ABSTRACT Prompted by earlier findings that the Rac1-related isoform Rac1b inhibits transforming growth factor (TGF)-β1-induced canonical Smad signalling, we studied here whether Rac1b also

impacts TGF-β1-dependent non-Smad signalling such as the MKK6-p38 and MEK-ERK mitogen-activated protein kinase (MAPK) pathways and epithelial-mesenchymal transition (EMT). Transient

depletion of Rac1b protein in pancreatic cancer cells by RNA interference increased the extent and duration of TGF-β1-induced phosphorylation of p38 MAPK in a Smad4-independent manner. Rac1b

depletion also strongly increased basal ERK activation - independent of the kinase function of the TGF-β type I receptor ALK5 - and sensitised cells towards further upregulation of

cell migration. Rac1b depletion also enhanced TGF-β1 autoinduction of crucial TGF-β pathway components and decreased that of TGF-β pathway inhibitors. Our results show that Rac1b antagonises

TGF-β1-dependent EMT by inhibiting MKK6-p38 and MEK-ERK signalling and by controlling gene expression in a way that favors attenuation of TGF-β signalling. SIMILAR CONTENT BEING VIEWED BY

OTHERS RNF12 IS REGULATED BY AKT PHOSPHORYLATION AND PROMOTES TGF-Β DRIVEN BREAST CANCER METASTASIS Article Open access 10 January 2022 PHARMACOLOGICAL BLOCKAGE OF TRANSFORMING GROWTH

FACTOR-Β SIGNALLING BY A TRAF2- AND NCK-INTERACTING KINASE INHIBITOR, NCB-0846 Article Open access 27 November 2020 SMAD2/3 MEDIATE ONCOGENIC EFFECTS OF TGF-Β IN THE ABSENCE OF SMAD4 Article

Open access 07 October 2022 INTRODUCTION Pancreatic ductal adenocarcinoma (PDAC) is one of the most deadliest diseases for which no curative therapies are available to date. To successfully

establish prevention and treatment strategies for this disease, a better understanding of the molecular events underlying PDAC tumourigenesis is mandatory. Transgenic mouse models have

shown that aggressive PDAC develops after pancreas-specific inhibition of transforming growth factor-beta (TGF-β) signalling in cooperation with active K-Ras expression1. However, the

effector pathways of the TGF-β/K-Ras crosstalk remain elusive. Data from a _K-ras_ G12D murine model with pancreas-specific ablation of _RAC1_ suggested that the protein product(s) of _RAC1_

is a crucial mediator of TGF-β/K-Ras-driven tumourigenesis since it prevented tumour development and significantly prolonged survival in these mice2. Although the oncogenic role of _Rac1_

in this context has clearly been established, data interpretation remains problematic as _Rac1_ gives rise to two different proteins, Rac1 and its splice variant, Rac1b. Rac1b differs from

Rac1 by inclusion of a short exon (exon 3b, comprising 19 amino acids) close to the switch II region3,4. As a consequence of this modification, Rac1b has been found to have an accelerated

GDP/GTP exchange and delayed GTP hydrolysis5 and to differ from Rac1 in certain signalling and functional properties. Rac1b does not interact with RhoGDI or p21-activated kinase and does not

induce lamellipodia formation6, but retains the potential to increase cellular reactive oxygen species7. Since Rac1b is expressed at a much lower level than Rac1 in cells, it is normally

not detected in immunoblot analyses and thus not analysed. Moreover, because of inevitable co-deletion of Rac1b upon _RAC1_ ablation, the antitumour effects observed in the above mentioned

mouse model cannot be ascribed unequivocally to the absence of Rac1. A solution to this dilemma would be a selective depletion of exclusively one of both isoforms, however, such data are not

yet available. As far as Rac1 is concerned, we have shown earlier that Rac1 promotes TGF-β1 signalling in PDAC-derived cell lines towards a pro-metastatic outcome by enhancing

TGF-β1-induced Smad2 activation, epithelial-mesenchymal transition (EMT), and random cell migration and invasion8. Recently, we have detected Rac1b protein in tumour tissues of PDAC patients

with expression being most prominent in the tumour cell fraction. Intriguingly, high Rac1b expression correlated with fewer metastases and significantly prolonged survival times compared to

patients that lacked Rac1b expression in their tumour cells9. These finding argue in favor of an antimetastatic - and thus Rac1 antagonistic - effect of Rac1b in the context of a

TGF-β1-rich microenvironment. It was therefore of interest to study _in vitro_ i) how Rac1b controls tumour cell responses to TGF-β that are associated with malignant conversion such as EMT

and cell migration/invasion and ii) which signalling pathways are targetted by Rac1b. In keeping with the idea that Rac1b represents an endogenous inhibitor of Rac1, we observed earlier that

Rac1b inhibits TGF-β1-induced random cell migration and suppresses the C-terminal phosphorylation, and thus activation, of both Smad2 and Smad39. TGF-β-induced activation of Smad complexes

has crucial roles during induction of EMT10,11. However, whereas Smad4 and Smad3 promote EMT, Smad2 can inhibit it12. Hence, negative regulation of Smad2 _and_ Smad3 activation would not

explain the effect, if any, of Rac1b on TGF-β-induced EMT. Various studies have shown that TGF-β1-dependent control of EMT and mesenchymal traits such as matrix production and cell motility

may not only depend on canonical Smad- but also on non-canonical Smad and non-Smad signalling, sometimes in a tissue and cell-type specific manner13,14,15. Non-Smad signalling during EMT

leads to activation of Rho GTPases16, mitogen-activated protein kinase (MAPK) pathways, and the PI3 kinase-Akt-mTOR pathway13,14,15. The MKK3/6-p3810,11,13,17 and the MEK-extracellular

signal-regulated kinase (ERK) MAPK pathways10,11,14,18 control non-transcription changes/gene reprogramming and during EMT cooperate with Smad-mediated gene expression, _e.g_. through the

transcription factor ATF219, but may also directly regulate the stabilities and activities of Smads15. The ubiquitin ligase TRAF6 binds to the TGF-β type I receptor ALK5 and mediates

Smad-independent activation of the MAPKKK TAK1, the MAPKKs MKK3/MKK6, and the JNK/p38 MAPKs20. The ERK pathway is stimulated by activated ALK5 through tyrosine phosphorylation of the adaptor

protein Shc, allowing docking of the Grb2-Sos1 complex which subsequently leads to downstream activation of the Ras-Raf-MEK-ERK pathway21. In light of the previously observed (negative)

regulation of TGF-β1-induced Smad2/3 activation and random cell migration in PDAC-derived cells8,9, it is conceivable that Rac1b also affects activity of TGF-β1-dependent MKK6-p38 and/or

MEK-ERK signalling as well as adoption of a mesenchymal phenotype. It should be mentioned, however, that due to the Smad- and ALK5 serine/threonine kinase-independent activation of p38 and

ERK1/2, respectively, the effects of Rac1b on the Smad signalling pathway are not predictive of the effects of Rac1b on either the MKK6-p38 or the MEK-ERK signalling pathway. To study the

effect of Rac1b on TGF-β1-induced p38 and ERK activation and on EMT, we primarily utilized the human PDAC-derived and TGF-β1-responsive cell lines Panc1, Colo357 and IMIM-PC1. In some

experiments we included non-tumourigenic cells of pancreatic and non-pancreatic origin to evaluate whether the observed effects were tumour and tissue-specific, respectively. METHODS

ANTIBODIES AND REAGENTS The following primary antibodies were used: Anti-phospho-p38 (#9211), anti-p38 (#9212), anti-phospho-ERK1/2 (#4370), anti-E-Cadherin (#3195), anti-Snail (#4719) (all

from Cell Signaling Technology, Frankfurt/Main, Germany), anti-HSP90 (both #sc-7947 and #sc-13119), anti-MKK6 (sc-6073), anti-TGF-β receptor I/ALK5 (V22, #sc-398), anti-TGF-β1 (3C11,

#sc-130348) (all from Santa Cruz Biotechnology, Heidelberg, Germany), anti-ERK1/2 (#AF1576, R&D Systems, Wiesbaden, Germany) anti-Rac1b (#09-271, Merck Millipore, Darmstadt, Germany),

anti-Rac1 (#610650), anti-Cip1/WAF1 (#610233) (both from BD Biosciences, Heidelberg, Germany), anti-β-actin (#A1978, Sigma, Deisenhofen, Germany). anti-Flag M2 (F3165, Sigma), HRP-linked

anti-rabbit (#7074), anti-mouse (#7076) and anti-rat (#7077) secondary antibodies were from Cell Signaling Technology, anti-goat secondary antibody (#ab6741) was from Abcam (Cambridge, UK).

Recombinant human (rh) TGF-β1 (#300-023) was purchased from ReliaTech (Wolfenbüttel, Germany) and used at a concentration of 5 ng/ml. The p38 inhibitor SB203580, the MEK1 inhibitor U0126 and

the ALK5 inhibitor SB431542 were purchased from Calbiochem and used at a concentration of 10 μM (SB203580, UO126) and 5 μM (SB431542). Treatment of cells with these inhibitors for up to 48 

h had no gross effect on cell viability. CELL CULTURE AND GENERATION OF PANC1 CELLS ECTOPICALLY EXPRESSING HA-RAC1B OR DOMINANT-NEGATIVE MUTANTS OF MKK6 OR ALK5 Panc1 and Colo357 human PDAC

cells were originally obtained from ATCC (Manassas, VA). Another PDAC-derived cell line, IMIM-PC1, was obtained from P. Real (University of Madrid) and kindly supplied by A. Menke

(University of Giessen). Panc1 and Colo357 cells were cultured in RPMI 1640 supplemented with 10% heat-inactivated fetal bovine serum (FBS), 1% Penicillin-Streptomycin-Glutamine (Life

Technologies) and 1% sodium pyruvate (Merck Millipore). IMIM-PC1 cells and HaCaT immortalized human keratinocytes (ATCC) were maintained in DMEM containing the same supplements. The human

ductal pancreatic epithelial cell line HPDE6c7 was a kind gift of S. Sebens (University of Kiel) and was cultured as described22. The generation of Panc1 cell clones ectopically expressing

HA-Rac1b9, dominant-negative mutants of MKK6 (MKK6KA, carrying K82A substitution)23,24, ALK5 (ALK5KR, carrying K232R substitution)25 or Flag-Smad41-514 26 was described in detail earlier. In

all cases, cells were stably transduced using the retroviral vector TJBA5bMoLink-neo, followed by selection of successfully transduced cells with G418 (700 µg/ml) and generation of

individual cell clones using limited dilution. Pooled empty-vector transductants served as a control. Ectopic expression of the mutant proteins was verified by immunoblotting for Rac1b, MKK6

and ALK5, respectively. CELL COUNTING The number of Panc1 cells with spindle-shaped morphology after Rac1b siRNA transfection and TGF-β1 treatment was counted per visual field by two

investigators in a blinded fashion. QUANTITATIVE REAL-TIME RT-PCR (QPCR) ANALYSIS Total RNA was extracted from Panc1 cells grown in 24-well plates using PeqGold RNAPure (Peqlab, Erlangen,

Germany) and purified according to manufacturer’s instructions. For each sample, 2.5 μg RNA were subjected to reverse transcription for 1 h at 37 °C, using 200 U M-MLV Reverse Transcriptase

and 2.5 μM random hexamers (both from Life Technologies) in a total volume of 20 μl. Relative mRNA expression of target genes was quantified by real-time PCR on an I-Cycler (BioRad) using

Maxima SYBR Green Mastermix (Thermo Fisher Scientific). Data were normalized to the expression of TATA-box-binding protein (TBP) for each sample. For PCR primers see Supplementary Table S1.

TRANSIENT TRANSFECTION OF SIRNAS On day 1 and 2 after seeding into 6, 12 or 24-well plates (NunclonTM Delta Surface, Nunc, Roskilde, Denmark) cells were transfected with 50 nM of siRNA

specific for Rac1b or scrambled control, siRNA specific for Rac1 + Rac1b (ON-TARGETplus SMARTpool, a mixture of four prevalidated siRNAs) or matched negative control (non-target control

SMARTpool), both purchased from GE Healthcare Dharmacon (Epsom, UK) or a pool of three different and validated siRNAs to ALK5 (Validated Stealth RNAiTM siRNA (Set of 3) HSS110695, HSS110696,

HSS110697) or matched negative control (Life Technologies), or siRNA to BGN or matched negative control (Qiagen, CA) for 4 h using Lipofectamine RNAiMAX (Life Technologies) at a

concentration of 0.5%. REPORTER GENE ASSAY For reporter gene assays, Panc1 cells were seeded in 96-well plates (NunclonTM Delta Surface) and cotransfected on the following day serum-free for

4 h with Lipofectamine 2000 (Life Technologies) and 50 nM of BGN siRNA or control siRNA. Afterwards, cells received standard growth medium. Twenty-four h after the start of the first

transfection, cells underwent a second round of transfection with 50 nM each of BGN or control siRNA plus 100 ng/well of the TGF-β-responsive luciferase reporter plasmid p3TP-Lux and 25

ng/ml pRL-TK-Luc, a vector encoding Renilla luciferase (Promega, Heidelberg, Germany). On the next day, cells were stimulated with 5 ng/ml TGF-β1 for 24 h and then lysed in Glo lysis buffer

(Promega) and subjected to dual luciferase measurement with the Dual Luciferase Assay System according to the manufacturer’s protocol (Promega). [3H]-THYMIDINE INCORPORATION ASSAY This assay

was performed exactly as described in detail earlier8. CELL LYSIS AND IMMUNOBLOTTING Confluent cells were washed once with ice-cold PBS and lysed with 1x PhosphoSafe lysis buffer (Merck

Millipore). Cell lysates were sonicated and centrifuged for 10 min at 14.000 × g and 4 °C following determination of total protein concentration in supernatants using BioRad DC Protein

Assay. Samples containing equal amounts of protein were prepared using 3x SDS Sample Buffer and 125 mM Dithiothreitol (both from New England Biolabs), subjected to gel electrophoresis using

BioRad mini-PROTEAN TGX any-kD precast gels and blotted to 0.45 μm PVDF membranes. Membranes were blocked with nonfat dry milk (Carl Roth GmbH, Karlsruhe, Germany) or BSA (Sigma-Aldrich) and

incubated with primary antibodies at 4 °C overnight. HRP-linked secondary antibodies and Amersham ECL Prime Detection Reagent (GE Healthcare) were used for detection of proteins on a BioRad

ChemiDoc XRS imaging system. RotiFree stripping buffer (Carl Roth GmbH) was used for membrane stripping. Signal intensities were quantified by densitometry and computed with either NIH

image J or Image Lab (version 5.2.1, BioRad). ELISA FOR TGF-Β1 Twenty-four h after the second transfection with either control siRNA or TGF-β1 siRNA Panc1 cells received fresh medium

containing 0.5% FBS and culture supernatants were allowed to be conditioned for another 24 h. Aliquots from the culture supernatants were cleared by centrifugation, appropriately diluted and

subjected to a TGF-β1-specific ELISA (Human/Mouse TGF beta1 ELISA Ready-SET-Go!, eBioscience/Affymetrix Inc. San Diego, CA) according to the manufacturer’s instructions. The detection limit

was 25 pg/ml. Data for total TGF-β1 were normalised to the cell number from the respective well. REAL-TIME CELL MIGRATION ASSAYS The xCELLigence® DP system (ACEA Biosciences, distributed by

OLS, Bremen, Germany) was used to measure random migratory activity of wild-type Panc1 cells and stably transduced Panc1 cell clones. Cells were seeded in 6-well plates, treated as desired

and then serum-starved (standard growth medium containing 0.5% FBS) for 24 h prior to transferring the cells to the assay. For all assays, RPMI with 1% FBS and a final concentration of

TGF-β1 of 5 ng/ml in both the upper and lower chambers of the CIM plates-16 was used. CIM plates-16 were prepared according to the instruction manual and previous descriptions9,27,28. The

underside of the upper chambers of the CIM plate-16 was coated with 30 μl of collagen I (400 μg/ml) and allowed to dry for at least 2 h prior to plate assembly. 40,000 cells were loaded into

each well of the upper chamber immediately after addition of TGF-β1 to the cell suspensions. Data acquisition was at 15 min intervals and analysis done with the RTCA software. STATISTICAL

ANALYSIS Statistical significance was calculated using the Mann-Whitney u test. Results were considered significant at _p_ < 0.05 (*). Higher levels of significance were _p_ < 0.01

(**) and _p_ < 0.001 (***). DATA AVAILABILITY STATEMENT All data generated or analysed during this study are included in this published article (and its Supplementary Information files).

RESULTS RAC1B SILENCING INCREASES THE DURATION OF THE P38 MAPK PHOSPHORYLATION RESPONSE TO TGF-Β1 We have previously shown that TGF-β1-induced activation of both Smad2 and p38 MAPK is

Rac1-dependent in PDAC-derived cells8,29. Since Rac1b inhibits Smad2/3 activation, we hypothesised that Rac1b might also inhibit p38 activation. To study this in more detail, we first

depleted the PDAC cell lines Panc1 and Colo357 of cellular Rac1b protein by transfection of a siRNA targetting exon 3b unique to Rac1b9, or an irrelevant control siRNA, and subsequently

exposed the cells for various times to rhTGF-β1. In Panc1 cells, the levels of phospho-p38 (p-p38) increased in both control siRNA and Rac1b siRNA-transfected cells and in both groups the

differences became significant at 2 h of TGF-β1 treatment (Fig. 1A). At the 4 h time point p-p38 levels in control siRNA-transfected cells had returned to baseline levels while those of

Rac1b siRNA-transfected cells remained elevated for up to 12 h (Fig. 1A). A very similar and delayed activation of p38 by TGF-β with peak levels of p-p38 at 2 h and a subsequent decline has

been described earlier in (non-transfected) Panc1 cells30. In Colo357 cells, p-p38 levels peaked at 1 h of TGF-β1 treatment in both control siRNA and Rac1b siRNA-transfected cells and

declined thereafter but remained elevated over those in non-TGF-β1-treated controls (Supplementary Fig. S1). As in Panc1 cells, the p-p38 levels in Rac1b-depleted at all time points appeared

higher than in control cells, however, only after 4 and 8 h of TGF-β1 stimulation these differences reached statistical significance (Supplementary Fig. S1). To analyse whether or not the

Rac1b effect was specific for PDAC-derived cells, we studied the effects of Rac1b depletion in HaCaT keratinocytes (Fig. 1B). Upon an 1-h TGF-β1 treatment, p-p38 levels in both control siRNA

and Rac1b siRNA-transfected cells rose dramatically and thereafter declined but for Rac1b siRNA remained elevated over non-stimulated controls even after 4 h of TGF-β1 exposure (Fig. 1B,

lane 2 _vs_. 8). After 1 h but not after 2 and 4 h, the TGF-β1 stimulated levels of p-p38 were significantly higher in Rac1b siRNA than in control siRNA-transfected cells (Fig. 1B, lane 3

_vs_. 4). Further analysis in Panc1 cells revealed that the increase in p-p38 levels in control siRNA-transfected cells after 2 h of TGF-β1 treatment was blunted by ectopic expression of a

kinase-dead ALK5 mutant (ALK5KR) (Fig. 1C). ALK5KR also prevented the TGF-β1-induced upregulation of p-p38 in Rac1b-depleted cells (Fig. 1C, lane 8 _vs_. 10) and suppressed p-p38 levels in

non-TGF-β1-treated control cells (Fig. 1C, compare lane 1 with lane 7 and lane 2 with lane 8). The latter effect may reflect disruption of an autocrine TGF-β signalling loop (see

Discussion). Since the ALK5 kinase phosphorylates both Smads and p38 and phosphorylation of p38 in some cell types has been shown to be Smad-dependent, we sought to reveal whether in Panc1

cells the Rac1b effect on TGF-β1-induced p38 activation is Smad-dependent. To clarify this issue, we employed Panc1 cells stably expressing a dominant-negative mutant of Smad4

(Smad41-514)26. We have demonstrated previously that these cells have defective Smad signaling as they failed to respond to TGF-β1 treatment with upregulation of biglycan (BGN)26. When we

depleted these cells of Rac1b, we observed higher abundance of p-p38 in non-TGF-β1 stimulated Smad41-514 expressing cells vs the corresponding empty vector control cells (Fig. 1D). However,

no difference in the response of Smad41-514 expressing cells compared to the vector control cells was seen after 2 and 4 h of TGF-β1 stimulation although there seems to be a trend towards

higher p38 activation in Panc1- Smad41-514 cells at 2 h time point (Fig. 1D). Panc1 cells have been shown to secrete large amounts of TGF-β1 and stimulate themselves in an autocrine

fashion31. To assess the relative effects of endogenous TGF-β1 vs. rhTGF-β1 on p38 activation upon Rac1b-depletion, we cotransfected Panc1 cells with siRNAs specific for TGF-β1 and Rac1b and

determined the levels of p-p38 as above. As shown in Fig. 1E, disrupting endogenous TGF-β1 secretion (verified by a TGF-β1-specific ELISA, Fig. 1E) and hence the autocrine stimulatory

feedback loop strongly reduced p-p38 levels in both rhTGF-β1-treated cells (compare lanes 7 and 8 with lanes 5 and 6, and lanes 11 and 12 with lanes 9 and 10) and in non-rhTGF-β1-treated

control cells (compare lanes 3 and 4 with lanes 1 and 2) (Fig. 1E). However, while the increase in p-p38 levels between control and Rac1b siRNA-transfected cells was maintained in TGF-β1

siRNA-transfected cells after the 2 h-treatment with rhTGF-β1 (lanes 7 and 8) it was lost after 4 h of treatment (lanes 11 and 12). Together, these data confirmed our assumption that Rac1b

inhibits both an early (<2 h, in Colo357 and HaCaT cells) and a late (2–12 h, in Panc1 cells) phosphorylation response of p38 to rhTGF-β1 stimulation. Moreover, dominant-negative

inhibition of ALK5 but not Smad4 in Panc1 cells blunted the TGF-β1-induced upregulation of p-p38 in both control and Rac1b siRNA-transfected cells. Finally, a large fraction of p-p38 under

basal conditions is induced by endogenous TGF-β1 and endogenous and rhTGF-β1 cooperate to maintain p-p38 levels in Rac1b-depleted cells higher than in control cells beyond 2 h of stimulation

with rhTGF-β1. RAC1B DEPLETION RESULTS IN AN INCREASE IN BASAL AND TGF-Β1-INDUCED P-ERK1/2 LEVELS In PDAC cells ERK1/2 MAPK signalling is activated downstream of Ras and Rac1 but may also

be stimulated by TGF-β through the intrinsic tyrosine kinase activity of ALK521. To test whether Rac1b is involved in ERK activation, we monitored by immunoblotting the phosphorylation

status of ERK1/2 in various cell lines (Panc1, Colo357, IMIM-PC1 and HaCaT) after siRNA-mediated, selective depletion of endogenous Rac1b. In control siRNA-transfected Panc1 cells we

observed an only moderate (2-fold) induction of p-ERK1/2 over a TGF-β1 stimulation period of 18 h (Fig. 2A, lanes 1 _vs_. 7). Rac1b depletion in non-stimulated Panc1 cells enhanced p-ERK1/2

levels by 15-fold (Fig. 2A, lanes 1 _vs_. 2). TGF-β1 stimulation of Rac1b-depleted cells for 12 and 18 h increased ERK levels by 1.9 and 2.1-fold, respectively, relative to non-stimulated

control cells (Fig. 2A, lane 2 _vs_. 6, and lane 2 _vs_. 8). In control siRNA-transfected Colo357 and IMIM-PC1 cells p-ERK1/2 was either not induced or induced 2-fold, respectively, within 1

 h of TGF-β1 stimulation (Fig. 2B, lane 1 _vs_. 3). Rac1b depletion, however, enhanced p-ERK1/2 levels in non-stimulated Colo357 and IMIM-PC1 cells by 1.6-fold and 2.2-fold, respectively

(Fig. 2B, lane 1 _vs_. 2). Importantly, a 1-h treatment with TGF-β1 further enhanced the p-ERK1/2 levels in Rac1b-depleted Colo357 and IMIM-PC1 cells 1.8 and 1.4-fold, respectively, relative

to non-stimulated controls (Fig. 2B, lane 2 _vs_. 4). It could be speculated that the changes in p-ERK levels were stronger if a more complete inhibition of Rac1b protein levels would have

been obtained in these two cell lines. Very similar results with respect to the effect of Rac1b depletion on TGF-β1 stimulated ERK activation was seen in HaCaT cells (Supplementary Fig. S2).

To test if the Rac1b siRNA-mediated increase in p-ERK1/2 levels in non-stimulated cells was the result of relief from a TGF-β autocrine feedback loop, we blocked TGF-β signalling in Panc1

cells by either ectopic expression of kinase-inactive ALK5KR (Fig. 2C) or by siRNA to ALK5 (Fig. 2D). In Panc1-ALK5KR cells, p-ERK1/2 levels were not different from those in control cells

after Rac1b depletion (Fig. 2C). Likewise, in Panc1 cells codepleted of ALK5 and Rac1b protein by siRNA levels of p-ERK1/2 appeared enhanced rather than decreased relative to cells only

depleted of Rac1b (Fig. 2D, lane 3 _vs_. 7). However, as expected, ALK5 siRNA prevented appearance of the TGF-β1-induced fracton of p-ERK1/2 protein in Rac1b siRNA-transfected cells (Fig. 

2D, lane 4 _vs_. 8). To study the impact of combined Rac1 + Rac1b depletion on ERK activation, we transfected Panc1 cells with a Rac1 siRNA (which targets both Rac1 and Rac1b). In contrast

to selective depletion of Rac1b, codepletion of Rac1 _and_ Rac1b did not result in a statistically significant increase in p-ERK1/2 levels (Supplementary Fig. S3), suggesting that while

Rac1b inhibition promotes ERK1/2 activation under basal conditions, concomitant inhibition of Rac1 has the opposite effect and was able to override the Rac1b siRNA effect. ECTOPIC

OVEREXPRESSION OF RAC1B DECREASES TGF-Β-DEPENDENT ERK1/2 ACTIVATION AND INHIBITS TGF-Β TARGET GENE EXPRESSION In a previous study, we have shown that stable ectopic overexpression of

HA-Rac1b in PDAC cells suppresses TGF-β1-mediated cell migration9. To study if this effect is associated with a concomitant downregulation of signalling pathways promoting cell motility such

as Ras-MEK-ERK1/2, we monitored p-ERK1/2 levels in two previously characterised independent HA-Rac1b expressing clones of the Panc1 cell line9 after 12 h of TGF-β1 treatment (the time of

maximal ERK activation, see Fig. 2A). To this end, p-ERK1/2 levels were significantly reduced in both clones relative to empty vector controls clones (Fig. 3A). To analyse whether ectopic

Rac1b also impacts regulation of other TGF-β target genes we studied in the same cells regulation of E-cadherin, an EMT-associated gene that is downregulated by this growth factor, and

regulation of proteinase-activated receptor 2 (PAR2), a G protein-coupled receptor implicated in tumour cell invasion/metastasis32 and required for a full-blown TGF-β response28. In line

with previous results on TGF-β-dependent expression of Slug9, downregulation of E-cadherin and upregulation of PAR2 by TGF-β1 was greatly reduced in both clones (Fig. 3B). Together with the

data presented in Fig. 2, this indicates that Rac1b is a negative regulator of TGF-β1-induced ERK activation and gene expression in PDAC-derived cells. RNAI-MEDIATED DEPLETION OF RAC1B

ENHANCES THE EXPRESSION OF GENES ASSOCIATED WITH TGF-Β1-INDUCED EMT Cell migration is known to be associated with the process of EMT which involves distinct changes in cellular morphology

and gene expression. Having shown that depletion of Rac1b enhanced TGF-β1-induced cell migration while its ectopic overexpression partially prevented downregulation of E-cadherin (see Fig. 

3B), we considered the possibility that Rac1b depletion also enhances EMT induced by TGF-β1. To study this in more detail, we again depleted Panc1 cells of cellular Rac1b protein by RNAi and

subsequently exposed the cells for 48 h to TGF-β1. Interestingly, Rac1b depletion enhanced the ability of TGF-β1 to induce cell scattering and the appearance of cells with a spindle-shaped

morphology (Fig. 4A). To analyse if these changes correspond to alterations in cell adhesion molecules or transcription factors orchestrating the EMT process, we performed immunoblot

analysis for E-cadherin and Snail, two proteins that are down- and upregulated, respectively, during TGF-β-induced EMT. Strikingly, E-cadherin and Snail protein levels in control

siRNA-transfected cells decreased and increased, respectively, by TGF-β1 treatment (Fig. 4B). Upon Rac1b depletion E-cadherin levels were dramatically reduced independent of TGF-β while

those of Snail were strongly increased only in TGF-β1-treated cells (Fig. 4B, lanes 4–6). Interestingly, cells could be rescued from the Rac1b effect when transfected with an siRNA directed

against both Rac1b and Rac1 (Fig. 4B, lanes 7–9). We then employed qPCR analysis to clarify whether Rac1b regulation of E-Cadherin and Snail as well as other EMT/migration-associated TGF-β

target genes was evident at the transcriptional level. To this end, Rac1b regulation of E-cadherin and Snail mRNA essentially mirrored the immunoblot data (Fig. 4C). A similar regulatory

pattern as for Snail mRNA was also seen for Slug and ZEB-1 mRNA. Rac1b depletion sensitised Panc1 cells to Slug and ZEB-1 mRNA induction by TGF-β1 that was higher and lower, respectively,

than for Snail (Fig. 4C). Likewise, mRNA of matrix metalloproteinase 9 (MMP9) and plasminogen activator-inhibitor type I (PAI-1), two proteins involved in TGF-β-induced invasion of PDAC

cells33, were both markedly upregulated after Rac1b knockdown while PAI-1 but not MMP9 mRNA levels in Rac1b-depleted cells were further enhanced by concomitant stimulation with rhTGF-β1 for

24 h (Fig. 4C). INHIBITION OF P38 AND ERK1/2 MAPK SIGNALLING PREVENTS TGF-Β1 DEPENDENT HYPEREXPRESSION OF EMT MARKER EXPRESSION AFTER RAC1B SILENCING As shown above, Rac1b knockdown

sensitised the Snail, Slug and PAI-1 genes in Panc1 cells to hyperinduction by TGF-β1 which may in part reflect a relief from Rac1b inhibition of Smad2/3 C phosphorylation9. However,

TGF-β1-mediated control of EMT and cell motility in PDAC cells does not only require Smad, but also crosstalk with p38 MAPK - and in pancreatic cancer cells with activating K-Ras mutations -

with the Ras-Raf-MEK-ERK-signalling cascade. We therefore analysed if the dramatic increase in gene expression of Snail, Slug, PAI-1, and MMP-9 after Rac1b removal was sensitive to

inhibition of p38 and/or ERK1/2 MAPK signalling. Panc1 cells depleted of Rac1b were treated or not with TGF-β1 in the presence or absence of the specific p38 MAPK inhibitor SB203580 (10 μM),

the MEK inhibitor U0126 (10 μM), or the ALK5 inhibitor SB431542 (5 μM) as control. Results show that in Rac1b-depleted cells both SB203580, U0126, and SB431542 as control, were able to

reduce the TGF-β1 effect on Snail mRNA (Fig. 5A) and Snail protein (Supplementary Fig. S4) expression, while the stimulatory effect of Rac1b depletion on TGF-β1-induced Slug was only

sensitive to SB203580 and SB431542 (Fig. 5A). In contrast, TGF-β1/Rac1b siRNA-mediated hyperstimulation of PAI-1 was not affected by SB203580 but was sensitive to U0126 and SB431542 (data

not shown). TGF-β-ALK5 signalling to _MMP9_ has been reported to be dependent on MEK-ERK but not JNK, p38 or Smad434. Interestingly, MMP9 mRNA upregulation after Rac1b siRNA transfection was

completely prevented by U0126 in both control and Rac1b-depleted cells, while SB203580 and SB431542 had no major effect (Fig. 5A). We then strived to confirm the results from pharmacologic

inhibition by ectopic expression of MKK6KA, a dominant negative mutant of the p38 upstream activator MKK6. In agreement with the data from pharmacological inhibition, Rac1b siRNA + 

TGF-β1-induced hyperexpression of Snail and Slug was strongly reduced in both MKK6KA clones compared to empty vector expressing control cells. In contrast and in agreement with the small

molecule inhibition data (see Fig. 5A), Rac1b siRNA + TGF-β1-induced upregulation of PAI-1 mRNA (not shown) and MMP9 mRNA (Fig. 5B) in MKK6KA clones did not differ significantly from vector

control cells. As expected, the stimulatory effect of TGF-β1 on Snail and Slug expression in both MKK6KA and empty vector expressing Panc1 cells was blunted upon transfection with an ALK5

siRNA (Fig. 5B). In summary, hyperinduction of TGF-β target gene expression by Rac1b depletion in combination with TGF-β1 treatment was differentially relieved by various inhibitors

reflecting involvement of the specific non-Smad/MAPK pathways involved. While hyperinduction of Snail mRNA was blocked by inhibition of either MKK6-p38 or MEK-ERK signalling, Rac1b

siRNA-mediated overexpression of Slug mRNA was only blocked by inhibition of MKK6-p38 signalling and that of PAI-1 and MMP9 mRNA only by inhibition of MEK-ERK signalling. INHIBITION OF P38

OR ERK1/2 MAPK SIGNALLING ABOLISHES THE PROMIGRATORY EFFECT OF RAC1B SILENCING Cell migration and invasion are considered EMT-associated processes that are controlled by both Smad and

non-Smad, _e.g_. MAPK signalling. We therefore hypothesised that Rac1b exerts its antimigratory/antiinvasive effects via inhibition of either the MKK6-p38 and/or the MEK-ERK pathway. When

Rac1b-depleted cells were subjected to xCELLigence® technology-based real-time cell migration assays, the promigratory effect of TGF-β1 was dramatically enhanced (Fig. 6A, green _vs_. red

curve). Notably, treatment of cells with SB203580 during the assay reduced the TGF-β1-induced migration of the Rac1b siRNA-transfected cells at the 24 h time point by 59 ± 20.5% (n = 3, p 

< 0.05) compared to siRac1b-transfected cells treated with vehicle (Fig. 6A, green _vs_. magenta curve). We then repeated the migration assays in MKK6KA expressing Panc1 cells. The

relative migratory activity of TGF-β1-treated Rac1b siRNA-transfected MKK6KA expressing clones #11 and #12 at the 24 h time point was reduced by 60 ± 8% (n = 3, p < 0.05) and 46 ± 28% (n 

= 3, p < 0.05), respectively, compared to Rac1b siRNA-transfected empty vector expressing control cells (Fig. 6B, green _vs_. magenta curve. Note that for the sake of clarity only the

curves for TGF-β1-treated cells are shown). The MAPKs ERK1/2 are also crucial in TGF-β-dependent EMT and cell motility of PDAC-derived cells18. Interestingly, the MEK inhibitor U0126

completely blocked the pro-migratory TGF-β1 effect in Rac1b-depleted cells (Fig. 6C, note that only the curves for TGF-β1-treated cells are shown). In IMIM-PC1 cells, Rac1b depletion also

enhanced the TGF-β1 effect on random cell migration only in vehicle treated cells (Fig. 6D, green _vs_. magenta curve, significant at 4 h and all later time points) but not in U0126-treated

cells (Fig. 6D, red _vs_. blue curve, no significant difference at any time point). Together, these data indicate that Rac1b depletion-induced enhancement of the promigratory TGF-β1 effect

in PDAC cells could be blocked by inhibition of either p38 or ERK1/2 signalling. RAC1B DEPLETION ENHANCES TGF-Β1-INDUCED GROWTH ARREST Data so far indicate that Rac1b through inhibiting p38

and ERK activation suppresses EMT-associated changes in morphology, gene expression, and cell motility. Another prominent response to TGF-β, growth inhibition, has also been implicated in

TGF-β-induced EMT and according to our hypothesis should also be under negative control by Rac1b. Preliminary evidence for this came from the microscopy and cell counting experiments in

which it was observed that cultures treated with a combination of Rac1b siRNA and TGF-β1 appeared more sparse (see Fig. 4A). To this end, siRNA-mediated depletion of Rac1b resulted a much

stronger growth inhibitory effect of TGF-β1 in both Panc1 cells and in the non-tumourigenic pancreatic ductal epithelial cell line HPDE6c7 as measured by DNA synthesis via [3H]-thymidine

incorporation (Fig. 7A). Interestingly, the greater ability of TGF-β1 to inhibit DNA synthesis in Rac1b-depleted Panc1 cells correlated with elevated protein levels of the cyclin-dependent

kinase inhibitor p21WAF1 (Fig. 7B). Together, these data show that Rac1b acts as an inhibitor of TGF-β1-induced growth inhibition. RAC1B DEPLETION ALTERS THE EXPRESSION OF POSITIVE AND

NEGATIVE REGULATORS OF TGF-Β SIGNALLING TOWARDS AN INHIBITORY OUTCOME Data so far indicate that Rac1b is a potent inhibitor of TGF-β/ALK5-induced EMT, expression of EMT/migration-associated

genes as well as activation of Smad2/39 and p38 and ERK MAPKs (this study). Rac1b might thus control the TGF-β pathway in a more direct and subtle fashion, _i.e_. by regulating the

expression of the TGF-β receptor(s) or TGF-β ligand(s). Strikingly, we consistently observed an upregulation of ALK5 protein abundance following silencing of Rac1b (Fig. 1C: ectopically

expressed ALK5KR, lanes 7–12, and Fig. 2D: endogenous ALK5, lanes 1 and 2 vs. lanes 3 and 4). A more thorough and quantitative analysis at both the RNA and protein levels confirmed this

observation (Fig. 8A). Moreover, TGF-β1 exhibits autoinduction of its expression (Fig. 8B) and, likewise, the TGF-β1-dependent levels of TGF-β1 mRNA (Fig. 8A, graph) and protein (Fig. 8B,

immunoblot) were further enhanced upon Rac1b depletion. Of interest, PAR2 was dramatically increased when Rac1b depletion was combined with TGF-β1 exposure (Fig. 8C) and these data were in

line with impaired TGF-β1 induction of PAR2 mRNA in HA-Rac1b overexpressing cells (see Fig. 3B). Above, we observed a greatly prolonged period of p38 activation in response to TGF-β1

stimulation in Rac1b-depleted cells which might be explained by defective termination of TGF-β signalling. We therefore tested whether Rac1b also affects the expression of negative

regulators of TGF-β signalling such as the inhibitory Smad, Smad7. Smad7 terminates excess signalling activity of activated ALK5 by preventing the binding of Smad2/3 and by enhancing

internalization and ubiquitin-mediated degradation of ALK535. Intriguingly, Rac1b depletion resulted in a _decrease_ in Smad7 expression in TGF-β1-treated cells (Fig. 8D), suggesting that

Rac1b, in addition to inhibiting initiation of TGF-β signalling, also favors its termination through Smad7 upregulation. Activation of the TGF-β signalling pathway is primarily determined by

the abundance, bioavailability, and access of the TGF-β ligand to the receptors. Hence, cells may protect themselves from overstimulation by soluble TGF-β by secreting factors such as BGN

that can bind and sequester TGF-β, prevent its binding to TβRII/ALK5 and thus neutralise its biological activity36. Given this function, we hypothesised that i) BGN is _positively_

controlled by Rac1b and if so ii) down-regulation of BGN should be able to mimic the Rac1b siRNA effect and amplify TGF-β1-induced reporter gene activity and cell migration. To test this

hypothesis, we chose Panc1 cells because BGN expression is dramatically induced by TGF-β1 in these cells within a period of 24 h26. Strikingly, and as shown above for Smad7, BGN induction

was strongly _reduced_ after Rac1b silencing in both untreated control cells and in cells treated for 24 h with TGF-β1 (Fig. 8E). Moreover, silencing _BGN_ with a specific siRNA was

associated with enhanced TGF-β/Smad transcriptional activity of the TGF-β-dependent luciferase reporter gene p3TP-Lux (Fig. 8F). Moreover, in both untreated and TGF-β1-treated Panc1 cells

BGN depletion caused a marked _increase_ in their migratory activity (Fig. 8G). Together, these findings suggest that Rac1b interferes with TGF-β pathway activation in several ways, _e.g_.

by inhibiting the expression of positive regulators (ALK5, TGF-β1, PAR2) and by promoting the expression of negative ones such as Smad7 and BGN. DISCUSSION We have shown previously that

Rac1b interfered with TGF-β1-dependent activation of Smad2/3 and random cell migration9. Since TGF-β induction of cell motility in PDAC-derived cells involves Smad and non-Smad signalling

pathways, which may be activated independently from each other, it was of interest to reveal whether TGF-β1-induced p38 and/or ERK signalling, too, are subject to (negative) control by

Rac1b. Notably, Rac1b depletion in Panc1 cells in a Smad4-independent manner prevented the decline in p-p38 levels seen at 4 h of TGF-β1 stimulation and extended the duration of the p38

phosphorylation response to TGF-β1 to 12 h. Of note, this time course correlated closely with the kinetics of the migratory response of this cell line to TGF-β1 (see Fig. 6A,B). A similar

albeit less pronounced effect of Rac1b depletion was seen in Colo357 cells after 4 and 8 h of TGF-β stimulation but not earlier which may be due to less effective inhibition of Rac1b protein

(due to a generally lower transfection efficiency) in this cell line when compared to Panc1 cells (see Supplementary Fig. S1). In HaCaT cells, Rac1b depletion had no effect on basal p-p38

levels but potentiated the TGF-β1-induced p-p38 levels at the 1 h time point (see Fig. 1B). Activation of ERK1/2 by TGF-β1 in PDAC cells is somewhat variable and can comprise an early and/or

a late activation episode depending on the cell type18. However, Rac1b depletion led to a strong increase in basal p-ERK1/2 levels and this was further enhanced by rhTGF-β1 treatment (see

Fig. 2A,B). In contrast to Panc1 cells, a rapid activation (within 1 h) is seen in both control and Rac1b siRNA-transfected Colo357 and IMIM-PC1 cells (see Fig. 2B, compare lanes 1 and 3),

and in HaCaT cells (see Supplementary Fig. S2). The strong ERK activation upon RNAi-mediated downregulation of Rac1b protein remained unaltered in Panc1 cells ectopically expressing

ALK5K232R (see Fig. 2C), transfected with ALK5 siRNA (Fig. 2D), or treated with the ALK5 inhibitor SB431542 (not shown), suggesting that the Rac1b siRNA-induced upregulation of p-ERK1/2

levels is independent of the TGF-β/ALK5 pathway. In contrast, the additional increase in ERK1/2 phosphorylation elicited by rhTGF-β1 in Rac1b-depleted cells was blunted by ALK5 protein

depletion (see Fig. 2D). In addition, we analysed the effects of ectopic overexpression of HA-Rac1b in individual clones of Panc1 cells. In two independent clones we observed a _reduced_

ability of TGF-β1 to induce ERK1/2 activation relative to empty vector controls (see Fig. 3). Hence, our data so far point to a crucial role of Rac1b in negative control of TGF-β1-dependent

activation of p38 and ERK in PDAC and non-PDAC cells. The ability of Rac1b to effectively suppress non-Smad TGF-β signalling may underlie its potent anti-EMT and antimigratory activity.

Interestingly, a negative effect of Rac1b on MEK-ERK signalling has also been observed upon neurotrophin 3 (NT3) stimulation of human bone marrow-derived stromal cells37. A major issue of

this study was whether selectively inhibiting the generation of Rac1b protein would affect the cells’ response to TGF-β1-induced EMT. Notably, Rac1b depletion enhanced the ability of TGF-β1

to induce morphological changes such as a spindle-shaped morphology and altered the expression of various epithelial and mesenchymal EMT marker genes. Specifically, Rac1b depletion alone

(without addition of rhTGF-β1) dramatically enhanced downregulation of E-cadherin and upregulation of Slug, MMP9, and PAI-1 expression. Moreover, cells depleted of Rac1b protein reacted with

a much stronger induction by rhTGF-β1 of Snail, Slug, ZEB-1, and PAI-1 mRNA. For E-cadherin and Snail we confirmed this regulatory pattern at the protein level (Fig. 4B). Taken together

with the inhibitory effect of Rac1b on TGF-β1-dependent Smad2/3 activation9, the data implicate Rac1b as a protein that helps to maintain an epithelial phenotype thereby protecting

PDAC-derived cells from TGF-β1-induced mesenchymal conversion. Using pharmacological and dominant-negative inhibition strategies, we show that blocking p38 signalling partially relieved the

Rac1b siRNA-induced hypererinduction by TGF-β1 for Snail and Slug, but not PAI-1 and MMP9 (see Fig. 5). Blocking Ras-Raf-MEK-ERK signalling with U0126 completely or partially relieved the

Rac1b siRNA-induced hypererinduction by TGF-β1 for Snail, PAI-1, and MMP9, respectively, but not Slug. With respect to the signalling pathways involved, regulation of MMP9 deserves

particular attention since TGF-β-ALK5 signalling to _MMP9_ is dependent solely on MEK-ERK but not JNK, p38 or Smad434. This has been confirmed here by the dramatic upregulation of MMP9 mRNA

after Rac1b siRNA transfection (Figs 4C and 5A) and its almost complete relief by U0126 in both control and Rac1b-depleted Panc1 cells (Fig. 5A) and by the failure of MKK6KA expression (Fig.

 5B), SB203580 and SB431542 (Fig. 5A) to mimic the U0126 effect. The failure of SB431542 to decrease the MMP9 mRNA levels in Rac1b-depleted and TGF-β1-treated cells is likely due to the fact

that ERK activation by TGF-β is mediated by the tyrosine kinase rather than the serine/threonine kinase function of ALK521. Our data on p38 involvement in TGF-β regulation of E-cadherin,

Snail, and PAI-1 are in agreement with published data38,39. Moreover, we revealed differences in TGF-β-MAPK signalling to Snail and Slug in PDAC cells. Our data show that by inhibiting p38

and ERK signalling, Rac1b can suppress the response of critical regulators of TGF-β-induced EMT program. To confirm the important role of non-Smad-mediated TGF-β signalling in Rac1b

regulation of EMT, we additionally evaluated the effects of the above mentioned p38 and ERK inhibitors on TGF-β1-stimulated cell migration. In agreement with the gene expression data, the

strong increase in TGF-β1-induced migratory activity resulting from Rac1b depletion could be relieved by treatment of cells with SB203580 or U0126, or by ectopic expression of a

dominant-negative mutant of MKK6. EMT program has been associated with enhanced growth inhibition40 which contributes to the known chemo- and radioresistant phenotype of PDAC cells.

Interestingly, Rac1b appears to act as an inhibitor of TGF-β1-dependent growth arrest in malignant and benign pancreatic ductal epithelial cells (see Fig. 7A), probably by suppressing the

expression of p21WAF1, a potent cell cycle inhibitor in PDAC cells41. This matches the role of Rac1b as an EMT inhibitor and suggests the exciting possibility that Rac1b can be utilized as a

chemo-/radiosensitzer in targeted therapies for PDAC. In this context it is noteworthy that inhibition of TGF-β signalling has been discussed as a potential strategy to improve success of

radiotherapy42 and of novel therapies based on the use of TRAIL43. Prompted by the negative effects of Rac1b on EMT-associated gene expression, migration, and growth inhibition as well as on

Smad and MAPK signaling, we pursued the idea that Rac1b targets central components of the TGF-β signalling pathway. Intriguingly, Rac1b appears to restrict expression of TGF-β1 ligand,

ALK5, and PAR2, a G protein-coupled receptor that was shown recently by us to be indispensable for Smad activation and TGF-β1-dependent cell motility28. While the TGF-β1, ALK5 and PAR2 genes

are all under negative control by Rac1b, two endogenous inhibitors of TGF-β signalling, namely Smad7 and BGN, turned out to be positively regulated by Rac1b reinforcing the concept of Rac1b

being an effective inhibitor of TGF-β signalling. The observation that ALK5 protein was more abundant in Rac1-depleted cells is particularly interesting since altering receptor expression

is a prominent mechanism through which tumour cells can modulate their sensitivity to TGF-β44. PDAC cells, particularly Panc1 and to a lower extent Colo357 and IMIM-PC1 cells, are known to

secrete large amounts of TGF-β1 into the culture medium and to autostimulate themselves31. Since in Panc1 cells exposed to rhTGF-β1 Rac1b controls TGF-β1 mRNA and protein expression in a

negative fashion, its depletion is expected to enhance this autocrine loop by relieving TGF-β-Smad and MAPK signalling from inhibition. The presence of an autocrine feedback loop is also

suggested by the observation that the p-p38 levels in both non-rhTGF-β1 stimulated control siRNA and Rac1b siRNA-transfected cells were reduced upon dominant-negative inhibition of the ALK5

kinase (see Fig. 1C) and siRNA-mediated depletion of endogenous TGF-β1 (see Fig. 1E). With respect to its possible role in tumour progression, our data suggest that Rac1b is required for

maintaining an epithelial phenotype by preventing TGF-β1 from inducing EMT and thus a mesenchymal and potentially invasive phenotype in the tumour cells. These results provide a molecular

correlate for the antimetastatic function proposed earlier by us on the basis of higher expression of Rac1b in long-time _vs_. short-time survivors among PDAC patients9. Crosstalk of TGF-β

with K-Ras signalling has been shown to be central to tumourigenesis of PDAC1, however, therapeutically targetting K-Ras may not be feasible as available Ras inhibitors have largely failed

to block the protumourigenic effects of oncogenic K-Ras. Since Rac1 is a well-known downstream target of Ras and mediator of TGF-β1-induced EMT and cell motility, our observation that Rac1b

can antagonise Rac1 function and help to maintain an epithelial phenotype in PDAC-derived cells is intriguing. Therapeutically increasing the generation of Rac1b over that of Rac1 in the

tumour tissue, _e.g_. by shifting the splice ratio45 could be a means to block some undesired effects of hyperactive Ras and may represent a promising strategy for alleviating the

protumourigenic effects of TGF-β. REFERENCES * Ijichi, H. _et al_. Aggressive pancreatic ductal adenocarcinoma in mice caused by pancreas-specific blockade of transforming growth factor-beta

signaling in cooperation with active Kras expression. _Genes Dev._ 20, 3147–60 (2006). Article  CAS  PubMed  PubMed Central  Google Scholar  * Heid, I. _et al_. Early requirement of Rac1 in

a mouse model of pancreatic cancer. _Gastroenterology_ 141, 719–730 (2011). Article  CAS  PubMed  Google Scholar  * Jordan, P., Brazao, R., Boavida, M. G., Gespach, C. & Chastre, E.

Cloning of a novel human Rac1b splice variant with increased expression in colorectal tumors. _Oncogene_ 18, 6835–6839 (1999). Article  CAS  PubMed  Google Scholar  * Schnelzer, A. _et al_.

Rac1 in human breast cancer: overexpression, mutation analysis, and characterization of a new isoform, Rac1b. _Oncogene_ 19, 3013–3020 (2000). Article  CAS  PubMed  Google Scholar  *

Haeusler, L. C. _et al_. Purification and biochemical properties of Rac1, 2, 3 and the splice variant Rac1b. _Methods Enzymol._ 406, 1–11 (2006). Article  CAS  PubMed  Google Scholar  *

Matos, P., Collard, J. G. & Jordan, P. Tumor-related alternatively spliced Rac1b is not regulated by Rho-GDP dissociation inhibitors and exhibits selective downstream signaling. _J Biol

Chem._ 278, 50442–50448 (2003). Article  CAS  PubMed  Google Scholar  * Radisky, D. C. _et al_. Rac1b and reactive oxygen species mediate MMP-3-induced EMT and genomic instability. _Nature_

436, 123–127 (2005). Article  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Ungefroren, H. _et al_. Differential roles of Smad2 and Smad3 in the regulation of TGF-β1-mediated growth

inhibition and cell migration in pancreatic ductal adenocarcinoma cells: control by Rac1. _Mol Cancer_ 10, 67 (2011). Article  CAS  PubMed  PubMed Central  Google Scholar  * Ungefroren, H.

_et al_. Rac1b negatively regulates TGF-β1-induced cell motility in pancreatic ductal epithelial cells by suppressing Smad signalling. _Oncotarget_ 5, 277–290 (2014). Article  PubMed  Google

Scholar  * Derynck, R., Muthusamy, B. P. & Saeteurn, K. Y. Signaling pathway cooperation in TGF-β-induced epithelial-mesenchymal transition. _Curr Opin Cell Biol._ 31, 56–66 (2014).

Article  CAS  PubMed  PubMed Central  Google Scholar  * Heldin, C. H., Vanlandewijck, M. & Moustakas, A. Regulation of EMT by TGFβ in cancer. _FEBS Lett._ 586, 1959–1970 (2012). Article

  CAS  PubMed  Google Scholar  * Hoot, K. E. _et al_. Keratinocyte-specific Smad2 ablation results in increased epithelial-mesenchymal transition during skin cancer formation and

progression. _J Clin Invest._ 118, 2722–2732 (2008). CAS  PubMed  PubMed Central  Google Scholar  * Bakin, A. V., Rinehart, C., Tomlinson, A. K. & Arteaga, C. L. p38 mitogen-activated

protein kinase is required for TGFbeta-mediated fibroblastic transdifferentiation and cell migration. _J Cell Sci._ 115, 3193–3206 (2002). CAS  PubMed  Google Scholar  * Imamichi, Y. _et

al_. TGF beta-induced focal complex formation in epithelial cells is mediated by activated ERK and JNK MAP kinases and is independent of Smad4. _Biol Chem._ 386, 225–236 (2005). Article  CAS

  PubMed  Google Scholar  * Mu, Y., Gudey, S. K. & Landström, M. Non-Smad signaling pathways. _Cell Tissue Res._ 347, 11–20 (2012). Article  CAS  PubMed  Google Scholar  * Ungefroren,

H., Witte, D. & Lehnert, H. The role of small GTPases of the Rho/Rac family in TGF-β-induced EMT and cell motility in cancer. _Dev Dy_n. _Apr_ 8, https://doi.org/10.1002/dvdy.24505.

[Epub ahead of print] (2017) * Kolosova, I., Nethery, D. & Kern, J. A. Role of Smad2/3 and p38 MAP kinase in TGF-β1-induced epithelial-mesenchymal transition of pulmonary epithelial

cells. _J Cell Physiol_. 226, 1248–1254 (2011). * Ellenrieder, V. _et al_. Transforming growth factor beta1 treatment leads to an epithelial-mesenchymal transdifferentiation of pancreatic

cancer cells requiring extracellular signal-regulated kinase 2 activation. _Cancer Res_. 61, 4222-4228 (2001). * Sano, Y. _et al_. ATF-2 is a common nuclear target of Smad and TAK1 pathways

in transforming growth factor-β signaling. _J Biol Chem._ 274, 8949–8957 (1999). Article  CAS  PubMed  Google Scholar  * Yamashita, M. _et al_. TRAF6 mediates Smad-independent activation of

JNK and p38 by TGF-β. _Mol Cell_ 31, 918–924 (2008). Article  CAS  PubMed  PubMed Central  Google Scholar  * Lee, M. K. _et al_. TGF-beta activates Erk MAP kinase signalling through direct

phosphorylation of ShcA. _EMBO J._ 26, 3957–3967 (2007). Article  CAS  PubMed  PubMed Central  Google Scholar  * Liu, N., Furukawa, T., Kobari, M. & Tsao, M. S. Comparative phenotypic

studies of duct epithelial cell lines derived from normal human pancreas and pancreatic carcinoma. _Am J Pathol._ 153, 263–269 (1998). Article  CAS  PubMed  PubMed Central  Google Scholar  *

Ungefroren, H., Lenschow, W., Chen, W. B., Faendrich, F. & Kalthoff, H. Regulation of biglycan gene expression by transforming growth factor-beta requires MKK6-p38 mitogen-activated

protein kinase signaling downstream of Smad signaling. _J Biol Chem._ 278, 11041–11049 (2003). Article  CAS  PubMed  Google Scholar  * Raingeaud, J., Whitmarsh, A. J., Barrett, T., Dérijard,

B. & Davis, R. J. MKK3- and MKK6-regulated gene expression is mediated by the p38 mitogen-activated protein kinase signal transduction pathway. _Mol Cell Biol_. 16, 1247-1255 (1996). *

Ungefroren, H., Groth, S., Ruhnke, M., Kalthoff, H. & Fändrich, F. Transforming growth factor-beta (TGF-beta) type I receptor/ALK5-dependent activation of the GADD45beta gene mediates

the induction of biglycan expression by TGF-beta. _J Biol Chem._ 280, 2644–2652 (2005). Article  CAS  PubMed  Google Scholar  * Chen, W. B. _et al_. Smad4/DPC4-dependent regulation of

biglycan gene expression by transforming growth factor-beta in pancreatic tumor cells. _J Biol Chem._ 277, 36118–36128 (2002). Article  CAS  PubMed  Google Scholar  * Mandel, K. _et al_.

Characterization of spontaneous and TGF-β-induced cell motility of primary human normal and neoplastic mammary cells _in vitro_ using novel real-time technology. _PLoS One_ 8, e56591 (2013).

Article  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Zeeh, F. _et al_. Proteinase-activatedreceptor 2 promotes TGF-β-dependent cell motility in pancreatic cancer cells by sustaining

expression of the TGF-β type I receptor ALK5. _Oncotarget_ 7, 41095–41109 (2016). Article  PubMed  PubMed Central  Google Scholar  * Groth, S., Schulze, M., Kalthoff, H., Fandrich, F. &

Ungefroren, H. Adhesion and Rac1-dependent regulation of biglycan gene expression by transforming growth factor-beta. Evidence for oxidative signaling through NADPH oxidase. _J Biol Chem._

280, 33190–33199 (2005). Article  CAS  PubMed  Google Scholar  * Takekawa, M. _et al_. Smad-dependent GADD45beta expression mediates delayed activation of p38 MAP kinase by TGF-beta. _EMBO

J._ 21, 6473–6482 (2002). Article  CAS  PubMed  PubMed Central  Google Scholar  * Carl, C. _et al_. Ionizing radiation induces a motile phenotype in human carcinoma cells _in vitro_ through

hyperactivation of the TGF-beta signaling pathway. _Cell Mol Life Sci._ 73, 427–443 (2016). Article  CAS  PubMed  Google Scholar  * Ramachandran, R. & Hollenberg, M. D. Proteinases and

signalling: pathophysiological and therapeutic implications via PARs and more. _Br J Pharmacol._ 153(Suppl 1), S263–282 (2008). CAS  PubMed  Google Scholar  * Ellenrieder, V. _et al_.

TGF-beta-induced invasiveness of pancreatic cancer cells is mediated by matrix metalloproteinase-2 and the urokinase plasminogen activator system. _Int J Cancer_ 93, 204–211 (2001). Article

  CAS  PubMed  Google Scholar  * Safina, A., Vandette, E. & Bakin, A. V. ALK5 promotes tumor angiogenesis by upregulating matrix metalloproteinase-9 in tumor cells. _Oncogene._ 26,

2407–22 (2007). Article  CAS  PubMed  Google Scholar  * Yan, X. & Chen, Y. G. Smad7: not only a regulator, but also a cross-talk mediator of TGF-β signalling. _Biochem J._ 434, 1–410

(2011). Article  CAS  PubMed  Google Scholar  * Droguett, R., Cabello-Verrugio, C., Riquelme, C. & Brandan, E. Extracellular proteoglycans modify TGF-beta bio-availability attenuating

its signaling during skeletal muscle differentiation. _Matrix Biol._ 25, 332–341 (2006). Article  CAS  PubMed  Google Scholar  * Curtis, K. M., Gomez, L. A. & Schiller, P. C. Rac1b

regulates NT3-stimulated Mek-Erk signaling, directing marrow-isolated adult multilineage inducible (MIAMI) cells toward an early neuronal phenotype. _Mol Cell Neurosci._ 49, 138–148 (2012).

Article  CAS  PubMed  Google Scholar  * Strippoli, R. _et al_. p38 maintains E-cadherin expression by modulating TAK1-NF-kappa B during epithelial-to-mesenchymal transition. _J Cell Sci._

123, 4321–4331 (2010). Article  CAS  PubMed  Google Scholar  * Liu, R. M. _et al_. Postlethwait EM. Oxidative modification of nuclear mitogen-activated protein kinase phosphatase 1 is

involved in transforming growth factor beta1-induced expression of plasminogen activator inhibitor 1 in fibroblasts. _J Biol Chem._ 285, 16239–16247 (2010). Article  CAS  PubMed  PubMed

Central  Google Scholar  * Zheng, X. _et al_. Epithelial-to-mesenchymal transition is dispensable for metastasis but induces chemoresistance in pancreatic cancer. _Nature_ 527, 525–530

(2015). Article  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Datto, M. B., Yu, Y. & Wang, X. F. Functional analysis of the transforming growth factor beta responsive elements in

the WAF1/Cip1/p21 promoter. _J Biol Chem._ 270, 28623–28628 (1995). Article  CAS  PubMed  Google Scholar  * Du, S. & Barcellos-Hoff, M. H. Tumors as organs: biologically augmenting

radiation therapy by inhibiting transforming growth factor β activity in carcinomas. _Semin Radiat Oncol._ 23, 242–251 (2013). Article  PubMed  PubMed Central  Google Scholar  * Radke, D. I.

_et al_. Negative control of TRAIL-R1 signaling by transforming growth factor β1 in pancreatic tumor cells involves Smad-dependent down regulation of TRAIL-R1. _Cell Signal._ 28, 1652–1662

(2016). Article  CAS  PubMed  Google Scholar  * Herhaus, L. & Sapkota, G. P. The emerging roles of deubiquitylating enzymes (DUBs) in the TGFβ and BMP pathways. _Cell Signal._ 26,

2186–2192 (2014). Article  CAS  PubMed  PubMed Central  Google Scholar  * Gonçalves, V. _et al_. Phosphorylation of SRSF1 by SRPK1 regulates alternative splicing of tumor-related Rac1b in

colorectal cells. _RNA_ 20, 474–482 (2014). Erratum in: _RNA_ 22, 166. Henriques, Andreia [corrected to Henriques, Andreia F A]; Pereira, Joana [corrected to Pereira, Joana F S] (2016).

Download references ACKNOWLEDGEMENTS We are indebted to H. Albrecht and S. Grammerstorf-Rosche for excellent technical assistance and Dr. A. Menke (University of Giessen, Germany) for kindly

donating IMIM-PC1 cells. AUTHOR INFORMATION AUTHORS AND AFFILIATIONS * First Department of Medicine, University Hospital Schleswig-Holstein (UKSH), Campus Lübeck, and University of Lübeck,

23538, Lübeck, Germany David Witte, Hannah Otterbein, Maria Förster, Hendrik Lehnert & Hendrik Ungefroren * Signal Transduction of Cellular Motility, Internal Medicine V,

Justus-Liebig-University Giessen, 35392, Giessen, Germany Klaudia Giehl * Department of Hematology and Oncology, Freiburg University Medical Center, Albert-Ludwigs-University, 79106,

Freiburg i.Br., Germany Robert Zeiser * Department of General and Thoracic Surgery, UKSH, Campus Kiel, 24105, Kiel, Germany Hendrik Ungefroren Authors * David Witte View author publications

You can also search for this author inPubMed Google Scholar * Hannah Otterbein View author publications You can also search for this author inPubMed Google Scholar * Maria Förster View

author publications You can also search for this author inPubMed Google Scholar * Klaudia Giehl View author publications You can also search for this author inPubMed Google Scholar * Robert

Zeiser View author publications You can also search for this author inPubMed Google Scholar * Hendrik Lehnert View author publications You can also search for this author inPubMed Google

Scholar * Hendrik Ungefroren View author publications You can also search for this author inPubMed Google Scholar CONTRIBUTIONS D.W., M.F., H.O., and H.U. performed experiments and data

analysis. D.W. and H.U. supervised the study design and tests and drafted the manuscript. K.G., R.Z., and H.L. critically read and edited the manuscript. H.U. designed and directed the

study. All authors read and approved the final manuscript. CORRESPONDING AUTHOR Correspondence to Hendrik Ungefroren. ETHICS DECLARATIONS COMPETING INTERESTS The authors declare that they

have no competing interests. ADDITIONAL INFORMATION PUBLISHER'S NOTE: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional

affiliations. ELECTRONIC SUPPLEMENTARY MATERIAL SUPPLEMENTARY FIGURES AND TABLES RIGHTS AND PERMISSIONS OPEN ACCESS This article is licensed under a Creative Commons Attribution 4.0

International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the

source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative

Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by

statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

http://creativecommons.org/licenses/by/4.0/. Reprints and permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Witte, D., Otterbein, H., Förster, M. _et al._ Negative regulation of

TGF-β1-induced MKK6-p38 and MEK-ERK signalling and epithelial-mesenchymal transition by Rac1b. _Sci Rep_ 7, 17313 (2017). https://doi.org/10.1038/s41598-017-15170-6 Download citation *

Received: 04 May 2017 * Accepted: 23 October 2017 * Published: 11 December 2017 * DOI: https://doi.org/10.1038/s41598-017-15170-6 SHARE THIS ARTICLE Anyone you share the following link with

will be able to read this content: Get shareable link Sorry, a shareable link is not currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt

content-sharing initiative