Skip to main content

An epigenome-wide analysis of socioeconomic position and tumor DNA methylation in breast cancer patients

Abstract

Background

Disadvantaged socioeconomic position (SEP), including lower educational attainment and household income, may influence cancer risk and outcomes. We hypothesized that DNA methylation could function as an intermediary epigenetic mechanism that internalizes and reflects the biological impact of SEP.

Methods

Based on tumor DNA methylation data from the Illumina 450 K array from 694 breast cancer patients in the Women’s Circle of Health Study, we conducted an epigenome-wide analysis in relation to educational attainment and household income. Functional impact of the identified CpG sites was explored in silico using data from publicly available databases.

Results

We identified 25 CpG sites associated with household income at an array-wide significance level, but none with educational attainment. Two of the top CpG sites, cg00452016 and cg01667837, were in promoter regions of NNT and GPR37, respectively, with multiple epigenetic regulatory features identified in each region. NNT is involved in β-adrenergic stress signaling and inflammatory responses, whereas GPR37 is involved in neurological and immune responses. For both loci, gene expression was inversely correlated to the levels of DNA methylation. The associations were consistent between Black and White women and did not differ by tumor estrogen receptor (ER) status.

Conclusions

In a large breast cancer patient population, we discovered evidence of the significant biological impact of household income on the tumor DNA methylome, including genes in the β-adrenergic stress and immune response pathways. Our findings support biological effects of socioeconomic status on tumor tissues, which might be relevant to cancer development and progression.

Introduction

There has been growing public and research interest in the deleterious effects of psychosocial stress on human health, including associations with cancer incidence and mortality [1]. As lower socioeconomic position (SEP) can bring about challenges such as financial hardship, individuals from the lower end of the socioeconomic spectrum might more likely experience chronic psychosocial stress, which can lead to physiological and biological changes [2, 3].

One biological process that may be responsive to psychosocial stressors is DNA methylation. As one of the most well-studied epigenetic mechanisms, DNA methylation modulates regional gene expression and activity by adding or removing the methyl group at the CpG dinucleotides [4, 5]. Prior studies reported that DNA methylation was modifiable by various external exposures, such as cigarette smoking [6], body mass index and waist circumference [7], and nutrients of maternal diet [8]. Although the magnitude of DNA methylation changes associated with these factors is generally smaller than the differences observed between breast tumor vs. normal tissues or estrogen receptor (ER) positive vs. negative breast tumors, such epigenetic changes may still confer biologically and clinically significant effects. Moreover, the plasticity and dynamic changes of DNA methylation in response to various exposures make it a relevant intermediary mechanism linking exposures to internal biological effects. In recent years, there has been growing research interest in this burgeoning field of socioepigenomics that focuses on how psychosocial stressors get “under the skin” through epigenetic mechanisms. In an earlier study, it was found that the most socioeconomically disadvantaged group had 17% lower global DNA methylation content in blood than the least disadvantaged group [9]. Occupation has also been associated with DNA methylation, and one study found that workers doing manual labor had 24% lower levels of DNA methylation in blood samples than workers not doing manual labor [9]. Recently, several studies examined life-course SEP with DNA methylation of genes involved in inflammation and stress reactivity and found some genes were hypermethylated in individuals of low SEP [10,11,12,13,14,15,16,17,18].

Notably, most published socioepigenomics studies used DNA derived from peripheral blood [9,10,11,12,13,14,15,16,17]. Interpreting these results for relevance to solid tumors is challenging because of tissue specificity of DNA methylation and uncertain correlations between systemic and local effects of psychosocial stressors. Here, we conducted an epigenome-wide analysis of two of the commonly assessed SEP factors, total household income and educational attainment, using tumor DNA methylation data from a large population of breast cancer patients.

Methods and materials

Patient population

Data and breast tumor tissue samples were from participants in the Women’s Circle of Health Study (WCHS), a case–control study designed to investigate risk factors for aggressive breast cancer in Black and White women. Details on study design and participant recruitment have been described previously [19, 20]. In brief, Black and White women with primary, histologically confirmed invasive breast cancer or ductal carcinoma in situ, ages 20–75 years, were recruited and interviewed from two sites, one in NYC based at Mount Sinai School of Medicine (MSSM), and one in NJ based at The Cancer Institute of New Jersey (CINJ). Data and biospecimens were sent to Roswell Park Comprehensive Cancer Center (RPCCC) in Buffalo, NY, for processing and storage. Controls were frequency matched to cases on age and race and enrolled initially by random digital dialing and later through community health events. In-home interviews were conducted to obtain data on known and suspected risk factors for breast cancer, including educational attainment and total household income. As part of the informed consent, > 95% participants signed a release for their pathology reports and archived specimens in form of formalin-fixed, paraffin-embedded (FFPE) tumor blocks, which were obtained from the pathology departments of the treating hospitals. Data on tumor pathological features, including estrogen receptor (ER) status, were extracted from the pathology reports. This study protocol was approved by Institutional Review Boards (IRBs) at all participating institutes.

DNA methylation assay and bioinformatic processing

DNA was extracted from FFPE breast tumor tissues as previously described [21]. Fragmented DNA was repaired using the Infinium HD DNA Restoration kit before bisulfite conversion. Real-time PCR assays using Illumina Infinium HD FFPE QC Kit (Infinium HD FFPE QC Assay Protocol, Illumina) were used for quality control (QC) assay of FFPE samples. The quality cycle threshold (QCT) value was calculated by subtracting the average Cq of Illumina QC standard from the average Cq value determined for each FFPE sample. Illumina recommends that a QCT value ≤ 5 be utilized for optimal assay performance.

Genome-wide DNA methylation analysis was performed by Roswell Park Genomics Shared Resource (GSR) using the Illumina Infinium humanMethylation450 BeadChip platform according to the manufacturer’s protocol. To minimize batch effects, DNA samples were randomized across plates according to age, ancestry, ER status, and FFPE sample type (slide, punch, or curl) using the OSAT program. The methylation level of each CpG site, expressed as β value, ranged from 0 (unmethylated) to 1 (methylated). The 450 K array data were subjected to rigorous sample and locus-specific quality control criteria, SWAN normalization, and correction for batch effects using the ComBat algorithm [22]. Low-quality probes (probes with detection p value > 0.05 in more than half of samples) and samples with poor detection p values (samples with detection p values < 1 × 10–5 at more than 75% of CpG loci) were removed using the IMA package [23]. We used Bowtie 2 for sequence alignment [24]. Probes known to map ambiguously and that contain single nucleotide polymorphisms were also removed, leaving the final dataset containing 276,108 CpG loci in 694 tumor samples for final analyses [25, 26].

Annotation to regions around CpG sites and analysis of publicly available breast cancer cohort data (TCGA-BRCA)

Genomic and epigenomic context of each CpG site was visualized using the UCSC genome browser. The vertebrate conservation tracks and the transcription factor binding site (TFBS) tracks were surveyed using the UCSC phastCons base-wise conservation database and the UCSC HMR Conserved Transcription Factor Binding Site database, respectively. DNaseI hypersensitive sites indicative of open chromatin structure and transcriptional activity were extracted from ENCODE reference epigenomes. Chromatin-based features were similarly extracted from the Roadmap and ENCODE Project, and inferred chromatin states (ChromHMM) were derived from histone modification profiles. To examine the associations of DNA methylation and gene expression, publicly available RNA-seq and HM450 DNA methylation data from 780 breast cancer patients in The Cancer Genome Atlas (TCGA) were analyzed.

Statistical analysis

Educational attainment and total household income, which are commonly collected in epidemiological studies, were used to capture SEP in the analysis. Log-transformed beta-value of each CpG probe passing the QC filters was analyzed with education and income in linear model using limma R package. The following factors were included in multivariable regression model: age at diagnosis, self-reported race, and tumor ER status. We also added body mass index (BMI), smoking, breastfeeding, menopausal status, and first-degree family history of breast cancer into the model, which did not substantially change the results. Thus, the parsimonious models were used. For sensitivity analysis, we also included recruiting site in the regression model.

Total household income and educational attainment levels were first modeled dichotomously (income below vs. above $50,000; education below vs. above college). For CpG probes that were significant at an array-wide level (p value < 1.8e−7 after Bonferroni correction for testing 276,108 CpG loci), the two SEP factors were further categorized into a 5-level variable (income: 1, $20,000 and below, 2, between $20,000 and $34,999, 3, between $35,000 and $49,999, 4, between $50,000 and $89,999, 5, $90,000 and above; education: 1, below high school, 2, high school or equal, 3, college, 4, college graduate, 5, postgraduate and above) and re-analyzed for dose–response relationships. Subgroup analyses were performed after stratification by race and tumor ER status. Aside from the epigenome-wide analysis, a p value threshold of 0.05 was used for statistical significance. All analyses were performed using R v4.1.1.

Results

Patient population

Table 1 summarizes the demographic and socioeconomic characteristics of the study population with tumor DNA methylation data stratified by family income levels, including 245 cases in low family income group (≤ $50,000) and 314 cases in relative higher family income group (> $50,000). While no differences in first-degree family history of breast cancer (with first-degree family history 16.3% vs 19.4%, p value = 0.34) or ER status distribution (ER-negative status 25.3% vs. 21.3%, p value = 0.27) were found between the two income groups, low family income group had higher BMI (29.6 vs. 27.5 kg/m2, p value = 0.001), higher prevalence of obesity (46.6% vs. 31.6%, p value < 0.001), slightly older age at diagnosis (52.8 vs. 50.5 years, p value = 0.01), less cases in menopausal status (44.1% vs. 57.3%, p value < 0.02), and younger age at first birth (22.4 vs. 26.9 years, p value = 0.004). In comparison with high family income group, low family income group women were also less likely to complete college or attain postgraduate education (48.1% vs. 80.9%, p value < 0.001). Lastly, there was much higher portion of Black women in low family income group than in relatively higher family income group (73.5% vs. 38.2%, p value < 0.001).

Table 1 Descriptive characteristics of breast cancer patients in WCHS with methylation data

Breast tumor DNA methylation with socioeconomic status

In the analysis of household income (> $50,000 vs. ≤ $50,000) in relation to DNA methylation, 25 CpG sites reached Bonferroni significance (p value = 1.8e−7) after adjusting for age, race, and ER status (Fig. 1). In contrast, no CpG site was associated with educational attainment (college and above vs. no college, Additional file 1: Fig. S1). To examine linear “dose–response” between household income and tumor DNA methylation levels, a 5-level income was tested, and the results, together with those from the 2-level income tests, are summarized in Table 2. The direction of associations with all the 25 CpG sites remained consistent, three of which remained array-wide significant, including cg00452016 in nicotinamide nucleotide transhydrogenase (NNT), cg04990372 in MIR1259, and cg01667837 in G protein-coupled receptor 37 (GPR37). After further adjustment for study sites, 5 of the 25 CpG sites were associated with the 2-level income at the array-wide significance level, although all of the associations remained in the same direction with similar effect size (Additional file 5: Table S1).

Fig. 1
figure 1

Manhattan plot for epigenome-wide associations of family income and breast tumor DNA methylation. Linear regression between family income (2-level) and DNA methylation levels was assessed, adjusting for age, race, and tumor estrogen receptor (ER) status. Bonferroni corrections were used to adjust for multiple comparison. The adjusted p values were log-transformed, negated, and plotted based on chromosomal position. The red horizontal line indicates the cutoff significance level of 1.8e−7

Table 2 CpG loci exhibit significant association between family income and breast tumor DNA methylation

When the analyses of the 2-level income were stratified by race or tumor ER status, the results did not differ between the subgroups, with the associations remained in the same direction, although some of the findings in White or ER-negative groups were not statistically significant (p value > 0.05), possibly due to smaller sample size (Additional file 5: Tables S2 and S3). As shown in Fig. 2, the methylation levels of the three array-wide significant CpG sites were higher among those with higher income levels. The trends were again similar between Black and White subgroups, as well as between ER-positive and ER-negative subtypes (Fig. 3), although the difference in methylation status of cg00452016 was not significant in ER-negative subtype.

Fig. 2
figure 2

Top 3 CpG loci from association analysis between 5-level family income and breast tumor DNA methylation. Linear regression was performed, adjusting for age at diagnosis, race, and tumor estrogen receptor (ER) status. The adjusted p values are shown. A cg00452016 close to NNT; B cg01667837 close to GPR37; C cg04990372 close to MIR1259

Fig. 3
figure 3

Differential methylation levels for cg00452016 (A, B), cg01667837 (C, D), and cg04990372 (E, F) between dichotomized income groups. For each of the CpG sites, linear regression was used to test association between income levels and DNA methylation within estrogen receptor (ER)+ or ER− subgroups, or within Black or White subgroups. Regression models were adjusted for age at diagnosis, and race (when stratified by ER status) or ER status (when stratified by race). The adjusted p values are shown. The average methylation level differences between the two ethnic groups and the two ER status groups were also tested using Mann–Whitney U test and the p values are labeled on top

Functional exploration of SEP-associated CpG methylation

In publicly accessible data from TCGA breast cancer cohort, the methylation levels of cg00452016 in NNT (rho = -0.16, p value = 2.6e−5) and cg01667837 in GPR37 (rho = -0.19, p value = 6.3e−7) were inversely correlated with their corresponding gene expression (Additional file 2: Fig. S2). Additional file 3: Fig. S3 and Additional file 4: Fig. S4 show the regulatory features of the genomic regions where cg00452016 and cg01667837 reside, respectively. Both regions contain multiple cis-elements consisting of active promoter of the chromHMM state, DNaseI hypersensitive sites, suggesting multiple cis-epigenetic regulatory mechanisms in these regions.

Discussion

In an epigenome-wide analysis of SEP factors with DNA methylation in breast tumors, we identified 25 CpG sites across the genome associated with total household income, but none with educational attainment. Two of the top CpG sites, cg00452016 and cg01667837 located in the promoter region of NNT and GPR37, respectively, are of particular interest for the multiple epigenetic regulatory features in these regions, their inverse correlations with mRNA expression in breast tumor tissues, and moreover, their putative biological functions in mediating the biological impact of SEP.

One top hit in our association analysis was mapped to NNT, which encodes an enzyme playing crucial roles in maintaining mitochondrial redox balance in a wide variety of organs including the brain. Defective NNT function has been implicated in heart failure [27], neurodegenerative diseases [28], and a rare congenital condition known as familial glucocorticoid deficiency (FGD) [29]. FGD is characterized by an inability of the adrenal cortex to produce cortisol, a key stress hormone molecule that could be triggered by chronic socioeconomic stress in patients from the low SEP group [30]. Herein, cortisol mediates β-adrenergic stress signaling and inflammatory responses within individuals of lower SEP, explaining increased susceptibility to chronic diseases [31]. Thus, income-dependent methylation of NNT might blunt the rising cortisol levels and chronic inflammation that would otherwise be observed in the lower income group.

Another top hit, GPR37, is an orphan G protein-coupled receptor found with high abundance in the central nervous system (CNS). Deregulated GPR37-mediated G protein linked signaling pathway has been implicated in neurological disorders including Parkinson’s disease and autism spectrum disorder [32, 33], as well as psychiatric diseases including bipolar disorder (BPD) and major depression disorder (MDD) [34]. Interestingly, GPR37 levels were decreased in MDD but increased in BPD. Gpr37-knockout female mice showed significantly increased anxiety and depression-like behaviors [35]. Outside the CNS, GPR37 was found in macrophages in the immune system. Activation of GPR37 promotes macrophage phagocytosis and clearance of pathogens [36], and Gpr37-deficient mice showed delayed resolution of inflammatory pain featured by dysregulation of proinflammatory and anti-inflammatory cytokines [37]. Aberrant expression of GPR37 was also implicated in several cancers, including gastric cancer, multiple myeloma, and hepatocellular carcinoma [38,39,40,41], yet little is known of its involvement in breast cancer. Interestingly, recent studies have demonstrated increased β-adrenergic autonomous sympathetic neuronal activities and their immunosuppression effects in breast cancer. Because GPR37 is a receptor for prosaposin [42], a potent neurotrophic factor with activities in inhibition of neuron apoptosis and promoting neurite outgrowth and neuron regeneration [43], GPR37 might be involved in transducing β-adrenergic signaling in response to chronic psychosocial stress in breast tumor tissues. In patients with low household income, demethylation of GPR37 and subsequent increased expression as found in our study might lead to elevated β-adrenergic stress signaling and immunosuppression in the tumor microenvironment.

As lower SEP might lead to chronic psychosocial stress [1], we hypothesized that it could impact cancer development and progression through two biological pathways: alterations in β-adrenergic stress response and immunosuppression [44, 45]. Long-term and continuous stress disturbs the neuroendocrine balance through the hypothalamic–pituitary–adrenal axis, perpetuating a state of psychological and physiological ‘burnout’. Chronic stress also causes high levels of chronic inflammation and suppresses different facets of immune functions, creating a pro-cancer immunological milieu [46]. Our agnostic epigenome-wide analysis now provides new evidence to suggest that stress-response and immunosuppression pathways might be important in understanding how socioeconomic status impacts tumor development and progression.

It should be noted that family income and educational attainment used in our analysis are only two crude proxies of psychosocial stressors, and by no means the only measures in terms of the scope, depth and complexity of stressors from various sources. They may not reflect the actual perceived stress or individual coping capability, either. Nevertheless, the identification of NNT and GPR37 in our study lends support to the validity of household income as a commonly available measure of socioeconomic stress. Although we set out to perform an epigenome-wide analysis without a priori hypothesis, the two most significant genes came from stress-related neuroendocrine response and immune response pathways, the two central biological mechanisms that were previously proposed [44, 45]. This highlights the importance of these two pathways to regulate the biological effects of psychosocial stress. Moreover, the associations of family income and DNA methylation at these two CpG sites appeared to be linear, consistent with the previous findings of a graded association between SEP and health, where each improvement in education, income, occupation, or wealth was associated with better health outcomes [47].

Other limitations of this study include a relatively small sample size that would not allow us to perform separate analysis by race or tumor subtypes, the use of the 450 K methylation array instead of the newer EPIC assay or methylation sequencing, and the one-time assessment of SEP and DNA methylation. It is also important to note that the fold change of methylation status of the CpG sites associated with family income in our study was much smaller than those by tumor vs. normal or by tumor ER status. Future studies are warranted to replicate our findings and to investigate the downstream biological and clinical significance of the methylation changes and whether the findings can be extrapolated to other cancer types.

In conclusion, in a racially diverse breast cancer patient population, we discovered evidence of biological impact of household income on tumor DNA methylation, particularly on genes involved in the β-adrenergic stress and immune response pathways. Because most previous socioepigenomic studies relied on blood samples to assess the systemic impact of socioeconomic status with indirect relevance to local tumor tissues, our new findings support biological effects on tumor tissues by SEP, which are likely to modulate cancer development and progression.

Availability of data and materials

DNA methylation data and phenotype data analyzed in this study have been deposited in GEO under accession number GSE226569.

References

  1. Chida Y, Hamer M, Wardle J, Steptoe A. Do stress-related psychosocial factors contribute to cancer incidence and survival? Nat Clin Pract Oncol. 2008;5(8):466–75. https://doi.org/10.1038/ncponc1134.

    Article  PubMed  Google Scholar 

  2. Cole SW. Human social genomics. PLoS Genet. 2014;10(8):e1004601. https://doi.org/10.1371/journal.pgen.1004601.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. Slavich GM, Cole SW. The emerging field of human social genomics. Clin Psychol Sci. 2013;1(3):331–48. https://doi.org/10.1177/2167702613478594.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Jin B, Li Y, Robertson KD. DNA methylation: superior or subordinate in the epigenetic hierarchy? Genes Cancer. 2011;2(6):607–17. https://doi.org/10.1177/1947601910393957.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Szyf M, McGowan P, Meaney MJ. The social environment and the epigenome. Environ Mol Mutagen. 2008;49(1):46–60. https://doi.org/10.1002/em.20357.

    Article  CAS  PubMed  Google Scholar 

  6. Zhang Y, Elgizouli M, Schottker B, Holleczek B, Nieters A, Brenner H. Smoking-associated DNA methylation markers predict lung cancer incidence. Clin Epigenet. 2016;8:127. https://doi.org/10.1186/s13148-016-0292-4.

    Article  CAS  Google Scholar 

  7. Aslibekyan S, Demerath EW, Mendelson M, Zhi D, Guan W, Liang L, et al. Epigenome-wide study identifies novel methylation loci associated with body mass index and waist circumference. Obesity (Silver Spring). 2015;23(7):1493–501. https://doi.org/10.1002/oby.21111.

    Article  CAS  PubMed  Google Scholar 

  8. Dolinoy DC, Weidman JR, Waterland RA, Jirtle RL. Maternal genistein alters coat color and protects Avy mouse offspring from obesity by modifying the fetal epigenome. Environ Health Perspect. 2006;114(4):567–72. https://doi.org/10.1289/ehp.8700.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. McGuinness D, McGlynn LM, Johnson PC, MacIntyre A, Batty GD, Burns H, et al. Socio-economic status is associated with epigenetic differences in the pSoBid cohort. Int J Epidemiol. 2012;41(1):151–60. https://doi.org/10.1093/ije/dyr215.

    Article  PubMed  Google Scholar 

  10. Stringhini S, Polidoro S, Sacerdote C, Kelly RS, van Veldhoven K, Agnoli C, et al. Life-course socioeconomic status and DNA methylation of genes regulating inflammation. Int J Epidemiol. 2015;44(4):1320–30. https://doi.org/10.1093/ije/dyv060.

    Article  PubMed  Google Scholar 

  11. Needham BL, Smith JA, Zhao W, Wang X, Mukherjee B, Kardia SL, et al. Life course socioeconomic status and DNA methylation in genes related to stress reactivity and inflammation: the multi-ethnic study of atherosclerosis. Epigenetics. 2015;10(10):958–69. https://doi.org/10.1080/15592294.2015.1085139.

    Article  PubMed  PubMed Central  Google Scholar 

  12. Swartz JR, Hariri AR, Williamson DE. An epigenetic mechanism links socioeconomic status to changes in depression-related brain function in high-risk adolescents. Mol Psychiatry. 2017;22(2):209–14. https://doi.org/10.1038/mp.2016.82.

    Article  CAS  PubMed  Google Scholar 

  13. McDade TW, Ryan C, Jones MJ, MacIsaac JL, Morin AM, Meyer JM, et al. Social and physical environments early in development predict DNA methylation of inflammatory genes in young adulthood. Proc Natl Acad Sci USA. 2017;114(29):7611–6. https://doi.org/10.1073/pnas.1620661114.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Santos HP Jr, Bhattacharya A, Martin EM, Addo K, Psioda M, Smeester L, et al. Epigenome-wide DNA methylation in placentas from preterm infants: association with maternal socioeconomic status. Epigenetics. 2019;14(8):751–65. https://doi.org/10.1080/15592294.2019.1614743.

    Article  PubMed  PubMed Central  Google Scholar 

  15. McDade TW, Ryan CP, Jones MJ, Hoke MK, Borja J, Miller GE, et al. Genome-wide analysis of DNA methylation in relation to socioeconomic status during development and early adulthood. Am J Phys Anthropol. 2019;169(1):3–11. https://doi.org/10.1002/ajpa.23800.

    Article  PubMed  Google Scholar 

  16. Laubach ZM, Perng W, Cardenas A, Rifas-Shiman SL, Oken E, DeMeo D, et al. Socioeconomic status and DNA methylation from birth through mid-childhood: a prospective study in Project Viva. Epigenomics. 2019;11(12):1413–27. https://doi.org/10.2217/epi-2019-0040.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Reuben A, Sugden K, Arseneault L, Corcoran DL, Danese A, Fisher HL, et al. Association of neighborhood disadvantage in childhood with DNA methylation in young adulthood. JAMA Netw Open. 2020;3(6):e206095. https://doi.org/10.1001/jamanetworkopen.2020.6095.

    Article  PubMed  PubMed Central  Google Scholar 

  18. Cerutti JK, Lussier AA, Zhu Y, Liu J, Dunn EC. Associations between indicators of socioeconomic position and DNA methylation: a systematic review. medRxiv. 2021. https://doi.org/10.1101/2021.01.21.21250199.

    Article  Google Scholar 

  19. Ambrosone CB, Ciupak GL, Bandera EV, Jandorf L, Bovbjerg DH, Zirpoli G, et al. Conducting molecular epidemiological research in the age of HIPAA: a multi-institutional case–control study of breast cancer in African-American and European-American Women. J Oncol. 2009;2009:871250. https://doi.org/10.1155/2009/871250.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Bandera EV, Chandran U, Zirpoli G, McCann SE, Ciupak G, Ambrosone CB. Rethinking sources of representative controls for the conduct of case-control studies in minority populations. BMC Med Res Methodol. 2013;13:71. https://doi.org/10.1186/1471-2288-13-71.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Espinal AC, Buas MF, Wang D, Cheng DT, Sucheston-Campbell L, Hu Q, et al. FOXA1 hypermethylation: link between parity and ER-negative breast cancer in African American women? Breast Cancer Res Treat. 2017. https://doi.org/10.1007/s10549-017-4418-y.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Leek JT, Johnson WE, Parker HS, Jaffe AE, Storey JD. The sva package for removing batch effects and other unwanted variation in high-throughput experiments. Bioinformatics. 2012;28(6):882–3. https://doi.org/10.1093/bioinformatics/bts034.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Wang D, Yan L, Hu Q, Sucheston LE, Higgins MJ, Ambrosone CB, et al. IMA: an R package for high-throughput analysis of Illumina’s 450K Infinium methylation data. Bioinformatics. 2012;28(5):729–30. https://doi.org/10.1093/bioinformatics/bts013.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9(4):357–9. https://doi.org/10.1038/nmeth.1923.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Chen YA, Lemire M, Choufani S, Butcher DT, Grafodatskaya D, Zanke BW, et al. Discovery of cross-reactive probes and polymorphic CpGs in the Illumina Infinium HumanMethylation450 microarray. Epigenetics. 2013;8(2):203–9. https://doi.org/10.4161/epi.23470.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Zhang X, Mu W, Zhang W. On the analysis of the illumina 450k array data: probes ambiguously mapped to the human genome. Front Genet. 2012;3:73. https://doi.org/10.3389/fgene.2012.00073.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Sheeran FL, Rydstrom J, Shakhparonov MI, Pestov NB, Pepe S. Diminished NADPH transhydrogenase activity and mitochondrial redox regulation in human failing myocardium. Biochim Biophys Acta. 2010;1797(6–7):1138–48. https://doi.org/10.1016/j.bbabio.2010.04.002.

    Article  CAS  PubMed  Google Scholar 

  28. Lopert P, Patel M. Nicotinamide nucleotide transhydrogenase (Nnt) links the substrate requirement in brain mitochondria for hydrogen peroxide removal to the thioredoxin/peroxiredoxin (Trx/Prx) system. J Biol Chem. 2014;289(22):15611–20. https://doi.org/10.1074/jbc.M113.533653.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Meimaridou E, Kowalczyk J, Guasti L, Hughes CR, Wagner F, Frommolt P, et al. Mutations in NNT encoding nicotinamide nucleotide transhydrogenase cause familial glucocorticoid deficiency. Nat Genet. 2012;44(7):740–2. https://doi.org/10.1038/ng.2299.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Cohen S, Doyle WJ, Baum A. Socioeconomic status is associated with stress hormones. Psychosom Med. 2006;68(3):414–20. https://doi.org/10.1097/01.psy.0000221236.37158.b9.

    Article  PubMed  Google Scholar 

  31. Miller GE, Chen E, Fok AK, Walker H, Lim A, Nicholls EF, et al. Low early-life social class leaves a biological residue manifested by decreased glucocorticoid and increased proinflammatory signaling. Proc Natl Acad Sci USA. 2009;106(34):14716–21. https://doi.org/10.1073/pnas.0902971106.

    Article  PubMed  PubMed Central  Google Scholar 

  32. Marazziti D, Golini E, Mandillo S, Magrelli A, Witke W, Matteoni R, et al. Altered dopamine signaling and MPTP resistance in mice lacking the Parkinson’s disease-associated GPR37/parkin-associated endothelin-like receptor. Proc Natl Acad Sci USA. 2004;101(27):10189–94. https://doi.org/10.1073/pnas.0403661101.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Fujita-Jimbo E, Yu ZL, Li H, Yamagata T, Mori M, Momoi T, et al. Mutation in Parkinson disease-associated, G-protein-coupled receptor 37 (GPR37/PaelR) is related to autism spectrum disorder. PLoS ONE. 2012;7(12):e51155. https://doi.org/10.1371/journal.pone.0051155.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Tomita H, Ziegler ME, Kim HB, Evans SJ, Choudary PV, Li JZ, et al. G protein-linked signaling pathways in bipolar and major depressive disorders. Front Genet. 2013;4:297. https://doi.org/10.3389/fgene.2013.00297.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Mandillo S, Golini E, Marazziti D, Di Pietro C, Matteoni R, Tocchini-Valentini GP. Mice lacking the Parkinson’s related GPR37/PAEL receptor show non-motor behavioral phenotypes: age and gender effect. Genes Brain Behav. 2013;12(4):465–77. https://doi.org/10.1111/gbb.12041.

    Article  CAS  PubMed  Google Scholar 

  36. Bang S, Donnelly CR, Luo X, Toro-Moreno M, Tao X, Wang Z, et al. Activation of GPR37 in macrophages confers protection against infection-induced sepsis and pain-like behaviour in mice. Nat Commun. 2021;12(1):1704. https://doi.org/10.1038/s41467-021-21940-8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Bang S, Xie YK, Zhang ZJ, Wang Z, Xu ZZ, Ji RR. GPR37 regulates macrophage phagocytosis and resolution of inflammatory pain. J Clin Investig. 2018;128(8):3568–82. https://doi.org/10.1172/JCI99888.

    Article  PubMed  PubMed Central  Google Scholar 

  38. Wang H, Hu L, Zang M, Zhang B, Duan Y, Fan Z, et al. REG4 promotes peritoneal metastasis of gastric cancer through GPR37. Oncotarget. 2016;7(19):27874–88. https://doi.org/10.18632/oncotarget.8442.

    Article  PubMed  PubMed Central  Google Scholar 

  39. Yang Y, Nishimura I, Imai Y, Takahashi R, Lu B. Parkin suppresses dopaminergic neuron-selective neurotoxicity induced by Pael-R in Drosophila. Neuron. 2003;37(6):911–24. https://doi.org/10.1016/s0896-6273(03)00143-0.

    Article  CAS  PubMed  Google Scholar 

  40. Huang X, Wang Y, Nan X, He S, Xu X, Zhu X, et al. The role of the orphan G protein-coupled receptor 37 (GPR37) in multiple myeloma cells. Leuk Res. 2014;38(2):225–35. https://doi.org/10.1016/j.leukres.2013.11.007.

    Article  CAS  PubMed  Google Scholar 

  41. Liu F, Zhu C, Huang X, Cai J, Wang H, Wang X, et al. A low level of GPR37 is associated with human hepatocellular carcinoma progression and poor patient survival. Pathol Res Pract. 2014;210(12):885–92. https://doi.org/10.1016/j.prp.2014.07.011.

    Article  CAS  PubMed  Google Scholar 

  42. Meyer RC, Giddens MM, Schaefer SA, Hall RA. GPR37 and GPR37L1 are receptors for the neuroprotective and glioprotective factors prosaptide and prosaposin. Proc Natl Acad Sci USA. 2013;110(23):9529–34. https://doi.org/10.1073/pnas.1219004110.

    Article  PubMed  PubMed Central  Google Scholar 

  43. O’Brien JS, Carson GS, Seo HC, Hiraiwa M, Kishimoto Y. Identification of prosaposin as a neurotrophic factor. Proc Natl Acad Sci USA. 1994;91(20):9593–6. https://doi.org/10.1073/pnas.91.20.9593.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Glaser R, Kiecolt-Glaser JK. Stress-induced immune dysfunction: implications for health. Nat Rev Immunol. 2005;5(3):243–51. https://doi.org/10.1038/nri1571.

    Article  CAS  PubMed  Google Scholar 

  45. Antoni MH, Lutgendorf SK, Cole SW, Dhabhar FS, Sephton SE, McDonald PG, et al. The influence of bio-behavioural factors on tumour biology: pathways and mechanisms. Nat Rev Cancer. 2006;6(3):240–8. https://doi.org/10.1038/nrc1820.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Chen M, Singh AK, Repasky EA. Highlighting the potential for chronic stress to minimize therapeutic responses to radiotherapy through increased immunosuppression and radiation resistance. Cancers (Basel). 2020;12(12):3853. https://doi.org/10.3390/cancers12123853.

    Article  CAS  PubMed  Google Scholar 

  47. Adler NE, Stewart J. Health disparities across the lifespan: meaning, methods, and mechanisms. Ann N Y Acad Sci. 2010;1186:5–23. https://doi.org/10.1111/j.1749-6632.2009.05337.x.

    Article  PubMed  Google Scholar 

Download references

Funding

This work was supported by Roswell Park Alliance Foundation Herd of Hope Award, the Breast Cancer Research Foundation, and grants from the National Institutes of Health/National Cancer Institute (R01 CA1332641, P01 CA151135, R01 CA100598) and the US Army Medical Research and Material Command (DAMD-17-01-1-0334), and a gift from the Philip L. Hubbell family. The Roswell Park Comprehensive Cancer Center DataBank and BioRepository (DBBR) and the Genomics Shared Resource (GSR) are Roswell Park CCSG Shared Resources (NIH P30 CA016056).

Author information

Authors and Affiliations

Authors

Contributions

CBA and SY conceived and supervised the projects; QH and SL finished acquisition of the genome-wide methylation data; JC and SY designed data analysis plan, did all statistical analysis and produced all figures; JC and SY interpreted analysis results and wrote original manuscript; MJH, SS, SJM, LY, TK, C-CH, EB, AKS, EAR, EGB, reviewed and edited the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Song Yao.

Ethics declarations

Ethics approval and consent to participate

This study was approved by the institutional review boards of Roswell Park Comprehensive Cancer Center and Rutgers Cancer Institute of New Jersey. Participants in the three studies gave written informed consent for their archived tumor specimens to be obtained from clinical laboratories and used for research related to breast cancer. All methods were carried out in accordance with relevant guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Fig. S1.

Manhattan plot for epigenome-wide associations of educational attainment and breast tumor DNA methylation. Linear regression between education (2-level) and DNA methylation levels were assessed using limma package after adjusting age, race, and estrogen receptor status. Bonferroni corrections were used to adjust for multiple comparison. The adjusted p values were log-transformed, negated, and plotted based on chromosomal position. The red horizontal line indicates the cutoff significance level of 1.8e−7.

Additional file 2: Fig. S2.

Inverse correlation between DNA methylation and gene expression levels in the TCGA BRCA cohorts (780 participants) for cg00452016/NNT (A) and cg01667837/GPR37 (B). Adjusted p values from Spearman test are shown.

Additional file 3: Fig. S3.

Integrative annotation of molecular features of the genomic regions of cg00452016/NNT. From top to bottom, the tracks show: NNT; the cg00452016 site highlighted in vertical cyan line; measurements of evolutionary conservation from alignments of 100 vertebrate species; conserved transcriptional binding sites; the hypersensitive DNaseI sites profile of HMEC and MCF cell lines of the ENCODE project; the HMEC chromHMM tracks indicating putative active (bright red) promoters, strong enhancer (orange), strong transcript (green), as well as putative weak enhancers (yellow); histone modifications of H3K27ac, H3K4me1/3 in various breast cell types.

Additional file 4: Fig. S4.

Integrative annotation of molecular features of the genomic regions of cg01667837/GPR37. From top to bottom, the tracks showing: GPR37; the cg01667837 site highlighted in vertical cyan line; measurements of evolutionary conservation from alignments of 100 vertebrate species; conserved transcriptional binding sites; the hypersensitive DNaseI sites profile of HMEC and MCF cell lines of the ENCODE project; the HMEC chromHMM tracks indicating putative active (bright red) promoters, strong enhancer (orange), strong transcript (green), as well as putative weak enhancers (yellow); histone modifications of H3K27ac, H3K4me1/3 in various breast cell types.

Additional file 5. Supplemental Table S1.

CpG loci exhibit significant association between family income and breast tumor DNA methylation. Linear regression was performed using limma to test associations between DNA methylation levels at each CpG site and family income levels (2- or 5-levels), adjusted for age at diagnosis (continuous values), race (Black vs. White), recruitment sites, and tumor estrogen receptor (ER) status (positive vs. negative). Supplemental Table S2. Associations between family income and methylation levels at significant CpG sites by race. Linear regression was performed using limma to test associations between DNA methylation levels at each CpG site and family income levels (2-levels) in Black and White subgroups separately, adjusted for age at diagnosis (continuous values), recruitment sites, and tumor estrogen receptor (ER) status (positive vs. negative). Supplemental Table S3. Associations between family income and methylation levels at significant CpG sites by tumor estrogen receptor status. Linear regression was performed using limma to test associations between DNA methylation levels at each CpG site and family income levels family income levels (2-levels) in estrogen receptor (ER) status (positive and negative) subgroups separately, adjusted for age at diagnosis (continuous values), recruitment sites, and race (Black vs. White).

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Chen, J., Long, M.D., Sribenja, S. et al. An epigenome-wide analysis of socioeconomic position and tumor DNA methylation in breast cancer patients. Clin Epigenet 15, 68 (2023). https://doi.org/10.1186/s13148-023-01470-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13148-023-01470-4

Keywords