Urine Proteome in Distinguishing Hepatic Steatosis in Patients with Metabolic-Associated Fatty Liver Disease

Figures

• 

  Figure 1. Proteomics study by using MAFLD patients’ urine samples. (A) The experimental design of urine profiling in proteomics for distinguishing mild/severe hepatic steatosis in MAFLD patients. Quantitative data were expressed as mean ± SD. For all laboratory measures and for continuous demographics: 1-way analysis of variance test with Bonferroni adjustments. Proportions: percentage. (B) Hepatic steatosis was measured by MRI-PDFF. (C) Baseline characteristics of included participants in discovery proteomics set according to mild hepatic steatosis (liver fat content 5–10%) and severe hepatic steatosis (liver fat content > 10%). ALT, alanine transaminase; AST, aspartic transaminase; ALP, alkaline phosphatase; GGT, γ-glutamyl transferase; TG, triglycerides; HbA1c, glycosylated hemoglobin, type A1C; INS, insulin released test; HOMA-IR, homeostasis model assessment-insulin resistance; and FIB-4, fibrosis 4 score.

• 

  Figure 2. Identification and quantification of MAFLD urine samples from mild/severe steatosis and healthy controls. (A) The Venn diagram for the identified urine proteins from the healthy volunteers and mild and severe steatosis MAFLD patients. (B) The dynamic range of the intensity-based absolute quantification (iBAQ) algorithm of abundance of identified proteins from healthy volunteers and mild and severe steatosis MAFLD patients. (C) Principal component analysis (PCA) of urine proteome of MAFLD patients and healthy controls.

• 

  Figure 3. The clustering heatmap analyses differentiate mild hepatic steatosis and severe steatosis of MAFLD patients from healthy controls in proteomic features. Patients were ordered in increasing severity of hepatic steatosis measured by MRI-PDFF (from left to right). Hierarchical clustering was performed only on the 53 proteins significant by ANOVA.

• 

  Figure 4. Two proteins were finally filtered for distinguishing mild/severe steatosis. (A) Among 12 clusters, cluster 3 included consistently up-regulated proteins from healthy controls to mild hepatic steatosis and to severe hepatic steatosis patients, whereas clusters 1 and 8 included consistently down-regulated proteins. (B) The GO analysis and expression (log2) of the filtered proteins (red color for up-regulated and green for down-regulated). (C) The Pearson association of 9 unique filtered proteins with clinical parameters. (D) The expression intensity (Un-log2) of selected ORM1 and ceruloplasmin, which were most strongly correlated with liver fat content (measured by MRI-PDFF), fibrosis, ALT, triglyceride, glucose, HOMA-IR, and C-reactive protein. (E) The diagnostic accuracy of ORM1 and ceruloplasmin for hepatic steatosis from healthy controls (right panel) and severe hepatic steatosis from mild hepatic steatosis patients (right panel) was calculated based on the expression intensity of proteomic data.

• 

  Figure 5. Validation of ceruloplasmin and ORMA1 in urine samples. (A) Baseline characteristics of participants included in the validation set according to mild hepatic steatosis (liver fat content 5–10%) and severe hepatic steatosis (liver fat content > 10%). Quantitative data were expressed as mean ± SD. For all laboratory measures and for continuous demographics: 1-way analysis of variance test with Bonferroni adjustments. Proportions: percentage. (B) ELISA. (C) Western blot; figure below was calculated by relative absorbance units and normalized to urine creatinine excretion. (D) The ROC curve (using ELISA results) for distinguishing mild steatosis and severe steatosis. NC, healthy control; MS, mild steatosis; SS, severe steatosis.