Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2016 Sep 8;109(1):djw188.
doi: 10.1093/jnci/djw188. Print 2017 Jan.

Effects of In Utero Exposure to Ethinyl Estradiol on Tamoxifen Resistance and Breast Cancer Recurrence in a Preclinical Model

Leena Hilakivi-Clarke  1 , Anni Wärri  1   2 , Kerrie B Bouker  1 , Xiyuan Zhang  1 , Katherine L Cook  1   3 , Lu Jin  1 , Alan Zwart  1 , Nguyen Nguyen  1 , Rong Hu  1 , M Idalia Cruz  1 , Sonia de Assis  1 , Xiao Wang  4 , Jason Xuan  4 , Yue Wang  4 , Bryan Wehrenberg  5 , Robert Clarke  1
Affiliations

Effects of In Utero Exposure to Ethinyl Estradiol on Tamoxifen Resistance and Breast Cancer Recurrence in a Preclinical Model

Leena Hilakivi-Clarke et al. J Natl Cancer Inst. .

Abstract

Background: Responses to endocrine therapies vary among patients with estrogen receptor (ER+) breast cancer. We studied whether in utero exposure to endocrine-disrupting compounds might explain these variations.

Methods: We describe a novel ER+ breast cancer model to study de novo and acquired tamoxifen (TAM) resistance. Pregnant Sprague Dawley rats were exposed to 0 or 0.1 ppm ethinyl estradiol (EE2), and the response of 9,12-dimethylbenz[a]anthracene (DMBA)-induced mammary tumors to 15 mg/kg TAM, with (n = 17 tumors in the controls and n = 20 tumors in EE2 offspring) or without 1.2 g/kg valproic acid and 5 mg/kg hydralazine (n = 24 tumors in the controls and n = 32 tumors in EE2 offspring) in the female offspring, was assessed. One-sided Chi2 tests were used to calculate P values. Comparisons of differentially expressed genes between mammary tumors in in utero EE2-exposed and control rats, and between anti-estrogen-resistant LCC9 and -sensitive LCC1 human breast cancer cells, were also performed.

Results: In our preclinical model, 54.2% of mammary tumors in the control rats exhibited a complete response to TAM, of which 23.1% acquired resistance with continued anti-estrogen treatment and recurred. Mammary tumors in the EE2 offspring were statistically significantly less likely to respond to TAM (P = .047) and recur (P = .007). In the EE2 offspring, but not in controls, adding valproic acid and hydralazine to TAM prevented recurrence (P < .001). Three downregulated and hypermethylated genes (KLF4, LGALS3, MICB) and one upregulated gene (ETV4) were identified in EE2 tumors and LCC9 breast cancer cells, and valproic acid and hydralazine normalized the altered expression of all four genes.

Conclusions: Resistance to TAM may be preprogrammed by in utero exposure to high estrogen levels and mediated through reversible epigenetic alterations in genes associated with epithelial-mesenchymal transition and tumor immune responses.

PubMed Disclaimer

Figures

Figure 1.
Figure 1.
Response of mammary tumors to tamoxifen. A) 9,12-dimethylbenz[a]anthracene (DMBA)-induced mammary tumor growth in rats that were exposed to 0.1 ppm EE2 or vehicle control in utero. Rats either received no treatment during tumor monitoring, or when a tumor reached a size of 13 mm in diameter they received 15 mg/kg/day TAM in diet, or 1.2 g/kg/day valproic acid and 5 mg/kg/day hydralazine in drinking water.P values indicate statistically significant differences between the C and EE2 groups; one-sided Chi2-test. NS = not statistically significant. The number of tumors studied in each group varied from 24 to 59. B) Effect of adding VA+H as a second line of treatment on the development of new de novo and acquired TAM-resistant tumors in the control and in utero EE2-exposed rats is shown. P values indicate statistically significant differences between the C and EE2 groups; one-sided Chi2-test. The number of tumors was 17 in the control group and 20 in the EE2 group. C) Panel shows the expression of ERα protein, assessed by western blot, in the DMBA-induced mammary adenocarcinomas of control and in utero EE2-exposed rats that either did not receive any treatment during tumor monitoring or received TAM or TAM and VA+H. The number of tumors studied in each group was three to four. Means (SDs) are shown. C = control; EE2 = ethinyl estradiol; ER = estrogen receptor; H = hydralazine; NS = not statistically significant; TAM = tamoxifen; VA = valproic acid.
Figure 2.
Figure 2.
Differences in gene signaling in mammary tumors between control and in utero EE2–exposed rats. A) Heat map showing differences in the expression of the 48 genes in common between the LCC1 (n = 3) vs LCC9 cells (n = 3) and between mammary tumors in the control (n = 4) vs EE2 (n = 4)-exposed offspring. Genes associated with poor response to tamoxifen in at least one of three GEO datasets (Zhang et al. [GSE12093] ], Sotiriou et al. [GSE2990] ], and Loi et al. [GSE6532] [32–34]) and with differential expression verified by quantitative polymerase chain reaction are listed at the side of the heat maps. B) Results of differential dependency network analysis to identify unique connections among differentially expressed genes in LCC9 cells/tumors in EE2 rats vs in LCC1 cells/mammary tumors in control rats. Connections shown in green exist only in LCC9 cells/mammary tumors in EE2 rats, while connections shown in red exist only in LCC1 cells/tumors in control rats. Circles in orange indicate upregulation of gene expression, and circles in blue indicate downregulation in LCC9 cells/mammary tumors in EE2 rats. Nodes are circled in orange. DDN = differential dependency network; EE2 = ethinyl estradiol.
Figure 3.
Figure 3.
Changes in the expression of ETV4, KLF4, LGALS3, and MICB by valproic acid and hydralazine and their DNA methylation patterns. A) Reversal of upregulation of ETV4, and downregulation of KLF4, LGALS4, and MICB expression in LCC9 cells treated with vehicle (V) or 4-hydroxy-tamoxifen, by 1 mM valproic acid and 5 μM hydralazine (n = 5 per group). P values indicate statistically significant differences (two-sided two-way analysis of variance) in gene expression between LCC1 and LCC9 cells, and between LCC9 cells, with or without VA+H. Differences between V and TAM: *P < .05; †P < .001 using the two-sided Tukey test. Means (SDs) are shown. B) Differences are shown in the methylation patterns in promoter regions and first exon of ETV4, KLF4, LGALS4, and MICB genes between LCC1 cells compared with LCC9 cells (n = 5 per group). Bars represent CpG islands; those marked with * are different between LCC1 and LCC9 cells (P < .05, two-sided Student’s t test). H = hydralazine; TAM = tamoxifen; VA = valproic acid.
Figure 3.
Figure 3.
B) continued.
Figure 4.
Figure 4.
Effect of valproic acid and hydralazine on tamoxifen –resistant human breast cancer cells. Response of anti-estrogen-resistant LCC9 human breast cancer cells to 4-hydroxy-tamoxifen (10–1,000 nM) alone or TAM plus 1 mM valproic acid with and without 5 μM hydralazine. Bars marked * (P < .05), † (P < .01), or ‡ (P < .001) are statistically significantly different from no valproic acid or hydralazine control (two-sided one-way analysis of variance, followed by two-sided Tukey test). P values also are given for statistically significant differences between valproic acid– or valproic acid and hydralazine–treated cells. Data were generated using three replicates of each exposure. Means (SDs) are shown. TAM = tamoxifen.
Figure 5.
Figure 5.
Effect of silencing by siRNA of A) ETV4 in LCC9 cells and B) LGALS3 in LCC1 cells on tamoxifen resistance and correlation of gene expression with response to TAM or letrozole. Both LCC9 and LCC1 cells were treated with different doses of 4-hydroxy-tamoxifen or vehicle for seven days, and then cells were stained with 0.5% crystal violet. Experiments were repeated five times (n = 5 in each group). Two-sided two-way analysis of variance values between control siRNA and LGALS3 or ETV4 siRNA are shown. Means (SDs) are provided. C) Association of ETV4, LGALS3 (galectin-3), and KLF4 expression with recurrence-free survival in estrogen receptor–positive (ER+) breast cancer. Data obtained using tumors from ER+ breast cancers in patients treated with TAM in combined GEO datasets: Zhang et al. (GSE12093) (31), Sotiriou et al. (GSE2990) (30), and Loi et al. (GSE6532) (32–34). D) Association of KLF4 expression with response to 90 days’ neoadjuvant treatment (letrozole 2.5 mg/day, oral) (GSE20181) (36,37). Bar marked with * is statistically significantly different using the two-sided Tukey test from the baseline of women responding to treatment: P < .05. Means (SDs) are shown. CI = confidence interval; HR = hazard ratio; NR = no response to treatment; R = responding to treatment; TAM = tamoxifen.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Palmer JR, Wise LA, Hatch EE, et al. Prenatal diethylstilbestrol exposure and risk of breast cancer. Cancer Epidemiol Biomarkers Prev. 2006;15(8):1509–1514. - PubMed
    1. Hoover RN, Hyer M, Pfeiffer RM, et al. Adverse health outcomes in women exposed in utero to diethylstilbestrol. N Engl J Med. 2011;365(14):1304–1314. - PubMed
    1. Hilakivi-Clarke L, Clarke R, Onojafe I, Raygada M, Cho E, Lippman ME. A maternal diet high in n-6 polyunsaturated fats alters mammary gland development, puberty onset, and breast cancer risk among female rat offspring. Proc Natl Acad Sci U S A. 1997;94:9372–9377. - PMC - PubMed
    1. Hilakivi-Clarke L. Maternal exposure to diethylstilbestrol during pregnancy and increased breast cancer risk in daughters. Breast Cancer Res. 2014;16(2):208. - PMC - PubMed
    1. de Assis S, Warri A, Cruz MI, et al. High-fat or ethinyl-oestradiol intake during pregnancy increases mammary cancer risk in several generations of offspring. Nat Commun. 2012;3:1053. - PMC - PubMed

MeSH terms