Inhibiting the Immunophilin FKBP12: A Potential Therapeutic Approach to Iron Overload/Low Hepcidin Disorders

Introduction

Hepcidin negatively regulates body iron by binding and degrading the iron exporter ferroportin. Its expression is mainly controlled by the liver BMP-SMAD pathway whose activation requires BMP ligands (BMP2 and BMP6), constitutively active BMP type II receptors and the type I receptors ALK2 and ALK3. The co-receptor hemojuvelin (HJV) further potentiates the signaling. Mutations in key genes of the pathway impair hepcidin synthesis in Hereditary Hemochromatosis (HH), one of the most severe form being due to HJV mutations. As current therapies are symptomatic and do not correct hepcidin deficiency, alternative targeted therapies to increase hepcidin production are needed.

We have demonstrated that the immunophilin FKBP12 binds ALK2 in hepatoma cells to prevent uncontrolled activation of the BMP pathway in the absence of ligands. We also demonstrated that FKBP12 sequestration by immunosuppressive drugs, such as FK506 (tacrolimus, TAC) or rapamycin, upregulate hepcidin in vitro. The role of FKBP12 is maintained in vivo since acute TAC treatment of wild type (WT) mice increases hepcidin expression (Colucci et al., Blood 2017).

The aim of this study is to investigate whether pharmacologic inactivation of FKBP12, by TAC or antisense oligonucleotides (ASO), upregulates hepcidin for therapeutic purposes.

Methods

Primary hepatocytes isolated from WT, Hjv and Tfr2 KO mice (sv129/j background) were treated with increasing concentrations of TAC. Hepcidin and Id1 expression was investigated by qRT-PCR. Nine-weeks-old Hjv KO male mice were treated for 28 days with TAC (0.37 mg/h) delivered through surgically implanted mini-osmotic pumps. Six-weeks-old WT mice, were treated twice a week for 6 weeks with 50 mg/kg of Fkbp12 or control ASO. Mice were sacrificed and analyzed for iron, CBC, erythropoiesis and liver expression of hepcidin and BMP-SMAD target genes.

Results

TAC treatment of primary HCs from Hjv KO (Colucci et al., Blood 2018) and Tfr2 KO mice upregulates hepcidin as in WT mice, suggesting that HJV and TFR2 are dispensable for FKBP12-dependent hepcidin regulation and providing the proof of principle for FKBP12 targeting in HH. First, we explored a drug repurposing approach in the severe Hjv KO mice by chronic subcutaneous delivery of suboptimal, non-immunosuppressive TAC doses. Treatment of Hjv KO mice with TAC upregulates hepcidin via BMP-SMAD pathway activation, as assessed by Id1 and Smad7 upregulation. Since TAC also inhibits calcineurin upon FKBP12 sequestration and to avoid potential off-targets effect, FKBP12 was inactivated by ASO. Fkbp12 ASO treatment of WT mice decreases Fkbp12 expression by about 70-80% in liver, spleen and kidney, but not in the bone marrow. Fkbp12 ASO-treated mice exhibit microcytic anemia, decreased serum iron and upregulation of liver BMP-SMAD target genes. However, hepcidin remains inappropriately high considering the low serum iron. Fkbp12 ASO-treated mice show increased spleen immature erythroid precursors and increased expression of erythroferrone (Erfe). This is due to the unexpected effect of FKBP12 on spleen erythropoiesis and likely explains the lack of hepcidin upregulation.

Conclusions

1. Chronic TAC treatment in Hjv KO mice, which causes FKBP12 sequestration, improves hepcidin expression via BMP-SMAD pathway and favors spleen iron retention, suggesting that this "drug repurposing approach" may be beneficial for all iron overload disorders exhibiting low hepcidin.

2. ASO-Fkbp12 in WT mice efficiently downregulates hepatic Fkbp12, upregulates the BMP-SMAD pathway but leaves hepcidin unchanged compared to control mice. We hypothesize that this result reflects the concomitant hepcidin inhibition due to decreased serum iron and increased Erfe expression. The latter is a consequence of spleen Fkbp12 reduction in ASO-treated mice, suggesting that a hepatocyte targeting Fkbp12 ASO is required for therapeutic purposes.

3. Since Fkbp12 binds and inhibits ALK2 (Colucci et al., Blood 2017), our in vitro and in vivo data suggest that HJV and TFR2 functionally interact with ALK3 but not with ALK2.