Diet-gene interactions and PUFA metabolism: a potential contributor to health disparities and human diseases

Abstract

The "modern western" diet (MWD) has increased the onset and progression of chronic human diseases as qualitatively and quantitatively maladaptive dietary components give rise to obesity and destructive gene-diet interactions. There has been a three-fold increase in dietary levels of the omega-6 (n-6) 18 carbon (C18), polyunsaturated fatty acid (PUFA) linoleic acid (LA; 18:2n-6), with the addition of cooking oils and processed foods to the MWD. Intense debate has emerged regarding the impact of this increase on human health. Recent studies have uncovered population-related genetic variation in the LCPUFA biosynthetic pathway (especially within the fatty acid desaturase gene (FADS) cluster) that is associated with levels of circulating and tissue PUFAs and several biomarkers and clinical endpoints of cardiovascular disease (CVD). Importantly, populations of African descent have higher frequencies of variants associated with elevated levels of arachidonic acid (ARA), CVD biomarkers and disease endpoints. Additionally, nutrigenomic interactions between dietary n-6 PUFAs and variants in genes that encode for enzymes that mobilize and metabolize ARA to eicosanoids have been identified. These observations raise important questions of whether gene-PUFA interactions are differentially driving the risk of cardiovascular and other diseases in diverse populations, and contributing to health disparities, especially in African American populations.

Figure 1

LCPUFA Biosynthetic Pathways. LCPUFA are derived from C18PUFAs (LA and ALA obtained from the mammalian diet) by alternate desaturation (red/orange enzymes) and elongation (blue enzymes) steps. These enzymes utilize both n-6 and n-3 substrates. n-3 LCPUFAs undergo further metabolism through a β-oxidation step (green box) to the generate DHA.

Figure 2

LCPUFA Cellular Metabolism. The incorporation of LCPUFAs into phospholipids and their release occur by a complex network of enzymatic activities and related genes. Free fatty acids are activated by conjugation with Coenzyme A (CoA) by gene products derived from up to five genes before incorporation into and remodelling through various phospholipids by enzymes differing in both LCPUFA and phospholipid acceptor specificities. Upon cell activation, phospholipases (A₂, C and D) cleave esterified fatty acids from membrane phospholipids. Certain LCPUFAs (such as ARA) can then act as substrates for the eicosanoid-generating enzymes COX1/2, ALOX and P450.

Figure 3

Associations between FADS1 variants and intermediate, molecular phenotypes (IPs), such as metabolic traits. This figure illustrates that the strength of an association is related to the molecular step(s) and metabolite(s) most impacted by genetic variants. In this case, FADS1 variants are most strongly associated with the conversion DGLA to ARA, and thus there are strong associations between the ratio of ARA-containing phospholipids to DGLA-containing phospholipids. This in turn impacts levels of circulating and cellular phospholipid, but the association with FADS1 variants is not as strong because more molecular steps are involved. Levels of phospholipids in turn affects cholesterol levels in lipoprotein particles (as phospholipids and cholesterol esters within lipoprotein particles contain PUFAs) but again the associations are weaker due to the numerous factors (FADS1 variants being only one) that effect total cholesterol levels. Adapted from Gieger et al. [98]

Figure 4

Dramatic differences in the frequency of derived alleles in a 100 kb region surrounding rs174537 in the Human Genome Diversity Panel Data. (A) is the physical location of the SNPs and genes in the region; (B) is SNP name and the specific derived allele in parenthesis; (C) is the derived allele frequency in orange in 52 populations clustered based on geography; (D) reflects the allele associated with increased LC-PUFA metabolism from published studies.; (E) is the detailed overview of rs174537; (F) shows three distributions within the Tangier Island Population. Adapted from Mathias et al. [149] (Panels A–E) and from Mathias et al [111]. (Panel F).