Table 1.
Characteristics of the patient population.
Fig 1.
Fatty acids that showed significantly different levels in serum between patients with NAFLD activity scores (NAS) 1–4 (n = 68) and patients with NAS ≥5 (n = 38).
(A) Saturated fatty acids (SFAs), pentadecanoic (15:0) and heptadecanoic (17:0) acid, (B) monounsaturated fatty acids (MUFAs), palmitelaidic (16:1n7t), palmitoleic (16:1n7c) and vaccenic (18:1n7c) acid. Fatty acids were measured as percentage of the phospholipid fraction. P-value was calculated using two-tailed student’s T-test.
Fig 2.
Pentadecanoic acid (15:0), vaccenic acid (18:1n7c) and docosapentaenoic acid (22:5n3) serum levels in patients.
Isolated steatosis (n = 27), NASH patients with fibrosis stage 0–1 (n = 45) and NASH patients with fibrosis stage 2–4 (n = 34). Fatty acids were measured as percentage of the phospholipid fraction. Statistical significance was done using ANOVA test and then pairwise comparison using Tukey’s test was performed.
Fig 3.
Correlation of fatty acids with NASH parameters.
(A) Correlation of pentadecanoic (15:0), heptadecanoic (17:0) acid and palmitelaidic (16:1n7t) with NAFLD activity scores (NAS), (B) correlation of pentadecanoic (15:0), heptadecanoic (17:0) acid and palmitelaidic (16:1n7t) with hepatocyte ballooning scores, (C) correlation of vaccenic acid (18:1n7c) with liver inflammation scores, and (D) correlation of vaccenic acid (18:1n7c) with fibrosis scores. Fatty acid levels are presented as a percentage of the phospholipid fraction. Statistical significance was done using ANOVA test and then pairwise comparison using Tukey’s test was performed. Spearman correlation was calculated using GraphPad Prism 6 and Bonferroni correction was used to select significant correlations.
Fig 4.
Correlation of fatty acid levels with parameters of liver injury and metabolic syndrome.
(A) Correlation of pentadecanoic (15:0) with fasting glucose levels and aspartate transaminase (AST), (B) correlation of heptadecanoic (17:0) with AST and alanine transaminase (ALT) levels, (C) correlation of palmitoleic acid (16:1n7c) with AST levels, (D) correlation of vaccenic acid (18:1n7c) with ferritin levels, and (E) correlation of docosapentaenoic acid (22:5n3) with albumin levels. Spearman correlation was calculated using GraphPad Prism 6 and Bonferroni correction was used to select the significant correlations.
Fig 5.
Receiver operating characteristic (ROC) analysis.
(A) based on NAFLD activity scores (NAS)–patients with NAS 1–4 versus patients with NAS ≥5. (B) Based on fibrosis scores—patients with fibrosis 0–1 versus patients with fibrosis 2–4.
Table 2.
Comparison of physiological parameters with 15:0 treatments in vivo.
Fig 6.
Effects of 15:0-supplemented MCD diet in vivo.
(A) NAS scores from H&E stained slides. (B) Histologic evaluation by H&E and (C) Periodic acid Schiff with diastase (PAS-D) staining in liver from mice treated with MCD or with MCD+15:0. (D) AST and ALT levels in mice treated with MCD or with MCD+15:0.