Báo cáo khóa học: Nerve growth factor mediates activation of the Smad pathway in PC12 cells doc

Nerve growth factor mediates activation of the Smad pathway
in PC12 cells
Marion Lutz
1
, Kerstin Krieglstein
2
, Simone Schmitt
1
, Peter ten Dijke
3
, Walter Sebald
1
, Andrea Wizenmann
4
and Petra Knaus
1
1
Department of Physiological Chemistry II, Biocenter, University of Wu
¨
rzburg, Germany;
2
Department of Anatomy, University
of Go
¨
ttingen, Germany;
3
Division of Cellular Biochemistry, the Netherlands Cancer Institute, Amsterdam, the Netherlands;
4
JRG (of) Developmental Neurobiology, Biocenter, University of Wu
¨
rzburg, Germany

Ligand-induced oligomerization of receptors is a key step
in initiating growth factor signaling. Nevertheless, complex
biological responses often require additional trans-signaling
mechanisms involving two or more signaling cascades. For
cells of neuronal origin, it was shown that neurotrophic
effects evoked by nerve growth factor or other neurotro-
phins depend highly on the cooperativity with cytokines that
belong to the transforming growth factor b (TGF-b)
superfamily. We found that rat pheochromocytoma cells,
which represent a model system for neuronal differentiation,
are unresponsive to TGF-b1 due to limiting levels of its
receptor, TbRII. However, stimulation with nerve growth
factor leads to activation of the Smad pathway independent
of TGF-b. In contrast to TGF-b signaling, activation of
Smad3 by nerve growth factor does not occur via phospho-
rylation of the C-terminal SSXS-motif, but leads to hetero-
meric complex formation with Smad4, nuclear translocation
of Smad3 and transcriptional activation of Smad-dependent
reporter genes. This response is direct and does not require
de novo protein synthesis, as shown by cycloheximide treat-
ment. This initiation of transcription is dependent on func-
tional tyrosine kinase receptors and can be blocked by
Smad7. These data provide further evidence that the Smad
proteins are not exclusively activated by the classical TGF-b
triggered mechanism. The potential of NGF to activate the
Smad pathway independent of TGF-b represents an import-
ant regulatory mechanism with special relevance for the
development and function of neuronal cells or of other NGF-
sensitive cells, in particular those that are TGF-b-resistant.
Keywords:PC12cells;Smads;crosstalk;nervegrowth

factor; transforming growth factor-b.
Proteins of the transforming growth factor b (TGF-b)
family are multifunctional cytokines that display a very
broad range of biological activities including cell prolifer-
ation, differentiation and apoptosis [1]. TGF-bs are ubi-
quitously expressed and act on virtually all tissues, thereby
causing distinct cell-specific effects depending on the present
composition of receptors, Smad proteins and DNA-binding
partners [2,3]. Referring to cell populations of neuronal
origin, TGF-bs are described to possess neurotrophic effects
when acting in concert with other cytokines or neurotro-
phins [4,5]. Signals mediated by TGF-b are propagated by
two receptor serine/threonine kinases designated as TGF-b
type I (TbRI) and type II (TbRII) receptors [6,7]. The type
II receptors comprise TbRII [8] and its alternative splice
variant TbRII-B [9]. The initial binding of TGF-b1to
TbRII is followed by recruitment and activation of TbRI
[10]. Receptor-associated Smads (R-Smads) involved in
TGF-b signaling (Smad2 and Smad3) are phosphorylated
at the C-terminal SSXS-motif [11,12], interact with the
common mediator Smad4 [13] and translocate to the
nucleus to mediate specific transcriptional responses
[14,15]. Although Smad2 and Smad3 share 92% amino
acid identity, they are functionally distinct. A short amino
acid sequence in the MAD homology 1 (MH1) domain of
Smad2 is responsible for its inability to bind DNA [16,17].
However, Smad3 can directly bind to a specific DNA
sequence termed the Smad binding element (SBE). These
distinct properties account for activation of different subsets
of target genes by either Smad2 or Smad3.

Various proteins have been identified that negatively
influence TGF-b signaling at different levels [18]. One of
these proteins, the inhibitory Smad7, mediates its antagon-
istic effects by stable interaction with TbRI, thus preventing
the transient contact of R-Smads with the receptor and
blocking the proceeding cascade [19,20]. In addition, Smad7
was shown to recruit the E3 ubiquitin ligases Smurf1 and
Smurf2 to TbRI, thereby triggering degradation of the
TGF-b receptor complex [21,22]. Interestingly, expression
of Smad7 is rapidly induced in response to TGF-b1[23]and
therefore plays a crucial role in regulating TGF-b signaling.
Correspondence to P. Knaus, Physiological Chemistry II, Biocenter,
University of Wu
¨
rzburg, 97074 Wu
¨
rzburg, Germany.
Fax: + 49 931 888 4113, Tel.: + 49 931 888 4127,
E-mail:
Abbreviations:TGF-b, transforming growth factor-b;TbR, TGF-b
receptor type; NGF, nerve growth factor; R-Smads, receptor-associ-
ated Smads; MH, Mad homology domain; SBE, Smad binding
element; TrkA, tyrosine kinase receptor; ECD, extracellular domain;
MAPK, mitogen activated protein kinase; RSK, receptor serine/
threonine kinase; BMP, bone morphogenetic protein; GFP, green
fluorescent protein.
(Received 5 November 2003, accepted 15 January 2004)
Eur. J. Biochem. 271, 920–931 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.03994.x
In contrast to TGF-b signaling through receptor serine/
threonine kinases, nerve growth factor (NGF) signals via a

tyrosine kinase receptor (TrkA) which initiates multiple
pathways, the most prominent being the Ras/Raf-MAPK
pathway [24,25]. Besides TrkA, NGF can also bind to a
low-affinity neurotrophin receptor, p75
NTR
,whichisa
member of the tumor necrosis factor cytokine receptor
family [26].
It is increasingly evident that signal transduction in
general does not only occur in a linear fashion but rather
comprises a complex network of signaling pathways that
mutually influence their activity [27]. TGF-b family
members are implicated in multiple interdependent signals
between pathways originating from receptor serine/threo-
nine kinases and receptor tyrosine kinases. Cellular
responses induced by bone morphogenetic protein
(BMP) for example, can be impaired by epidermal growth
factor and hepatocyte growth factor, which lead to the
phosphorylation of Smad1 in the linker region and thus
prevent Smad1 nuclear translocation [28]. Direct effects
of signaling intermediates were shown for the c-Jun
N-terminal kinase and protein kinase C. c-Jun N-terminal
kinase phosphorylates Smad3 outside the SSXS-motif,
thus supporting nuclear transport of Smad3 [29]. Protein
kinase C however, abrogates direct DNA binding of
Smad3 by serine phosphorylation in the MH1 domain
[30]. This indicates that Smads are not restricted to TGF-
b/BMP pathways; rather, they represent a point of
convergence of various signals and their activation is a
precise contextually regulated process.

Here we demonstrate that in rat pheochromocytoma
cells (PC12), NGF stimulation results in activation of the
Smad cascade. This Smad activation is independent of
TGF-b1 and occurs by a mechanism which is different
from that induced by TGF-b1 in that it does not lead to
C-terminal phosphorylation of R-Smads. However, NGF
rapidly triggers association between Smad3 and Smad4,
translocation to the nucleus and gene expression. NGF-
mediated transcriptional activation of TGF-b responsive
reporter constructs requires the presence of functional
TrkA receptors and can be impaired by the inhibitory
Smad7.
Materials and methods
Antibodies and reagents
The monoclonal antibody against TGF-b1, -b2and-b3
(clone #1D11) was purchased from R&D Systems. Details
of polyclonal antisera against Smad2 (anti-S2), Smad3
(anti-S3) and C-terminally phosphorylated forms of
Smad1 (anti-PS1) and Smad2 (anti-PS2) were published
previously [31,32]. The P-Smad1 antibody shows cross-
reactivity to phosphorylated Smad3 and can thus be used
as an Ôanti-P-Smad3Õ (anti-PS3) [33]. The monoclonal
antibody Smad2/3 was purchased from BD Biosciences
and the monoclonal antibody Smad4 (B-8) was obtained
from Santa Cruz. Peroxidase-coupled goat anti-(rabbit
IgG) Ig was obtained from Dianova. Doxycycline was
purchased from Sigma. Human TGF-b1wasfromR&D
Systems and mouse NGF (2.5S) from Alomone labs
(Jerusalem, Israel).
Coating of culture dishes with rat tail collagen

Collagen was isolated from rat tails according to standard
protocols.
1
Appropriate measures were taken to minimize
animal pain and discomfort according to the European
Communities Council Directive of 24 November 1986
(86/609/EEC). Culture dishes were incubated with a
solution of 38 lgÆmL
)1
collagen in 0.1% (v/v) acetic acid
for at least 1 h followed by thorough washing with sterile
dH
2
O and medium without supplements.
Neutralization of TGF-b
All TGF-b isoforms were neutralized by the addition of
either monoclonal antibodies against TGF-b1, -b2, -b3
(20 lgÆmL
)1
) or a 100- or 1000-fold molar excess of the
soluble extracellular domain (ECD) of TbRII-B (TbRII-B-
ECD), kindly provided by J. Nickel (Biocenter, University
of Wu
¨
rzburg, Germany).
DNA constructs
Smad2 and Smad7 constructs were published previously
[20,34]. The Smad3 construct was kindly provided by
R. Derynck (University of California at San Francisco, CA,
USA). Smad1, Smad4 and Smad4 (DSAD) constructs were

a gift from M. de Caestecker (National Cancer Institute,
Bethesda, MD, USA) [35]. The NGF receptor constructs
(TrkA and TrkA-K538R) were obtained from M. Chao
(Skirball Institute of Biomolecular Medicine, New York,
NY, USA).
Cell culture and transient transfection
PC12 cells [36] were cultured in RPMI supplemented with
10% (v/v) horse serum, 5% (v/v) fetal bovine serum and
antibiotics
2
(penicillin, 100 UÆmL
)1
and streptomycin,
100 lgÆmL
)1
). Transient transfection of PC12 cells was
performed using LipofectAMINE
TM
(Life Technologies
Inc.) according to the manufacturer’s protocol. 293T cells
were maintained in minimum Eagle’s medium (MEM)
containing 10% (v/v) fetal bovine serum and antibiotics.
Transfection of 293T cells was performed using calcium
phosphate–DNA coprecipitation [37]. L6 rat myoblasts
were cultured in DMEM with 10% (v/v) fetal bovine serum.
MLEC cells [38] were cultured in DMEM supplemented
with 10% (v/v) fetal bovine serum and 250 lgÆmL
)1
geneticin.
Retroviral constructs and transduction of PC12 cells

by retroviral transfer
The retroviral construct pMX-GFP-Smad3 was a kind
gift from Y. Henis (Tel Aviv University, Israel) [39].
The construct for N-terminally HA-tagged TbRII-wt
was subcloned into the retroviral vector pczCFG-EGIRT
(D. Lindemann, unpublished results)
3
downstream of a
tetracycline-inducible cytomegalovirus (CMV) minimal
promoter [40]. Cells were transduced by infection with
helper-free VSV-G pseudotyped retroviruses as described
previously [37]. Briefly, 293T cells were cotransfected with
the retroviral construct and plasmids for gag-pol and
Ó FEBS 2004 Crosstalk between NGF and TGF-b pathways (Eur. J. Biochem. 271) 921
VSV-G. Twenty-four hours post-transfection, cells were
treatedwith10m
M
sodium-butyrate for 10 h. Infection of
the target cells was performed 48 h and 72 h after transfec-
tion. Because the retroviral sequences contain the gfp gene,
infected cells could be selected by FACS sorting.
Luciferase reporter gene assays
For reporter gene assays, different TGF-b responsive
elements were used. The p3TP-luc(+) reporter is derived
from the p3TP-luc construct [7] but contains a modified luc
gene (pSP-luc(+), Promega). The pSBE reporter [41] serves
as a readout for TGF-b as well as BMP signaling whereas
p3TP-luc(+) and (CAGA)
12
-luc constructs respond speci-

fically to TGF-b mediated signals [42,43].
PC12 cells were plated on collagen-coated 6-well plates
prior to cotransfection with reporter constructs, pRL-TK
(Renilla luciferase, Promega) and with the indicated receptor
or Smad constructs. The total amount of DNA was kept
constant by the addition of empty vector (pcDNA3).
Twenty-four hours post-transfection, cells were starved in
medium containing 0.2% (v/v) horse serum for 4 h,
followed by stimulation with either 2 n
M
NGF or 200 p
M
TGF-b1 for an additional 24 h. Cell lysis and luciferase
measurement were performed according to manufacturer’s
instructions (Dual Luciferase Assay system, Promega) and
data were normalized using Renilla luminescence. The value
of unstimulated transfected cells was set to one and all other
values were calculated accordingly.
Smad-Western blotting
To investigate Smad phosphorylation, cells were seeded on
Petri dishes, starved for 4 h in medium containing 0.2%
(v/v) serum and stimulated with 2 n
M
NGF or 200 p
M
TGF-b1 for 30 min. Cells were lysed in TNE lysis-buffer
[20 m
M
Tris/HCl pH 7.4, 150 m
M

NaCl, 1% (v/v) Triton
X-100, 1 m
M
EDTA, 1 m
M
phenylmethanesulfonyl fluor-
ide] containing protease inhibitors (Complete
TM
, Roche)
and phosphatase inhibitors (50 m
M
NaF, 10 m
M
Na
4
P
2
O
7
and 1 m
M
Na
3
VO
4
). Equal amounts of cell lysates were
analyzed by immunoblotting using anti-PS2, anti-Smad2,
anti-PS3 or anti-Smad3. Immunoreactive proteins were
visualized by enhanced chemiluminescence
4

.
Co-immunoprecipitation of Smad proteins
PC12 cells (5 · 10
6
) were starved in medium containing
0.2% (v/v) horse serum for 4 h. Ligand stimulation was
carried out for the indicated periods of time using 2 n
M
NGF. Cells were lysed in a buffer containing NaCl/P
i
pH 7.4, 0.5% (v/v) Triton X-100, 1 m
M
EDTA, phospha-
tase inhibitors and protease inhibitors. Cell lysates were
incubated with monoclonal antibody Smad2/3 or Smad4
for 2 h, followed by incubation with protein-A sepharose
beads overnight at 4 °C. The beads were washed twice with
lysis buffer and twice with NaCl/P
i
before immunocom-
plexes were eluted by boiling in SDS sample buffer for
5 min. Following separation by SDS/PAGE and electro-
transfer to a nitrocellulose membrane, proteins were
immunoblotted with Smad4 or Smad2/3 antibodies as
appropriate.
Nuclear and cytoplasmic fractionation
PC12 cells and L6 cells were starved in low serum medium
for 4 h. Ligand stimulation was performed for the indicated
periods of time with 2 n
M

NGF or 200 p
M
TGF-b1,
respectively. Control cells were stimulated with 2 n
M
NGF
for 1 h following a 1 h treatment with cycloheximide
(5 lgÆmL
)1
in dimethylsulfoxide) or dimethylsulfoxide only.
Cells were washed with NaCl/P
i
, centrifuged (1000 g,4°C,
10min) and the cell pellet was then resuspended in hypotonic
buffer (10 m
M
Hepes pH 7.9, 1.5 m
M
MgCl
2
,10m
M
KCl,
protease inhibitors). Cells were vortexed thoroughly and cell
lysis was followed by microscopy until 90% of the cells were
lysed. Following centrifugation (1000 g,4°C, 10 min), the
supernatant was referred to as the cytoplasmic fraction. The
pellet containing the nuclei was resuspended in high salt
buffer [20 m
M

Hepes pH 7.9, 25% (v/v) glycerol, 420 m
M
NaCl
2
,1.5m
M
MgCl
2
,0.2m
M
EDTA, protease inhibitors].
Extraction of nuclear proteins was achieved by vortexing
this solution thoroughly, incubating for 30 min on ice and
subsequent centrifugation (25 000 g,4°C, 20 min). The
supernatant was collected and represents the nuclear
fraction.
Nuclear translocation
PC12 cells were plated on collagen-coated dishes. Starvation
in a low serum medium for 4 h was followed by stimulation
with 2 n
M
NGF for the indicated times. Control cells were
stimulated with 2 n
M
NGF for 30 min following a 1 h
treatment with cycloheximide (5 lgÆmL
)1
in dimethylsulf-
oxide) or dimethylsulfoxide only. The cells were then
washed with NaCl/P

i
and 3% (v/v) BSA, fixed in 4% (v/v)
paraformaldehyde and 0.2% (v/v) TX-100 for 10 min at
room temperature. After washing with NaCl/P
i
containing
3% (v/v) BSA, Smad2/3 staining was performed with
antibodies from BD Biosciences. Nuclei were stained by
the addition of 1 lgÆmL
)1
Hoechst 33342 for 2 min. The
subcellular distribution of Smad2/3 was then analyzed by
confocal microscopy.
Results
NGF mediates the activation of Smad-dependent
reporter genes independently of TGF-b
Survival of neuronal cells is described to be synergistically
promoted by TGF-bs and neurotrophic factors (e.g. NGF)
[44,45]. To examine whether NGF has the potential to
modulate the Smad pathway in PC12 cells, we performed
reporter gene assays using luciferase constructs containing
promoter elements that are responsive to proteins of the
TGF-b superfamily, i.e. pSBE-luc, p3TP-luc(+) and
(CAGA)
12
-luc [7,41,42]. As indicated in Fig. 1A, PC12 cells
show a significant increase of transcriptional activation after
stimulation with NGF on all tested Smad-dependent
reporter constructs. In contrast, TGF-b1 is not able to
induce transcription from these reporters in PC12 cells.

Different approaches were chosen to exclude that this NGF-
mediated transcriptional response is a secondary effect,
caused for instance by NGF-triggered secretion of TGF-b1
as published previously [46]. First, the reporter gene assay
922 M. Lutz et al.(Eur. J. Biochem. 271) Ó FEBS 2004
was carried out in the presence of monoclonal antibodies
against TGF-b1, -b2and-b3 to neutralize all TGF-b
isoforms (Fig. 1B). Stimulation with NGF leads to a
significant increase in transcriptional activity which is not
impaired by the presence of neutralizing TGF-b antibodies.
TGF-b, however, does not induce luciferase activity, neither
Fig. 1. NGF mediates transcription from
TGF-b responsive reporter genes. (A) PC12
cells were plated on collagen-coated dishes and
cotransfected with pRL-TK and pSBE-luc,
p3TP-luc(+) or (CAGA)
12
-luc. Following
starvation for 4 h in medium containing 0.2%
(v/v) horse serum, cells were stimulated for
24 h with either 2 n
M
(50 ngÆmL
)1
)NGF
(black bars) or 200 p
M
TGF-b1(graybars),or
were left untreated (white bars). Cell lysates
were prepared and the luciferase activity was

measured. Data were normalized as described
in Materials and methods, and error bars
represent the SD evaluated from three inde-
pendent experiments. (B) Transfection and
starvation of PC12 cells was carried out as
described in (A) using pSBE-luc as the repor-
ter construct. Subsequently, cells were treated
with 2 n
M
NGF (black bars) or 200 p
M
TGF-
b1 (gray bars) either in presence or in absence
of 20 lgÆmL
)1
TGF-b1, -b2, -b3antibody.
After 24 h, luciferase activity was recorded
and the data were evaluated as described
above. (C) L6 cells stably expressing
GFP-Smad3 were starved for 4 h followed by
treatment with either 20 lgÆmL
)1
anti-
(TGF-b1, -b2, -b3) (lane 2), 200 p
M
TGF-b1
(lane 3) or 200 p
M
TGF-b1 together with
20 lgÆmL

)1
anti-(TGF-b1, -b2, -b3) (lane 4).
Cell lysates were analysed for C-terminally
phosphorylated Smad3 (upper panel) or for
total amounts of Smad3 (lower panel). (D)
PC12 cells were cotransfected with pSBE and
pRL-TK. The day after transfection, PC12
cells were starved as described previously and
stimulated with either 2 n
M
NGF (black bars),
200 p
M
TGF-b1 (gray bars) or were left
untreated (white bars). In addition, the
experiment was carried out in the absence or
presence of the soluble extracellular domain of
TbRII-B (TbRII-B-ECD). TbRII-B-ECD
was used at 100-fold and 1000-fold molar
excess as indicated. Twenty-four hours after
stimulation, the supernatant of the PC12 cells
was transferred to L6 cells (lower panel) which
were equivalently transfected and starved
prior to addition of the supernatant. PC12
cells (upper panel) as well as L6 cells (lower
panel) were lysed and tested for luciferase
activity. Error bars represent the SD from two
independent experiments.
Ó FEBS 2004 Crosstalk between NGF and TGF-b pathways (Eur. J. Biochem. 271) 923
in the presence nor in the absence of antibodies. The

neutralizing capacity of the TGF-b1, -b2and-b3 antibodies
was verified in L6 cells stably transduced with a GFP–
Smad3 construct (Fig. 1C). In the absence of TGF-b1, -b2
and -b3 antibodies, stimulation with TGF-b1leadsto
C-terminal phosphorylation of Smad3, whereas treatment
with TGF-b1 together with neutralizing antibodies impedes
the phosphorylation of Smad3. Second, we performed a
complementary experiment using the soluble ECD of
TbRII-B, a TGF-b type II receptor splice variant that
binds all three TGF-b isoforms [9]. TbRII-B-ECD was
added in two different concentrations (100- and 1000-fold
molar excess) to PC12 cells transfected with the pSBE-luc
reporter in the presence of NGF or TGF-b1(Fig.1D,
upper panel). The TbRII-B-ECD did not abolish NGF-
mediated Smad activation in PC12 cells. Next, we harvested
the supernatant of PC12 cells treated in this way and placed
it on the TGF-b sensitive L6 myoblast cell line that was
equally transfected with pSBE-luc (Fig. 1D, lower panel).
As there was no detectable increase in luciferase activity in
TGF-b-sensitive L6 cells that were treated with the super-
natant from NGF stimulated PC12 cells (Fig. 1D, lower
panel, bar 2), we can conclude that NGF treatment of PC12
cells does not lead to the production of active TGF-b.In
contrast, the TGF-b-treated PC12 cell supernatant results in
reporter activation in L6 cells (Fig. 1D, lower panel, bar 3).
This activation is almost completely blocked by a 1000-fold
molar excess of TbRII-B-ECD (Fig. 1D, lower panel, bar 9).
Furthermore, we measured the amount of TGF-b that
is produced in response to NGF and checked whether this
TGF-b is present in an active or latent form. The

quantification of TGF-b was performed by using the
MLEC cell line, which stably expresses a luciferase
reporter gene under the control of a truncated PAI-1
promoter [38]. We found that NGF leads to production
of only marginal amounts ( 1.1 p
M
)ofactiveTGF-b
(data not shown). This is in accordance with the results
presented above (Fig. 1D, lower panel). Taken together,
all approaches clearly demonstrate that NGF mediates the
activation of the Smad pathway independently of the
TGF-b ligand.
TGF-b resistance is due to limiting amounts of TbRII
in PC12 cells
The expression of TGF-b receptors, particularly TbRII, in
PC12 cells is controversially discussed in the literature
[47,48]. Phosphorylation studies and reporter gene assays
demonstrate that TbRII represents the limiting component
of TGF-b1 signaling in PC12 cells. Smad2 phosphorylation
wasanalysedinparentalPC12cellsandinstablePC12cells
expressing TbRII-wt under the control of a doxycycline-
inducible promoter. Figure 2A shows that treatment with
TGF-b1 results in C-terminal phosphorylation only in
PC12 cells that ectopically express TbRII but not in
parental PC12 cells. However, in response to NGF, there
is no phosphorylation of Smad2 at the SSXS-motif.
Luciferase assays also show that transient transfection of
PC12 cells with TbRII but not TbRI constructs leads to
increased responsiveness to TGF-b1 (Fig. 2B), again indi-
cating that TbRII is the limiting component of TGF-b1

signaling in PC12 cells.
Mechanism of Smad activation by NGF
The mechanism of Smad reporter activation was investigated
by using luciferase constructs that allowed us to distinguish
between signals originating from different R-Smads, i.e.
pSBE-luc, p3TP(+)-luc and (CAGA)
12
-luc [7,41,42] (see
Materials and methods). Using p3TP-luc(+), NGF-induced
reporter activation was investigated after ectopic expression
of various R-Smad constructs (Fig. 3A). From all R-Smads
tested, Smad3 shows the most prominent induction of
transcription after NGF stimulation. Similar results were
obtained with the (CAGA)
12
-luc reporter (data not shown).
Given that in TGF-b signaling, phosphorylation of the
C-terminal serine residues (SSXS) is essential for dissoci-
ation from the type I receptor and for heteromeric complex
formation with Smad4 [34,49], we investigated C-terminal
phosphorylation in response to TGF-b1aswellasNGF.
The phosphorylation pattern of Smad3 resulting from
stimulation with either TGF-b or NGF was analysed in
Fig. 2. PC12 cells express low levels of endogenous TbRII. (A)
C-terminal phosphorylation of Smad2 was investigated in either par-
ental PC12 cells (lanes 1–3) or PC12 cells stably expressing TbRII-wt
(lanes 4–6). Cells were kept in media containing 1 lgÆmL
)1
doxycycline
for 3 days to induce expression of TbRII-wt. Following starvation,

cells were treated either with 2 n
M
NGF(lanes2and5)orwith200p
M
TGF-b1 (lanes 3 and 6) or were left untreated (lanes 1 and 4). Cell
lysates were examined for Smad2 phosphorylation by immunoblotting
using an antibody raised against the phosphorylated SSXS-motif of
Smad2 (anti-PS2). (B) PC12 cells were plated on collagen-coated dishes
and were cotransfected with pRL-TK, pSBE and the indicated
expression constructs. Luciferase activity was measured after starva-
tion and treatment for 24 h with control medium (white bars) or with
medium containing 200 p
M
TGF-b1 (black bars).
924 M. Lutz et al.(Eur. J. Biochem. 271) Ó FEBS 2004
TGF-b-responsive L6 rat myoblasts and in PC12 cells,
which were both stably transduced with a retroviral GFP–
Smad3 construct [39]. Immunoblotting using an antibody
that specifically recognizes C-terminally phosphorylated
Smad3 [32,33] revealed that phosphorylation of both the
heterologous GFP–Smad3 and the endogenous Smad3
protein occurs in L6 cells after treatment with TGF-b1, but
not with NGF (Fig. 3B). In PC12 cells, however, we did not
detect any phosphorylation at the SSXS-motif of Smad3;
neither in response to NGF nor in response to TGF-b1. The
low amounts of TbRII expressed in PC12 cells can account
for the lack of phosphorylation upon TGF-b stimulation.
This suggests that NGF-mediated Smad activation occurs
independently from
5

phosphorylation at the C-terminal
SSXS-motif.
Involvement of Smad4 in NGF-triggered activation
of the Smad signaling cascade
In TGF-b1-induced signaling, R-Smads form heteromeric
complexes with Smad4 following the activation by TbRI.
Reporter gene assays using the p3TP(+)-luc construct were
performed to determine whether NGF-mediated activation
of Smad response elements is also Smad4-dependent. PC12
cells were transfected with Smad3, Smad4 or a functionally
inactive Smad4 variant – Smad4(DSAD) – either alone or
in the indicated combinations. Smad4(DSAD) lacks amino
acids 274–321 which encode the Smad activation domain
(SAD) [35,50]. Figure 4A demonstrates that ectopic expres-
sion of Smad3 results in efficient transcriptional activation
of Smad-dependent reporter genes, whereas neither Smad4
nor the mutant Smad4 show an effect on luciferase
induction when expressed alone. Coexpression of Smad3
and Smad4, however, enhances the Smad3 effect. In
Fig. 3. Mode of Smad activation by NGF. (A) Induction of specific
R-Smads was investigated using the p3TP-luc(+) reporter. PC12 cells
were transiently transfected with the indicated Smad constructs. Fol-
lowing starvation, cells were left untreated (white bars) or were stimu-
lated with 2 n
M
NGF (black bars). Data were normalized as described
inMaterialsandmethods.Errorbarswerecalculatedfromthree
independent measurements. (B) Phosphorylation of Smad3 was tested
in TGF-b responsive L6 rat myoblasts (lanes 1–4) and PC12 cells (lanes
5–8) that were transduced with pMX-GFP-Smad3 (lanes 2–4 and 6–8)

using retroviral transfer. Cells were starved for 4 h, followed by sti-
mulation with 2 n
M
NGF (lanes 3 and 7) or 200 p
M
TGF-b1(lanes4
and 8) and cell lysates were analysed for C-terminally phosphorylated
Smad3 (upper panel) or for total amounts of Smad3 (lower panel).
Fig. 4. Role of Smad4 in NGF-triggered activation of the Smad path-
way. (A) The role of functional Smad4 was investigated by reporter
gene assays using p3TP-luc(+). PC12 cells were cotransfected with the
luciferase constructs and with different combinations of Smad3 and
Smad4 variants as indicated. Following starvation and treatment for
24 h with either 2 n
M
NGF (black bars) or control medium (white
bars), cell lysates were prepared and used for luminescence measure-
ment. (B) NGF-induced association of Smad3 and Smad4 was
assessed by coimmunoprecipitation studies. PC12 cells were starved
and treated with 2 n
M
NGF for the indicated periods of time. Smad3
was immunoprecipitated from cell lysates using the anti-Smad2/3 Ig
and analyzed by Western blotting using anti-Smad4 Igs. Additionally,
the experiment was repeated with immunoprecipitation of Smad4
followed by detection of Smad3 by Western blotting (upper panels).
Total amounts of protein were verified by immunoblotting proteins of
total lysates using the appropriate antibodies (lower panels).
Ó FEBS 2004 Crosstalk between NGF and TGF-b pathways (Eur. J. Biochem. 271) 925
contrast, coexpression of the functionally inactive Smad4

variant – Smad4(DSAD) – largely prevents Smad3-
mediated reporter gene activation. These results were also
confirmed by using the (CAGA)
12
-luc reporter (data not
shown).
Furthermore, co-immunoprecipitation studies confirmed
that NGF stimulation of PC12 cells leads to interaction
between Smad3 and Smad4 (Fig. 4B). Whereas in the
absence of ligand there is no heteromeric complex forma-
tion, NGF treatment triggers association of the Smad
proteins within 30 min, indicating that NGF directly
activates Smad signaling.
NGF stimulation rapidly initiates nuclear accumulation
of Smad3
To assess whether Smad3 translocates to the nucleus in
response to NGF treatment, nuclear extracts were investi-
gated for the content of Smad3 protein and the cellular
distribution of Smad3
6
was determined in whole cells.
Nuclear extracts were prepared from L6 rat myoblasts that
were stimulated with TGF-b1 (Fig. 5, upper panel) and
from PC12 cells at several time points after NGF treatment
(Fig. 5
7
, middle panel). In both cell lines, Smad3 can be
detected in the nuclear fraction after 5 min of ligand
stimulation and the amount of nuclear Smad3 increases
with prolonged growth factor treatment, reaching a maxi-

mum after 15–30 min of stimulation with TGF-b1inL6
cells and after 1 h of stimulation with NGF in PC12 cells.
Referring to Smad4, there are significant levels of protein
in the nucleus of both cell lines already in the absence of
ligand. Cycloheximide treatment for 1 h prior to the
addition of NGF indicated that Smad3 was directly
stimulated by NGF for nuclear translocation, with no
de novo protein synthesis required (Fig. 5, lower panel).
Next, we investigated the nuclear transport of Smad3 in
PC12 cells by confocal microscopy. Staining with Hoe-
chst 33342 was performed to visualize the nuclei. Without
NGF treatment, Smad3 is distributed throughout the whole
cell (Fig. 6, first row). Stimulation with NGF for 30 min
shows a strong decrease of cytoplasmic Smad3 staining and
accumulation of Smad3 in the nucleus, and NGF treatment
for 3 h results in a solely nuclear localization of Smad3
(Fig. 6, second and third rows, respectively). Cycloheximide
treatment indicated that Smad3 was immediately stimulated
by NGF for nuclear translocation; this process does not
require de novo protein synthesis (Fig. 6, rows 4 and 5). This
agrees with the data that we obtained by cellular fraction-
ation of parental PC12 cells (Fig. 5).
NGF-mediated activation of Smad reporter constructs
can be efficiently abrogated by either kinase-dead TrkA
receptors or by the inhibitory Smad7 protein
To assess the involvement of TrkA receptors in the
activation of the Smad pathway, different TrkA variants
were tested in reporter gene assays using the pSBE-luc
construct (Fig. 7). In PC12 cells transfected with wild-type
TrkA (TrkA-wt), the basal level of luciferase activity is

elevated already but the signal can be potently enhanced by
stimulation with NGF. Transfection of the TrkA variant
(TrkA–K538A) that carries a mutation resulting in the
inactivation of the tyrosine kinase activity causes a signifi-
cant reduction of responsiveness. An even stronger inhi-
bitory effect can be observed after cotransfection of the
wild-type TrkA receptor together with Smad7. The antago-
nizing impact of Smad7 becomes additionally apparent by
the strong inhibitory effect on endogenous signaling that is
elicited following expression of ectopic Smad7 (Fig. 7, lanes
2 and 4). These results suggest functional TrkA receptors
to be necessary for NGF-mediated activation of Smad-
dependent reporter genes and demonstrate the inhibitory
role of Smad7 on this NGF-mediated effect.
Discussion
Originally, Smad proteins were exclusively attributed to
pathways activated by TGF-b family members but it
becomes increasingly evident that multiple signaling cas-
cades originating from other receptor systems are involved
in modulating Smad signaling [14,27,30,51,52]. In the
present work, we demonstrate that in PC12 cells NGF-
stimulated signaling via the TrkA receptor leads to activa-
tion of the Smad pathway. NGF-mediated Smad activation
is independent of TGF-b ligand and occurs by a mechanism
which is different from that induced by TGF-b.
PC12 rat pheochromocytoma cells represent a widely
used model system to investigate neuronal differentiation
that is initiated following stimulation with NGF [36].
Fig. 5.
10

NGF induces nuclear accumulation of Smad3. L6 cells and PC12
cells were starved for 4 h in medium containing 0.2% (v/v) fetal bovine
serum or horse serum, respectively, and nuclear fractions were pre-
pared at various time points after exposure to either 200 p
M
TGF-b1
or 2 n
M
NGF as indicated. Proteins contained in the nuclear fraction
were subjected to SDS/PAGE, electrotransferred to nitrocellulose and
immunoblotted with monoclonal antibodies to Smad2/3 or Smad4.
The purity of the cytoplasmic and nuclear fractions was confirmed
by immunoblotting with an anti-lamin serum. Control cells were
stimulated with 2 n
M
NGF for 1 h following a 1 h treatment with
cycloheximide (lower panel). Nuclear fractions were probed with anti-
Smad2/3 and subsequently with anti-lamin.
926 M. Lutz et al.(Eur. J. Biochem. 271) Ó FEBS 2004
Although TGF-bs do not promote survival or differenti-
ation of neuronal populations on their own, they elicit a
neurotrophic potential if they are applied together with
other cytokines (GDNF, GDF-5) or neurotrophins (NGF,
NT-3) [4,5], suggesting that the signaling cascades of
TGF-bs and neurotrophins are somehow interdependent.
However, ectopic expression of inhibitors of the TGF-b/
Smad pathway such as Smad7 or neutralizing TGF-b
antibodies did not prevent NGF-induced neurite formation
(data not shown), suggesting that the Smad pathway that
Fig. 6. Smad3 nuclear translocation in PC12 cells using confocal microscopy. Cells were plated on collagen-coated dishes, starved in medium

containing 0.2% (v/v) horse serum for 4 h and stimulated for with 2 n
M
NGF for 30 min (second row), 3 h (third row) or were left untreated (first
row). Control cells were stimulated with 2 n
M
NGFfor30minfollowinga1htreatmentwithcycloheximide(CHX;fourthrow)ordimethyl-
sulfoxide (DMSO; fifth row). Cells were fixed, nuclei were stained with Hoechst 33342 for 2 min and the cells were analysed by confocal
microscopy. The projection of multiple sections is seen on the left for each panel to visualize the morphology of the cells. The middle row shows
staining of Smad3 and on the right an overlay of Smad3 and Hoechst staining is seen.
Ó FEBS 2004 Crosstalk between NGF and TGF-b pathways (Eur. J. Biochem. 271) 927
can be activated by NGF is mainly important for other
cellular responses.
PC12 cells show transcriptional activation of TGF-
b-responsive reporter genes upon NGF but not TGF-b1
stimulation (Fig. 1). Low amounts of TbRII expressed in
these cells can account for TbRII being the limiting factor
for proper TGF-b1 signaling, which is in accordance with
results showing that ectopic expression of TbRII restores
TGF-b responsiveness (Fig. 2). As an earlier report shows
upregulation of TGF-b by NGF [46], we investigated
whether NGF-triggered Smad activation is caused by
autocrine action of TGF-b. Considering the limiting
amount of TbRII discussed above, PC12 cells are equally
resistant to signals evoked by either exogenous or autocrine
TGF-b. Furthermore, even if all TGF-b isoforms are
neutralized by the addition of antibodies or the soluble
extracellular domain of TbRII-B (TbRII-B-ECD), NGF is
still capable of activating Smad-dependent reporter genes
(Fig. 1B,D). Concerning the amounts of secreted TGF-b,we
found that besides latent (i.e. biologically inactive) TGF-b,

only marginal amounts of active TGF-b can be detected
in the supernatant of NGF-treated PC12 cells. TGF-b is
synthesized as a precursor proprotein that is cleaved during
secretion. However, the mature TGF-b remains associated
with its propeptide thereby forming a latent complex until
activation [53]. Because TGF-b activation takes place in the
extracellular compartment, intracellular signaling events
initiated by autocrine TGF-b can be excluded. Taken
together, the effects of NGF on the Smad pathway are
independent of TGF-b ligand.
To characterize the point of convergence and the mode of
Smad activation, C-terminal phosphorylation, heteromeric
complex formation with Smad4 and nuclear translocation
of R-Smads was investigated. Comparison of signaling
through different R-Smad proteins revealed that Smad3
mediates the most potent activation (Fig. 3A). As Smad2
contains an additional exon in the MH1 domain that is not
present in Smad3, it lacks the capacity to bind directly to
DNA [16,17], which might explain the different behavior of
the two TGF-b-activated Smads in response to NGF.
Although treatment of PC12 cells with NGF results in
the activation of Smad-dependent reporter genes, the
preceeding signaling events are not identical to those that
are known from TGF-b signaling as NGF does not induce
phosphorylation of the C-terminal SSXS-motif (Fig. 3B).
Recent publications describe alternative mechanisms of
Smad activation which are likewise independent of
C-terminal phosphorylation: c-Jun N-terminal kinase was
showntoberapidlyactivatedbyTGF-b stimulation in a
Smad-independent manner and to cause initial phosphory-

lation of Smad3 at sites other than the SSXS motif. This
modification in turn promotes TbRI-dependent C-terminal
phosphorylation of Smad3 [29]. The mitogen-activated
protein kinase kinase kinase was shown to trigger
phosphorylation outside the C-terminal motif, which
results in enhanced transcriptional activity of Smad2 in
endothelial cells [54]. These examples support our findings
that activation of Smad proteins can occur independently
of C-terminal phosphorylation. Besides direct phosphory-
lation events, NGF potentially triggers other modifications
of R-Smads resulting in Smad nuclear translocation and
transcriptional activation.
As the NGF-initiated processes were shown to be
dependent on functional Smad4 proteins (Fig. 4A) and to
lead to heteromeric complex formation between Smad3
and Smad4 (Fig. 4B), we assume that the presence of the
SSXS-motif is crucial to allow interaction between
R-Smads and Smad4, even if the C-terminal serines are
not phosphorylated. Recent reports show that phosphory-
lation of the SSXS-motif enhances heteromeric complex
formation and stabilizes the assembly of the Smad homo-
and hetero-oligomers. Nevertheless, Smad3 and Smad4
were shown to heterotrimerize in the absence of phos-
phorylation [55,56].
Thus it remains to be elucidated whether phosphorylation
of other residues or different modifications causes the same
or even a distinct oligomerization pattern of Smads.
Nuclear translocation of Smad3 could be confirmed by
the appearance of the Smad3 protein in nuclear extracts
following NGF stimulation (Fig. 5) and by investigation of

the cellular distribution of Smad3 by confocal microscopy
(Fig. 6). The observation, that Smad3 appears in the
nuclear fraction of PC12 cells after only 5 min of NGF
stimulation and reaches a maximum after 1 h hints of a
direct effect of NGF on Smad proteins. This is also
confirmed by cycloheximide treatment (Figs 5 and 6),
demonstrating that no de novo protein synthesis is required
for NGF-mediated nuclear translocation of Smad3.
Whereas Smad3 is not present in the nucleus in the absence
of NGF, Smad4 can be found in the nucleus regardless of
ligand stimulation. This is in accordance with the findings
that Smad4 continuously shuttles between the cytoplasm
and the nucleus [57]. R-Smads, however, underlie cytoplas-
mic retention in the absence of ligand due to their
interaction with Smad anchor for receptor activation
8
[58,59] or microtubules [60]. Confocal microscopy studies
additionally confirm the NGF-induced nuclear transloca-
tion of Smad3 within 30 min (Fig. 6). While untreated cells
reveal Smad3 staining throughout the whole cell, stimula-
tion with NGF for 30–60 min provokes complete nuclear
accumulation of Smad3.
To define the role of the high-affinity NGF receptor,
TrkA, we ectopically expressed functionally inactive NGF
receptors in PC12 cells and found that the tyrosine kinase
Fig. 7. Functional TrkA receptors are essential for the activation of
Smad-dependent reporters by NGFPC12. Cells were transfected with
pSBE-luc and the indicated DNA constructs. Total amounts of DNA
were kept constant by the addition of empty vector (pcDNA3). The
experiment was carried out as described in Fig. 1 and error bars are

calculated from three independent measurements.
928 M. Lutz et al.(Eur. J. Biochem. 271) Ó FEBS 2004
function of the TrkA receptor is required for the activation
of Smad-dependent reporter constructs by NGF (Fig. 7).
Interestingly, expression of Smad7 results in an almost
complete loss of transcriptional activity, even when it is
coexpressed with functional TrkA receptors. This demon-
strates that Smad7 functions downstream of TrkA to block
Smad signaling. Different scenarios of Smad7-mediated
signal abrogation have been previously described. Smad7 is
capable of blocking Smad signaling at the receptor level by
interaction with activated TbRI [19] or by recruiting the
E3 ubiquitin ligases Smurf1 and Smurf2 to the receptors,
resulting in enhanced turnover of TGF-b receptors [21,22].
Furthermore, Smad7 was shown to interfere with signal
transduction by interaction with cytoplasmic proteins such
as TAB1 [61] or mitogen-activated protein kinase kinase
kinase [54]. These distinct antagonizing mechanisms of
Smad7 open up the question whether Smad7 blocks NGF-
induced Smad signaling at the receptor level or by
interaction with other proteins. As dominant-negative
TGF-b receptor mutants did not block NGF-induced
Smad activation (data not shown), they seem to be
dispensible for NGF-mediated signals, and therefore a
mechanism that involves Smad7 interaction with cytoplas-
micproteinsisfavored.
The Alk7 type I receptor is highly similar in its
intracellular domain to TbRI and the constitutively active
form of Alk7 was shown also to induce Smad2/3 phos-
phorylation. Studies in PC12 cells have indicated that Alk7

signaling augments differentiation response to NGF [48].
The ligand for this receptor, however, is presently unknown.
In conclusion, we describe here that NGF stimulation of
PC12 cells results in activation of the Smad pathway
independently of TGF-b1. This activation is direct and
results in nuclear translocation of Smad3 within only
30 min of NGF treatment. Binding of NGF to its high-
affinity receptor TrkA induces activation of Smad3,
heteromeric complex formation with Smad4, nuclear trans-
location and transcriptional activation. However, unlike
TGF-b1 signaling, this process does not include phosphory-
lation of the C-terminal SSXS-motif of the R-Smad. Based
on the diverse mechanism of Smad activation by either
TGF-b1 or NGF, specific subsets of target genes might be
induced.
The potential of NGF to activate the Smad pathway
independently of TGF-b might be of special importance in
regulating the expression of genes that are essential for
the development and function of neuronal cells or other
NGF-sensitive cells, in particular those which are TGF-b
resistant.
Acknowledgements
R. Derynck, M. de Caestecker and M. Chao are gratefully acknow-
ledged for expression vectors and D. Lindemann, Y. Henis and X. Liu
for retroviral vector constructs. We thank F. Neubauer for generating
the p3TP-luc(+) construct and J. Nickel for providing the TbRII-
B-ECD. We are grateful to J. Fey for preparation of collagen and to
Y. Kehl for excellent technical assistance. We also acknowledge
S.Hassel,R.Scha
¨

fer and M. Sammar for helpful discussions. This
work was supported by the Deutsche Forschungsgemeinschaft (DFG)
grant Kn332/3–2 to P. Knaus and EEC, Project 171R ERB-
FMRXCT980216 to P. ten Dijke. M. Lutz was supported by GK 181.
References
1. Zimmerman, C.M. & Padgett, R.W. (2000) Transforming growth
factor-b signaling mediators and modulators. Gene 249, 17–30.
2. Massague
´
, J. (1998) TGF-b signal transduction. Annu. Rev. Bio-
chem. 67, 753–791.
3. Piek, E., Heldin, C H. & ten Dijke, P. (1999) Specificity, diversity,
andregulationinTGF-b superfamily signaling. FASEB J. 13,
2105–2124.
4. Krieglstein, K., Henheik, P., Farkas, L., Jaszai, J., Galter, D.,
Krohn, K. & Unsicker, K. (1998) Glial cell line-derived neuro-
trophic factor requires transforming growth factor-b for exerting
its full neurotrophic potential on peripheral and CNS neurons.
J. Neurosci. 18, 9822–9834.
5. Bo
¨
ttner, M., Krieglstein, K. & Unsicker, K. (2000) The trans-
forming growth factor-bs: Structure, signaling, and roles in
nervous system development and functions. J. Neurochem. 75,
2227–2240.
6. ten Dijke, P., Miyazono, K. & Heldin, C H. (1996) Signaling via
hetero-oligomeric complexes of type I and type II serine/threonine
kinase receptors. Curr. Opin. Cell Biol. 8, 139–145.
7. Wrana, J.L., Attisano, L., Carcamo, J., Zentella, A., Doody, J.,
Laiho, M., Wang, X.F. & Massague

´
, J. (1992) TGF-b signals
through a heteromeric protein kinase receptor complex. Cell 71,
1003–1014.
8. Lin, H.Y., Wang, X.F., Ng-Eaton, E., Weinberg, R.A. & Lodish,
H.F. (1992) Expression cloning of the TGF-b type II receptor, a
functional transmembrane serine/threonine kinase. Cell 68, 775–785.
9. Rotzer, D., Roth, M., Lutz, M., Lindemann, D., Sebald, W. &
Knaus, P. (2001) Type III TGF-b receptor-independent signalling
of TGF-b2viaTbRII- B, an alternatively spliced TGF-b type II
receptor. EMBO J. 20, 480–490.
10. Wrana, J.L., Attisano, L., Wieser, R., Ventura, F. & Massague
´
,J.
(1994) Mechanism of activation of the TGF-b receptor. Nature
370, 341–347.
11. Heldin, C H., Miyazono, K. & ten Dijke, P. (1997) TGF-b sig-
nalling from cell membrane to nucleus through Smad proteins.
Nature 390, 465–471.
12. Massague
´
, J. & Chen, Y.G. (2000) Controlling TGF-b signaling.
Genes Dev. 14, 627–644.
13. Lagna, G., Hata, A., Hemmati-Brivanlou, A. & Massague
´
,J.
(1996)PartnershipbetweenDPC4andSmadproteinsinTGF-b
signalling pathways. Nature 383, 832–836.
14. Zhang, Y. & Derynck, R. (1999) Regulation of Smad signalling by
protein associations and signalling crosstalk. Trends Cell Biol. 9,

274–279.
15. Itoh, S., Itoh, F., Goumas, M J. & ten Dijke, P. (2000) Signaling
of transforming growth factor-b family members through Smad
proteins. Eur. J. Biochem. 267, 6954–6967.
16. Dennler, S., Huet, S. & Gauthier, J.M. (1999) A short amino-acid
sequence in MH1 domain is responsible for functional differences
between Smad2 and Smad3. Oncogene 18, 1643–1648.
17. Yagi,K.,Goto,D.,Hamamoto,T.,Takenoshita,S.,Kato,M.&
Miyazono, K. (1999) Alternatively spliced variant of Smad2
lacking exon 3. Comparison with wild-type Smad2 and Smad3.
J. Biol. Chem. 274, 703–709.
18. ten Dijke, P., Miyazono, K. & Heldin, C H. (2000) Signaling
inputs converge on nuclear effectors in TGF-b signaling. Trends
Biochem. Sci. 25, 64–70.
19. Hayashi,H.,Abdollah,S.,Qiu,Y.,Cai,J.,Xu,Y.Y.,Grinnell,
B.W., Richardson, M.A., Topper, J.N., Gimbrone, M.A. Jr,
Wrana, J.L. & Falb, D. (1997) The MAD-related protein Smad7
associates with the TGF-b receptor and functions as an antagonist
of TGF-b signaling. Cell 89, 1165–1173.
20. Nakao, A., Afrakhte, M., More
´
n, A., Nakayama, T., Christian,
J.L., Heuchel, R., Itoh, S., Kawabata, M., Heldin, N.E., Heldin,
Ó FEBS 2004 Crosstalk between NGF and TGF-b pathways (Eur. J. Biochem. 271) 929
C H. & ten Dijke, P. (1997) Identification of Smad7, a TGF-
b-inducible antagonist of TGF-b signalling. Nature 389, 631–635.
21. Ebisawa, T., Fukuchi, M., Murakami, T.C., Tanaka, K., Imam-
ura, T. & Miyazono, K. (2001) Smurf1 interacts with transforming
growth factor-b type I receptor through Smad7 and induces
receptor degradation. J. Biol. Chem. 276, 12477–12480.

22. Kavsak, P., Rasmussen, R.K., Causing, C.G., Bonni, S., Zhu, H.,
Thomsen, G.H. & Wrana, J.L. (2000) Smad7 binds to Smurf2 to
form an E3 ubiquitin ligase that targets the TGF-b receptor for
degradation. Mol. Cell 6, 1365–1375.
23. Afrakhte, M., More
´
n, A., Jossan, S., Itoh, S., Sampath, K.,
Westermark, B., Heldin, C H., Heldin, N.E. & ten Dijke, P.
(1998) Induction of inhibitory Smad6 and Smad7 mRNA by
TGF-b family members. Biochem. Biophys. Res. Commun. 249,
505–511.
24. Friedman, W.J. & Greene, L.A. (1999) Neurotrophin signaling via
Trks and p75. Exp. Cell Res. 253, 131–142.
25. Kaplan, D.R. & Miller, F.D. (2000) Neurotrophin signal trans-
duction in the nervous system. Curr. Opin. Neurobiol. 10, 381–391.
26. Dechant, G. & Barde, Y.A. (1997) Signalling through the neuro-
trophin receptor p75
NTR
. Curr. Opin. Neurobiol. 7, 413–418.
27. Massague
´
, J. (2000) How cells read TGF-b signals. Nat. Rev. Mol.
Cell. Biol. 1, 169–178.
28.Kretzschmar,M.,Doody,J.&Massague
´
, J. (1997) Opposing
BMP and EGF signalling pathways converge on the TGF-b
family mediator Smad1. Nature 389, 618–622.
29. Engel, M.E., McDonnell, M.A., Law, B.K. & Moses, H.L.
(1999) Interdependent Smad and JNK signaling in transforming

growth factor-b-mediated transcription. J. Biol. Chem. 274,
37413–37420.
30. Yakymovych, I., ten Dijke, P., Heldin, C H. & Souchelnytskyi, S.
(2001) Regulation of Smad signaling by protein kinase C. FASEB
J. 15, 553–555.
31. Nakao, A., Imamura, T., Souchelnytskyi, S., Kawabata, M.,
Ishisaki, A., Oeda, E., Tamaki, K., Hanai, J., Heldin, C H.,
Miyazono,K.&tenDijke,P.(1997)TGF-b receptor-mediated
signalling through Smad2, Smad3 and Smad4. EMBO J. 16,
5353–5362.
32. Ishisaki, A., Yamato, K., Hashimoto, S., Nakao, A., Tamaki, K.,
Nonaka, K., ten Dijke, P., Sugino, H. & Nishihara, T. (1999)
Differential inhibition of Smad6 and Smad7 on bone morpho-
genetic protein- and activin-mediated growth arrest and apoptosis
in B cells. J. Biol. Chem. 274, 13637–13642.
33. Dooley, S., Delvoux, B., Streckert, M., Bonzel, L., Stopa, M., ten
Dijke, P. & Gressner, A.M. (2001) Transforming growth factor-b
signal transduction in hepatic stellate cells via Smad2/3 phos-
phorylation, a pathway that is abrogated during in vitro pro-
gression to myofibroblasts. FEBS Lett. 502, 4–10.
34. Souchelnytskyi, S., Tamaki, K., Engstrom, U., Wernstedt, C., ten
Dijke, P. & Heldin, C H. (1997) Phosphorylation of Ser465 and
Ser467 in the C terminus of Smad2 mediates interaction with
Smad4 and is required for transforming growth factor-b signaling.
J. Biol. Chem. 272, 28107–28115.
35. de Caestecker, M.P., Hemmati, P., Larisch-Bloch, S., Ajmera, R.,
Roberts, A.B. & Lechleider, R.J. (1997) Characterization of
functional domains within Smad4/DPC4. J. Biol. Chem. 272,
13690–13696.
36. Greene, L.A. & Tischler, A. (1976) Establishment of a non-

adrenergic clonal line of rat adrenal pheochromocytoma cells
which respond to nerve growth factor. Proc. Natl Acad. Sci. USA
73, 2424–2428.
37. Pear, W.S., Nolan, G.P., Scott, M.L. & Baltimore, D. (1993)
Production of high-titer helper-free retroviruses by transient
transfection. Proc. Natl Acad. Sci. USA 90, 8392–8396.
38. Abe, M., Harpel, J.G., Metz, C.N., Nunes, I., Loskutoff, D.J. &
Rifkin, D.B. (1994) An assay for transforming growth factor-b
using cells transfected with a plasminogen activator inhibitor-1
promoter-luciferase construct. Anal. Biochem. 216, 276–284.
39. Liu, X., Sun, Y., Constantinescu, S.N., Karam, E., Weinberg,
R.A. & Lodish, H.F. (1997) Transforming growth factor-b-
induced phosphorylation of Smad3 is required for growth inhi-
bition and transcriptional induction in epithelial cells. Proc. Natl
Acad. Sci. USA 94, 10669–10674.
40. Kistner,A.,Gossen,M.,Zimmermann,F.,Jerecic,J.,Ullmer,C.,
Lubbert, H. & Bujard, H. (1996) Doxycycline-mediated quanti-
tative and tissue-specific control of gene expression in transgenic
mice. Proc.NatlAcad.Sci.USA93, 10933–10938.
41. Jonk, L.J., Itoh, S., Heldin, C H., ten Dijke, P. & Kruijer, W.
(1998) Identification and functional characterization of a Smad
binding element (SBE) in the JunB promoter that acts as a
transforming growth factor-b, activin, and bone morphogenetic
protein-inducible enhancer. J. Biol. Chem. 273, 21145–21152.
42. Dennler, S., Itoh, S., Vivien, D., ten Dijke, P., Huet, S. & Gau-
thier, J.M. (1998) Direct binding of Smad3 and Smad4 to critical
TGF-b-inducible elements in the promoter of human plasminogen
activator inhibitor-type 1 gene. EMBO J. 17, 3091–3100.
43. Yingling, J.M., Datto, M.B., Wong, C., Frederick, J.P., Liberati,
N.T. & Wang, X.F. (1997) Tumor suppressor Smad4 is a trans-

forming growth factor-b-inducible DNA binding protein. Mol.
Cell. Biol. 17, 7019–7028.
44. Krieglstein, K. & Unsicker, K. (1996) Distinct modulatory actions
of TGF-b and LIF on neurotrophin-mediated survival of devel-
oping sensory neurons. Neurochem. Res. 21, 843–850.
45. Unsicker, K. & Krieglstein, K. (2000) Co-activation of TGF-b and
cytokine signaling pathways are required for neurotrophic func-
tions. Cytokine Growth Factor Rev. 11, 97–102.
46. Kim, S.J., Park, K., Rudkin, B.B., Dey, B.R., Sporn, M.B. &
Roberts, A.B. (1994) Nerve growth factor induces transcription of
transforming growth factor-b1 through a specific promoter ele-
ment in PC12 cells. J. Biol. Chem. 269, 3739–3744.
47. Kimchi, A., Wang, X F., Weinberg, R.A., Cheifetz, S. & Mas-
sague
´
, J. (1988) Absence of TGF-b receptors and growth inhibi-
tory responses in retinoblastoma cells. Science 240, 196–199.
48. Jo
¨
rnvall, H., Blokzijl, A., Dijke, P. & Ibanez, C.F. (2001) The
orphan receptor serine/threonine kinase alk7 signals arrest of
proliferation and morphological differentiation in a neuronal cell
line. J. Biol. Chem. 276, 5140–5146.
49. Abdollah, S., Macı
´
as-Silva, M., Tsukazaki, T., Hayashi, H.,
Attisano, L. & Wrana, J.L. (1997) TbRI phosphorylation of Smad2
on Ser465 and Ser467 is required for Smad2-Smad4 complex
formation and signaling. J. Biol. Chem. 272, 27678–27685.
50. de Caestecker, M.P., Yahata, T., Wang, D., Parks, W.T., Huang,

S.,Hill,C.S.,Shioda,T.,Roberts,A.B.&Lechleider,R.J.(2000)
The Smad4 activation domain (SAD) is a proline-rich, p300-
dependent transcriptional activation domain. J. Biol. Chem. 275,
2115–2122.
51. Lutz, M. & Knaus, P. (2002) Integration of the TGF-b pathway
into the cellular signalling network. Cell. Signal. 14, 977–988
9
.
52. Sano, Y., Harada, J., Tashiro, S., Gotoh-Mandeville, R., Mae-
kawa, T. & Ishii, S. (1999) ATF-2 is a common nuclear target of
Smad and TAK1 pathways in transforming growth factor-b
signaling. J. Biol. Chem. 274, 8949–8957.
53. Gleizes,P.E.,Munger,J.S.,Nunes,I.,Harpel,J.G.,Mazzieri,R.,
Noguera, I. & Rifkin, D.B. (1997) TGF-b latency: biological sig-
nificance and mechanisms of activation. Stem Cells 15, 190–197.
54. Brown, J.D., DiChiara, M.R., Anderson, K.R., Gimbrone, M.A.
Jr & Topper, J.N. (1999) MEKK-1, a component of the stress
(stress-activated protein kinase/c- Jun N-terminal kinase) path-
way, can selectively activate Smad2-mediated transcriptional
activation in endothelial cells. J. Biol. Chem. 274, 8797–8805.
55. Chacko,B.M.,Qin,B.,Correia,J.J.,Lam,S.S.,deCaestecker,M.P.
& Lin, K. (2001) The L3 loop and C-terminal phosphorylation
930 M. Lutz et al.(Eur. J. Biochem. 271) Ó FEBS 2004
jointly define Smad protein trimerization. Nat. Struct. Biol. 8,
248–253.
56. Correia, J.J., Chacko, B.M., Lam, S.S. & Lin, K. (2001) Sedi-
mentation studies reveal a direct role of phosphorylation in
Smad3: Smad4 homo- and hetero-trimerization. Biochemistry 40,
1473–1482.
57. Pierreux, C.E., Nicolas, F.J. & Hill, C.S. (2000) Transforming

growth factor-b-independent shuttling of Smad4 between the
cytoplasm and nucleus. Mol. Cell. Biol. 20, 9041–9054.
58.Tsukazaki,T.,Chiang,T.A.,Davison,A.F.,Attisano,L.&
Wrana, J.L. (1998) SARA, a FYVE domain protein that recruits
Smad2 to the TGF-b receptor. Cell 95, 779–791.
59. Xu,L.,Chen,Y.G.&Massague
´
, J. (2000) The nuclear import
function of Smad2 is masked by SARA and unmasked by TGF-
b-dependent phosphorylation. Nat. Cell Biol. 2, 559–562.
60. Dong, C., Li, Z., Alvarez, R., Feng, X.H. & Goldschmidt-
Clermont, P.J. (2000) Microtubule binding to Smads may regulate
TGF-b activity. Mol. Cell. 5, 27–34.
61. Yanagisawa, M., Nakashima, K., Takeda, K., Ochiai, W.,
Takizawa, T., Ueno, M., Takizawa, M., Shibuya, H. & Taga, T.
(2001) Inhibition of BMP2-induced, TAK1 kinase-mediated
neurite outgrowth by Smad6 and Smad7. Genes Cells 6, 1091–
1099.
Ó FEBS 2004 Crosstalk between NGF and TGF-b pathways (Eur. J. Biochem. 271) 931