Progesterone receptor membrane component-1 regulates hepcidin biosynthesis

Abstract

Iron homeostasis is tightly regulated by the membrane iron exporter ferroportin and its regulatory peptide hormone hepcidin. The hepcidin/ferroportin axis is considered a promising therapeutic target for the treatment of diseases of iron overload or deficiency. Here, we conducted a chemical screen in zebrafish to identify small molecules that decrease ferroportin protein levels. The chemical screen led to the identification of 3 steroid molecules, epitiostanol, progesterone, and mifepristone, which decrease ferroportin levels by increasing the biosynthesis of hepcidin. These hepcidin-inducing steroids (HISs) did not activate known hepcidin-inducing pathways, including the BMP and JAK/STAT3 pathways. Progesterone receptor membrane component-1 (PGRMC1) was required for HIS-dependent increases in hepcidin biosynthesis, as PGRMC1 depletion in cultured hepatoma cells and zebrafish blocked the ability of HISs to increase hepcidin mRNA levels. Neutralizing antibodies directed against PGRMC1 attenuated the ability of HISs to induce hepcidin gene expression. Inhibiting the kinases of the SRC family, which are downstream of PGRMC1, blocked the ability of HISs to increase hepcidin mRNA levels. Furthermore, HIS treatment increased hepcidin biosynthesis in mice and humans. Together, these data indicate that PGRMC1 regulates hepcidin gene expression through an evolutionarily conserved mechanism. These studies have identified drug candidates and potential therapeutic targets for the treatment of diseases of abnormal iron metabolism.

Figures

Figure 6. HISs increase hepcidin biosynthesis in mice and humans.

(A) qPCR of hepcidin gene expression in mouse liver 10 hours after injection with vehicle or mifepristone. *P < 0.05, 2-tailed t test (n = 5 per group). (B) Serum hepcidin levels in women prior to progesterone treatment (day 1) or after receiving progesterone daily (day 6 and day 15) during a standard in vitro fertilization protocol using FET. #P < 0.001 compared with day 1 sample, †P < 0.01 compared with day 1 sample, **P < 0.01 compared with day 6 sample, paired t test with Bonferroni adjustment for multiple comparisons (n = 20 per group). All results expressed as mean ± SEM.

Figure 5. Progesterone, but not BMP6 or IL-6, requires SFK activity to increase hepcidin gene expression.

qPCR for hepcidin gene expression in HepG2 cells serum starved for 18 hours and then preincubated with the SFK inhibitor PP2 (10 μM) or control (DMSO) for 30 minutes and then exposed to (A) progesterone (30 μM for 8 hours), (B) BMP6 (20 ng/ml for 3 hours), or (C) IL-6 (100 ng/ml for 3 hours). #P < 0.001 compared with control treated cells exposed to vehicle, †P < 0.001 compared with control treated cells exposed to progesterone, 2-way ANOVA (n = 4 per group). All results expressed as mean ± SEM.

Figure 4. PGRMC1 is required for progesterone, but not BMP6 or IL-6, to increase hepcidin gene expression.

(A–C) qPCR for PGRMC1 gene expression in HepG2 cells transfected with scrambled siRNAs (siControl) or siRNAs directed against PGRMC1 (siPGRMC1) and exposed to vehicle, (A) progesterone (30 μM for 8 hours), (B) BMP6 (20 ng/ml for 3 hours), or (C) IL-6 (100 ng/ml for 3 hours. *P < 0.01, #P < 0.001 compared with siControl cells treated with vehicle, 2-way ANOVA (n = 4 per group). (D–F) qPCR for hepcidin gene expression in HepG2 cells transfected with siControl or siPGRMC1 siRNAs and treated with vehicle, (D) progesterone, (E) BMP6, or (F) IL-6 (identical concentrations and incubation times as above). (G–I) qPCR for hepcidin gene expression in HepG2 cells preincubated with control IgG or anti-PGRMC1 antibodies (PGRMC1 Ab) for 3 hours and then exposed to vehicle or (G) progesterone, (H) BMP6, or (I) IL-6 (identical concentrations and incubation times as above). #P < 0.001, *P < 0.01, compared with siControl cells (D–F) or IgG-treated cells (G–I) exposed to vehicle; **P < 0.001 compared with siControl cells (D) or IgG-treated cells (G) exposed to progesterone, 2-way ANOVA (n = 4 per group). (J–L) A plasmid containing the pgrmc1 morpholino oligonucleotide target site upstream of an EGFP coding sequence was injected (J) alone, (K) with control morpholino oligonucleotides (control MO), or (L) with pgrmc1 morpholino oligonucleotides (pgrmc1 MO) into 1-cell stage zebrafish embryos, and fluorescence was examined after 8 hours (original magnification, ×2.5). (M) qPCR for hepcidin gene expression in 3-day-old wild-type zebrafish embryos injected with control morpholino oligonucleotides or pgrmc1 morpholino oligonucleotides and treated with vehicle or progesterone (5 μM for 12 hours). #P < 0.001 compared with control morpholino oligonucleotides embryos treated with vehicle, **P < 0.001 compared with control morpholino oligonucleotides embryos treated with progesterone, 2-way ANOVA (n = 5 per group). qPCR results expressed as mean ± SEM.

Figure 3. HISs do not activate the BMP or STAT3 signaling pathway.

(A and B) SMAD1/5 phosphorylation (P-SMAD 1/5) and total SMAD1 (T-SMAD 1) levels in HepG2 cells exposed to vehicle (V) or BMP6 (20 ng/ml) for 2 hours and to progesterone (30 μM) or mifepristone (30 μM) for up to 24 hours. (C and D) STAT3 phosphorylation and total STAT3 levels in HepG2 cells exposed to vehicle or IL-6 (100 ng/ml) for 2 hours and to progesterone (30 μM) or mifepristone (30 μM) for up to 24 hours. Representative Western blots from 3 independent experiments are shown in A–D. (E) Treatment of HepG2 cells with BMP6 (20 ng/ml) for 24 hours increased Id1 promoter activity (BRE-luc), while treatment with progesterone (30 μM) or mifepristone (30 μM) had no effect on Id1 promoter activity. (F and G) Treatment of HepG2 cells for 24 hours with (F) BMP6 (20 ng/ml) or (G) IL-6 (100 ng/ml) increased hepcidin promoter activity nearly 15-fold and 5-fold, respectively, as measured by the hepcidin luciferase promoter assay (Hep-luc), while treatment of HepG2 cells with progesterone (30 μM) or mifepristone (30 μM) did not affect hepcidin promoter activity. Results in E–G are expressed as mean ± SEM, #P < 0.001 compared with vehicle treated, ANOVA (n = 6 per group). (H and I) qPCR for id1 and liv1 gene expression in zebrafish. Three-day-old wild-type zebrafish larvae were treated with progesterone (5 μM) or mifepristone (5 μM) for 12 hours and then lysed prior to qPCR analysis. Results are expressed as mean ± SEM (n = 4 per group).

Figure 2. Epitiostanol, progesterone, and mifepristone increase hepcidin gene expression.

(A–C) qPCR for hepcidin gene expression in zebrafish larvae treated with vehicle (DMSO) or 10 μM epitiostanol (dissolved in DMSO) (n = 5 per group), vehicle (ethanol) or 5 μM progesterone (n = 3 per group), and vehicle (ethanol) or 5 μM mifepristone (n = 4 per group), respectively. Three-day-old zebrafish larvae were treated with each steroid for 12 hours and then lysed for analysis. †P < 0.05 compared with vehicle treated, 2-tailed t test. (D–F) qPCR for hepcidin gene expression in human HepG2 cells treated with vehicle (DMSO) or 30 μM epitiostanol dissolved in DMSO (for 24 hours), vehicle (ethanol) or 30 μM progesterone (for 8 hours), and vehicle (ethanol) or 30 μM mifepristone (for 8 hours), respectively. Results are expressed as mean ± SEM, *P < 0.01 compared with vehicle treated, #P < 0.001 compared with vehicle treated, 2-tailed t test (n = 4 per group for HepG2 cells). (G and H) qPCR for hepcidin gene expression in HepG2 cells pretreated with 10 μM actinomycin D for 30 minutes and then incubated with vehicle (ethanol), 30 μM progesterone, or 30 μM mifepristone for 8 hours. Results are expressed as mean ± SEM, #P < 0.001 compared with control vehicle treated, ##P < 0.001 compared with control cells treated with progesterone or mifepristone, 2-way ANOVA (n = 4 per group).

Figure 1. A small-molecule screen in transgenic zebrafish identifies compounds that promote the degradation of ferroportin.

(A–D) Fluorescent dissection microscopy and confocal microscopy of Fpn-GFP transgenic zebrafish larvae treated with vehicle (0.1% DMSO) or 10 μM epitiostanol (dissolved in DMSO). Three-day-old transgenic embryos were soaked in E3 buffer containing either vehicle or epitiostanol for 12 hours before imaging (original magnification, ×4 [A and B]; ×40 [C and D]). High-magnification images of boxed regions in A and B are shown in C and D, respectively. (E–I) Iron staining of wild-type and ferroportin transgenic fish (ubi) treated with DMSO or epitiostanol (epi). Three-day-old embryos were treated with DMSO or 10 μM epitiostanol dissolved in DMSO for 12 hours and then fixed. Enhanced Prussian blue staining was then performed on each embryo. Arrows indicate iron deposition in the caudal hematopoietic tissue (original magnification, ×8 [E–H]). Iron staining was quantified by ImageJ based on the pixel intensity, as shown in I and described in the Methods. Results are expressed as mean ± SEM, *P < 0.001 ubi plus DMSO (n = 8), compared with wild-type plus DMSO (n = 8), wild-type plus epitiostanol (n = 14), or ubi plus epitiostanol (n = 13), ANOVA. (J) Chemical structure of some compounds that were tested in this screen. (K–P) Fluorescent microscopy (original magnification, ×4) of ferroportin transgenic fish treated with the steroid compounds listed in J. Three-day-old transgenic embryos were treated with vehicle (DMSO, n = 13), 10 μM epitiostanol dissolved in DMSO (n = 11), 10 μM mifepristone (n = 11), 10 μM progesterone (n = 9), or 10 μM dexamethasone (n = 13) for 12 hours, respectively. (P) The intensity of fluorescent staining was quantified for each treatment group and is further described in the Methods. Results are expressed as mean ± SEM, *P < 0.01, compared with the vehicle control, ANOVA.