Báo cáo khoa học: Identification of intracellular target proteins of the calcium-signaling protein S100A12 pdf

Identification of intracellular target proteins of the calcium-signaling
protein S100A12
Takashi Hatakeyama
1
, Miki Okada
1
, Seiko Shimamoto
1
, Yasuo Kubota
2
and Ryoji Kobayashi
1
Departments of
1
Signal Transduction Sciences and
2
Dermatology, Kagawa University Faculty of Medicine, Japan
In this report, we have focused our attention on i dentifying
intracellular mammalian proteins that bind S100A12 in a
Ca
2+
-dependent manner. Using S100A12 affinity chroma-
tography, we have identified cytosolic NADP
+
-dependent
isocitrate dehydrogenase (IDH), fructose-1,6-bisphosphate
aldolase A (aldolase), glyceraldehyde-3-phosphate dehy-
drogenese (GAPDH), annexin V, S100A9, and S100A12
itself as S100A12-binding proteins . Immunoprecipitation
experiments indicated th e formation o f stable c omplexes
between S100A12 and IDH, aldolase, GAPDH, annexin V

and S10 0A9 in vivo. Surface plasmon resonance analysis
showed that the b inding to S100A12, of S100A12, S100A9
and annexin V, was strictly Ca
2+
-dependent, whereas that of
GAPDH and IDH was only weakly Ca
2+
-dependent. To
localize the site of S100A12 interaction, we examined the
binding of a series of C-terminal t runcation mutants to the
S100A12-immobilized sensor chip. The results indicated t hat
the S100A12-binding site on S100A12 itself is located at
the C-terminus (residues 87–92). However, cross-linking
experiments with the truncation mutants in dicated that
residues 87–92 were not essential for S100A12 dimerization.
Thus, the interaction between S100A12 and S100A9 or
immobilized S100A12 should not be viewed as a typical S100
homo- or heterodimerization model. Ca
2+
-dependent
affinity chromatography revealed that C-terminal residues
75–92 are not necessary for the interaction of S100A12 with
IDH, aldolase, GAPDH and annexin V. To analyze the
functional properties of S100A12, we studied its action in
protein folding reactions in vitro. T he thermal agg regation of
IDH or GAPDH was facilitated by S100A12 in the a bsence
of Ca
2+
, whereas in the presence of Ca
2+

the protein sup-
pressed the aggregation of aldolase t o less t han 50%. T hese
results suggest that S100A12 may have a chaperone/anti-
chaperone-like f unction which is Ca
2+
-dependent.
Keywords: annexins; affinity chromatography; Ca
2+
-bind-
ing p roteins; molecular chaperone; surface plasmon reson-
ance ana lysis.
Ca
2+
, a second messenger, is involved in regulating various
cellular responses through a class of Ca
2+
-binding proteins,
including calmodulin (CaM), troponin C and S100 pro-
teins. These proteins exhibit a general structural principle o f
so-called EF-hand. They undergo a conformational c hange
through C a
2+
binding and consequently interact with their
target proteins.
S100 proteins were first considered to exist mainly in the
nervous system, but subsequently new members of the S100
family were identified in other tissues [1,2]. Based on amino
acid sequence similarity and other molecular properties, 20
different proteins have been assigne d to the family [2]. The y
have been found to interact in a Ca

2+
-dependent manner
with a diverse group of proteins, i ncluding those involved i n
cell proliferation and differentiation, cellular architecture,
signal transduction, and intracellular metabolism. For
example, S100A1 interacts with guanylate cyclase [3] and
glycogen phosphorylase [4]; S100A4 interacts with nonmus-
cle m yosin and tropomyosin [5,6]; S100B, S100A1, S100A8
and S100A9 interact with intermediate filaments [7–10];
S100A1 and S100B interact with microtubules [11–13],
twitchin kinase [14] and fruct ose-1,6-bisphosphate aldolase
[15]; S100B and S100A1 interact with N dr-kinase [ 16] and
phosphoglucomutase [17]; a nd S100A1, S100A6, S100A10,
S100A11 and S100B inteact with annexins [18–22].
S100A12 (also known as calgranulin C) has been detec-
ted in large amounts in neutrophils [23] and lung [24].
Recently, S100A12 from bovine lung was identified as a
ligand for the receptor for advanced glycation end-products
(RAGE), and anti-S100A12 immunogloblin was found
to substantially block leucocyte recruitment in delayed
hypersensitivity and colit is [25]. M ore recently, Yang et al.
reported the chemotactic properties of S100A12 for neu-
trophils and monocytes [26]. These reports indicate that
extracellular S 100A12 may contribute to the pathogenesis of
inflammatory responses. The X-ray structure of S100A12
hasbeensolvedintwocrystalforms:R
3
and P2
1
[27]. In the

R
3
crystal form, S100A12 is a dimer, and in the P2
1
crystal
form the dimers a re arranged as a h examer. Ca
2+
-depend-
ent hexamer formation could facilitate the binding of
S100A12 t o i ts rece ptor (RAGE) and r esult i n r eceptor
oligomerization [28]. The protein has been found to show
Correspondence to R. Kob ayashi, Department of S ignal Transduction
Sciences, Kagawa University Faculty of Medicine, 1750–1 Ikenobe,
Miki-cho, Kita-gun, Kagawa 761–0793, Japan.
Fax/Tel.: +81 87 8912249; E-mail:
Abbreviations: Aldolase, fructose-1,6-bisphosphate aldolase A; BS
3
,
bis-(sulfosuccinimidyl) suberate; CaM, calmodulin; GAPDH, glycer-
aldehyde-3-phosphate dehydrogenase; Hsp, heat shock protein; IDH,
NADP
+
-dependent isocitrate dehydrogenase; NF-jB, nuclear factor-
jB; NHS, N-hydroxysuccinamide; RAGE, receptor for advanced
glycation end-products; SPR, surface plasmon resonance.
(Received 1 8 March 2004, revised 21 July 2004,
accepted 2 A ugust 2004)
Eur. J. Biochem. 271, 3765–3775 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04318.x
the highest homologies to S100A8 (40%) and to S100A9
(46%) [29]. Upon Ca

2+
binding, S100A12 undergoes a
conformational c hange, supporting the idea t hat this p rotein
is involved in intracellular C a
2+
signal transduction [30].
In the present study, we have focused our attention on
identifying i ntracellular m ammalian proteins that bind
S100A12 in a Ca
2+
-dependent manner. Utilizing
S100A12-coupled Sepharose in affinity chromatography,
we have identified cytosolic NADP
+
-dependent isocitrate
dehydrogenase (IDH), fructose-1,6-bisphosphate aldolase
A (aldolase), glyceraldehyde-3-phosphate dehydrogenase
(GAPDH), annexin V, S100A9 and S100A12 itself as
S100A12-binding proteins.
Experimental procedures
Materials
Recombinant rat S100A1 and S100B were prepared as
described previously [31]. Recombinant human annexin V
was a gift from A. Iwasaki
1
(Kowa C o., L td, N agoya, Japan)
and recombinant human S100A9 a gift from K. Fukuda
(Asahi Kasei Corp., Osaka, Japan). Aldolase was obtained
from Sigma-Aldrich, IDH and GAPDH from Roche
Diagnostics, N-hydroxysuccinimide (NHS)

2
-activated Seph-
arose 4 Fast Flow and Protein-G Sepharose from Amer-
sham Biosciences, and bis-(sulfosuccinimidyl) suberate
(BS
3
) from Pierce Biotechnology, Inc. Anti-aldolas e i mmu-
nogloblulin was purchased from Roche Diagnostics, anti-
IDH a nd anti-GAPDH immunoglobulin from Nordic
Immunological Laboratories, and anti-annexin V and
anti-S100A9 immunoglobulin from Santa C ruz Biotechno-
logy, Inc. All other r eagents were at l east of analytical grade.
Construction and purification of S100A12 and its
truncation mutants
The bovine lung S100A12 coding sequence was obtained b y
PCR using the corresponding full-length cDNA, w hich was
cloned from a bovine lung Uni-Zap XR library [18], and
mutations were also induced by PCR. The oligonucleotides
employed as PCR primers installed an Nde Irestriction
enzyme site 5 ¢ to the ATG start c odon (sense primer), and
different stop codons and an adjacent BamHI restriction
enzyme site at the 3¢ end of the cDNA (antisense primers).
The P CR products were gel-purified and cloned into
pT7Blue T-Vector (Novagen). The nucleotide sequence of
each mutant was confirmed by automated DNA sequencing
(model 377; Applied Biosystems). An NdeI–BamHI frag-
ment of wild-type and all mutant cDNAs was ligated into
the same restriction en zyme site of pET-11a (Nov agen) and
then the plasmid was introduced into competent Escheri-
chia coli BL 21(DE3) (Stratagene).

Bacterial expression and purification was carried out as
described previously [32]. All proteins were judged to be
greater than 9 5% pure, based on 1 2% Tricine/SDS/PAGE
[33].
Affinity chromatography
The affinity matrix was prepared as follows: 2 mg of
recombinant wild-type S 100A12 and its t runcation mutants
was suspended in 2 mL of NaCl/P
i
(PBS) ( pH 7.0) and
mixed with a slurry (2 mL) of NHS-activated Sepharose.
Coupling w as allowed to proceed for 5 h at r oom tempera-
ture on a rocking table. The gel was then washed once with
NaCl/P
i
and incubated for 4 h at room temperature with
50 mL of 1
M
Tris/HCl ( pH 7.5). Af finity chromatography
was performed as described previously [24]. Briefly, bovine
lung (5 g) was homogenized with 6 volumes of 20 m
M
Tris/
HCl, 0.1 m
M
EGTA (pH 7.5) a nd centrifuged at 1 8 000 g
for 2 0 min at 4 °C. The s upernatant w as adjust ed to a Ca
2+
concentration of 1 m
M

with 1
M
CaCl
2
, followed by paper
filtration. The extract was applied to the affinity column
(2 mL of bed v olume), equilibrated in, and w ashed with, 20
volumes of the same buffer as used for homogenization and
adjusted to the same Ca
2+
concentration a s for the
supernatant. After washing the column, bound proteins
were eluted first with Tris buf fer containing 5 m
M
EGTA
alone, and then with the addition of 1
M
NaCl. The eluted
proteins were analyzed by SDS/PAGE.
Protein sequencing
Protein samples were subjected to SDS/PAGE and trans-
ferred to poly(vinylidene difluoride). After staining with
Ponceau S [34], the protein bands were cut out and digested
with lysylendopeptidase [35]. After overnight incubation at
37 °C, the proteolytic fragments were separated by HPLC
(model L C-10A; Shimadzu) in a C18 reverse-phase column
(TSK-gel ODS-80; Tosoh) with a linear gradient o f 0–80%
(v/v) a cetonitrile in the presence of 0 .1% (v/v) trifluoroacetic
acid at a flow rate of 1 mLÆmin
)1

. The amino acid sequence
of each proteolytic fragment was determined with an
automated protein sequencer (model PPSQ-21; Shimadzu).
Antibody production
Antibody to bovine S100A12 was produced by intramus-
cular injection of 1 mg of recombinant S100A12 protein
(emulsified in Freund’s complete adjuvant) into rabbits.
Booster shots of t he antigen in Freund’s i ncomplete
adjuvant were given three times at 2-week intervals. The
rabbits were bled 10 days after the last injection and the
specificity of antibody was check ed by Western blotting.
Immunoprecipitation and Western blotting
Bovine lung (1 g) was homogenized in five volumes of
20 m
M
Tris/HCl, p H 7.6, 0.1
M
KCl, 0.2 m
M
phenyl-
methanesulfonyl fluoride with 1 m
M
CaCl
2
and t hen
centrifuged at 15 000 g for 30 min. The extract (100 lL)
was incubated for 1 h at 25 °Cwith1lg of anti-S100A12
immunoglobulin. Protein G–Sepharose (20 lL) was added
to the sample, which was then rotated for 3 h. Sepharose
beads were washed four times and processed for SDS/

PAGE. Immunoprecipitates were analyzed by Western
blotting with antibodies specific for IDH, aldolase,
GAPDH, annexin V, S100A12 or S100A9.
Surface plasmon resonance (SPR) analysis
SPR analysis of real-time protein–protein interactions was
carried out using BIAcore X (Biacore, Inc.). All steps were
3766 T. Hatakeyama et al. (Eur. J. Biochem. 271) Ó FEBS 2004
performed at 25 °C. Recombinant S100A12 was covalently
linked to carboxymethylated dextran on the surface of
the s ensor chip , CM5, by amine coupling, according to
the manufacturer’s instructions. R ecombinant S100A12
(50 lgÆmL
)1
,60lL) was immobilized in 20 m
M
sodium
acetate, 0.15
M
NaCl, pH 6.0, until 13 000 response units
(1.45 p mol) were bound and a st able baseline was obtained.
As a control for nonspecific binding, a surface without
S100A12 (untreated surface) was tested in all experiments.
Samples of S100A12 and its mutants (27 n
M
,54n
M
,
108 n
M
, 216 n

M
, 436 n
M
and 873 n
M
), S100A9 (0.23 l
M
,
0.46 l
M
,0.93l
M
,1.85l
M
and 3.70 l
M
), annexin V
(0.35 l
M
,0.70 l
M
,1.39 l
M
,2.78 l
M
and 5.56 l
M
), GAPDH
(1.50 l
M

,3.0l
M
,4.5l
M
,6.00l
M
,9.00l
M
and 12.00 l
M
)
and IDH (1 .31 l
M
,2.63l
M
,5.26l
M
,7.89l
M
and
10.52 l
M
), were prepared in a running buffer containing
10 m
M
HEPES, pH 7.0, 150 m
M
NaCl, and 0.005% (v/v)
Tween-20 in the p resence of 1 m
M

CaCl
2
or 1 m
M
EGTA at
a flow rate of 20 lLÆmin
)1
. The S100A12-coupled sensor
chip was regenerated between protein injections with a brief
(60s)washin5m
M
EGTA, 2
M
NaCl until the baseline
returned to its preinjection level. The equilibrium dissoci-
ation constant, K
D
, was deduced from the kinetics of the
binding from the on (k
a
) and off (k
d
)rates(K
D
¼ k
d
/k
a
).
The rate constants were determined using

BIAEVALUATION
3.0 software using numerical integration algorithms.
Dimer formation and cross-linking of S100A12
and its mutants
Purified wild-type S100A12 and its mutant proteins were
preincubated each at a concentration of 0.5 mgÆmL
)1
in
20 m
M
HEPES (pH 7.0) in the presence of 1 m
M
CaCl
2
or
2m
M
EGTA for 20 min at 25 °C. Subsequently, aliquots
were incubated with 5 m
M
BS
3
crosslinker for 30 min at
25 °C. Reactions were quenched by adding Tris to a final
concentration of 50 m
M
. Proteins were visualized by 12%
Tricine/SDS/PAGE.
Gel filtration
Gel filtration was carried out on a Superdex 75 column (HR

10/30; Amersham Biosciences)-connected FPLC chroma-
tography system (Amersham Biosciences). The hydro-
dynamic analysis was carried out with 50 lgofthe
respective protein. The column buffer w as 20 m
M
HEPES,
pH 7 .0, 150 m
M
NaCl and 1 m
M
CaCl
2
or 1 m
M
EGTA.
The column w as operated at a flow rate of 0.5 mLÆmin
)1
and
a p rotein was detected using th e Bradford a ssay ( Bio-Rad).
The column was calibrated using transferrin (81 kDa),
ovalbumin (43 k Da), myoglobin (17.6 kDa), ribonuclease
A (13.7 kDa) and aprot inin (6.5 kDa).
Chaperone-like activity assay
The thermal-induced aggregation of IDH, aldolase and
GAPDH was performed according to established protocols
[36]. Protein solutions with or without S100A12 were mixed
at room temperature and then heated at 65 °C for 5 min in
a thermal cycler. The light scattering of the solution a t
488 n m was measured in a spectrophotometer (Model UV -
1600; Shimadzu).

Results
Identification of target proteins for S100A12
in bovine lung
Affinity chromatography with recombinant S100A12 cou-
pled to Sepharose was used to determine whether cytosolic
proteins of bovine lung were able to bind S100A12 in a
Ca
2+
-dependent manner (Fig. 1). Lung extract with 1 m
M
CaCl
2
added was applied to the S100A12 affinity column.
When Ca
2+
-dependent binding proteins were released
following application of a buffer c ontaining 5 m
M
EGTA,
a distinct protein peak was eluted f rom the column
(Fig. 1 A). The proteins found in the peak fraction migrated
at 38, 36, 35, 34, 23 and 6.5 kDa on SDS/PAGE ( Fig. 1B).
To identify these S100A12-binding proteins, a mino acid
sequencing o f the proteins was performed. The proteins
were digested with lysylendopeptidase, and the proteolytic
products were separated by reverse-phase HPLC on a C18
column. Amino acid sequence analysis was performed on
several major peptides. Referring to the NBRF protein
sequence database and the SwissProt database, the
sequences of 38, 36, 35, 34, 23 and 6.5 kDa proteins

matched IDH, aldolase, GAPDH, annexin V, S 100A9 and
S100A12, respectively (Fig. 1C). Of the 12 determined
amino acid sequences of the 23 kDa protein, 11 residues
were found to be identical t o those of bovine S 100A9, except
for one residue at position 41. The sequences of th e other
proteins were completely identical to those of each protein
compared. To i dentify, more strongly, binding proteins of
S100A12, we also eluted the S100A12-Sepharose column
with a 1
M
NaCl-containing buffer, after elution with the
EGTA-containing buffer. As shown in lane 2 of Fig. 1B,
trace amounts of h igh molecular mass proteins (ranging
from 45 to 100 kDa) were detected.
S100A12 interacts with IDH, aldolase, GAPDH, annexin V
and S100A9
in vivo
We performed co-immunoprecipitation experim ents to
demonstrate that in vitro S100A12 and target protein
associations also occur under physiological conditions in
mammalian cells. A physical interaction between S100A12
and IDH, aldolase, GAPDH, annexin V or S100A9 was
demonstrated by the observation that I DH, aldolase,
GAPDH, annexin V or S100A9 immunoprecipitate with
the a nti-S100A12 immunoglob ulin from bovine lung
extracts (Fig. 2). S100A12, IDH, aldolase, GAPDH, ann-
exin V and S100A9 were not detected after immunopreci-
pitation with rabbit preimmune serum or other nonspecific
antibodies (data not shown). These experiments indicate the
formation of stable complexes between S100A12 and IDH,

aldolase, GAPDH, annexin V or S100A9 under physiolo-
gical c onditions in mammalian cells, although t he possibility
that the complexes were formed after lysis cannot be
excluded.
Affinity determination of the interaction between
S100A12 and its target proteins
To study t he real-time binding kinetics of S100A 12,
S100A9, annexin V, GAPDH and IDH to S100 A12,
Ó FEBS 2004 S100A12 target proteins (Eur. J. Biochem. 271) 3767
recombinant S100A12 was immobilized on a biosensor
chip surface and the protein complex formation was
analyzed by SPR. The sensorgrams of the interactions are
shown i n F ig. 3 . The binding of S100A12, S100A9 and
annexin V to immobilized S100A12 was strictly Ca
2+
dependent (Fig. 3A–C). These observations are consistent
with the results of affinity chromatograp hy experiments
as described above. In contrast, the binding of GAPDH
and IDH showed weak Ca
2+
dependency (Fig. 3D,E).
The discrepancy between the Ca
2+
-sensitive elution of
GAPDH and IDH from the S100A12-Sepharose and the
relative Ca
2+
insensitivity in the SPR a nalyses, remains t o
be explained. The immobilization of S 100A12 to a
CM-sensor chip via its primary amines could have

modified the Ca
2+
-dependent conformational change of
the protein. The binding curves of S100A12, S100A9,
annexin V, GAPDH, and IDH were fitted to the bivalent
binding model. When the b inding curves were fitted t o the
bivalent model, the Ôgoodness of fitÕ was indicated by a v
2
value of < 1.0. All other models had a v
2
of > 1.0,
indicating higher nonrandom deviation from the fitted
curve. The binding affinities determined for these target
proteins to S100A12 are summarized in Table 1. S100A12
bound to immobilized S100A12 approximately 100-fold
more strongly than it bound to S100A9 or annexin V, and
about 10-fold more strongly than it bound to GAPDH or
IDH. There is a discrepancy between the relatively low-
affinity binding of S100A9 and annexin V to S100A12 and
the complex formation of these proteins observed by the
affinity chromatography (Fig. 1) and the coimmunopre-
cipitation experiments (Fig. 2). T his discrepancy appears t o
be explained by the high levels of S100A9 and annexin V
expressed in bovine lung [24,37]. As aldolase has a s trong
net positive charge (pI ¼ 8.9), it bound nonspecifically to
the CM5 sensor chip, presumably owing to electrostatic
interaction to the surface (data not shown).
Proteins Sequences
a : NADP -dependent isocitrate dehydrogenase
(IDH)

b : fructose-1,6-bisphosphate aldolase A
(aldolase)
c : glyceraldehyde 3-phosphate dehydrogenase
(GAPDH)
d : annexin V
e : S100A9
f : S100A12
ISGGSVVEM
5 ISGGSVVEM 13
IFPYVELDLHSYD
31 IFPYVELDLHSYD 43

ADDGRPFPQVIK
87 ADDGRPFPQVIK 98
RVIISAPSADAPMFV
116 RVIISAPSADAPMFV 130

FNGTVK
53 FNGTVK 58

QVYEEEYGSSLEDD
127 QVYEEEYGSSLEDD 140

QLVQK
35 QLVQK 39

EQPNFLK
40 ELPNFLK 46

IFQDLDADK

56 IFQDLDADK 64

TAHIDIHK
84 TAHIDIHK 91
0.6
0.3
0
Fraction number
5 mM EGTA 5 mM EGTA/1 M NaCl
MW
97.4
66.2
45.0
31.0
21.5
6.5
a
b
c
d
e
f
0 10 20 30 40
EGTA
EGTA/NaCl
14.4
A
C
B
Protein concentration (A

595
)
(kDa)
+
Fig. 1. Identification o f S100A12-binding proteins fro m bovine lung. (A) Elution p rofile obtained f rom bovine lung extract a fter S100A12 affinity
chromatography. (B) Tricine/SDS/PAGE (12%) of E GTA and EGTA/NaCl eluates. The 38, 36, 35, 3 4, 23 and 6.5 kDa proteins were tentatively
nameda,b,c,d,eandf,respectively.(C)Partialaminoacidsequencesof a–f. The amin o acid se qu ence(s) o f protein a w as (were) compared with
those of NADP
+
-dependent isoc itrate dehydro genase (IDH), p rotein b with fructose-1,6-bisphosphate aldolase A (aldolase), p rotein c with
glyceraldehyde-3-phosphate dehydrogenase (GAPDH), protein d with annexin V, protein e with S100A9, and protein f with S100A12. The
numbers indicate residue numbers for each protein.
3768 T. Hatakeyama et al. (Eur. J. Biochem. 271) Ó FEBS 2004
Mapping of the S100A12 binding site on S100A12 itself
To localize the site of S100A12 interaction on immobilized
S100A12, we examined the binding of a series o f C-terminal
truncation mutants to wild-type S100A12 coupled to the
sensor chip in the presence o f C a
2+
. A schematic r epre sen -
tation of the amino acid sequences of the four mutants is
given in F ig. 4A. Figure 4B and Table 1 show the kinetics
of the interactions between the t runcation mutants and
wild-type S100A12. C89 bound readily to the immobilized
wild-type S100A12 in the presence of Ca
2+
, while further
truncation resulted in a significant reduction of the binding.
C74 did not bind at all t o the immobilized S100A12. Owing
to a faster dissociation rate, the C89 fragment bound to the

immobilized S100A12 about 10-fold more weakly than full-
length S100A12. C86 and C79 had much lower affinities for
the immobilized S100A12 t han the full-length S100A12.
C74 did not bind to the immobilized S100A12. The results
of these experiments suggest that the S100A12-binding site
on S100A12 itself is located at the extreme C terminus
(residues 87–92) of the protein.
Dimerization of S100A12 mutants
To determine whether the C-terminal e xtension (residue
87–92) truncation mutant of S100A12 affects the ability to
dimerize, chemical cross-linking experiments were employed
using BS
3
. The results clearly indicated that wild-type
S100A12 and all of the S100A12 mutants, except C74, form
homodimers (Fig. 5). The formation of homodimers was
found to occur both in the presence and absence of Ca
2+
,
suggesting that Ca
2+
is not essential for S100A12 dimeri-
zation. Furthermore, these data indicate that the C-terminal
residues 87–92 are not critical for dimerization.
In cross-linking experiments, C74 did not migrate into t he
gel because of possible aggregation. The total absence of
C-terminal extension, including helix IV, m ight cause
instability of the molecule.
Gel filtration
To circumvent the use of a chemical cross-linking reagent,

which could stabilize rather weak and nonspecific protein
interactions, we next employed analytical gel filtration to
analyze the hydrodynamic p arameters of S 100A12 and its
truncation mutants in solution. Results f or wild-type
S100A12, and the C89, C86 and C74 mutants, are
summarized in Fig. 6. When compared with elution posi-
tions of marker proteins, the elution position of wild-type
S100A12 corresponds to  30.3 kDa. This differs slightly
from the 21.3 kDa position expected for a dimeric S100A12
of perfectly globular shape, and indicates that dimeric
S100A12 s hows s ome deviations from a globular shape. The
dimeric nature of wild-type S100A12 in the 30.3 kDa peak
was verified by gel filtration of covale ntly cross-linked wild-
type S100A12 dimers, which showed an elution position
indistinguishable from that of the noncross-linked protein
(data not shown). When compared with the wild-type
S100A12, C89 and C86 mutants show almost identical
elution profiles, i.e. they elute in a s ymmetrical peak at the
dimer position. When analyzed by analytical gel filtration,
the elution position of the C74 mutant corresponds to
 129 kDa, in dicating possible aggregation.
Interaction of C-terminal truncation mutants
of S100A12 and the S100A12 target proteins
To further determine the role of S100A12 C-terminal
residues i n S100A12 target protein interactions, we prepared
affinity matrices coupledwith C-terminal truncation mutants
andusedthemfortheCa
2+
-dependent affinity chromato-
graphy o f bovine l ung extracts (Fig. 7). Immobilized C89, as

well as immobilized wild-type S100A12, were able to interact
with all t he target proteins such as IDH, aldolase, GAPDH,
annexin V, S100A9, and S100A12. However, immobilized
C86 and C74 mutants readily bound to IDH, aldolase,
GAPDH and annexin V, b ut not at all to S100A9 or
S100A12. These results demonstrate that C-terminal residues
75–92 are not necessary for the interaction of S100A12 with
IDH, aldolase, GAPDH a nd annexin V.
Effects of S100A12 on the heat-induced aggregation
of IDH, aldolase, and GAPDH
To analyze the functional properties of S100A12, we studied
its action on protein-folding reactions in vitro (Fig. 8). The
thermal unfolding and aggregation of IDH, aldolase, and
GAPDH was used as a typical assay system. Heat ing 3 l
M
IDH, aldolase o r G APDH at 65 °Ccausedprotein
aggregations in less than 5 min. In t he absence of
S100A12, 1 m
M
Ca
2+
had essentially no effect on the
unbound bound
IP
anti-S100A12
IDH
aldolase
GAPDH
annexin V
S100A9

S100A12
WB: anti-IDH
WB: anti-aldolase
WB: anti-GAPDH
WB: anti-annexin V
WB: anti-S100A9
WB: anti-S100A12
41.6 -
kDa
28.2 -
14.8 -
41.6 -
28.2 -
41.6 -
41.6 -
41.6 -
Fig. 2. S100A12 interaction with NADP
+
-dependent IDH, aldolase,
GAPDH, annexin V, S 100A9 and S100A12 in vivo. Bovine lung
extracts were immunoprecipitated with an anti-S100A12 immuno-
globulin in the presence o f 1 m
M
CaCl
2
. The immunoprecipitates were
analyzed by Western blotting with anti-IDH, anti-aldolase, anti-
GAPDH, anti-annexin V, anti-S100A9 and anti-S100A12 immuno -
globulins. IP, i mmunoprecipitation, WB, Western blotting. Molecular
mass standards a re shown at t he left.

Ó FEBS 2004 S100A12 target proteins (Eur. J. Biochem. 271) 3769
thermal aggregation of IDH (Fig. 8A), aldolase (Fig. 8B)
and GAPDH (Fig. 8C). However, Ca
2+
had a dramatic
effect on the chaperone/antichaperone behavior of
S100A12. The thermal aggregation of IDH or GAPDH
was greatly facilitated by S100A12 in the absence of Ca
2+
(Fig. 8 A,C), whereas S 100A12 suppressed the aggregation
of aldolase to less than 50% in the presence of Ca
2+
(Fig. 8B). S100A12 alone caused slight optical changes in
the light-scattering analysis (Fig. 8). These behaviors of
S100A12 are similar to those of some of the well-studied
molecular chaperones or antichaperones [36].
Discussion
Recently, S100A12 was demonstrated to trigger a signal
transduction cascade by binding to the cell-surface recepto r
RAGE in endothelium, lymphocytes, and mononuclear
phagocytes, resulting in activation of the transcription
factor nuclear factor-jB(NF-jB) [25]. Based on these
findings, a molecular model for the involvement of extra-
cellular S100 p roteins in i nflammatory processes w as
proposed [25]. I dentification of the in vivo S100A12 target
proteins, a nd characterization of their m ode of interaction
with S100A12, is essential to d etermine the exact contribu-
tion of the S 100A12-dependent signaling pathways in
cellular function. S100 proteins have been implicated in
intracellular and extracellular regulatory activities. Mem-

bers of this protein family have been shown to interact with
several target proteins w ithin cells, t hereby regu lating
enzyme activities, cell growth and differentiation, and
Ca
2+
homeostasis. In the present study, we have focused
our attention on identifying intracellular S100A12-binding
proteins and presented evidence showing that IDH, aldo-
lase, G APDH, annexin V, S100A9 a nd S100A12 a re
putative S100A12 target proteins. An overlap between t he
specificities of different S100 proteins and CaM for inter-
acting target proteins, peptides, and drugs is often observed
[14,16,38]. Hence, aldolase, GADPH and annexin V in
0 50 100
Time (sec)
Responses (RU)
450
350
250
150
50
0
0 50 100
0 50 100
0 50 100
0 50 100
250
200
150
100

50
0
Responses (RU)
Time (sec)
250
200
150
100
50
0
Responses (RU)
250
200
150
100
50
0
Responses (RU)
250
200
150
100
50
0
Responses (RU)
(S100A12) (S100A9)
(annexin V)
(GAPDH)
(IDH)
Time (sec)

Time (sec)
Time (sec)
Ca
2+
Ca
2+
Ca
2+
Ca
2+
Ca
2+
EGTA
EGTA
EGTA
EGTA
EGTA
AB
D
C
E
Fig. 3. The b inding of S100A12 (A), S100A9 (B), annexin V (C), GAPDH (D), and IDH (E) to immobilized S100A12. Overlayplotofthereal-time
binding data r ecord ed upon interaction of i mmobiliz ed S100A12 with its target proteins. S10 0A12 was cova lently attached to the surface of th e
CM5 senso r ch ip, as described i n t he Exp erime ntal pro cedures. T he trac e s re flect the amount of p rote in c omplex formed at the surface over time.
Each sensorgram consists of an association phase and a dissociation phase during the sample flow, reflecting the decay of the complex when the
target protein was removed from the flow cell. S100A12 (873 n
M
), S100A9 (3.70 l
M
), annexin V (5.56 l

M
), GAPDH (12.0 l
M
)andIDH
(10.52 l
M
) were injected at a flow rate of 20 lLÆmin
)1
in the presence of 1 m
M
Ca
2+
or 2 m
M
EGTA.
Table 1. The affinity kinetics of interaction between S100A12 and its
target proteins in the presence of 1 m
M
CaCl
2
. Therateconstants(k
d
and k
a
) were derived from analysis of the dissociation and association
phases, respectively, of the s ensorgrams. K
D
was calculated from the
ratio o f k
d

and k
a
. G APDH, g lyceraldehyde-3-phosphate dehydroge-
nase;IDH,NADP
+
-dependent isoc itrate dehydro genase. WT , wild-
type S100A12; C89, C86, C79 and C74, truncated mutants of
S100A12.
Analyte K
D
(
M
) k
a
(1/Ms) k
d
(1/s)
S100A12 4.09 · 10
)9
1.32 · 10
5
5.40 · 10
)4
S100A9 1.52 · 10
)7
2.03 · 10
4
3.08 · 10
)3
Annexin V 6.21 · 10

)7
6.30 · 10
3
3.91 · 10
)3
GAPDH 8.27 · 10
)8
9.43 · 10
3
7.80 · 10
)4
IDH – 8.51 · 10
3
–
S100A12 (WT) 2.18 · 10
)9
6.23 · 10
4
1.36 · 10
)4
C89 6.62 · 10
)8
7.27 · 10
4
4.81 · 10
)3
C86 1.86 · 10
)7
1.23 · 10
4

2.29 · 10
)3
C79 – – –
C74 – – –
3770 T. Hatakeyama et al. (Eur. J. Biochem. 271) Ó FEBS 2004
crude cell extracts that bind t o S100A12–Sepharose beads
also bind to other S100 proteins or CaM [15,18–22,39].
IDH, S100A9 and S100A12 are the only proteins identified,
to date, that demonstrate strict interactions with S100A12
and a bsolutely n o binding to CaM (data not shown). The
results of specific binding proteins of S100A12, examined in
the present study, are briefly d iscussed below.
S100A12 and S100A9
Complex formation is regarded to be an essential pre-
requisite for the biological function of S100 proteins [1]. A
characteristic property of S100 proteins is their tendency to
dimerize. The question raised by the interaction between
S100A12 and S100A12 itself o r S100A9 c oncerns homo- or
A
B
Fig. 4. Binding of the C-terminal trun cation mutants of S100A12 to immobilized S100A12. (A) Amino ac id sequences of the wild-type S100A 12 (WT;
92 amino acids) a nd the C -terminal trun cation mut ants (C89, C 86, C79 a nd C74). S haded boxe s indicate fi rst and s econd Ca
2+
-binding loops o f
S100A12. (B) Overlay plot of sensorgrams recorded on injection of WT, C89, C86, C79 and C74 (873 n
M
each) over surfaces containing
immobilized S100A12 in the presence of Ca
2+
.

CaCl
2

− + − − + − − + − − + −
EGTA − − + − − + − − + − − +
MW
97.4
66.2
45.0
31.0
21.5
(kDa
)
dimer
monomer
WT C89 C86 C74
6.5
14.4
a b c
a b c
a b c
a b c
Fig. 5. Dimerization of wild-type S100A12
(WT) and its C -terminal truncation mutants.
S100A12 dimerization, as revealed by chem-
ical cross-linking. WT, C89, C86 and C74
(5 mg) were cross-linked using 2 m
M
bis-
(sulfosuccinimidyl) suberate (BS

3
) cross-
linking agent. D etails of the assays are
described in the Experimental procedures.
Lane a, noncross-linked proteins; lane b,
proteins cross-linked in the p re sence of 1 m
M
CaCl
2
, and lane c, proteins cross-linked in the
presence of 2 m
M
EGTA. The symbols Ô+Õ
and Ô–Õ indicate the presence and absence,
respectively, o f the pertinent reactant in
the reaction mixture. B lank arrows indicate
possible aggregation.
Ó FEBS 2004 S100A12 target proteins (Eur. J. Biochem. 271) 3771
hetero-dimerization processes. Deletion of the C-terminus
(residues 87–92) o f S100A12 was found to be sufficient to
abrogate interaction with S100A12 (Fig. 4). To study the
influence of this moiety on the dimerization, a set of
C-terminal deletion experiments were also carried out. T he
deletion of the C-terminal six residues did not prevent
S100A12 homodimerization (Figs 5 and 6). It is therefore
possible to state that S100A12 interaction with S100A12
itself or S100A9 has a similarity with the i nteraction
between S100B (or S 100A1) and its target proteins [40,41].
Recently, Ost erloh et al. reported that the C-terminal
domain (residues 88–90) of S100A1 is essential for TRTK

peptide bind ing, but dispensable for dimerization [40]. More
recently, Deloulme et al. s howed that deletion of the
extreme C-terminal residues (84–91) of S100B had no
effects on its dimerization but significantly decreased the
interaction between S100B and S100A6 or S100A11 [41].
Further d eletions which affect he lix IV completely abolished
both S100A12 and S100A9 interactions, and S100A12
dimerization, con firming the importance of t his r egion fo r
S100A12 complex formation. The observed Ca
2+
depend-
ency of the interaction of S100A12 with S100A 12 itself or
S100A9, b ut not of the S100A12 dimerization, also supports
such a model.
Annexin V
Members of the annexin family have been shown to bind
proteins of the S100 family. Annexin I has been shown to
bind S100 A11 [ 22], annexin II binds S100A10 [ 42], S100A6
[43] and S100B [19], annexin XI binds S100A6 [20], and
annexin V and VI bind S100A1 and S100B [18]. The
annexin II–S100A10 interaction is Ca
2+
independent, while
other annexin–S100 protein interactions are shown to be
Ca
2+
-dependent [44]. The mode of interaction o f annexin V
with S100A12 differs from the mod e of a nnexin II–
S100A10 i nteraction. In the c ase o f t he latter, the C-terminal
domain of S100A10 was proven to be critical for the

recognition of annexin [45]. However, in the case of the
annexin XI–S100A6 interaction, helix I of S100A6 was
found to be e ssential for the recognition o f annexin XI [46].
Similarly, helix I o f S100A11 was shown to be critical in the
recognition of annexin I [22]. Thus, there are a number o f
modes of interaction between members of the annexin
family and those of t he S100 family. Although t he precise
physiological significance of these interactions is not known,
Garbuglia et al. reported that binding of annexin VI, but
not annexin V, to S100A1 and S100B blocks the ability of
these two S100 proteins to inhibit glial fibrillary acidic
protein and desmin assemblies [18].
Enzymes
Previous biochemical studies have identified many potential
S100 target enzymes, such as glycogen phosphorylase [4],
twitchin kinase [14], aldolase [15], Ndr kinase [16], phospho-
glucomutase [17] and G APDH [ 47], some e xhibiting
Ca
2+
-dependent modulation while others show Ca
2+
-
independent modulation. Both the affinity chromatography
and the BIAcore experiments showed a high-affinity inter-
action between IDH, aldolase and GAPDH, and immobi-
lized S100A12. These observations prompted us to identify
Fraction (ml)
Fraction (ml)
Fraction (ml)
Fraction (ml)

V
0
81kDa
43kDa
17.6kDa
13.7kDa
6.5kDa
WT
C89
C86
C74
0
0.05
0.1
0.15
0.2
51015
0
0.05
0.1
0.15
0.2
51015
0
0.05
0.1
0.15
0.2
51015
0

0.05
0.1
0.15
0.2
51015
V
0
81kDa
43kDa
17.6kDa
13.7kDa
6.5kDa
V
0
81kDa
43kDa
17.6kDa
13.7kDa
6.5kDa
V
0
81kDa
43kDa
17.6kDa
13.7kDa
6.5kDa
Absorbance at 595nmAbsorbance at 595nmAbsorbance at 595nmAbsorbance at 595nm
Fig. 6. Analytical gel filtration of wild-type S100A12 (WT) and its
C-terminal truncation mutants. Gel filtration of the purified proteins
(WT, C89, C86 and C74) was carried out on a Superdex 75 c olum n i n

the absen ce of Ca
2+
. The respective protein peaks are given as a
function of volume. E lution volumes of the marker proteins tranferrin
(81 kDa), ovalu min (43 k Da), myoglob in (17.6 kDa), ribonuclease A
(13.7 kDa) and ap rotinin (6.5 kDa) are indicated for comparison.
3772 T. Hatakeyama et al. (Eur. J. Biochem. 271) Ó FEBS 2004
residues on S 100A12 involved i n the int eraction of S 100A12
with the enzymes IDH, aldolase and GAPDH. While the
C-terminal domain of S100A12 was found to be essential for
the interaction of imm obilized S100A12 with S100A12 itself
or S100A9, this domain was not n ecessary for the interaction
of S100A12 with the three enzymes. Several S100 proteins
have also been shown to regulate other enzyme activ-
ities. For example, Zimmer et al. reported the activation
of aldolase by S100A1 and S100B [15], the inhibition of
glycogen phosphorylase by S100A1 [4], the activation of
WT C89 C86 C74 MW
S100A12
S100A9
IDH
aldolase
GAPDH
annexin V
(kDa
)
31.0
21.5
14.4
6.5

66.2
45.0
97.4
Fig. 7. Role of C-terminal residues of S100A12
in S100 A12 target p rotein inter actions. Affinity
chromatographies with wild-type S100A12
(WT), C89, C86 and C74 coupled to NHS
4
-
activated Sepharose were used to determin e
whether c ytosolic proteins o f bovine lung were
able to bind to S100A12 in a Ca
2+
-dependent
manner. Lung extract (3 mL) with 1 m
M
CaCl
2
added was applied to the affinity
columns (0.5 m L of the bed volume). After
washing, the b ound proteins were e luted with
SDS sample buffer. The individual fractions
(5 lL) were then subjected to S DS/PAGE.
0 10 50
0.4
0.3
0.2
0.1
0
0 10 50

0 10 50
0.5
0.4
0.3
0.2
0.1
0
0.6
0.4
0.2
0
Absorbance at 488 nm
Absorbance at 488 nm
S100A12 (µg/ml)
S100A12 (µg/ml)
Absorbance at 488 nm
S100A12 (µg/ml)
(IDH) (aldolase) (GAPDH)
AB C
Fig. 8. Effects of S100A12 on th e thermal aggregation of NADP
+
-dependent IDH, aldolase and GAPDH. Light s cattering owing to t hermal (65 °C)
aggregation was measured as a bsorbance at 488 nm (vertical axis) in a spectrophotometer. Thermal aggregation of IDH (A), aldolase (B), and
GAPDH (C), 3 l
M
each, in the p resence (0.2 m
M
CaCl
2
, close d t riangles) or a bsen ce (2 m

M
EGTA, op en t rian gles) of C a
2+
, with the addition o f
S100A12 a t various d ifferen t concentrations. O pen and c losed circles i ndicate the prese nce (1 m
M
CaCl
2
) a nd ab sence ( 2 m
M
EGTA), respectively,
of Ca
2+
without the addition of target p rotein (IDH, aldolase or GAPDH). Each value represents the m ean and SD of three e xperiments.
Ó FEBS 2004 S100A12 target proteins (Eur. J. Biochem. 271) 3773
phosphoglucomutase by S100B [17], and the inhibition of
phosphoglucomutase by S100A1 [17]. Filipek et al. showed
that S100A6 had no influence on the V
max
and K
m
of
GAPDH [47]. These reports suggest that S100 proteins
might have a role in the regulation of energy metabolism.
Further experiments are required to draw definitive conclu-
sions about the physiological relevance o f these data.
Molecular c haperones consist of several groups of
proteins that suppress the aggregation of unstable inter-
mediates of proteins and are implicated in protein f olding,
protein targeting to membranes, protein renaturation,

subcellular transport and degradation [48]. In contrast,
some proteins, such as disulfide isomerase [ 49] and t protein
[50], facilitate the misfolding and aggregation of their
substrate proteins. These functions, termed Ôanti-chaperone
activityÕ, may provide a common mechanism for aggregate
formation in the cell [49]. Recently, we have identified
S100A1, b ut not CaM, as a potent m olecular chaperone and
a new member of the heat shock protein (Hsp)70/Hsp90
multichaperone complex [51]. In the present report, we
investigated whether S100A12 could also have a c haperone-
like o r an antichaperone-like a ctivity. We used IDH,
aldolase, and G APDH as substrates to examine t he possible
chaperone/antichaperone function of S100A12 in vitro.In
the absence, but not in the presence, of Ca
2+
, S100A12
could facilitate the thermal aggregation o f IDH and
GAPDH. In contrast, in the presence of Ca
2+
, S100A12
served as a chaperone and inhibited the thermal aggregation
of aldolase. Although it is still not clear whether or not
S100A12 is an essential factor for t he folding and unfolding
of the S100A12 target p roteins in vivo , t hes e re su lts s ugge st
that S100A12 may have a chaperone/antichaperone-like
function via Ca
2+
-dependent formation of higher order
oligomers [28].
Acknowledgements

WethankDrAkioIwasaki(DepartmentofCellBiology,KowaCo.,
Ltd) for r ecombinant human annexin V and Dr K ouichirou Fukuda
(Asahi-kasei Co., Ltd) for r ecombinant human S100A9. We also thank
Kayoko Yamashita and Aiko Itoh for excellent technical assistance.
References
1. Do nato, R. (2001) S100: a multigenic family of calcium-modula-
ted p roteins of the EF -hand type with i ntracellular a nd extr acel-
lular functional r oles. Int. J. Bio chem. Cell Biol. 33, 6 37–668.
2. Heizmann, C.W., Fritz, G. & Scha
¨
fer, B.W. (2002) S100 proteins:
structure, functions and pathology. Front. B iosci. 7, 1356–1368.
3. Du da, T., Goraczniak, R .M. & Sharma, R.K. (1996) M olecular
characterization of S100A1–S100B p rotein in re tina a nd its ac ti-
vation mechanism of bovine photoreceptor guanylate cyclase.
Biochemistry 35 , 6263–6266.
4. Zimm er, D.B. & D ubuisso n, J.G. ( 1993) Identification of an S100
target p rotein : glycogen phosphorylase . Cell Calcium 14 , 323–332.
5. Kriajevska, M.V., Cardenas, M .N., Grigorian, M.S., Ambartsu-
mian, N.S., Geogiev, G.P. & Lukanidin, E.M. (1994) Non-muscle
myosin heavy chain as a possib le target for protein encoded by
metastasis-relat ed mts-1 gene. J. Biol. C hem. 269, 19679–19682.
6. Takenaga, K ., N akamura, Y ., Sakiyama, S ., H asegawa, Y., Sato,
K. & Endo, H. (1994) Binding of pEL98 protein, an S100-related
calcium-binding protein, to nonmuscle tropomyosin. J. Cell Biol.
124, 757–768.
7. Roth, J., Burwinkel, F., van den Bos, C., Goebeler, M., Vollmer,
E. & Sorg, C. (1993) MRP8 and MRP14, S100-like proteins
associated with myeloid differentiation, are translocated to plasma
membranes and interm ediate filaments in a calcium-dependent

manner. Blood 82, 187 5–1883.
8. Garbuglia, M., Verzini, I., G iamban co, A., Spreca , R . & Donato,
R. (1996) Effects of calcium-binding proteins (S100a (o), S100a,
S100b) on desmin assembly in vitro. FASEB J. 10, 317–324.
9. Bianchi,R.,Giambanco,I.&Donato,R.(1993)S100protein,but
not calmodulin, binds to the glial fibrillary acidic protein and
inhibits its polymerization in a Ca
2+
-dependent manner. J. Biol.
Chem. 268, 12669–12674
3
.
10. Bianchi, R., Verzini, M., Ga rbuglia, M., Giambanco, I. &
Donato, R. (1994) Mechanism of S 100 protein-depen dent inhibi-
tion of glial fibrillary acidic protein (GFAP ) polymerization.
Biochim. Biophys. Acta 1223, 354–360.
11. Sorci, G., Agneletti, A.L., Bianchi, R. & Donato, R. (1998)
Association of S100B with intermediate filaments and micro-
tubules in g lial cells. Biochim. Biophys. Acta 144 8, 277–289.
12. Do nato, R. (1985) Calcium-sensitivity of brain microtubule
proteins in the presence of S100 proteins. Cell Calcium 6,
343–361.
13. Do nato, R . ( 1988) C alcium-independent, pH-regulated e ffects of
S100 proteins on assembly-disassembly of brain microtubule
protein in vitro. J. Bio l. Chem. 263, 106–110.
14. Heierhorst, J., Kobe, B., Feil, S. C.,Parker,M.W.,Benians,G.M.,
Weiss, K.R. & K emp, B.E. (1996) Ca
2+
/S100 regulation of g iant
protein kinases. Nature 380, 636–639.

15. Zimmer, D.B. & van Eldik, L.J. (1986) Identification o f a mole-
cular target for the calcium-mod ulated protein S100: fructose-
1,6-bisphosphate aldolase. J. Biol. Chem. 261, 11424–11428.
16. Millw ard, T .A., Heizmann, C.W., Scha
¨
fer, B.W. & Hemmings,
B.A. (1998) Ca lcium regulation of Ndr protein kinase mediated by
S100 calcium-binding proteins. EMBO J. 17, 5913–5922.
17. Landar,A.,Caddell,G.,Chessher,J.&Zimmer,D.B.(1996)
Identification of an S100A1/S100B target protein: phosphoglu-
comutase. Cell Calcium 20 , 279–285.
18. Garbuglia, M., Verzini, M., Hofmann, A., Huber, R. & D onato,
R. (2000) S100A1 and S100B interactions with annexins. Biochim.
Biophys. Acta 1498, 192–206.
19. Hagiwara, M ., Ochiai, M., O wada, K., T anaka, T. & Hidaka, H.
(1988) Modulation of tyro sine phosphorylation of p36 and oth er
substrates by t he S100 protein. J. Biol. Chem. 263, 6438–6441.
20. To kumitsu, H., Mizutani, A., Minami, H., Kobayashi, R. &
Hidaka, H . (1992) A calcyclin-associated protein is a newly iden-
tified member of the Ca
2+
/phospholipid-binding proteins, ann-
exin family. J. Biol. Chem. 267, 8 919–8924.
21. Johnsson, N., Nguyen-Van, P., Soling, H.D. & We ber, K. (1986)
Functionally distinct serine phosphorylation sites of p36, the cel-
lular substrate of retroviral protein kinase; differential inhibition
of reassociation with p11. EMBO J. 5, 3455–3460.
22. Mailliard, W.S., Haigler, H.T. & Schlaepfer, D.D. (1996) Cal-
cium-dependent binding of S100C to the N-terminal domain of
annexin I. J. Bio l. Chem. 271, 719– 725.

23. Guignard, F., Mauel, J. & Markert, M. (1995) Identification and
characterization of a novel human neutrophil protein related to
the S100 family. Biochem. J. 309, 395–401.
24. Shishibori, T., O yama, Y., Matsushita, O., Yamashita, K.,
Furuichi,H.,Okabe,A.,Maeta,H.,Hata,Y.&Kobayashi,R.
(1999) Three distinct anti-allergic drugs, amlexanox, cromolyn and
tranilast, bind to S100A12 and S100A13 of the S100 protein
family. Biochem. J. 33 8 , 583–589.
25. Hofmann,M.A.,Drury,S.,Fu,C.,Qu,W.,Taguchi,A.,Lu,Y.,
Avila, C., Kambham, N., Bierhaus, A., Nawroth, P., Neurath,
M.F., Slattery, T., Beach, D., McClary, J., Nagashima, M.,
Morser, J., Stern, D . & Schmidt, A.M. (1999) RA GE mediates a
3774 T. Hatakeyama et al. (Eur. J. Biochem. 271) Ó FEBS 2004
novel proinflammatory axis: a central cell surface recep tor for
S100/calgranulin polypeptide s. Cell 97, 889–901.
26. Yang, Z ., Tao, T ., R aftery, M.J., Youssef, P., D i G irolamo, N. &
Geczy, C.L. (2001) Proinflammatory properties of the human
S100 protein S100A12. J. Leukoc. Biol. 69, 986–994.
27. Moroz, O.V., Antson, A.A., Murshudov, G.N., Maitland, N.J.,
Dodson, G.G., Wilson, K.S., Skib shoj, I., Lukanidin , E. &
Bronstein, I.B. (2001) The three-dimensional structure of
S100A12. Acta Crystallogr. D 57, 20–29.
28. Moroz, O.V., Antson, A.A., Dodson, E.J., Burrell, H.J., Grist,
S.J., Lloyd, R .M., M aitland, N.J., Dodson, G.G., Wilson, K.S.,
Lukanidin, E. & Bronstein, I .B. (2002) The structure of S100A12
in a hexameric form and its proposed role in receptor signalling.
Acta Crystallogr. D 58 , 407–413.
29. Ilg, E.C., Troxler, H., Bu
¨
rgisser, D.M., Kuster, T., M a rke rt, M.,

Guignard, F., Hunziker, P., Birchler,N.&Heizmann,C.W.
(1996) Amino acid sequence determination of human S100A12
(p6, calgranulin C, CGRP, CAAF1) by tandem mass spectro-
metry. Biochem. Biophys. Res. C ommun. 225, 1 46–150.
30. Dell’Angelica, E.C., Schleicher, C.H. & Santome
´
, J.A. (1994)
Primary structure and bind ing properties of c algranu lin C, a no vel
S100-like calcium-binding protein from pig granulocytes. J. Biol.
Chem. 269, 28929–28936.
31. Okada, M., Tokumitsu, H., Kubota, Y. & Kobayashi, R. (2002)
Interaction of S 100 proteins with the antiallergic drugs, olopata-
dine, amlexanox, and cro molyn: identificatio n of pu tative drug
binding sites on S100A1 protein. Biochem. Biophys. R es. Commun.
292, 1023–1030.
32. Yamashita, K., Oyama, Y., Shishibori, T., Matsushita, O., Okabe,
A. & K obayashi, R. (1999) Purification of b ovine S100A12 from
recombinant Escherichia coli. Protein Expr. Purif. 16, 47–52.
33. Sc ha
¨
gger, H. & von Jagow, G. (1987) Tricine-sodium dodecyl
sulfate-polyacrylamide gel electrophoresis for the separation of
proteins in the range from 1 to 100 kDa. Anal. Biochem. 166,
368–379.
34. Salinorich, O. & Montelaro, R.C. (1986) Reversible staining and
peptide mapping of proteins transferred to nitrocellulose after
separation b y sodium dodecyl sulfa te-polyacrylamid e gel electro-
phoresis. Anal. B iochem. 156, 341 –347.
35. Fern andez, J., Demott, M., Atherton, D . & Mische, S.M. (1992)
Internal protein sequence analysis: enzymatic digestion for less

than 10 micrograms of protein bou nd to polyvinyliden e difluoride
or nitrocellulose membranes. Anal. Biochem. 201, 255–264.
36. Kim, T.D., Paik, S.R., Yang, C.H. & Kim, J. (2000) Structural
changes in alpha-synuclein affect its chaperone-like activity
in vitro. Protein Sci. 9, 2489–2496.
37. Oyama, Y., Shishibori, T., Yamashita, K ., Naya, T ., Nakagiri, S.,
Maeta, H. & Kobayashi, R. (1997) Two distinct anti-allergic
drugs, amlexanox and cromolyn, bind to the same k inds of
calcium binding proteins, except calmodulin, in bovine lung
extract. Biochem. Biop hys. Res. C ommun. 240, 341–347.
38. Marshak, D.R., Lukas, T.J. & Watterson, D.M. (1985) Drug–
protein interactions: binding of chlorp romazine to calmodulin,
calmodulin fragments, and related calcium binding proteins.
Biochemistry 24, 144–150.
39. Christova, T.Y., Orosz, F. & Ovadi, J. (1996) Interaction be tween
D
-glycerald ehyd e-3-p hosp hate dehyd rog enase and calm odulin.
Biochem. Biophys. Res. C ommun. 228, 272–277.
40. Osterloh, D., Ivanenkov, V .V. & Gerke, V. (1998) H ydrophobic
residues in the C-terminal region of S100 A1 are essential for target
protein binding but not for dimerization. Cell Calcium 24,
137–151.
41. De loulme, J .C., Assard, N., Mbel e, G.O., Mangin, C., Kuwano,
R. & B audier, J. ( 2000) S100A6 a nd S100A11 are specific targets
of the calcium- and z inc-bindi ng S100B p rotein in vivo. J. Biol.
Chem. 275, 35302–35310.
42. Gerke, V. & Weber, K. (1985) Calcium-dependent conformational
changes in the 36-kDa subu nit of intestinal protein I related to the
cellular 36-kDa target of Rous sarcoma virus tyrosine kinase.
J. Biol. Chem. 260, 1688–1695.

43. Filipek, A., Gerke, V., Weber, K. & Kuznicki, J. (1991) Char-
acterization of the cell-cycle-regulated protein calcyclin from
Ehrlich ascites tumor c ells. Identificatio n of two binding prote ins
obtained by Ca
2+
-dependent affinity chromatography. Eur. J.
Biochem. 195, 795–800.
44. Ge rke, V. & Weber, K. (1985) The regulatory chain in the p36-kd
substrate complex of viral tyrosine-specific protein kinases is
related in sequence to the S-100 protein of glial cells. EMBO J. 4,
2917–2920.
45. Kube, E., Becker, T., Weber, K. & Gerke, V. (1992) Protein–
protein interac tion stud ied b y s ite-directe d mu tagenesis. Charac -
terization of th e annexin II-binding site on p11, a m em ber of the
S100 protein family. J. Biol. Chem. 267, 14175–14182.
46. Watanabe, M., Ando, Y., Tokumitsu, H. & Hidaka, H. (1993)
Binding site of annexin XI on the ca lcy cl in molecule. Biochem.
Biophys. Res. Commun. 196, 1376–1382.
47. Filipek, A ., Wojda, U. & Lesniak, W. (1995) Interaction of cal-
cyclin and its cyanogen bromide fragmen ts with annexin II and
glyceraldehyde 3-phosphate. Int. J. Biochem. Ce ll Biol. 27, 1123–
1131.
48. Hartl, F.U. & Hayer-Hartl, M. (2002) Molecular chaperones in
the cytosol: from nascent chain to folded prote in. Science 295,
1852–1858.
49. Puig, A. & Gilbert, H.F. (1994) Protein disulfide isomerase
exhibits chaperone a nd anti-cha perone activity in the oxidative
refolding of lysozyme. J. Biol. Chem. 269, 7764–7771.
50. Chen, Y.H., He, R .Q., Liu, Y., Liu, Y. & Xue, Z.G. (2000) Effect
of human neuronal t o n d enaturati on a nd r eactivation o f rabbit

muscle
D
-glyceraldehyde-3-phosphate dehydrogenase. Biochem. J.
351, 233–240.
51. Okada, M., Hatakeyama, T., I toh, H., Tokuta, N., Tokumitsu, H.
& Kobayashi, R . ( 2004) S100A1 is a novel m olecular chaperone
and a member of the Hsp70/H sp90 multichaperone complex.
J. Biol. Chem. 279, 4221–4233.
Ó FEBS 2004 S100A12 target proteins (Eur. J. Biochem. 271) 3775