17β-Estradiol inhibits iron hormone hepcidin through an estrogen responsive element half-site

Abstract

Interaction of estrogen with iron at the systemic level is long suspected, but direct evidence linking the two is limited. In the present study, we examined the effects of 17β-estradiol (E2) on hepcidin, a key negative regulator of iron absorption from the liver. We found that transcription of hepcidin was suppressed by E2 treatment in human liver HuH7 and HepG2 cells, and this down-regulation was blocked by E2 antagonist ICI 182780. Chromatin immunoprecipitation, deletion, and EMSA detected a functional estrogen responsive element half-site that is located between -2474 and -2462 upstream from the start of transcription of the hepcidin gene. After cloning the human hepcidin promoter into the pGL3Luc-Reporter vector, luciferase activity was also down-regulated by E2 treatment in HepG2 cells. E2 reduced hepcidin mRNA in wild-type mice as well as in hemochromatosis Fe gene knockout mice. In summary, our data suggest that hepcidin inhibition by E2 is to increase iron uptake, a mechanism to compensate iron loss during menstruation. This mechanism may also contribute to increased iron stores in oral contraceptive users.

Figures

Fig. 1.

Prediction of ERE in the promoter region of the hepcidin gene. A, The human hepcidin promoter region contains four ERE. B, Comparison of the ERE sequence of the hepcidin gene with the consensus sequence of human c-Fos, PTTG1IP, and CYP1B1 genes. The invariant sequence is underlined. C, Illustration of ERE half-site, SP-1, and AP-1/2-binding site sequences in human hepcidin promoter region between −2500 and −2422. Nucleotide numbers are denoted with the transcription start site assigned as +1.

Fig. 2.

Inhibition of hepcidin mRNA expression by E2 in human liver HuH7 and HepG2 cells. A, HuH7 cells were treated with 10 and 100 nm E2 and 10 ng/ml IL-6 and mRNA levels of hepcidin were measured by RT-PCR. One representative gel from three independent experiments was shown. B, HepG2 cells were treated with E2 and hepcidin mRNA levels were measured by qPCR (n = 3). C, Inhibition of hepcidin mRNA by E2 and ICI 182780 pretreatment abolished E2-mediated hepcidin inhibition in HepG2 cells at 4- and 24-h treatments, respectively (n = 3). Y-axis was presented as a log scale. *, Significantly different from the untreated control.

Fig. 3.

Identification of functional ERE in the promoter region of hepcidin. A, Recruitment of ERα by E2 to the ERE site between −2630 and −2072 of the human hepcidin gene. After treating HuH7 cells with 100 nm E2 for 24 h, fragmented chromatin was immunoprecipitated with anti-ERα antibody and amplified by PCR using oligonucleotide primers flanking the ERE site. B, WT and the ERE half-site-deleted mutant were transfected into HepG2 cells. These cells were treated with E2 for 24 h and then analyzed for luciferase activity. The luciferase activity was normalized to Renilla luciferase activity, and the bar graph represents mean ± sd of three independent experiments. *, Significantly different from the untreated control.

Fig. 4.

Binding of nuclear proteins to the ERE of the hepcidin gene by EMSA. HepG2 cells were treated with or without 100 nm E2 for 24 h, and nuclear proteins were isolated and incubated with biotin-labeled mutant and WT oligonucleotides containing the ERE half-site of the hepcidin gene as described in Table 1. The resulting complexes were resolved by nondenaturing polyacrylamide gel electrophoresis. For competition EMSA, a 200-fold excess of unlabeled nonspecific or specific oligonucleotides harboring the ERE half-site of the hepcidin gene was added during the preincubation period. For supershift EMSA, anti-ERα antibody or normal human IgG was added. The arrows indicate supershift, ERE binding complex, and free probe.

Fig. 5.

Inhibition of hepcidin transcription in vivo by E2. WT and Hfe gene knockout mice were ip injected with 100 nmol E2/kg bodyweight (six mice per group). After killing at 24 h, liver samples were isolated and used for RNA extractions. Levels of mouse hepcidin-1 (A) and hepcidin-2 (B) mRNA were measured by qPCR. Data were presented as means ± sd. Y-axis was presented as a log scale. *, Significantly different from the untreated controls.

Fig. 6.

Alteration of body iron homeostasis in vivo by E2. WT mice (n = 6) were treated with E2 as described in the legend of Fig. 5. After killing at 24 h, blood samples were collected for measurements of serum iron (A) by ferrozine assay. Bone marrow samples were isolated from tibia and femora and used for qPCR analyses of TfR1 (B) and DMT-1 (C). Data were presented as means ± sd *, Significantly different from the untreated controls (ctr).