JNK mediates TGF-b1-induced epithelial mesenchymal
transdifferentiation of mouse transformed keratinocytes
Juan F. Santiban ˜ez
Laboratorio de Biologia Celular, Instituto de Nutricio ´n y Tecnologı ´a de los Alimentos, INTA, Universidad de Chile, Casilla 138, Santiago 11, Chile
terminal kinases (JNK) pathway in the TGF-b1 stimulation of
urokinase-type plasminogen activator (uPA), initial stages of
epithelial-mesenchymal transdifferentiation (EMT) and cell
migration. TGF-b1 induces JNK phosphorylation, c-Jun trans-
activation and AP1 activation. The involvement of JNK was
evaluated using dominant negative mutants SEK-1 AL, JNK
and cJun, depletion of JNK1,2 proteins by treatment of cells
with antisense oligonucleotides, as well as the chemical inhibitor
SP600125. Our results demonstrated that the JNK pathway is
required in the TGF-b1 enhancement of uPA, fibronectin, E-cad-
herin delocalization, actin re-organization and vimentin expres-
sion, concomitant with the induction of cell migration. These
results allow us to suggest a role of JNK in the TGF-b1 induction
of EMT in relation with the stimulation of malignant properties
of mouse transformed keratinocytes.
In this study we analyzed the role of the c-Jun N-
Keywords: JNK; TGF-b1; uPA; Keratinocytes; EMT
The process of epithelial-mesenchymal transdifferentiation
(EMT) is characterized by a set of transient phenotypic
changes often associated with the acquisition of migratory
properties by cancer cells and provides a means for cancer
propagation through out the organism [1,2].
The TGF-b super-family is implicated in the regulation of
cell proliferation,differentiation, migration, extracellular
matrix production, apoptosis and tumorigenesis . TGF-b
binds to the functional complex of the TGF-b family of recep-
tors at the cell surface [4,5], which, in turn, activate the Smads
and MAP kinases pathways including Ras, Erk1/2, and
JNK1 [6–12]. Transforming growth factor-b1 (TGF-b1) has
been postulated to have a dual role in tumour progression
by acting as tumor suppressor in the early stages of carcino-
genesis, and as pro-oncogenic in the last stages of metastatic
disease [13,14]; it also induces EMT of transformed cells
TGF-b1 increases plasminogen activator like-urokinase
(uPA) expression, which is regulated at the transcriptional
level by the AP-1 transcription factor [17,18]. The transcription
factor c-Jun (AP-1 component) is consequently activated
by the c-Jun NH2-terminal kinase (JNK) .
Although Ras-MAP kinases and the Smad signal pathways
contribute to malignant enhancement by TGF-b1 [10,20,21],
the mechanisms that mediate TGF-b1 transformed cell re-
sponses have not been fully elucidated. In the present study,
we have evaluated the role of the JNK pathway in the stimu-
lation of uPA, cell migration and EMT by TGF-b1. We found
that TGF-b1 activates JNK MAP kinase and transactivates
the c-Jun and AP1 complex. In addition, the inhibition of
the JNK pathway was found to affect uPA and cell migration
and consequently the initial step of malignant EMT of mouse
skin transformed cells.
2. Material and methods
2.1. Cell cultures and treatment conditions
The PDV cell line  was cultured as described . Cells were
incubated with TGF-b1 (Calbiochem-Novabiochem, La Jolla, CA)
at a final concentration of 10 ng/ml for the indicated period of time.
The chemical JNK inhibitor SP600125 (10 lM) provided by Calbio-
chem was added 30 min before the addition of TGF-b1.
AP-1-Luc and p-c-Fos-luc were provided by Dr. A. Corbı ´ (Centro
de Investigaciones Biolo ´gicas, Madrid, Spain), GAL4-c-Jun was pro-
vided by Dr. JL. Jameson (Northwestern University, Chicago, IL).
Vector pFA-Luc (5· GAL4-binding element) was purchased from
Stratagene. The pcDNA3.1 blank vector was obtained from Invitrogen
(Carlsbad, CA). The p-4.8 uPa-Luc luciferase reporter plasmid
(?4.8 kb of murine uPA promoter) was provided by Dr. P Munoz-
Canoves (Center for Genomic Regulation (CRG), Barcelona, Spain).
Dominant negative SEK1 AL (MEK4 mutant) was kindly provided
by Dr. J. R. Woodgett (York University, Toronto, Ontario, Canada).
Dominant Negative JNK and cJun were kindly provided by J. M. Re-
dondo (Instituto Severo Ochoa, UAM, Madrid, Spain).
The anti-phospho JNK monoclonal antibody, anti-JNK1 rabbit
polyclonal antibody, anti-p38 and anti-fibronectin monoclonal anti-
body were purchased from Santa Cruz Biotechnology (Santa Cruz,
CA). The anti-cadherin-E rat monoclonal antibody was kindly pro-
vided by Dr. M. Quintanilla (Instituto de Investigaciones Biome ´dicas,
Madrid, Spain). The anti-vimentin and anti a-tubulin monoclonal
antibody (Sigma, St. Louis, Mo).
Cells seeded on coverslips were fixed with 4% p-formaldehyde for
10 min at room temperature. For JNK1,2, F-actin and vimentin immu-
JNK, cJun N-terminal kinase; uPA, urokinase type plasminogen
activator; TGF-b1, Transforming growth factor-b1; Fn, Fibronectin;
E-cad, E-cadherin; vim, vimentin
E-mail address: email@example.com
nostaining cell monolayers were permeabilized with 0.1% Triton-X 100
for 2 min. at room temperature. F-actin was stained using Phalloidin-
Alexa Fluor (Molecular Probes, Eugene, OR), and secondary rabbit,
mouse and rat antibodies were coupled to FITC (Sigma, St. Louis,
Proteins were separated by SDS–PAGE and then transferred to
nitrocellulose membranes (BIORAD, Hercules, CA) which were
blocked in 4% milk (diluted in Tris-buffered saline and 0.5% Tween
20) and incubated with the appropriate antibody at 4 ?C overnight.
The targeted proteins were detected by enhanced chemiluminescence
as indicated by the manufacturers (Pierce).
2.6. Transient transfections and reporter gene measurements
For luciferase assays, PDV cells were transfected with superfect
(Quiagen) following themanufacturer’s instructions. Typically,
2 · 105cells were plated in each well of a 24-well plate. The next day
cells were transfected with 500 ng/well of each specific luciferase con-
struction, together with 25 ng/well SV40-b-Gal RL (Promega) as inter-
nal control for transfection efficiency. After 24 h of TGF-b1-treatment
cells were lysed and luciferase activity determined. For the GAL4 fu-
sion transactivation luciferase determinations, cells were co-transfected
with 0.5 lg of pFA-Luc and 0.2 lg of Gal4-cJun1.
2.7. Oligodeoxynucleotide treatments
JNK1 antisense (TCACGCTTGCTTCTGCTCAT) and JNK2 anti-
sense (TCACATTTACTGTCGCTCAT) phosphorothioate-modified
oligodeoxynucleotides  were synthesized and purified by Isogen
Bioscience BV (Maarssen, The Netherlands). As control, scrambled
S-oligo were used. A 1:1 mixture of antisense oligonucleotides for both
JNK 1 and 2 was added to the cells (50% confluent) and treated as de-
scribed by Santibanez et al. . The media containing oligos was
2.8. Zymographic and migration assay
The uPA secreted activity of cell cultures was determined by casein-
olytic zymography and the migration by wounded assay as previously
3.1. TGF-b1 induces JNK activation and c-Jun and AP1
transactivation in PDV cells
To examine whether TGF-b1 activates the JNK pathway,
we performed JNK phosphorylation, AP1 and cJun transacti-
vation assays. TGF-b1 increased JNK phosphorylation within
60 min. (Fig. 1A), reached its maximum (?5.0 fold time) at
120 min and started to decrease after 4 hours. This rapid and
( increase over control)
0 30 60 120 240
C SP SEK AL
C SP SEK AL
C SP SEK Al
+TGF- 1 1
Fig. 1. TGF-b1 induces on JNK Phosphorylation, AP1 and c-Jun transactivation in PDV cells. (A) Cell lysates were immunobloted and revealed
with either anti-p-JNK or JNK antibodies. Two independent experiments were performed and a representative is shown. Bottom part: densitometric
scans of results shown in top part. Cells were transiently transfected with Gal-4-cJun/pRf-Luc (B), AP1-luc (C) or c-Fos-luc (D). Dominants negative
SEK-AL, JNK or cJun were co-transfected, or pre-treated with JNK inhibitor SP600125 as in indicated points. Transfections and assays were
performed independently three times, each run in triplicate.
J.F. Santiban ˜ez
transient change suggests a direct effect of TGF-b1 on the JNK
pathway. Afterwards, we analyzed the capacity of TGF-b1 to
transactivate the Gal-4-c-Jun and AP1 costructs. TGF-b1 en-
hanced significantly c-Jun transactivation, which was strongly
inhibited by the dominant negatives SEK1 (SEK1-Al) and
JNK (DNJNK) or JNK inhibitor SP600125 (Fig. 1B). A sim-
ilar effect was observed in the TGF-b1-increased AP1 transac-
tivation which in addition was inhibited by the dominant
negative to c-Jun (DncJun) (Fig. 1C). To analyze the specificity
of JNK activation by TGF-b1 in PDV cells we analyzed the c-
Fos promoter, preferentially activated by the ERK1,2 pathway
in PDV cells (data not shown). TGF-b1 stimulated the c-Fos
promoter transactivation, and was not significantly inhibited
by SEK-AL or SP600125 (Fig. 1D). Together, these results
suggest that the TGF-b1 activates the JNK pathway in PDV
3.2. TGF-b1-induced uPA expression is mediated by JNK
To examine the role of JNK in the enhancement of cell
malignant transformation by TGF-b1, we investigated the
involvement of JNK in the regulation of upa, a gene known
to be induced by TGF-b1, and highly involved in the invasion
and migration of transformed cells [10,23]. To test whether
JNK participates in TGF-b1-enhaced uPA, we transiently
transfected PDV cells with mouse uPA promoter, and assayed
uPA activities in conditioned media by caseinolytic zymogra-
phy. After 24 h of TGF-b1 treatment of the uPA promoter
the activation was strongly inhibited by co-transfection with
either upstream SEK1 AL, dominant negative JNK or cJun
constructs, as well as by the chemical JNK inhibitor
SP600125 (Fig. 2A). As observed in Fig. 2B control cells se-
creted low levels of uPA and 24 h of TGF-b1 treatment of cells
induced an approximately 2.5 fold increase of a single band Mr
45Kd corresponding to mouse uPA. The TGF-b1-increased
uPA secreted activity was strongly suppressed by SP600125
(Fig. 2B). Thus, our results indicate that JNK activity was in-
volved in the TGF-b1-induced upregulation of both uPA tran-
scription and its enzimatic activity.
3.3. JNK mediates the TGF-b1-stimulated epithelial-
mesenchymal transdifferentation and cell migration
TGF-b1 is a potent inductor of EMT; this process occurs
associated with several changes in epithelial markers to the
mesenchymal phenotype [24,25]. We analyzed the effect of
JNK modulation on TGF-b1-induced EMT in PDV cells.
For this, we used two experimental approaches by applying
C SP SEK AL DNJNK DNcJun
p-4.8 uPa- Luc
TGF- 1 -
+ - +
uPA activity (%)
Fig. 2. The uPA promoter activation and secreted activity stimulated
by TGF-b1 were blocked by JNK inhibition. (A) Cells were transiently
transfected with p-4.8 Luc mouse uPA reporter gene. Dominants
negative for SEK-AL, JNK and cJun were co-transfected. (B)
Conditioned serum-free media from TGF-b1-stimulated cells (T) with
or without SP600125 (SP) treatment were subject to caseinolysis
zimographic analysis. Three independent experiments were performed
and a representative one is shown here.
Fig. 3. JNK mediates TGF-b1-induced epithelial-mesenchymal trans-
differentiation PDV cells and cell migration. (A) Cells seeded at 0.5 106
in six well plates were treated with JNK1,2 antisense-oligonucleotides,
AS, (5 and 20 lM), or scramble oligonucleotides, S, (20 lM)n. After
48 h of treatment cells were lysates and subject to Western blot analysis
for JNK1,2 and p38. Two independent experiments were performed
and a representative one is shown. (B) Cells seeded in coverslips were
treated as (A), except AS was only to 20 lM, and were immunostained
to JNK1,2. Photograph is representative of two independent experi-
ments. (C) Cells seeded in coverslips were immunostained to cellular-
derived FN (a–e) (1000·), E-cadherin (e–h) (400·), stained for F-actin
using phalloidin-Alexafluor 546 (i–l) (400·), and immunostained for
vimentin expression (m–p) (400·). Cells were stimulated with TGF-b1
(T) for 48 h, except vimentin that was stimulated for 96 h. 10 lM of
SP6001 (SP) and 20 lM JNK1,2 antisense oligonucleotides (AS) were
used respectively. All photographs are representative fields of at least
two or three independent experiments. (D and E) Statistical analysis
for E-cadherine loss contact and vimentin expression. For E- cadherine
the percentage of untreated cells was defined as 1, and the cells that lost
E-cadherin contact are shown as relative value against the control. For
vimentin a percentage of positive cells are represented. Values shown
are the means (+/? S.E.M.) of ten different fields of two independent
experiments. (F) Western blot for vimentin expression. Cells were
seeded as (A), and treated for four days with TGF-b1 in presence or
absence of 10 lM of SP6001 or 20 lM JNK antisense oligonucleotides,
a lysates cells samples were subject to Western blot analysis for
vimentin (Vm) and a-tubulin (a-Tub) as a control protein. Molecular
weight markers are indicated on the left. Two independent experiments
were performed and a representative is shown. (G) PDV cells
monolayer were wounded and stimulated with TGF-b (T) in the
presence or absence of SP600125 (SP) or AS JNK1,2 oligonucleotides
(20 lM). Photograph is representative of two independent experi-
J.F. Santiban ˜ez
antisense S-oligo strategie to cells depleted of JNK1,2 protein,
and JNK1,2 inhibition by using SP600125. As observed in
Fig. 3A, antisense oligonucleotides inhibited the expression
of JNK1,2 less intensely at 5 lM and strongly at 20 lM (about
90%), while the sense oligonucleotides had no effect on JNK
expression. As a control we analyzed the expression of the
p38 MAPK, and found no changes in p38 protein levels in cells
pretreated with the JNK sense or antisense oligos. Similar re-
sults were observed by immunofluorescence analysis, where
PDV cells treated with JNK antisense oligonucleotides dis-
played lower levels of immunoreactivity to the JNK1,2 anti-
body (Fig. 3B). We next tested whether JNK participates in
fibronectin (FN) production and in fibrillogenesis in PDV cells
stimulated by TGF-b1. After TGF-b1 treatment cells dis-
played an organized extracellular matrix of FN fibrils on their
surface (arrows, Fig. 3C (b)); this effect was strongly inhibited
by antisense oligos or SP600125 treatment (Fig. 3C (d and f)).
PDV cells pre-treated with the sense oligos responded to TGF-
b1 similarly to control cells (data not shown).
Cells treated with TGF-b1 displayed decrease of E-cadherin
cell–cell adhesion, showing a punctuated pattern distributed
along the cells surface (Fig. 3C (h)) as compared with control
cells (Fig. 3C (e)); this effect did not occur with JNK depletion
by antisense oligos or inhibition by SP600125, whereas the
intercellular distribution of E-cadherin was maintained after
TGF-b1 treatment, similarly to control cells (Fig. 3C (i and k)).
PDV cells showed strong F-actin staining distributed mainly
along cortical structures; after TGF-b1 treatment F-actin was
mainly organized in a network that resembles that of more mo-
tile cells (Fig. 3C (m and n)). Pre-treatment with JNK antisense
oligonucleotides or SP600125 exerted an inhibitory effect on
the F-actin reorganization in PDV cells, although these contin-
ued displaying cortical F-actin upon TGF-b1 treatment
(Fig. 3C (p and r)).
TGF-b1 treatment for four days induced cells to display a
vimentin cytoskeleton, unlike control cells that were mainly
negative for vimentin staining (Fig. 3C (s and u)). This effect
of TGF-b1 was strongly blocked by the JNK antisense oligo-
nucleotides or SP600125 inhibitor (Fig. 3C (m–p)). The effect
of interference on JNK by TGF-b1-induced EMT was con-
firmed by statistical analysis of the loss of E-cadherine cells
contact and vimentin expression; with TGF-b1 treatment
about 85% of the cells do not have E-cadherine cell contacts.
JNK antisense oligos or the JNK inhibitor block the effect of
TGF-b1 and cells remain in contact like the cells without
growth factor addition (Fig. 3D). During TGF-b1-treatment
about 76% of the cells expressed vimentin compared with con-
trol cells (4.7%); with antisense oligos or the inhibitor to JNK
this effect was strongly blocked to level comparable to that of
untreated cells (Fig. 3E). Furthermore, both antisense oligos
and JNK inhibitor blocked TGF-b1-induced vimentin when
analysed by Wester blotting (Fig. 3F), confirming the results
showed in Fig. 3C and E. Thus, these results indicate that
JNK1,2 play an important role in TGF-b1 induced EMT.
PDV cells were allowed to migrate in a wound healing assay
using TGF-b1; this growth factor induced cells to quickly close
the wound area after 24 h. However, cells treated with the anti-
sense oligonucleotides or SP6001 did not respond to TGF-b1-
induced motility, suggesting that JNK participates in TGF-b1-
induced migration (Fig. 3G).
Fig. 3 (continued)
J.F. Santiban ˜ez
In this study, we demonstrate that TGF-b1 induces uPA,
EMT and cellular migration in transformed mouse keratino-
cytes in vitro, using the JNK pathway as mediator. Treatment
of PDV cells with TGF-b1 activates the JNK pathway as evi-
denced by JNK phosphorylation and c-Jun and AP1 transac-
tivation (Fig. 1A). The kinetics of JNK phosphorylation
suggested its direct activation by TGF-b1; this is in agreement
with a previous report by Yue and co-workers . JNK acti-
vation by TGF-b1 is independent of TbRI-induced Smad acti-
vation, because ALK-5 mutant in the L45 loops of the kinase
domain, disabled to activate Smads, retained the capacity to
activate JNK .
We previously demonstrated that TGF-b1 induces uPA
expression, a key component of cell malignant transformation
[10,28]. The role of JNK on TGF-b1-induced uPA has not
been fully understood. Our results indicate that the JNK path-
way mediates the TGF-b1-induced uPA expression, because
dominant negatives of SEK1, JNK, cJun, and the JNK inhib-
itor SP600125 inhibit uPA-promoter transactivation; Further-
production. uPA expression is subject to regulation at the tran-
scriptional level by AP1. In addition, as it has already been de-
scribed c-Jun participates directly in uPA expression induced
by TGF-b1, and ATF2/c-Jun heterodimer is the predominant
form that promotes uPA expression . The promoter con-
struction used in this study has also been reported to be
Cadherine-E cell contact
relative to control
C SP AS
β ββ β1 - + -+ -+
C SP AS
Vimentin positive cells %
TGF- 1 - + -+ -+
TGF-β β1 1
TGF-β β1 - + -+ -+
C SP AS
Fig. 3 (continued)
J.F. Santiban ˜ez
effectively activated by overexpression of wild type c-Jun and
inhibited by mutated c-Jun .
TGF-b1-enhanced FN expression was dramatically blocked
by the JNK depletion using antisense oligonucleotides or inhi-
bition by SP600125 (Fig. 3C (a–f)). This is in agreement with
Hocevar et al.  who demonstrated that FN expression
was stimulated by TGF-b1 and was JNK-dependent in human
fibrosarcoma cells. TGF-b1-induced actin cytoskeleton reorga-
nization was mediated by JNK pathway (Fig. 3C (m–r)). How-
ever a direct link between actin organization and JNK has not
been clearly established. P150-Spir, which acts as an initiator
of the actin polymerization and is a downstream target of
JNK, may represent a direct link between JNK and actin orga-
At the initial stages of EMT, junctional protein complexes
are remodelled and functionally disintegrated [24,32]. After a
short interval TGF-b1 induces delocalization and loss of E-
cadherin cell contact in about 85% in PDV cells. This effect
of TGF-b1 on E-cadherin based cell-cell adhesion was abol-
ished by antisense oligos or SP600125, and remained close to
the actin cytoskeleton in the cell cortex (Fig. 3C (g–I) and
D). The mechanisms by which JNK could mediate TGF-b1-in-
duced E-cadherin delocalization remain unknown; it is possi-
ble that the JNK induced actin rearrangement could exert a
destabilizing effect on E-cadherin interaction with b and b-
catenin. E-cadherin plays a key role in epithelial integrity
[23,33], and mey act as a tumour-suppressor protein because
this downregulation plays a significant role in multistage carci-
Associated with the decrease of E-cadherin-based cell–cell
contacts, a concomitant increase in vimentin expression was
observed after TGF-b1 treatment (Fig. 3C (m–p), E and F);
this display of a mesenchymal type of cytoskeleton was depen-
dent on JNK expression and activity. Most invasive and/or
metastatic carcinomas are characterized by EMT, in which
the epithelial phenotype, exhibiting strong cell–cell junctions
and polarity, is replaced by a mesenchymal phenotype with
reduced cell–cell interactions and increased motility .
The cell migratory capabilities induced by TGF-b1 were
strongly dependent on JNK activity as shown in wound
healing assays (Fig. 3G); this is probably the result of JNK
inhibition upon TGF-b1-enhanced uPA, actin reorganization
and loss of E-cadherin cell adhesion. In summary, in this
study we show that JNK could mediate one of the early steps
of malignant EMT induced by TGF-b1 concomitant with a
more aggressive cell phenotype (uPA production and cell
migration), and we further suggest that the JNK pathway is
part of a network of signal pathways that mediate the
TGF-b1 stimulated malignant transformation of epithelial
Acknowledgements: Author thanks to M. Bozic, Oscar Brunser, C.
Bernabeu and M. Quintanilla for their support, suggestions and com-
ments. Funded by: FONDECYT Grant Nos. 3000045 and 1050476,
and DID (Universidad de Chile) I003-99/2 to (JFS).
 Hanahan, D. and Weinberg, R.A. (2000) The hallmarks of cancer.
Cell 100 (1), 57–70.
 Martinez-Arias, A. (2001) Epithelial mesenchymal interactions in
cancer development. Cell 105, 425–431.
 Massague, J. and Wotton, D. (2003) Transcriptional control by
the TGF-beta/Smad signaling system. EMBO J. 19 (8), 1745–1754.
 Dennler, S., Goumans, M.J. and ten Dijke, P. (2002) Transform-
ing growth factor beta signal transduction. J. Leukoc. Biol. 71 (5),
 Shi, Y. and Massague, J. (2003) Mechanisms of TGF-beta signal-
ing from cell membrane to the nucleus. Cell 113 (6), 685–700.
 Derynck, R. and Zhang, Y.E. (2003) Smad-dependent and Smad-
independent pathways in TGF-beta family signalling. Nature 425
 Mulder, K.M. and Morris, S.L. (1992) Activation of p21ras by
transforming growth factor beta in epithelial cells. J. Biol. Chem.
267 (8), 5029–5031.
 Iglesias, M., Frontelo, P., Gamallo, C. and Quintanilla, M. (2000)
Blockade of Smad4 in transformed keratinocytes containing a
Ras oncogene leads to hyperactivation of the Ras-dependent Erk
signalling pathway associated with progression to undifferentiated
carcinomas. Oncogene 19 (36), 4134–4145.
 MT, Mulder KM. 1997. Transforming growth factor-beta
signaling in epithelial cells. Pharmacol Ther. 75(1):21–41.
 Santibanez, J.F., Iglesias, M., Frontelo, P., Martinez, J. and
Quintanilla, M. (2000) Involvement of the Ras/MAPK signaling
pathway in the modulation of urokinase production and cellular
invasiveness by transforming growth factor-beta(1) in trans-
formed keratinocytes. Biochem. Biophys. Res. Commun. 273
 Frey, R.S. and Mulder, K.M. (1997) Involvement of extracellular
signal-regulated kinase 2 and stress-activated protein kinase/Jun
N-terminal kinase activation by transforming growth factor b in
the negative growth control of breast cancer cells. Cancer Res. 57,
 Zhou, G., Lee, S.C., Yao, Z. and Tan, T.H. (1999) Hematopoietic
progenitor kinase 1 is a component of transforming growth factor
beta-induced c-Jun N-terminal kinase signaling cascade. J. Biol.
Chem. 274 (19), 13133–13138.
 Roberts, A.B. and Wakefield, L.M. (2003) The two faces of
transforming growth factor beta in carcinogenesis. Proc. Natl.
Acad. Sci. USA 100 (15), 8621–8623.
 Oft, M., Heider, K.H. and Beug, H. (1998) TGFbeta signaling is
necessary for carcinoma cell invasiveness and metastasis. Curr.
Biol. 8 (23), 1243–1252.
 Caulin, C., Scholl, F.G., Frontelo, P., Gamallo, C. and Quinta-
nilla, M. (1995) Chronic exposure of cultured transformed mouse
epidermal cells to transforming growth factor-beta 1 induces an
epithelial-mesenchymal transdifferentiation and a spindle tumoral
phenotype. Cell Growth Differ. 6 (8), 1027–1035.
 Oft, M., Peli, J., Rudaz, C., Schwarz, H., Beug, H. and
Reichmann, E. (1996) TGF-beta1 and Ha-Ras collaborate in
modulating the phenotypic plasticity and invasiveness of epithelial
tumor cells. Genes Dev. 10 (19), 2462–2477.
 De Cesare, D., Vallone, D., Caracciolo, A., Sassone-Corsi, P.,
Nerlov, C. and Verde, P. (1995) Heterodimerization of c-Jun with
ATF-2 and c-Fos is required for positive and negative regulation
of the human urokinase enhancer. Oncogene 11 (2), 365–376.
 Himelstein, B.P., Lee, E.J., Sato, H., Seiki, M. and Muschel, R.J.
(1997) Transcriptional activation of the matrix metalloproteinase-
9 gene in an H-ras and v-myc transformed rat embryo cell line.
Oncogene 14 (16), 1995–1998.
 Karin, M. (1996) The regulation of AP-1 activity by mitogen-
activated protein kinases. Philos. Trans. R Soc. Lond. B Biol. Sci.
351 (1336), 127–134.
 Piek, E., Moustakas, A., Kurisaki, A., Heldin, C.H. and ten
Dijke, P. (1999) TGF-(beta) type I receptor/ALK-5 and Smad
proteins mediate epithelial to mesenchymal transdifferentiation in
NMuMG breast epithelial cells. J. Cell Sci. 112 (Pt 24), 4557–
 Oft, M., Akhurst, R.J. and Balmain, A. (2002) Metastasis is
driven by sequential elevation of H-ras and Smad2 levels. Nat.
Cell Biol. 4 (7), 487–494.
 Diaz-Guerra, M., Haddow, S., Bauluz, C., Jorcano, J.L., Cano,
A., Balmain, A. and Quintanilla, M. (1992) Expression of simple
epithelial cytokeratins in mouse epidermal keratinocytes harbor-
ing Harvey ras gene alterations. Cancer Res. 52 (3), 680–687.
 Thiery, J.P. (2002) Epithelial-mesenchymal transitions in tumour
progression. Nat. Rev. Cancer 2 (6), 442–454.
J.F. Santiban ˜ez
 Boyer, B., Valles, A.M. and Edme, N. (2000) Induction and Download full-text
regulation of epithelial-mesenchymal transitions. Biochem. Phar-
macol. 60 (8), 1091–1099.
 Gotzmann, J., Mikula, M., Eger, A., Schulte-Hermann, R.,
Foisner, R., Beug, H. and Mikulits, W. (2004) Molecular aspects
of epithelial cell plasticity: implications for local tumor invasion
and metastasis. Mutat. Res. 566 (1), 9–20, Review.
 Yue, J., Sun, B., Liu, G. and Mulder, K.M. (2004) Requirement
of TGF-beta receptor-dependent activation of c-Jun N-terminal
kinases (JNKs)/stress-activated protein kinases (Sapks) for TGF-
beta up-regulation of the urokinase-type plasminogen activator
receptor. J. Cell Physiol. 199 (2), 284–292.
 Itoh, S., Thorikay, M., Kowanetz, M., Moustakas, A., Itoh, F.,
Heldin, C.H. and ten Dijke, P. (2003) Elucidation of Smad
requirement in transforming growth factor-beta type I receptor-
induced responses. J. Biol. Chem. 278 (6), 3751–3761.
 Tanaka, Y., Kobayashi, H., Suzuki, M., Kanayama, N. and
Terao, T. (2004) Transforming growth factor-beta1-dependent
urokinase up-regulation and promotion of invasion are involved
in Src-MAPK-dependent signaling in human ovarian cancer cells.
J. Biol. Chem. 279 (10), 8567–8576.
 Miralles, F., Parra, M., Caelles, C., Nagamine, Y., Felez, J. and
Munoz-Canoves, P. (1998) UV irradiation induces the murine
urokinase-type plasminogen activator gene via the c-Jun N-
terminal kinase signaling pathway: requirement of an AP1
enhancer element. Mol. Cell Biol. 18 (8), 4537–4547.
 Hocevar, B.A., Brown, T.L. and Howe, P.H. (1999) TGF-beta
induces fibronectin synthesis through a c-Jun N-terminal kinase-
dependent, Smad4-independent pathway. EMBO J. 18 (5), 1345–
 Otto, I.M., Raabe, T., Rennefahrt, U.E., Bork, P., Rapp, U.R.
and Kerkhoff, E. (2000) The p150-Spir protein provides a link
between c-Jun N-terminal kinase function and actin reorganiza-
tion. Curr. Biol. 10 (6), 345–348.
 Bissell, M.J. and Radisky, D. (2001) Putting tumours in context.
Nat. Rev. Cancer. 1 (1), 46–54.
 Christofori, G. and Semb, H. (1999) The role of the cell-adhesion
molecule E-cadherin as a tumour-suppressor gene. Trends
Biochem. Sci. 24 (2), 73–76.
 Birchmeier, W. and Behrens, J. (1994) Cadherin expression in
carcinomas: role in the formation of cell junctions and the pre-
vention of invasiveness. Biochim. Biophys. Acta. 1198 (1), 11–26.
 Dandachi, N., Hauser-Kronberger, C., More, E., Wiesener, B.,
Hacker, G.W., Dietze, O. and Wirl, G. (2001) Co-expression of
tenascin-C and vimentin in human breast cancer cells indicates
phenotypic transdifferentiation during tumour progression: cor-
relation with histopathological parameters, hormone receptors,
and oncoproteins. J. Pathol. 193 (2), 181–189.
 Nagata, Y., Takahashi, N., Davis, R.J. and Todokoro, K. (1998)
Activation of p38 MAP kinase and JNK but not ERK is required
for erythropoietin-induced erythroid differentiation. Blood 92 (6),
J.F. Santiban ˜ez