relationship, red colour – positive relationship. *Significant relationship after FDR correction and significance level of 0.05.

Figure 2

Heatmap visualization based on Spearman correlation between variables of interest (column wise) and inflammation-associated proteome (row wise). Colour key indicates strength of a relationship: blue colour – negative relationship, red colour – positive relationship. *Significant relationship after FDR correction and significance level of 0.05.

Figure 3

Heatmap visualization based on Spearman correlation between variables of interest (column wise) and lipidome (row wise). Colour key indicates strength of a relationship: blue color – negative relationship, red colour – positive relationship. *Significant relationship after FDR correction and significance level of 0.05.

Figure 4

Heatmap visualization based on Spearman correlation between variables of interest (column wise) and metabolome (row wise). Colour key indicates strength of a relationship: blue colour – negative relationship, red colour – positive relationship. *Significant relationship after FDR correction and significance level of 0.05.

Figure 5

RF analysis (combined dataset) – variable importance based on mean decrease in accuracy for liver, pericardial, epicardial and myocardial fat.