Dietary PUFAs drive diverse system-level changes in lipid metabolism

Abstract

Objective: Polyunsaturated fatty acid (PUFA) supplements have been trialled as a treatment for a number of conditions and produced a variety of results. This variety is ascribed to the supplements, that often comprise a mixture of fatty acids, and to different effects in different organs. In this study, we tested the hypothesis that the supplementation of individual PUFAs has system-level effects that are dependent on the molecular structure of the PUFA.

Methods: We undertook a network analysis using Lipid Traffic Analysis to identify both local and system-level changes in lipid metabolism using publicly available lipidomics data from a mouse model of supplementation with FA(20:4n-6), FA(20:5n-3), and FA(22:6n-3); arachidonic acid, eicosapentaenoic acid, and docosahexaenoic acid, respectively. Lipid Traffic Analysis is a new computational/bioinformatics tool that uses the spatial distribution of lipids to pinpoint changes or differences in control of metabolism, thereby suggesting mechanistic reasons for differences in observed lipid metabolism.

Results: There was strong evidence for changes to lipid metabolism driven by and dependent on the structure of the supplemented PUFA. Phosphatidylcholine and triglycerides showed a change in the variety more than the total number of variables, whereas phosphatidylethanolamine and phosphatidylinositol showed considerable change in both which variables and the number of them, in a highly PUFA-dependent manner. There was also evidence for changes to the endogenous biosynthesis of fatty acids and to both the elongation and desaturation of fatty acids.

Conclusions: These results show that the full biological impact of PUFA supplementation is far wider than any single-organ effect and implies that supplementation and dosing with PUFAs require a system-level assessment.

Keywords: Lipid metabolism; Metabolic network; PUFA supplementation; Traffic analysis.

Figure 1

The mouse model of PUFA supplementation used in this study. This shows the tissue network of the mouse model of PUFA supplementation. The arrows show the metabolic connections between compartments.

Figure 2

Traffic Analysis of phospholipids in a mouse model of PUFA supplementation. Panel A, Switch Analysis of phosphatidylethanolamine (PE) variables. Panel B, Switch Analysis of phosphatidylcholine (PC) variables. Pie charts show the number of variables of the appropriate head group in the relevant tissue(s). Large inset pie charts show the B-type species (lipids found in two neighbouring compartments), whereas U-type lipids (lipids found only in one compartment) are depicted with smaller pie charts. The Jaccard-Tanimoto coefficients (J) and probability (p) values that describe the similarity between sets of variables. Translucent pie charts indicate those in which only the number of variables differs between groups.

Figure 3

Traffic Analysis of lipids in a mouse model of PUFA supplementation. Panel A, Switch Analysis of phosphatidylinositol (PI) variables; Panel B, Switch Analysis of triglyceride (TG) variables. Pie charts show the number of variables of the appropriate head group in the relevant tissue(s). Large inset pie charts show the B-type species (lipids found in two neighbouring compartments), whereas U-type lipids (lipids found only in one compartment) are depicted with smaller pie charts. The Jaccard-Tanimoto coefficients (J) and probability (p) values that describe the similarity between sets of variables. Translucent pie charts indicate those in which only the number of variables differs between groups.

Figure 4

Modifications to FA metabolism associated with supplementation with PUFAs. Panel A, Abundance of TGs associated with de novo lipogenesis, TG(46:0, 46:1, 48:0, 48:1, 48:2, 50:1) [37], shown as the mean with 1.5 IQR. Panel B, the activity of elongases ELOVL2/5 on FA(20:5) expressed as the ratio of the abundance of PC(18:0/22:5) to PC(18:0/20:5) and PC(16:0/22:5) to PC(16:0/20:5), with the latter marked∗. The box plots represent the values for mean, standard deviation and spread for n = 4 or 5 per group.