Metformin represses self-renewal of the human breast carcinoma stem cells via inhibition of estrogen receptor-mediated OCT4 expression

Abstract

Metformin, a Type II diabetic treatment drug, which inhibits transcription of gluconeogenesis genes, has recently been shown to lower the risk of some diabetes-related tumors, including breast cancer. Recently, "cancer stem cells" have been demonstrated to sustain the growth of tumors and are resistant to therapy. To test the hypothesis that metformin might be reducing the risk to breast cancers, the human breast carcinoma cell line, MCF-7, grown in 3-dimensional mammospheres which represent human breast cancer stem cell population, were treated with various known and suspected breast cancer chemicals with and without non-cytotoxic concentrations of metformin. Using OCT4 expression as a marker for the cancer stem cells, the number and size were measured in these cells. Results demonstrated that TCDD (100 nM) and bisphenol A (10 µM) increased the number and size of the mammospheres, as did estrogen (10 nM E2). By monitoring a cancer stem cell marker, OCT4, the stimulation by these chemicals was correlated with the increased expression of OCT4. On the other hand, metformin at 1 and 10 mM concentration dramatically reduced the size and number of mammospheres. Results also demonstrated the metformin reduced the expression of OCT4 in E2 & TCDD mammospheres but not in the bisphenol A mammospheres, suggesting different mechanisms of action of the bisphenol A on human breast carcinoma cells. In addition, these results support the use of 3-dimensional human breast cancer stem cells as a means to screen for potential human breast tumor promoters and breast chemopreventive and chemotherapeutic agents.

Figure 1. ER positive (A–D and F–H) and negative (E) human breast cells in phenol red-contained (A–E) or phenol red-free MEBM (F–H), expression level of OCT4 mRNA in passaged MCF-7 mammospheres (I), and several ER+ breast cancer mammospheres cultured in MEBM with or without phenol red (J).

; (A) MCF-7; (B), (F) M13SV1; (C), (G) M13SV1 R2; (D), (H) M13SV1 R2N1; (E) MDA-MB-231. The magnification was X 200. Scale bar represents 50 µm in length.

Figure 2. Effect of E2 on MCF-7 mammospheres.

(A), (B) Mammosphere formation was increased by 10 nM E2 treatment. Data were presented as the number of mammospheres per 1,000 seeded cells at 5d (mean ± SD., n = 3). The magnification was X 200. Scale bar represents 10 µm in length. *, P<0.05; ***, P<0.001. (C) 10 nM and 20 nM E2 induced OCT4 expression dramatically in RT-PCR.

Figure 3. Direct pathways of E2 on MCF-7 mammosphere proliferation.

(A) 100 nM ICI 182,780 suppressed the effect of E2 on increment of the size and number of mammospheres. The magnification was X 200. Scale bar represents 10 µm in length. (B) Number of mammospheres was decreased by co-treatment with ICI 182,780. Data were presented as the number of mammospheres per 1,000 seeded cells at 5d (mean ± S.D., n = 3). **, P<0.01; ***, P<0.001. (C) 100 nM ICI 182,780 could repress the OCT4 induction by 10 nM E2. (D) Immunocytochemical detection of OCT4 in MCF-7 mammospheres, compared to the non-treated control group (Control), 10 nM E2-treated (E2), and 100 nM ICI 182,780 with 10 nM E2 (E2 + ICI). The magnification was X 200. Scale bar represents 10 µm in length.

Figure 4. Indirect pathways of E2 on MCF-7 mammosphere proliferation.

(A) ROS detection with DCF-DA. MCF-7 mammospheres treated with 10 nM E2 (left), 100 nM E2 (center), or in concert with 10 µM NAC for 7d (right). The magnification was X 200. Scale bar represents 10 µm in length. (B) The change of mammosphere number by co-treatment with 10 µM NAC. Results were expressed as the number of mammospheres per 1,000 seeded cells at 5d (mean ± SD, n = 3). (C) Secondary MCF-7 mammospheres treated with 100 nM E2 (upper panels) in concert with 10 µM NAC (lower panels). The original magnifications were X 200 except left panel of (C) (X 100). Scale bar represents 10 µm in length.

Figure 5. Regulation of MCF-7 mammosphere formation by breast cancer promoters and metformin.

(A) TCDD increased proliferation of MCF-7 cells in a dose responsive manner (upper), however, BPA did not (middle). On the other hand, metformin treatment decreased MCF-7 proliferation (mean ± SD, n = 3). *, P<0.05; **, P<0.01; ***, P<0.001. (B, C) Metformin further decreased the size (B) and the number (C) of MCF-7 mammosphere formation enhanced by E2, TCDD or BPA (mean ± SD, n = 3). The magnification was X 200. Scale bar represents 50 µm in length. *, P<0.05; **, P<0.01; ***, P<0.001.

Figure 6. Regulation of OCT4 expression by metformin in MCF-7 cells.

(A) E2 and TCDD increases OCT4 expression levels in MCF-7 cell line, however, BPA did not (mean ± SD, n = 3). **, P<0.01; ***, P<0.001. (B) Schematics of primer design for chromatin immunoprecipitation to detect putative ERE sequences in OCT4 promoter regions. Arrow heads indicate locations of putative ERE sequences. DE, distal enhancer; PE, proximal enhancer; PP, proximal promoter. (C) Chromatin immunoprecipitation to assess ER alpha binding at putative ERE sequences in OCT4 promoter region suggested that a putative ERE sequence at -3544kb from OCT4 transcription starting site was bound to ER alpha (mean ± SD, n = 3). **, P<0.01; ***, P<0.001. (D) The ERE sequence at -3544kb was enriched with ER alpha by treatment of E2 and TCDD compared to control and BPA treatment group. The enrichment was attenuated by co-treatment of metformin (mean ± SD, n = 3). *, P<0.05; **, P<0.01; ***, P<0.001.