DNA methylation age is elevated in breast tissue of healthy women

Abstract

Background: Limited evidence suggests that female breast tissue ages faster than other parts of the body according to an epigenetic biomarker of aging known as the "epigenetic clock." However, it is unknown whether breast tissue samples from healthy women show a similar accelerated aging effect relative to other tissues, and what could drive this acceleration. The goal of this study is to validate our initial finding of advanced DNA methylation (DNAm) age in breast tissue, by directly comparing it to that of peripheral blood tissue from the same individuals, and to do a preliminary assessment of hormonal factors that could explain the difference.

Methods: We utilized n = 80 breast and 80 matching blood tissue samples collected from 40 healthy female participants of the Susan G. Komen Tissue Bank at the Indiana University Simon Cancer Center who donated these samples at two time points spaced at least a year apart. DNA methylation levels (Illumina 450K platform) were used to estimate the DNAm age.

Results: DNAm age was highly correlated with chronological age in both peripheral blood (r = 0.94, p < 0.0001) and breast tissues (r = 0.86, p < 0.0001). A measure of epigenetic age acceleration (age-adjusted DNAm Age) was substantially increased in breast relative to peripheral blood tissue (p = 1.6 × 10^-11). The difference between DNAm age of breast and blood decreased with advancing chronologic age (r = -0.53, p = 4.4 × 10^-4).

Conclusions: Our data clearly demonstrate that female breast tissue has a higher epigenetic age than blood collected from the same subject. We also observe that the degree of elevation in breast diminishes with advancing age. Future larger studies will be needed to examine associations between epigenetic age acceleration and cumulative hormone exposure.

Keywords: Biomarker of aging; Breast cancer; Cell cycling; DNA methylation; Epigenetics; Estrogen; Tissue aging.

Fig. 1

Epigenetic clock analysis of breast tissue and peripheral blood. a Scatter plot of DNAm age (y-axis) versus chronological age (x-axis) for breast tissue (red) and peripheral blood (black) samples at baseline visit, correlation coefficient = 0.8, p value = 5.5 × 10⁻¹⁹. Regression lines demonstrate the linear relationship between DNAm age and chronologic age in all samples (a), in breast (b, ρ = 0.86, p value = 1.2 × 10⁻¹²) and in blood (c, ρ = 0.94, p value = 2.4 × 10⁻¹⁹) separately. d This bar plot shows the relationship between epigenetic age acceleration and tissue type, using data at baseline visit, demonstrating the mean value (y-axis) and one standard error, with p value = 1.6 × 10⁻¹¹ from the results from a non-parametric group comparison test (Kruskal–Wallis)

Fig. 2

Tissue difference in DNAm age between breast and blood. a Frequency distribution of the absolute difference DNAm age breast and DNAm age blood at first visit. On average, DNAm age of breast is 11.4 years older than that of blood (SD = 7.1, range −7.9 to 23.5 years). b While DNAm age of breast is significantly higher than blood at younger ages, the gap closes with advancing age. This scatter plot shows a strong inverse correlation between chronologic age and the absolute difference DNAm age breast and DNAm age blood (correlation coefficient = −0.53, p = 4.4 × 10⁻⁴), c The absolute difference between DNAm age breast and DNAm age blood is higher in pre-menopausal women (mean 13.9, SD 6.3 years) compared with post-menopausal women (mean 8.3, SD 6.9 years, with p value = 0.0098 from non-parametric group comparison Kruskal–Wallis testing). d The rate of change in DNAm age is higher in blood than breast tissue. This bar plot shows the relationship between rate of change in DNAm age and tissue type, demonstrating the mean value (y-axis) and one standard error. While the DNAm age in breast starts out higher, DNAm age in blood eventually catches up to breast tissue, with results from a non-parametric group comparison test reveal a significantly faster rise in DNAm age in blood than breast over time, with p value = 0.038 (Kruskal–Wallis)

Fig. 3

Individual trajectories of DNAm age with advancing chronologic age. This plot shows the longitudinal changes in DNAm age of blood (left panel) and breast (right panel) for each individual at visits 1 and 2

Fig. 4

DNA methylation levels inside and outside of CpG islands. a Effect of chronological age on CpG methylation versus chromosomal location in a blood and b breast tissue. The y-axis of the bar plot shows the log (base 10)-transformed correlation test p value (whose sign reflects the sign of the correlation). c, d Mean methylation levels in CpG islands versus chronological age in c blood and d breast tissue. e, f Mean methylation levels outside of CpG islands versus chronological age in e blood and f breast tissue. g Mean methylation in CpG islands versus tissue type. h Mean methylation outside of CpG islands versus tissue type