Metabolomic and Lipidomic Analysis of the Heart of Peroxisome Proliferator-Activated Receptor-γ Coactivator 1-β Knock Out Mice on a High Fat Diet

Abstract

The peroxisome proliferator-activated receptor-γ coactivators (PGC-1) are transcriptional coactivators with an important role in mitochondrial biogenesis and regulation of genes involved in the electron transport chain and oxidative phosphorylation in oxidative tissues including cardiac tissue. These coactivators are thought to play a key role in the development of obesity, type 2 diabetes and the metabolic syndrome. In this study we have used a combined metabolomic and lipidomic analysis of cardiac tissue from the PGC-1β null mouse to examine the effects of a high fat diet on this organ. Multivariate statistics readily separated tissue from PGC-1β null mice from their wild type controls either in gender specific models or in combined datasets. This was associated with an increase in creatine and a decrease in taurine in the null mouse, and an increase in myristic acid and a reduction in long chain polyunsaturated fatty acids for both genders. The most profound changes were detected by liquid chromatography mass spectrometry analysis of intact lipids with the tissue from the null mouse having a profound increase in a number of triglycerides. The metabolomic and lipodomic changes indicate PGC-1β has a profound influence on cardiac metabolism.

Figure 1

Analysis of the lipidome by GC-MS and LC-MS. A) A typical GC-MS chromatogram of fatty acid methyl esters in the organic fraction. Typical LC-MS data with identification. B) Total ion chromatogram of the LC-MS analysis of intact lipids. Phospholipids elute between 3.5 and 5.5 minutes and triacylglycerols between 6.5 and 8.5 minutes. C) Single scan mass spectrum from the LC-MS at 4.41 minutes D) extracted ion chromatogram for 748.5 E) MS-MS of 748.5 at 4.4 minutes. Characteristic fragments for a phosphoethanolamine with two fatty acids 22:0 and 22:6.

Figure 2

Typical GC-MS chromatogram of the MSTFA derivatized metabolites from the aqueous fraction of the heart tissue.

Figure 3

Summary plots for the multivariate analysis of the aqueous fraction for the total dataset comparing genotype, regardless of gender. A) Scores plot of the NMR spectra comparing tissue from the PGC-1β null mice compared with wild type controls (Q²= 79%). B) S-plot where the loading is plotted against the correlation between the variable and the score for the NMR spectral dataset. Signals in (i) are due to creatine (ii) are due to phospohcholine, glutamic acid, orotic acid, proline and ATP peaks in (iii) are due to taurine, alanine and succinate. C) Scores plot of GC-MS analysis of the aqueous fraction comparing tissue from the PGC-1β null mice and. wild type controls (Q²= 62%). D) S-plot where the loading is plotted against the correlation between the variable and the score for the GC-MS dataset.

Figure 4

Summary plots for the multivariate analysis of the lipid fraction for the total dataset comparing genotype, regardless of gender. Analysis of organic GC-MS data. A) Scores plot of the GC-MS analysis of the lipid fraction to measure total fatty acid content in tissue from the PGC-1β null mice compared with wild type controls (Q²= 48%). B) Loadings plot sorted by loading value for the model in (A). The variables marked by the ovals were relevant to the model. Error bars represent 95% confidence interval as determined by jack-knifing. Analysis of LC-MS data. C) Scores plot comparing the LC-MS analysis of intact lipids with two injections for each sample comparing tissue from the PGC-1β null mice and the wild type controls (Q²= 89%). Two samples of each male KO, male WT, female KO and female WT were predicted correctly as shown in the plot. D) Loadings plot sorted by loadings value for the model in (D). The variables marked by the boxes were relevant to the model and identified. Error bars represent 95% confidence interval as determined by jack-knifing.