a-1 Antitrypsin binds preprohepcidin intracellularly and
prohepcidin in the serum
Edina Pandur
1
, Judit Nagy
1,2
, Viktor S. Poo
´
r
1,2
,A
´
kos Sarnyai
2
, Andra
´
s Husza
´
r
1
, Attila Miseta
2
and
Katalin Sipos
1
1 Department of Forensic Medicine, University of Pe
´
cs, Hungary
2 Institute of Laboratory Medicine, University of Pe
´
cs, Hungary

Tel: +36 72 536 230
E-mail:
(Received 13 November 2008, revised 22
January 2009, accepted 26 January 2009)
doi:10.1111/j.1742-4658.2009.06937.x
Recent discoveries have indicated that the hormone hepcidin plays a major
role in the control of iron homeostasis. Hepcidin regulates the iron level in
the blood through the interaction with ferroportin, an iron exporter
molecule, causing its internalization and degradation. As a result, hepcidin
increases cellular iron sequestration, and decreases the iron concentration
in the plasma. Only mature hepcidin (result of the cleavage of prohepcidin
by furin proteases) has biological activity; however, prohepcidin, the
prohormone form, is also present in the plasma. In this study, we aimed to
identify new protein–protein interactions of preprohepcidin, prohepcidin
and hepcidin using the BacterioMatch two-hybrid system. Screening assays
were carried out on a human liver cDNA library. Preprohepcidin screening
gave the following results: a-1 antitrypsin, transthyretin and a-1-acid
glycoprotein showed strong interactions with preprohepcidin. We further
confirmed and examined the a-1 antitrypsin binding in vitro (glutathione
S-transferase, pull down, coimmunoprecipitation, MALDI-TOF) and
in vivo (ELISA, cross-linking assay). Our results demonstrated that the
serine protease inhibitor a-1 antitrypsin binds preprohepcidin within the
cell during maturation. Furthermore, a-1 antitrypsin binds prohepcidin
significantly in the plasma. This observation may explain the presence of
prohormone in the circulation, as well as the post-translational regulation
of the mature hormone level in the blood. In addition, the lack of cleavage
protection in patients with a-1 antitrypsin deficiency may be the reason for
the disturbance in their iron homeostasis.
Abbreviations
A1AT, a-1 antitrypsin; AA, amino acid; CTCK, carbenicillin–tetracycline–chloramphenicol–kanamycin; CTKXi, chloramphenicol–tetracycline–

tion, the hormone causes the reduction of DMT1
(divalent metal ion transporter 1) expression [20]. As a
result, hepcidin increases cellular iron sequestration in
hepatocytes and macrophages, and reduces the iron
level in the plasma. The known signals for the induc-
tion of hepcidin synthesis are the elevation of the
plasma iron level, inflammation and bacterial invasions
[21–25].
To date, the only proven interaction of hepcidin is
with the iron exporter molecule ferroportin. We were
interested in whether we could identify new protein–
protein interactions of preprohepcidin, prohepcidin
and hepcidin in vivo. For these experiments, we used
the BacterioMatch system, a two-hybrid screening
assay system developed in bacteria. The most consis-
tent and strongest interaction occurred with the serine
protease inhibitor a-1 antitrypsin (A1AT). This associ-
ation was further tested by both in vivo and in vitro
methods to evaluate its significance.
Results
In vivo interactions of preprohepcidin and
hepcidin with hepatocyte proteins
The reporter strain of the BacterioMatch two-hybrid
system harbours two reporter genes: lacZ and carbeni-
cillin resistance genes. These genes are transcribed by
RNA polymerase if the bait and target proteins, which
are expressed by the bait plasmid (pBT) and target
plasmid (pTRG), interact. In the case of transcrip-
tional activation, bacterial colonies will be blue on
5-bromo-4-chloroindol-3-yl-b-d-galactoside (Gal-X)

to be with A1AT, a member of the serine protease
inhibitor (serpin) family. A1AT was ‘fished out’ at the
screenings more times than any other interacting protein
(one-third of all sequenced cDNA clones), indicating
Fig. 1. Structure and maturation of preprohepcidin. The first 24 AAs serve as a signal sequence for secretion. To generate the mature
25-AA hepcidin peptide, there is a furin cleavage site in the C-terminal part of prohepcidin.
E. Pandur et al. In vivo interactions of preprohepcidin and prohepcidin
FEBS Journal 276 (2009) 2012–2021 ª 2009 The Authors Journal compilation ª 2009 FEBS 2013
a consistent and potentially relevant interaction with
preprohepcidin. However, a more abundant represen-
tation of A1AT clones, when compared with other
positive clones, cannot be excluded. The strong binding
between preprohepcidin and A1AT was confirmed when
the latter was cloned into pTRG, and cotransformed
with preprohepcidin expressing pBT into BacterioMatch
competent cells. These cells were able to grow on CTCK
plates in the presence of 500 lgÆmL
)1
carbenicillin con-
centration. As furin, a serine protease involved in the
maturation of hepcidin, is also inhibited by A1AT, we
considered this as a potentially important observation.
This cotransformation was repeated with the same
protease inhibitor expressed in pTRG, and either the
60-AA prohepcidin (without the targeting sequence) or
the 25-AA-containing mature hepcidin cloned into
pBT. We detected the growth of the cotransformed
BacterioMatch strain on carbenicillin indicator
(CTCK) plate in the case of prohepcidin (60 AA), but
not with mature hepcidin. We found that the protease

with anti-A1AT IgG. The in vitro binding of the two
molecules appeared to be specific, as GST-carrying
affinity columns produced only negligible quantities of
A1AT tethering (Fig. 2).
Hepcidin expression causes parallel alterations in
A1AT mRNA levels
Next, we studied the influence of the overexpression or
downregulation of preprohepcidin on the A1AT
mRNA level. We transfected WRL68 cells with prep-
rohepcidin ⁄ pTriex3-Neo plasmid and were able to
demonstrate a 470-fold increase in the copy number of
preprohepcidin mRNA by real-time quantitative PCR.
Using antisense RNA, we reduced the preprohepcidin
mRNA level to 63% (Fig. 3A). The same samples were
processed for A1AT mRNA level measurement. We
found that the A1AT mRNA level increased by more
than two-fold when preprohepcidin was overexpres-
sed. Even more significantly, the 37% decrease in
preprohepcidin expression caused by antisense RNA
Table 2. In vivo interactions of a-1 antitrypsin with preprohepcidin,
prohepcidin and mature hepcidin.
Insert in pBT Insert in pTRG
Growth on
CTCK plates
a
Preprohepcidin a-1 Antitrypsin +++
Prohepcidin a-1 Antitrypsin +++
Hepcidin a-1 Antitrypsin )
a
Colony growth was classified as follows: ), no growth; +++,

anti-A1AT, illustrating effective cross-linking between
the two molecules, but there was no signal after
control pTriex3-Neo plasmid transfection (Fig. 4).
Prohepcidin binds to A1AT in the serum
Next, we studied the interaction of prohepcidin and
plasma A1AT in the circulation. We carried out ultra-
filtration assays with sera collected from presumably
healthy volunteers. After measuring the prohepcidin
level with ELISA, the serum was filtered through a
30 kDa cut-off membrane and the prohepcidin level
was determined in the filtrate (first ultrafiltrate).
Prohepcidin itself did not bind to the filter of the
Microcon tube, and A1AT did not appear in the serum
ultrafiltrate (data not shown). We found that the serum
prohepcidin level was 210 lgÆL
)1
, whereas the first
Fig. 3. Changes in mRNA levels of preprohepcidin and A1AT
caused by preprohepcidin overexpression or preprohepcidin
silencing with antisense technique in cultured WRL68 cells. mRNA
levels were determined by a real-time PCR method, and expression
ratios were calculated using b-actin as reference gene. Values
represent the mean ± standard error of the mean (SEM) of three
independent experiments. (A) Preprohepcidin mRNA levels follow-
ing the two different treatments of cell cultures. (B) A1AT mRNA
levels displayed parallel changes to the amount of preprohepcidin
mRNA. *P < 0.01 versus untreated cells.
Fig. 4. Cross-linking of A1AT with preprohepcidin. Cultured Huh7
cells were transfected with His-tagged preprohepcidin-expressing
plasmid and then treated with the cross-linker DSS. Protein

)1
A1AT to
the first serum ultrafiltrate. The prohepcidin concentra-
tion in the second ultrafiltrate was further reduced to
46.6 lgÆL
)1
(22% of the total), or to 65% of the first
filtrate (Fig. 5).
To reveal the specificity of the preceding binding
reaction, we performed coimmunoprecipitation assays.
We attached A1AT antibody to a column of CNBr-
activated Sepharose beads, and incubated this with
serum. Sepharose beads were washed and A1AT-asso-
ciated proteins were eluted with Laemmli buffer. Next,
we probed the eluent with anti-hepcidin IgG. Results
of the dot blot displayed strong positive signals, indi-
cating that A1AT and prohepcidin associated in vivo
in the serum. Ultrafiltrated ‘free’ prohepcidin by itself
gave no binding to the activated Sepharose beads
(Fig. 6).
Similar affinity purification was carried out using
the ZipTip method, in which A1AT antibody was
attached to the C18 column of ZipTip and incubated
with serum, as in the previous experiment. The eluted
sample was analysed on a MALDI-TOF mass
spectrometer. The spectrum was compared with that
obtained in the case of bacterially expressed His-tagged
prohepcidin with a molecular weight of 7760.08 Da.
In the latter case, two major peaks appeared in the
spectrum, at m ⁄ z 1410.96 and 6349.12. The peak at

hormone. Indeed, data in the literature show that the
inherited mutations of A1AT are associated with
increased iron accumulation and liver disease [31]. One
of the effects of A1AT modifications is hyperferritina-
emia [32]. A possible explanation is that the mutated
protease inhibitor does not protect prohepcidin
sufficiently. Consequently, more mature hepcidin is
produced, which binds to ferroportin, causing
Prohepcidin (ng·mL
–1
)
Fig. 5. Serum ultrafiltration assay. Human serum with a known
A1AT level was centrifuged in a Microcon YM-30 tube. The
ultrafiltrate (first ultrafiltrate) was incubated with additional 1.5 gÆ L
)1
A1AT and centrifuged again (second ultrafiltrate). Prohepcidin levels
of the original serum, first and second ultrafiltrates were deter-
mined with the Hepcidin Prohormone ELISA kit. Values are
displayed as means ± standard error of the mean (SEM) of three
different experiments. *P < 0.01 versus serum; **P < 0.01 versus
first ultrafiltrate.
A
C
B
Fig. 6. Identification of A1AT–prohepcidin binding with coimmuno-
precipitation. Anti-A1AT IgG was coupled to CNBr-activated Sepha-
rose 4B beads and utilized for the purification of A1AT-associated
protein complexes from serum. Eluted proteins were analysed
by dot blotting with the application of anti-hepcidin IgG. (A) Serum
ultrafiltrate with ‘free’ (unbound) prohormone was incubated with

cytes. Consequently, the availability of A1AT for
prohepcidin may actually decrease in acute inflamma-
tion. Further studies are needed to substantiate this
hypothesis.
Additional significant interactions of preprohepcidin
involve the a-1 acid protein and transthyretin. The
former is a major plasma protein, but its physiological
functions have not yet been elucidated; the latter is a
thyroid-binding transfer protein. The blood level of
a-1 acid protein is elevated in different conditions
associated with acute and chronic inflammation [40].
Chronic inflammation is frequently associated with
tissue iron overload, as well as with anaemia [5,41,42].
Weaker interactions of the preprohormone of
unknown relevance were also found with some intra-
cellular proteins.
Fig. 7. ZipTip affinity purification and mass spectrometric analysis of A1AT-bound serum prohepcidin. ZipTip C18 with bound anti-A1AT was
applied to purify the A1AT-coupled prohepcidin from human serum. The eluted samples were analysed using a MALDI-TOF mass spectrom-
eter. (A) Bacterially expressed prohepcidin–His fusion protein was used as a prohepcidin standard in the mass spectrometric analysis. The
molecular weight of prohepcidin–His was 7760.08 Da. The peaks at m ⁄ z 1410.96 and 6349.12 correspond to two fragments of the prohepcidin–
His protein. The peak at m ⁄ z 1410.96 represents the 6· His and 5 AAs of the C-terminal end of prohepcidin (MCCKTHHHHHH). (B) Identification
of the affinity-purified prohepcidin from serum. The peak at m ⁄ z 6349.14 demonstrates the same fragmentation of prohepcidin as described
above.
E. Pandur et al. In vivo interactions of preprohepcidin and prohepcidin
FEBS Journal 276 (2009) 2012–2021 ª 2009 The Authors Journal compilation ª 2009 FEBS 2017
None of the proteins which contacted preprohepci-
din interacted with the mature form of hepcidin in
BacterioMatch screening. This further supports the
possibility that these proteins may play a role in the
protection of prohepcidin from the serine protease

sidase inhibitor (CTKXi) agar plates and incubated
overnight at 30 °C. These indicator plates were supple-
mented with chloramphenicol (34 lgÆmL
)1
), tetracycline
(15lgÆmL
)1
), kanamycin (50 lgÆmL
)1
), Gal-X (80 lgÆ mL
)1
)
and 0.2 mm b-galactosidase inhibitor (i). Plasmids pro-
vided in the kit were used as positive control; recombinant
pBT cotransformed with positive control pTRG+ was
used as negative control.
Colonies that appeared blue on Gal-X-containing indica-
tor plates were streaked onto LB–CTCK agar plates
(containing 250–500 lgÆmL
)1
carbenicillin instead of
Gal-X) for assay validation. The plates were incubated at
30 °C overnight and the growth rates of the colonies from
screening were compared with the growth rates of controls.
Protein–protein interaction validation
Single colonies were taken from LB–CTCK agar plates and
inoculated into 10 mL of LB supplemented with TCK. The
cultures were incubated at 30 °C overnight with shaking at
140 r.p.m. Recombinant pBT and cDNA containing pTRG
were isolated using a QIAprep Spin Miniprep Kit (Qiagen

carbenicillin.
GST fusion protein binding assay
The preprohepcidin coding cDNA was cloned into pGex4T-1
(expression of preprohepcidin was demonstrated by western
blot, applying anti-hepcidin IgG; Fig. S1) and A1AT was
cloned into pET51b(+). The constructs were then trans-
formed into E. coli BL21. GST or GST–preprohepcidin
fusion protein was produced in E. coli BL21 after induction
with 0.5 m m isopropyl thio- b -d-galactoside for 2 h at 30 °C.
Cells were harvested by centrifugation and resuspended into
STE (10 mm Tris ⁄ HCl, pH 8, 150 mm NaCl, 1 mm EDTA).
The bacteria were lysed by mild sonication at 4 °CinSTE
with a final concentration of 1.5% sarcosyl. The supernatant
In vivo interactions of preprohepcidin and prohepcidin E. Pandur et al.
2018 FEBS Journal 276 (2009) 2012–2021 ª 2009 The Authors Journal compilation ª 2009 FEBS
was gently mixed with STE-washed Glutathione–Sepha-
rose 4B beads (Amersham Biosciences, Uppsala, Sweden) at
4 °C for 1 h. GST proteins bound to beads were collected by
centrifugation at 3000 g, followed by three successive washes
with STE. In vitro protein–protein interaction assay (GST
pull-down) was carried out by incubating 30 lL of GST–
preprohepcidin and GST beads with an equal volume of
A1AT expressing BL21 lysate for 1 h in 5 mL of binding buf-
fer (50 mm Tris ⁄ HCl, pH 8, 50 mm NaCl, 10% glycerol,
0.1% Triton X-100). After centrifugation, the beads were
washed three times with binding buffer, resuspended in
30 lLof4· Laemmli buffer and centrifuged. The super-
natant was loaded onto 8% SDS-PAGE and transferred by
electroblotting to nitrocellulose membranes (Hybond C;
Amersham Pharmacia Biotech, Uppsala, Sweden). The pro-

tics), with 200 nm final concentrations of each primer.
Dissociation curves were generated after each quantitative
PCR run to ensure that a single specific product was ampli-
fied. Both target and reference genes were amplified with
efficiencies near 100% and within 5% of each other. For
the relative gene expression analysis, the 2
DDCt
(Livak)
method was used. The expression level of the gene of
interest was compared with the level of b-actin in each
sample. These relative expression rates were then compared
between the treated and untreated samples.
In vivo cross-linking
Huh7 (HPACC) cells (10
7
) were cultured in MEM supple-
mented with 10% fetal bovine serum and transiently trans-
fected with 30 lg preprohepcidin ⁄ pTriex-Neo3 (insert with
C-terminal His-tag) for 24 h with Transfectin reagent
(BioRad). In fresh medium, a specific cross-linker (DSS;
Sigma-Aldrich Corporation, St Louis, MO, USA) was
added to the cells in a 0.2 mm final concentration for
30 min at room temperature. The reaction was stopped with
50 mm Tris ⁄ HCl (pH 7.4). Cells were collected and washed
twice with NaCl ⁄ P
i
and then lysed on ice for 1 h in 150 lL
of lysis buffer (50 mm Hepes, pH 7.4, 150 mm NaCl, 1 mm
MgCl
2

of incubation buffer and then eluted with 20 lL4· Laemmli.
The total eluted volume was dotted onto nitrocellulose
membrane (Hybond C) and probed with anti-hepcidin IgG
(Alpha Diagnostic, San Antonio, TX, USA). Synthetic
E. Pandur et al. In vivo interactions of preprohepcidin and prohepcidin
FEBS Journal 276 (2009) 2012–2021 ª 2009 The Authors Journal compilation ª 2009 FEBS 2019
hepcidin peptide (Sigma-Aldrich Corporation) was used as
positive control for the western blot. Serum ultrafiltrate with
unbound prohormone served as a negative control for the
experiment.
Mass spectrometry after ZipTip C18 affinity
purification
ZipTip (Millipore Corp.) was washed ten times with 20 lL
of 50% acetonitrile, 0.1% trifluoroacetic acid and then
incubated with 20 lL of anti-A1AT IgG. After incubation,
the tip was washed three times with 10 lL of NaCl ⁄ P
i
(pH 7.4) and blocked with 20 lLof20mgÆmL
)1
bovine
serum albumin, and washed again with NaCl ⁄ P
i
. Human
serum (15 lL) concentrated with Microcon YM-30 was
pipetted up and down for 5 min. The proteins which did
not bind to the column were eliminated with two NaCl ⁄ P
i
and Milli-Q water (produced by Milli-Q Element Ultrapure
Water System; Millipore Corp.) washing steps. The elution
was carried out with 3 lL of 1% trifluoroacetic acid. The

metabolism. Cell 117, 285–297.
3 Papanikolaou G, Tzilianos M, Christakis JI, Bogdanos
D, Tsimirika K, MacFarlane J, Goldberg YP, Sakellaro-
poulos N, Ganz T & Nemeth E (2005) Hepcidin in iron
overload disorders. Blood 105, 4103–4105.
4 Beutler E (2007) Iron storage disease: facts, fiction and
progress. Blood Cells Mol Dis 39, 140–147.
5 Ganz T (2003) Hepcidin, a key regulator of iron metab-
olism and mediator of anemia of inflammation. Blood
102, 783–788.
6 Robson KJ, Merryweather-Clarke AT, Cadet E,
Viprakasit V, Zaahl MG, Pointon JJ, Weatherall DJ &
Rochette J (2004) Recent advances in understanding
haemochromatosis: a transition state. J Med Genet 41,
721–730.
7 Dunn LL, Rahmanto YS & Richardson DR (2007) Iron
uptake and metabolism in the new millennium. Trends
Cell Biol 17, 93–100.
8 Swinkels DW, Janssen MC, Bergmans J & Marx JJ
(2006) Hereditary hemochromatosis: genetic complexity
and new diagnostic approaches. Clin Chem 52, 950–968.
9 Courselaud B, Pigeon C, Inoue Y, Inoue J, Gonzalez
FJ, Leroyer P, Gilot D, Boudjema K, Guguen-Guillo-
uzo C, Brissot P et al. (2002) C ⁄ EBPalpha regulates
hepatic transcription of hepcidin, an antimicrobial
peptide and regulator of iron metabolism. Cross-talk
between C ⁄ EBP pathway and iron metabolism. J Biol
Chem 277, 41163–41170.
10 Verga Falzacappa MV & Muckenthaler MU (2005)
Hepcidin: iron-hormone and anti-microbial peptide.

regulates cellular iron efflux by binding to ferroportin
and inducing its internalization. Science 306, 2090–2093.
19 De DI, Ward DM, Langelier C, Vaughn MB, Nemeth
E, Sundquist WI, Ganz T, Musci G & Kaplan J (2007)
The molecular mechanism of hepcidin-mediated ferro-
portin down-regulation. Mol Biol Cell 18, 2569–2578.
20 Mena NP, Esparza A, Tapia V, Valdes P & Nunez MT
(2008) Hepcidin inhibits apical iron uptake in intestinal
cells. Am J Physiol Gastrointest Liver Physiol 294,
G192–G198.
21 Nemeth E, Valore EV, Territo M, Schiller G, Lichten-
stein A & Ganz T (2003) Hepcidin, a putative mediator
of anemia of inflammation, is a type II acute-phase
protein. Blood 101, 2461–2463.
22 Wrighting DM & Andrews NC (2006) Interleukin-6
induces hepcidin expression through STAT3. Blood 108,
3204–3209.
23 Verga Falzacappa MV, Vujic SM, Kessler R, Stolte J,
Hentze MW & Muckenthaler MU (2007) STAT3 medi-
ates hepatic hepcidin expression and its inflammatory
stimulation. Blood 109, 353–358.
24 Fleming RE (2007) Hepcidin activation during
inflammation: make it STAT. Gastroenterology 132,
447–449.
25 Nicolas G, Chauvet C, Viatte L, Danan JL, Bigard X,
Devaux I, Beaumont C, Kahn A & Vaulont S (2002)
The gene encoding the iron regulatory peptide hepcidin
is regulated by anemia, hypoxia, and inflammation.
J Clin Invest 110, 1037–1044.
26 Leong WI & Lonnerdal B (2004) Hepcidin, the recently

(2008) Serum pro-hepcidin levels in rheumatoid arthritis
and systemic lupus erythematosus. Inflammation 31,
146–153.
36 Jaroszewicz J, Rogalska M & Flisiak R (2008) Serum
prohepcidin reflects the degree of liver function impair-
ment in liver cirrhosis. Biomarkers 13, 478–485.
37 Constante M, Jiang W, Wang D, Raymond VA, Bilo-
deau M & Santos MM (2006) Distinct requirements for
Hfe in basal and induced hepcidin levels in iron over-
load and inflammation. Am J Physiol Gastrointest Liver
Physiol 291, G229–G237.
38 Oates PS (2007) The role of hepcidin and ferroportin in
iron absorption. Histol Histopathol 22, 791–804.
39 Kemna EH, Kartikasari AE, van Tits LJ, Pickkers P,
Tjalsma H & Swinkels DW (2008) Regulation of
hepcidin: insights from biochemical analyses on
human serum samples. Blood Cells Mol Dis 40,
339–346.
40 Hochepied T, Berger FG, Baumann H & Libert C
(2003) Alpha(1)-acid glycoprotein: an acute phase
protein with inflammatory and immunomodulating
properties. Cytokine Growth Factor Rev 14, 25–34.
41 Weiss G (2008) Iron metabolism in the anemia of
chronic disease. Biochim Biophys Acta, doi:10.1016/
j.bbagen.2008.08.006 (in press).
42 Kartikasari AE, Roelofs R, Schaeps RM, Kemna EH,
Peters WH, Swinkels DW & Tjalsma H (2008) Secretion
of bioactive hepcidin-25 by liver cells correlates with its
gene transcription and points towards synergism
between iron and inflammation signaling pathways.