The role of ubiquitination in hepcidin-independent and hepcidin-dependent degradation of ferroportin

Abstract

The iron exporter ferroportin (Fpn) is essential to transfer iron from cells to plasma. Systemic iron homeostasis in vertebrates is regulated by the hepcidin-mediated internalization of Fpn. Here, we demonstrate a second route for Fpn internalization; when cytosolic iron levels are low, Fpn is internalized in a hepcidin-independent manner dependent upon the E3 ubiquitin ligase Nedd4-2 and the Nedd4-2 binding protein Nfdip-1. Retention of cell-surface Fpn through reductions in Nedd4-2 results in cell death through depletion of cytosolic iron. Nedd4-2 is also required for internalization of Fpn in the absence of ferroxidase activity as well as for the entry of hepcidin-induced Fpn into the multivesicular body. C. elegans lacks hepcidin genes, and C. elegans Fpn expressed in mammalian cells is not internalized by hepcidin but is internalized in response to iron deprivation in a Nedd4-2-dependent manner, supporting the hypothesis that Nedd4-2-induced internalization of Fpn is evolutionarily conserved.

Figure 1. Depletion of cytosolic iron leads to Fpn degradation

(A) HEK293T Fpn-GFP cells were incubated with 10 μM Ponasterone A (PoA) to induce Fpn-GFP. Fpn-GFP localization was examined by epifluorescence microscopy after 18h, 24h and 36h of incubation. The percent of plasma membrane fluorescence per cell was determined as described in the experimental procedures. Error bars represent the standard error of the mean of three separate experiments. Fpn levels were analyzed by Western blot using rabbit or mouse antibodies against Fpn or tubulin followed by peroxidase-conjugated goat anti-rabbit/mouse IgG. (B) HEK293T cells were transiently transfected with WT Fpn-GFP, , Fpn(Y302-303F)-GFP Fpn(K253A)-GFP, Fpn(N174I)-GFP or Fpn(C326Y)-GFP expressed under a CMV promoter. Eighteen and 36h after transfection Fpn-GFP localization was examined by epifluorescence microscopy. The percentage of cells showing cell surface fluorescence versus intracellular fluorescence was determined as described in Experimental Procedures. Error bars represent the standard error of the mean of three separate experiments. Fpn levels were analyzed by Western blot using rabbit or mouse antibodies against GFP followed by peroxidase-conjugated goat anti-rabbit IgG. (C) HEK293T cells expressing Fpn-GFP were incubated with 10μM Ponasterone A (PoA) in presence or absence of FAC (10 μM Fe) or 100 μM DFX added six h post induction. Fpn-GFP localization was examined by epifluorescence microscopy and the percent of plasma membrane fluorescence per cell was determined as described in the Experimental Procedures. Cells were biotinylated using the impermeable sulfo-NHS-SS-biotin. After biotinylation, cells were solubilized and biotinylated proteins were affinity purified using streptavidin affinity gel. Affinity-purified samples and the flow through were analyzed by Western blot using a rabbit anti-GFP followed by peroxidase-conjugated goat anti-rabbit IgG. (D) HEK293T cells were transfected with WT Fpn-GFP or Fpn(K253A)-GFP and pCMV-dynaminK44A. Eighteen h post transfection cells were incubated in the presence or absence of 100 μM DFX. After four h cells were lysed and Fpn was immunoprecipitated using anti-GFP antibodies. Immunoprecipitated samples were analyzed for GFP and ubiquitination. Error bars represent the standard error of the mean of three separate experiments. * = p <0.05, **= p <0.005 and ns = not significance.

Figure 2. Cytosolic iron depletion leads to a Nedd4-2 dependent ubiquitination of Fpn

(A) HEK293TFpn-GFP cells were transfected with either non-specific siRNA (N.S.) or human Nedd4-2 specific siRNA oligonucleotide pools using OligofectAMINE. Forty-eight h later cells were induced to express Fpn-GFP and after 18h and 36h Fpn-GFP localization was examined by epifluorescence microscopy (panel i). Cells were biotinylated using sulfo-NHS-SS-biotin. After biotinylation, cells were solubilized and biotinylated proteins were affinity purified using streptavidin affinity gel. Affinity-purified samples and the flow through were analyzed by Western blot using a rabbit anti-GFP followed by peroxidase-conjugated goat anti-rabbit IgG (panel ii). Silencing was assessed by Western blot with tubulin as a loading control (panel ii). The percent of plasma membrane fluorescence per cell was determined as described in the experimental procedures (panel iiii). Error bars represent the standard error of the mean of three separate experiments. Cells treated as above were transfected with siRNA resistant mouse Nedd4-2 expressed under the control of the CMV promoter (panel iii). (B) HEK293TFpn-GFP cells were silenced as in A. Forty-eight h later cells were incubated in the presence or absence of 100 μM DFX and Fpn-GFP localization was examined by epifluorescence microscopy six h later and images quantified as in A. (C) Mouse bone marrow macrophages were transfected with either nonspecific (N.S.) or mouse Nedd4-2 specific siRNA pools using OligofectAMINE. Forty-eight h later cells were incubated with 250 μM BCS. Fpn localization was examined by immunofluorescence microscopy 18 h later. The percentage of cells showing cell surface fluorescence versus intracellular fluorescence was determined as described in Experimental Procedures. Silencing efficiency was assessed by Western blot. Error bars represent the standard error of the mean of three separate experiments. * = p <0.05 and ns = not significance.

Figure 3. Nedd4-2 is required for entry of hepcidin-internalized Fpn into the MVB

HEK293TFpn-GFP cells were transfected with either NS siRNA or human Nedd4-2 specific siRNA oligonucleotide pools using OligofectAMINE. Forty-eight h later cells were induced to express Fpn-GFP. Eighteen h post Fpn-GFP induction cells were incubated in presence or absence of 1 μg/ml hepcidin for 30 min and Fpn-GFP localization was examined by epifluorescence microscopy. Silencing efficiency and Fpn levels were assessed by Western blot.

Figure 4. Nedd4-2 internalization requires the Nedd4-2 binding protein Ndfip1

(A) Mouse bone marrow macrophages were incubated with either NS or human Ndfip-1 specific siRNA oligonucleotide pools using OligofectAMINE. Forty-eight h post silencing cells were incubated with or without 100 μM DFX for an additional four h. Silencing and Fpn levels were assessed by Western blot with tubulin as a loading control. (B) Mouse bone marrow macrophages were incubated with either NS or human Ndfip-2 specific siRNA oligonucleotide pools, treated as in A and silencing and Fpn levels assessed. (C) Mouse bone marrow macrophages obtained from mice homozygous for a targeted deletion in Ndfip1, were transfected with pEGFP-WTFpn. Eighteen h later cells were incubated in the presence or absence of 100 μM DFX or 1 μg/ml hepcidin for four h. Fpn-GFP localization was examined by epifluorescence microscopy. The percentage of cells showing cell surface fluorescence versus intracellular fluorescence was determined as described in Experimental Procedures. Error bars represent the standard error of the mean of three separate experiments. **= p <0.005.

Figure 5. Iron-deprivation-induced Fpn degradation results from loss of transport substrate

HEK293T cells expressing Fpn-GFP were incubated with 10 μM Ponasterone A. After 18h 100 μM DFX and 50 μM ZnSO₄ was added and cells incubated at 37°C for four h. Fpn-GFP localization was examined by epifluorescence microscopy. The percentage of cells showing cell surface fluorescence versus intracellular fluorescence was determined as described in Experimental Procedures. Error bars represent the standard error of the mean of three separate experiments. Cells were biotinylated using sulfo-NHS-SS-biotin. After biotinylation, cells were solubilized and biotinylated proteins were affinity purified using streptavidin affinity gel. Affinity-purified samples and the flow through were analyzed by Western blot using a rabbit anti-GFP followed by peroxidase-conjugated goat anti-rabbit IgG. **= p <0.005.

Figure 6. An inability to downregulate Fpn leads to apoptosis

Mouse bone marrow macrophages were incubated with NS siRNA or human Nedd4-2 specific siRNA oligonucleotide pools using OligofectAMINE. Forty-eight h post silencing cells were incubated in the absence (A) or presence of 100 μM DFX (B) or FAC (10 μM Fe) (C). Twelve or 18h later cells were incubated with Hoechst 33342 dye for 15 min at 37°C. Cells were analyzed using a spectrofluorometer with excitation of 346 and emission of 460 nm. All data were normalized to cell protein. Error bars represent the standard error of the mean of three separate experiments. **= p <0.005. * = p <0.05, **= p <0.005 and ns = not significance.

Figure 7. Hepcidin-independent internalization is evolutionarily conserved

(A) Sequence alignment of WT mouse Fpn and WT C. elegans Fpn. The hepcidin binding domain (HBD) is highlighted in the box. (B) HEK293T cells were transfected with empty vector pEGFP, WT mouse pEGFP-Fpn (mFpn) or WT C. elegans pEGFP-Fpn (cFpn). After transfection cells were incubated in presence of FAC (10μM Fe). Eighteen h post transfection iron was removed and cells were incubated for further 18h (+/−FAC). Cells were solubilized and ferritin levels measured by ELISA. (C) HEK293T cells were transfected WT cFpn. Eighteen h post transfection cells were incubated with or without 1 μg/ml hepcidin, 100 μM DFX or 50 μM ZnSO₄ plus 100 μM DFX for six h. Fpn-GFP localization was examined by epifluorescence microscopy. Fpn and tubulin level were analyzed by Western blot using tubulin as loading control. The percentage of cells showing cell surface fluorescence versus intracellular fluorescence was determined as described in Experimental Procedures. Error bars represent the standard error of the mean of three separate experiments. * = p <0.05 and **= p <0.005.