FGF-2, IL-1b and TGF-b regulate fibroblast expression
of S100A8
Farid Rahimi, Kenneth Hsu, Yasumi Endoh and Carolyn L. Geczy
Inflammatory Diseases Research Unit, School of Medical Sciences, University of New South Wales, Sydney, Australia
Fibroblasts are heterogeneous stromal resident cells
that participate in wound healing, fibrosis ⁄ scarring
and immune ⁄ inflammatory processes [1,2] by contri-
buting to leukocyte recruitment ⁄ accumulation, angio-
genesis, matrix metabolism, and protection against
oxidative damage [3,4]. Numerous factors inclu-
ding extracellular matrix (ECM) components, some
cytokines, prostaglandins, reactive oxygen species
(ROS), and growth factors [5] modulate fibroblast
function.
Keywords
FGF-2; fibroblasts; interleukin-1b; S100A8
gene; TGF-b
Correspondence
C. Geczy, Inflammatory Diseases Research
Unit, School of Medical Sciences, The
University of New South Wales, Sydney,
NSW 2052, Australia
Fax: + 61 293851389
Tel: + 61 293851599
E-mail:
Website: />(Received 3 March 2005, revised 28 March
2005, accepted 5 April 2005)
doi:10.1111/j.1742-4658.2005.04703.x
Growth factors, including fibroblast growth factor-2 (FGF-2) and transform-
ing growth factor-b (TGF-b) regulate fibroblast function, differentiation and
proliferation. S100A8 and S100A9 are members of the S100 family of Ca

HRP, horseradish-peroxidase; IFNc, interferon c; IL-1b, interleukin-1b; JNK, c-Jun N-terminal kinase; LDL, low-density lipoprotein; LPS,
endotoxin; mS100A8, murine S100A8; MAPK, mitogen-activated protein kinase; MEK, MAPK kinase; mOxS100A8, HOCl-oxidized murine
S100A8; PKC, protein kinase C; ROS, reactive oxygen species; SPF, splenic ‘primary’ fibroblast-like cells; TGF-b, transforming growth
factor-b; TNF, tumor-necrosis factor.
FEBS Journal 272 (2005) 2811–2827 ª 2005 FEBS 2811
S100 Ca
2+
-binding proteins may have evolved from
a calmodulin ancestor [6] and are implicated in numer-
ous intra ⁄ extracellular processes and can act as Ca
2+
sensors [7]. Some (e.g. S100A6 (calcyclin) [8], S100A4
[9] and S100A11 [10,11]) have been implicated in fibro-
blast growth and differentiation. S100A12 acts as
a pro-inflammatory ‘cytokine’ [12,13] and S100A12
mRNA is induced by interleukin-1a (IL-1a) and tumor
necrosis factor (TNF) in bovine corneal fibroblasts
[14]. S100B is also expressed by fibroblasts [15] and
may be involved in regulation of growth arrest and
apoptosis [16], stimulation of cell proliferation [17]
and protection against apoptosis [18]. Other S100
proteins also have proliferative and anti-apoptotic
effects [8,9,19].
S100A8 and S100A9 (calgranulins A and B; MRP8
and MRP14) have been associated with leukocyte dif-
ferentiation, inflammation and wound healing [20,21].
They are proposed to be involved in reorganization of
the keratin cytoskeleton and differentiation of kera-
tinocytes and in antibacterial or antioxidant defense in
the wounded or normal epidermis [20–22]. Some func-

cellular responses to oxidative stress. In particular in
keratinocytes, S100A2 oxidation and translocation
were proposed as early markers of oxidative stress and
were markedly attenuated in malignant keratinocytes,
favoring a role in oxidant defense rather than in tumor
proliferation [33].
Here we show that factors important in wound heal-
ing regulate S100A8, but not S100A9, in fibroblasts.
Fibroblast growth factor-2 (FGF-2) and IL-1b
strongly induced S100A8 via a MAPK-dependent
pathway. Responses to FGF-2 were amplified by hep-
arin and there was strong synergy between FGF-2 and
IL1b. The protein was cytoplasmic. TGF-b suppressed
S100A8 induction by FGF-2 but not by IL-1b, sug-
gesting important regulatory differences, and promoter
analysis confirmed different enhancer elements regula-
ting induction by IL-1b and FGF-2. In a rat incisional
wound, immunohistochemical studies showed S100A8
expression in fibroblast- and macrophage-like cells,
keratinocytes and neutrophils in the incision area. We
propose that S100A8 may be involved in pathways
regulating fibroblast growth and differentiation, pos-
sibly by regulating intracellular redox, at sites of
inflammation and ⁄ or repair ⁄ remodeling.
Results
FGF-2 induces S100A8 mRNA in 3T3 fibroblasts
Initially, induction of murine S100A8 (mS100A8)
mRNA in 3T3 fibroblasts was variable, suggesting
that, like microvascular endothelial cells [34], cell–cell
contact may be important. Fibroblast monolayers

and heparin induced strong responses that were maxi-
mal with 3 nm FGF-2 (Fig. 2A); 6 and 15 nm FGF-2
induced mS100A8 mRNA levels that were  80 and
60% of maximal expression, respectively (Fig. 2A),
suggesting production of a suppressor. Heparin gener-
ated maximal responses at 1 and 10 IUÆmL
)1
(Fig. 2B)
whereas higher amounts (50 and 100 IUÆmL
)1
) reduced
responses to  72 and 48% of maximum, respectively,
in cells costimulated with 1.5 nm FGF-2 (Fig. 2B)
possibly due to soluble heparin-mediated inhibition of
FGF-2-receptor binding [35]. For subsequent experi-
ments, 1.5 nm FGF-2 with 1 IUÆmL
)1
heparin were
used to stimulate confluent fibroblasts.
S100A8 mRNA induction in 3T3 fibroblasts activa-
ted with FGF-2 was evident after 8 h and in the pres-
ence of heparin, mRNA levels increased in parallel up
to 12 h when the response to FGF-2 was maximal and
gradually declined over 36 h (Fig. 2C). Potentiation by
heparin was most apparent at 18 h, when mRNA lev-
els were approximately double those with FGF-2 alone
at 12 h. In FGF-2-heparin-stimulated cells, S100A8
mRNA levels were maintained for up to 36 h and
declined to 20% of maximum by 48 h (Fig. 2C).
Because the mS100A8 gene in elicited macrophages

). (C) Kinetics of induction of S100A8
mRNA. 3T3 cells treated with FGF-2 (1.5 n
M) or FGF-2 ⁄ heparin
(1 IUÆmL
)1
) were harvested at the times indicated and Northern
analysis performed. Similar results observed in three different
experiments.
F. Rahimi et al. Expression and regulation of S100A8 in fibroblasts
FEBS Journal 272 (2005) 2811–2827 ª 2005 FEBS 2813
1000 ngÆmL
)1
), interferon c (IFNc, 500 UÆmL
)1
), or
TNF (30 ngÆmL
)1
) were tested. No S100A8 mRNA
was detected in 3T3 cells stimulated for 24 h although
mRNA was substantially augmented with a combina-
tion of LPS, IFNc, and TNF and confluence influ-
enced levels (not shown). Induction was maximal after
24 h and decreased to  47% of maximum by 48 h
(not shown).
IL-1b-stimulated 3T3 fibroblasts express S100A8
mRNA
IL-1b is a strong inducer of the S100A8 gene in micro-
vascular endothelial cells [34]. IL-1b induced S100A8
mRNA in confluent 3T3 fibroblasts in a dose- and
time-dependent manner. As little as 1 UÆmL

)1
) induced mRNA levels
similar to those induced by FGF-2 and heparin. The
magnitude of synergy was more apparent, with  580-
fold more S100A8 mRNA in cells costimulated with
FGF-2, heparin and IL-1b compared to those stimula-
ted with FGF-2 and heparin.
AB
D
C
Fig. 3. Effects of IL-1b on S100A8 mRNA
induction. (A) Northern analysis of mRNA
from confluent 3T3 cells stimulated with the
given doses of IL-1b for 24 h. Results repre-
sent three experiments. (B) 3T3 cells stimu-
lated with 10 UÆmL
)1
IL-1b were harvested
at the times indicated. The line graph indi-
cates normalized levels of S100A8 mRNA.
Similar results observed in two experi-
ments. (C) Northern blot analysis of conflu-
ent 3T3 cells stimulated for 24 h with FGF-2
(F, 1.5 n
M) and heparin (H, 1 IUÆmL
)1
)inthe
presence of increasing doses of IL-1b; data
representative of at least three different
experiments. (D) Confluent SPF stimulated

fibroblast-like cells (SPF) and bone marrow-derived
fibroblast-like cells (BMF) were analyzed. Confluent
SPF contained no detectable S100A8 mRNA by
northern analysis (Fig. 3D) and cells stimulated for 24 h
with IL-1b (samples 8 and 9), FGF-2 (sample 10), or
FGF-2 and heparin (samples 4 and 5) did not express
the gene. IL-1b and FGF-2 combined (samples 6 and 7,
Fig. 3D) weakly induced mRNA ( 38–47% of maxi-
mum), responses at 24 h were maximal with 6 nm FGF-2,
1IUÆmL
)1
heparin and 10 UÆmL
)1
IL-1b (sample 3)
and half-maximal with 2 UÆmL
)1
IL-1b and 1.5 nm
FGF-2 (sample 2, Fig. 3D). The same pattern was
observed with BMF stimulated with FGF-2, heparin
and IL-1b (Fig. 6D). RT-PCR confirmed that S100A9
was not induced in primary murine fibroblast-like cells
by FGF-2, heparin and IL-1b stimulants (not shown).
Promoter analysis in 3T3 fibroblasts
To examine mechanisms of transcriptional regulation
of the S100A8 gene by IL-1b, and FGF-2 plus hep-
arin, 5¢-flanking sequences upstream of the transcrip-
tion initiation site, untranslated intron 1 and sequences
upstream of exon 1, were used to evaluate activities of
deletion constructs after transient transfection into 3T3
cells (Fig. 4). Marked differences between FGF-2-

FEBS Journal 272 (2005) 2811–2827 ª 2005 FEBS 2815
protein (C ⁄ EBP), Ets, and E box are located within
this region. Constructs not containing the 1st exon and
intron ()178–0 bp) generated positive, but relatively
weak luciferase activities but in the same proportions,
indicating that this region still contains elements essen-
tial for gene induction by FGF-2.
The MAPK pathway is involved in S100A8 mRNA
induction by FGF-2 and IL-1b in 3T3 fibroblasts
The MAPK pathway is implicated in S100A8 mRNA
induction by IFNc and LPS in macrophages [26]. In
3T3 cells, PD 098059 (MAPK kinase (MEK) inhibitor,
50 lm) suppressed FGF-2- and FGF-2-heparin-
induced S100A8 mRNA by  65 and 62%, respect-
ively, and SB 202190 [c-Jun N-terminal kinase
(JNK) ⁄ p38 inhibitor, 10 lm]by 50% (Fig. 5A).
Similarly, PD 098059 (50 or 75 lm) or SB 202190 (10
or 20 lm) reduced responses to IL-1b by  70–83%
(Fig. 5B) indicating converging pathways in FGF-2- ⁄
heparin- and IL-1b-induced S100A8 mRNA induction
in fibroblasts.
Cycloheximide completely abrogated the S100A8
gene in 3T3 cells stimulated with FGF-2 ± heparin
(Fig. 5C) or IL-1b (Fig. 5D), indicating a requirement
for de novo protein synthesis.
TGF-b suppresses FGF-2-induced mS100A8 mRNA
in 3T3 and primary fibroblasts
Because TGF-b modulates fibroblast function and phe-
notype [38,39], its effect on S100A8 gene expression
was tested. 3T3 cells cultured with TGF-b, heparin,

C
D
Fig. 5. Involvement of the MAPK pathways and de-novo protein
synthesis in induction of S100A8 mRNA in fibroblasts. (A) 3T3 cells
stimulated (24 h) with FGF-2 (F, 1.5 n
M) or FGF-2 and heparin
(H, 1 IUÆmL
)1
) with ⁄ without 4-h preincubation with PD 098059
(PD, 50 l
M) or SB 202190 (SB, 10 lM) as indicated. (B) 3T3 cells sti-
mulated (24 h) with IL-1b (10 UÆmL
)1
) with ⁄ without 4-h preincuba-
tion with PD 098059 (50–75 l
M) and SB 202190 (10–20 lM). Data
represent three experiments. (C) 3T3 fibroblasts stimulated with
FGF-2 (F, 1.5 n
M) and heparin (H, 1 IUÆmL
)1
) for 24 h with or without
4-h preincubation with 5 lgÆmL
)1
cycloheximide (CHX) as indicated.
C, mRNA from unstimulated cells; BM, murine bone-marrow RNA.
(D) 3T3 cells stimulated with IL-1b (10 UÆmL
)1
) for 24 h with or with-
out 4-h preincubation with cycloheximide (5 and 10 lgÆmL
)1

disulfide-linked complexes and the characteristic mono-
meric mass of 10 kDa (rA8; Fig. 7A) was confirmed.
In an alternative approach, supernatants and lysates
of activated cells were concentrated using an affinity
A
B
C
D
Fig. 6. TGF-b regulates S100A8 mRNA induction by FGF-2. (A) Nor-
thern analysis of mRNA from 3T3 cells stimulated with FGF-2
(F, 1.5 n
M), heparin (H, 1 IUÆmL
)1
), TGF-b (T, 0.08 nM) alone or in the
combinations indicated for 24 h; C, unstimulated cells. (B) RT-PCR
analysis of the RNA samples used in (A). Levels of S100A8 mRNA
were compared to those of HPRT in the corresponding samples and
ratios of S100A8 mRNA ⁄ HPRT are indicated for each sample. (C)
Cells stimulated with FGF-2 (F, 1.5 n
M) ⁄ heparin (H, 1 IUÆmL
)1
) with
or without IL-1b (10 UÆmL
)1
) in the presence or absence of increas-
ing concentrations of TGF-b (8 p
M-800 pM) as indicated. BM, total
bone-marrow RNA. (D) Confluent BMF treated for 24 h with the indi-
cated doses of FGF-2, heparin, IL-1b and TGF-b.
A

phase HPLC. Recombinant mS100A8 elutes as a single
peak at 19.8–20 min (not shown). Because no major
peak were obtained (possibly due to low levels),
fractions were collected between 17.25 and 21.25 min
(covering the expected retention-time range for native
mS100A8 monomer and dimer), and lyophilized. No
mS100A8 was detected in fractions from supernatants
(not shown) or lysates (lane 1, Fig. 7B) of unstimu-
lated cells. Western blotting of the three fractions from
lysates of stimulated cells collected over 18.25–20.25 min
(lane 2, fraction collected at 18.25–19.25 min, Fig. 7B)
contained components of molecular mass 20 kDa, with
the same migration profile as dimeric mS100A8,
contained in the positive control (lane 3, Fig. 7B). No
mS100A8 monomer (10 kDa) was detected. This was
unexpected as the same conditions have yielded mono-
meric and complexed forms of mS100A8 in other
stimulated cell types [34,41].
No S100A8 was found in unstimulated 3T3 cells
stained with preimmune IgG or an antibody against
HOCl-oxidized mS100A8 (mOxS100A8) (Fig. 8A,B,
respectively); 4,6-diamidino-2-phenylindole (DAPI)-
stained nuclei were evident. Similarly, stimulated cells
stained with the nonimmune IgG were unreactive
(Fig. 8C). Approximately 10–30% of 3T3 cells stimula-
ted with FGF-2 ⁄ heparin ⁄ IL-1b showed bright cyto-
plasmic fluorescence (Fig. 8D). When costimulated
with TGF-b, S100A8-positive cells dropped to 5% of
total (Fig. 8E). Confocal microscopy clearly showed
localization of S100A8 (red fluorescence) in the cyto-

wounded rat skin 2 days post injury showed an inten-
sely anti-S100A8-immunoreactive scab, containing
S100A8-positive neutrophils (Fig. 9Aa), impinging on
the injured, and surrounding the normal epidermis. In
normal epidermis, the superficial more differentiated
keratinocytes reacted with anti-mS100A8 more inten-
sely than those in the stratum basale which contains
the proliferating keratinocytes. In the dermis, small
dilated microvessels containing intensely stained
S100A8-positive neutrophils were evident, representing
the in situ positive controls (Fig. 9Ab). Based on
morphology, extravascular S100A8-positive cells were
identified as macrophages or extravasating neutro-
phils. Some spindle-shaped fibroblast-like cells closely
apposed to, and aligned with collagen fibers were also
relatively intensely mS100A8-positive (Fig. 9Ab)
although staining was heterogeneous. These were iden-
tified as fibroblast-like cells, based on morphology,
Fig. 9. Immunohistochemical localization of S100A8 in rat dermal wounds. (A) Immunostaining of rat dermal wound 2 days after injury. (Aa)
Low-power (200 ·) view of the wounded dermis beneath the neutrophil-rich S100A8-positive scab (S). Black arrows indicate S100A8-positive
fibroblast-like cells apposed to collagen fibers indicated by C. Neovessels and sebaceous glands are indicated by V and SG, respectively.
Delineated inset in (Aa) corresponds to (Ab) (400 ·) where neutrophils and macrophage-like cells are indicated by red and blue arrows,
respectively. Black arrows point to S100A8-positive fibroblast-like cells. (B) Staining of rat wound 4 days after injury. Anti-S100A8 IgG
(Ba, Bb) and nonimmune IgG (Bc) were used. (Ba) The scab (S) is evident along the wounded epidermis and encroaching normal uninjured
keratinocytes (K). Some hair follicles (H), small vessels (V) and sebaceous glands (SG) surrounded by collagen (C) fibers are evident. (Bb) An
area of granulation tissue showing macrophage-like (blue arrows) and fibroblast-like cells (black arrows) around neovessels (V) were S100A8-
positive. Some S100A8-negative fibroblast-like cells are indicated by asterisks. (Bc) Immunostaining of the 4-day wound with nonimmune
IgG. (C) Anti-mS100A8 immunostaining of rat wound 7 days after injury. (Ca) Low-power view of an area of mature granulation tissue and
scar formation rich in collagen fibers and fibroblasts stained with anti-mS100A8 IgG. The epidermal keratinocytes are evident to the right of
the wounded area (K). The inset shown in (Ca) corresponds to (Cb) where S100A8-negative fibroblast-like cells are indicated by arrows.

are influenced by mediators that regulate inflamma-
tion, wound healing, fibrosis, oxidative processes and
vascular remodeling. This study provides the first evi-
dence that S100A8 is regulated in fibroblasts by partic-
ular growth factors, further supporting its role in
inflammatory processes.
FGFs stimulate proliferation of many cell types
involved in wound healing including endothelial cells,
fibroblasts and keratinocytes and are essential for sur-
vival, replication, differentiation and migration of var-
ious cell types during embryonic and fetal development
[42,43]. Interestingly, S100A8 has roles in cell migra-
tion [24,25] and S100A8, but not S100A9 [29,44], is
essential for embryonic development [45]. FGF-2
up-regulated S100A8, but not S100A9, mRNA in 3T3
and primary fibroblasts. A fragment of human S100A8
(S100A8
21)45
) is chemotactic for fibroblast-like perio-
dontal ligament cells [46] and this may be a function
of S100A8 released as a consequence of wound healing
and in an inflammatory environment. S100A8 induc-
tion was confluence-dependent (Fig. 1), a requirement
similar to its induction in microvascular endothel-
ial cells [34]. Although fibroblasts do not normally
establish contacts in vivo, in culture they establish gap
junctions [37] indicating metabolic interdependence.
Cell–cell ⁄ cell–ECM contacts may provide additional
signals for optimal induction of the S100A8 gene
in vitro and because of the potentiation exhibited by

gene coinduction. The stability of S100A8 protein in
neutrophils is suggested to be dependent on S100A9
coexpression [29], but, like the situation in activated
murine macrophages [26,27] and keratinocytes [22],
S100A9 was not coinduced with S100A8 in any fibro-
blast type tested, strongly supporting our proposal that
S100A8 ⁄ S100A9 coexpression is not mandatory for the
function or stability of murine S100A8. This may not
be the case with the human proteins that are generally
coexpressed [49] although S100A8 alone was detected
in the human LDL proteome by mass spectrometry
and peptide mass fingerprinting [28].
Negligible S100A8 was secreted in response to IL-1b
and FGF-2 ⁄ heparin (Fig. 7A). This is in stark contrast
with activated macrophages which secrete high levels
in response to various stimulants, particularly LPS
with IL-10 or prostaglandin E
2
[26,27], but similar to
IL-1-activated microvascular endothelial cells [34] and
Expression and regulation of S100A8 in fibroblasts F. Rahimi et al.
2820 FEBS Journal 272 (2005) 2811–2827 ª 2005 FEBS
UVA-irradiated keratinocytes [22], where the protein is
located in the nuclei and cytoplasm of a proportion of
cultured cells. The high levels of S100A8 mRNA found
in FGF-2 ⁄ heparin-activated fibroblasts costimulated
with IL-1b correlated well with protein levels detected
by western blotting (Fig. 7), and S100A8-positive
fibroblasts were more numerous (Fig. 8). S100A8
induced by FGF-2 ⁄ heparin ⁄ IL-1b was mainly cyto-

-dependent mechanisms in macrophages [26,53],
and these results suggest complex control involving
converging pathways for its induction in fibroblasts.
Promoter deletion analyses indicated differential
regulation of S100A8 induction by IL-1b and FGF-
2 ⁄ heparin (Fig. 4). The minimal promoter required for
FGF-2- ⁄ heparin-induced responses was restricted to
the region from )94 to +465 (Fig. 4). Deletion of the
region from )178 to )94 bp, which contains activated
protein-1, Ets and C ⁄ EBP motifs, totally negated
activity induced by LPS⁄ IL-10 in macrophages [27],
indicating distinct responsive elements activated by
FGF-2 ⁄ heparin in fibroblasts and LPS ⁄ IL-10 in macr-
ophages. IL-1b-responsive elements may be distinct
from FGF-2-responsive elements as they were not
located within )917 to +465 bp and preliminary
experiments with a construct spanning 4.9 kb upstream
of the transcriptional start site strongly indicate differ-
ences in transcriptional regulation. The strong synergy
generated by the combination of FGF-2 ⁄ heparin and
IL-1b may be mediated through distinct enhancer ⁄
responsive elements in the two regions or IL-1b may
enhance S100A8 mRNA stability.
TGF-b accelerates deposition, remodeling and mat-
uration of collagen later in the healing process of der-
mal wounds [54,55]. Differentiation of fibroblasts from
a proliferative phenotype to collagen-producing and
then to myofibroblastic phenotypes is believed to be
mediated by TGF-b [39]. IL-1b and FGF-2 promote
proliferation [56,57], whereas TGF-b promotes differ-

inflammation in the growing scar tissue and less oxida-
tive stress as levels of tissue antioxidants including
glutathione, catalase, superoxide dismutase, glutathione-
S-transferase, and glutathione peroxidase partially or
completely recover as healing progresses [62]. This
correlated well with the immunohistochemical studies
in the healing rat wound. Some spindle-shaped fibro-
blast-like cells in the granulation tissue were S100A8-
positive 2 and 4 days after dermal injury whereas
almost all fibroblasts were negative 7 days post injury
(Fig. 9).
The increasing importance of various S100 proteins
in regulating oxidative processes is emerging [33]
and in the mouse, S100A8, but not S100A9, was
induced in dermal keratinocytes following oxida-
tive stress and by ultraviolet A irradiation, and gene
F. Rahimi et al. Expression and regulation of S100A8 in fibroblasts
FEBS Journal 272 (2005) 2811–2827 ª 2005 FEBS 2821
induction was dependent on generation of ROS [22].
Interestingly, under nonreducing ⁄ nondenaturing condi-
tions, only mS100A8 dimer was detected in activated
fibroblast lysates (Fig. 7B), a structural modification
induced in S100A8 by oxidants such as peroxide [31],
suggesting a function partially analogous to S100A2
[63]. S100A2 also forms disulfide bonds in response to
H
2
O
2
and is sensitive to early cellular responses to oxi-

one-S-transferase fusion protein expressed in Escherichia
coli as described previously [67]. For production of poly-
clonal antibodies, S100A8 and mOxS100A8 (50 lg) [30]
bound to nitrocellulose particles in Freund’s complete adju-
vant (Sigma, St Louis, MO, USA) were prepared as des-
cribed [67], and injected intradermally into New Zealand
white rabbits. Rabbits were boosted after 4 and 8 weeks
with 100 lg mOxS100A8 or mS100A8 in incomplete Fre-
und’s adjuvant (Sigma). IgG from pooled sera was purified
by Protein A-Sepharose (Amersham Pharmacia Biotech,
Buckinghamshire, UK). Titer and reactivities of IgG with
mS100A8 ⁄ mOxS100A8 were tested by immunoblotting and
ELISA as described [40]. Pre-immune sera did not react
with mS100A8 or mOxS100A8. Anti-mOxS100A8 and anti-
mS100A8 did not cross-react with S100A9 monomer but
recognized the murine and rat S100A8 monomer and oxi-
dized forms from 20 kDa to <100 kDa.
Cytokines, growth factors and chemicals were obtained
from the following sources: recombinant human FGF-2
and recombinant human TGF-b1 were from Sigma, murine
IL-1b from Genzyme (Cambridge, MA), and murine IFNc
from Genzyme, or Genentech, or Sigma. TNF was from
Genzyme or Sigma, LPS (E. coli, 055:B5) was from Difco
(Detroit, MI). The MAPK inhibitors, PD 098059 and SB
202190, Dulbecco’s phosphate-buffered saline (DPBS), col-
lagenase, N-2-hydroxyethylpiperazine-N¢-2-ethanesulfonic
acid (Hepes), sodium pyruvate, l-glutamine, penicillin,
streptomycin, trypsin-EDTA solution, Triton X-100, Tween
20, DTT, ovalbumin, BSA, saponin, normal goat serum,
agarose-bound caprine anti-rabbit IgG, ActD, cyclohexi-

were removed, minced into small pieces, and digested for
20 min at 37 °C in collagenase (2 mgÆmL
)1
) in DPBS sup-
plemented with CaCl
2
(1 mm) then passed through a tissue
strainer (70 lm, BD Biosciences, San Jose, CA, USA), and
washed three times in RPMI 1640 containing 10% (v ⁄ v)
heated (56 °C, 30 min) FBS, Hepes (10 mm), sodium pyru-
vate (1 mm), l-glutamine (2 mm), 100 UÆmL
)1
penicillin,
100 lgÆmL
)1
streptomycin, 2-mercaptoethanol (50 lm), and
NaHCO
3
(19 mm), hereafter referred to as RPMI culture
medium (RPMI CM). Cells were resuspended in RPMI
CM in tissue-culture dishes (100 mm diameter) and main-
tained at 37 °C in humidified 5% CO
2
in air. Non-adherent
cells and debris were removed by replenishing RPMI CM
after 24, 48 and 72 h. Cells were replenished with fresh
medium every 3 days until they reached confluence. SPF
from passages 2 through 8 were used in the experiments.
Primary fibroblasts obtained by this method do not contain
macrophages or other cells of hematopoietic origin [68] and

was added to arrest digestion, cells harvested and seeded
onto tissue-culture flasks or 60-mm dishes (BD Biosciences)
for maintenance or stimulation (2–3 · 10
6
cells per dish).
Generally, cells grown to 100% confluence were activated
with the desired stimulants after a DPBS wash and replen-
ishment with fresh CM. Cell viability by Trypan-blue exclu-
sion was > 90–97%.
To determine whether de novo protein synthesis was
required for induction of S100A8 mRNA, confluent 3T3
cells were preincubated with the protein synthesis inhibitor,
cycloheximide (5–10 lgÆ mL
)1
), for 4 h before stimulation
with activators and RNA analysis performed after 24 h.
For inhibition of RNA synthesis, fibroblasts were stimula-
ted with particular stimulants for 20 h before an additional
20-h incubation with ActD (10 lgÆmL
)1
) and RNA levels
analyzed before and after ActD addition. To determine the
role of the MAPK pathway in S100A8 mRNA induction,
3T3 cells were preincubated for 4 h with inhibitors (PD
098059, the MEK inhibitor, 50–75 lm; SB 202190,
JNK ⁄ p38 inhibitor, 10–20 lm) before stimulation.
RNA extraction and Northern analysis
Total cellular RNA from 2.5 to 3.5 · 10
6
fibroblasts was

Reporter assays
The mS100A8-promoter-luciferase-fused reporter plasmids
were described previously [27]. Sub-confluent 3T3 cells
grown in 24-well plates were transiently transfected with
0.8 lg firefly luciferase constructs, or the parent plasmid
pGL2-basic DNA, or the pGL2 promoter DNA com-
bined with 0.03 lg of the control pRL-TK plasmid
(Promega) in the presence of LipofectAMINE. After
24 h, cells were replenished with fresh DMEM CM and
stimulated with FGF-2 (1.5 nm) + heparin (1 IUÆmL
)1
)
or with IL-1b (10 UÆmL
)1
) for another 48 h. Firefly and
Renilla luciferase activities in cell extracts (20 lL) were
assayed using the Dual-Luciferase Assay System in a TD-
20 ⁄ 20 Luminometer (Turner Design, Sunnyvale, CA,
USA) following manufacturer’s instructions. Promoter
activity was normalized to Renilla luciferase which resul-
ted in reproducible and constant values relative to pGL2
promoter.
Protein purification and Western blotting
Control or stimulated (28–30 h) 3T3 cells (20–25 · 10
6
)
were washed twice with DPBS after removal of super-
natants, lysed in DPBS containing 1% (v ⁄ v) Triton X-100,
50 mm Tris ⁄ HCl, pH 8.0 and Complete Protease Inhibitors,
subjected to three freeze-thaw cycles and sonicated on ice

(Nalge Nunc International, Roskilde, Denmark) were acti-
vated with stimulants for 30 h, rinsed thrice in NaCl ⁄ P
i
,
slides air-dried then re-hydrated in NaCl ⁄ P
i
(10 min), fixed
in 4% paraformaldehyde (10 min) and permeabilized
(10 min) with 0.5% saponin ⁄ 0.1% BSA in NaCl ⁄ P
i
. After
blocking with normal goat serum (20% w ⁄ v in saponin–
BSA solution, 20 min), and incubation with anti-
mOxS100A8 IgG or preimmune rabbit IgG (both at
10 lgÆmL
)1
in saponin–BSA) overnight at 4 °C, slides were
rinsed in NaCl ⁄ P
i
and incubated with anti-rabbit IgG-
Alexa-Fluor-568 (1 : 200, in saponin–BSA) for 1 h in the
dark at 25 °C, rinsed in NaCl ⁄ P
i
(3 · 5 min) and nuclei
stained with DAPI (0.3 nm in NaCl ⁄ P
i
, 10 min in the dark).
Slides were mounted in Vectashield Mounting Medium
(Vector Laboratories, Burlingame, CA, USA), cover-slipped
and examined using a Leica DM IRB inverted confocal

(3% v ⁄ v in NaCl ⁄ P
i
, 30 min) and 10% normal
goat serum (30 min), incubated for 1 h (25 °C) with rabbit
anti-mS100A8 IgG or nonimmune rabbit IgG (both
10 lgÆmL
)1
) in NaCl ⁄ P
i
containing 0.05% saponin ⁄ 0.1%
BSA. Optimal IgG concentrations were predetermined by
titration. Slides were washed with NaCl ⁄ P
i
and bound anti-
body detected after 30-min incubation with biotinylated
caprine anti-rabbit IgG (Dako) and streptavidin-peroxidase
(Kirkegaard & Perry Laboratories, Inc. Gaithersburg, MA,
USA). Visualization was with diaminobenzidine preceding
counterstaining with hematoxylin. For other controls, pri-
mary or secondary antibodies were omitted.
Acknowledgements
The authors are obliged to Professor Rolf Howlett for
the preparation and donation of wound specimens
and Drs M. Raftery, R. Passey, N. Tedla, Z. Yang,
W. X. Yan and H. Cai for technical help ⁄ advice.
Funding from the National Health and Medical
Research Council of Australia is acknowledged; FR
held an Australian Postgraduate Award.
References
1 Stadelmann WK, Digenis AG & Tobin GR (1998) Phy-

tion. J Invest Dermatol 116, 956–963.
10 Sakaguchi M, Miyazaki M, Inoue Y, Tsuji T, Kouchi H,
Tanaka T, Yamada H & Namba M (2000) Relationship
between contact inhibition and intranuclear S100C of
normal human fibroblasts. J Cell Biol 149, 1193–1206.
11 Sakaguchi M, Tsuji T, Inoue Y, Miyazaki M, Namba
M, Yamada H & Tanaka T (2001) Loss of nuclear loca-
lization of the S100C protein in immortalized human
fibroblasts. Radiat Res 155, 208–214.
12 Yang Z, Tao T, Raftery MJ, Youssef P, Di Girolamo
N & Geczy CL (2001) Proinflammatory properties of
the human S100 protein S100A12. J Leukoc Biol 69,
986–994.
13 Hofmann MA, Drury S, Fu C, Qu W, Taguchi A, Lu Y,
Avila C, Kambham N, Bierhaus A, Nawroth P, Neurath
MF, Slattery T, Beach D, McClary J, Nagashima M,
Expression and regulation of S100A8 in fibroblasts F. Rahimi et al.
2824 FEBS Journal 272 (2005) 2811–2827 ª 2005 FEBS
Morser J, Stern D & Schmidt AM (1999) RAGE med-
iates a novel proinflammatory axis: a central cell surface
receptor for S100 ⁄ calgranulin polypeptides. Cell 97, 889–
901.
14 Gottsch JD, Li Q, Ashraf F, O’Brien TP, Stark WJ &
Liu SH (1999) Cytokine-induced calgranulin C expres-
sion in keratocytes. Clin Immunol 91, 34–40.
15 Shiras A, Shepal V, Bhonde M & Shastry P (2001) S100
expression in primary mouse fibroblast cultures in vitro.
Cell Dev Biol Anim 37, 172–174.
16 Scotto C, Deloulme JC, Rousseau D, Chambaz E &
Baudier J (1998) Calcium and S100B regulation of p53-

by ultraviolet A irradiation is dependent on reactive
oxygen intermediates. J Invest Dermatol 121, 1168–1174.
23 Ahmad A, Bayley DL, He S & Stockley RA (2003)
Myeloid related protein-8 ⁄ 14 stimulates interleukin-8
production in airway epithelial cells. Am J Respir Cell
Mol Biol 29, 523–530.
24 Lackmann M, Rajasekariah P, Iismaa SE, Jones G,
Cornish CJ, Hu S, Simpson RJ, Moritz RL & Geczy
CL (1993) Identification of a chemotactic domain of the
pro-inflammatory S100 protein CP-10. J Immunol 150,
2981–2991.
25 Vogl T, Ludwig S, Goebeler M, Strey A, Thorey IS,
Reichelt R, Foell D, Gerke V, Manitz MP, Nacken
W, Werner S, Sorg C & Roth J (2004) MRP8 and
MRP14 control microtubule reorganization during
transendothelial migration of phagocytes. Blood 104,
4260–4268.
26 Xu K & Geczy CL (2000) IFN-c and TNF regulate
macrophage expression of the chemotactic S100 protein
S100A8. J Immunol 164, 4916–4923.
27 Xu K, Yen T & Geczy CL (2001) IL-10 up-regulates
macrophage expression of the S100 protein S100A8.
J Immunol 166, 6358–6366.
28 Karlsson H, Leanderson P, Tagesson C & Lindahl M
(2005) Lipoproteomics I: Mapping of proteins in low-
density lipoprotein using two-dimensional gel
electrophoresis and mass spectrometry. Proteomics 5,
551–565.
29 Hobbs JA, May R, Tanousis K, McNeill E, Mathies M,
Gebhardt C, Henderson R, Robinson MJ & Hogg N

fibroblast heterogeneity and the role of fibroblast subpo-
pulations in fibrosis. Clin Immunol Immunopathol 72,
283–292.
37 Sappino AP, Schurch W & Gabbiani G (1990) Differen-
tiation repertoire of fibroblastic cells: expression of
cytoskeletal proteins as marker of phenotypic modula-
tions. Lab Invest 63, 144–161.
38 Kurosaka H, Kurosaka D, Kato K, Mashima Y &
Tanaka Y (1998) Transforming growth factor-b1 pro-
motes contraction of collagen gel by bovine corneal
F. Rahimi et al. Expression and regulation of S100A8 in fibroblasts
FEBS Journal 272 (2005) 2811–2827 ª 2005 FEBS 2825
fibroblasts through differentiation of myofibroblasts.
Invest Ophthalmol Vis Sci 39, 699–704.
39 Mattey DL, Dawes PT, Nixon NB & Slater H (1997)
Transforming growth factor b1 and interleukin 4
induced a smooth muscle actin expression and myofi-
broblast-like differentiation in human synovial fibro-
blasts in vitro: modulation by basic fibroblast growth
factor. Ann Rheum Dis 56, 426–431.
40 Kocher M, Kenny PA, Farram E, Abdul Majid KB,
Finlay-Jones JJ & Geczy CL (1996) Functional chemo-
tactic factor CP-10 and MRP-14 are abundant in mur-
ine abscesses. Infect Immun 64, 1342–1350.
41 Raftery MJ, Harrison CA, Alewood P, Jones A &
Geczy CL (1996) Isolation of the murine S100 protein
MRP14 (14 kDa migration-inhibitory-factor-related
protein) from activated spleen cells: characterization of
post-translational modifications and zinc binding.
Biochem J 316, 285–293.

lier K, Compan D, Bernard B, Lahfa M, Leclaire J,
Medaisko C, Mehul B, Seite S, Mauviel A & Dubertret
L (2003) Modulation of gene expression induced in
human epidermis by environmental stress in vivo.
J Invest Dermatol 121, 1447–1458.
50 Lu G, Beuerman RW, Zhao S, Sun G, Nguyen DH,
Ma S & Kline DG (1997) Tumor necrosis factor-a and
interleukin-1 induce activation of MAP kinase and SAP
kinase in human neuroma fibroblasts. Neurochem Int
30, 401–410.
51 O’Neill LA & Greene C (1998) Signal transduction
pathways activated by the IL-1 receptor family: ancient
signaling machinery in mammals, insects, and plants.
J Leukoc Biol 63, 650–657.
52 Cobb MH (1999) MAP kinase pathways. Prog Biophys
Mol Biol 71, 479–500.
53 Roth J, Goebeler M, Wrocklage V, van den Bos C &
Sorg C (1994) Expression of the calcium-binding pro-
teins MRP8 and MRP14 in monocytes is regulated by a
calcium-induced suppressor mechanism. Biochem J 301
(3), 655–660.
54 Gold LI, Sung JJ, Siebert JW & Longaker MT (1997)
Type I (RI) and type II (RII) receptors for transforming
growth factor-b isoforms are expressed subsequent to
transforming growth factor-b ligands during excisional
wound repair. Am J Pathol 150, 209–222.
55 Pierce GF, Tarpley JE, Yanagihara D, Mustoe TA, Fox
GM & Thomason A (1992) Platelet-derived growth fac-
tor (BB homodimer), transforming growth factor-b1,
and basic fibroblast growth factor in dermal wound

antioxidant defence enzymes in a healing cutaneous
wound. Free Radic Res 26 , 93–101.
Expression and regulation of S100A8 in fibroblasts F. Rahimi et al.
2826 FEBS Journal 272 (2005) 2811–2827 ª 2005 FEBS
63 Deshpande R, Woods TL, Fu J, Zhang T, Stoll SW &
Elder JT (2000) Biochemical characterization of S100A2
in human keratinocytes: subcellular localization, dimeri-
zation, and oxidative cross-linking. J Invest Dermatol
115, 477–485.
64 Pani G, Colavitti R, Bedogni B, Anzevino R, Borrello S
& Galeotti T (2000) A redox signaling mechanism for
density-dependent inhibition of cell growth. J Biol Chem
275, 38891–38899.
65 Moussazadeh M & Greene JJ (2000) Redox regulation
of casein kinase II autophosphorylation and its effect
on Jun-DNA binding. Cell Mol Biol (Noisy-le-Grand)
46, 1265–1275.
66 Thannickal VJ, Day RM, Klinz SG, Bastien MC, Lari-
os JM & Fanburg BL (2000) Ras-dependent and -inde-
pendent regulation of reactive oxygen species by
mitogenic growth factors and TGF-b1. FASEB J 14,
1741–1748.
67 Iismaa SE, Hu S, Kocher M, Lackmann M, Harrison
CA, Thliveris S & Geczy CL (1994) Recombinant and
cellular expression of the murine chemotactic protein,
CP-10. DNA Cell Biol 13, 183–192.
68 Pechhold K, Patterson NB, Craighead N, Lee KP, June
CH & Harlan DM (1997) Inflammatory cytokines IFN-
c plus TNF-a induce regulated expression of CD80
(B7–1) but not CD86 (B7–2) on murine fibroblasts.