The Relationship between Dietary Fatty Acids and Inflammatory Genes on the Obese Phenotype and Serum Lipids

Obesity, a chronic low-grade inflammatory condition is associated with the development of many comorbidities including dyslipidemia. This review examines interactions between single nucleotide polymorphisms (SNP) in the inflammatory genes tumor necrosis alpha (TNFA) and interleukin-6 (IL-6) and dietary fatty acids, and their relationship with obesity and serum lipid levels. In summary, dietary fatty acids, in particular saturated fatty acids and the omega-3 and omega-6 polyunsaturated fatty acids, impact the expression of the cytokine genes TNFA and IL-6, and alter TNFα and IL-6 production. In addition, sequence variants in these genes have also been shown to alter their gene expression and plasma levels, and are associated with obesity, measures of adiposity and serum lipid concentrations. When interactions between dietary fatty acids and TNFA and IL-6 SNPs on obesity and serum lipid were analyzed, both the quantity and quality of dietary fatty acids modulated the relationship between TNFA and IL-6 SNPs on obesity and serum lipid profiles, thereby impacting the association between phenotype and genotype. Researching these diet–gene interactions more extensively, and understanding the role of ethnicity as a confounder in these relationships, may contribute to a better understanding of the inter-individual variability in the obese phenotype.


Introduction
Obesity, a chronic low-grade inflammatory condition, is associated with the development of many comorbidities, including cardiovascular disease (CVD), type 2 diabetes, and a number of cancers [1]. As more is understood about obesity, the complexity of this chronic disorder becomes more apparent, exhibiting a multi-factorial etiology [2]. Lifestyle factors such as diet and exercise continue to be recognized to play an important role in the development and progression of obesity and its comorbidities. However, genetic variation is also known to contribute to the obese phenotype. Lifestyle factors, including dietary components, such as fatty acids, interact with genetic variants to regulate the development and progression of obesity and its comorbidities. These complex interactions may explain differences observed in the obese phenotype and its comorbidities that vary both within and across populations [3][4][5][6].

Figure 1.
A proposed schematic diagram for obesity-associated low-grade inflammation, and the relationship of diet-gene interactions on obesity and dyslipidemia. Adipocytes become hypertrophic through over-nutrition. Expansion of adipose tissue in obesity leads to a subsequent increase in the production of chemokines by the adipocytes, resulting in increasing macrophage infiltration and enhanced production of pro-inflammatory cytokines, such as TNFα and IL-6. Obesity-associated low-grade inflammation results in an increase in serum trigycerides, and LDL-C concentrations and is associated with dyslipidemia. Environmental factors and DNA sequence variations in inflammatory genes, interact to impact molecular processes of the inflammatory pathway, serum lipids and the obese phenotype. APOA, Apolipoprotein A; APOB, Apolipoprotein B; APOE, Apolipoprotein E; CRP, C-reactive protein; IL-6, interleukin-6; LDL-C, low-density lipoprotein cholesterol; MCP-1, monocyte chemotactic protein-1; PAI, plasminogen activated inhibitor; PPARγ, peroxisome proliferator-activated receptor gamma; TAG, triglycerides; TNFα, tumor necrosis factor.
The aim of this review is to illustrate the current state of knowledge regarding the interaction between dietary fatty acids and cytokines associated with obesity and serum lipids. Dietary fatty acids modulate the regulation and production of tumor necrosis factor alpha (TNFα) and interleukin-6 (IL-6), thereby influencing inflammatory status. Furthermore, emerging research suggests that diet-gene interactions play an important role in the regulation of these inflammatory cytokines, impacting the obese phenotype. This review focuses on interactions between single nucleotide polymorphisms (SNPs) in the inflammatory genes TNFA and IL-6, and dietary fatty acids, and their relationship with obesity and serum lipid levels as proof-of-principle examples. Figure 1 illustrates the development of obesity-associated low-grade inflammation and the impact of diet-gene interactions on obesity and dyslipidemia.

Relationship between Inflammation, Obesity and Lipid Metabolism
White adipose tissue constitutes the majority of adipose tissue (AT) [7] and can be divided into two fractions; the adipocyte fraction composed of mature adipocytes, and the stroma vascular fraction (SVF) composed of many other cell types, including preadipocytes, macrophages, endothelial cells and fibroblasts [8]. Although AT is considered a metabolically active endocrine organ, the macrophages present in the SVF are the primary source of obesity-induced inflammation [9][10][11]. The number of macrophages present in the SVF is directly correlated with the level of adiposity and adipocyte size [10,12]. Adipocyte hypertrophy, results in increased chemokine secretion e.g., transforming growth factor β1, soluble ICAM, and monocyte chemotactic protein-1 (MCP-1) [10]. There is a subsequent increase in the infiltration of macrophages, which in turn secrete cytokines such as IL-6 and TNFα. Since AT expansion is characterized by increased macrophage infiltration, these cells are responsible for almost all TNFα and significant amounts of IL-6 secreted by AT [10].
Low-grade chronic inflammation also mediates all stages of atherosclerosis from initiation through to the complications of thrombosis. So much so, that elevated inflammatory markers are used to define risk of atherosclerotic complications, independent of myocardial damage [13]. One of the proposed mechanisms, linking inflammatory processes to the development of atheroma, is the low-density lipoprotein (LDL) oxidation hypothesis [14]. LDL in the intima, the innermost layer of an artery or vein, undergoes oxidative modification, inducing expression of proinflammatory cytokines, chemokines and other mediators of inflammation. Very-low density lipoprotein (VLDL) and intermediate-density molecules are also believed to undergo oxidative modification and may themselves activate the inflammatory processes of the vascular endothelial cells [15] through activation of nuclear factor kappa-light-chain-enhancer of activated B cells (NFκB) [16].
Genetic sequence variants-known as polymorphisms-within the promoter region of inflammatory genes influence gene transcription, altering protein production. Functional polymorphisms have been reported in the TNFA, IL-1, IL-6 and lymphotoxin-α (LTA) inflammatory genes, altering cytokine production [17][18][19][20]. These polymorphisms have been shown to interact with dietary fatty acids to regulate production and secretion of cytokines, predisposing an individual to inflammation and altering obesity and associated comorbidity risk [6,21].
In a recent review article by Phillips [22], it has been suggested that inconsistencies in previous studies on the influence of cytokine polymorphisms in the risk of obesity, diabetes and the metabolic syndrome (MetS), may be in part explained by the fact that cytokines act in a complex network, and that studying single genes may not provide full insight [22]. The recent LIPGENE-SU.VI.MAX MetS case control study reported by Phillips et al., examined the relationship between LTA, IL-6 and TNFA gene variants and MetS [23]. They reported that the G allele of the TNFA -308 G > A SNP and the minor A allele of the LTA rs915654 SNP were associated with a 20%-40% higher MetS risk [23], but the combined effect of carrying both risk genotypes further increased MetS risk. It was also shown that total plasma PUFA/SFA levels modified the observed additive genetic effects of IL-6, TNFA and LTA [23], highlighting the importance of studying nutrigenetic SNP-SNP or gene-gene interactions.

Dietary Fatty Acids and Inflammation
Dietary fatty acids have received considerable attention for their ability to regulate inflammatory gene expression and secretion. It has been proposed that dietary fatty acids affect inflammatory processes through the modulation of transcription factors such as NFκB and peroxisome proliferator-activated receptor gamma (PPARγ) [24,25]. PPARγ inhibits NFκB, and both transcription factors are sensitive to dietary fatty acids [26].
There is general agreement that increasing dietary SFA intake, especially in overweight or obese individuals, is associated with raised inflammatory markers [27], predominately by activating the toll-like receptor 4 (TLR4) pathway. The TLR4 pathway is expressed in both subcutaneous adipose tissue (SAT) and visceral adipose tissue (VAT). SFAs serve as ligands for TLR4, inducing inflammatory changes in both adipocytes and macrophages through NFκB activation [28], increasing adipocytokine gene expression and production [29,30]. It has been shown that when adipocytes are exposed to the SFA, palmitic acid, IL-6 mRNA expression and protein production increased, most likely through activation of NFκB [31,32]. Similarly, monocytes were directly activated when exposed to SFA, especially lauric acid [30,33].
The n-6 PUFA, linoleic acid (LA), constitutes the majority of PUFA intake in the western diet. LA is the precursor of the n-6 PUFA arachidonic acid (AA). A high LA intake is considered proinflammatory, however the evidence to support this is contradictory and not conclusive [27,[34][35][36]. AA intake in the diet is low relative to LA intake, its metabolic precursor. However, AA is the most prevalent n-6 PUFA in inflammatory cell membranes and is the substrate for the synthesis of the proinflammatory eicosanoids, including prostaglandin E2 (PGE 2 ) and 4-series leukotrienes, associated with inflammatory processes [27]. Despite this, studies investigating the impact of AA on inflammatory markers are inconclusive, and few human intervention studies have reported on the effect of dietary intake of AA on low-grade inflammation [37][38][39].
The n-3 PUFA, alpha-linolenic acid (ALA) is an 18-carbon n-3 essential fatty acid common in canola, soybean oil and some nuts, but in greatest concentrations in flaxseed and flaxseed oil [40]. ALA is elongated and desaturated to eicosapentanoic acid (EPA) and further to decosahexanoic acid (DHA), however the efficiency of this conversion has been debated [41]. Association studies between dietary intake of ALA and inflammatory markers suggest a modest anti-inflammatory effect of ALA [27,34,36,[42][43][44].
The long chain n-3 PUFAs, EPA and DHA are found in seafood, especially oily fish and in some algal oils. Is it proposed that n-3 PUFAs affect inflammation mainly through altered eicosanoid production, but potentially also impacting cell signaling and gene expression [27,42]. When EPA and DHA are incorporated into human inflammatory cells, this is partly at the expense of AA, providing less substrate for eicosanoid production [27]. Culture systems, animal models and human intervention studies have been generally consistent in recognizing the anti-inflammatory actions of n-3 PUFAs [24,25,42,[45][46][47][48].

Tumor Necrosis Factor-α
TNFα was the first cytokine associated with inflammation [20]. TNFα is overexpressed in the AT of obese individuals, with greater expression in visceral fat than in subcutaneous fat [1,49]. TNFα acts in a paracrine manner, suggesting that circulating TNFα levels may not be indicative of actual TNFα levels [50]. Despite this caution, elevated circulating levels of TNFα have been observed in the obese phenotype, which decrease with weight loss [51]. It was originally proposed that adipocytes were the principle source of raised TNFα levels in obesity, however, it is now well recognized that TNFα is abundantly produced by macrophages in the SVF [1,10,49,50]. TNFα has numerous effects in AT including regulating adipogenesis, lipid metabolism and insulin signaling [6,11]. In addition to its ability to increase other pro-inflammatory cytokines [49,52], TNFα has also been associated with a reduction in anti-inflammatory adipokines such as adiponectin [53,54]. Conclusively, TNFα plays a powerful role in regulating inflammatory pathways, favoring an overall inflammatory state [6].
The influence of TNFα on lipid metabolism is complex with the underlying mechanisms still unclear. However, Chen et al. in his review describes four signal pathways involved in TNFα-mediated lipid metabolism [55]. The effects of TNFα's on lipid metabolism occur in different cells, tissues, and organs and include a number of metabolic processes. TNFα induces lipolysis, increasing free fatty acid (FFA) production. In addition, TNFα regulates cholesterol metabolism and other adipocyte-derived adipokines such as leptin, adiponectin, etc. which may also alter lipid metabolism [55].
A number of studies, in cell culture, rodent and human clinical models have confirmed the relationship between TNFα and lipid metabolism [55]. In clinical patients with dyslipidemia, TNFα levels were altered. Compared with healthy subjects, patients with hyperlipoproteinemia showed higher TNFα levels and raised total cholesterol (T-C), triglycerides (TAG), and LDL-C concentrations. Furthermore, after treatment with fenofibrate, T-C, TAG and VLDL-C decreased, which correlated with decreasing TNFα concentrations [56]. It has also been shown that cholesterol-lowering drugs such as simvastatin and atorvastatin decrease TNFα concentrations [57,58], and that blockading TNFα production improves lipid metabolism [59]. Lastly, administration of TNFα interferes with plasma lipid levels [55,60,61]. Feingold et al. have shown that an increase in hepatic VLDL-TAG secretion induced by TNFα is due to both the stimulation of hepatic de novo fatty acid synthesis and an increase in lipolysis [60].

TNFα and Dietary Fatty Acids
Several studies have investigated the effect of dietary fatty acids on TNFα concentrations and TNFA gene expression in cell, animal and human models (Table 1). Plasma TNFα levels and TNFA gene expression increased in 3T3-L1 adipocytes incubated with the SFA palmitic acid, whereas incubation with the MUFA, oleic acid and n-3 PUFA, DHA had no effect [62]. In rodent studies, supplementing the diet with n-3 PUFA decreased TNFA gene expression in mice [63]. In another study, rats were fed a standard diet (18% energy from protein, 76% as carbohydrate (CHO) and 6% of energy as lipid) or a high-fat cafeteria diet (9% energy as protein, 29% as CHO and 62% as lipid). The high fat diet increased both body weight and fat mass, however when these rats were supplemented with EPA they gained less weight, decreased their food intake and increased leptin production. TNFA gene expression was also increased by the high fat diet, but not in the rats supplemented with EPA [64].
Similar results were reported in human studies; Caucasians supplemented with ALA in the form of flaxseed oil in domestic food preparation for four weeks, experienced a 30% reduction in TNFα production. When these subjects were exposed to further supplementation with fish oil (9 g/day) for an additional four weeks, TNFα production was reduced by up to 70%. There was a significant inverse exponential relation between TNFα synthesis and mononuclear cell content of EPA [65]. Similarly, Endres et al. [66] used a radioimmunoassay to measure TNFα produced in vitro by stimulated peripheral-blood mononuclear cells. They reported that supplementation with n-3 PUFA (18 g/day) for six weeks decreased TNFα levels, but these levels returned to baseline levels once supplementation was stopped [66]. However, Grimble et al. demonstrated that the change in TNFα production in response to n-3 PUFA supplementation depended on the subjects' plasma TNFα levels prior to supplementation, with subjects with lower pre-supplementation TNFα levels showing the greatest decrease in TNFα levels post-supplementation [67]. It is possible that the inherent inflammatory status, potentially due to a pre-existing condition such as obesity, could determine the extent of the inflammatory response to different dietary fatty acids. In addition to the independent influence of dietary fatty acids on TNFα production, variation in the TNFA gene may also contribute to the individual variability observed in TNFα production and TNFA gene expression.

TNFA Gene Variants, Obesity and Serum Lipids
Several SNPs have been identified in the promoter region of the TNFA gene, however the TNFA -308 G > A (rs1800629) and -238 G > A (rs361525) SNPs are most commonly associated with measures of adiposity, obesity risk and serum lipids ( Table 2). The A allele of the functional -308 G > A SNP results in a 2-fold increase in TNFA transcription, with a subsequent increase in TNFα production [20]. Several studies have reported that carriers of the pro-inflammatory -308 A allele (AA and GA genotypes) reported a higher body mass index (BMI) and/or percent body fat than those with the GG genotype [68][69][70][71][72][73]. In a recent meta-analysis by Yu et al., including 48 eligible studies, the -308 GA + AA genotypes were associated with an increased risk of obesity (OR, 1.19; 95% CI, 1.02-1.39) [21]. This result was consistent with the study by Sookoian et al., who performed a meta-analysis of 31 observational studies with a total of 3562 individuals and showed that individuals with the GA + AA genotypes had a 23% elevated risk of obesity compared with the GG genotype (OR, 1.23; 95% CI, 1.05-1.45) [74]. In comparison to the -308 G > A SNP, only a few studies have investigated the association between the -238 G > A SNP and obesity ( Table 2). In our laboratory it was found that black South African women with the -238 A allele had greater body fat % than those with the GG genotype [5]. To our knowledge only two papers have found an independent association between the -308 G > A SNP and serum lipid concentrations. In Caucasian men, the -308 A allele was associated with increased TAG [75], and in Polish Caucasian men and women the AA genotype was associated with lower high-density lipoprotein cholesterol (HDL-C) concentrations compared to the GG genotype [76].      (2255) and Meta-analysis (26944)


No association between -174 G > C polymorphism and adiposity.
 T allele of the IVS3 +281 G > T polymorphism associated with adiposity when part of a haplotype.   Not all studies have however shown an association between the -308 G > A and -238 G > A SNPs and obesity [3,4,74,[83][84][85][86]. Further, to our knowledge, no independent associations have been reported between the -238 G > A SNP and serum lipid concentrations. While the A allele of the -308 G > A and -238 G > A SNPs appear to be associated with the obese phenotype and serum lipid concentrations, it is highly likely that genetic variation in the TNFA gene may provide only a partial explanation with regards the inter-individual variability observed and the heterogeneity of the results in these studies. Other variables such as ethnicity, gender, diet, lifestyle and environmental factors may modulate these associations and contribute to the different results observed.

TNFA Gene and Diet Interactions on Obesity and Serum Lipids
The TNFA -308 G > A and -238 G > A SNPs have been shown to modulate the relationship between dietary fat intake on obesity and serum lipid profiles in different populations (Table 3). Despite the many studies showing independent association between the TNFA SNPs and obesity (Table 2), to our knowledge only two studies have investigated diet-gene interactions. Nieters et al. found that German Caucasian men and women with the -308 A allele, who were in the highest tertile for intake of the n-6 PUFAs LA and AA (%E), had an increased obesity risk [71]. More recently, we have reported that the odds of obesity for black South African (SA) women with the -308 A allele increased with total dietary fat intake (%E) [3] (Figure 2). This interaction was not observed in white SA women [4].


In black women, with increasing P:S ratio and n-6:n-3 PUFA ratio, HDL-C decreased, and T-C:HDL-C ratio increased in those with the GA genotype but not the GG genotype.  In black women, with increasing n-3 PUFA intake T-C:HDL-C ratio decreased in those with the GA genotype, but not in those with the GG genotype.  In white SA women, with increasing EPA (%E) intake, LDL-C decreased in the GG genotype but not the GA genotype.
[5]    G allele carriers showed higher TAG and VLDL-C, and higher fasting and post-glucose load free fatty acids levels than C allele carriers. Only one study has reported a diet-gene interaction between the -238 G > A SNP, dietary intake and obesity (Table 3). We found that in black SA women, with increasing total fat, SFA and MUFA intake (%E), weight increased in those with the -238 GA genotype, but not the GG genotype [5]. In contrast, a number of studies have shown interactions between the -308 G > A and -238 G > A SNPs and dietary fat intake on serum lipid profiles (Table 3). In ethnically diverse diabetic Canadians, Fontaine-Bisson et al. reported that PUFA intake was inversely associated with HDL-C concentrations in -308 A allele carriers, but positively associated with HDL-C concentrations in -238 A allele carriers [94]. In a combined analysis in healthy non-diabetic Canadians, they also reported that in individuals with the -308 GA + AA and -238 GG genotypes, an inverse relationship was observed between HDL-C concentrations and n-3, n-6 and total PUFA intake [95].
More recently, we reported interactions between dietary fat intake and the TNFA -308 G > A and -238 G > A SNPs on serum lipid profiles in black and white SA women. With increasing dietary fat intake, serum lipids increased in black women with the -308 GA + AA genotypes; however, with increasing n-3 PUFA and ALA intake, total cholesterol:HDL-cholesterol ratio (T-C:HDL-C ratio) decreased in black SA women [3,4]. For the -238 G > A SNP, in black SA women with increasing polyunsaturated:saturated fat ratio and n-6:n-3 ratio, HDL-C concentrations decreased, and T-C:HDL-C ratio increased in those with the -238 GA genotype but not the GG genotype. However, with increasing n-3 PUFA, T-C:HDL-C ratio decreased in those with the -238 GA genotype, but not in those with the GG genotype. In white SA women, with increasing EPA intake, LDL-C concentrations decreased in those with the -238 GG genotype but not the GA genotype [5]. Interactions between the -238 G > A SNP and dietary fat intake on serum lipids differed depending on the type of dietary fatty acid consumed, and were not the same in black and white SA women suggesting that ethnicity as well as diet-gene interactions contribute to the complexity and heterogeneity observed within serum lipid profiles.

Interleukin-6
During an acute incident, the cytokine IL-6 acts as an anti-inflammatory cytokine, but in a chronic inflammatory condition, IL-6 is pro-inflammatory, as well as being a key regulator of hepatic C-reactive protein (CRP) production [99,100]. IL-6 is secreted by a variety of cells. However approximately 30% of total IL-6 is produced by AT and macrophages that have infiltrated WAT produce approximately 50% of AT derived IL-6 [99][100][101]. It is therefore not surprising that higher circulating concentrations of IL-6 have been associated with obesity, especially visceral fat deposition [102][103][104], and decrease in response to weight loss [102,105].
IL-6 is associated with lipid metabolism and plays a role in the development of atherosclerosis through a number of different mechanisms causing metabolic and endothelial dysfunction. [13,101]. IL-6 impairs insulin action, elevating lipolysis, and increasing FFA release [106]. The increase in FFAs reduces nitric oxide (NO) bioavailability and impairs vasodilation [107]. Insulin resistance and subsequent hyperglycemia also increases lipoprotein oxidation and subsequent increased expression of adipocytokines [108]. In rats administrated IL-6, serum TAG, T-C levels and hepatic TAG secretion increased [109]. Similarly, in human studies, increased IL-6 levels have been associated with an increase in serum TAG and FFAs, and low HDL-C concentrations [110,111].

IL-6 and Dietary Fatty Acids
Several studies have investigated the effect of different dietary fatty acids on IL-6 production and IL-6 gene expression in cell, animal and human models, with all studies showing similar results (Table 1). An in vitro study reported that IL-6 plasma levels and IL-6 gene expression increased when 3T3-L1 adipocytes were incubated with the SFA, palmitic acid, whereas no effect was observed for lauric acid (C12:0) and the n-3 PUFA, DHA [31]. Similarly, macrophages treated with EPA or DHA showed a decrease in LPS stimulated IL-6 mRNA and IL-6 production. In agreement with these findings, rats fed a diet high in n-6 PUFA-rich sunflower oil showed moderate IL-6 release from adipocytes, which was less than when fed a SFA-rich diet, but greater than when fed a MUFA-rich diet [79]. In addition, in mice fed a low fat diet or one of two high-fat diets, consisting of either unsaturated soybean oil or saturated palmitic acid for 16 weeks, MCP-1 expression in AT tissue increased for both high-fat diets compared to the low-fat diet. However, the high saturated fat diet also showed a three-fold increase in IL-6 expression in AT not observed in the soybean oil diet [78].
In human studies (Table 1), increased long-chain n-3 PUFA intake and fish consumption were associated with decreased plasma IL-6 concentrations and other inflammatory markers (matrix metalloproteinase-3, CRP, soluble intercellular adhesion molecule-1) in men from the Multi-Ethnic Study of Atherosclerosis cohort [81]. Furthermore, Ferrucci et al. reported that plasma levels of PUFAs, especially n-3 PUFAs, were independently associated with lower levels of pro-inflammatory markers and higher levels of anti-inflammatory markers [82]. A controlled feeding trial investigating the effects of SFA and MUFA-rich diets on serum lipid concentrations and whole-genome microarray gene expression profiles of AT, found that consuming a SFA-enriched diet for eight weeks resulted in increased expression of genes involved in inflammatory processes in AT including IL-6, and NFκB signaling, whereas the MUFA-enriched diet led to a more anti-inflammatory gene expression profile, accompanied by a decrease in serum LDL-C concentration [80].

IL-6 Gene Variants, Obesity and Serum Lipids
There is growing scientific evidence reporting associations between DNA sequence variants within the IL-6 gene and increased risk of obesity and dyslipidemia (Table 2) [88,91,99,101,112,113]. The most frequently studied IL-6 SNP is -174 G > C (rs1800795). This is a functional SNP, with most studies showing the -174 C allele to be associated with raised IL-6 and CRP concentrations in mostly Caucasian populations [21,91,114,115]. This SNP is rare in individuals of African ancestry, is not informative, and has therefore not been studied in these populations. Association studies between the -174 G > C SNP, obesity and dyslipidemia have yielded conflicting results. A recent meta-analysis found the -174 G > C SNP was associated with obesity [21]. However, this finding was not reproduced in two other meta-analyses [88,89]. The relationship between the -174 G > C SNP appears to be complex in that while the -174 C allele appears to be the risk allele associated with obesity [21], it is the -174 G allele that is associated with raised serum lipids concentrations [87,92,93], despite the -174 C allele being shown to be associated with raised IL-6 and CRP levels [21,91,114,115]. The lack of consistent results in these association studies may be due to interactions between multiple SNPs on the IL-6 gene that may modulate these relationships. This was illustrated by Qi et al., who found no independent association between the -174 G > C SNP and obesity, but did identify an association between an IL-6 haplotype containing the -174 G > C SNP and adiposity in healthy American men and women [89].
A number of studies have reported associations between the -174 G > C SNP and serum lipid profiles ( Table 2). Though not consistent for all the studies [93], for most, in Caucasian men and women the -174 G allele was associated with higher T-C, LDL-C, VLDL-C and TAG concentrations compared to the C allele [88,92,93]. Riikola et al. suggested several possible reasons that may explain the inconclusiveness of study results [87]. These include differences in body mass and body composition, as well as metabolic and pharmacological interference in the different study cohorts. They also suggested that the effect of IL-6 on serum lipids may differ depending on the phase of development of atherosclerosis. Age may also impact these associations as subtle allelic effects observable in young populations may be masked by stronger life-long diet and lifestyle covariates in older populations [88].
Preliminary results from our laboratory found no independent associations between the -174 G > C SNP and obesity or serum lipid profiles in white SA women. However, we did find associations between the less researched SNPs IL-6 IVS4 +869 A > G (rs2069845) and IVS3 +281 G > T (rs1554606) and obesity and serum lipid profiles in black and white SA women. The IVS4 +869 G allele was associated with greater waist and fat mass in black SA women, and the IVS3 +281 T allele was associated with lower TAG concentrations in white SA women compared to those with the GG genotype [90] (Table 2).

IL-6 Gene and Diet Interactions on Obesity and Serum Lipids
As previously described, studies have shown that dietary fat intake modulates the relationship between TNFA gene variants and obesity and serum lipid profiles (Table 3). However, to our knowledge, only two published studies have reported on the relationship between any IL-6 SNP and dietary intake, and both studies investigated IL-6 -174 G > C. Corpeleijn et al. reported that the ability to increase fat oxidation after a high fat load was increased in obese European Caucasians with the -174 C allele [96]. In Spanish men and women with a high CVD risk, the -174 CC genotype was associated with higher levels of adiposity at baseline, however after three years of nutritional intervention, those with the -174 CC genotype following a Mediterranean-style diet, had the greatest reduction in body weight [97].
To our knowledge only a single study has reported an interaction between the -174 G > C SNP and serum lipids levels. In this study, those individuals with the G allele had higher post-glucose load TAG and VLDL-C concentrations, and higher post-glucose load FFA levels than C allele carriers [92].
Preliminary results from our laboratory have recently identified a number of interactions between three IL-6 SNPs (IL-6 -174 G > C, IVS3 +281 G > T, and IVS4 +869 A > G) and dietary fatty acids on obesity and serum lipid levels in both black and white SA women [98] (Table 3). In white SA women, with increasing n-3 PUFA intake and decreasing n-6:n-3 ratio, BMI decreased in those with the -174 C allele, IVS3 +281 T allele and IVS4 +869 AG genotype. In black SA women, with increasing dietary fat intake, adiposity decreased in those with the IVS3 +281 TT and IVS4 +869 GG genotypes and increased in the IVS3 +281 GT + GG and IVS4 +869 AA or AG genotypes (Figure 3). This figure illustrates that the type of fatty acid consumed in the diet interacts with IL-6 genotypes to modulate measures of obesity. When analyzing serum lipid profiles in white SA women, we observed that with increasing n-3 PUFA; TAG and T-C:HDL-C ratio decreased and HDL-C increased in those with the -174 C allele, and T-C:HDL-C ratio decreased in the IVS3 +281 TT genotype compared to the GG and GT genotype, with HDL-C increasing with each IVS3 +281 T allele. In contrast, HDL-C decreased with each IVS4 +869 G allele. In black SA women, with increasing total fat intake, TAG and T-C:HDL-C ratio increased in those with the IVS4 +869 G allele and decreased in the AA genotype [98]. These results suggest that both the quantity and quality of dietary fatty acids modulate the relationship between IL-6 SNPs on measures of obesity and serum lipids, and that these effects may differ according to ethnic group.

The Role of Ethnicity and Gender as Confounders
It has been reported that while the prevalence of obesity may be high in a given population, the prevalence of obesity-associated comorbidities may also differ between ethnic groups [75,[116][117][118][119][120]. As an example; while both African American and black SA women have a higher prevalence of obesity [117,[121][122][123], they have less atherogenic lipid profiles than their white counterparts; characterized by low TAG, T-C, and LDL-C concentrations [118,[124][125][126].
Differences between ethnicities comprise complex etiologies. Ethnicity incorporates a variety of different components including; genetic variation, diet and lifestyle, as well as cultural, behavioral and socio-demographic conditions [120]. In the example of ethnic differences in lipid profiles described above, ethnic variability may be observed in dietary intake, body fat distribution (VAT vs. SAT) [127,128], sequence variation in genes such as apolipoprotein E (APOE) [129], inflammatory gene expression (TNFA and IL-6) [130,131], and genotype frequencies. Genotype and allele frequencies of SNPs discussed in this review paper have been compared in Table 4 for European, British in England and Scotland, African, and African American populations. These SNPs are polymorphic in all these populations. However, there is a very low frequency of the IL-6 -174 C allele in populations of African descent [132].
Of importance, dietary fatty acid intake may differ [133], as well as physical activity levels. In considering differences between ethnic groups resident in developed and developing countries, attitudes to food as well as the quality of the urban environment will also pay a role [120].
In a systematic review on the influence of ethnicity on the relationship between n-3 PUFA intake and CVD, Patel et al. concluded that ethnicity is a factor that accounts for inconsistencies in study results. Specifically, some of the effects of n-3 PUFA are limited to cultures with very high n-3 PUFA intake, and in turn, ethnicity moderates the efficiency with which PUFAs are derived from the diet [133].
Another key consideration in reviewing dietary intake and diet-gene interactions is the impact that certain genes have on dietary fatty acid metabolism, and how these may differ between ethnic groups. Of interest is the fatty acid desaturase gene (FADS), which codes for enzymes in PUFA metabolism. Lu et al. reported how genetic variation in the FADS1 gene interacted with dietary intake of both n-3 and n-6 PUFA to affect T-C and HDL-C concentrations [134]. Furthermore, it has recently been shown that FADS SNPs altered the capacity of different ethnic groups to synthesize long-chain fatty acids [135,136]. Specifically, Sergeant et al. found that FADS genotype frequencies differed significantly between African Americans and European Americans, and SNPs in the FADS genes meant that African Americans were able to more efficiently convert the n-6 fatty acid LA to the proinflammatory fatty acid AA, resulting in higher circulating AA levels, and potentially a more deleterious impact of a diet high in LA in this ethnic group [136]. In this review we have identified for both the TNFA and IL-6 genes, independent associations and diet-gene interactions that differed between ethnic groups. It is likely that the contributing factors described above may partly explain these differences. Like ethnicity, it is likely that the gender (sex) of study participants may impact genotype-phenotype interactions. A number of studies have observed gene-diet-gender interactions, whereby an interaction was identified in one sex, but not another. For example, the FINGEN study, which examined the effect of long chain n-3 PUFA supplementation and APOE genotype on plasma lipids, reported greater TAG lowering effects following dietary intervention in APOE4 males than in females [139]. Further, results from the Framingham Heart Study showed that dietary PUFA intake modulates the effect of the apolipoprotein A1 (APOA1) -75 G > A polymorphism (rs670) on plasma HDL-C concentrations in women but not in men [140]. In addition, Phillips et al. have shown that the TCF7L2 rs7903146 polymorphism influences MetS risk, and is impacted by both gender and dietary SFA intake [141]. Phillips et al. have also identified associations between genetic variants of the apolipoprotein B and APOA1 gene and MetS risk, however the modulation of MetS risk by dietary fat intake observed in the entire cohort was observed in the male high-fat consumers only [142].

Conclusions
This review suggests that DNA sequence variations in genes involved in inflammation may interact with environmental exposures, such as dietary intake, to modulate an individual's susceptibility to developing obesity and its comorbidities. In summary, dietary fatty acids, in particular SFAs and the n-3 and n-6 PUFAs, impact the expression of the cytokine genes TNFA and IL-6, and alter TNFα and IL-6 production. In addition, sequence variants in these genes have also been shown to alter their gene expression and plasma levels, and are associated with obesity, measures of adiposity and serum lipid concentrations. When interactions between dietary fatty acids with TNFA and IL-6 SNPs on obesity and serum lipid were analyzed, it became evident that both the quantity and quality of dietary fatty acids modulate the relationship between TNFA and IL-6 SNPS on obesity and serum lipid profiles, thereby impacting the association between phenotype and genotype. The inter-individual variability in the obese phenotype and the inconsistencies in study results may be better understood by researching these diet-gene interactions more extensively.
Although this review focused on only two of the inflammatory cytokines, many other adipokines and chemokines have been associated with obesity (e.g., adiponectin, IL-1, IL-10, and MCP-1), and are sensitive to dietary fatty acid intake and should be studied further. In addition, it is also important to unravel the molecular mechanisms that govern the impact of dietary fatty acids on inflammation, which remain unclear and appear to act via multiple pathways. While it is known that different groups of dietary fatty acids (SFA, MUFA, and n-3, n-6 and n-9 PUFA) differ in their effect on inflammatory gene expression and plasma levels [27], these results are not conclusive, especially with regards the n-6 PUFA class of fatty acids which have at times been shown to act in both a pro-and anti-inflammatory manner, and appear to be modulated by gene variants.
The nutrigenetic interactions described in this review are complex and it is difficult to assess the magnitude of their impact in managing an individual's diet. Furthermore, the inter-ethnic variability reported should caution us with regards singular dietary recommendations, as it cannot be assumed that fatty acids and other nutrient metabolism is uniform for all populations. In addition to diet-single polymorphism interactions, it is also important to understand the combined effect of a number of polymorphisms on the same or different genes interacting with the environment. For this reason, the identification and analysis of haplotype-nutrient interactions may provide additional insights in future research. The future study of nutrigenomics offers the opportunity to clarify the underlying molecular mechanisms governing the interactions between dietary fatty acids and the inflammatory phenotype, potentially elucidating the observed differences between different ethnic groups and genders in developing population-specific dietary recommendations.