Abstract
There is evidence that epigenetic changes occur early in breast carcinogenesis; we hypothesized that early life exposures associated with breast cancer would be associated with epigenetic alterations in breast tumors. In particularly, we examined DNA methylation patterns in breast tumors in association with several early life exposures in a population-based case-control study. Promoter methylation of E-cadherin, p16 and RAR-β2 genes was assessed in archived tumor blocks from 803 cases with real time methylation-specific PCR. Unconditional logistic regression was used for case-case comparisons of those with and without promoter methylation. We found no differences in the prevalence of DNA methylation of the individual genes by age at menarche, age at first live birth, and weight at age 20. In case-case comparisons of premenopausal breast cancer, lower birth weight was associated with increased likelihood of E - cadherin promoter methylation (OR = 2.79, 95% CI, 1.15–6.82, for ≤ 2.5 vs 2.6–2.9 kg); higher adult height with RAR-β2 methylation (OR =3.34, 95% CI, 1.19–9.39, for ≥ 1.65 vs <1.60 m); and not having been breast fed with p16 methylation (OR = 2.75, 95% CI, 1.14–6.62). Among postmenopausal breast cancers, birth order was associated with increased likelihood of p16 promoter methylation. Being other than first in the birth order was inversely associated with likelihood of ≥1 of the 3 genes being methylated for premenopausal breast cancers, but positively associated with methylation in postmenopausal women. These results suggest that there may be alterations in methylation associated with early life exposures that persist into adulthood and affect breast cancer risk.
Keywords: early life exposure, breast cancer, epidemiology, epigenetics, promoter methylation
Introduction
There is evidence that fetal and early life exposures play a critical role in development of breast cancer in adulthood (1). The development of the breast is a progressive process, initiated in the embryonic period, continuing throughout childhood particularly during puberty, with terminal differentiation during the first full-term pregnancy (2, 3). Age at menarche and age of first live birth are well-established risk factors for breast cancer. Other exposures related to the in utero environment including infant birth size (birth weight and birth length), mother’s age at delivery, birth order, and mother’s pre-eclampsia may also affect risk (4–6). There is some evidence suggesting an inverse association of exposure to breast milk in infancy with breast cancer risk (7, 8). Childhood and adolescent body mass index (BMI) also appear to be inversely associated with breast cancer risk (9, 10). Moreover, there is a consistently positive association between height and breast cancer (1, 11). The biological mechanisms underlying the observed associations are not clear; one possible mechanism is that there are epigenetic alterations resulting from these exposures which play a role in breast carcinogenesis.
Commonly found in human neoplasias, epigenetic alterations in response to early life exposures may influence the risk of adult diseases (2, 12, 13) , including cancer. Both global hypomethylation and gene-specific promoter hypermethylation are prominent features of breast tumors (14). The E-cadherin, p16, and RARβ2 genes are tumor suppressor genes involved in key cellular processes including cell cycle regulation, cell-to-cell interaction, hormone- and receptor-mediated cell signaling, apoptosis and angiogenesis (15). Promoter methylation of these genes occurs frequently in breast cancer cell lines and breast tumors (16–18). A recent animal study reported that in utero and lactational exposure to environmental contaminants decreased expression of p16 gene and slightly increased methylation of CpG sites in the promoter region of p16 in prepubertal female offspring rats (19). In addition, there is both in vivo and in vitro evidence that prenatal and early postnatal nutrition can influence patterns of DNA methylation and cause changes in gene expression in the offspring (12, 13). In a study from the Netherlands, individuals periconceptionally exposed to famine during the Dutch Hunger Winter of 1944–45 were found to have persistent hypomethylation of IGF2 compared with their unexposed, same-sex siblings (20). To our knowledge, there are no studies of the associations of early life exposures with DNA methylation in breast tumors.
To better understand the relationship of early life exposures with gene-specific promoter hypermethylation in breast tumor, we evaluated DNA promoter methylation status of E-cadherin, p16, and RARβ2 genes in breast tumors in a population-based study.
Materials and Methods
Study population
Briefly, the Western New York Exposures and Breast Cancer Study (WEB Study) was a population-based case-control study, including female residents of Erie and Niagara counties in western New York State who were diagnosed with primary, histologically confirmed, incident breast cancer between 1996 and 2001, and were 35–79 years of age. This report includes data from cases only. Among 1,638 eligible cases, 1,170 (72%) participated. All participants provided informed consent, and the protocol was approved by the Institutional Review Boards of all the participating institutions.
Extensive in-person interviews and self-administered questionnaires were administered to participants, including queries regarding demographics factors, family history of breast cancer, medical history, alcohol drinking and smoking history, physical activity and other breast cancer risk factors. Information on menstrual and reproductive history included age at menarche, outcome and duration of each pregnancy, and age at each live birth. Participants were also queried regarding their birth weight and whether they were breastfed by their mother or not. Additionally, participants were asked to recall their body weight for each decade of their lives from age 20 until 12–24 months prior to diagnosis for cases. Current height was measured by trained interviewers according to a standardized protocol. Participants provided information on the age of their mother at the time of their birth and on their place in the birth order in their families. Data included information on menstruation and menopausal status. Women were considered postmenopausal if their menses had ceased permanently and naturally, or if they had undergone any of the following conditions: a bilateral oophorectomy, a hysterectomy without removal of the ovaries and were older than 50, or radiation or other medical treatment which resulted in permanent cessation of their menses and were older than 55.
Information on tumor size, histological grade, and cancer stage was abstracted from medical records by trained research nurses using a standardized protocol. Estrogen receptor (ER) status was determined by a single pathologist, by immunohistochemical analysis as described previously (21)
Tumor block promoter methylation determination
Archived tumor blocks were successfully obtained from 920 (78.6%) of all participant breast cancer cases. Tumor samples were microdissected from fixed microscope slides. Bisulfite modification was performed on 2µg of tumor DNA isolated from the dissected tissue in accordance with methods described elsewhere (21, 22). Promoter methylation of E- cadherin, p16, and RAR-β2 was determined by the fluorescence-based version of methylation-specific PCR using real time PCR amplification of bisulfite converted DNA in an ABI 7900HT real time PCR system as previously described (21, 22). Briefly, each reaction contained 5 µl of Taqman Universal Master Mix, 4.5 µl of bisulfite-treated DNA and 0.5 µl of a 60× assay by design premix containing the primers and probes that were designed for each respective gene (Applied Biosystems, Carlsbad CA); primers and probes sequences were published previously (21). As a control to check for modified viable DNA, we used an assay for the ACTB gene with primers and probes specifically designed for CpG free sites within the gene sequence, thus amplifying the modified DNA regardless of the methylation status. If the ACTB result was negative (i.e. no amplification signal was detected), the DNA was not used in subsequent assays, and re-modification was attempted; the other 3 genes were assayed only if ACTB was positive. Each individual DNA sample was assayed in triplicate for each gene for quality control purposes. Additionally, as a positive control, universally methylated DNA (CpGenome; Norcross, GA) was used along with water blanks as a negative control. We had successful promoter methylation results for 803 cases.
Statistical analysis
Early life characteristics of participating cases with and without promoter methylation of each specific gene were compared using ANOVA for continuous variables and the χ2 test for categorical variables. Birth weight was classified into three categories: <2.5kg, 2.6–3.9kg, and >3.9kg. Maternal age at delivery was categorized into three groups <25 years, 25–30 years, and ≥30 years. Tertile distributions among cases were used to categorize data on age at menarche and adult height. Age at first birth was classified into three groups <23 year, 23–30 years, and ≥30 years among parous cases. Weight at age 20 year was divided into two groups with a cutoff based on the median in controls.
Unconditional logistic regression was used for case-case comparisons of those with and without promoter methylation to evaluate odds ratios (ORs) and 95% confidence intervals (95% CIs) for the associations of early life exposures with the likelihood of promoter methylation. All analyses were adjusted by age, education level, and race. Because we had previously found an association of methylation with ER status (21), we also adjusted for ER status. Potential confounding effect by PR status and other known risk factors of breast cancer, including family history of breast cancer, were further examined, and no appreciable confounding was observed. We also found that maternal age at delivery did not modify the effect of both birth weight and birth order, and adjusting it categorically produced similar results. Those results are not shown. We evaluated associations stratifying on menopausal status. All statistical tests were based on two-sided probability. Statistical analyses were conducting using SAS, Version 9.2 (SAS Institute, Cary, NC).
Results
We previously reported case control associations between early life exposures and breast cancer (10). Analyses here were limited to case-case comparisons in relation to DNA methylation.
Demographic characteristics of cases with and without promoter methylation of E - cadherin, p16, and RAR-β2 genes have been shown in detail elsewhere (22). Briefly, there were no differences in the methylation frequency by age at diagnosis, race, education, menopausal status, age at menopause, and recent body mass index (BMI). Table 1 shows comparisons of cases with and without promoter methylation of the three genes for early life exposure factors. After adjusting for current age, cases with promoter methylation of E - cadherin gene were younger age at first birth than those without methylation of this gene. There were no differences in other early life exposure factors comparing cases with or without methylation of E - cadherin, p16, or RAR-β2 genes.
Table 1.
Descriptive characteristics of breast cancer cases and controls by hypermethylation status of E-cadherin, p16, and RAR-β2, WEB Study 1996–20011
| E-cadherin | p16 | RAR-β2 | ||||
|---|---|---|---|---|---|---|
| M (n = 161) 2 | UM 2 | M (n = 208) 2 | UM 2 | M (n = 221) 2 | UM 2 | |
| Age, y | 58.0 ± 11.8 | 57.4 ± 11.2 | 58.0 ± 11.2 | 57.4 ± 11.3 | 57.4 ± 11.3 | 57.6 ± 11.3 |
| Postmenopausal | 111 (68.9%) | 455 (70.9%) | 147 (70.7%) | 419 (70.4%) | 157 (71.0%) | 409 (70.3%) |
| Early life exposure factors 3 | ||||||
| Birth weight (kg) | ||||||
| ≤2.5 (Below average) | 21 (13.9%) | 75 (12.6%) | 19 (9.9%) | 77 (13.9%) | 25 (12.6%) | 71 (13.0%) |
| 2.6–3.9 (Average) | 110 (72.9%) | 446 (75.0%) | 152 (79.6%) | 404 (72.8%) | 145 (72.9%) | 411 (75.1%) |
| >3.9 (Above average) | 20 (13.2%) | 74 (12.4%) | 20 (10.5%) | 74 (13.3%) | 29 (14.5%) | 65 (11.95) |
| Birth order | ||||||
| First born | 52 (32.5%) | 222 (35.0%) | 61 (29.6%) | 213 (36.2%) | 78 (35.8%) | 196 (34.0%) |
| Other than first born | 108 (67.5%) | 412 (65.0%) | 145 (70.4%) | 375 (63.8%) | 140 (64.2%) | 380 (66.0%) |
| Having been breastfed | ||||||
| Yes | 62 (48.1%) | 251 (49.2%) | 81 (49.1%) | 232 (48.9%) | 76 (42.9%) | 237 (51.3%) |
| No | 67 (51.9%) | 259 (50.8%) | 84 (50.9%) | 242 (51.1%) | 101 (57.1%) | 225 (48.7%) |
| Maternal age at delivery (yrs) | ||||||
| <25 | 55 (36.7%) | 251 (41.0%) | 76 (39.0%) | 230 (40.5%) | 81 (39.1%) | 225 (40.4%) |
| 25–30 | 48 (32.0%) | 184 (30.0%) | 58 (29.7%) | 174 (30.6%) | 71 (34.3%) | 161 (29.0%) |
| >30 | 47 (31.3%) | 178 (29.0%) | 61 (31.3%) | 164 (28.9%) | 55 (26.6%) | 170 (30.6%) |
| Age at menarche, y | 12.5 ± 1.5 | 12.6 ± 1.6 | 12.6 ± 1.6 | 12.6 ± 1.6 | 12.6 ± 1.5 | 12.5 ± 1.6 |
| Age at first live birth, y4 | 23.4 ± 4.9† | 24.5 ± 4.8 | 24.0 ± 4.7 | 24.4 ± 4.9 | 24.0 ± 4.5 | 24.3 ± 5.0 |
| Nulliparous women | 30 (18.6%) | 111 (17.3%) | 42 (20.2%) | 99 (16.7%) | 33 (14.9%) | 108 (18.6%) |
| Parity 4 | 2.9 ± 1.5 | 2.8 ± 1.4 | 3.0 ± 1.5 | 2.8 ± 1.4 | 2.9 ± 1.5 | 2.8 ± 1.4 |
| Adult height (m) | 1.62 ± 0.07 | 1.62 ± 0.07 | 1.63 ± 0.06 | 1.62 ± 0.07 | 1.63 ± 0.06 | 1.62 ± 0.07 |
| Weight at 20 years (kg) | 55.2 ± 8.8 | 55.3 ± 9.3 | 55.2 ± 9.0 | 55.3 ± 9.2 | 56.1 ± 10.7 | 54.5 ± 8.5 |
Data are shown as Means ± SD (continuous variables) or N (%) (categorical variables). All continuous variables were analyzed with ANOVA tests, and all categorical variables were analyzed with chi-square.
Subjects with missing values were excluded from the analysis.
M – methylated; UM – un-methylated.
Comparisons of early life exposure factors between participating cases with and without promoter methylation of specific gene were adjusted for age.
Among parous women.
Comparison of methylated to unmethylated cases, p < 0.05.
Likelihood of promoter methylation in premenopausal breast tumors by early life exposures are shown in Table 2. There was increased likelihood of tumors with E – cadherin methylation among those with lower birth weight (≤2.5 kg) compared to average birth weight (2.6–3.9 kg) (OR = 2.79, 95% CI, 1.15–6.82). Compared to those who had been breastfed, there was increased likelihood of methylation for p16 gene in premenopausal tumors of cases who had not been breastfed (OR = 2.75, 95% CI, 1.14–6.62). There was a reduction of likelihood of p16 methylation among participants reporting maternal age at delivery between 25–30 years compared to those with earlier maternal age at delivery (OR = 0.41, 95% CI, 0.18–0.90). Additionally, adult height was associated with increased likelihood of RAR-β2 methylation, the adjusted OR and 95% CI for the comparison of the highest compared to the lowest tertile was 3.34 (95% CI, 1.19–9.39). There was no association of birth order of the participant in her family, her own age at menarche, age of first live birth, or weight at age 20 with the likelihood of promoter methylation of E –cadherin, p16, or RAR-β2 gene in premenopausal breast tumors.
Table 2.
Association between early life exposures and promoter methylation of E-cadherin, p16, and RAR-β2 genes in breast tumors among premenopausal women: case-case comparisons
| E-cadherin | p16 | RAR-β2 | |||||||
|---|---|---|---|---|---|---|---|---|---|
| M | UM | OR (95% CI) 1 | M | UM | OR (95% CI) 1 | M | UM | OR (95% CI) 1 | |
| Birth weight (kg) | |||||||||
| ≤2.5 | 10 | 18 | 2.79 (1.15–6.82) | 6 | 22 | 0.70 (0.27–1.85) | 7 | 21 | 1.00 (0.39–2.57) |
| 2.6–3.9 | 28 | 135 | 1.0 | 45 | 118 | 1.0 | 42 | 121 | 1.0 |
| >3.9 | 9 | 26 | 1.69 (0.70–4.05) | 8 | 27 | 0.79 (0.33–1.88) | 12 | 23 | 1.61 (0.72–3.60) |
| Birth order | |||||||||
| First born | 14 | 49 | 1.0 | 20 | 43 | 1.0 | 23 | 40 | 1.0 |
| Other than first born | 35 | 134 | 0.83 (0.40–1.69) | 40 | 129 | 0.68 (0.35–1.30) | 40 | 129 | 0.52 (0.27–1.02) |
| Maternal age at delivery (yrs) | |||||||||
| <25 | 15 | 70 | 1.0 | 27 | 58 | 1.0 | 22 | 63 | 1.0 |
| 25–30 | 18 | 53 | 1.85 (0.81–4.20) | 12 | 59 | 0.41 (0.18–0.90) | 18 | 53 | 0.76 (0.35–1.63) |
| >30 | 13 | 54 | 1.25 (0.54–2.89) | 19 | 48 | 0.82 (0.41–1.67) | 20 | 47 | 1.21 (0.58–2.49) |
| Having been breastfed | |||||||||
| Yes | 9 | 43 | 1.0 | 9 | 43 | 1.0 | 12 | 40 | 1.0 |
| No | 34 | 119 | 1.21 (0.50–2.93) | 45 | 108 | 2.75 (1.14–6.62) | 47 | 106 | 1.18 (0.53–2.62) |
| Age at menarche (yrs) | |||||||||
| <12 | 11 | 48 | 0.81 (0.37–1.81) | 18 | 41 | 1.66 (0.81–3.38) | 15 | 44 | 0.83 (0.40–1.75) |
| 12–13 | 27 | 95 | 1.0 | 27 | 95 | 1.0 | 32 | 90 | 1.0 |
| >13 | 12 | 44 | 0.97 (0.44–2.14) | 16 | 40 | 1.42 (0.68–2.96) | 17 | 39 | 1.28 (0.62–2.64) |
| Age at first live birth (yrs) | |||||||||
| <23 | 18 | 51 | 1.0 | 16 | 53 | 1.0 | 20 | 49 | 1.0 |
| 23–29 | 20 | 70 | 0.92 (0.42–2.00) | 23 | 67 | 0.81 (0.41–1.61) | 23 | 67 | 0.63 (0.30–1.33) |
| ≥30 | 5 | 34 | 0.48 (0.15–1.50) | 11 | 28 | 1.28 (0.61–2.66) | 11 | 28 | 0.80 (0.31–2.04) |
| Adult height (m) | |||||||||
| <1.60 | 10 | 40 | 1.02 | 11 | 39 | 1.02 | 7 | 43 | 1.02 |
| 1.60–1.64 | 16 | 62 | 0.94 (0.37–2.37) | 19 | 59 | 1.11 (0.46–2.68) | 23 | 55 | 3.00 (1.13–7.97) |
| ≥1.65 | 24 | 85 | 1.04 (0.40–2.74) | 31 | 78 | 1.39 (0.56–3.46) | 34 | 75 | 3.34 (1.19–9.39) |
| Weight at 20 years (kg) | |||||||||
| <54.4 | 22 | 79 | 1.0 | 23 | 78 | 1.0 | 27 | 74 | 1.0 |
| ≥54.4 | 28 | 108 | 0.97 (0.51–1.85) | 38 | 98 | 1.29 (0.71–2.37) | 37 | 99 | 1.08 (0.59–1.96) |
M – methylated; UM – un-methylated
Odds ratios and 95% confidence intervals adjusted for age, race, education, and ER status.
Odds ratios and 95% confidence intervals adjusted for age, race, education, ER status, weight at 20s, weight 2 years ago, and maternal height.
Table 3 shows results among postmenopausal women. There was greater likelihood of p16 promoter methylation associated with being other than first born in the birth order (OR = 1.78, 95% CI, 1.17–2.70). Age at first live birth between 23–29 years was associated with reduced likelihood of E – cadherin methylation, while age at first live birth ≥ 30 years was not associated with likelihood E – cadherin methylation. There were no association of the likelihood of RAR-β2 methylation with any of the early life exposures; case-case comparisons were generally close to the null.
Table 3.
Association between early life exposures and promoter methylation of E-cadherin, p16, and RAR-β2 genes in breast tumors among postmenopausal women: case-case comparisons
| E-cadherin | p16 | RAR-β2 | |||||||
|---|---|---|---|---|---|---|---|---|---|
| M | UM | OR (95% CI) 1 | M | UM | OR (95% CI) 1 | M | UM | OR (95% CI) 1 | |
| Birth weight (kg) | |||||||||
| ≤2.5 | 11 | 57 | 0.77 (0.38–1.54) | 13 | 55 | 0.66 (0.34–1.26) | 18 | 50 | 1.03 (0.57–1.85) |
| 2.6–3.9 | 82 | 311 | 1.0 | 107 | 286 | 1.0 | 103 | 290 | 1.0 |
| >3.9 | 11 | 48 | 0.86 (0.42–1.73) | 12 | 47 | 0.68 (0.35–1.34) | 17 | 42 | 1.20 (0.65–2.22) |
| Birth order | |||||||||
| First born | 38 | 173 | 1.0 | 41 | 170 | 1.0 | 55 | 156 | 1.0 |
| Other than first born | 73 | 278 | 1.18 (0.76–1.83) | 105 | 246 | 1.78 (1.17–2.70) | 100 | 251 | 1.14 (0.77–1.69) |
| Maternal age at delivery (yrs) | |||||||||
| <25 | 40 | 181 | 1.0 | 49 | 172 | 1.0 | 59 | 162 | 1.0 |
| 25–30 | 30 | 131 | 1.02 (0.60–1.73) | 46 | 115 | 1.37 (0.86–2.20) | 53 | 108 | 1.35 (0.86–2.11) |
| >30 | 34 | 124 | 1.20 (0.72–2.02) | 42 | 116 | 1.26 (0.78–2.04) | 35 | 123 | 0.80 (0.49–1.29) |
| Having been breastfed | |||||||||
| Yes | 53 | 208 | 1.0 | 72 | 189 | 1.0 | 64 | 197 | 1.0 |
| No | 33 | 140 | 1.06 (0.64–1.77) | 39 | 134 | 0.79 (0.49–1.26) | 54 | 119 | 1.30 (0.83–2.04) |
| Age at menarche (yrs) | |||||||||
| <12 | 25 | 103 | 0.94 (0.56–1.58) | 32 | 96 | 0.93 (0.58–1.49) | 33 | 95 | 0.88 (0.55–1.40) |
| 12–13 | 64 | 249 | 1.0 | 85 | 228 | 1.0 | 89 | 224 | 1.0 |
| >13 | 22 | 103 | 0.77 (0.45–1.33) | 30 | 95 | 0.83 (0.51–1.35) | 35 | 90 | 1.00 (0.63–1.60) |
| Age at first live birth (yrs) | |||||||||
| <23 | 49 | 163 | 1.0 | 59 | 153 | 1.0 | 62 | 150 | 1.0 |
| 23–29 | 31 | 169 | 0.56 (0.33–0.95) | 49 | 151 | 0.89 (0.56–1.42) | 58 | 142 | 1.00 (0.64–1.56) |
| ≥30 | 8 | 44 | 0.56 (0.24–1.28) | 8 | 44 | 0.51 (0.22–1.16) | 14 | 38 | 0.90 (0.45–1.81) |
| Adult height (m) | |||||||||
| <1.60 | 40 | 156 | 1.02 | 49 | 147 | 1.02 | 60 | 136 | 1.02 |
| 1.60–1.64 | 32 | 152 | 0.94 (0.55–1.61) | 50 | 134 | 1.18 (0.73–1.91) | 48 | 136 | 0.78 (0.49–1.26) |
| ≥1.65 | 39 | 147 | 1.24 (0.71–2.18) | 48 | 138 | 1.17 (0.70–1.98) | 49 | 137 | 0.71 (0.43–1.17) |
| Weight at 20 years (kg) | |||||||||
| <54.4 | 53 | 219 | 1.0 | 70 | 202 | 1.0 | 76 | 196 | 1.0 |
| ≥54.4 | 58 | 236 | 1.04 (0.68–1.58) | 77 | 217 | 1.08 (0.74–1.58) | 81 | 213 | 0.97 (0.67–1.41) |
M – methylated; UM – un-methylated
Odds ratios and 95% confidence intervals adjusted for age, race, education, and ER status.
Odds ratios and 95% confidence intervals adjusted for age, race, education, ER status, weight at 20s, weight 2 years ago, and maternal height.
In addition, we investigated the associations of those early life exposure factors with the likelihood of promoter methylation in at least one of the three genes in tumors stratified on menopausal status (Table 4). Among premenopausal cases, adult height was associated with increased likelihood of tumors with promoter methylation of at least one gene (OR = 2.88, 95% CI, 1.25–6.61, for the highest vs lowest tertile). Among premenopausal cases, being other than first born was associated with reduced likelihood of promoter methylation of at least one gene (OR = 0.38, 95% CI, 0.19–0.77); while a positive association between birth order and likelihood of promoter methylation in at least one gene (OR= 1.77, 95% CI, 1.23–2.55, for being other than first born compared to first born) was observed among postmenopausal breast cases. Additionally, among postmenopausal breast cases, there was an inverse association between age at first live birth and likelihood of promoter methylation in at least one gene (OR = 0.51, 95% CI, 0.27–0.97, for age at first live birth ≥30 vs <23 years). Other early life exposures were not associated with the likelihood of promoter methylation in at least one of the three genes in tumors both pre- and postmenopausal women.
Table 4.
Case-case comparisons of early life exposure and promoter methylation of at least one gene in breast tumors across menopausal status
| Any vs None | Any vs None | |||||
|---|---|---|---|---|---|---|
| Any one | None | OR (95% CI) 1 | Any one | None | OR (95% CI) 1 | |
| PREMENOPAUSAL | POSTMENOPAUSAL | |||||
| Birth weight (kg) | ||||||
| ≤2.5 (Below average) | 20 | 8 | 1.47 (0.60–3.57) | 39 | 59 | 0.87 (0.40–1.15) |
| 2.6–3.9 (Average) | 104 | 89 | 1.0 | 267 | 126 | 1.0 |
| >3.9 (Above average) | 28 | 7 | 2.31 (0.94–5.66) | 37 | 22 | 0.82 (0.46–1.47) |
| Birth order | ||||||
| First born | 51 | 12 | 1.0 | 124 | 87 | 1.0 |
| Other than first born | 106 | 63 | 0.38 (0.19–0.77) | 250 | 101 | 1.77 (1.23–2.55) |
| Maternal age at delivery (yrs) | ||||||
| <25 | 60 | 25 | 1.0 | 135 | 86 | 1.0 |
| 25–30 | 45 | 26 | 0.67 (0.33–1.36) | 116 | 45 | 1.59 (1.02–2.47) |
| >30 | 45 | 22 | 0.84 (0.42–1.69) | 104 | 54 | 1.21 (0.79–1.87) |
| Having been breastfed | ||||||
| Yes | 30 | 22 | 1.0 | 172 | 89 | 1.0 |
| No | 112 | 41 | 1.87 (0.91–3.83) | 115 | 58 | 1.02 (0.67–1.57) |
| Age at menarche (yrs) | ||||||
| <12 | 40 | 19 | 1.15 (0.59–2.25) | 85 | 43 | 0.95 (0.61–1.47) |
| 12–13 | 78 | 44 | 1.0 | 213 | 100 | 1.0 |
| >13 | 42 | 14 | 1.78 (0.86–3.67) | 79 | 16 | 0.78 (0.50–1.21) |
| Age at first live birth (yrs) | ||||||
| <23 | 47 | 22 | 1.0 | 152 | 60 | 1.0 |
| 23–29 | 60 | 30 | 0.93 (0.46–1.88) | 126 | 74 | 0.68 (0.44–1.05) |
| ≥30 | 26 | 13 | 0.97 (0.40–2.35) | 29 | 23 | 0.51 (0.27–0.97) |
| Adult height (m) | ||||||
| <1.60 | 27 | 23 | 1.02 | 136 | 60 | 1.02 |
| 1.60–1.64 | 51 | 27 | 1.69 (0.79–3.63) | 119 | 65 | 0.84 (0.53–1.32) |
| ≥1.65 | 82 | 27 | 2.88 (1.25–6.61) | 122 | 64 | 0.83 (0.51–1.34) |
| Weight at 20 years (kg) | ||||||
| <54.4 | 65 | 34 | 1.0 | 179 | 93 | 1.0 |
| ≥54.4 | 95 | 43 | 1.17 (0.67–2.05) | 198 | 96 | 1.08 (0.76–1.54) |
Odds ratios and 95% confidence intervals adjusted for age, race, education, and ER status.
Odds ratios and 95% confidence intervals adjusted for age, race, education, ER status, weight at 20s, weight 2 years ago, and maternal height.
Discussion
To our knowledge, this study is the first to assess associations of early life exposures with promoter methylation in breast tumors in a large population-based study. We found positive associations between lower birth weight and E – cadherin promoter methylation, between not having been breastfed and p16 promoter methylation, and between adult height and RAR-β2 methylation in premenopausal breast tumors. Among postmenopausal cases, being other than first born was positively associated with the likelihood of p16 promoter methylation. The likelihood of promoter methylation of the individual genes, E - cadherin, p16, or RAR-β2 did not differ by maternal age at delivery, age at menarche, age of first live birth, and weight at age 20 year in either strata of menopause. While none of the exposures we studied was consistently associated with alterations of all three of the genes studied, our findings suggested changes in methylation for individual genes in association with particular exposures.
Unlike most organs where most development occurs during embryogenesis and the early postnatal period, development of the breast occurs from the in utero period, to infancy through puberty (2); breast tissue is not fully differentiated until after the first full-term pregnancy (2, 3). Therefore, prenatal and early life exposures might influence susceptibility to breast cancer in adulthood. Most previous studies found a positive association between higher than average birth weight and risk of premenopausal breast cancer (4–6, 9). However, results from several studies also suggested increased breast cancer risk associated with lower than average birth weight (<2.5 kg or 5.5 pounds), particularly in young women (4, 23, 24). Birth weight, a proxy for intrauterine environment, has been linked with maternal hormones and other maternal exposures, including nutrition, alcohol and tobacco smoking.
There is evidence that early life exposures may impact methylation in tissues. Recent studies reported a significant correlation between global LINE-1 methylation in cord blood and low birth weight (25). Maternal smoking during pregnancy is strongly associated with increased risk of low birth weight (26), and has been found to be inversely associated with genomic DNA demethylation in adulthood (27). Results from the Children’s Health Study showed that children exposed to prenatal tobacco smoke had significantly increased promoter methylation of AXL and PTPRO genes (28). Poor maternal nutrition or reduced intake of micronutrients during gestation could also contribute to low birth weight (20, 29); animal studies have shown histone modification and aberrant DNA methylation of different genes in rats exposed to intrauterine under-nutrition (30). In the current study, we observed an association between lower birth weight and increased likelihood of E – cadherin promoter methylation in premenopausal breast tumors. These data support the hypothesis that prenatal environment/ development may lead to aberrant DNA methylation, and further influence adult disease risk.
We found that not having been breastfed was significantly associated with increased likelihood of p16 promoter methylation in premenopausal breast tumors. Some studies have found an inverse association between being breastfed in infancy and premenopausal breast cancer risk (7, 10, 31), while others have not (32–34). A protective effect associated with breastfeeding in infancy may be partly related to exposure to growth factors, enzymes and hormones in breast milk (35). Early introduction of cow milk proteins (complementary feeding) has been proposed to induce intestinal mucosa inflammation and increase gut permeability (36). Inflammation has been found to alter the DNA promoter methylation pattern and critical gene regulation in several studies (37, 38). The mechanism by which early life exposure to breast milk and/ or complementary feeding may influence breast cancer risk in adulthood is still unclear. Further studies, including animal studies, are needed to examine the association of not being breastfed in infancy with aberrant DNA methylation.
Several previous studies found inverse associations of birth order and breast cancer risk, either overall or among premenopausal women (4, 39). Hormonal levels varying according to parity may explain the potential associations, estrogen levels appear to be lower during successive pregnancies compared to first pregnancies (4). There is growing evidence that higher estrogen or xenoestrogen exposure during early life could induce aberrant DNA methylation (13, 40). In the current study, we found reduced likelihood of promoter methylation of at least one gene with being other than first born among premenopausal women, possibly related to this difference in estrogen exposure. However, the possible explanation for different finding for postmenopausal women is not known.
Adult height is due in part to childhood nutrition and health (41). Childhood malnutrition can influence DNA methylation, impacting nutrients involved in one-carbon metabolism (2). Hughes et al (42) found that individuals exposed to famine during early life (adolescence and early adulthood) had decreased risk of developing a CpG island methylator phenotype (CIMP) colorectal tumor later in life, and the degree of hypermethylation was inversely associated with early life exposure to energy restriction. Height is also influenced by genetic factors (43). We found that greater adult height was associated with increased likelihood of RAR-β2 promoter methylation in premenopausal tumors; findings did not appreciably change with adjustment for maternal height (data not shown).
As for any study of this kind, our results need to be considered in the context of the strengths and weaknesses of the data. The strengths of our study include the population-based study design, a relatively large sample of archived tumor tissues, measurement of anthropometric factors by trained interviewers, and detailed information on possible confounders and disease characteristics. However, several limitations should be considered. While the number of tumor blocks available for our study was large, the statistical power in some subgroups of our study remained limited, reducing our ability to identify weak associations. Another concern was that our inability to obtain paraffin-embedded tumor tissues for all breast cancer cases may have led to a selection bias; however, cases with and without breast tumor tissue had similar distributions of early life exposure factors. Additionally, there were similar distributions of early life exposure factors between cases with tumor tissue but no methylation results and cases with tumor tissue and methylation results. There may have been errors in self-reported early life exposures. There is evidence that there is not much bias in recall of reproductive history in case-control studies of breast cancer (44). In our study, participant’s birth certificates were also collected for 80 women as a validation of self-reported birth weight; reports by the study participants were well correlated with birth certificate data (r = 0.62) (10). Furthermore, our previous results on associations of breast cancer risk with perinatal exposures have been consistent with other studies (1, 10). It is unlikely that recall of early life exposures would be related to gene promoter methylation; these case-case comparisons may be affected by misclassification but are not likely affected by biased recall. There may be confounding of other unmeasured factors. In particularly, we did not have information on maternal or paternal education. Further studies in other populations are needed to confirm our findings and to elucidate the underlying biological mechanisms.
There are some concerns with limitations of the measurement of methylation. We examined methylation for three genes that are known to be frequently methylated in breast tumors and are known to be significant in three important pathways in breast carcinogenesis. However, we clearly are limited by the study of a small number of genes. Expansion of our findings to a large number of genes or imprinted genes will be important. Furthermore, the methodology used in our study was limited to examination of a single CpG island in the promoter regions. It is assumed that these single regions are sentinels for gene silencing and methylation of other CpG islands, especially in tumors; but it is possible that in some women, these genes are hypermethylated in CpG sequences that we did not study. Finally, we used real-time methylation-specific PCR to assess promoter methylation that can increase the specificity of the PCR by interrogating more than one CpG. More specifically, we used a fluorescence-based version of the methylation-specific PCR technique because of its increased throughput by eliminating the need for gel electrophoresis (45). This method has been found to be 10 times more sensitive than the classic methylation-specific PCR method and able to detect methylated sequences from an excess of 10,000-fold unmethylated alleles (45). Although we are aware of the limitation of this technique given its qualitative nature compared with other quantitative methods, such as pyrosequencing and the fact that it interrogates a limited number of CpG sites, we have followed stringent quality control criteria to ensure confidence in results. Moreover, recent findings showed that results from methylation-specific PCR are highly correlated with other quantitative methods. By using the highly specific real time methylation-specific PCR, it is likely that our results would be reproduced by other methods (46).
Results from our study support the hypothesis that perinatal and early life exposures may impact DNA methylation thereby influencing development of breast cancer in later life. Replication in other populations and the exploration of methylation of a larger number of sites are necessary to better understand this finding of potential significance in understanding the mechanism for exposures in early life and breast cancer risk.
Acknowledgement
This study would not have been possible without the support of all the study participants and the research staff of the WEB Study. This work was supported in part by the National Institute on Alcohol Abuse and Alcoholism (P50-AA09802), the Department of Defense (DAMD 179616202, DAMD 17030446), and the National Cancer Institute (R01CA 092040).
Footnotes
Conflicts of interest: All authors have no conflicts of interests.
Statement of Interest: None
References
- 1.Okasha M, McCarron P, Gunnell D, Smith GD. Exposures in childhood, adolescence and early adulthood and breast cancer risk: a systematic review of the literature. Breast Cancer Res Treat. 2003;78:223–276. doi: 10.1023/a:1022988918755. [DOI] [PubMed] [Google Scholar]
- 2.Burdge GC, Lillycrop KA, Jackson AA. Nutrition in early life, and risk of cancer and metabolic disease: alternative endings in an epigenetic tale? Brit J Nutr. 2009;101:619–630. doi: 10.1017/S0007114508145883. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Russo J, Russo IH. Development of the human breast. Maturitas. 2004;49:2–15. doi: 10.1016/j.maturitas.2004.04.011. [DOI] [PubMed] [Google Scholar]
- 4.Park SK, Kang D, McGlynn KA, et al. Intrauterine environments and breast cancer risk: meta-analysis and systematic review. Breast Cancer Res. 2008;10:R8. doi: 10.1186/bcr1850. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Xu X, Dailey AB, Peoples-Sheps M, Talbott EO, Li N, Roth J. Birth weight as a risk factor for breast cancer: a meta-analysis of 18 epidemiological studies. J Womens Health. 2009;18:1169–1178. doi: 10.1089/jwh.2008.1034. [DOI] [PubMed] [Google Scholar]
- 6.Xue F, Michels KB. Intrauterine factors and risk of breast cancer: a systematic review and meta-analysis of current evidence. Lancet Oncol. 2007;8:1088–1100. doi: 10.1016/S1470-2045(07)70377-7. [DOI] [PubMed] [Google Scholar]
- 7.Martin RM, Middleton N, Gunnell D, Owen CG, Smith GD. Breast-feeding and cancer: The Boyd Orr Cohort and a systematic review with meta-analysis. J Natl Cancer Inst. 2005;97:1446–1457. doi: 10.1093/jnci/dji291. [DOI] [PubMed] [Google Scholar]
- 8.Forman MR, Cantwell MM, Ronckers C, Zhang Y. Through the looking glass at early-life exposures and breast cancer risk. Cancer Invest. 2005;23:609–624. doi: 10.1080/07357900500283093. [DOI] [PubMed] [Google Scholar]
- 9.Ruder EH, Dorgan JF, Kranz S, Kris-Etherton PM, Hartman TJ. Examining breast cancer growth and lifestyle risk factors: early life, childhood, and adolescence. Clinical Breast Cancer. 2008;8:334–342. doi: 10.3816/CBC.2008.n.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Barba M, McCann SE, Nie J, et al. Perinatal exposures and breast cancer risk in the Western New York Exposures and Breast Cancer (WEB) Study. Cancer Causes Control. 2006;17:395–401. doi: 10.1007/s10552-005-0481-5. [DOI] [PubMed] [Google Scholar]
- 11.Green J, Cairns BJ, Casabonne D, Wright FL, Reeves G, Beral V. for the Million Women Study collaborators. Height and cancer incidence in the Million Women Study: prospective cohort, and meta-analysis of prospective studies of height and total cancer risk. Lancet Oncol. 2011;12:785–794. doi: 10.1016/S1470-2045(11)70154-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Waterland RA. Is epigenetics an important link between early life events and adult disease? Horm Res. 2009;(Suppl 1):13–16. doi: 10.1159/000178030. [DOI] [PubMed] [Google Scholar]
- 13.Bernal AJ, Jirtle RL. Epigenomic disruption: the effects of early developmental exposures. Birth Defects Research (Part A) 2010;88:938–944. doi: 10.1002/bdra.20685. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Szyf M, Pakneshan P, Rabbani SA. DNA methylation and breast cancer. Biochem Parmacol. 2004;68:1187–1197. doi: 10.1016/j.bcp.2004.04.030. [DOI] [PubMed] [Google Scholar]
- 15.Esteller M. Epigenetics in cancer. N Engl J Med. 2008;358:1148–1159. doi: 10.1056/NEJMra072067. [DOI] [PubMed] [Google Scholar]
- 16.Herman JG, Merlo A, Mao L, et al. Inactivation of the CDKN2/p16/MTS1 gene is frequently associated with aberrant DNA methylation in all common human cancers. Cancer Res. 1995;55:4525–4530. [PubMed] [Google Scholar]
- 17.Arapshian A, Kuppumbatti YS, Mira-y-Lopez R. Methylation of conserved CpG sites neighboring the beta retinoic acid response element may mediate retinoic acid receptor beta gene silencing in MCF-7 breast cancer cells. Oncogene. 2000;19:4066–4070. doi: 10.1038/sj.onc.1203734. [DOI] [PubMed] [Google Scholar]
- 18.Caldeira JR, Prando EC, Quevedo FC, Neto FA, Rainho CA, Rogatto SR. CDH1 promoter hypermethylation and E-cadherin protein expression in infiltrating breast cancer. BMC Cancer. 2006;6:48. doi: 10.1186/1471-2407-6-48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Desaulniers D, Xiao GH, Lian H, et al. Effects of mixtures of polychlorinated biphenyls, methylmercury, and organochlorine pesticides on hepatic DNA methylation in prepubertal female Sprague-Dawley rats. Int J Toxicol. 2009;28:294–307. doi: 10.1177/1091581809337918. [DOI] [PubMed] [Google Scholar]
- 20.Heijmans BT, Tobi EW, Stein AD, et al. Persistent epigenetic differences associated with prenatal exposure to famine in humans. Proc Natl Acad Sci U S A. 2008;105:17046–17049. doi: 10.1073/pnas.0806560105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Tao MH, Shields PG, Nie JMA, et al. DNA hypermethylation and clinicopathological features in breast cancer: the Western New York Exposures and Breast Cancer (WEB) Study. Breast Cancer Res Treat. 2009;114:559–568. doi: 10.1007/s10549-008-0028-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Tao MH, Marian C, Nie J, et al. Body mass and DNA promoter methylation in breast tumors in the Western New York Exposures and Breast Cancer Study. Am J Clin Nutr. 2011;94:831–838. doi: 10.3945/ajcn.110.009365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Mellemkjaer L, Olsen ML, Sørensen HT, Thulstrup AM, Olsen J, Olsen JH. Birth weight and risk of early-onset breast cancer (Denmark) Cancer Causes Control. 2003;14:61–64. doi: 10.1023/a:1022570305704. [DOI] [PubMed] [Google Scholar]
- 24.Innes K, Byers T, Schymura M. Birth characteristics and subsequent risk for breast cancer in very young women. Am J Epidemiol. 2000;152:1121–1128. doi: 10.1093/aje/152.12.1121. [DOI] [PubMed] [Google Scholar]
- 25.Fryer AA, Emes RD, Ismail KM, et al. Quantitative, high-resolution epigenetic profiling of CpG loci identifies associations with cord blood plasma homocysteine and birth weight in humans. Epigenetics. 2011;6:86–94. doi: 10.4161/epi.6.1.13392. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.La Merrill M, Stein CR, Landrigan P, Engel SM, Savitz DA. Prepregnancy body mass index, smoking during pregnancy, and infant birth weight. Ann Epidemiol. 2011 Mar 19; doi: 10.1016/j.annepidem.2010.11.012. [Epub ahead of print] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Flom J, Ferris J, Gonzalez K, Santella R, Terry M. Prenatal tobacco smoke exposure and genome wide methylation in adulthood. Cancer Epidemiol Biomarkers Prev. 2011;20:2518–2523. doi: 10.1158/1055-9965.EPI-11-0553. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Breton CV, Byun HM, Wenten M, Pan F, Yang A, Gilliland FD. Prenatal tobacco smoke exposure affects global and gene-specific DNA methylation. Am J Respir Crit Care Med. 2009;180:462–447. doi: 10.1164/rccm.200901-0135OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Shah PS, Ohlsson A. Knowledge Synthesis Group on Determinants of Low Birth Weight and Preterm Births. Effects of prenatal multimicronutrient supplementation on pregnancy outcomes: a meta-analysis. CMAJ. 2009;180:E99–E108. doi: 10.1503/cmaj.081777. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Attig L, Gabory A, Junien C. Early nutrition and epigenetic programming: chasing shadows. Curr Opin Clin Nutr Metab Care. 2010;13:284–293. doi: 10.1097/MCO.0b013e328338aa61. [DOI] [PubMed] [Google Scholar]
- 31.Freudenheim JL, Marshall JR, Graham S, et al. Exposure to breastmilk in infancy and the risk of breast cancer. Epidemiology. 1995;5:324–331. doi: 10.1097/00001648-199405000-00011. [DOI] [PubMed] [Google Scholar]
- 32.Wise LA, Titus-Ernstoff L, Newcomb PA, et al. Exposure to breast milk in infancy and risk of breast cancer. Cancer Causes Control. 2009;20:1083–1089. doi: 10.1007/s10552-009-9332-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Michels KB, Trichopoulos D, Rosner BA, et al. Being breastfed in infancy and breast cancer incidence in adult life: results from the two nurses’ health studies. Am J Epidemiol. 2001;153:275–283. doi: 10.1093/aje/153.3.275. [DOI] [PubMed] [Google Scholar]
- 34.Titus-Ernstoff L, Egan KM, Newcomb PA, et al. Exposure to breast milk in infancy and adult breast cancer risk. J Natl Cancer Inst. 1998;90:921–924. doi: 10.1093/jnci/90.12.921. [DOI] [PubMed] [Google Scholar]
- 35.Rodriguez-Palmero M, Koletzko B, Kunz C, Jensen R. Nutritional and biochemical properties of human milk: II. Lipids, micronutrients, and bioactive factors. Clin Perinatol. 1999;26:335–359. [PubMed] [Google Scholar]
- 36.Knip M, Virtanen SM, Akerblom HK. Infant feeding and the risk of type 1 diabetes. Am J Clin Nutr. 2010;91:1506S–1513S. doi: 10.3945/ajcn.2010.28701C. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Kang GH, Lee HJ, Hwang KS, Lee S, Kim JH, Kim JS. Aberrant CpG island hypermethylation of chronic gastritis, in relation to aging, gender, intestinal metaplasia, and chronic inflammation. Am J Pathol. 2003;163:1551–1556. doi: 10.1016/S0002-9440(10)63511-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Issa JP, Ahuja N, Toyota M, Bronner MP, Brentnall TA. Accelerated age-related CpG island methylation in ulcerative colitis. Cancer Res. 2001;61:3573–3577. [PubMed] [Google Scholar]
- 39.Hsieh CC, Tzonou A, Trichopoulos D. Birth order and breast cancer risk. Cancer Causes Control. 1991;2:95–98. doi: 10.1007/BF00053127. [DOI] [PubMed] [Google Scholar]
- 40.Fernandez SV, Russo J. Estrogen and xenoestrogens in breast cancer. Toxicol Pathol. 2010;38:110–122. doi: 10.1177/0192623309354108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Victora CG, Adair L, Fall C, et al. Maternal and Child Undernutrition Study Group. Maternal and child undernutrition: consequences for adult health and human capital. Lancet. 2008;371:340–357. doi: 10.1016/S0140-6736(07)61692-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Hughes LA, van den Brandt PA, de Bruïne AP, et al. Early life exposure to famine and colorectal cancer risk: a role for epigenetic mechanisms. PLoS One. 2009;4:e7951. doi: 10.1371/journal.pone.0007951. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Li L, Manor O, Power C. Early environment and child-to-adult growth trajectories in the 1958 British birth cohort. Am J Clin Nutr. 2004;80:185–192. doi: 10.1093/ajcn/80.1.185. [DOI] [PubMed] [Google Scholar]
- 44.Slanger T, Mutschelknauss E, Kropp S, Braendle W, Flesch-Janys D, Chang-Claude J. Test-retest reliability of self-reported reproductive and lifestyle data in the context of a German case-control study on breast cancer and postmenopausal hormone therapy. Ann Epidemiol. 2007;17:993–998. doi: 10.1016/j.annepidem.2007.07.094. [DOI] [PubMed] [Google Scholar]
- 45.Eads CA, Danenberg KD, Kawakami K, et al. MethyLight: a high-throughput assay to measure DNA methylation. Nucleic Acids Res. 2000;28:E32. doi: 10.1093/nar/28.8.e32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Lee ES, Issa JP, Roberts DB, et al. Quantitative promoter hypermethylation analysis of cancer-related genes in salivary gland carcinomas: comparison with methylation-specific PCR technique and clinical significance. Clin Cancer Res. 2008;14(9):2664–2672. doi: 10.1158/1078-0432.CCR-07-1232. [DOI] [PubMed] [Google Scholar]