Figure 1.
Heat map of diet effects on liver.
The heat map represents the fold-change for each metabolite relative to control chow-fed versus WD-fed mice. The WD was supplemented with olive (O), EPA (E), DHA (D) or EPA and DHA (E + D). Results are sorted by fold-change within each pathway.
Figure 2.
Diet effects on metabolic pathways.
Panel A: Number of metabolites significantly changed in each pathway by all diets. Panel B: Percent of metabolites in each pathway that were induced or suppressed in WD + O fed mice relative to chow fed mice. Panel C: Percent of metabolites in each pathway that were induced or suppressed in all WD + C20-22 n-3 PUFA fed mice relative to WD + O fed mice.
Figure 3.
Volcano plots of diet effects on hepatic metabolites.
Volcano plots were prepared using MetaboAnalyst (http://www.metaboanalyst.ca). The following groups (8 mice/group) were examined, Panel A: Chow versus WD + O; Panel B: WD + O versus WD + D. Results were plotted as log2 fold-change versus -log10 p-value. Several metabolites are labeled in each plot.
Figure 4.
Diet effects on plasma endotoxin.
Plasma endotoxin was quantified as described in Materials and Methods. Results are presented as Plasma Endotoxin, Fold Change. Mean ± SD relative to chow fed mice; *, P ≤ 0.05 versus chow.
Figure 5.
Diet effects on hepatic markers of inflammation, SFA, MUFA and damage.
Linear regression analysis of hepatic palmitoyl-sphingomyelin (Fold Change relative to chow) versus hepatic MCP1 mRNA expression (fold change relative to chow) (Panel A); hepatic total MUFA content (µmoles total MUFA/mg protein) (Panel B); hepatic palmitate (16:0) (µmoles /mg protein) (Panel C); and plasma AST (units (U)/ml of plasma) (Panel D). Palmitoyl-sphingomyelin was quantified in the metabolomic analysis while hepatic MCP1, MUFA, palmitate, and plasma AST were quantified and reported previously [8]. Each data point in Panels A-D represents the relative abundance of palmitoyl-sphingomyelin and hepatic MCP1 mRNA, MUFA, 16:0 or plasma AST for each animal. The groups are colored-coded to facilitate visualization of the distribution in each group.
Figure 6.
Diet effects on sphingolipid metabolites and enzymes involved in sphingolipid synthesis.
Panel A: Pathway for de novo sphingomyelin synthesis. Panel B: Hepatic palmitoyl-sphingomyelin and metabolites involved in sphingomyelin synthesis (sphinganine) and ceramide degradation (sphingosine). Results are represented as Metabolites-Fold Change relative to chow; mean ± SD, n=8 per group. Panel C: RNA expression of key enzymes involved in sphingomyelin synthesis; mean ± SD. [serine-palmitoyl transferase long chain base subunit-1 & 2 (SPTLC1 & 2) and phosphatidylcholine:ceramide choline phosphotransferase 1 & 2 (SGMS1 & 2)]; *, p ≤ 0.05 versus chow; #, p ≤ 0.05 versus WD + O.
Figure 7.
Diet effects on hepatic one-carbon, choline and glutathione metabolism.
Panel A: Pathways for one-carbon, choline, glutathione and sphingomyelin metabolism. Panel B: Metabolites quantified by the metabolomic analysis were expressed as Metabolite-Fold Change and represented as mean ± SD, n=8 per group; *, p ≤ 0.05 versus chow; #, p ≤ 0.05 versus WD + O.
Figure 8.
Diet effects on glucose metabolism and de novo MUFA synthesis.
Panel A: Glucose conversion to saturated and monounsaturated fatty acids. Panel B: Metabolites involved in glucose metabolism were quantified by the metabolomic analysis and expressed as Metabolite-Fold Change; mean ± SD, n=8 per group. Panel C: Metabolites involved in de novo MUFA synthesis were quantified by the metabolomic analysis and expressed as Metabolite-Fold Change; mean ± SD, n=8 per group; *, p ≤ 0.05 versus chow; #, p≤ 0.05 versus WD + O.
Figure 9.
Diet effects on the expression of PPARγ2 and enzymes involved in MUFA synthesis.
Panel A: Expression of enzymes involved in MUFA synthesis and the nuclear receptor peroxisome proliferator activated receptor γ2 (PPARγ2). Results are represented as mRNA Abundance-Fold Change, relative to chow-fed mice; mean ± SD, n=8/group. [Fatty acid synthase (FASN); ATP citrate lyase (ACL); stearoyl CoA desaturase 1 (SCD1), fatty acid elongase (Elovl)] Panel B: Hepatic nuclear abundance of PPARγ2 expressed as Fold Change relative to chow-fed mice; mean ± SD; n=8/group; *, P ≤ 0.05 versus chow; #, P ≤ 0.05 versus WD + O.
Figure 10.
Diet effects on hepatic fat content.
Panel A: Abundance of total hepatic SFA, MUFA, n-6 PUFA, and n-3 PUFA fatty acids analyzed from total lipid extracts using gas chromatography [8]. The sum of the fatty acids in each group (SFA, MUFA, n-3 and n-6 PUFA) is presented to illustrate the cumulative effects of diet on hepatic fat. Results are presented as total µmoles of fatty acid/g protein; mean + SD in each fatty acid class; n=8/group. Panel B: Distribution of SFA, MUFA, n-6 PUFA, and n-3 PUFA in hepatic phospholipids. Hepatic phospholipids were fractionated by solid phase separation, saponified and methylated for GC analysis (see Fig, S1 for the quality of the separation of phospholipids from total lipids). Results are expressed as Fatty Acid Mole% in each fatty acid class, i.e., SFA, MUFA, N3 and N6-PUFA; mean + SD, n=8/group.
Figure 11.
Diet effects on hepatic phospholipid fatty acids.
Individual fatty acids in the phospholipid fraction were quantified as described above and represented as Fatty Acid-Mole%, mean + SD, n=8/group; *, P ≤ 0.05 versus chow; #, P ≤ 0.05 versus WD + O.
Figure 12.
Diet effects on hepatic lysophospholipids.
Panels A and B: Lysophospholipids were fractionated and quantified by the metabolomic analysis (LC/MS) as described in the Methods. Panel A: Lysophospholipids with acyl chains in the sn-1 position; Panel B: Lysophospholipids with acyl chains in the sn-2 position. The lysophospholipids are represented as Metabolite-Fold Change relative to chow; mean ± SD; n=8/group; *, P ≤ 0.05 versus chow; #, P ≤ 0.05 versus WD + O.
Figure 13.
Diet effects on the expression of enzymes involved in membrane remodeling.
Expression of enzymes involved in the incorporation of fatty acyl chains into phospholipids (lysophosphatidylcholine acyl transferase subtypes, LPCAT1-4) and excision of fatty acids from the sn-2 position of phospholipids (phospholipase A2 subtypes, iPLA2γ and PLA2γ6) were quantified. Results are represented as mRNA Abundance-Fold Change, mean ± SD; n=8/group; *, p ≤ 0.05 versus chow; #, p ≤ 0.05 versus WD + O.
Figure 14.
Pathways for the formation of oxidized lipids from arachidonic acid (20:4,n-6) [Panel A] and eicosapentaenoic acid (20:5,n-3)[Panel B].
See text for explanation.
Figure 15.
Diet effects on hepatic oxidized lipids.
N-6 and n-3 PUFA and oxidized fatty acids were quantified by the metabolomic analysis (Methods). Results are expressed and Metabolite-Fold Change relative to the chow-fed group, mean ± SD, n=8/group; *, p ≤ 0.05 versus chow; #, p ≤ 0.05 versus WD + O.
Figure 16.
Expression of enzymes involved in fatty acid oxidation.
Panel A: Hepatic expression of cyclooxygenases (COX1 & 2), lipoxygenases (5-. 12- & 15-LOX). Panel B: Hepatic expression of cytochrome P450-C2 (CYP2C subtypes) and soluble and microsomal epoxide hydroxylases (EPHX1 and 2). Results are represented as mRNA Abundance-Fold Change, mean ± SD; n=8/group; *, p ≤ 0.05 versus chow; #, p ≤ 0.05 versus WD + O.
Figure 17.
Diet effects on urinary isoprostanes and hepatic α-tocopherol, ascorbate and Nrf2.
Panel A: Levels of 24-Hour urinary F2- and F3-IsoPs were quantified as described [17]. F2-IsoPs are derived from arachidonic acid (20:4 n-6) and F3-IsoPs are derived from eicosapentaenoic acid (20:5 n-3) (Fig. 14). Results are represented as Urinary Isoprostanes (ng/mg creatinine) mean + SD, n=3; urine from 2 pools of mice (3 mice/pool) from each diet group were assayed. Panel B: Hepatic α-tocopherol (vitamin E) and ascorbate (vitamin C) were quantified by the metabolomic analysis (Methods) and represented as Metabolite-Fold Change, relative to chow-fed mice; mean ± SD; n=8/group. Panel C: Hepatic nuclear abundance of Nrf2. Hepatic nuclear extracts were assayed for Nrf2 and the loading control protein, TATA-binding protein (TBP) using methods previously described [17]. Nrf2 nuclear abundance was normalized to TBP for each sample. Results are represented as Nrf2 Nuclear Abundance-Fold Change, mean + SD, n=8 group; *, P ≤ 0.05 versus chow; #, P ≤ 0.05 versus WD + O.
Figure 18.
Diet effects on S-lactoylglutathione and metabolites from carbohydrate and lipid oxidation.
Panel A: Pathway of methylglyoxal formation and detoxification. Panel B: Hepatic abundance of S-lactoylglutathione was quantified by the metabolomic analysis (Methods). Results are represented as S-Lactoylglutathione-Fold Change, relative to chow-fed mice; mean ± SD, n=8/group. Panel C: Expression of enzymes involved in S-lactoylglutathione formation and degradation, i.e., glyoxalase 1 (Glo 1) and glyoxalase 2 (Glo 2). Results are represented as mRNA Abundance-Fold Change, relative to chow-fed mice; mean + SD, n=8/group.
Figure 19.
Diet effects on hepatic abundance of glycolytic metabolites that can be converted to methylglyoxal and S-lactoyl-glutathione.
Hepatic metabolites were quantified by the metabolomic analysis (Methods) and represented as Metabolite-Fold Change, relative to chow-fed mice; mean ± SD, n=8/group; *, P ≤ 0.05 versus chow; #, P ≤ 0.05 versus WD + O.