Abstract

Mutations leading to abrogation of matriptase-2 proteolytic activity in humans are associated with an iron-refractory iron deficiency anemia (IRIDA) due to elevated hepcidin levels. Here we describe two novel heterozygous mutations within the matriptase-2 ( TMPRSS6 ) gene of monozygotic twin girls exhibiting an IRIDA phenotype. The first is the frameshift mutation (P686fs) caused by the insertion of the four nucleotides CCCC in exon 16 (2172_2173insCCCC) that is predicted to terminate translation before the catalytic serine. The second mutation is the di-nucleotide substitution c.467C>A and c.468C>T in exon 3 that causes the missense mutation A118D in the SEA domain of the extracellular stem region of matriptase-2. Functional analysis of both variant matriptase-2 proteases has revealed that they lead to ineffective suppression of hepcidin transcription. We also demonstrate that the A118D SEA domain mutation causes an intra-molecular structural imbalance that impairs matriptase-2 activation. Collectively, these results extend the pattern of TMPRSS6 mutations associated with IRIDA and functionally demonstrate that mutations affecting protease regions other than the catalytic domain may have a profound impact in the regulatory role of matriptase-2 during iron deficiency.

INTRODUCTION

The type II transmembrane serine proteases (TTSPs) are a family of enzymes with more than 20 members currently identified in humans and mice ( 1–3 ). In addition to their characteristic type II transmembrane spanning region, the TTSPs share a number of common structural features, including a serine protease domain, a variable length stem region consisting of a mosaic of structural domains and a short cytoplasmic tail. Based on their amino acid sequences, TTSPs are likely synthesized as single chain zymogens before proteolytic activation following an arginine or lysine residue present in their highly conserved activation domains. Once activated, TTSPs are predicted to remain membrane-bound through a conserved disulphide bond linking the pro- and catalytic domains. Peri-cellular proteolysis via cell surface localized proteases is now recognized as an essential pathway through which cells interact with their immediate micro-environment ( 4 ). Cell-surface proteolysis regulates the transduction of extracellular stimuli across the cell membrane ( 5 , 6 ), release of bioactive growth factors, cytokines and peptide hormones, in addition to facilitating interactions with neighboring cells and proteins of the basement membrane and extracellular matrix ( 4 ). Increasingly, the TTSP family is proving to be integral components of key physiological events such as digestion (enteropeptidase) ( 7 ), hypertension (corin) ( 8 ) and auditory function (hepsin and TMPRSS3) ( 9 , 10 ).

Recently, the TTSP family member called matriptase-2 ( 11–13 ) has emerged as an essential regulator of body iron levels through its proteolytic processing of hemojuvelin and causal suppression of hepcidin ( 14 )—a central regulator of iron homeostasis ( 15 ). Hepcidin orchestrates body iron levels by limiting iron egress from duodenal enterocytes and macrophages to serum transferrin through its degradation of the iron transporter, ferroportin ( 16 ). Accordingly, mice deficient in matriptase-2 possess elevated levels of hepcidin and reduced duodenal expression of ferroportin that is accompanied by intestinal retention of dietary iron and presentation of iron-limited anemia ( 17 , 18 ). In accordance with its role in coordinating body iron levels, alterations in the genes encoding hepcidin ( 19 ) or its key regulators ( 20–22 ) induce iron overload syndromes such as hereditary hemochromatosis (HH). Consistently, HH disorders result from inadequate hepcidin production relative to body iron stores ( 23 ). Conversely, elevated hepcidin levels have been described in iron deficiency anemia patients that are insensitive to oral iron therapy and display an incomplete hematological recovery with parenteral iron administrations, a condition termed iron-refractory iron deficiency anemia [IRIDA (MIM 206200)] ( 24 , 25 ). Finberg et al . ( 24 ) provided the first description of human matriptase-2 mutations in IRIDA patients, which was subsequently complemented by two similar reports ( 25 , 26 ). In their entirety the reports describe heterozygous and homozygous biallelic human matriptase-2 mutations in just 14 IRIDA patients from European, African and African American ancestries. The mutations currently include frameshift, splice junction, missense and nonsense mutations, but all of them are distal to exon 6 and predominately encode matriptase-2 proteins which lack functional protease domains.

Here we describe two novel heterozygous mutations within the matriptase-2 ( TMPRSS6 ) gene of monozygotic twin girls exhibiting IRIDA phenotype and elevated hepcidin levels. The first is the frameshift mutation P686fs, caused by the insertion of four nucleotides in exon 16 (2172_2173insCCCC) of TMPRSS6 that is predicted to introduce a premature stop codon and terminate translation before the catalytic serine, generating a proteolytically deficient matriptase-2 variant. The second is a di-nucleotide substitution in exon 3 that causes the missense mutation A118D in the sea urchin sperm protein, enteropeptidase, agrin (SEA) domain of the extracellular stem region of matriptase-2. Mutations, such as the identified 2172_2173insCCCC, that produce proteolytically inactive matriptase-2 enzymes have been strongly linked to an IRIDA phenotype. Consequently, we first sought to delineate the less apparent molecular implications of the SEA domain mutation. In other SEA domain-containing membrane proteins, auto-proteolysis after a glycine residue within a conserved motif (e.g. GSVVV) releases these molecules from the cell surface ( 27 ). Generation of homology models of this domain in its wild-type and mutant forms displayed an intra-molecular structural imbalance in the A118D SEA domain mutant which in vitro correlated to a disruption in enzyme activation. Further functional analysis of the A118D and P686fs mutant proteases showed that both mutants are unable to suppress hepcidin transcription to the levels of the wild-type matriptase-2, providing proof-of-principle for the high hepcidin levels recorded in the studied probands.

RESULTS

Patient clinical synopsis

Written informed consent from the family was obtained and the present study completed in compliance with Spanish national ethical standards on human experimentation. Proband-1 and -2 are monozygotic twins of a non-consanguineous Spanish couple. The mother of the probands has a family history of thalassemia. Prior to our consultation, at the age of 1 year they were both diagnosed with microcytic anemia. Previous attempts to correct anemia through ingestion of ferrous sulphate proved ineffective. Consistently, negative iron absorption was demonstrated in both twins at the age of 4. Further, parenteral delivery of iron had minimal success, maintaining hemoglobin levels between 80 and 85 g/l. The probands first came to our attention at the ages of 15, both displaying physical symptoms of iron deficiency, including asthenia, koilonychia, pale lips and hair loss. The probands showed microcytic, hypochromic anemia, low serum iron, normal total iron binding capacity (TIBC), low saturation index, normal to high serum ferritin and high levels of soluble transferrin receptor (Table  1 ). Clinical analysis discarded thalassemia, sideroblastic anemia and the common forms of congenital dyserythropoietic anemias as possible causes. Dissimilar to oral iron therapy, administration of 200 mg iron glucose 3 times per week intravenously, improved serum levels of iron, ferritin and hemoglobin in both probands (Table  2 ), although MCV, MCH and transferrin saturation remained within the range for iron deficiency. Further, the anemia of proband-1 proved difficult to control due to hypermenorrhea and required further intravenous iron treatment. Subsequent to the initial intravenous iron treatment and normalization of hemoglobin, assessment of liver iron content by Magnetic Resonance demonstrated elevated iron stores of 200 µmol/g dry tissue and 150 µmol/g dry tissue for proband-1 and -2, respectively. Hepatic enzymes were not modified. Currently, proband-1 is receiving erythropoietin to mobilize iron from the liver to increase erythropoiesis. Using this approach serum ferritin has decreased from 688 µg/l to 191 µg/l and liver iron to 95 µmol/g dry tissue. This treatment also caused hemoglobin to reach a maximum of 131 g/l. Further, when serum ferritin levels approached 181 µg/l, hemoglobin decreased to 112 g/l and mucosal symptoms of iron deficiency reappeared. Consequently, iron sucrose was added to erythropoietin and the situation reverted. Proband-2 is currently without treatment and continues to exhibit normalized hematological parameters (last hemoglobin 133 g/l). Curiously, both parents showed minimal abnormalities in iron parameters.

Table 1.

Kindred clinical and biochemical parameters

  Family members 5
 
 Mother Father Proband 1 Proband 2 
Age at evaluation 48 50 15 15 
Sex 
Hb (g/l) 130 143 98 119 
MCV (fl) 89 84 65 77 
MCH (pg) 28.6 27 19.6 25 
RDW (%) 13.6 13.6 17.9 14.2 
Serum iron (μmol/l) 10 
Serum ferritin (μg/l) 54 409 123 388 
TIBC (μmol/l) 72 59 53 51 
Tf saturation (%) 14 15 
Soluble TfRs (mg/l) 4.8 13.5 6.6 
Urine hepcidin (ng/mg) ND ND  5069 a  5350 a 
  Family members 5
 
 Mother Father Proband 1 Proband 2 
Age at evaluation 48 50 15 15 
Sex 
Hb (g/l) 130 143 98 119 
MCV (fl) 89 84 65 77 
MCH (pg) 28.6 27 19.6 25 
RDW (%) 13.6 13.6 17.9 14.2 
Serum iron (μmol/l) 10 
Serum ferritin (μg/l) 54 409 123 388 
TIBC (μmol/l) 72 59 53 51 
Tf saturation (%) 14 15 
Soluble TfRs (mg/l) 4.8 13.5 6.6 
Urine hepcidin (ng/mg) ND ND  5069 a  5350 a 

Transferrin saturation was calculated by dividing the serum iron level by the total iron binding capacity (TIBC) and multiplying by 100. Urinary hepcidin values are presented as hepcidin/creatine ratios with a reference range of 71–1567 ng/mg for healthy adult controls. Hb, hemoglobin; MCV, mean corpuscular volume; MCH, mean corpuscular hemoglobin; ND, not determined; RDW, red blood cell distribution width; Tf, transferrin; TfRs, transferrin receptors.

a Hepcidin was measured after parenteral iron was administered.

Table 2.

The proband's hematological indices before and following intravenous administration of iron sucrose

  Proband 1
 
Proband 2
 
 Before 6 months after Before 6 months after 
Hb (g/l) 98 114 119 129 
MCV (fl) 65 75 77 81 
MCH (pg) 19.6 23.3 25 25.8 
RDW (%) 17.9 15.4 14.2 13.5 
Serum iron (μmol/l) 
Serum ferritin (mg/l) 123 365 388 586 
TIBC (μmol/l) 53 51 51 45 
Tf saturation (%) 12 
Soluble TfRs (mg/l) 13.5 6.6 ND 
  Proband 1
 
Proband 2
 
 Before 6 months after Before 6 months after 
Hb (g/l) 98 114 119 129 
MCV (fl) 65 75 77 81 
MCH (pg) 19.6 23.3 25 25.8 
RDW (%) 17.9 15.4 14.2 13.5 
Serum iron (μmol/l) 
Serum ferritin (mg/l) 123 365 388 586 
TIBC (μmol/l) 53 51 51 45 
Tf saturation (%) 12 
Soluble TfRs (mg/l) 13.5 6.6 ND 

Analysis of urinary hepcidin levels in patients with iron-refractory anemia

The post-natal presentation of anemia in conjunction with an inability to supplement serum iron through oral administration and partial response to intravenous iron treatments are consistent with an IRIDA phenotype. Deviating from the undetectable urinary hepcidin levels in individuals with iron deficiency anemia, IRIDA patients exhibit inappropriately high urinary hepcidin levels in relation to their iron status. Consequently, we sought to determine the hepcidin levels of the probands in an attempt to delineate potential causes of their anemia. Urine was collected from both probands and screened using a competitive ELISA approach (Intrinsic LifeSciences, CA, USA). As displayed in Table  1 , proband-1 and -2 evidenced extremely high levels of hepcidin in the urine (5069 and 5350 ng/mg creatine, respectively) that were consistent with prior reports of urinary hepcidin levels from IRIDA patients.

Identification of novel TMPRSS6 mutations in IRIDA patients

The striking phenotypic and hematological similarities of the probands and our recently described Tmprss6−/− mice—which exhibit microcytic anemia and elevated hepcidin levels—in conjunction with emerging evidence of matriptase-2 genetic abnormalities in IRIDA patients, led us to examine the genetic status of matriptase-2 in the probands. Nucleotide sequences of the 18 TMPRSS6 exons and their adjacent intron/exon boundaries were obtained through PCR amplification of paternal and proband genomic DNA using primers complementary to flanking intronic sequences, followed by automated sequencing of the PCR products. Sequencing analysis demonstrated that each proband harbors two heterozygous mutations. The first mutated allele is transmitted recessively from the father and arises from the di-nucleotide substitution c.467C>A and c.468C>T in exon 3 (nucleotide numbers refer to mRNA GenBank entry NM_153609) (Fig.  1 ), resulting in the missense mutation A118D within the SEA domain of the translated protein (Fig.  2 A; amino acid number refers to protein GenBank entry NP_705837). The second TMPRSS6 mutant allele is inherited recessively from the mother and originates from an insertion of the four nucleotides CCCC after 2172C in exon 16 (Fig.  1 ). This insertion is predicted to cause a frameshift mutation after proline 686 during translation of the protein and results in a premature stop codon which leads to the loss of the catalytic serine and conserved disulphide bond that anchors the activated protease domain to the stem region (Fig.  2 A).

Figure 1.

Family pedigree of the female twin sisters affected by iron-refractory iron deficiency anemia. The TMPRSS6 mutations identified by automated sequencing are displayed with the wild-type (WT) allele sequence beside the appropriate heterozygous affected parent as chromatographs. The missense mutation of Alanine 118 to Aspartate (A118D) results from the heterozygous substitutions c.467C>A and c.468C>T in exon 3. The frameshift mutation immediately following Proline 686 (P686fs) arises from an insertion of the four nucleotides CCCC between bases 2172 and 2173 in exon 16.

Figure 1.

Family pedigree of the female twin sisters affected by iron-refractory iron deficiency anemia. The TMPRSS6 mutations identified by automated sequencing are displayed with the wild-type (WT) allele sequence beside the appropriate heterozygous affected parent as chromatographs. The missense mutation of Alanine 118 to Aspartate (A118D) results from the heterozygous substitutions c.467C>A and c.468C>T in exon 3. The frameshift mutation immediately following Proline 686 (P686fs) arises from an insertion of the four nucleotides CCCC between bases 2172 and 2173 in exon 16.

Figure 2.

Mutation domain locations in matriptase-2. ( A ) Schematic representation of the predicted domain structure of the matriptase-2 protein with the corresponding location of the A118D and P686fs mutations. The matriptase-2 sequence was analyzed for modular protein domains using the SMART algorithm (ExPASy Proteomics Tools website) and the provided numbers refer to the amino acid position in the preproenzyme. ( B ) Matriptase-2 SEA domain sequences from the indicated species were aligned using ClustalX 2.0.10 software ( 46 ). Residues corresponding to human alanine 118 are boxed. Amino acids have been classified as hydrophobic (black), charged (red) or polar (green). Sequences were obtained from the following GenBank entries: Human NP_705837; Macaque XP_001085319; Dog XP_531743; Cow XP_871580; Mouse NP_082178; Rat XP_235768.

Figure 2.

Mutation domain locations in matriptase-2. ( A ) Schematic representation of the predicted domain structure of the matriptase-2 protein with the corresponding location of the A118D and P686fs mutations. The matriptase-2 sequence was analyzed for modular protein domains using the SMART algorithm (ExPASy Proteomics Tools website) and the provided numbers refer to the amino acid position in the preproenzyme. ( B ) Matriptase-2 SEA domain sequences from the indicated species were aligned using ClustalX 2.0.10 software ( 46 ). Residues corresponding to human alanine 118 are boxed. Amino acids have been classified as hydrophobic (black), charged (red) or polar (green). Sequences were obtained from the following GenBank entries: Human NP_705837; Macaque XP_001085319; Dog XP_531743; Cow XP_871580; Mouse NP_082178; Rat XP_235768.

Activation impairment in the SEA and protease domain matriptase-2 mutants

According to the above results, the two TMPRSS6 mutations found in the analyzed family affect two different regions of matriptase-2 and may have distinct functional consequences. The missense point mutation lies in a protein domain which has not been previously found altered in IRIDA patients, whereas the frameshift mutation results in the partial loss of the catalytic domain, including residues essential for proteolysis. It is now well documented that mutations leading to abrogation of matriptase-2 proteolytic activity are associated with an IRIDA phenotype, presumably due to absent peri-cellular processing of hemojuvelin and inadequate hepcidin suppression. Further, Silvestri et al . ( 28 ) have recently reported IRIDA patients with loss-of-function mutations in the second LDLa and CUB domains of matriptase-2. To gain insight into the possible implications of the SEA domain mutation, we first evaluated the potential impact of the A118D mutation on matriptase-2 function by using the SIFT and PMut programs ( 29 , 30 ). When the protein sequence of the SEA domain was interrogated using the PMut program the A118D mutation was predicted to be pathological (score of 0.94). In agreement with this finding, using the SIFT algorithm the normalized probability for an A118D substitution at the SEA domain of matriptase-2 was 0.02 with a median sequence conservation of 3.02, thus predicting a deleterious effect for the mutated protein. Furthermore, protein sequence analysis of the SEA domains from human, macaque, cow, dog, rat and mouse demonstrates that alanine 118 (based on human residue numbering) is absolutely conserved across these species (Fig.  2 B) and suggests that this evolutionary conserved residue may be important for matriptase-2 function.

Having established some preliminary evidence that the SEA domain mutation is potentially deleterious for matriptase-2 function, we next sought to determine the presence of molecular alterations by generating homology models of this domain in its wild-type and mutant forms (Fig.  3 ). As the A118D mutation is predicted to force the insertion of an aspartic acid residue into the hydrophobic core of the protein, we performed energy minimizations of both the wild-type (Fig.  3 A) and mutant (Fig.  3 B) matriptase-2 SEA domains. Critically, the preeminent difference between the wild-type and mutant SEA domains is destabilization of the h1 α-helix in the mutant form. This suggests that the A118D mutation in matriptase-2 may affect the folding of its SEA domain and as a consequence lead to structural destabilization. Interestingly, some wild-type minimized models show a significant spatial shift at the auto-activation loop of this domain (Fig.  3 A, green circle) which is not found in the mutant form. Collectively, our homology modelling suggests that the A118D mutation may affect enzyme function through structural alterations within the SEA domain.

Figure 3.

Predicted structural consequences of the SEA domain mutation. Energy-minimized homology models of ( A ) wild-type and ( B ) mutant matriptase-2 SEA domains. Structures are shown as ribbons, and residue 118 is displayed as a ball and stick. Molecular shifts after energy minimization are displayed with the color code, where regions in red have been shifted farther away than regions in blue.

Figure 3.

Predicted structural consequences of the SEA domain mutation. Energy-minimized homology models of ( A ) wild-type and ( B ) mutant matriptase-2 SEA domains. Structures are shown as ribbons, and residue 118 is displayed as a ball and stick. Molecular shifts after energy minimization are displayed with the color code, where regions in red have been shifted farther away than regions in blue.

To further validate this observation, we initially chose to analyze in vitro the expression and cellular processing of the A118D matriptase-2 variant protease in comparison to the wild-type enzyme. In parallel, we performed the same studies with the P686fs matriptase-2 variant identified in the probands. To this purpose, the full-length cDNA encoding wild-type matriptase-2 (WT) and the A118D and P686fs matriptase-2 cDNAs, obtained through site-directed mutagenesis, were transiently transfected into HepG2 cells and immunofluorescence was used to examine the cellular localization in permeabilized (whole cell; Fig.  4 A top panels) and non-permeabilized (plasma membrane; Fig.  4 A bottom panels) cells. In permeabilized HepG2 cells, the A118D and P686fs proteins displayed a similar diffuse expression pattern throughout the cytoplasm as the wild-type protein. Importantly, in non-permeabilized cells the mutant proteins showed distinct membrane staining like that observed for wild-type matriptase-2 (marked by white arrowheads) and demonstrate that the A118D and P686fs matriptase-2 variants are expressed on the plasma membrane.

Figure 4.

Cellular expression and localization of wild-type and mutant matriptase-2 proteins. ( A ) Confocal microscopy analysis of permeabilized (upper panel;+perm) and non-permeabilized (lower panel; - perm) HepG2 cells transiently transfected with pcDNA3 only (mock), pcDNA3 matriptase-2 wild-type-FLAG (WT), pcDNA3 matriptase-2 A118D-FLAG (A118D) or pcDNA3 matriptase-2 P686fs-FLAG (P686fs) expression constructs. Cells were incubated with an anti-FLAG antibody (green) and stained with DAPI (blue) to identify nuclei. Cell surface localized matriptase-2 proteins in non-permeabilized cells are indicated by arrowheads. The white scale bar represents 10 µm. ( B ) Western blot analysis probing with an anti-FLAG antibody of whole cell lysates (upper left panel) and conditioned media (upper right panels) from HepG2 cells transiently transfected with pcDNA3 only (mock), pcDNA3-matriptase-2 wild-type-FLAG (WT), pcDNA3 matriptase-2 A118D-FLAG (A118D) or pcDNA3 matriptase-2 P686fs-FLAG (P686fs). Lysates (50 µg) and conditioned media, which are displayed in separated panels due to differences in protein amounts (left panel represents 100 µg of media and the right panel 50 µg of media), were electrophoresed under reducing conditions. Membranes were also probed with an anti-beta actin antibody to demonstrate equal loading in cell lysates (lower left panel) and an absence of intracellular protein contamination (lower right panels). Asterisk denotes non-specific band.

Figure 4.

Cellular expression and localization of wild-type and mutant matriptase-2 proteins. ( A ) Confocal microscopy analysis of permeabilized (upper panel;+perm) and non-permeabilized (lower panel; - perm) HepG2 cells transiently transfected with pcDNA3 only (mock), pcDNA3 matriptase-2 wild-type-FLAG (WT), pcDNA3 matriptase-2 A118D-FLAG (A118D) or pcDNA3 matriptase-2 P686fs-FLAG (P686fs) expression constructs. Cells were incubated with an anti-FLAG antibody (green) and stained with DAPI (blue) to identify nuclei. Cell surface localized matriptase-2 proteins in non-permeabilized cells are indicated by arrowheads. The white scale bar represents 10 µm. ( B ) Western blot analysis probing with an anti-FLAG antibody of whole cell lysates (upper left panel) and conditioned media (upper right panels) from HepG2 cells transiently transfected with pcDNA3 only (mock), pcDNA3-matriptase-2 wild-type-FLAG (WT), pcDNA3 matriptase-2 A118D-FLAG (A118D) or pcDNA3 matriptase-2 P686fs-FLAG (P686fs). Lysates (50 µg) and conditioned media, which are displayed in separated panels due to differences in protein amounts (left panel represents 100 µg of media and the right panel 50 µg of media), were electrophoresed under reducing conditions. Membranes were also probed with an anti-beta actin antibody to demonstrate equal loading in cell lysates (lower left panel) and an absence of intracellular protein contamination (lower right panels). Asterisk denotes non-specific band.

Having established that the expression and cell surface localization of the A118D and P686fs matriptase-2 variants remains unchanged, we next sought to examine if the mutations affect enzyme maturation in vitro . Expression constructs encoding wild-type matriptase-2 and the A118D and P686fs mutants were transiently transfected into HepG2 cells and whole cell lysates and conditioned media were collected after 48 h. Western blot analysis under reducing conditions of whole cell lysates using an anti-flag antibody predominately detected wild-type and A118D matriptase-2 as ∼105 kDa proteins, while the P686fs matriptase-2 variant was detected as an 100 kDa protein, consistent with the frame-shift effecting the loss of 51 amino acids (Fig.  4 B; upper left panel). Further, the presence of the FLAG-epitope at the carboxyl termini of the full-length proteins and the absence of a smaller molecular weight immunoreactive species that would represent the proteolytic domain under reducing conditions illustrates that wild-type and mutant matriptase-2 variants are present as zymogen enzymes at the cell surface.

This result prompted us to examine if matriptase-2 requires cell surface shedding for enzyme activation. Western blot analysis under reducing conditions of concentrated media from wild-type, A118D and P686fs matriptase-2 transfected HepG2 cells using an anti-FLAG antibody demonstrated that two wild-type matriptase-2 isoforms were present in the media as ∼95 and 30 kDa proteins (Fig.  4 B; upper right panel). In accordance with predicted sizes and the carboxyl location of the FLAG-epitope, these bands likely represent shed zymogen enzyme (95 kDa) and the active protease domain that has been separated from the stem region through disruption of its disulphide linkage (30 kDa). Further, when shed wild-type matriptase-2 was examined under non-reducing conditions preserving the potential disulphide linkage that an activated protease domain would possess with the stem region, a single high molecular weight species was detected, supporting further the identity of the 30 kDa band as the active protease domain (data not shown). These results suggest a multi-step activation sequence for matriptase-2, which first requires release from the cell surface before activation occurs. Importantly, Western blot analysis of cell media from A118D matriptase-2 transfected cells detected only a single 100 kDa protein and demonstrates that activation of matriptase-2 is impaired in enzymes containing the SEA domain mutation. Similarly, the P686fs mutant matriptase-2 was found in the media as a single 95 kDa isoform with no apparent release of truncated protease domain being observed. Collectively, these data suggest that while surface shedding of the mutant matriptase-2 proteases remains unaffected, subsequent stages of the enzymes activation sequence are disrupted as a direct result of the A118D and P686fs mutations.

A118D and P686fs mutations lead to impaired hepcidin transcription in vitro

As matriptase-2 lowers hepcidin transcription through proteolytic processing of the bone morphogenetic protein (BMP) co-receptor hemojuvelin, we next examined if impaired activation of the A118D variant and the proteolytic inactivity of the P686fs variant had an impact on hemojuvelin stimulated hepcidin transcription. Hep3B cells were transiently transfected with luciferase under the control of the HAMP promoter, as well as cDNAs for hemojuvelin and either wild-type matriptase-2 (WT), A118D mutant matriptase-2 (A118D), P686fs mutant matriptase-2 (P686fs) or a vector only control. As shown in Figure  5 , under basal conditions, expression of the A118D and P686fs matriptase-2 mutants displayed a statistically significant ( P = 0.02 and P = 0.05, respectively) inability to reduce hepcidin transcription in comparison to the wild-type enzyme. This observation was similarly reflected in Hep3B cells that were stimulated with BMP-2, with A118D and P686fs variants both unable to reduce hepcidin transcription ( P = 0.04 and P = 0.03, respectively) to the extent as that measured for wild-type matriptase-2. Interestingly under basal conditions, despite a less effective hepcidin transcriptional reduction when compared with the wild-type enzyme, the A118 variant demonstrated statistically significant hepcidin suppression when compared with the vector only control ( P = 0.02). While not statistically significant ( P = 0.07), under basal conditions the P686fs variant demonstrated a similar trend of hepcidin suppression. However, when stimulated with BMP-2, both matriptase-2 variants displayed a complete inability to reduce hepcidin transcription.

Figure 5.

Reduced hepcidin suppression by the A118D and P686fs mutant matriptase-2 enzymes. The hepcidin promoter luciferase reporter construct and pRL-TK construct (transfection efficiency control) were transiently transfected into the Hep3B cell line simultaneously with expression vectors encoding hemojuvelin, in addition to either wild-type matriptase-2 (WT), A118D mutant matriptase-2 (A118D), P686fs mutant matriptase-2 (P686fs) or no insert (vector control). The transcriptional activity of the hepcidin promoter under basal and BMP-2 stimulated conditions was quantified by determining the Firefly (hepcidin)/Renilla (transfection control) luciferase activities. Results are reported as the mean ± SEM of the relative luciferase activity for cells transfected with WT, A118D or P686fs matriptase-2, compared with cells transfected with vector only. The experiment was performed in triplicate. Student's t- test was used for estimation of statistical significance with the calculated P -values displayed.

Figure 5.

Reduced hepcidin suppression by the A118D and P686fs mutant matriptase-2 enzymes. The hepcidin promoter luciferase reporter construct and pRL-TK construct (transfection efficiency control) were transiently transfected into the Hep3B cell line simultaneously with expression vectors encoding hemojuvelin, in addition to either wild-type matriptase-2 (WT), A118D mutant matriptase-2 (A118D), P686fs mutant matriptase-2 (P686fs) or no insert (vector control). The transcriptional activity of the hepcidin promoter under basal and BMP-2 stimulated conditions was quantified by determining the Firefly (hepcidin)/Renilla (transfection control) luciferase activities. Results are reported as the mean ± SEM of the relative luciferase activity for cells transfected with WT, A118D or P686fs matriptase-2, compared with cells transfected with vector only. The experiment was performed in triplicate. Student's t- test was used for estimation of statistical significance with the calculated P -values displayed.

DISCUSSION

Here we present clinical, genetic and biochemical evidence of an IRIDA phenotype that arises from two novel heterozygous mutations within the TMPRSS6 gene. The studied subjects, monozygotic twin girls of Spanish descent, presented numerous hematological abnormalities upon initial consultation, including microcytic anemia that was accompanied by low serum iron, normal total iron binding capacity, low saturation index, normal to high serum ferritin and high levels of soluble transferrin receptor. Our finding of highly elevated levels of urinary hepcidin in the probands explains the ineffectiveness of prior attempts to correct iron levels through oral supplementation, as high hepcidin levels would increase ferroportin internalization and degradation from the surface of enterocytes, lowering intestinal iron transport to serum. Further, circumvention of intestinal iron absorption via intravenous iron administrations corrected several of the hematological deficiencies, despite iron loading reticuloendothelial cells ( 31 ). Potentially, intravenous iron treatment delivers sufficient iron to maintain an almost normalized level of erythropoiesis. Differences in iron requirements between both probands could be due to hypermenorrhea of proband-1. It is important to point out that the liver demonstrated a high level of iron loading after intravenous iron administrations. Interestingly, this iron was partially used in new erythropoiesis after erythropoietin treatment. Used judiciously, erythropoietin and intravenous iron were able to maintain hemoglobin levels while avoiding potentially harmful iron overload.

Previously, high hepcidin levels in IRIDA patients have been linked to loss-of-function matriptase-2 mutations resulting in inadequate hepcidin suppression ( 24–26 ). In accordance, matriptase-2 deficient Tmprss6−/− and Mask murine models have iron-limited anemia due to elevated hepcidin levels reducing duodenal ferroportin expression and sequestering dietary iron within the intestine ( 17 , 18 ). Our analysis of the coding regions of the TMPRSS6 gene in the probands identified the presence of two novel heterozygous mutations. The mutations arise from a frameshift mutation caused by the insertion of four nucleotides CCCC in exon 16 (2172_2173insCCCC) that is predicted to introduce a premature stop codon and terminate translation before the catalytic serine. The second mutation is the di-nucleotide substitution c.467C>A and c.468C>T in exon 3 that causes the missense mutation A118D in the SEA domain of the extracellular stem region of matriptase-2. In humans and mice, the elimination of matriptase-2 proteolytic function through direct mutations within the enzyme protease domain, or exclusion of residues essential for proteolytic function, is causally linked to IRIDA phenotype development due to absent hepcidin suppression ( 14 ). Accordingly, the TMPRSS6 exon 16 loss-of-function mutation identified in this study was a clear candidate as being responsible for the high hepcidin levels detected in the probands. However, the functional consequences of the other variant allele (alanine to aspartic acid substitution at amino acid position 118, exon 3) were less apparent. This prompted us to delineate the associated molecular and function alterations caused by the A118D missense mutation within the SEA domain of the matriptase-2 extracellular stem region.

The stem regions of the TTSPs contain as many as 11 structural domains that potentially mediate regulatory and/or protein interaction roles. These include LDLa (low density lipoprotein receptor class A) domains; SR (group A scavenger receptor) domains; frizzled domains; CUB (Cls/Clr, urchin embryonic growth factor and bone morphogenic protein 1) domains; MAM (a meprin, A5 antigen, and receptor protein phosphatase μ) domains; and SEA domains ( 2 , 3 ). Although the roles of individual domains have not yet been elucidated, the importance of stem regions for normal TTSP biochemical function is illustrated by studies demonstrating their roles in zymogen conversion, substrate recognition and proteolytic activity ( 32–35 ). Consistently, in this study we have demonstrated that the stem region mutation A118D disrupts matriptase-2 zymogen activation in vitro through a predicted intra-molecular structural imbalance. The degree to which the A118D mutation affects matriptase-2 function may be potentially linked to its SEA domain location. In other SEA domain-containing transmembrane proteins, autoproteolysis, following a glycine residue within a GSVVV consensus motif, releases these molecules from the cell surface ( 27 ). Proteolysis within this conserved motif which is located near the middle of the SEA domain, is catalysed by conformational stress and the hydroxyl-group of the serine residue present within the motif ( 27 ). Relevantly, the highly homologous TTSP family member, matriptase, undergoes a complex and not yet fully elucidated activation sequence that is dependent upon initial SEA domain cleavage and subsequent endo-proteolysis by an unknown protease(s) ( 36–38 ). Further, shedding from the cell surface for murine matriptase ( 37 ) and porcine enteropeptidase ( 39 ) is mediated through proteolytic processing within their SEA domains. Critically, although we demonstrate herein that the A118D matriptase-2 variant is shed from the cell surface like its wild-type counterpart, the mutated SEA domain appears to disrupt the ensuing enzyme activation sequence, as only the wild-type enzyme releases an appropriately size fragment representing the activated protease domain. Interestingly, we demonstrate that cell surface shedding is a potential requirement for enzyme activation as only the zymogen form of matriptase-2 was detected in cell lysates. We can also speculate that the larger molecular weight of the shed A118D matriptase-2 zymogen (100 kDa) in comparison to the wild-type shed zymogen (95 kDa), indicates that in addition to inceptive proteolysis for cell surface release, matriptase-2 activation requires further endo-proteolytic cleavage(s). Furthermore, the inability of the proteolytic deficient P686fs mutant to undergo activation following cell surface release signifies that matriptase-2 requires a structurally intact protease domain and/or proteolytic activity for enzyme maturation. Consistently, we have previously reported auto-activation of a recombinant matriptase-2 protease domain produced in bacteria ( 11 ). Overall, these data highlight an analogous activation mechanism for matriptase and matriptase-2 that requires multiple endo-proteolytic events and the SEA and protease domains of the enzyme.

The arrest of enzyme activation catalyzed by the A118D and P686fs mutations forms a blockade in matriptase-2 substrate proteolysis. Recently, matriptase-2 has been shown to degrade hemojuvelin in vitro , which in turn leads to hepcidin transcriptional suppression ( 14 ). Hemojuvelin is synthesized by hepatocytes as a membrane GPI-linked protein that behaves as a co-receptor for BMP-2, -4 and -6 hepcidin stimulation in vitro ( 40–42 ), while in vivo BMP-6 has recently been described as the physiological regulator of hepcidin ( 43 , 44 ). BMP stimuli, transduced via SMAD proteins, are the primary activators of hepcidin expression ( 45 ). In this work, we have demonstrated in vitro that in comparison to wild-type matriptase-2, the A118D and P686fs variant enzymes have a markedly impaired ability to reduce basal and BMP-2 stimulated hepcidin transcription. Unexpectedly, both mutant proteases displayed a degree of hepcidin transcriptional suppression under basal conditions. This is potentially due to the mutant enzymes behaving in dominant negative fashions, binding and sequestering hemojuvelin from interactions with endogenous agonists. However when cells were stimulated with exogenous BMP-2 both mutant proteases displayed a complete inability to reduce hepcidin transcription. Supporting our findings, a study published during revision of this manuscript describes that mutations in the LDLa and CUB domains of the matriptase-2 stem region, impair the enzymes ability to lower hepcidin transcription ( 28 ).

In summary, the cumulative effect of inheriting the TMPRSS6 alleles containing the four nucleotide insertion in exon 16 (2172_2173insCCCC) and the di-nucleotide substitution in exon 3 (c.467C>A and c.468C>T) is a catalytically inactive population of matriptase-2 enzymes, due to absent proteolytic residues and impaired zymogen activation, respectively. The probands high hepcidin levels and resulting IRIDA phenotype, potentially derives from an inability to process hemojuvelin to ensure maintenance of hepcidin within the precise boundaries for normal physiological function. Further, our description of a SEA domain mutation that is linked to a human disease condition is a novel observation for the TTSP family that improves our understanding of the functional importance the individual domains within TTSP stem regions have for enzyme activity.

MATERIALS AND METHODS

Cell culture and reagents

HepG2 and Hep3B cell lines were obtained from American Type Culture Collection (ATCC; Manassas, VA) and were grown in Dulbecco's modified Eagle's medium containing 10% fetal calf serum (FCS), 100 units/ml penicillin and 100 µg/ml streptomycin (Invitrogen) and propagated in 95% air, 5% CO 2 at 37°C.

cDNA expression constructs

A full-length human matriptase-2 cDNA with the inclusion of a C-terminal FLAG epitope was cloned into pcDNA3 as previously described. ( 11 ) The A118D matriptase-2 variants were generated by site-directed mutagenesis of the wild-type cDNA using the QuikChange XL site-directed mutagenesis kit (Stratagene) and the oligonucleotides 5′-gcagtgaaaccgccaaagatcagaagatgctcaaggag-3′ and 5′ctccttgagcatcttctgatctttggcggtttcactg c-3′. The P686fs variant was generated by site-directed mutagenesis using the oligonucleotides 5′-ccgtgcgccccccccgtctgcctgcc-3′ and 5′-ggcaggcagacggggggggcgcacgg-3′ followed by PCR amplification to include a C-terminal FLAG epitope using the oligonucleotides 5′-ctgagctgtctcggcacccacttgcagt-3′ and 5′-gttacttgtcatcgtcgtccttgtagtctccccctgacaggcatcc-3′, before insertion into pcDNA3. All generated expression constructs were sequence verified. A pGL2-basic reporter vector containing a 2.9 kb fragment of the human hepcidin promoter and a pcDNA3.1 expression plasmid containing the hemojuvelin cDNA were both generous gifts from Prof. Clara Camaschella (Vita-Salute San Raffaele University, Istituto di Ricovero e Cura a Carattere Scientifico San Raffaele, Milan, Italy).

TMPRSS6 mutation screening

Genomic DNA was isolated from whole blood using the blood and cell DNA midi kit (Qiagen) and quantified using a NanoDrop ND-1000 spectrophotometer (NanoDrop technologies). Using genomic DNA as template, each of the exons of all family members was individually amplified by PCR using primers ( Supplementary Material, Table S1 ) targeting intronic sequences located ∼40 bp from each of the 36 exon/intron boundaries of the TMPRSS6 gene. Each of the PCR products was purified using a high pure PCR product purification kit (Roche) and sequences determined using fluorescent dye-terminator chemistry (Macrogen). Resulting sequences were analyzed by Mutation Surveyor software (Softgenetics) using the GenBank entry NM_153609 as a reference. Identified TMPRSS6 mutations were ruled out as common polymorphic changes by sequencing the corresponding exons in 60 Spanish individuals with normal hematological indices.

Urinary hepcidin analysis

Duplicate samples of urine were obtained from each proband, snap frozen and sent to Intrinsic LifeSciences for analysis of bioactive hepcidin levels using a competitive enzyme-linked immunoassay (ELISA). In 24 normal subjects urinary hepcidin levels ranged from 71 to 1567 ng/mg creatine (M. Westerman, Intrinsic LifeSciences, La Jolla, CA, personal communication).

Matriptase-2 mutation analysis

The potential impact of an amino acid change in matriptase-2 function was predicted in silico using the PMut (available at http://mmb2.pcb.ub.es:8080/PMut/ ) and SIFT algorithms (Sorting Intolerant From Tolerant, available at http://blocks.fhcrc.org/sift/SIFT.html ) ( 29 ). PMut is a web-based tool that calculates the pathological significance of single point amino acid mutations using neural networks, with those resulting output scores greater than 0.5 considered to be pathological. SIFT is a computational web-based approach that calculates a score, ranging from 0 to 1, which corresponds to the probability that a particular change in an amino acid is tolerated. Amino acids with probabilities <0.05 are predicted to be deleterious, whereas 1 is neutral.

Molecular modeling of matriptase-2

The sequences of wild type and mutant SEA domains of matriptase-2 were manually threaded over a known SEA domain structure (PDB ID 2E7V) by using Deep View. The resulting alignment was then sent to the Swiss Model server. The resulting models were subjected to energy minimization with the 2005.06 release of MOE (Molecular Operation Environment). We used the MMFF94× force field with calculation of partial charges and with a 0.05 gradient. Several slightly modified starting conditions were used to get robust results. This included the mutation of different small aliphatic residues to aspartic acid.

Confocal microscopy

Cells plated on sterile glass coverslips were allowed to adhere overnight, and then transfected with either pcDNA3 containing no cDNA (vector control), wild-type matriptase-2, A118D matriptase-2 or P686fs matriptase-2 expression constructs. After 48 h, cells were fixed with 1% (v/v) formaldehyde and where indicated cells were permeabilized by incubation with 1% (v/v) Triton X-100 in PBS. Cells were washed, blocked with 5% (v/v) normal goat serum in PBS, followed by overnight incubation at 4°C with an anti-FLAG antibody (Cell Signaling; 1:200) in blocking buffer. Following multiple washes the cells were incubated with an Alexa Fluor 488 nm-conjugated goat anti-rabbit IgG antibody (1:500) for 1 h at room temperature. Nuclei were stained by incubating cells with 4',6-diamidino-2-phenylindole (DAPI), after which, coverslips were mounted on slides and cells imaged with a Leica SP2 confocal microscope (Leica Microsystems). Images were processed using Adobe Photoshop CS3 and displayed using CorelDraw.

Western blot analysis

HepG2 cells at 80% confluency were transiently transfected with plasmids encoding wild-type matriptase-2 or the A118D or P686fs matriptase-2 variants using Lipofectamine Plus (Invitrogen) according to the manufacturer's instructions. After 18 h, the transfection media was replaced with serum-free media for a further 24 h, and then the media was cleared and concentrated using a 10 kDa molecular mass cutoff ultra-filtration column (Amicon Ultra; Millipore). Whole cell lysates were collected in a buffer containing Triton X-100 (1%, v/v), 50 m m Tris–HCl (pH 7.4), NaCl (150 m m ), and protease inhibitor mixture (Roche Applied Science) and protein concentrations were determined by bicinchoninic acid assay (Pierce). Equal amounts of lysates (50 µg) or concentrated media (50 µg or 100 µg as indicated) were separated by SDS–PAGE and transferred to a nitrocellulose membrane that was blocked with 5% powdered milk. Membranes were incubated overnight at 4°C with an anti-FLAG (1:1000) or anti-beta actin (Sigma; 1:10 000) antibody and then washed before incubation with species-appropriate HRP-conjugated secondary antibodies for 1 h at room temperature. Membranes were analyzed using an LAS-3000 mini imaging system (Fujifilm).

Luciferase assays

Hep3B cells at 80% confluency in 12-well plates (Becton Dickson) were transiently transfected with 1.5 µg hepcidin promoter luciferase reporter construct in combination with 20 ng of pRL-TK Renilla luciferase vector (Promega) to monitor transfection efficiency. In addition to the reporter and transfection control constructs, 1 µg of the hemojuvelin expression construct and constructs encoding either wild-type matriptase-2, A118D and P686fs matriptase-2 variants or pcDNA3 containing no cDNA (vector control) were also transfected. Eighteen hours post-transfection, cells were serum starved (basal) or treated with 50 ng/ml BMP-2 (R&D systems) in the absence of serum for a further 24 h period. Cells were then lysed and the transcriptional activity of the hepcidin promoter quantified by determining the Firefly (hepcidin)/Renilla (transfection control) luciferase activities. Results are reported as the mean ± SEM of the relative luciferase activity for cells transfected with wild-type matriptase-2 or the A118D or P686fs matriptase-2 variants, compared with cells transfected with vector only. The experiment was performed in triplicate. Student's t -test was used for estimation of statistical significance.

SUPPLEMENTARY MATERIAL

Supplementary Material is available at HMG online .

FUNDING

This work was supported by grants from Ministerio de Ciencia e Innovación-Spain, Fundación ‘M. Botín’ and the European Union (FP7 MicroEnviMet). The Instituto Universitario de Oncología is supported by Obra Social Cajastur and Acción Transversal del Cáncer-RTICC.

ACKNOWLEDGEMENTS

We thank Dr L.M. Sánchez and Dr A.R. Folgueras for helpful comments and S. Alvarez and D. Alvarez for excellent technical assistance.

Conflict of Interest statement. None declared.

REFERENCES

1
Puente
X.S.
Sanchez
L.M.
Overall
C.M.
Lopez-Otin
C.
Human and mouse proteases: a comparative genomic approach
Nat. Rev. Genet.
 , 
2003
, vol. 
4
 (pg. 
544
-
558
)
2
Szabo
R.
Bugge
T.H.
Type II transmembrane serine proteases in development and disease
Int. J. Biochem. Cell Biol.
 , 
2008
, vol. 
40
 (pg. 
1297
-
1316
)
3
Hooper
J.D.
Clements
J.A.
Quigley
J.P.
Antalis
T.M.
Type II transmembrane serine proteases. Insights into an emerging class of cell surface proteolytic enzymes
J. Biol. Chem.
 , 
2001
, vol. 
276
 (pg. 
857
-
860
)
4
Netzel-Arnett
S.
Hooper
J.D.
Szabo
R.
Madison
E.L.
Quigley
J.P.
Bugge
T.H.
Antalis
T.M.
Membrane anchored serine proteases: a rapidly expanding group of cell surface proteolytic enzymes with potential roles in cancer
Cancer Metastasis Rev.
 , 
2003
, vol. 
22
 (pg. 
237
-
258
)
5
Ramsay
A.J.
Dong
Y.
Hunt
M.L.
Linn
M.
Samaratunga
H.
Clements
J.A.
Hooper
J.D.
Kallikrein-related peptidase 4 (KLK4) initiates intracellular signaling via protease-activated receptors (PARs). KLK4 and PAR-2 are co-expressed during prostate cancer progression
J. Biol. Chem.
 , 
2008
, vol. 
283
 (pg. 
12293
-
12304
)
6
Ramsay
A.J.
Reid
J.C.
Adams
M.N.
Samaratunga
H.
Dong
Y.
Clements
J.A.
Hooper
J.D.
Prostatic trypsin-like kallikrein-related peptidases (KLKs) and other prostate-expressed tryptic proteinases as regulators of signalling via proteinase-activated receptors (PARs)
Biol. Chem.
 , 
2008
, vol. 
389
 (pg. 
653
-
668
)
7
Kitamoto
Y.
Yuan
X.
Wu
Q.
McCourt
D.W.
Sadler
J.E.
Enterokinase, the initiator of intestinal digestion, is a mosaic protease composed of a distinctive assortment of domains
Proc. Natl Acad. Sci. USA
 , 
1994
, vol. 
91
 (pg. 
7588
-
7592
)
8
Chan
J.C.
Knudson
O.
Wu
F.
Morser
J.
Dole
W.P.
Wu
Q.
Hypertension in mice lacking the proatrial natriuretic peptide convertase corin
Proc. Natl Acad. Sci. USA
 , 
2005
, vol. 
102
 (pg. 
785
-
790
)
9
Guipponi
M.
Tan
J.
Cannon
P.Z.
Donley
L.
Crewther
P.
Clarke
M.
Wu
Q.
Shepherd
R.K.
Scott
H.S.
Mice deficient for the type II transmembrane serine protease, TMPRSS1/hepsin, exhibit profound hearing loss
Am. J. Pathol.
 , 
2007
, vol. 
171
 (pg. 
608
-
616
)
10
Guipponi
M.
Toh
M.Y.
Tan
J.
Park
D.
Hanson
K.
Ballana
E.
Kwong
D.
Cannon
P.Z.
Wu
Q.
Gout
A.
, et al.  . 
An integrated genetic and functional analysis of the role of type II transmembrane serine proteases (TMPRSSs) in hearing loss
Hum. Mutat.
 , 
2008
, vol. 
29
 (pg. 
130
-
141
)
11
Velasco
G.
Cal
S.
Quesada
V.
Sanchez
L.M.
Lopez-Otin
C.
Matriptase-2, a membrane-bound mosaic serine proteinase predominantly expressed in human liver and showing degrading activity against extracellular matrix proteins
J. Biol. Chem.
 , 
2002
, vol. 
277
 (pg. 
37637
-
37646
)
12
Hooper
J.D.
Campagnolo
L.
Goodarzi
G.
Truong
T.N.
Stuhlmann
H.
Quigley
J.P.
Mouse matriptase-2: identification, characterization and comparative mRNA expression analysis with mouse hepsin in adult and embryonic tissues
Biochem. J.
 , 
2003
, vol. 
373
 (pg. 
689
-
702
)
13
Ramsay
A.J.
Reid
J.C.
Velasco
G.
Quigley
J.P.
Hooper
J.D.
The type II transmembrane serine protease matriptase-2–identification, structural features, enzymology, expression pattern and potential roles
Front. Biosci.
 , 
2008
, vol. 
13
 (pg. 
569
-
579
)
14
Silvestri
L.
Pagani
A.
Nai
A.
De Domenico
I.
Kaplan
J.
Camaschella
C.
The serine protease matriptase-2 (TMPRSS6) inhibits hepcidin activation by cleaving membrane hemojuvelin
Cell Metab.
 , 
2008
, vol. 
8
 (pg. 
502
-
511
)
15
Ramsay
A.J.
Hooper
J.D.
Folgueras
A.R.
Velasco
G.
Lopez-Otin
C.
Matriptase-2 (TMPRSS6): a proteolytic regulator of iron homeostasis
Haematologica
 , 
2009
, vol. 
94
 (pg. 
840
-
849
)
16
Nemeth
E.
Tuttle
M.S.
Powelson
J.
Vaughn
M.B.
Donovan
A.
Ward
D.M.
Ganz
T.
Kaplan
J.
Hepcidin regulates cellular iron efflux by binding to ferroportin and inducing its internalization
Science
 , 
2004
, vol. 
306
 (pg. 
2090
-
2093
)
17
Folgueras
A.R.
de Lara
F.M.
Pendas
A.M.
Garabaya
C.
Rodriguez
F.
Astudillo
A.
Bernal
T.
Cabanillas
R.
Lopez-Otin
C.
Velasco
G.
Membrane-bound serine protease matriptase-2 (Tmprss6) is an essential regulator of iron homeostasis
Blood
 , 
2008
, vol. 
112
 (pg. 
2539
-
2545
)
18
Du
X.
She
E.
Gelbart
T.
Truksa
J.
Lee
P.
Xia
Y.
Khovananth
K.
Mudd
S.
Mann
N.
Moresco
E.M.
, et al.  . 
The serine protease TMPRSS6 is required to sense iron deficiency
Science
 , 
2008
, vol. 
320
 (pg. 
1088
-
1092
)
19
Roetto
A.
Papanikolaou
G.
Politou
M.
Alberti
F.
Girelli
D.
Christakis
J.
Loukopoulos
D.
Camaschella
C.
Mutant antimicrobial peptide hepcidin is associated with severe juvenile hemochromatosis
Nat. Genet.
 , 
2003
, vol. 
33
 (pg. 
21
-
22
)
20
Feder
J.N.
Gnirke
A.
Thomas
W.
Tsuchihashi
Z.
Ruddy
D.A.
Basava
A.
Dormishian
F.
Domingo
R.
Jr
Ellis
M.C.
Fullan
A.
, et al.  . 
A novel MHC class I-like gene is mutated in patients with hereditary haemochromatosis
Nat. Genet.
 , 
1996
, vol. 
13
 (pg. 
399
-
408
)
21
Camaschella
C.
Roetto
A.
Cali
A.
De Gobbi
M.
Garozzo
G.
Carella
M.
Majorano
N.
Totaro
A.
Gasparini
P.
The gene TFR2 is mutated in a new type of haemochromatosis mapping to 7q22
Nat. Genet.
 , 
2000
, vol. 
25
 (pg. 
14
-
15
)
22
Papanikolaou
G.
Samuels
M.E.
Ludwig
E.H.
MacDonald
M.L.
Franchini
P.L.
Dube
M.P.
Andres
L.
MacFarlane
J.
Sakellaropoulos
N.
Politou
M.
, et al.  . 
Mutations in HFE2 cause iron overload in chromosome 1q-linked juvenile hemochromatosis
Nat. Genet.
 , 
2004
, vol. 
36
 (pg. 
77
-
82
)
23
Nemeth
E.
Ganz
T.
Regulation of iron metabolism by hepcidin
Annu. Rev. Nutr.
 , 
2006
, vol. 
26
 (pg. 
323
-
342
)
24
Finberg
K.E.
Heeney
M.M.
Campagna
D.R.
Aydinok
Y.
Pearson
H.A.
Hartman
K.R.
Mayo
M.M.
Samuel
S.M.
Strouse
J.J.
Markianos
K.
, et al.  . 
Mutations in TMPRSS6 cause iron-refractory iron deficiency anemia (IRIDA)
Nat. Genet.
 , 
2008
, vol. 
40
 (pg. 
569
-
571
)
25
Melis
M.A.
Cau
M.
Congiu
R.
Sole
G.
Barella
S.
Cao
A.
Westerman
M.
Cazzola
M.
Galanello
R.
A mutation in the TMPRSS6 gene, encoding a transmembrane serine protease that suppresses hepcidin production, in familial iron deficiency anemia refractory to oral iron
Haematologica
 , 
2008
, vol. 
93
 (pg. 
1473
-
1479
)
26
Guillem
F.
Lawson
S.
Kannengiesser
C.
Westerman
M.
Beaumont
C.
Grandchamp
B.
Two nonsense mutations in the TMPRSS6 gene in a patient with microcytic anemia and iron deficiency
Blood
 , 
2008
, vol. 
112
 (pg. 
2089
-
2091
)
27
Macao
B.
Johansson
D.G.
Hansson
G.C.
Hard
T.
Autoproteolysis coupled to protein folding in the SEA domain of the membrane-bound MUC1 mucin
Nat. Struct. Mol. Biol.
 , 
2006
, vol. 
13
 (pg. 
71
-
76
)
28
Silvestri
L.
Guillem
F.
Pagani
A.
Nai
A.
Oudin
C.
Silva
M.
Toutain
F.
Kannengiesser
C.
Beaumont
C.
Camaschella
C.
, et al.  . 
Molecular mechanisms of the defective hepcidin inhibition in TMPRSS6 mutations associated with iron-refractory iron deficiency anemia
Blood
 , 
2009
, vol. 
113
 (pg. 
5605
-
5608
)
29
Ng
P.C.
Henikoff
S.
SIFT: Predicting amino acid changes that affect protein function
Nucleic Acids Res.
 , 
2003
, vol. 
31
 (pg. 
3812
-
3814
)
30
Ferrer-Costa
C.
Gelpi
J.L.
Zamakola
L.
Parraga
I.
de la Cruz
X.
Orozco
M.
PMUT: a web-based tool for the annotation of pathological mutations on proteins
Bioinformatics
 , 
2005
, vol. 
21
 (pg. 
3176
-
3178
)
31
Martini
A.
Ravelli
A.
Di Fuccia
G.
Rosti
V.
Cazzola
M.
Barosi
G.
Intravenous iron therapy for severe anaemia in systemic-onset juvenile chronic arthritis
Lancet
 , 
1994
, vol. 
344
 (pg. 
1052
-
1054
)
32
Lee
M.S.
Tseng
I.C.
Wang
Y.
Kiyomiya
K.
Johnson
M.D.
Dickson
R.B.
Lin
C.Y.
Autoactivation of matriptase in vitro: requirement for biomembrane and LDL receptor domain
Am. J. Physiol. Cell. Physiol.
 , 
2007
, vol. 
293
 (pg. 
C95
-
C105
)
33
Lu
D.
Yuan
X.
Zheng
X.
Sadler
J.E.
Bovine proenteropeptidase is activated by trypsin, and the specificity of enteropeptidase depends on the heavy chain
J. Biol. Chem.
 , 
1997
, vol. 
272
 (pg. 
31293
-
31300
)
34
Knappe
S.
Wu
F.
Masikat
M.R.
Morser
J.
Wu
Q.
Functional analysis of the transmembrane domain and activation cleavage of human corin: design and characterization of a soluble corin
J. Biol. Chem.
 , 
2003
, vol. 
278
 (pg. 
52363
-
52370
)
35
Knappe
S.
Wu
F.
Madlansacay
M.R.
Wu
Q.
Identification of domain structures in the propeptide of corin essential for the processing of proatrial natriuretic peptide
J. Biol. Chem.
 , 
2004
, vol. 
279
 (pg. 
34464
-
34471
)
36
Oberst
M.D.
Williams
C.A.
Dickson
R.B.
Johnson
M.D.
Lin
C.Y.
The activation of matriptase requires its noncatalytic domains, serine protease domain, and its cognate inhibitor
J. Biol. Chem.
 , 
2003
, vol. 
278
 (pg. 
26773
-
26779
)
37
Cho
E.G.
Kim
M.G.
Kim
C.
Kim
S.R.
Seong
I.S.
Chung
C.
Schwartz
R.H.
Park
D.
N-terminal processing is essential for release of epithin, a mouse type II membrane serine protease
J. Biol. Chem.
 , 
2001
, vol. 
276
 (pg. 
44581
-
44589
)
38
Kim
C.
Cho
Y.
Kang
C.H.
Kim
M.G.
Lee
H.
Cho
E.G.
Park
D.
Filamin is essential for shedding of the transmembrane serine protease, epithin
EMBO Rep.
 , 
2005
, vol. 
6
 (pg. 
1045
-
1051
)
39
Matsushima
M.
Ichinose
M.
Yahagi
N.
Kakei
N.
Tsukada
S.
Miki
K.
Kurokawa
K.
Tashiro
K.
Shiokawa
K.
Shinomiya
K.
, et al.  . 
Structural characterization of porcine enteropeptidase
J. Biol. Chem.
 , 
1994
, vol. 
269
 (pg. 
19976
-
19982
)
40
Xia
Y.
Babitt
J.L.
Sidis
Y.
Chung
R.T.
Lin
H.Y.
Hemojuvelin regulates hepcidin expression via a selective subset of BMP ligands and receptors independently of neogenin
Blood
 , 
2008
, vol. 
111
 (pg. 
5195
-
5204
)
41
Babitt
J.L.
Huang
F.W.
Xia
Y.
Sidis
Y.
Andrews
N.C.
Lin
H.Y.
Modulation of bone morphogenetic protein signaling in vivo regulates systemic iron balance
J. Clin. Invest.
 , 
2007
, vol. 
117
 (pg. 
1933
-
1939
)
42
Babitt
J.L.
Huang
F.W.
Wrighting
D.M.
Xia
Y.
Sidis
Y.
Samad
T.A.
Campagna
J.A.
Chung
R.T.
Schneyer
A.L.
Woolf
C.J.
, et al.  . 
Bone morphogenetic protein signaling by hemojuvelin regulates hepcidin expression
Nat. Genet.
 , 
2006
, vol. 
38
 (pg. 
531
-
539
)
43
Meynard
D.
Kautz
L.
Darnaud
V.
Canonne-Hergaux
F.
Coppin
H.
Roth
M.P.
Lack of the bone morphogenetic protein BMP6 induces massive iron overload
Nat. Genet.
 , 
2009
, vol. 
41
 (pg. 
478
-
481
)
44
Andriopoulos
B.
Jr
Corradini
E.
Xia
Y.
Faasse
S.A.
Chen
S.
Grgurevic
L.
Knutson
M.D.
Pietrangelo
A.
Vukicevic
S.
Lin
H.Y.
, et al.  . 
BMP6 is a key endogenous regulator of hepcidin expression and iron metabolism
Nat. Genet.
 , 
2009
, vol. 
41
 (pg. 
482
-
487
)
45
Wang
R.H.
Li
C.
Xu
X.
Zheng
Y.
Xiao
C.
Zerfas
P.
Cooperman
S.
Eckhaus
M.
Rouault
T.
Mishra
L.
, et al.  . 
A role of SMAD4 in iron metabolism through the positive regulation of hepcidin expression
Cell Metab.
 , 
2005
, vol. 
2
 (pg. 
399
-
409
)
46
Larkin
M.A.
Blackshields
G.
Brown
N.P.
Chenna
R.
McGettigan
P.A.
McWilliam
H.
Valentin
F.
Wallace
I.M.
Wilm
A.
Lopez
R.
, et al.  . 
Clustal W and Clustal X version 2.0
Bioinformatics
 , 
2007
, vol. 
23
 (pg. 
2947
-
2948
)