Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Apr 25.
Published in final edited form as: Mol Nutr Food Res. 2015 Nov 23;60(2):410–419. doi: 10.1002/mnfr.201500436

The effects of omega-3 polyunsaturated fatty acids and genetic variants on methylation levels of the interleukin-6 gene promoter

Yiyi Ma 1,2, Caren E Smith 1, Chao-Qiang Lai 1, Marguerite R Irvin 3, Laurence D Parnell 1, Yu-Chi Lee 1, Lucia D Pham 1, Stella Aslibekyan 3, Steven A Claas 3, Michael Y Tsai 4, Ingrid B Borecki 5, Edmond K Kabagambe 6, José M Ordovás 1,7,8,*, Devin M Absher 9,*, Donna K Arnett 3,*
PMCID: PMC4844557  NIHMSID: NIHMS772890  PMID: 26518637

Abstract

Scope

Omega-3 PUFAs (n-3 PUFAs) reduce IL-6 gene expression, but their effects on transcription regulatory mechanisms are unknown. We aimed to conduct an integrated analysis with both population and in vitro studies to systematically explore the relationships among n-3 PUFA, DNA methylation, single nucleotide polymorphisms (SNPs), gene expression, and protein concentration of IL6.

Methods and results

Using data in the Genetics of Lipid Lowering Drugs and Diet Network (GOLDN) study and the Encyclopedia of DNA Elements (ENCODE) consortium, we found that higher methylation of IL6 promoter cg01770232 was associated with higher IL-6 plasma concentration (p = 0.03) and greater IL6 gene expression (p = 0.0005). Higher circulating total n-3 PUFA was associated with lower cg01770232 methylation (p = 0.007) and lower IL-6 concentration (p = 0.02). Moreover, an allele of IL6 rs2961298 was associated with higher cg01770232 methylation (p = 2.55 × 10−7). The association between n-3 PUFA and cg01770232 methylation was dependent on rs2961298 genotype (p = 0.02), but higher total n-3 PUFA was associated with lower cg01770232 methylation in the heterozygotes (p = 0.04) not in the homozygotes.

Conclusion

Higher n-3 PUFA is associated with lower methylation at IL6 promoter, which may be modified by IL6 SNPs.

Keywords: DNA methylation, Gene-by-environment interaction, Genetic variant, Interleukin-6, n-3 polyunsaturated fatty acids

1 Introduction

Omega-3 PUFAs (n-3 PUFAs) ameliorate systematic inflammation through reduction of plasma concentration of IL-6 [1–3], for which higher concentrations are associated with cardiovascular disease (CVD) [4]. Further, supplementation with both eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) for 8 wk was found to reduce not only the plasma concentration of IL-6 but also its gene expression in adipose tissue [5].

Gene expression of interleukin-6 (IL6) has been shown to be regulated by DNA methylation of its promoter region [6–9]. The observed effect of IL6 methylation on its gene expression may be related to the presence of potential binding sites for the methyl CpG binding protein 2 (MeCP2), which are located from positions −666 to −426 relative to the transcription start site [9]. Therefore, the effect of n-3 PUFA on IL6 promoter methylation as a mechanism to reduce concentration of this inflammatory biomarker is intriguing.

Adding to the complexity of the modulation of n-3 PUFA on IL6 methylation is the role of genetic sequence variation. For example, single nucleotide polymorphisms (SNPs) were found to affect DNA methylation across the whole genome [10]. Moreover, IL6 SNPs were shown to interact with different environmental factors, including diet [11], smoking [12,13], and social position to determine plasma IL-6 concentration [14]. This evidence demonstrates the modifying role of IL6 sequence variants on the effect of environmental factors on plasma IL-6 concentration. Therefore, the objective of this study is to explore the relationships between n-3 PUFA, genetic variants, DNA methylation, and gene expression at IL6 locus.

To achieve this objective, we first identified the methylation site relevant to both transcription and translation within the IL6 locus based on the DNA methylation-expression relationship across all 17 human cell types available with both methylation and gene expression at IL6 from the Encyclopedia of DNA Elements (ENCODE) consortium, and the DNA methylation–protein relationship across 848 participants of the Genetics of Lipid Lowering Drugs and Diet Network (GOLDN) study. We further investigated in GOLDN participants the association between n-3 PUFA and methylation levels of the identified CpG site, and evaluated whether the observed association could be further modified by IL6 SNPs.

2 Materials and methods

2.1 Study population

The GOLDN study, designed to evaluate genetic factors that modulate lipid responses to diet and fenofibrate treatment, recruited participants from the National Heart, Lung, and Blood Institute Family Heart Study [15]. The study design and methodology were described previously [16]. Study protocol was approved by the Human Studies Committee of Institutional Review Board at the University of Minnesota, University of Utah, and Tufts University/New England Medical Center. All participants provided written informed consent. The current analysis consisted of 848 individuals (459 men and 389 women) after removing those with missing variables and those taking hormone replacement therapies because of their mixed effects on IL-6 [17].

2.2 Biochemical measurements

Blood samples from each participant were collected, stored frozen at −70°C, and analyzed at the same time to eliminate interassay variability. IL-6, IL-2-soluble receptor α, tumor necrosis factor (TNF) α, and monocyte chemoattractant protein 1 (MCP-1) were measured using quantitative sandwich enzyme immunoassay techniques (ELISA kit assays, R&D System Inc., Minneapolis, MN, USA) as described previously [18]. High-sensitivity C-reactive protein was measured using a latex particle enhanced immunoturbidimetric assay (Kamiya Biomedical Company, Seattle, WA, USA) as described previously [19]. Plasma adiponectin was measured using competitive RIA (Linco Research, St Charles, MO, USA) as described previously [20]. Fatty acids in the erythrocyte membranes were measured by a capillary Varian CP7420 100 m column with a Hewlett Packard 5890 gas chromatograph equipped with a HP6890A autosampler [21]. The measurements of fatty acids were reliable and have been validated against a diet history questionnaire [22,23].

2.3 Genotyping and DNA methylation in GOLDN

The region of interest, referred as IL6 locus, was defined as the region from 1 kb upstream of the CpG island of IL6 promoter to 1 kb downstream of the 3′ untranslated region of IL6 (Fig. 1A). At IL6 locus, genotypes for 39 SNPs were obtained using the Affymetrix Genome-Wide Human SNP array 6.0 (Affymetrix, Santa Clara, CA, USA) with the genomic DNA extracted from blood samples using Gentra Puregene Blood Kits (Gentra Systems, Inc., Minneapolis, MN, USA).

Figure 1.

Figure 1

Open in a new tab

Methylation of CpG site cg01770232 within IL6 promoter is associated with both plasma IL-6 concentration and IL6 gene expression. (A) The genomic structure of IL6 locus, in which exons, CpG island, potential binding site for MeCP2, CpG sites, and SNPs are represented by black box, gray box, gray circle, black dots, and black bars, respectively. (B) The Pearson correlation between methylation level of cg01770232 and IL6 gene expression across 17 human cell types in ENCODE, in which each black dot represents one cell type and the gray line represents the corresponding regression line, and the correlation coefficient (r) and its corresponding p-value are shown at the bottom right corner. (C) The association of quartiles of methylation level of cg01770232 with plasma-logged IL-6 concentration (geometric means ± 95% CI) in GOLDN. p-Value is obtained from the generalized linear model adjusting for first four principal components for cellular purity and population structure, age, sex, center, pedigree, smoking, alcohol intake, total energy intake, physical activity, intake of vitamin B-12 and folate, acute inflammatory conditions (infection or fever), chronic diseases (abdominal obesity, CVD, diabetes, and hypertension), other inflammatory markers (MCP-1, adiponectin, high-sensitivity C-reactive protein, tumor necrosis factor α, and IL-2-soluble receptor α). TSS, transcription start site; UTR, untranslated region.

Two CpG sites (cg01770232 and cg26061582) were included into the analysis because they are located within the potential binding site for the MeCP2 [9]. Their methylation levels were measured using Infinium Human Methylation 450K BeadChip (Illumina, San Diego, CA) [24] with 500 ng sodium bisulfite treated DNA (Zymo Research Corporation, Irvine, CA, USA) extracted from CD4+ T-cells (Qiagen, Germantown, MD, USA) [24], which were isolated from the stored buffy coats collected at the same time erythrocyte fatty acids and plasma cytokines were measured. CD4+ T cells were selected because of its relevance to the proinflammatory cytokines and the established temporal stability [25].

2.4 Gene expression and DNA methylation data from ENCODE

In ENCODE, methylation and gene expression data at IL6 locus were available for 17 human cell types, including one human primary hepatocyte and 16 human cell lines (Supporting Information Table 1), and all of these 17 available human cell types were included into the correlation analysis. Methylation data were downloaded (1 September, 2014) from UCSC genome browser HAIB Methyl450 track (http://hgdownload.cse.ucsc.edu/goldenPath/hg19/encodeDCC/wgEncodeHaibMethyl450/). Gene expression data were downloaded (1 September, 2014) from UCSC genome browser Duke Affymetrix Exon Array track (http://hgdownload.cse.ucsc.edu/goldenPath/hg19/encodeDCC/wgEncodeDukeAffyExon/).

2.5 Statistical methods

In ENCODE, Pearson correlation was conducted between methylation of the candidate CpG sites and IL6 gene expression. In GOLDN, χ2 tests, analysis of variance, and analysis of covariance analyses were conducted to examine the differences in population characteristics and potential confounding factors by erythrocyte total n-3 PUFA, categorized in quartiles. Generalized linear models were applied to test the main associations and interactions among the methylation levels of both CpG sites, plasma IL-6 concentration, genotypes of IL6 SNPs, and erythrocyte n-3 PUFA, including total n-3 PUFA, EPA, and DHA. The methylation levels of CpG sites were represented as quartiles for the exposure variables and as continuous variables for the outcome variable. The analysis adjusted for the potential confounding factors including pedigree, principal components of cellular purity [26] and population structure, age, sex, study center, smoking, alcohol intake, total energy intake, physical activity, vitamin B-12 intake, folate intake, acute inflammatory conditions (infection or fever), chronic diseases known to affect IL-6 such as abdominal obesity [27], CVD [4], diabetes [28], and hyper-tension [29], and other inflammatory markers known to affect IL-6, including high-sensitivity C-reactive protein [22], tumor necrosis factor α [30], MCP-1 [31], IL-2-soluble receptor α [22], and adiponectin [32]. Log transformation was performed for those variables not following a normal distribution. All data were analyzed using SAS (version 9.3 for Windows; SAS Institute, Inc. Cary, NC, USA). A two-tail p-value of <0.05 was considered statistically significant.

3 Results

3.1 Population characteristics

Population characteristics were compared across quartiles of erythrocyte total n-3 PUFA (Table 1). Individuals in the quartiles with higher erythrocyte total n-3 PUFA tended to be older, consumed more dietary n-3 PUFA and less total energy, and had lower concentration of MCP-1(p < 0.05) than those in the lower quartiles. Also, the quartiles with higher levels of erythrocyte n-3 PUFA tended to contain fewer smokers, and more individuals with abdominal obesity than those with lower levels (p < 0.05). Individuals in the quartiles with higher n-3 PUFA tended to have lower plasma IL-6 concentration (p = 0.02).

Table 1.

Population characteristics of GOLDN

Variable Erythrocyte n-3 PUFAs
 p
Q1 (n = 212) Q2 (n = 212) Q3 (n = 212) Q4 (n = 212)
Erythrocyte n-3 PUFA Total n-3 PUFA (% total fatty acids)a) 4.65 5.25 5.96 7.08
EPA (% total fatty acids)b) 0.39 (0.09) 0.45 (0.10) 0.53 (0.12) 0.80 (0.48) <0.0001
DHA (% total fatty acids)b) 2.10 (0.30) 2.59 (0.30) 3.10 (0.32) 4.12 (0.73) <0.0001
Dietary n-3 PUFA intake Total n-3 PUFAs (% total energy intake/day)a) 0.026 (0.02) 0.03 (0.02) 0.04 (0.03) 0.06 (0.05) <0.0001
EPA (% total energy intake/day)a) 0.006 (0.006) 0.008 (0.007) 0.01 (0.009) 0.02 (0.02) <0.0001
DHA (% total energy intake/day)a) 0.017 (0.01) 0.02 (0.01) 0.025 (0.02) 0.04 (0.03) <0.0001
Demographics Age (y)b) 42 (14) 45 (16) 51 (15) 57 (15) <0.0001
Men (n)c) 124 (58) 119 (56) 107 (50) 109 (51) 0.29
Lifestyle parameters Current smoker (n)c) 27 (13) 24 (11) 12 (6) 4 (2) 0.0003
Drinker (n)c) 114 (54) 112 (53) 93 (44) 93 (44) 0.06
Total energy intake (kcal/day)b) 2313.95 (1496.83) 2322.94 (1325.29) 2182.88 (1193.17) 1898.38 (836.63) 0.001
Vitamin B-12 intake (mcg)b) 5.62 (3.74) 5.84 (3.76) 5.56 (3.33) 5.18 (3.47) 0.29
Folate intake (mcg)b) 423.61 (235.84) 416.24 (263.21) 421.98 (229.78) 404.28 (198.43) 0.83
Disease status Infection or fever (n)d) 2 (1) 5 (2) 4 (2) 2 (1) 0.45
Abdominal obesity (n)d) 86 (41) 121 (57) 114 (54) 108 (51) 0.01
Hypertension (n)d) 30 (14) 52 (25) 54 (26) 83 (39) 0.40
Diabetes (n)d) 8 (4) 19 (9) 12 (6) 25 (12) 0.98
CVD (n)d) 5 (2) 10 (5) 11 (5) 24 (11) 0.29
Inflammatory markers IL-6 (pg/mL)e) 1.75 (1.13) 1.75 (1.12) 1.65 (1.11) 1.57 (1.12) 0.02
hsCRP (mg/dl)f) 0.11 (1.09) 0.15 (1.11) 0.11 (1.09) 0.12 (1.09) 0.69
TNF-α (pg/mL)f) 2.92 (1.03) 3.08 (1.03) 2.91 (1.03) 3.06 (1.05) 0.57
MCP-1 (pg/mL)f) 212.66 (1.02) 210.69 (1.03) 200.22 (1.02) 198.26 (1.02) 0.006
Adiponectin (ng/mL)f) 7762.81 (1.05) 6460.1 (1.05) 6930.59 (1.04) 6915.36 (1.05) 0.22
IL2sR-α (pg/mL)f) 1000.05 (1.03) 1020.96 (1.03) 959.39 (1.03) 964.97 (1.03) 0.16
Open in a new tab
a)

Data are medians of each quartile of total n-3 PUFAs.

b)

Data are means (SDs) within each quartile of total n-3 PUFAs and p-values are obtained from analysis of variance test.

c)

Data are n (%) within each quartile of total n-3 PUFAs and p-values are obtained from Chi-square test.

d)

Data are means (SDs) within each quartile of total n-3 PUFAs and p-values are obtained from logistic regression adjusting for the first four principal components for cellular purity and population structure, age, sex, center, smoking, alcohol intake, total energy intake, physical activity, and intake of vitamin B-12 and folate.

e)

Data are least squared adjusted means (standard errors) within each quartile of total n-3 PUFAs and p-values are obtained from analysis of covariance test adjusting for the first four principal components for cellular purity and population structure, age, sex, center, pedigree, smoking, alcohol intake, total energy intake, physical activity, intake of vitamin B-12 and folate, acute inflammatory conditions (infection or fever), chronic disease status (including abdominal obesity, CVD, diabetes, and hypertension), and other inflammatory markers (including hsCRP, TNF-α, MCP-1, IL2sR-α, and adiponectin).

f)

Data are least squared adjusted means (standard errors) within each quartile of total n-3 PUFAs, p-values are obtained from analysis of covariance test adjusting for the first four principal components for cellular purity and population structure, age, sex, center, pedigree, smoking, alcohol intake, total energy intake, physical activity, intake of vitamin B-12 and folate.

hsCRP, high-sensitivity C-reactive protein, IL2sR-α, IL-2-soluble receptor α, TNF-α, tumor necrosis factor.

3.2 Cg01770232 methylation with protein concentration and gene expression of IL6

Two CpG sites located within the potential binding site for the MeCP2 (cg01770232 and cg26061582) were included in the analysis (Fig. 1A). Methylation level of cg01770232 was correlated with IL6 gene expression across 17 human cell types in ENCODE (r = 0.75 and p = 0.0005; Fig. 1B). In GOLDN, methylation level of cg01770232 was also associated with plasma IL-6 concentration (β = 0.22 and p = 0.03; Fig. 1C). We did not find any significant associations for cg26061582 (data not shown).

3.3 n-3 PUFA with protein concentration and cg01770232 methylation of IL6

We compared associations of n-3 PUFA with the methylation level of cg01770232, and also with plasma IL-6 concentration in GOLDN (Fig. 2). Total n-3 PUFA and DHA were negatively associated with plasma IL-6 (p = 0.02 and 0.01 respectively), while the negative association with EPA did not reach statistical significance. In parallel with the association patterns observed between total n-3 PUFA, EPA, and DHA for the outcome of plasma IL-6 concentrations, the same fatty acids were also negatively associated with methylation level of cg01770232 (p = 0.007 and 0.01 for total n-3 PUFA and DHA, respectively), and the negative association with EPA reached a borderline significance (p = 0.05).

Figure 2.

Figure 2

Open in a new tab

Associations of erythrocyte n-3 PUFA with plasma IL-6 and DNA methylation level of cg01770232 in GOLDN. Based on adjusted generalized linear models, predicted plasma concentration of IL-6 (log transformed) is plotted against total n-3 PUFA (A), EPA (C), and DHA (E), and predicted methylation level of cg01770232 is also plotted against total n-3 PUFA (B), EPA (D), and DHA (F). β- and p-values represent the regression coefficients and statistical significance, respectively, and both of them are obtained from the generalized linear model adjusting for first four principal components for cellular purity and population structure, age, sex, center, pedigree, smoking, drinking, total energy intake, physical activity, intake of vitamin B-12 and folate, acute inflammatory conditions (infection or fever), chronic disease status (including abdominal obesity, CVD, diabetes, and hypertension), other inflammatory markers (MCP-1, adiponectin, high-sensitivity C-reactive protein, tumor necrosis factor α, and IL-2-soluble receptor α).

3.4 Rs2961298 with cg01770232 methylation at IL6

Since we have test associations between 39 SNPs within the region of interest and methylation of cg01770232, the Bonferroni-adjusted significance level should be 0.001, which was obtained by dividing 0.05 by 39. Evidence for associations between all 39 SNPs within IL6 locus and the methylation level of cg01770232 is presented in Fig. 3 and Supporting Information Table 2. Rs2961298, an SNP close to the CpG island upstream of IL6 gene, exhibited the most robust association with the methylation level of cg01770232 (p = 2.55 × 10−7).

Figure 3.

Figure 3

Open in a new tab

Genetic associations with the methylation level of cg01770232 in GOLDN within the IL6 locus. SNPs are plotted by position within IL6 locus against association with the methylation level of cg01770232 (−log10 p-value). Estimated recombination rates (from 1000 Genomes Pilot 1 CEU) are plotted in gray line to reflect the local linkage disequilibrium structure. The SNPs surrounding the most significant SNP (rs2961298), represented by dark gray diamonds, are plotted as gray-coded squares to reflect their linkage disequilibrium r2 with this SNP, which is estimated based on 1000 Genomes Pilot 1 CEU database.

3.5 n-3 PUFA, rs2961298, and cg01770232 methylation

Because of the identified associations of both SNP rs2961298 and n-3 PUFA with the methylation level of cg01770232, the potential interactions between fatty acids and SNPs were further explored (Fig. 4). Rs2961298 was shown to significantly modify the association between cg01770232 methylation and total n-3 PUFA (pinteraction = 0.02), EPA (pinteraction = 0.01), and DHA (pinteraction = 0.05). However, we found a significant lowering effect of n-3 PUFA on cg01770232 methylation only in the rs2961298 heterozygotes (AC; ptrend = 0.04, 0.06, and 0.05 for total n-3 PUFA, EPA, and DHA, respectively), not in the homozygotes (AA and CC).

Figure 4.

Figure 4

Open in a new tab

Interaction between n-3 PUFA and rs2961298 modulating the methylation level of cg01770232 in GOLDN. Based on adjusted generalized linear models, predicted methylation levels of cg01770232 (log transformed) by genotype of rs2961298 is plotted against total n-3 PUFA (A), EPA (B), and DHA (C), adjusted for first four principal components for cellular purity and population structure, age, sex, center, pedigree, smoking, alcohol intake, total energy intake, physical activity, intake of vitamin B-12 and folate, acute inflammatory conditions (infection or fever), chronic disease status (including abdominal obesity, CVD, diabetes, and hypertension), and other inflammatory markers (MCP-1, adiponectin, high-sensitivity C-reactive protein, tumor necrosis factor α, and IL-2-soluble receptor α). p-values indicate the statistical significance of the adjusted interaction term and adjusted regression coefficients (represented as betas) in the regression line corresponding to three genotype groups of rs2961298 (diamond for CC, square for AC, and triangle for AA).

4 Discussion

In the current study, we identified a transcriptionally and translationally relevant CpG site within the IL6 promoter, cg01770232, and its association with n-3 PUFA in erythrocyte membranes, which was further modified by IL6 SNP rs2961298. This is the first study to integrate analysis of datasets from both a population cohort and in vitro experiments to systematically explore the associations between n-3 PUFA and different components of IL6, including epigenetic modification, genetic variants, gene expression, and protein concentration. Our consistent association findings provide potential mechanistic hypotheses, calling for further experimental demonstrations.

Our finding of the positive correlation between cg01770232 methylation and IL6 gene expression is consistent with the observations of those promoters with low content of CG because the CG content within 500 bp of cg01770232 is just 1.4%. It was shown that the hypermethylation status of some of the CG poor promoters was correlated with greater gene expression [33–35], and more active promoter status [36]. Also, the positive correlation between methylation and gene expression shown by the ENCODE datasets was in line with the positive correlation between methylation and IL-6 protein concentration in GOLDN population. Moreover, the parallel negative associations between n-3 PUFA and both IL-6 protein concentration and IL6 methylation in GOLDN population further provided indirect evidence for the observed positive correlation between IL6 methylation and gene expression. Although a previous publication reported a negative association between methylation and expression [9], our study may provide additional solid evidence based on a greater number and variety of cell types from data in ENCODE, and a more unbiased statistical analysis. Finally, the unbiased nature of the publically available datasets from ENCODE consortium may also help to increase the level of scientific evidence for our findings.

Although we have obtained consistent associations among n-3 PUFA, IL6 methylation, gene expression, and IL-6 protein concentration, we do not have the power to detect whether the effect of n-3 PUFA on IL-6 protein concentration is through its effect on IL6 methylation because of our observational study design. Potential mechanisms to account for the observed association between n-3 PUFA and IL6 methylation are unknown. One possibility may be the potential interactions between n-3 PUFA and MeCP2, the protein binding to the methylated cg01770232 site at IL6 locus. This mechanistic possibility is supported by previous studies reporting protective effect of n-3 PUFA on Rett Syndrome, a neurodevelopmental disorder mainly caused by mutations in MECP2 gene [37, 38]. Of course, there could be also many other possibilities that are beyond the scope of the current study.

Our finding that only subjects with heterozygous geno-type of rs2961298 have the significant lowering effects of total n-3 PUFA on IL6 promoter methylation suggests that another SNP sharing linkage disequilibrium with rs2961298 might be responsible for the differential methylation. The phenomenon of genotype-dependent methylation is widely spread across the whole human genome [10], but the specific mechanism driving the allele-specific differential methylation at IL6 is unclear. It could be possible that different genotypes may have different affinity for DNA-binding proteins with downstream effects on DNA methylation [39]. It could also be possible that DNA sequence had direct effects on the propensity of DNA methylation [40].

Our study was limited by its cross-sectional design, from which only associations rather than causality can be established. Further longitudinal and intervention studies are needed in order to confirm and solidify these findings. Although we have analyzed all the publically available cell types in ENCODE to remove the potential of selection bias to some degree, we admit that there could still be some remaining selection bias because not all the cell types were included in ENCODE. Lack of replication and inability to directly assess relationships with gene expression in GOLDN should also be taken into consideration. Additionally, our findings can be appropriately generalized only to those populations with similar intakes of n-3 PUFA.

An important strength of the current study is that it provides evidence for potential mechanisms by which long-term intake of n-3 PUFA is modulating genotype-based variability in disease biomarkers, because the erythrocyte measurement of n-3 PUFA reflects long-term intake [41] and the measurements were validated to be correlated with dietary intake (Table 1). While previous studies reported the interactions between sequence variants and environmental factor that modulate plasma IL-6 concentration [11–14], our work suggests that the modification effect of IL6 SNPs on the relationship between n-3 PUFA and IL-6 may occur through IL6 methylation and gene expression. Specifically, SNP rs2961298 interacted significantly with total n-3 PUFA to modulate methylation status of the identified functional CpG site, cg01770232. When evaluated in the same human population, n-3 PUFAs were found to be associated with methylation level of cg01770232 in a genotype-specific manner. Our findings may lead to novel mechanistic explanations of the link between n-3 PUFA and IL-6, and could identify potential new diagnostic and therapeutic targets for disease prevention.

Supplementary Material

SUPP TABLE 1
SUPP Table 2

Acknowledgments

Y.M. designed research including project conception and development of overall research plan; C.E.S., and J.M.O. designed research including study oversight; Y.M., C.Q.L., M.R.I., and Y.C.L. conducted research; Y.M. analyzed data; Y.M. and C.E.S. wrote the paper. D.K.A. and D.A. provided essential data. L.D.P., L.D.P., S.A., M.Y.T., I.B.B., and E.K.K. contributed to manuscript editing. Y.M. had primary responsibility for final content.

This study is funded by National Heart Lung and Blood Institute Grant U01HL072524-04 and 5R01HL1043135-04, and the National Institute of Neurological Disorders and Stroke Grant T32NS054584. C.E.S. is supported by K08 HL112845. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture. The USDA is an equal opportunity provider and employer. This material is based upon work supported by the U.S. Department of Agriculture, under agreement No. 58-1950-0-014. All authors read and approved the final manuscript. Any opinions, findings, conclusion, or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the U.S. Department of Agriculture.

Abbreviations

CVD

cardiovascular disease

DHA

docosahexaenoic acid

ENCODE

Encyclopedia of DNA Elements

EPA

eicosapentaenoic acid

GOLDN

Genetics of Lipid Lowering Drugs and Diet Network

IL2sR-α

interleukin 2 soluble receptor α

MCP-1

monocyte chemoattractant protein 1

MeCP2

methyl CpG binding protein 2

n-3 PUFA

omega-3 PUFA

SNP

single nucleotide polymorphism

TNF-α

tumor necrosis factor α

Footnotes

Additional supporting information may be found in the online version of this article at the publisher's web-site

The authors have declared no conflict of interest.

References

  • 1.Farzaneh-Far R, Harris WS, Garg S, Na B, et al. Inverse association of erythrocyte n-3 fatty acid levels with inflammatory biomarkers in patients with stable coronary artery disease: The Heart and Soul Study. Atherosclerosis. 2009;205:538–543. doi: 10.1016/j.atherosclerosis.2008.12.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Trebble T, Arden NK, Stroud MA, Wootton SA, et al. Inhibition of tumour necrosis factor-alpha and interleukin 6 production by mononuclear cells following dietary fish-oil supplementation in healthy men and response to antioxidant co-supplementation. Br. J. Nutr. 2003;90:405–412. doi: 10.1079/bjn2003892. [DOI] [PubMed] [Google Scholar]
  • 3.Kelley DS, Siegel D, Fedor DM, Adkins Y, et al. DHA supplementation decreases serum C-reactive protein and other markers of inflammation in hypertriglyceridemic men. J. Nutr. 2009;139:495–501. doi: 10.3945/jn.108.100354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Cesari M, Penninx BW, Newman AB, Kritchevsky SB, et al. Inflammatory markers and onset of cardiovascular events: results from the Health ABC study. Circulation. 2003;108:2317–2322. doi: 10.1161/01.CIR.0000097109.90783.FC. [DOI] [PubMed] [Google Scholar]
  • 5.Itariu BK, Zeyda M, Hochbrugger EE, Neuhofer A, et al. Long-chain n-3 PUFAs reduce adipose tissue and systemic inflammation in severely obese nondiabetic patients: a randomized controlled trial. Am. J. Clin. Nutr. 2012;96:1137–1149. doi: 10.3945/ajcn.112.037432. [DOI] [PubMed] [Google Scholar]
  • 6.Tekpli X, Landvik NE, Anmarkud KH, Skaug V, et al. DNA methylation at promoter regions of inter-leukin 1B, interleukin 6, and interleukin 8 in non-small cell lung cancer. Cancer Immunol. Immunother. 2013;62:337–345. doi: 10.1007/s00262-012-1340-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Stefani FA, Viana MB, Dupim AC, Brito JA, et al. Expression, polymorphism and methylation pattern of interleukin-6 in periodontal tissues. Immunobiology. 2013;218:1012–1017. doi: 10.1016/j.imbio.2012.12.001. [DOI] [PubMed] [Google Scholar]
  • 8.Nile CJ, Read RC, Akil M, Duff GW, et al. Methylation status of a single CpG site in the IL6 promoter is related to IL6 messenger RNA levels and rheumatoid arthritis. Arthritis Rheum. 2008;58:2686–2693. doi: 10.1002/art.23758. [DOI] [PubMed] [Google Scholar]
  • 9.Dandrea M, Donadelli M, Costanzo C, Scarpa A, et al. MeCP2/H3meK9 are involved in IL-6 gene silencing in pancreatic adenocarcinoma cell lines. Nucleic Acids Res. 2009;37:6681–6690. doi: 10.1093/nar/gkp723. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Zhi D, Aslibekyan S, Irvin MR, Claas SA, et al. SNPs located at CpG sites modulate genome-epigenome interaction. Epigenetics. 2013;8:802–806. doi: 10.4161/epi.25501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Kanoni S, Dedoussis GV, Herbein G, Fulop T, et al. Assessment of gene-nutrient interactions on inflammatory status of the elderly with the use of a zinc diet score–ZINCAGE study. J. Nutr. Biochem. 2010;21:526–531. doi: 10.1016/j.jnutbio.2009.02.011. [DOI] [PubMed] [Google Scholar]
  • 12.Shin KK, Jang Y, Koh SJ, Chae JS, et al. Influence of the IL-6 -572C>G polymorphism on inflammatory markers according to cigarette smoking in Korean healthy men. Cytokine. 2007;39:116–122. doi: 10.1016/j.cyto.2007.06.005. [DOI] [PubMed] [Google Scholar]
  • 13.Sunyer J, Forastiere F, Pekkanen J, Plana E, et al. Interaction between smoking and the interleukin-6 gene affects systemic levels of inflammatory biomarkers. Nicotine Tob. Res. 2009;11:1347–1353. doi: 10.1093/ntr/ntp144. [DOI] [PubMed] [Google Scholar]
  • 14.Sanderson SC, Kumari M, Brunner EJ, Miller MA, et al. Association between IL6 gene variants -174G>C and -572G>C and serum IL-6 levels: interactions with social position in the Whitehall II cohort. Atherosclerosis. 2009;204:459–464. doi: 10.1016/j.atherosclerosis.2008.09.019. [DOI] [PubMed] [Google Scholar]
  • 15.Higgins M, Province M, Heiss G, Eckfeldt J, et al. NHLBI Family Heart Study: objectives and design. Am. J. Epidemiol. 1996;143:1219–1228. doi: 10.1093/oxfordjournals.aje.a008709. [DOI] [PubMed] [Google Scholar]
  • 16.Corella D, Arnett DK, Tsai MY, Kabagambe EK, et al. The -256T>C polymorphism in the apolipoprotein A-II gene promoter is associated with body mass index and food in-take in the genetics of lipid lowering drugs and diet network study. Clin. Chem. 2007;53:1144–1152. doi: 10.1373/clinchem.2006.084863. [DOI] [PubMed] [Google Scholar]
  • 17.Georgiadou P, Sbarouni E. Effect of hormone replacement therapy on inflammatory biomarkers. Adv. Clin. Chem. 2009;47:59–93. doi: 10.1016/s0065-2423(09)47003-3. [DOI] [PubMed] [Google Scholar]
  • 18.Tsai MY, Hanson NQ, Straka RJ, Hoke TR, et al. Effect of influenza vaccine on markers of inflammation and lipid profile. J. Lab. Clin. Med. 2005;145:323–327. doi: 10.1016/j.lab.2005.03.009. [DOI] [PubMed] [Google Scholar]
  • 19.Aslibekyan S, Kabagambe EK, Irvin MR, Straka RJ, et al. A genome-wide association study of inflammatory biomarker changes in response to fenofibrate treatment in the Genetics of Lipid Lowering Drug and Diet Network. Pharmacogenet. Genomics. 2012;22:191–197. doi: 10.1097/FPC.0b013e32834fdd41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Shen J, Arnett DK, Peacock JM, Parnell LD, et al. Interleukin1beta genetic polymorphisms interact with polyun-saturated fatty acids to modulate risk of the metabolic syndrome. J. Nutr. 2007;137:1846–1851. doi: 10.1093/jn/137.8.1846. [DOI] [PubMed] [Google Scholar]
  • 21.Cao J, Schwichtenberg KA, Hanson NQ, Tsai MY. Incorporation and clearance of omega-3 fatty acids in erythrocyte membranes and plasma phospholipids. Clin. Chem. 2006;52:2265–2272. doi: 10.1373/clinchem.2006.072322. [DOI] [PubMed] [Google Scholar]
  • 22.Kabagambe EK, Ordovas JM, Tsai MY, Borecki IB, et al. Smoking, inflammatory patterns and post-prandial hypertriglyceridemia. Atherosclerosis. 2009;203:633–639. doi: 10.1016/j.atherosclerosis.2008.08.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Kabagambe EK, Tsai MY, Hopkins PN, Ordovas JM, et al. Erythrocyte fatty acid composition and the metabolic syndrome: a National Heart, Lung, and Blood Institute GOLDN study. Clin. Chem. 2008;54:154–162. doi: 10.1373/clinchem.2007.095059. [DOI] [PubMed] [Google Scholar]
  • 24.Ma Y, Smith CE, Lai CQ, Irvin MR, et al. Genetic variants modify the effect of age on APOE methylation in the Genetics of Lipid Lowering Drugs and Diet Network study. Aging Cell. 2015;14:49–59. doi: 10.1111/acel.12293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Flanagan JM, Brook MN, Orr N, Tomczyk K, et al. Temporal stability and determinants of white blood cell DNA methylation in the breakthrough generations study. Cancer Epidemiol. Biomarkers Prev. 2015;24:221–229. doi: 10.1158/1055-9965.EPI-14-0767. [DOI] [PubMed] [Google Scholar]
  • 26.Frazier-Wood AC, Aslibekyan S, Absher DM, Hopkins PH, et al. Methylation at CPT1A locus is associated with lipoprotein subfraction profiles. J. Lipid Res. 2014;55:1324–1330. doi: 10.1194/jlr.M048504. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Brydon L, Wright CE, O'Donnell K, Zachary I, et al. Stress-induced cytokine responses and central adiposity in young women. Int. J. Obes. 2008;32:443–450. doi: 10.1038/sj.ijo.0803767. [DOI] [PubMed] [Google Scholar]
  • 28.Duncan BB, Schmidt MI, Pankow JS, Ballantyne CM, et al. Low-grade systemic inflammation and the development of type 2 diabetes: the atherosclerosis risk in communities study. Diabetes. 2003;52:1799–1805. doi: 10.2337/diabetes.52.7.1799. [DOI] [PubMed] [Google Scholar]
  • 29.Zhang W, Wang W, Yu H, Zhang Y, et al. Interleukin 6 underlies angiotensin II-induced hypertension and chronic renal damage. Hypertension. 2012;59:136–144. doi: 10.1161/HYPERTENSIONAHA.111.173328. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Schindler R, Mancilla J, Endres S, Ghorbani R, et al. Correlations and interactions in the production of interleukin-6 (IL-6), IL-1, and tumor necrosis factor (TNF) in human blood mononuclear cells: IL-6 suppresses IL-1 and TNF. Blood. 1990;75:40–47. [PubMed] [Google Scholar]
  • 31.Kim CS, Park HS, Kawada T, Kim JH, et al. Circulating levels of MCP-1 and IL-8 are elevated in human obese subjects and associated with obesity-related parameters. Int. J. Obes. 2006;30:1347–1355. doi: 10.1038/sj.ijo.0803259. [DOI] [PubMed] [Google Scholar]
  • 32.Abke S, Neumeier M, Weigert J, Wehrwein G, et al. Adiponectin-induced secretion of interleukin-6 (IL-6), monocyte chemotactic protein-1 (MCP-1, CCL2) and interleukin-8 (IL-8, CXCL8) is impaired in monocytes from patients with type I diabetes. Cardiovasc. Diabetol. 2006;5:17. doi: 10.1186/1475-2840-5-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Rishi V, Bhattacharya P, Chatterjee R, Rozenberg J, et al. CpG methylation of half-CRE sequences creates C/EBPalpha binding sites that activate some tissue-specific genes. Proc. Natl. Acad. Sci. U S A. 2010;107:20311–20316. doi: 10.1073/pnas.1008688107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Flower K, Thomas D, Heather J, Ramasubramanyan S, et al. Epigenetic control of viral life-cycle by a DNA-methylation dependent transcription factor. PLoS One. 2011;6:e25922. doi: 10.1371/journal.pone.0025922. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Vinokurova S, von Knebel Doeberitz M. Differential methylation of the HPV 16 upstream regulatory region during epithelial differentiation and neoplastic transformation. PLoS One. 2011;6:e24451. doi: 10.1371/journal.pone.0024451. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Weber M, Hellmann I, Stadler MB, Ramos L, et al. Distribution, silencing potential and evolutionary impact of promoter DNA methylation in the human genome. Nat. Genet. 2007;39:457–466. doi: 10.1038/ng1990. [DOI] [PubMed] [Google Scholar]
  • 37.De Felice C, Cortelazzo A, Signorini C, Guerranti R, et al. Effects of omega-3 polyunsaturated fatty acids on plasma proteome in Rett syndrome. Mediators Inflamm. 2013;2013:723269. doi: 10.1155/2013/723269. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.De Felice C, Signorini C, Leoncini S, Pecorelli A, et al. The role of oxidative stress in Rett syndrome: an overview. Ann. N Y Acad. Sci. 2012;1259:121–135. doi: 10.1111/j.1749-6632.2012.06611.x. [DOI] [PubMed] [Google Scholar]
  • 39.Kerkel K, Spadola A, Yuan E, Kosek J, et al. Genomic surveys by methylation-sensitive SNP analysis identify sequence-dependent allele-specific DNA methylation. Nat. Genet. 2008;40:904–908. doi: 10.1038/ng.174. [DOI] [PubMed] [Google Scholar]
  • 40.Lienert F, Wirbelauer C, Som I, Dean A, et al. Identification of genetic elements that autonomously determine DNA methylation states. Nat. Genet. 2011;43:1091–1097. doi: 10.1038/ng.946. [DOI] [PubMed] [Google Scholar]
  • 41.Smith CE, Follis JL, Nettleton JA, Foy M, et al. Dietary fatty acids modulate associations between genetic variants and circulating fatty acids in plasma and erythrocyte membranes: meta-analysis of nine studies in the CHARGE consortium. Mol. Nutr. Food Res. 2015;59:1373–1383. doi: 10.1002/mnfr.201400734. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

SUPP TABLE 1
NIHMS772890-supplement-SUPP_TABLE_1.docx (14.6KB, docx)
SUPP Table 2
NIHMS772890-supplement-SUPP_Table_2.docx (19.8KB, docx)

RESOURCES