? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 117 Number 7 July 2007
Modulation of bone morphogenetic
protein signaling in vivo regulates
systemic iron balance
Jodie L. Babitt,1 Franklin W. Huang,2 Yin Xia,1 Yisrael Sidis,1 Nancy C. Andrews,2 and Herbert Y. Lin1
1Program in Membrane Biology and Nephrology Division, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, USA.
2Children’s Hospital Boston and Dana-Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts, USA.
Anemia of chronic disease, also known as anemia of inflamma-
tion, is prevalent in patients with many systemic diseases including
autoimmune disorders, malignancy, and chronic kidney disease (1).
In these patients intestinal iron absorption is impaired and iron
remains sequestered in reticuloendothelial cells, leading to hypofer-
remia and anemia (1). Research over the last several years implicates
hepcidin excess in the pathogenesis of this disease (1–6). A key regu-
lator of systemic iron homeostasis (7), hepcidin is secreted by the
liver (2, 8, 9) and induces internalization and degradation of the
iron exporter ferroportin in absorptive enterocytes and reticuloen-
dothelial cells, thereby inhibiting iron absorption from the intestine
and iron release from reticuloendothelial cell stores (10). Hepcidin
expression is inhibited by anemia and hypoxia, thus increasing iron
availability when needed for erythropoiesis (3). Conversely, hepci-
din expression is induced by iron loading, thus providing a feed-
back mechanism to limit further iron absorption (2–4). Hepcidin
expression is also induced by inflammatory cytokines, and this is
thought to be the mechanism underlying the impaired intestinal
iron absorption, reticuloendothelial cell iron sequestration, and
hypoferremia characteristic of anemia of chronic disease (1–6).
While hepcidin excess has a role in anemia of chronic disease, inad-
equate hepcidin expression appears to be a common pathogenic
mechanism for the iron overload disorder hereditary hemochroma-
tosis as a result of mutations in the genes encoding hepcidin (HAMP),
hemojuvelin (HFE2), HFE, or transferrin receptor 2 (TFR2) (11–21).
In patients and animal models with iron overload due to mutations
in these genes, hepcidin levels are low, thereby leading to ferroportin
overactivity, increased intestinal iron absorption, increased reticulo-
endothelial cell iron release, elevated serum iron levels, and abnormal
tissue iron deposition (11–21). Although the mechanisms by which
mutations in HFE and TFR2 lead to low hepcidin levels remain
unclear, emerging evidence suggests that hemojuvelin functions as
a coreceptor for bone morphogenetic protein (BMP) signaling (22)
and that BMP/TGF-β superfamily signaling has a role in regulating
hepcidin expression and systemic iron balance (22–24).
Members of the BMP/TGF-β superfamily, including BMPs,
TGF-βs, growth and differentiation factors (GDFs), and activins,
initiate an intracellular signaling cascade by binding to a complex
of type I and type II serine threonine kinase receptors (25). The
activated receptor complex phosphorylates intracellular Smad
proteins, which then complex with common mediator Smad4.
Smad complexes translocate to the nucleus where they modulate
gene transcription. In general, BMPs and GDFs signal via one set
of Smad proteins (Smad1, Smad5, and Smad8), while TGF-βs and
activins signal via another set (Smad2 and Smad3).
Hemojuvelin (also known as RGMc) is a member of the repul-
sive guidance molecule (RGM) family, which includes RGMa and
DRAGON (also known as RGMb) (13, 26, 27). We have recently
demonstrated that, similar to RGMa and DRAGON (28, 29),
hemojuvelin functions as a BMP coreceptor that binds directly to
BMP-2 and BMP-4 and enhances cellular responses to BMP, but
not TGF-β, ligands (22). Furthermore, BMP-2 positively regulates
hepcidin expression (22, 24), and hemojuvelin increases hepcidin
induction in response to BMP-2 (22). Hemojuvelin mutants associ-
ated with juvenile hemochromatosis have impaired BMP signaling
ability, and hepatocytes from Hfe2–/– mice demonstrate blunted
hepcidin induction in response to BMP-2 (22). This suggests that
the mechanism for iron overload in patients with hemojuvelin
mutations is a result of decreased BMP signaling in the liver lead-
ing to decreased hepcidin expression.
Further evidence supporting a role for BMP signaling in regulat-
ing hepcidin expression and iron metabolism in vivo comes from
Nonstandard?abbreviations?used: BMP, bone morphogenetic protein; GDF,
growth and differentiation factor; HJV.Fc, soluble hemojuvelin comprised of the
extracellular domain of hemojuvelin fused to the Fc portion of human IgG;
pRL-TK, Renilla luciferase vector.
Conflict?of?interest: H.Y. Lin’s laboratory received research funding from
Citation?for?this?article: J. Clin. Invest. 117:1933–1939 (2007). doi:10.1172/JCI31342.
Related Commentary, page 1755
1934?The?Journal?of?Clinical?Investigation http://www.jci.org Volume 117 Number 7 July 2007
mice with a conditional liver-specific knockout of Smad4. These
mice have low hepcidin levels and develop iron overload (23). In
that study, both BMP-4 and TGF-β1 were shown to induce hep-
cidin expression in liver cells in vitro (23). Hepcidin induction by
BMP-9 has also been described (24).
Here we test a wide array of TGF-β superfamily members to
compare their relative abilities to regulate hepcidin expression in
vitro. We also investigate the ability of BMP-2 to positively regulate
hepcidin expression and reduce serum iron levels in vivo. Finally,
we generate soluble hemojuvelin, comprised of the extracellular
domain of hemojuvelin fused to the Fc portion of human IgG
(HJV.Fc) and examine the ability of soluble hemojuvelin to inhibit
hepatic BMP signaling, decrease hepcidin expression, increase fer-
roportin expression, mobilize reticuloendothelial iron stores, and
increase serum iron levels.
Selective regulation of hepcidin by BMP/TGF-β superfamily members.
TGF-β superfamily members were tested for their ability to regu-
late hepcidin using both a hepcidin promoter reporter assay
(Figure 1A) and quantitative real-time RT-PCR (Figure 1B) in Hep3B
hepatoma-derived cells. Relative concentrations of BMP/TGF-β
superfamily ligands used are similar to those previously used by oth-
ers to compare responses among superfamily ligands (23, 30, 31).
BMP-2, BMP-4, BMP-5, BMP-6, BMP-7, and BMP-9 robustly increased
hepcidin promoter luciferase activity 20- to 100-fold over baseline
and increased hepcidin mRNA expression by 160- to 1,100-fold.
In contrast, TGF-β1, -β2, and -β3 increased hepcidin expression by
only 1.5- to 3-fold over baseline by both methods. BMP-3, BMP-11,
GDF-5, GDF-6, and GDF-7 showed no or comparatively little hep-
cidin induction by both methods. Activin A increased hepcidin pro-
moter relative luciferase activity by 10-fold but increased hepcidin
mRNA expression to a lesser extent relative to BMP-2, BMP-4, BMP-5,
BMP-6, BMP-7, and BMP-9 as analyzed by real-time RT-PCR. Bio-
logic activity of all ligands was verified by luciferase assay using a
BMP-responsive firefly luciferase reporter (30) and a TGF-β/activin–
responsive firefly luciferase reporter (31). Results using both meth-
ods correlated well with each other, suggesting that the hepcidin
promoter luciferase assay is a good surrogate for hepcidin mRNA
expression by quantitative real-time RT-PCR. Thus, many TGF-β
superfamily members can positively regulate hepcidin expression in
vitro; however, BMP-2, BMP-4, BMP-5, BMP-6, BMP-7, and BMP-9
are much more potent regulators of hepcidin compared with other
superfamily members, including all 3 TGF-β ligands.
BMP-2 administration in vivo increases hepcidin expression and decreases
serum iron. We next investigated whether BMP-2 regulates hepcidin
expression and iron metabolism in vivo. Purified BMP-2 at 1 mg/kg
was injected retroorbitally into mice, followed by determination of
serum iron levels and hepatic hepcidin mRNA expression 4 hours
after injection. BMP-2 administration increased hepatic hepcidin
mRNA expression 1.8-fold over mice injected with vehicle alone
(Figure 2A; P = 0.02). BMP-2 administration also decreased serum
iron levels from 170 μg/dl to 114 μg/dl (Figure 2B; P = 0.02). This
is consistent with a role for BMP-2 as a positive regulator of hep-
cidin expression in vivo.
Soluble HJV.Fc selectively inhibits BMP signaling in vitro. Soluble recep-
tors such as the soluble TNF receptor etanercept have previously
been used to inhibit ligand activity in vitro and in vivo, presumably
by binding to ligands and preventing their interaction with mem-
brane-bound receptors (32). Interestingly, soluble hemojuvelin
has been detected in human sera and has been shown to inhibit
hepcidin expression in cultured cells, although the mechanism for
this inhibition was not investigated (33). We therefore generated
purified soluble HJV.Fc (Figure 3A), the murine homolog of which
Induction of hepcidin expression by TGF-β/BMP superfamily ligands.
(A) Hep3B cells were transfected with a hepcidin promoter firefly lucif-
erase reporter and a control pRL-TK. Transfected cells were incubated
either alone (control) or with 50 ng/ml BMP or GDF ligands, 5 ng/ml
TGF-β ligands, or 30 ng/ml activin A (ActA) as indicated. Cell lysates
were analyzed for luciferase activity. To control for transfection efficien-
cy, relative luciferase activity was calculated as the ratio of firefly lucif-
erase values to Renilla luciferase values and is expressed as the fold
increase compared with control. Results are reported as the mean ± SD
(n = 2–3 per group). (B) Hep3B cells were treated with BMP, GDF,
TGF-β, or activin A ligands as in A. Total RNA was analyzed by quan-
titative real-time RT-PCR for hepcidin mRNA expression and β-actin
mRNA expression. Samples were analyzed in triplicate and are report-
ed as the ratio of the mean values of hepcidin to β-actin.
BMP-2 administration in mice increases hepcidin mRNA expression and
decreases serum iron. 129S6/SvEvTac mice were injected retroorbitally
with 1 mg/kg BMP-2 (n = 8) or an equal volume of vehicle alone (n = 7).
Four hours after injection, blood and livers were harvested. (A) Total
mRNA was isolated from livers and analyzed by quantitative real-time
RT-PCR for hepcidin mRNA expression relative to expression of GAPDH
mRNA, which was used as an internal control. (B) Serum iron was mea-
sured by colorimetric assay. Results are reported as the mean ± SD.
*P = 0.02 for BMP-2–treated mice compared with controls.
? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 117 Number 7 July 2007
we have previously shown can bind to BMP-2 and BMP-4 ligands
(22). We then investigated whether HJV.Fc inhibited basal hepcidin
expression and BMP induction of hepcidin expression in vitro. In
hepatoma-derived HepG2 cells, which have higher basal hepcidin
expression, HJV.Fc inhibited basal hepcidin mRNA expression by
80% (Figure 3B; P = 0.03). These results are consistent with a prior
report using soluble hemojuvelin that does not contain an Fc tail
(33), suggesting that the Fc tail does not affect the function of
soluble hemojuvelin. HJV.Fc also inhibited BMP-2 induction of
hepcidin expression (Figure 3C; P = 0.009) and BMP-2–induced
activation of the hepcidin promoter in a dose-dependent fash-
ion (Figure 3D). HJV.Fc inhibition of BMP ligands was selective:
HJV.Fc inhibited more than 90% of hepcidin promoter activation
induced by BMP-2, BMP-4, BMP-5, and BMP-6 but did not inhibit
BMP-9 even at lower ligand concentrations (Figure 3D). There was
a trend toward low-level inhibition of BMP-7 (Figure 3D).
We have previously demonstrated that BMP-2 and BMP-4 are
endogenously expressed in HepG2 cells (22), and we hypothesized
that inhibition of these endogenous BMP ligands was the mecha-
nism by which HJV.Fc decreased basal hepcidin expression in
HepG2 cells. We therefore used RT-PCR to investigate whether other
BMP ligands are endogenously expressed in HepG2. We then tested
whether siRNA inhibition of these endogenously expressed BMP
ligands inhibited basal hepcidin expression in a manner similar to
HJV.Fc. BMP-2, BMP-4, and BMP-6 were endogenously expressed
in HepG2 cells, with BMP-4 being the most abundant (Figure 4A).
BMP-2, BMP-4, and BMP-6 siRNA each selectively and significantly
reduced endogenous ligand expression in HepG2 cells by 65%, 90%,
and 55%, respectively, as measured by real-time RT-PCR (Figure 4B).
BMP-2, BMP-4, and BMP-6 siRNA each significantly inhibited
basal hepcidin expression in HepG2 cells by approximately 10%
(P = 0.012), 35% (P = 0.0027), and 15% (P = 0.0026), respectively, as
Soluble HJV.Fc inhibits basal hepcidin expression and selectively
inhibits BMP induction of hepcidin expression. (A) Western blot of
purified soluble HJV.Fc fusion protein with anti-hemojuvelin antibody
(α-HJV) and anti-Fc antibody (α-Fc). (B and C) HepG2 cells were incu-
bated alone (control) or with 25 μg/ml HJV.Fc alone, 25 ng/ml BMP-2
alone, or a combination of HJV.Fc and BMP-2 as indicated. Total RNA
was isolated and quantitative real-time RT-PCR for hepcidin mRNA
relative to β-actin mRNA was performed as in Figure 1. Results are
reported as the mean ± SD (n = 3 per group; *P = 0.03 for HJV.Fc
compared with control; †P = 0.009 for HJV.Fc plus BMP-2 compared
with BMP-2 alone). (D) Hep3B cells were transfected with the hepci-
din promoter luciferase construct and pRL-TK. Transfected cells were
incubated alone, with 5 ng/ml BMP-9, 50 ng/ml BMP-5, or 25 ng/ml
BMP-2, BMP-4, BMP-6, or BMP-7 ligands, or with the BMP ligands
plus 0.2 to 25 μg/ml HJV.Fc as indicated, followed by measurement of
relative luciferase activity as in Figure 1. Results are reported as the
mean ± SD of the percent decrease in relative luciferase activity for
cells treated with BMP ligands in combination with HJV.Fc compared
with cells treated with BMP ligands alone (n = 2 per group).
siRNA inhibition of endogenous BMP ligands decreases basal hepcidin expression. (A) Expression of endogenous BMP ligands in HepG2 cells as
measured by RT-PCR. Purified plasmid cDNAs expressing BMP ligands were used as positive controls. (B and C) HepG2 cells were transfected with
BMP ligand siRNAs or a control scrambled siRNA as indicated. Total RNA was analyzed for BMP ligand expression (B) or hepcidin expression (C)
relative to β-actin expression by real-time quantitative RT-PCR. Results are reported as the mean ± SD of the percent decrease in the ratio of hepcidin
or BMP ligand to β-actin for cells treated with various BMP siRNAs compared with cells treated with control siRNA; n = 3–6 per group; *P < 0.05.
1936?The?Journal?of?Clinical?Investigation http://www.jci.org Volume 117 Number 7 July 2007
measured by real-time RT-PCR (Figure 4C). As a negative control,
neither a control siRNA nor a BMP-7–specific siRNA inhibited basal
hepcidin expression. The relative ability of each ligand to inhibit
basal hepcidin correlated with the relative mRNA abundance of the
ligand and the strength of siRNA inhibition of ligand expression.
These data suggest that endogenous BMP-2, BMP-4, and BMP-6
ligands all contribute to basal hepcidin expression in HepG2 cells.
These data are consistent with our hypothesis that the mechanism
by which HJV.Fc inhibits basal hepcidin expression in these cells is
by inhibiting endogenous BMP signaling, presumably by binding
and sequestering endogenously produced BMP ligands and prevent-
ing their interaction with BMP type I and type II receptors.
Soluble HJV.Fc inhibits hepatic BMP signaling, inhibits hepcidin expres-
sion, increases ferroportin expression, mobilizes reticuloendothelial cell iron
stores, and increases serum iron in vivo. To test whether HJV.Fc admin-
istration could regulate hepcidin expression and iron metabolism
in vivo, mice were injected with 25 mg/kg purified HJV.Fc or an
equal volume of normal saline by i.p. injection 3 times weekly
for 3 weeks. Western blot analysis of liver lysates from these mice
showed decreased phosphorylated Smad1, Smad5, and Smad8
expression relative to total Smad1 expression in HJV.Fc-treated
mice compared with control mice (Figure 5A; P = 0.0497), dem-
onstrating that HJV.Fc decreases hepatic BMP signaling in vivo.
Quantitative real-time RT-PCR analysis revealed a 10-fold decrease
in hepatic hepcidin mRNA expression in HJV.Fc-treated mice com-
pared with control mice (Figure 5B; P = 0.003). Consistent with
the predicted effects of depressed hepcidin levels to increase ferro-
portin cell surface expression, increase intestinal iron absorption,
and increase release of iron from reticuloendothelial stores, HJV.Fc
treatment increased ferroportin expression in the spleen compared
with control mice, as measured by western blot (Figure 5C). HJV.Fc
treatment also increased serum iron levels from 177 ± 26 μg/dl to
309 ± 2 μg/dl (Figure 5D; P = 0.01) and increased serum transferrin
saturation from 70% to 100% (Figure 5E; P = 0.004). Furthermore,
HJV.Fc treatment increased hepatic tissue iron content by approxi-
mately 2-fold (Figure 5F; P = 0.03) and reduced splenic tissue iron
content by almost 60% (Figure 5G; P = 0.009).
Soluble HJV.Fc inhibits IL-6 induction of hepcidin expression. Inflam-
matory cytokines induce hepcidin expression, and this hepcidin
excess is thought to play a role in the anemia of chronic disease
(1–6). We therefore investigated whether HJV.Fc could inhibit hep-
cidin induction by the inflammatory cytokine IL-6. IL-6 increased
hepcidin expression 3.3-fold in HepG2 cells as measured by real-
time RT-PCR (Figure 6; P = 0.003). Hepcidin induction by IL-6 was
significantly abrogated when cells were incubated with HJV.Fc in
combination with IL-6 (Figure 6; P = 0.0006 compared with cells
treated with IL-6 alone).
Hepcidin deficiency is the common pathogenic mechanism for both
juvenile and adult forms of the genetic iron overload disorder hered-
itary hemochromatosis as a result of mutations in HAMP, HFE2,
TFR2, and HFE (11–21). We have previously shown that hemojuv-
elin acts as a coreceptor for BMP signaling and that BMP-2 signaling
induces hepcidin expression in vitro (22). Here we show that BMP-2
administration in mice increases hepcidin expression and reduces
serum iron levels in vivo. These data support our in vitro data that
BMP-2 can positively regulate hepcidin expression in vivo. The mod-
est induction of hepcidin expression in response to BMP-2 in vivo
compared with our in vitro findings is likely multifactorial. First,
the mice were maintained on a standard diet, where dietary iron is
replete, basal hepcidin levels are generally high, and hepcidin induc-
tion by well established regulators such as iron and LPS have been
reported to be absent or less robust compared with mice maintained
on an iron-deficient diet (4). Indeed, the degree of hepcidin induction
by BMP-2 in our study was similar to the 1.8-fold induction reported
Soluble HJV.Fc administration in mice decreases hepatic phosphorylated Smad1, Smad5, and Smad8 expression, decreases hepcidin expres-
sion, increases serum iron, increases liver iron content, and decreases spleen iron content. 129S6/SvEvTac mice received an i.p. injection of
25 mg/kg HJV.Fc or normal saline (control) 3 times weekly for 3 weeks. (A) Liver lysates were analyzed for phosphorylated Smad1, Smad5,
and Smad8 (α–p-Smad1/5/8) expression by western blot. Blots were stripped and reprobed for expression of total Smad1 and β-actin, which
were used as loading controls. Chemiluminescence was quantitated by IPLab Spectrum software for phosphorylated Smad1, Smad5, and
Smad8 expression relative to total Smad1 expression. (B) Total mRNA was isolated from livers and analyzed by quantitative real-time PCR
for hepcidin mRNA expression relative to GAPDH mRNA expression as an internal control. (C) Spleen membrane preparations were analyzed
for ferroportin expression by western blot. Blots were stripped and reprobed for expression of β-actin, which was used as a loading control.
(D and E) Measurement of serum iron (D) and transferrin saturation (Serum Tf sat; E). (F and G) Quantitation of liver (F) and spleen (G) tissue
iron content. Results are expressed as mean ± SD, n = 3 mice per group; *P = 0.0497, †P = 0.003, ‡P = 0.01, §P = 0.004, ¶P = 0.03, #P = 0.009
for HJV.Fc-treated mice compared with controls.
? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 117 Number 7 July 2007
after LPS administration in mice fed on a standard diet (34). Further-
more, BMPs typically act in an autocrine or paracrine manner in vivo,
while i.v. administered BMP-2 has been documented to be rapidly
eliminated from the systemic circulation (t1/2 = 16 min) (35). Thus, it
is likely that the systemically administered BMP-2 dose was not deliv-
ered efficiently to the liver. Nevertheless, the decrease in serum iron
suggests that the BMP-2–induced increase in hepcidin expression
was physiologically relevant and presumably reflects decreased iron
export from reticuloendothelial cells and duodenal enterocytes due
to hepcidin-induced internalization and degradation of ferroportin.
Although treatment with systemic BMP-2 itself may be impractical
due to its high cost and rapid elimination from the systemic circu-
lation, our data suggest that therapies that enhance hepatic BMP
signaling may provide alternative treatment strategies for managing
iron overload in patients with hereditary hemochromatosis.
Liver-specific conditional Smad4 knockout mice have reduced
hepcidin expression and total body iron overload, underscoring the
important role for TGF-β/BMP superfamily members in regulat-
ing hepcidin expression and iron metabolism in vivo (23). While
our data demonstrate that BMP-2 can positively regulate hepcidin
expression in vivo, it remains unknown which superfamily ligands
are the endogenous regulators of hepcidin expression, since Smad4
is the common downstream mediator for all superfamily members.
Indeed, many superfamily members are endogenously expressed in
the liver, including BMP-2, BMP-4, BMP-5, BMP-6, BMP-9, and all
3 TGF-β ligands (36–40 and J.L. Babitt, unpublished observations).
Whether the liver is in fact the source of the endogenous BMP super-
family ligands, however, remains uncertain. Our previous data show-
ing that hemojuvelin is a coreceptor for BMP, but not TGF-β, sig-
naling (22) suggests that members of the BMP subfamily are more
important than members of the TGF-β subfamily for regulating
iron metabolism in vivo. Here we show that many members of the
TGF-β superfamily can induce hepcidin mRNA expression in vitro.
However, our results suggest that a subset of BMP ligands including
BMP-2, BMP-4, BMP-5, BMP-6, BMP-7, and BMP-9 are much more
potent inducers of hepcidin expression than other ligands tested,
including all 3 TGF-β ligands. Our results showing 3-fold induction
of hepcidin expression by TGF-β1 are consistent with prior find-
ings (23); however, we found that BMP-4 and BMP-9 were much
more potent inducers of hepcidin expression compared with prior
studies (23–24). This may be in part related to differences in ligand
concentration (2- to 5-fold higher in our study) and/or differences
in cell lines used. Interestingly, although BMP-9 is expressed in the
liver (40) and can robustly increase hepcidin mRNA expression in
vitro, HJV.Fc was unable to inhibit BMP-9 activation of the hepcidin
promoter. HJV.Fc also had reduced ability to inhibit BMP-7 com-
pared with BMP-2, BMP-4, BMP-5, and BMP-6 ligands, consistent
with our prior findings (22). The ability of HJV.Fc to inhibit hepci-
din expression and increase serum iron in vivo therefore suggests
that BMP-2, BMP-4, BMP-5, and/or BMP-6 are good candidates for
endogenous regulators of hepcidin expression, while BMP-9, BMP-7,
and TGF-β ligands may not be important endogenous regulators of
hepcidin. Future research will be needed to definitively determine
which TGF-β/BMP superfamily ligands function as endogenous
regulators of hepcidin expression and systemic iron balance, as well
as determine the source of these ligands in vivo.
Anemia of chronic disease is associated with hypoferremia
and reticuloendothelial cell iron sequestration. Inflammatory
cytokines are potent inducers of hepcidin expression, and hepcidin
excess is believed to have a key role in the pathogenesis of anemia
in these patients (1–6). Presumably, inhibitors of hepcidin expres-
sion would allow for increased availability of iron from the diet and
increased mobilization of iron from the spleen, thereby improving
red blood cell production and ameliorating anemia. Here we pro-
vide in vivo evidence showing that soluble HJV.Fc inhibits BMP
signaling in the liver, inhibits hepcidin expression, increases fer-
roportin protein expression, decreases splenic iron stores, and
increase serum iron levels. This suggests that HJV.Fc is a potential
new treatment for anemia associated with hepcidin excess.
We hypothesize that inhibition of hepatic BMP signaling is the
predominant mechanism by which HJV.Fc inhibits hepcidin expres-
sion and regulates systemic iron balance in vivo. Indeed, inhibition of
endogenous BMP signaling in HepG2 cells using BMP siRNAs had
a similar effect on decreasing hepcidin expression as treatment with
HJV.Fc. Furthermore, loss of TGF-β/BMP superfamily signaling in
the liver is sufficient to reduce hepcidin expression and generate iron
overload, as demonstrated in mice with a liver-specific conditional
knockout of Smad4 (23). However, hemojuvelin has been shown to
bind to the receptor neogenin, a member of the deleted in colon can-
cer (DCC) receptor group, which has been reported to have a role in
diverse functions including cell survival, axonal guidance, and cel-
lular iron uptake (41). Whether this or other mechanisms contribute
to HJV.Fc inhibition of hepcidin and alterations of systemic iron bal-
ance remains to be determined. Treatment with HJV.Fc in our study
did not appear to have any other adverse effects on mice. Indeed,
regulation of hepcidin expression and iron metabolism appears to be
the principal role for TGF-β/BMP superfamily signaling in the adult
liver in vivo, since iron overload was the predominant phenotype of
liver-specific conditional Smad4 knockout mice (23).
Recent studies suggest that inflammatory mediators such as IL-6
regulate hepcidin expression through STAT3 (42–44). Although
the relationship between the IL-6/STAT3 and BMP/SMAD sig-
naling pathways in the regulation of hepcidin expression is still
poorly understood, mice with a liver-specific conditional knock-
out of Smad4 demonstrate attenuated hepcidin induction in
response to IL-6 (23). This suggests that BMP/TGF-β superfam-
ily signaling is also necessary for hepcidin excess in inflammatory
states and that inhibition of BMP signaling with HJV.Fc might
attenuate hepcidin excess induced by inflammatory states. Here
we show that HJV.Fc inhibits hepcidin induction in response to
Soluble HJV.Fc inhibits IL-6 induction of hepcidin expression. HepG2
cells were incubated for 16 hours alone (control), with 100 ng/ml IL-6,
or with 100 ng/ml IL-6 in combination with HJV.Fc after pre-incubation
with HJV.Fc for 1 hour. Total RNA was analyzed for hepcidin expres-
sion relative to β-actin expression by quantitative real-time RT-PCR.
Results are expressed as mean ± SD, n = 3 per group; *P = 0.003 for
IL-6–treated cells compared with control cells; †P = 0.0006 for cells
treated with HJV.Fc in combination with IL-6 compared with cells treat-
ed with IL-6 alone.
1938? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 117 Number 7 July 2007
the inflammatory cytokine IL-6, consistent with prior reports
for recombinant soluble hemojuvelin (33). Taken together, our
results suggest that HJV.Fc or other inhibitors of BMP signaling
may prove to be viable treatments for anemia of chronic disease
caused by inflammation.
cDNA subcloning. To generate cDNA encoding HJV.Fc, an upstream frag-
ment of human hemojuvelin (containing a preprotrypsin leader sequence
and FLAG tag) was digested SpeI/BstEII from the plasmid FLAG-HJV (22).
A downstream fragment of human hemojuvelin that does not include the
glycophosphatidylinositol domain was amplified by PCR from the plasmid
FLAG-HJV using the primers 5′-AGAAGGTGTATCAGGCTGAGGTGG-3′
and 5′-CAGCTCGAGTGAGGGGAAGAGATGCAGCTTCTC-3′, followed
by BstEII/XhoI digestion. Both fragments were ligated into the SpeI/XhoI
sites of Signal pIg plus (R&D Systems) in-frame with the Fc portion of
human IgG. Sequences were verified by bidirectional sequencing at the
DNA sequencing core facility of Massachusetts General Hospital.
Purification of HJV.Fc. HEK293 cells (catalog no. CRL-1573; ATCC) cul-
tured in RPMI medium 1640 (Invitrogen) supplemented with l-glutamine
(Invitrogen) and 10% FBS (Atlanta Biologicals) were stably transfected
with cDNA encoding HJV.Fc using Lipofectamine 2000 (Invitrogen)
according to the manufacturer’s instructions. Stably transfected cells
were selected and cultured in 1 mg/ml geneticin (Mediatech Inc.). HJV.Fc
was purified from the conditioned media of stably transfected cells by Bio
Express Cell Culture Services.
Animals. All animal protocols were approved by the Institutional
Animal Care and Use Committee at the Massachusetts General Hospi-
tal or Children’s Hospital Boston. Six- to 8-week-old 129S6/SvEvTac
mice (Taconic) were housed in the Massachusetts General Hospital or
Children’s Hospital Boston rodent facilities and fed on either the Prolab
5P75 Isopro RMH 3000 or Prolab RMH 3000 diet (LabDiet), each with
380 parts per million iron.
BMP injection. Mice were anesthetized with avertin (Sigma-Aldrich) and
given a single retroorbital injection of either 1 mg/kg of BMP-2 (kindly pro-
vided by Vicki Rosen, Harvard School of Dental Medicine, Boston, Massa-
chusetts, USA) in 0.1% BSA in 1× PBS or an equal volume of vehicle alone.
Dosage of BMP-2 was based on published data for BMP-7 showing that doses
ranging from 0.25–1.0 mg/kg i.v. or i.p. are effective in inhibiting fibrosis and
preserving renal function in several animal models of kidney injury (45–49).
Four hours after injection the mice were sacrificed, and blood and livers were
harvested for measurement of iron parameters and hepcidin expression.
HJV.Fc injection. Mice were injected with an i.p. dose of 25 mg/kg purified
HJV.Fc or an equal volume of normal saline 3 times per week for 3 weeks.
Twenty-four hours after the last injection, mice were sacrificed, and blood,
livers, and spleen were harvested for measurement of iron parameters,
phosphorylated Smad1, Smad5, and Smad8, and hepcidin expression.
Serum iron measurements. Blood was collected in BD Microtainer serum
separator tubes (Fischer Scientific), and serum was isolated according to
the manufacturer’s instructions. Serum iron and unsaturated iron-binding
capacity (UIBC) were measured by colorimetric assay using the Iron/UIBC
kit (Thermo Electron Corp.). Total iron-binding capacity (TIBC) was cal-
culated as the sum of serum iron and UIBC measurements, and transferrin
saturation percentage was calculated as serum iron / TIBC × 100.
Tissue iron measurement. Immediately after harvest, livers and spleen were
sectioned and weighed. Quantitative measurement of nonheme iron was
performed according to the method of Torrence and Bothwell (50). Results
are reported as micrograms iron/gram wet weight tissue.
Luciferase assay. Hepcidin promoter luciferase assays in hepatoma-
derived Hep3B cells were carried out using the Dual-Luciferase Reporter
Assay System (Promega) as previously described (22) with the following
modifications: for BMP/TGF-β stimulation assays, cells transfected with
the hepcidin promoter luciferase reporter and control Renilla lucifer-
ase vector (pRL-TK) were serum starved in α-MEM with l-glutamine
(Invitrogen) supplemented with 1% FBS for 6 hours, followed by stimu-
lation with 50 ng/ml BMP ligands, 30 ng/ml activin A, or 5 ng/ml TGF-β
ligands (R&D Systems) for 16 hours. Relative concentrations of BMP/
TGF-β superfamily ligands were similar to those previously used by oth-
ers to compare superfamily ligand responses (23, 30–31). For HJV.Fc
inhibition assays, cells transfected with the hepcidin promoter luciferase
reporter and pRL-TK were serum starved as above and incubated with
25 ng/ml BMP-2, BMP-4, BMP-6, or BMP-7 ligands, 50 ng/ml BMP-5,
or 5 ng/ml BMP-9 either alone or with 0.5–25.0 μg/ml of HJV.Fc for
16 hours. Relative concentrations of BMP ligands were chosen to elicit
similar degrees of hepcidin promoter relative luciferase activity. Experi-
ments using equal concentrations of ligands were also carried out and
had similar results (data not shown).
RT-PCR. Total RNA was isolated from HepG2 or Hep3B cells and was
analyzed for BMP2, BMP4, BMP5, BMP6, and BMP9 expression as previ-
ously described (29) using the primers BMP-2 forward and reverse, BMP-4
forward and reverse, BMP-6 forward and reverse, and BMP-9 forward and
reverse (see Supplemental Table 1 for primer sequences; supplemental data
available online with this article; doi:10.1172/31342DS1).
Quantitative real-time RT-PCR. Hep3B or HepG2 cells were serum
starved for 6 hours in α-MEM supplemented with 1% FBS and treated
for 16 hours with varying amounts of BMP/TGF-β superfamily ligands
or 100 ng/ml IL-6 in the absence or presence of 25 μg/ml purified HJV.
Fc. For BMP siRNA experiments, HepG2 cells were plated in 24-well
plates and transfected with 200 ng pcDNA3 (Invitrogen) and 40 nM
BMP-2, BMP-4, BMP-6, BMP-7, or control scramble siRNA (Ambion;
see Supplemental Table 2 for siRNA sequences) in α-MEM using Lipo-
fectamine 2000 (Invitrogen) according to the manufacturer’s instruc-
tions. Cells were serum starved overnight in α-MEM supplemented with
0.1% BSA. Total RNA was isolated from treated cells, and real-time quan-
titation of hepcidin relative to β-actin mRNA transcripts was performed
using 2-step quantitative real-time RT-PCR as previously described (22).
For BMP siRNA experiments, real-time quantitation of BMP2, BMP4,
and BMP6 relative to β-actin mRNA transcripts was also performed as
described above using the primers qBMP-2 forward, qBMP-2 reverse,
qBMP-4 forward, qBMP-4 reverse, qBMP-6 forward, and qBMP-6 reverse
(see Supplemental Table 3 for primer sequences). For mouse livers, total
RNA was isolated using the Illustra RNAspin Mini Kit (GE Healthcare)
according to the manufacturer’s instructions. Real-time quantification
of hepcidin (Hamp1) relative to Gapdh mRNA transcripts was performed
as described above using primers Hamp1 forward (6) Hamp1 reverse
(6), Gapdh forward, and Gapdh reverse (see Supplemental Table 3 for
Western blot. Western blot of purified HJV.Fc using anti-hemojuvelin
antibody (22) and anti-Fc antibody (Jackson ImmunoResearch Labo-
ratories) was performed as previously described (22). Western blot of
liver lysates for phosphorylated Smad1, Smad5, and Smad8 expression
relative to total Smad1 and β-actin expression was performed as previ-
ously described (22). Chemiluminescence was quantitated using IPLab
Spectrum software version 3.9.5 r2 (Scanalytics). For ferroportin assays,
spleen membrane preparations were prepared as previously described
(51). Protein concentrations were determined by BCA assay (Pierce). After
solubilization in 1× Laemmli buffer for 30 minutes at room temperature,
35 μg of protein per sample were separated by SDS-PAGE using pre-cast
NuPAGE Novex 4–12% Bis-Tris gels (Invitrogen) and transferred onto
PDVF membranes. Western blot was performed using anti-ferroportin
? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 117 Number 7 July 2007
antibody (kindly donated by Francois Canonne-Hergaux) as previously
described (52). Blots were stripped and reprobed for β-actin expression as
a loading control as previously described (22).
Statistics. A 2-tailed Student’s t test with P < 0.05 was used to determine
J.L. Babitt was supported in part by NIH grant K08 DK-075846.
F.W. Huang was supported in part by NIH grant T32 HL07623.
N.C. Andrews was supported in part by NIH grant RO1
DK-053813. H.Y. Lin was supported in part by NIH grants RO1
DK-69533 and RO1 DK-71837 and a grant from the Roche
Foundation for Anemia Research.
Received for publication December 26, 2006, and accepted in
revised form April 10, 2007.
Address correspondence to: Jodie L. Babitt, Program in Mem-
brane Biology, Division of Nephrology, Massachusetts General
Hospital, 185 Cambridge St., CPZN-8206, Boston, Massachu-
setts 02114, USA. Phone: (617) 643-3181; Fax: (617) 643-3182;
1. Weiss, G., and Goodnough, L.T. 2005. Anemia of
chronic disease. N. Engl. J. Med. 352:1011–1023.
2. Pigeon, C., et al. 2001. A new mouse liver-specific
gene, encoding a protein homologous to human
antimicrobial peptide hepcidin, is overexpressed
during iron overload. J. Biol. Chem. 276:7811–7819.
3. Nicolas, G., et al. 2002. The gene encoding the
iron regulatory peptide hepcidin is regulated by
anemia, hypoxia, and inflammation. J. Clin. Invest.
4. Nemeth, E., et al. 2004. IL-6 mediates hypoferremia
of inflammation by inducing the synthesis of the
iron regulatory hormone hepcidin. J. Clin. Invest.
5. Nemeth, E., et al. 2003. Hepcidin, a putative media-
tor of anemia of inflammation, is a type II acute-
phase protein. Blood. 101:2461–2463.
6. Lee, P., Peng, H., Gelbart, T., Wang, L., and Beutler,
E. 2005. Regulation of hepcidin transcription by
interleukin-1 and interleukin-6. Proc. Natl. Acad.
Sci. U. S. A. 102:1906–1910.
7. Hentze, M.W., Muckenthaler, M.U., and Andrews,
N.C. 2004. Balancing acts: molecular control of
mammalian iron metabolism. Cell. 117:285–297.
8. Krause, A., et al. 2000. LEAP-1, a novel highly disul-
fide-bonded human peptide, exhibits antimicrobial
activity. FEBS Lett. 480:147–150.
9. Park, C.H., Valore, E.V., Waring, A.J., and Ganz, T.
2001. Hepcidin, a urinary antimicrobial peptide syn-
thesized in the liver. J. Biol. Chem. 276:7806–7810.
10. Nemeth, E., et al. 2004. Hepcidin regulates cellular
iron efflux by binding to ferroportin and inducing
its internalization. Science. 306:2090–2093.
11. Pietrangelo, A. 2006. Hereditary hemochromatosis.
Biochim. Biophys. Acta. 1763:700–710.
12. Roetto, A., et al. 2003. Mutant antimicrobial
peptide hepcidin is associated with severe juvenile
hemochromatosis. Nat. Genet. 33:21–22.
13. Papanikolaou, G., et al. 2004. Mutations in HFE2
cause iron overload in chromosome 1q-linked juve-
nile hemochromatosis. Nat. Genet. 36:77–82.
14. Huang, F.W., Pinkus, J.L., Pinkus, G.S., Fleming, M.D.,
and Andrews, N.C. 2005. A mouse model of juvenile
hemochromatosis. J. Clin. Invest. 115:2187–2191.
15. Niederkofler, V., Salie, R., and Arber, S. 2005.
Hemojuvelin is essential for dietary iron sensing,
and its mutation leads to severe iron overload.
J. Clin. Invest. 115:2180–2186. doi:10.1172/JCI25683.
16. Ahmad, K.A., et al. 2002. Decreased liver hepcidin
expression in the Hfe knockout mouse. Blood Cells
Mol. Dis. 29:361–366.
17. Bridle, K.R., et al. 2003. Disrupted hepcidin regu-
lation in HFE-associated haemochromatosis and
the liver as a regulator of body iron homoeostasis.
18. Muckenthaler, M., et al. 2003. Regulatory defects in
liver and intestine implicate abnormal hepcidin and
Cybrd1 expression in mouse hemochromatosis.
Nat. Genet. 34:102–107.
19. Nicolas, G., et al. 2003. Constitutive hepcidin
expression prevents iron overload in a mouse
model of hemochromatosis. Nat. Genet. 34:97–101.
20. Kawabata, H., et al. 2005. Expression of hepcidin is
down-regulated in TfR2 mutant mice manifesting
a phenotype of hereditary hemochromatosis. Blood.
21. Nemeth, E., Roetto, A., Garozzo, G., Ganz, T., and
Camaschella, C. 2005. Hepcidin is decreased in
TFR2 hemochromatosis. Blood. 105:1803–1806.
22. Babitt, J.L., et al. 2006. Bone morphogenetic pro-
tein signaling by hemojuvelin regulates hepcidin
expression. Nat. Genet. 38:531–539.
23. Wang, R.H., et al. 2005. A role of SMAD4 in iron
metabolism through the positive regulation of
hepcidin expression. Cell Metab. 2:399–409.
24. Truksa, J., Peng, H., Lee, P., and Beutler, E. 2006.
Bone morphogenetic proteins 2, 4, and 9 stimulate
murine hepcidin 1 expression independently of
Hfe, transferrin receptor 2 (Tfr2), and IL-6. Proc.
Natl. Acad. Sci. U. S. A. 103:10289–10293.
25. Shi, Y., and Massague, J. 2003. Mechanisms of TGF-
beta signaling from cell membrane to the nucleus.
26. Monnier, P.P., et al. 2002. RGM is a repulsive guidance
molecule for retinal axons. Nature. 419:392–395.
27. Samad, T.A., et al. 2004. DRAGON: a member of
the repulsive guidance molecule-related family of
neuronal- and muscle-expressed membrane pro-
teins is regulated by DRG11 and has neuronal
adhesive properties. J. Neurosci. 24:2027–2036.
28. Samad, T.A., et al. 2005. DRAGON, a bone mor-
phogenetic protein co-receptor. J. Biol. Chem.
29. Babitt, J.L., et al. 2005. Repulsive guidance mol-
ecule (RGMa), a DRAGON homologue, is a bone
morphogenetic protein co-receptor. J. Biol. Chem.
30. Korchynskyi, O., and ten Dijke, P. 2002. Identifi-
cation and functional characterization of distinct
critically important bone morphogenetic protein-
specific response elements in the Id1 promoter.
J. Biol. Chem. 277:4883–4891.
31. Dennler, S., et al. 1998. Direct binding of Smad3
and Smad4 to critical TGF beta-inducible elements
in the promoter of human plasminogen activator
inhibitor-type 1 gene. EMBO J. 17:3091–3100.
32. Moreland, L.W., et al. 1997. Treatment of rheuma-
toid arthritis with a recombinant human tumor
necrosis factor receptor (p75)-Fc fusion protein.
N. Engl. J. Med. 337:141–147.
33. Lin, L., Goldberg, Y.P., and Ganz, T. 2005. Competi-
tive regulation of hepcidin mRNA by soluble and
cell-associated hemojuvelin. Blood. 106:2884–2889.
34. Roy, C.N., et al. 2004. An Hfe-dependent pathway
mediates hyposideremia in response to lipopolysac-
charide-induced inflammation in mice. Nat. Genet.
35. Wyeth Pharmaceuticals Inc. 2004. Summary of
safety and effectiveness data for INFUSE bone graft.
36. De Bleser, P.J., Niki, T., Rogiers, V., and Geerts, A.
1997. Transforming growth factor-beta gene expres-
sion in normal and fibrotic rat liver. J. Hepatol.
37. Kingsley, D.M. 1994. What do BMPs do in mam-
mals? Clues from the mouse short-ear mutation.
Trends Genet. 10:16–21.
38. Knittel, T., Fellmer, P., Muller, L., and Ramadori, G.
1997. Bone morphogenetic protein-6 is expressed
in nonparenchymal liver cells and upregulated by
transforming growth factor-beta 1. Exp. Cell Res.
39. Su, A.I., et al. 2002. Large-scale analysis of the
human and mouse transcriptomes. Proc. Natl. Acad.
Sci. U. S. A. 99:4465–4470.
40. Miller, A.F., Harvey, S.A., Thies, R.S., and Olson,
M.S. 2000. Bone morphogenetic protein-9. An
autocrine/paracrine cytokine in the liver. J. Biol.
41. Zhang, A.S., West, A.P., Jr. Wyman, A.E., Bjorkman,
P.J., and Enns, C.A. 2005. Interaction of hemojuv-
elin with neogenin results in iron accumulation in
human embryonic kidney 293 cells. J. Biol. Chem.
42. Wrighting, D.M., and Andrews, N.C. 2006. Interleu-
kin-6 induces hepcidin expression through STAT3.
43. Verga Falzacappa, M.V., et al. 2007. STAT3 medi-
ates hepatic hepcidin expression and its inflamma-
tory stimulation. Blood. 109:353–358.
44. Pietrangelo, A., et al. 2007. STAT3 is required for
IL-6-gp130-dependent activation of hepcidin in
vivo. Gastroenterology. 132:294–300.
45. Bosukonda, D., Shih, M.S., Sampath, K.T., and
Vukicevic, S. 2000. Characterization of receptors
for osteogenic protein-1/bone morphogenetic
protein-7 (OP-1/BMP-7) in rat kidneys. Kidney Int.
46. Hruska, K.A., et al. 2000. Osteogenic protein-1 prevents
renal fibrogenesis associated with ureteral obstruction.
Am. J. Physiol. Renal Physiol. 279:F130–F143.
47. Morrissey, J., et al. 2002. Bone morphogenetic
protein-7 improves renal fibrosis and accelerates
the return of renal function. J. Am. Soc. Nephrol.?
48. Vukicevic, S., et al. 1998. Osteogenic protein-1
(bone morphogenetic protein-7) reduces severity
of injury after ischemic acute renal failure in rat.
J. Clin. Invest. 102:202–214.
49. Zeisberg, M., et al. 2003. Bone morphogenic pro-
tein-7 inhibits progression of chronic renal fibrosis
associated with two genetic mouse models. Am. J.
Physiol. Renal Physiol. 285:F1060–F1067.
50. Torrance, J.D., and Bothwell, T.H. 1980. Tissue iron
stores. Methods Hematol. 1:90–115.
51. Canonne-Hergaux, F., Gruenheid, S., Ponka, P., and
Gros, P. 1999. Cellular and subcellular localization
of the Nramp2 iron transporter in the intestinal
brush border and regulation by dietary iron. Blood.
52. Canonne–Hergaux, F., Donovan, A., Delaby, C.,
Hui-jun, W., and Gros, P. 2005. Comparative
studies of duodenal and macrophage ferroportin
proteins. Am. J. Physiol. Gastrointest. Liver Physiol.