Abstract

Hemochromatosis type 4 is a rare form of primary iron overload transmitted as an autosomal dominant trait caused by mutations in the gene encoding the iron transport protein ferroportin 1 (SLC40A1). SLC40A1 mutations fall into two functional categories (loss- versus gain-of-function) underlying two distinct clinical entities (hemochromatosis type 4A versus type 4B). However, the vast majority of SLC40A1 mutations are rare missense variations, with only a few showing strong evidence of causality. The present study reports the results of an integrated approach collecting genetic and phenotypic data from 44 suspected hemochromatosis type 4 patients, with comprehensive structural and functional annotations. Causality was demonstrated for 10 missense variants, showing a clear dichotomy between the two hemochromatosis type 4 subtypes. Two subgroups of loss-of-function mutations were distinguished: one impairing cell-surface expression and one altering only iron egress. Additionally, a new gain-of-function mutation was identified, and the degradation of ferroportin on hepcidin binding was shown to probably depend on the integrity of a large extracellular loop outside of the hepcidin-binding domain. Eight further missense variations, on the other hand, were shown to have no discernible effects at either protein or RNA level; these were found in apparently isolated patients and were associated with a less severe phenotype. The present findings illustrate the importance of combining in silico and biochemical approaches to fully distinguish pathogenic SLC40A1 mutations from benign variants. This has profound implications for patient management.

INTRODUCTION

Hemochromatosis type 4 (OMIM #606069), also called ferroportin disease, is an inborn error in iron metabolism transmitted through autosomal dominant inheritance and associated with mutations in the solute-carrier family 40 member 1 (SLC40A1) gene. The disease is characterized by wide clinical heterogeneity, showing marked differences in type of target cell and iron deposition, sub-phenotypes and degree of penetrance (1–3). Although rare, hemochromatosis type 4 is the second most common cause of hereditary iron overload, after HFE-related hemochromatosis (4).

SLC40A1, also known as ferroportin 1 or FPN1 (Uni-Prot #Q9NP59), is the sole iron export protein reported in mammals and is expressed in all types of cell that handle major iron flow, including iron-recycling macrophages, iron-storing hepatocytes and absorptive enterocytes (5). SLC40A1 is predominantly regulated by the liver-derived peptide hepcidin, which binds SLC40A1 on the cell surface and induces internalization and degradation (6). Hepcidin-SLC40A1 interaction is critical in both normal iron homeostasis and common iron metabolism pathologies, including not only inherited disorders and iron overload but also non-genetic diseases and anemia (7–9).

SLC40A1 gene mutations fall into two functional categories, underlying two distinct clinical entities (hemochromatosis type 4A or classical ferroportin disease, versus type 4B or non-classical ferroportin disease) (3, 9, 10). Loss-of-function mutations, which are more frequent, are associated with hemochromatosis type 4A, characterized by relative plasma iron deficiency (manifesting in normal-to-low transferrin saturation) and preferential iron retention in reticuloendothelial cells (manifesting in high serum ferritin concentrations). As the disease progresses, iron is deposited in hepatocytes, normally accompanied by a rise in transferrin saturation. Aggressive phlebotomy regimens can be a problem in the early stages of the disease, when patients often display borderline anemia. Gain-of-function mutations, on the other hand, result in higher plasma iron concentrations (manifesting in elevated transferrin saturation levels), due to abnormal iron release from iron-recycling macrophages and enterocytes. Hyperferritinemia is secondary to the plasma iron burden and is mostly associated with iron deposits in liver parenchymal cells (hepatocytes). These biochemical and histological features mimic the natural history of typical HFE-related hemochromatosis. Hence, patients with ferroportin gain-of-function mutations usually respond well to phlebotomy.

Thus far, a total of 50 SLC40A1 variations have been shown to be associated with iron overload phenotypes (11–13): 46 missense mutations, three deletions (c.−59-45del, p.Val162del and p.Gly330del) and one splicing mutation (c.1402 G > A). To fully understand the allelic heterogeneity found at the SLC40A1 locus, it is noteworthy that few variations have been identified in multiple pedigrees worldwide (1, 12, 13), indicating that only private or population-specific variants contribute to disease . Additionally, a lack of genotype–phenotype correlation has been reported between unrelated patients and within families (1, 2, 13). Finally, it must be borne in mind that only a small number of variations have been investigated in vitro and that functional studies have produced conflicting results (14–19). Hence, despite identification in patients, not all 50 SLC40A1 variations can be firmly established as disease-causing mutations.

Establishing the pathogenicity of SLC40A1 variations is of major diagnostic importance, with implications for other family members and for therapeutic approaches targeting iron withdrawal. It can also aid in identifying individuals at high risk of liver fibrogenesis and carcinogenesis (20, 21). The present study provides evidence that an integrated strategy combining structural and functional annotations with clinical findings can reduce the risk of misdiagnosis.

RESULTS

Molecular analysis of 44 hyperferritinemia patients and frequency of SLC40A1 variations in a Caucasian control population

Eighteen heterozygous SLC40A1 variations (1 single amino acid deletion and 17 missense mutations) were identified in 44 individuals (22 female, 22 male; Table 1). All patients were negative for genotypes known to cause hemochromatosis types 1–3 (in the HFE, HJV, HAMP and TFR2 genes) and for mutations in the FTL gene 5′ untranslated region (hyperferritinemia-cataract syndrome; OMIM#600886) and exon 1 (hyperferritinemia without iron overload and cataract; OMIM#134790).

Table 1.

Description of the study sample: SLC40A1 variations, family relationships, biological and clinical data

Variation Family relationships Patient Gender Agea (years) Serum ferritin (µg/l) Transferrin saturation (%) AST (UI/l) ALT (UI/l) GGT (UI/l) CRP (mg/l) RBC (Tera/l) Hgb (g/dl) MCV (fl) HIC (µmol/g) Alcohol consumptionb Clinical observations 
p.Gly80Ser Index case FRA-05-005 49 702 24 22 29   12.1 91.0  Abstinent  
Daughter FRA-05-010 29 519 47 17 15   <0.3 4.6 14.9 96.0  Abstinent  
Son FRA-05-004 27 1200         130 Abstinent Fatigue 
p.Arg88Gly Index case FRA-03-002 41 4129 54 27 34 19  4.4 13.5 93.0  Abstinent  
Brother FRA-06-004 38 2840   52 24  4.5 13.9 88.4  Abstinent Fatigue 
p.Asp157Gly Index case FRA-06-007 64 4459 35 32 38 25 3.3 5.3 15.7 90.0  Moderate Arthralgia 
Daughter FRA-06-002 42 1590 20 17 28 19 0.6 4.4 13.0 91.0  Moderate Fatigue 
p.Val162del Index case FRA-02-009 31 2578 35 19 22 15  <2 3.7 11.4 95.7  Moderate  
Brother FRA-02-008 21 71.2 32 16 19 18  <2 5.1 15.2 89.8  Moderate Fatigue, arthralgia 
p.Val162del Index case FRA-02-014 45 977 14 25 23 17 4.6 13.1 91.0  Abstinent Fatigue, arthralgia 
Daughter FRA-02-012 18 1572  16 19 4.5 12.7 88.4  Abstinent  
p.Val162del Index case FRA-05-008 19 2268 25      16.6   Moderate  
Sister FRA-05-018 19 2198 18         Abstinent  
p.Asp181Val Index case FRA-01-010 37 1679 24 14 13 16 5.0 12.8 90.3  Moderate Fatigue 
Sister FRA-01-007 25 1173 28 19 15 13 26.6 4.5 13.6 87.7  Moderate Fatigue 
Uncle FRA-01-009 57 8515 49 17 21 51 4.6 15.1 96.1  Moderate Fatigue, hepatomegaly 
p.Leu233Pro Index case FRA-06-001 52 4293 76 46 91 35   14.4 96.0 >300 Moderate  
Son FRA-06-005 23 1552 40           
p.Gly490Asp Index case FRA-06-006 54 3890     1.1  12.3 80.0 160   
Daughter FRA-06-008 25 910 32           
p.Asp504Asn Index case FRA-05-002 60 1514 89 37 68 86  4.6 16.5 108.0 320 Moderate Hepatomegaly 
Brother FRA-05-011 52 1605 102 36 113   4.9 15.7 93.0    
 Unrelated patients                
p.Gly80Ser  FRA-02-013 46 3109 35     4.8 14.1 94.0  Negligible  
p.Arg88Gly  FRA-06-003 77 4910 39 27 35 10 3.7 11.4 93.0 150 Abstinent Arthralgia, hepatomegaly, type 2 diabetes 
p.Asp157Gly  FRA-02-007 66  >6000 53         Abstinent Fatigue 
p.Asp157Tyr  FRA-02-016 64 2213 35 34 36 30  5.2 11.7 73.0  Moderate  
p.Val162del  FRA-01-001 52 12178 37 29 28 23 6.5 17.0 77.6  Abstinent Fatigue, hepatomegaly 
p.Ile180Thr  FRA-05-009 37 604 30     4.5 13.9 90.5    
p.Asp181Val  FRA-01-006 36 1973 29 22 20 27 4.4 13.2 89.5  Moderate Fatigue 
p.Thr230Asn  FRA-02-003 64 1015 30 31 52 33  <1 5.2 15.4 91.0  Moderate Arthralgia 
p.Gln248His  FRA-03-001 35 1305 38 40 20 110 5.2  86.0  Moderate Fatigue 
p.Met266Thr  FRA-05-012 40 1910 44 69 48 582  4.7 16.0 101.0  Heavy  
p.Leu345Phe  FRA-04-009 53 620 41           
p.Ile351Val  FRA-01-005 60 1293 32 15 14 17 3.6 3.8 11.8 91.7  Abstinent Fatigue 
p.Pro443Leu  FRA-04-003  585  17 28 20 4.8 14.4 86.0  Abstinent Arthralgia 
p.Gly490Asp  FRA-02-005 53 4377 22 32 34 51 4.5 14.5 94.0  Moderate Fatigue 
p.Gly490Asp  FRA-05-006 32 1853 33 21 22    14.8   Moderate Fatigue, arthralgia 
p.Gly490Asp  FRA-05-013 49 2151 38 14 15 14  3.9 10.9 89.2  Abstinent Fatigue, arthralgia, hepatomegaly 
p.Gly490Asp  FRA-05-014 31 2700 23      14.8    Fatigue, arthralgia 
p.Gly490Asp  FRA-05-015 33 2900 12 13 13 15  4.9 11.9 76.5  Moderate Fatigue, arthralgia 
p.Gly490Asp  FRA-05-016 23 255 10 20 17 25 3.6 11.0 88.7  Moderate Fatigue, hepatomegaly 
p.Gly490Asp  FRA-05-017 47 4728 53 40 43 46  4.2 13.0 90.4  Abstinent Arthralgia 
p.Gly490Ser  FRA-05-007 27 1050 13     4.5      
p.Arg561Gly  FRA-04-010 71 1100 43     4.5      
Variation Family relationships Patient Gender Agea (years) Serum ferritin (µg/l) Transferrin saturation (%) AST (UI/l) ALT (UI/l) GGT (UI/l) CRP (mg/l) RBC (Tera/l) Hgb (g/dl) MCV (fl) HIC (µmol/g) Alcohol consumptionb Clinical observations 
p.Gly80Ser Index case FRA-05-005 49 702 24 22 29   12.1 91.0  Abstinent  
Daughter FRA-05-010 29 519 47 17 15   <0.3 4.6 14.9 96.0  Abstinent  
Son FRA-05-004 27 1200         130 Abstinent Fatigue 
p.Arg88Gly Index case FRA-03-002 41 4129 54 27 34 19  4.4 13.5 93.0  Abstinent  
Brother FRA-06-004 38 2840   52 24  4.5 13.9 88.4  Abstinent Fatigue 
p.Asp157Gly Index case FRA-06-007 64 4459 35 32 38 25 3.3 5.3 15.7 90.0  Moderate Arthralgia 
Daughter FRA-06-002 42 1590 20 17 28 19 0.6 4.4 13.0 91.0  Moderate Fatigue 
p.Val162del Index case FRA-02-009 31 2578 35 19 22 15  <2 3.7 11.4 95.7  Moderate  
Brother FRA-02-008 21 71.2 32 16 19 18  <2 5.1 15.2 89.8  Moderate Fatigue, arthralgia 
p.Val162del Index case FRA-02-014 45 977 14 25 23 17 4.6 13.1 91.0  Abstinent Fatigue, arthralgia 
Daughter FRA-02-012 18 1572  16 19 4.5 12.7 88.4  Abstinent  
p.Val162del Index case FRA-05-008 19 2268 25      16.6   Moderate  
Sister FRA-05-018 19 2198 18         Abstinent  
p.Asp181Val Index case FRA-01-010 37 1679 24 14 13 16 5.0 12.8 90.3  Moderate Fatigue 
Sister FRA-01-007 25 1173 28 19 15 13 26.6 4.5 13.6 87.7  Moderate Fatigue 
Uncle FRA-01-009 57 8515 49 17 21 51 4.6 15.1 96.1  Moderate Fatigue, hepatomegaly 
p.Leu233Pro Index case FRA-06-001 52 4293 76 46 91 35   14.4 96.0 >300 Moderate  
Son FRA-06-005 23 1552 40           
p.Gly490Asp Index case FRA-06-006 54 3890     1.1  12.3 80.0 160   
Daughter FRA-06-008 25 910 32           
p.Asp504Asn Index case FRA-05-002 60 1514 89 37 68 86  4.6 16.5 108.0 320 Moderate Hepatomegaly 
Brother FRA-05-011 52 1605 102 36 113   4.9 15.7 93.0    
 Unrelated patients                
p.Gly80Ser  FRA-02-013 46 3109 35     4.8 14.1 94.0  Negligible  
p.Arg88Gly  FRA-06-003 77 4910 39 27 35 10 3.7 11.4 93.0 150 Abstinent Arthralgia, hepatomegaly, type 2 diabetes 
p.Asp157Gly  FRA-02-007 66  >6000 53         Abstinent Fatigue 
p.Asp157Tyr  FRA-02-016 64 2213 35 34 36 30  5.2 11.7 73.0  Moderate  
p.Val162del  FRA-01-001 52 12178 37 29 28 23 6.5 17.0 77.6  Abstinent Fatigue, hepatomegaly 
p.Ile180Thr  FRA-05-009 37 604 30     4.5 13.9 90.5    
p.Asp181Val  FRA-01-006 36 1973 29 22 20 27 4.4 13.2 89.5  Moderate Fatigue 
p.Thr230Asn  FRA-02-003 64 1015 30 31 52 33  <1 5.2 15.4 91.0  Moderate Arthralgia 
p.Gln248His  FRA-03-001 35 1305 38 40 20 110 5.2  86.0  Moderate Fatigue 
p.Met266Thr  FRA-05-012 40 1910 44 69 48 582  4.7 16.0 101.0  Heavy  
p.Leu345Phe  FRA-04-009 53 620 41           
p.Ile351Val  FRA-01-005 60 1293 32 15 14 17 3.6 3.8 11.8 91.7  Abstinent Fatigue 
p.Pro443Leu  FRA-04-003  585  17 28 20 4.8 14.4 86.0  Abstinent Arthralgia 
p.Gly490Asp  FRA-02-005 53 4377 22 32 34 51 4.5 14.5 94.0  Moderate Fatigue 
p.Gly490Asp  FRA-05-006 32 1853 33 21 22    14.8   Moderate Fatigue, arthralgia 
p.Gly490Asp  FRA-05-013 49 2151 38 14 15 14  3.9 10.9 89.2  Abstinent Fatigue, arthralgia, hepatomegaly 
p.Gly490Asp  FRA-05-014 31 2700 23      14.8    Fatigue, arthralgia 
p.Gly490Asp  FRA-05-015 33 2900 12 13 13 15  4.9 11.9 76.5  Moderate Fatigue, arthralgia 
p.Gly490Asp  FRA-05-016 23 255 10 20 17 25 3.6 11.0 88.7  Moderate Fatigue, hepatomegaly 
p.Gly490Asp  FRA-05-017 47 4728 53 40 43 46  4.2 13.0 90.4  Abstinent Arthralgia 
p.Gly490Ser  FRA-05-007 27 1050 13     4.5      
p.Arg561Gly  FRA-04-010 71 1100 43     4.5      

AST, aspartate aminotransferase; ALT, alanine aminotransferase; GGT, gamma-glutamyltransferase; CRP, C-reactive protein; RBC, red blood cell; Hgb, hemoglobin; MCV, mean corpuscular volume; HIC, hepatic iron concentration.

aAge at diagnosis.

bModerate: up to 1 drink per day for women and up to 2 drinks per day for men; heavy: more than 1 drink per day for women and 2 drinks per day for men.

To our knowledge, the p.Asp157Tyr, p.Thr230Asn, p.Met266Thr, p.Leu345Phe, p.Ile351Val and p.Asp504Asn SLC40A1 variations have not been previously reported; the others were previously observed either in pedigrees or in single patients (11–13).

A total of 734 DNA samples from healthy subjects, exclusively from north-western France (Brittany), were investigated to control the frequency of the 18 SLC40A1 variations. None of the 18 mutated alleles was detected. Allele frequencies in the control population were estimated at 0.00034 on Wald's method .

Identification of nine loss-of-function mutations: distinction between SLC40A1 mislocation and intrinsic inability to export iron

The iron-exporting function of the 18 SLC40A1 variants was assessed by first assaying the cytosolic iron storage protein ferritin, the expression of which is tightly regulated by the intracellular iron pool. HEK293T cells were transiently transfected with plasmids encoding SLC40A1-V5 fusion proteins in the presence of holotransferrin (to increase background ferritin levels). The p.Ala77Asp missense mutation, which significantly damages ferroportin structure and is known to prevent cell-surface localization (15), was used as a negative control. Figure 1A shows that cells expressing the p.Ile180Thr, p.Thr230Asn, p.Glu248His, p.Met266Thr, p.Leu345Phe, p.Ileu351Val, p.Pro443Leu, p.Asp504Asn or p.Arg561Gly variant were iron-depleted, similar to those expressing wild-type (WT) SLC40A1, indicating that these nine variants did not alter the iron-exporting function of the cells. In contrast, cells expressing the p.Gly80Ser, p.Arg88Gly, p.Val162del, p.Asp157Gly, p.Asp157Tyr, p.Asp181Val, p.Leu233Pro, p.Gly490Asp or p.Gly490Ser variant were iron-loaded, similar to those expressing the p.Ala77Asp loss-of-function mutation. Comparable ferritin levels were found in cells transfected with the commercial pcDNA3.1-V5-His vector (no-SLC40A1).

Figure 1.

Effect of ferroportin variants on intracellular iron accumulation, cell-surface expression and iron export. (A) HEK293T cells were transiently transfected with CMV-regulated plasmids containing no ferroportin sequence (no-SLC40A1), full-length WT human ferroportin cDNA or a mutated sequence in the presence of 1 mg/ml holotransferrin. After 48 h, intracellular ferritin levels were determined by ELISA and normalized to the total protein concentration. Error bars represent the SD of three independent experiments (performed in triplicate). (B) HEK293T cells were transiently co-transfected with plasmids encoding either a V5-tagged ferroportin protein (WT or variant) or a V5-tagged HLA-A protein. HLA-A was used as a control and as a standard for normalization, being a cell-surface protein with no known role in iron metabolism. At 24 h after transfection, cell-surface proteins were selectively purified and analyzed by western blotting using a peroxidase-conjugated mouse anti-V5 antibody. Densitometric scans of SLC40A1 levels (normalized to HLA-A) are shown in the lower part of the figure. The error bars represent the standard deviation of three independent experiments. (C) HEK293T cells were grown in 20 µg/ml 55Fe-transferrin for 24 h before being washed and transiently transfected with WT or mutated SLC40A1-V5 expression plasmids. After 15 h, cells were washed and then serum starved for up to 36 h. The 55Fe exported into the supernatant was collected at various time points. Data are presented as percentage cellular radioactivity at time zero. Each point represents the mean SD; n = 3 in each group. The data are representative of three separate experiments (see below). (D) The graph shows the percentage of 55Fe collected from the supernatant of SLC40A1-transfected and no-SLC40A1-transfected cells at 36 h. The error bars represent the SD of three independent experiments (performed in triplicate; n = 9). P-values were calculated using Student's t-test. *P < 0.01 and **P < 0.001, compared with WT SLC40A1.

Figure 1.

Effect of ferroportin variants on intracellular iron accumulation, cell-surface expression and iron export. (A) HEK293T cells were transiently transfected with CMV-regulated plasmids containing no ferroportin sequence (no-SLC40A1), full-length WT human ferroportin cDNA or a mutated sequence in the presence of 1 mg/ml holotransferrin. After 48 h, intracellular ferritin levels were determined by ELISA and normalized to the total protein concentration. Error bars represent the SD of three independent experiments (performed in triplicate). (B) HEK293T cells were transiently co-transfected with plasmids encoding either a V5-tagged ferroportin protein (WT or variant) or a V5-tagged HLA-A protein. HLA-A was used as a control and as a standard for normalization, being a cell-surface protein with no known role in iron metabolism. At 24 h after transfection, cell-surface proteins were selectively purified and analyzed by western blotting using a peroxidase-conjugated mouse anti-V5 antibody. Densitometric scans of SLC40A1 levels (normalized to HLA-A) are shown in the lower part of the figure. The error bars represent the standard deviation of three independent experiments. (C) HEK293T cells were grown in 20 µg/ml 55Fe-transferrin for 24 h before being washed and transiently transfected with WT or mutated SLC40A1-V5 expression plasmids. After 15 h, cells were washed and then serum starved for up to 36 h. The 55Fe exported into the supernatant was collected at various time points. Data are presented as percentage cellular radioactivity at time zero. Each point represents the mean SD; n = 3 in each group. The data are representative of three separate experiments (see below). (D) The graph shows the percentage of 55Fe collected from the supernatant of SLC40A1-transfected and no-SLC40A1-transfected cells at 36 h. The error bars represent the SD of three independent experiments (performed in triplicate; n = 9). P-values were calculated using Student's t-test. *P < 0.01 and **P < 0.001, compared with WT SLC40A1.

Secondly, the cellular location of the nine iron-export defective variants was examined. Cultured HEK293T cells were transiently cotransfected with plasmids encoding either full-length human ferroportin or human leukocyte antigen (HLA)-A, both proteins being fused to a V5 epitope tag. Twenty-four hours after transfection, proteins were biotinylated, purified and analyzed by western blot and densitometry. HLA-A was used as control and standard for normalization, being a cell-surface protein with no known role in iron metabolism. As expected, WT SLC40A1 expression resulted in cell-surface localization, whereas the p.Ala77Asp mutant showed markedly reduced localization (Fig. 1B). Eight of the nine tested variants (p.Val162del, p.Gly80Ser, p.Asp157Gly, p.Asp157Tyr, p.Asp181Val, p.Leu233Pro, p.Gly490Asp and p.Gly490Ser) were also found to cause ferroportin mislocalization; that is, 60% or less of the protein was detected at the cell surface as compared with WT SLC40A1. Only the p.Arg88Gly variant properly reached the plasma membrane (101%).

The impact of the ferroportin p.Arg88Gly variant on the iron-export function was specifically re-investigated using radioactively labeled iron, as previously reported (15). Kinetics experiments revealed that the p.Arg88Gly variant was not able to export 55Fe in amounts comparable with WT ferroportin (Fig. 1C) but was more active than the p.Ala77Asp control; Student's t-test highlighted significant differences between both variants and WT ferroportin (Fig. 1D; P < 0.01 and 0.001, respectively) and between the two variants (P < 0.001).

Identification of a new gain-of-function mutation: p.Asp540Asn amino acid substitution

To investigate whether the p.Ile180Thr, p.Thr230Asn, p.Glu248His, p.Met266Thr, p.Leu345Phe, p.Ile351Val, p.Pro443Leu, p.Asp504Asn and p.Arg561Gly variants could modify response to hepcidin, transiently transfected HEK293T cells were initially treated with the full-length 25 amino acid mature hepcidin peptide for 16 h, and cell-surface ferroportin levels were assessed by western blot and densitometry. The p.Cys326Tyr mutant was used as positive control, completely ablating hepcidin binding and thus preventing ferroportin/hepcidin complex internalization (22). Adding hepcidin to cells expressing WT SLC40A1 resulted in the disappearance of ferroportin from the plasma membrane, while the p.Cys326Tyr mutant induced no change (Figs. 2A and B). The western blot pattern of the p.Asp504Asn variant, unlike the other eight tested variants, was similar to that of the p.Cys326Tyr mutant.

Figure 2.

Effect of ferroportin variants on hepcidin-induced internalization. (A and B) HEK293T cells were transiently co-transfected with CMV-regulated plasmids expressing HLA(A)-V5 and either WT SLC40A1-V5 or SLC40A1-V5 variants. At 16 h posttransfection, the cells were incubated in the presence or absence of hepcidin (0.5 µm; Biochem) for 3 h. Plasma membrane proteins were purified and analyzed by western blotting and densitometry. The data are expressed as percentage ferroportin in cells not treated by hepcidin, according to the formula 100 × (SLC40A1 hepcidin/SLC40A1 + hepcidin). Error bars are the SD of three independent experiments. (C) HEK293T cells expressing WT SLC40A1-V5 or variant SLC40A1-V5 were treated with [125I]-hepcidin and cell-associated radioactivity was measured. Each bar represents the average of three independent experiments (performed in triplicate). Data were normalized to the amount of radioactivity bound to WT SLC40A1-V5-expressing cells, and the amount bound to untransfected cells was subtracted as background for each point. P-values were calculated using Student's t-test. **P < 0.001 and ***P < 0.0001, compared with WT SLC40A1.

Figure 2.

Effect of ferroportin variants on hepcidin-induced internalization. (A and B) HEK293T cells were transiently co-transfected with CMV-regulated plasmids expressing HLA(A)-V5 and either WT SLC40A1-V5 or SLC40A1-V5 variants. At 16 h posttransfection, the cells were incubated in the presence or absence of hepcidin (0.5 µm; Biochem) for 3 h. Plasma membrane proteins were purified and analyzed by western blotting and densitometry. The data are expressed as percentage ferroportin in cells not treated by hepcidin, according to the formula 100 × (SLC40A1 hepcidin/SLC40A1 + hepcidin). Error bars are the SD of three independent experiments. (C) HEK293T cells expressing WT SLC40A1-V5 or variant SLC40A1-V5 were treated with [125I]-hepcidin and cell-associated radioactivity was measured. Each bar represents the average of three independent experiments (performed in triplicate). Data were normalized to the amount of radioactivity bound to WT SLC40A1-V5-expressing cells, and the amount bound to untransfected cells was subtracted as background for each point. P-values were calculated using Student's t-test. **P < 0.001 and ***P < 0.0001, compared with WT SLC40A1.

To confirm these results, radiolabeled hepcidin was added to ferroportin-overexpressing cells. As expected, after 3 h incubation with [125I]-hepcidin, the WT SLC40A1-V5 fusion protein was efficiently internalized, showing the highest level of intracellular radioactivity (Fig. 2C) whereas that of cells expressing the p.Cys326Tyr mutant was indistinguishable from the level in cells transfected with the commercial pcDNA3.1-V5-His vector (no-SLC40A1). Consistent with our previous findings, a marked decrease in [125I]-hepcidin accumulation was also observed in cells expressing the p.Asp504Asn variant (P < 0.001), whereas the p.Ile180Thr, p.Thr230Asn, p.Glu248His, p.Met266Thr, p.Leu345Phe, p.Ile351Val, p.Pro443Leu and p.Arg561Gly variants were found to increase intracellular [125I]-hepcidin in amounts similar to WT SLC40A1.

Structure–function analysis

We recently built a 3D model of ferroportin based on homology with the crystal structure of a bacterial protein (EmrD) of the major facilitator superfamily (MFS) (15). Combined with the results presented in Figures 1 and 2, this 3D model provided a novel opportunity for structure–function analysis.

As illustrated in Figure 3 and Supplementary Material, Figure S1 and discussed in detail below, a clear distribution into separate functional subclasses was observed:

  • Loss-of-function mutations (Fig. 3A). Most loss-of-function mutations were located at the cytoplasmic end of the membrane-spanning domains.

    • – Mutations affecting intracellular trafficking and cell-surface localization. The effects of the p.Val162del, p.Asp157Tyr and p.Gly490Asp mutants (red in Fig. 3A) on the structure of ferroportin were previously predicted (15). However, a novel experimental MFS 3D structure (YajR), in an outward-facing conformation (23), recently provided important new information on the probable role of p.Asp157, which could not be anticipated from previously observed conformations: p.Asp157, located in helix H4 and highly conserved in the MSF family, is predicted to be involved in a charge-relay system with the other two highly conserved amino acids (p.Asp84 and p.Arg88) of the MSF-specific or conserved A motif (between helices H2 and H3). This system, together with inter-domain charge-helix dipole interaction, probably plays a critical role in the stability of the outward-facing conformation and in regulating conformational transition. The p.Asp157Tyr substitution might thus play a role in protein stability, as demonstrated for YajR and other MSF proteins (23). p.Gly80 is located at the very end of helix H2 (orange in Fig. 3A), at a position that is strictly conserved not only across ferroportin proteins from different species but also across the various structures of MFS transporters (Supplementary Material, Figure S1). p.Gly80 bends helix H2, which, like the irregularities observed in other cavity-lining helices, is probably important to the structural flexibility of the protein and/or the tight helix–helix packing observed in the outward-facing conformation of MSF transporters (23); pGly80Ser substitution may therefore destabilize this local conformation. p.Leu233 is included in the long intracellular segment IS4 linking helices H6 and H7 (orange in Fig. 3A, not modeled). It should be noted that p.Leu233 constitutes part of a small cluster including other hydrophobic amino acids, which is typical of an extended structure and may form a short β-sheet with another predicted extended structure encompassing p.Val160-p.Val161-p.Val162. This local structure, which is potentially disturbed by Leu233Pro substitution, may therefore contribute to a specific pore-entry architecture. Finally, p.Asp181 is located within helix H5, at a position that is highly conserved across ferroportin proteins (orange in Fig. 3A). The Asp side-chain is oriented not toward the pore but toward the lipid bilayer, but is able to form a stabilizing salt bridge with the preceding p.Arg178 (top inset in Fig. 3A), which is also highly conserved and may form a hydrogen bond with p.Asn174; p.Asp181Val substitution may thus disturb the insertion of helix H5 in the lipid bilayer.

    • – Mutations preventing iron egress. p.Arg88, which is the only amino acid for which substitution resulting in a loss-of-function mutation does not impair cellular localization of the protein, is located at the extreme N-terminus of helix H3, at the end of the IS2 loop-linking helices H2 and H3 (pink in Fig. 3A). Following our model, p.Arg88 is involved in a tightly linked network of interactions stabilizing not only the IS2 loop between helices H2 and H3 (interaction with H2 p.Trp82) but also the IS4 loop between helices H6 and H7 (interactions with p.Met216 and p.Tyr220) (bottom inset in Fig. 3A). Thus, it is likely that substitution of p.Arg88 by a glycine residue has no structural impact, allowing this variant to be correctly folded and addressed to the plasma membrane (Fig. 1B). Rather, the arginine residue may indirectly play a functional role by maintaining the local conformation of loops on the intracellular side. Notably, p.Arg88 is located in the immediate vicinity of p.Ile152 and p.Asn174, at the intracellular end of the transmembrane domain, highlighting the importance of this region for iron egress (bottom inset Fig. 3A). Following our model, the side-chain of p.Ile152, at the intracellular end of helix H4, is oriented toward the pore, whereas that of p.Asn174, at the intracellular end of helix H5, points towards the lipid bilayer but may, like p.Asp181, form a hydrogen bond with p.Arg178.

  • The newly identified p.Asp504Asn gain-of-function mutation. p.Asp504 is located within a large loop (ES6) that was not modeled due to its divergence from the MFS template structure (light green in Fig. 3A). However, it should be noted that, as previously shown for p.His507 (15), p.Asp504 is located at the entrance to the iron transport channel, not far from the critical p.Cys326 residue (22, 24), which is consistent with the hepcidin resistance reported for both the p.Asp504Asn (Fig. 2) and p.His507Arg mutations (25).

  • Neutral substitutions. None of the 8 SLC40A1 substitutions that apparently do not alter the ferroportin iron export function or down-regulation by hepcidin are predicted to cause structural disturbance. As shown in Figure 3B and Supplementary Material, Figure S1, the p.Thr230, p.Glu248, p.Met266, p.Pro443 and p.Arg561 residues are included in loops linking the transmembrane helices or in the C-terminal tail (p.Arg561) at some distance from the ferroportin pore (light blue in Fig. 3A). These positions are therefore likely not critical for the folding or iron export function of ferroportin. The other three residues (p.Ile180, p.Leu345 and p.Ile351) are located on the hydrophobic side of transmembrane helices, oriented toward the lipid region (blue on Fig. 3B). The hydrophobic character is, however, conserved at these three positions for the observed mutations, which should therefore not be structurally damaging.

Figure 3.

Structural analysis of variants. The variants were localized within a ribbon representation (lateral view) of the previously reported 3D structural model of human ferroportin (15). Non-modeled loops are indicated by dotted lines. (A) Loss-of-function (red, orange and pink) and gain-of-function (green) mutations. Pink indicates amino acid positions at which variants did not impair membrane localization, in contrast to red and orange positions. The two insets focus on the environment of p.Asp181/pAsn174 (top) and p.Arg88 (bottom). (B) Positions of amino acids showing neutral variations. The inset provides an orthogonal view of the 3D model, viewed from the extracellular side.

Figure 3.

Structural analysis of variants. The variants were localized within a ribbon representation (lateral view) of the previously reported 3D structural model of human ferroportin (15). Non-modeled loops are indicated by dotted lines. (A) Loss-of-function (red, orange and pink) and gain-of-function (green) mutations. Pink indicates amino acid positions at which variants did not impair membrane localization, in contrast to red and orange positions. The two insets focus on the environment of p.Asp181/pAsn174 (top) and p.Arg88 (bottom). (B) Positions of amino acids showing neutral variations. The inset provides an orthogonal view of the 3D model, viewed from the extracellular side.

Combined use of in silico and in vitro splicing assays for interpretation of the 8 SLC40A1 variations with no impact at protein level

Analysis was extended by assessing the effects of the p.Ile180Thr, p.Thr230Asn, p.Glu248His, p.Met266Thr, p.Leu345Phe, p.Ile351Val, p.Pro443Leu and p.Arg561Gly variations on pre-mRNA splicing. Different algorithms and SLC40A1 minigenes were used; no abnormalities were found (see Supplementary Material, Results and Fig. S2). Results of in vitro investigations for the 18 ferroportin variants are summarized in Table 2.

Table 2.

Classification of the 18 ferroportin substitutions into loss-of-function versus gain-of-function mutations and neutral variants

Variation Functional category
 
Loss of function
 
Gain of function Neutral 
Abnormal cell-surface routing Reduced iron export Total resistance to hepcidin inhibition  
p.Gly80Ser   
p.Arg88Gly    
p.Asp157Gly   
p.Asp157Tyr   
p.Val162del   
p.Ile180Thr    
p.Asp181Val   
p.Thr230Asn    
p.Leu233Pro   
p.Gln248His    
p.Met266Thr    
p.Leu345Phe    
p.Ile351Val    
p.Pro443Leu    
p.Gly490Asp   
p.Gly490Ser   
p.Asp504Asn    
p.Arg561Gly    
Variation Functional category
 
Loss of function
 
Gain of function Neutral 
Abnormal cell-surface routing Reduced iron export Total resistance to hepcidin inhibition  
p.Gly80Ser   
p.Arg88Gly    
p.Asp157Gly   
p.Asp157Tyr   
p.Val162del   
p.Ile180Thr    
p.Asp181Val   
p.Thr230Asn    
p.Leu233Pro   
p.Gln248His    
p.Met266Thr    
p.Leu345Phe    
p.Ile351Val    
p.Pro443Leu    
p.Gly490Asp   
p.Gly490Ser   
p.Asp504Asn    
p.Arg561Gly    

Genotype–phenotype correlations

Familial recurrence of iron overload is an alternative means of assessing the clinical relevance of SLC40A1 variations. Variations detected in at least two relatives with hyperferritinemia (defined as serum ferritin >300 µg/l in males and >200 µg/l in females) were investigated: i.e. 8 of the 10 SLC40A1 variations with clear functional impact; exceptions were the p.Asp157Tyr and p.Gly490Ser loss-of-function mutations (Table 1). No family discordances were detected, apart from the well-known p.Val162del mutation, which was associated with distinct phenotypes in one sib pair: the index case (FRA-02-009) displayed significant hyperferritinemia (2578 µg/l), while her younger brother (FRA-02-008) showed normal iron values.

The p.Asp504Asn gain-of-function mutation was identified in two brothers with hyperferritinemia (>1500 µg/l) and was found to increase transferrin saturation levels (>85%). Abdominal MRI was available for the index case (Supplementary Material, Figure S3) and showed a marked reduction in signal intensity in the liver, consistent with iron overload (HIC estimated at 320 μmol/g). No signal reduction was observed in spleen or bone marrow.

In total, 34 patients harboring a loss-of-function mutation were identified: 14 males and 20 females. In most of these patients, hyperferritinemia contrasted with normal or low transferrin saturation levels. Serum ferritin concentrations correlated significantly and positively with age in both males and females (Supplementary Material, Fig. S4). Only five patients (two males and three females) displayed >45% transferrin saturation. It is noteworthy that four of these patients presented particularly high serum ferritin concentrations (hyperferritinemia >4000 µg/l), and that these four patients were among the oldest: the two males were diagnosed after 42 years of age, which was the median age in the group of 14 males carrying a loss-of-function mutation, and the two females after 52 years of age, which was the 75th age percentile in the group of 20 females carrying a loss-of-function mutation. MRI, performed in the 52-year-old female positive for the p.Leu233Pro mutation, found mixed iron overload with evidence of iron deposition in liver (HIC >300 μmol/g), spleen and bone marrow.

All of the neutral SLC40A1 variations came from a single patient. All of these patients presented moderate-to-high serum ferritin concentrations (620–1910 µg/l) and normal transferrin saturation levels. Whether hyperferritinemia was accompanied by iron overload was unknown, as none of the patients underwent liver biopsy or MRI, and no information was available on therapeutic phlebotomy. One 40-year-old man, positive for the new SLC40A1 p.Met266Thr substitution, declared heavy alcohol consumption and at diagnosis displayed a mean corpuscular volume of 101 fl, a gamma-glutamyl transpeptidase level of 582 IU/l and a ratio of serum aspartate aminotransferase to serum alanine aminotransferase of 1.4. No other obvious cause of acquired hyperferritinemia was found in patients with polymorphic SLC40A1 variants. Interestingly, males who harbored a neutral variant displayed significantly lower ferritin values than those who harbored a loss-of-function mutation (P = 0.0054; Fig. 4).

Figure 4.

Serum ferritin levels in male carriers of a loss-of-function mutation or a non-functional polymorphism at the SLC40A1 locus. Box and whisker plots showing median and interquartile range in the box (50th, 25th and 75th percentiles) and minimum to maximum range (whiskers). P-value was calculated using Mood's median test – Pearson's χ2-test.

Figure 4.

Serum ferritin levels in male carriers of a loss-of-function mutation or a non-functional polymorphism at the SLC40A1 locus. Box and whisker plots showing median and interquartile range in the box (50th, 25th and 75th percentiles) and minimum to maximum range (whiskers). P-value was calculated using Mood's median test – Pearson's χ2-test.

DISCUSSION

Lack of penetrance and the influence of gender and co-morbid factors have been postulated to explain part of the clinical heterogeneity attributed to SLC40A1 variations (1, 2), in addition to the well-recognized dichotomy between loss-of-function and gain-of-function mutations (1, 2). Although it seems reasonable that SLC40A1 mutations may not always be phenotype determinants and that other factors modulate the severity of iron overload, it is important to bear in mind that evidence for the causality of many non-synonymous changes and small deletions is at best circumstantial. Regardless of the conditions underlying phenotypic heterogeneity, there is an unmet need for more narrowly defined phenotypes and more effective functional characterization strategies.

The present study classified 9 SLC40A1 variants as loss-of-function mutations (Fig. 1). With few exceptions, the disease associated with these mutations involved isolated hyperferritinemia (Table 1). In spite of markedly elevated ferritin levels in the majority of patients (hyperferritinemia>10 001 000 µg/l in 28 of 34 patients), the clinical manifestations were rather moderate. They included: fatigue (n = 13), joint pain (n = 8), hepatomegaly (n = 1) and type 2 diabetes (n = 1). Hepatic iron overload was found on MRI in five patients, with no sign of liver fibrosis. Serum ferritin concentrations increased with age, independently of gender (Supplementary Material, Figure S4). Interestingly, the patients showing elevated transferrin saturation were among the oldest. MRI demonstrated a mixed pattern of iron accumulation in parenchymal and non-parenchymal cells in a 52-year-old woman with serum ferritin >4000 µg/l and 76% transferrin saturation. Overall, these findings support the idea that hemochromatosis type 4A consists in progressive iron overload with biochemical and histological hallmarks that are clearly different from those of classical HFE-hemochromatosis (4, 13).

Particular attention was paid to the p.Arg88Gly amino acid substitution: of the 10 loss-of-function mutations (including the p.Ala77Asp control), only the p.Arg88Gly ferroportin mutant was defective for iron egress while being normally addressed to the cell surface (Fig. 1B–D). In the literature, such a functional defect has been reported only for the p.Ile152Phe and p.Asn174Ile clinical mutations (15, 26, 27). Accordingly, the present structural predictions suggest that p.Arg88 is located in the immediate vicinity of the p.Ile152 and p.Asn174 residues in a region that might be directly involved in iron egress (Fig. 3B). Deciphering the precise function of the p.Arg88, p.Ile152, p.Asp174 and neighboring residues may improve understanding of the ferroportin iron export mechanism.

Apart from the three cases in the present study (Table 1), the p.Arg88Gly mutation was also reported in four French patients: three family members and one apparently unrelated individual (1, 28). It was systematically associated with classical ferroportin disease (hemochromatosis type 4A), although there were potential confounding factors (metabolic syndrome, excessive alcohol intake) in two patients (11). p.Arg88Gly substitution should thus be considered recurrent in the French population. Another mutation, causing replacement of glycine by the acidic polar residue aspartic acid at position 490 (p.Gly490Asp), appears to be particularly frequent in patients originating from the Reunion Island (nine patients in the present study; Table 1). It would be interesting to perform a population genetics study on this French island located in the Indian Ocean, which is likely to represent a genetic isolate.

Furthermore, a typical hemochromatosis phenotype was found in two brothers carrying the new p.Asp504Asn gain-of-function mutation (Table 1; Supplementary Material, Fig. S3). In vitro assessment suggested total resistance to hepcidin inhibition (Fig. 2), as previously demonstrated for the p.Cys326Tyr and p.Cys326Ser disease-causing mutations (22, 29). According to the present 3D model of ferroportin, aspartic acid 504 is part of an extracellular loop (ES6) at the entrance to the iron export channel (Fig. 3A). We previously reported that this extracellular loop could not be accurately modeled due to its size and relative divergence from the MFS structure template (15). Consequently, the distance between aspartic acid 504 and cysteine 326, which has proved critical for the docking of hepcidin with ferroportin (22, 24), is difficult to estimate. Similar clinical and functional observations were, however, made by Mayr et al. reporting p.His507Arg substitution in a third-generation Caucasian pedigree (25); testing hepcidin activity in presence of synthetic ferroportin peptides, they argued that the strongly conserved histidine 507 residue was directly involved in the interaction of hepcidin with ferroportin. A recent study on small peptides mimicking hepcidin activity showed that the hepcidin-binding domain (HBD) of ferroportin is larger than previously thought (24), extending to amino acids adjacent to the region between glycine 323 and serine 343 (30); the question of whether it also extends to residues outside the region between helix 7 and helix 8 and, particularly, to the extracellular loop containing the p.Arg504 and p.His507 residues (ES6), clearly remains open.

Complementary in vitro and in silico studies demonstrated that eight missense variations had no noticeable effects on ferroportin function and its interaction with hepcidin (Figs. 1–3) or on SLC40A1 pre-mRNA splicing (Supplementary Material, Figure S2). These non-synonymous ferroportin variants, identified in eight single patients showing moderate increases in serum ferritin and normal transferrin saturation (Table 1; Fig. 4) most likely corresponded to neutral polymorphisms. It should be borne in mind in this regard that elevated serum ferritin is frequent in clinical practice and can be caused by a variety of conditions that do not necessarily result in excess iron. Furthermore, in many patients, hyperferritinemia remains unexplained after assessment (31). The present study excluded mutations in the other four hemochromatosis genes (HFE, HJV, HAMP and TFR2) and in the l-ferritin-encoding gene, which has recently been associated with elevated levels of glycosylated serum ferritin without iron overload (32). On the other hand, alcohol abuse emerged as the most probable cause of hyperferritinemia in the patient carrying the new p.Met266Thr substitution. C-reactive protein levels were found to be normal in five other patients. There was not enough information to ascertain metabolic abnormalities or other obvious underlying causes of hyperferritinemia.

Another possible explanation, however, is that hypomorphic mutations were overlooked under the present experimental conditions. A supporting argument might be that the eight missense mutations are infrequent in the French population, so that the likelihood of any of them contributing to false diagnosis of hyperferritinemia can be expected to be low. On the other hand, the present control population was representative only of patients originating from the north-western part of France, likely underestimating the frequency of SLC40A1 variants in patients with different genetic histories; e.g. p.Glu248His substitution has a minor allele frequency of 5.7–6.6% in subjects of African-American ancestry (Supplementary Material, Table S1) and is thought to be even more frequent in native Africans (33).

It should be noted that the common p.Glu248His substitution is associated with a tendency for elevated serum ferritin in various African-American and native African groups (34–37), whereas the risk of iron overload associated with p.Glu248His has not been established (33, 36). Consistently with both the present observations and previous in vitro findings (18, 29, 38, 39), it has been suggested that the p.Glu248His polymorphism causes iron overload only when associated with other genetic and/or environmental factors (33); the p.Glu248His substitution is thereby supposed to have slight functional impact and not to be a major genetic determinant of iron excess. Whether or not this applies to the apparently less frequent variants identified in the present study, classified as neutral, will be hard to determine.

In summary, the present findings show that the functional consequences of missense mutations in the gene-encoding ferroportin are crucial for interpreting the observed clinical heterogeneity, further highlighting the need to clarify the various causal hypotheses found in the literature. The present study also demonstrates that interpreting missense mutations at the SLC40A1 locus will increase knowledge of structure–function relationships in ferroportin.

MATERIALS AND METHODS

Patients

The present study involved six laboratories in the French network for the molecular diagnosis of inherited iron overload disorders (University Hospitals of Amiens, Paris-Bichat, Bordeaux, Brest, Créteil and Marseille). Biological, clinical, socio-demographic, lifestyle and molecular genetic data were collected for all consenting patients with suggested molecular diagnosis of ferroportin disease or hemochromatosis type 4. Phenotypes were assessed retrospectively, based on a questionnaire completed by the physicians who had initially ordered molecular testing. When available, familial and abdominal MRI data were also collected. All 44 patients with complete molecular genetic data (i.e. Sanger sequencing of the HFE, HJV, HAMP, TFR2, SLC40A1 and FTL genes) were recruited over the period from July 2011 to July 2013. Informed consent for molecular studies was obtained from all patients, in accordance with the Declaration of Helsinki, and, in line with French ethical guidelines, the Clinical Research Ethics Committee of the University Hospital of Brest approved the study on October 25, 2010.

In vitro experiments and molecular modeling

In vitro investigations and structural predictions were performed as previously described (36–39) and are presented in detail in the Supplementary Material, Documents.

SUPPLEMENTARY MATERIAL

Supplementary Material is available at HMG online.

FUNDING

This work was supported by the Programme Hospitalier de Recherche Clinique French Hospital Clinical Research Program (PHRC National 2009; Brest University Hospital – UF0857). The authors thank Dr Virginie Scotet for statistical support, and Dr Jian-Min Chen and Pr. Cédric Le Marechal for critical reading.

Conflict of Interest statement. None declared.

REFERENCES

1
Le Lan
C.
Mosser
A.
Ropert
M.
Detivaud
L.
Loustaud-Ratti
V.
Vital-Durand
D.
Roget
L.
Bardou-Jacquet
E.
Turlin
B.
David
V.
, et al.  . 
Sex and acquired cofactors determine phenotypes of ferroportin disease
Gastroenterology
 , 
2011
, vol. 
140
 (pg. 
1199
-
1207
)
2
Pelucchi
S.
Mariani
R.
Salvioni
A.
Bonfadini
S.
Riva
A.
Bertola
F.
Trombini
P.
Piperno
A.
Novel mutations of the ferroportin gene (SLC40A1): analysis of 56 consecutive patients with unexplained iron overload
Clin. Genet.
 , 
2008
, vol. 
73
 (pg. 
171
-
178
)
3
Pietrangelo
A.
Hereditary hemochromatosis: pathogenesis, diagnosis, and treatment
Gastroenterology
 , 
2010
, vol. 
139
 (pg. 
393
-
408
)
4
Pietrangelo
A.
The ferroportin disease
Blood Cells. Mol. Dis.
 , 
2004
, vol. 
32
 (pg. 
131
-
138
)
5
Donovan
A.
Lima
C.A.
Pinkus
J.L.
Pinkus
G.S.
Zon
L.I.
Robine
S.
Andrews
N.C.
The iron exporter ferroportin/Slc40a1 is essential for iron homeostasis
Cell Metab.
 , 
2005
, vol. 
1
 (pg. 
191
-
200
)
6
Hentze
M.W.
Muckenthaler
M.U.
Galy
B.
Camaschella
C.
Two to tango: regulation of Mammalian iron metabolism
Cell
 , 
2010
, vol. 
142
 (pg. 
24
-
38
)
7
Ganz
T.
Nemeth
E.
The hepcidin-ferroportin system as a therapeutic target in anemias and iron overload disorders
Hematol. Am. Soc. Hematol. Educ. Program
 , 
2011
, vol. 
2011
 (pg. 
538
-
542
)
8
Pietrangelo
A.
Hepcidin in human iron disorders: therapeutic implications
J. Hepatol.
 , 
2011
, vol. 
54
 (pg. 
173
-
181
)
9
Camaschella
C.
Iron and hepcidin: a story of recycling and balance
Hematol. Am. Soc. Hematol. Educ. Program
 , 
2013
, vol. 
2013
 (pg. 
1
-
8
)
10
Anderson
G.J.
McLaren
G.D.
Iron Physiology and Pathophysiology in Humans
 , 
2012
New York
Humana Press
11
Détivaud
L.
Island
M.-L.
Jouanolle
A.-M.
Ropert
M.
Bardou-Jacquet
E.
Le Lan
C.
Mosser
A.
Leroyer
P.
Deugnier
Y.
David
V.
, et al.  . 
Ferroportin diseases: functional studies, a link between genetic and clinical phenotype
Hum. Mutat.
 , 
2013
, vol. 
34
 (pg. 
1529
-
1536
)
12
Le Gac
G.
Férec
C.
The molecular genetics of haemochromatosis
Eur. J. Hum. Genet.
 , 
2005
, vol. 
13
 (pg. 
1172
-
1185
)
13
Mayr
R.
Janecke
A.R.
Schranz
M.
Griffiths
W.J.H.
Vogel
W.
Pietrangelo
A.
Zoller
H.
Ferroportin disease: a systematic meta-analysis of clinical and molecular findings
J. Hepatol.
 , 
2010
, vol. 
53
 (pg. 
941
-
949
)
14
De Domenico
I.
Ward
D.M.
Nemeth
E.
Vaughn
M.B.
Musci
G.
Ganz
T.
Kaplan
J.
The molecular basis of ferroportin-linked hemochromatosis
Proc. Natl. Acad. Sci. USA
 , 
2005
, vol. 
102
 (pg. 
8955
-
8960
)
15
Le Gac
G.
Ka
C.
Joubrel
R.
Gourlaouen
I.
Lehn
P.
Mornon
J.-P.
Férec
C.
Callebaut
I.
Structure-function analysis of the human ferroportin iron exporter (SLC40A1): effect of hemochromatosis type 4 disease mutations and identification of critical residues
Hum. Mutat.
 , 
2013
, vol. 
34
 (pg. 
1371
-
1380
)
16
Gonçalves
A.S.
Muzeau
F.
Blaybel
R.
Hetet
G.
Driss
F.
Delaby
C.
Canonne-Hergaux
F.
Beaumont
C.
Wild-type and mutant ferroportins do not form oligomers in transfected cells
Biochem. J.
 , 
2006
, vol. 
396
 (pg. 
265
-
275
)
17
Rice
A.E.
Mendez
M.J.
Hokanson
C.A.
Rees
D.C.
Björkman
P.J.
Investigation of the biophysical and cell biological properties of ferroportin, a multipass integral membrane protein iron exporter
J. Mol. Biol.
 , 
2009
, vol. 
386
 (pg. 
717
-
732
)
18
Schimanski
L.M.
Drakesmith
H.
Merryweather-Clarke
A.T.
Viprakasit
V.
Edwards
J.P.
Sweetland
E.
Bastin
J.M.
Cowley
D.
Chinthammitr
Y.
Robson
K.J.H.
, et al.  . 
In vitro functional analysis of human ferroportin (FPN) and hemochromatosis-associated FPN mutations
Blood
 , 
2005
, vol. 
105
 (pg. 
4096
-
4102
)
19
Wallace
D.F.
Harris
J.M.
Subramaniam
V.N.
Functional analysis and theoretical modeling of ferroportin reveals clustering of mutations according to phenotype
Am. J. Physiol. Cell Physiol.
 , 
2010
, vol. 
298
 (pg. 
75
-
84
)
20
Corradini
E.
Ferrara
F.
Pollicino
T.
Vegetti
A.
Abbati
G.L.
Losi
L.
Raimondo
G.
Pietrangelo
A.
Disease progression and liver cancer in the ferroportin disease
Gut
 , 
2007
, vol. 
56
 (pg. 
1030
-
1032
)
21
Rosmorduc
O.
Wendum
D.
Arrivé
L.
Elnaggar
A.
Ennibi
K.
Hannoun
L.
Charlotte
F.
Grangé
J.-D.
Poupon
R.
Phenotypic expression of ferroportin disease in a family with the N144H mutation
Gastroentérologie Clin. Biol.
 , 
2008
, vol. 
32
 (pg. 
321
-
327
)
22
Fernandes
A.
Preza
G.C.
Phung
Y.
De Domenico
I.
Kaplan
J.
Ganz
T.
Nemeth
E.
The molecular basis of hepcidin-resistant hereditary hemochromatosis
Blood
 , 
2009
, vol. 
114
 (pg. 
437
-
443
)
23
Jiang
D.
Zhao
Y.
Wang
X.
Fan
J.
Heng
J.
Liu
X.
Feng
W.
Kang
X.
Huang
B.
Liu
J.
, et al.  . 
Structure of the YajR transporter suggests a transport mechanism based on the conserved motif A
Proc. Natl. Acad. Sci. USA
 , 
2013
, vol. 
110
 (pg. 
14664
-
14669
)
24
Preza
G.C.
Ruchala
P.
Pinon
R.
Ramos
E.
Qiao
B.
Peralta
M.A.
Sharma
S.
Waring
A.
Ganz
T.
Nemeth
E.
Minihepcidins are rationally designed small peptides that mimic hepcidin activity in mice and may be useful for the treatment of iron overload
J. Clin. Invest.
 , 
2011
, vol. 
121
 (pg. 
4880
-
4888
)
25
Mayr
R.
Griffiths
W.J.H.
Hermann
M.
McFarlane
I.
Halsall
D.J.
Finkenstedt
A.
Douds
A.
Davies
S.E.
Janecke
A.R.
Vogel
W.
, et al.  . 
Identification of mutations in SLC40A1 that affect ferroportin function and phenotype of human ferroportin iron overload
Gastroenterology
 , 
2011
, vol. 
140
 (pg. 
2056
-
2063
)
26
De Domenico
I.
McVey Ward
D.
Nemeth
E.
Ganz
T.
Corradini
E.
Ferrara
F.
Musci
G.
Pietrangelo
A.
Kaplan
J.
Molecular and clinical correlates in iron overload associated with mutations in ferroportin
Haematologica
 , 
2006
, vol. 
91
 (pg. 
1092
-
1095
)
27
Girelli
D.
De Domenico
I.
Bozzini
C.
Campostrini
N.
Busti
F.
Castagna
A.
Soriani
N.
Cremonesi
L.
Ferrari
M.
Colombari
R.
, et al.  . 
Clinical, pathological, and molecular correlates in ferroportin disease: a study of two novel mutations
J. Hepatol.
 , 
2008
, vol. 
49
 (pg. 
664
-
671
)
28
Cunat
S.
Giansily-Blaizot
M.
Bismuth
M.
Blanc
F.
Dereure
O.
Larrey
D.
Quellec
A.L.
Pouderoux
P.
Rose
C.
Raingeard
I.
, et al.  . 
Global sequencing approach for characterizing the molecular background of hereditary iron disorders
Clin. Chem.
 , 
2007
, vol. 
53
 (pg. 
2060
-
2069
)
29
Drakesmith
H.
Schimanski
L.M.
Ormerod
E.
Merryweather-Clarke
A.T.
Viprakasit
V.
Edwards
J.P.
Sweetland
E.
Bastin
J.M.
Cowley
D.
Chinthammitr
Y.
, et al.  . 
Resistance to hepcidin is conferred by hemochromatosis-associated mutations of ferroportin
Blood
 , 
2005
, vol. 
106
 (pg. 
1092
-
1097
)
30
De Domenico
I.
Nemeth
E.
Nelson
J.M.
Phillips
J.D.
Ajioka
R.S.
Kay
M.S.
Kushner
J.P.
Ganz
T.
Ward
D.M.
Kaplan
J.
The hepcidin-binding site on ferroportin is evolutionarily conserved
Cell Metab.
 , 
2008
, vol. 
8
 (pg. 
146
-
156
)
31
Adams
P.C.
Barton
J.C.
A diagnostic approach to hyperferritinemia with a non-elevated transferrin saturation
J. Hepatol.
 , 
2011
, vol. 
55
 (pg. 
453
-
458
)
32
Kannengiesser
C.
Jouanolle
A.-M.
Hetet
G.
Mosser
A.
Muzeau
F.
Henry
D.
Bardou-Jacquet
E.
Mornet
M.
Brissot
P.
Deugnier
Y.
, et al.  . 
A new missense mutation in the L ferritin coding sequence associated with elevated levels of glycosylated ferritin in serum and absence of iron overload
Haematologica
 , 
2009
, vol. 
94
 (pg. 
335
-
339
)
33
Barton
J.C.
Acton
R.T.
Lee
P.L.
West
C.
SLC40A1 Q248H allele frequencies and Q248H-associated risk of non-HFE iron overload in persons of sub-Saharan African descent
Blood Cells. Mol. Dis.
 , 
2007
, vol. 
39
 (pg. 
206
-
211
)
34
Beutler
E.
Barton
J.C.
Felitti
V.J.
Gelbart
T.
West
C.
Lee
P.L.
Waalen
J.
Vulpe
C.
Ferroportin 1 (SCL40A1) variant associated with iron overload in African-Americans
Blood Cells. Mol. Dis.
 , 
2003
, vol. 
31
 (pg. 
305
-
309
)
35
Gordeuk
V.R.
Caleffi
A.
Corradini
E.
Ferrara
F.
Jones
R.A.
Castro
O.
Onyekwere
O.
Kittles
R.
Pignatti
E.
Montosi
G.
, et al.  . 
Iron overload in Africans and African-Americans and a common mutation in the SCL40A1 (ferroportin 1) gene
Blood Cells. Mol. Dis.
 , 
2003
, vol. 
31
 (pg. 
299
-
304
)
36
McNamara
L.
Gordeuk
V.R.
MacPhail
A.P.
Ferroportin (Q248H) mutations in African families with dietary iron overload
J. Gastroenterol. Hepatol.
 , 
2005
, vol. 
20
 (pg. 
1855
-
1858
)
37
Rivers
C.A.
Barton
J.C.
Gordeuk
V.R.
Acton
R.T.
Speechley
M.R.
Snively
B.M.
Leiendecker-Foster
C.
Press
R.D.
Adams
P.C.
McLaren
G.D.
, et al.  . 
Association of ferroportin Q248H polymorphism with elevated levels of serum ferritin in African Americans in the Hemochromatosis and Iron Overload Screening (HEIRS) Study
Blood Cells. Mol. Dis.
 , 
2007
, vol. 
38
 (pg. 
247
-
252
)
38
McGregor
J.A.
Shayeghi
M.
Vulpe
C.D.
Anderson
G.J.
Pietrangelo
A.
Simpson
R.J.
McKie
A.T.
Impaired iron transport activity of ferroportin 1 in hereditary iron overload
J. Membr. Biol.
 , 
2005
, vol. 
206
 (pg. 
3
-
7
)
39
Nekhai
S.
Xu
M.
Foster
A.
Kasvosve
I.
Diaz
S.
Machado
R.F.
Castro
O.L.
Kato
G.J.
Taylor
J.G.
Gordeuk
V.R.
Reduced sensitivity of the ferroportin Q248H mutant to physiological concentrations of hepcidin
Haematologica
 , 
2013
, vol. 
98
 (pg. 
455
-
463
)

Author notes

On behalf of the French National Network for the Molecular Diagnosis of Inherited Iron-Overload Disorders and Clinical Investigators (Jean-Baptiste Nousbaum, Caroline de Kerguenec, Pascale Poullin, Dominique Capron, Georges Barjonet, Samir Berber, Olivier Rosmorduc, Nicolas Ferry, Denis Vital-Durand).

Supplementary data