MCB
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lee, M.-S.
Right arrow Articles by Lee, J. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lee, M.-S.
Right arrow Articles by Lee, J. W.
Molecular and Cellular Biology, August 2005, p. 6921-6936, Vol. 25, No. 16
0270-7306/05/$08.00+0     doi:10.1128/MCB.25.16.6921-6936.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.

The Signaling Network of Transforming Growth Factor ß1, Protein Kinase C{delta}, and Integrin Underlies the Spreading and Invasiveness of Gastric Carcinoma Cells

Mi-Sook Lee,1 Tae Young Kim,1 Yong-Bae Kim,2 Sung-Yul Lee,1 Seong-Gyu Ko,2,{dagger} Hyun-Soon Jong,2 Tae-You Kim,1,2 Yung-Jue Bang,1,2 and Jung Weon Lee1,2*

Cancer Research Institute, Departments of Molecular and Clinical Oncology and,1 Tumor Biology, College of Medicine, Seoul National University, Seoul 110-799, Republic of Korea2

Received 23 November 2004/ Returned for modification 15 February 2005/ Accepted 24 May 2005


   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Integrin-mediated cell adhesion and spreading enables cells to respond to extracellular stimuli for cellular functions. Using a gastric carcinoma cell line that is usually round in adhesion, we explored the mechanisms underlying the cell spreading process, separate from adhesion, and the biological consequences of the process. The cells exhibited spreading behavior through the collaboration of integrin-extracellular matrix interaction with a Smad-mediated transforming growth factor ß1 (TGFß1) pathway that is mediated by protein kinase C{delta} (PKC{delta}). TGFß1 treatment of the cells replated on extracellular matrix caused the expression and phosphorylation of PKC{delta}, which is required for expression and activation of integrins. Increased expression of integrins {alpha}2 and {alpha}3 correlated with the spreading, functioning in activation of focal adhesion molecules. Smad3, but not Smad2, overexpression enhanced the TGFß1 effects. Furthermore, TGFß1 treatment and PKC{delta} activity were required for increased motility on fibronectin and invasion through matrigel, indicating their correlation with the spreading behavior. Altogether, this study clearly evidenced that the signaling network, involving the Smad-dependent TGFß pathway, PKC{delta} expression and phosphorylation, and integrin expression and activation, regulates cell spreading, motility, and invasion of the SNU16mAd gastric carcinoma cell variant.


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Integrin-mediated adhesion to extracellular matrix (ECM) proteins allows cells to efficiently respond to extracellular stimuli for spreading, proliferation, migration, invasion, and gene transcription. This response is mediated by bidirectional signal transduction between extracellular and intracellular spaces that cross talks with other signal pathways (2, 3, 15, 22). Integrins, a family of cell adhesion receptors, are composed of an {alpha} and a ß subunit. They transduce direct signaling via engagements with ECM proteins, leading to the regulation of downstream intracellular signaling molecules. They also function in collaborative (indirect) signaling, in which integrins cosignal with other membrane receptor-mediated signal pathways (e.g., growth factor receptors, G-protein coupled receptors, or the transforming growth factor ß1 [TGFß1] signaling pathway) (4, 8, 17, 25, 38, 43).

TGFß1 is a multifunctional cytokine which inhibits cell growth and also mediates cell differentiation and metastasis. Activation of TGFß1 receptor complex by TGFß1 binding propagates intracellular signal transduction, involving Smad proteins, to regulate numerous developmental and homeostatic processes via regulations in gene induction (1). Smad7 is a major inhibitory Smad, which inhibits the TGFß1-mediated phosphorylation of R-Smad2 and R-Smad3 through competition with Smad2/3 for binding to the TGFß1 receptor (29). Recently, TGFß1 was demonstrated to activate a variety of intracellular signaling molecules, including mitogen-activated protein kinases (MAPKs) (9, 12, 45) and small GTPases (28), either by Smad-dependent or -independent signaling pathways (7). TGFß1 signaling modulates the expression of ECM proteins (14, 34) and integrins (27). Conversely, integrin-mediated signaling also regulates TGFß1 expression levels (16, 21). Although this collaborative relationship between integrin- and TGFß1-mediated signal pathways appears to be important for diverse cellular functions, mechanistic details underlying their collaboration and signaling network are largely unknown.

Protein kinase C{delta} (PKC{delta}) is a member of a novel family of the PKC families and can be activated by either diacylglycerol or phorbol ester (44). PKC{delta} has been shown to exert antitumorigenic or tumorigenic effects, depending on the nature of cellular stimuli (37). Although PKC{delta} has been implicated in ECM synthesis, as shown in a couple of previous studies (13, 46), the evidence is not conclusive for its roles in TGFß1-mediated regulation of cell functions.

So far, signaling networks consisting of integrins, TGFß1, and PKC (especially PKC{delta}) have not been thoroughly investigated, especially for cell spreading and invasiveness. In this study, we have attempted to mechanistically explore the signaling networks which regulate the cell spreading process, separately from the adhesion process. A gastric carcinoma cell line that is usually round in adhesion was used, so that stimuli-induced spreading was investigated with regards to signal cross talks between TGFß1, integrin, and PKC. We observed the signaling network in which Smad-dependent TGFß1 signaling to integrin-mediated signaling is mediated by expression and activation of PKC{delta}, leading to cell spreading. Furthermore, we also investigated the biological consequences inherent in the signal network-mediated cell spreading.


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Cells. SNU16mAd cells, a variant cell line enriched with adherent cells, were obtained from subsequent cultures by collecting adherent cells among mostly anchorage-independent SNU16{alpha}5 cells (18). SNU16mAd cells were cultured at 37°C and 5% CO2, in RPMI 1640 culture media containing 10% (vol/vol) fetal bovine serum and 0.2 mg/ml G418.

Cell lysate preparation and Western blots. Replating of SNU16mAd cells on diverse ECM-precoated dishes (10 µg/ml fibronectin [Fn], 10 µg/ml collagen type I, 10 µg/ml vitronectin, 10 µg/ml laminin I [Chemicon, Temecula, CA], or 10 µg/ml poly-L-lysine [PL; Sigma]) was done as explained previously (25). In certain cases, pharmacological inhibitors (12.5 µM GF-109203X or 10 µM rottlerin [10] [Calbiochem, San Diego, CA]) were pretreated, 30 min prior to the replating without or with TGFß1 treatment. Upon replating, TGFß1 (5 ng/ml; Chemicon) was added directly to the replating media, and the treatment lasted for 20 h or indicated periods. In cases of experiments with protein synthesis inhibition, 12 h after the replating, certain cells were treated with 10 µg/ml cycloheximide (Sigma), a protein synthesis inhibitor, with or without a concomitant 5 ng/ml TGFß1, followed by additional incubation for 8 h (retreated every 4 h) for a total of 20 h of incubation on Fn. In certain cases, cells were premixed with 10 µg/ml anti-integrin {alpha}2 (P1E6), {alpha}3 (P1B5), or {alpha}5 (P1D6) antibodies (Chemicon), 30 min before the replating on Fn and a concomitant TGFß1 treatment for 20 h. In cases in which adenovirus for either LacZ, FLAG-tagged Smad2, Smad3, Smad7 (25), PKC{delta}, or dominant-negative PKC{alpha} (K368R mutant) (kind gifts from J.-S. Chun, Gwangju Institute of Science and Technology, Gwangju, Korea) was separately infected, 24 h after the infection, cells were replated on ECM without or with TGFß1 treatment for 20 h. In cases in which the TGFß1 treatment periods were shorter (see Fig. 3A, 4C, and 6C), the indicated periods (x h) from the end of a total of 20 h of incubation were with TGFß1, after certain incubations (20 – x h) without TGFß1. In the case of PKC{delta} knockout via introduction of small interfering RNA (siRNA) (QIAGEN), siRNA against either PKC{delta} (to target AAG ATG AAG GAG GCG CTC AG; QIAGEN, catalog no. 1022453) or its negative control (AAT TCT CCG AAC GTG TCA CGT; QIAGEN, catalog no. 1022079) was separately transfected using Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA), according to the manufacturer's protocols. The target sequence of PKC{delta} siRNA was unique for PKC{delta} according to NCBI BLAST searches. In cases of integrin {alpha} subunit overexpression, pSF2-human integrin {alpha}2, {alpha}3, or pcDM{alpha}5 (23) or pEF-PKC{theta} was separately transfected as above. Twenty-four hours after the transfection, cells were replated on either Fn-precoated dishes or cover glasses in the absence or presence of TGFß1 treatment for 20 h. Cell lysates were prepared as described in the previous studies (24, 25). The lysates were used in Western blots using phospho-Y397FAK, phospho-Y925FAK, phospho-Y416Src, PKC{alpha}, PKC{delta}, c-Src (Santa Cruz Biotechnology, Santa Cruz, CA), phospho-Y118paxillin, phospho-PKCs (Cell Signaling Technology, Beverly, MA), FLAG (Sigma), integrins {alpha}2, {alpha}3, or {alpha}4, human laminin 5 (P3H9-2 clone) (Chemicon), focal adhesion kinase (FAK), paxillin, p130Cas, Nck, {alpha}-tubulin, {alpha}5 (BD Transduction Laboratories, San Jose, CA), or type I collagen (Biodesign, Saco, ME). In some cases, the membrane was stripped by incubation in a stripping buffer (62.5 mM Tris, pH 6.8, 2% sodium dodecyl sulfate [SDS], and 100 mM ß-mercaptoethanol) at 65°C for 30 min, washed for 1 h (3 times for 20 min) with Tris-based saline with 0.05% Tween-20 (TBST), reblocked with TBST containing 1% bovine serum albumin (BSA) plus 1% skim milk proteins, and then reprobed with another primary antibody.



View larger version (54K):
[in this window]
[in a new window]
  		FIG. 3. Integrin- and TGFß1-mediated cell spreading on fibronectin involves phosphorylation of focal adhesion molecules. Manipulation of cells and preparation of cell lysates were done the same as for Fig. 2. Cell lysates were analyzed by Western blotting or indirect immunofluorescent microscopy using primary antibodies against the indicated molecules. The data shown are representative of several experiments. (A) TGFß1 treatment increased phosphorylation of the FA molecules and expression and Ser643 phosphorylation of PKC{delta}, in a time-dependent manner. Cells were treated with 5 ng/ml TGFß1 for the indicated hours at the end of the 20-h incubation on fibronectin (i.e., 1 h indicates that TGFß1 was directly added to the replating media after 19 h on Fn and the whole-cell extracts were prepared after one additional hour of incubation). (B) Integrin- and TGFß1-mediated phosphorylation of the FA molecules was significantly reduced by PKC inhibition. Cells were pretreated with the indicated PKC inhibitors (12.5 µM GF-109203X [GF] or 10 µM rottlerin [Rot]), 30 min prior to the replating on Fn. (C) Expression of FRNK (dominant-negative FAK inhibitor), inactive c-Src (Y416F c-Src), or dominant-negative paxillin (Y31/118/157F paxillin) abolished the cell spreading. Cells were cotransfected with FRNK and GFP, Y416F c-Src and GFP, or Y31/118/157F paxillin and GFP. Two days later, cells were immunostained with rabbit anti-pY397FAK and then anti-rabbit IgG-conjugated TRITC. (D) FAK-p130Cas and p130Cas-Nck interactions correlated with the TGFß1 effects. Cell lysates were subjected to immunoprecipitation with mouse monoclonal anti-FAK or Nck antibodies, and the immunoprecipitates were used in immunoblots using antibodies against the indicated molecules. The data shown were representative of three isolated experiments.

 


View larger version (46K):
[in this window]
[in a new window]
  		FIG. 4. TGFß1-, PKC-, and integrin-mediated cell spreading required new protein synthesis. (A) Inhibition of protein synthesis abolished phosphorylation of the FA molecules (left) and cell spreading (right) by TGFß1 treatment. Cell replating and a concomitant TGFß1 treatment were done, as described above. Twelve hours after the replating, certain cells were treated with 10 µg/ml cycloheximide every 4 h during an additional 8-h incubation with TGFß1 treatment at 37°C. Cell images were taken using a phase-contrast microscope, and whole-cell lysates were prepared and used in Western blots as explained earlier. Data shown were representative of three independent experiments. (B) TGFß1-mediated effects on integrins and ECM expression levels. Cell lysates were immunoblotted for antibodies against indicated integrins or collagen I or {alpha}3 chain of human laminin 5 (LN5-{alpha}3) (19, 35). Data shown were representative of three isolated experiments. (C) Increases in integrin {alpha}2 and {alpha}3 expression levels by TGFß1 in a time-dependent manner. Cells from the indicated conditions were analyzed for integrin {alpha}2 or {alpha}3 expression by flow cytometry. Histograms are shown for controls with no primary antibody (C) and TGFß1-treatment for 0 h (1), 8 h (2), 12 h (3), and 20 h (4). (D) Blockage of TGFß1-induced integrin {alpha}2 or {alpha}3 expression by PKC inhibition. Histograms included are for no primary antibody control (C), no TGFß1 treatment (1), TGFß1 treatment for 20 h (2), and GF-109203X pretreatment 30 min prior to TGFß1 treatment for 20 h (3). (E) The TGFß1-mediated effects on integrin and PKC{delta} inductions were reduced by Smad7 overexpression. Twenty-four hours after the infection of cells with adenovirus encoding for ß-galactosidase (LacZ) or Flag-Smad7, cells were replated on Fn in the absence or presence of TGFß1 treatment for 20 h. Cell lysates were prepared and used in Western blots with antibodies against the indicated molecules. Data shown were representative of three independent experiments.

 






View larger version (336K):
[in this window]
[in a new window]
  		FIG. 6. Integrin- and TGFß1-mediated cell spreading requires PKC{delta} expression and activation. (A) Infection with adenovirus for PKC{delta} WT (Ad-PKC{delta}) increased the expression and Ser643 phosphorylation level of PKC{delta}. Cells were infected with various amounts of Ad-PKC{delta}. Two days later, cell lysates were prepared and used in Western blots for the indicated molecules. Data shown were representative of three independent experiments. (B) Increased activity and expression level of PKC{delta} by viral infection enhanced phosphorylation of the FA molecules. Cells were infected with a control adenoviral vector (Ad-LacZ) or Ad-PKC{delta}. Twenty-four hours later, cells were replated on Fn. Certain cells were pretreated with GF-109203X (GF) or Rottlerin (Rot), as described for Fig. 2. After the 20-h incubation on Fn, cell lysates were prepared and used in Western blots with antibodies against the indicated molecules. Data shown were representative of three different experiments. (C) Integrin- and TGFß1-mediated cell spreading on Fn was accelerated by Ad-PKC{delta} infection. Cells infected with Ad-LacZ or Ad-PKC{delta} were replated on Fn with a concomitant TGFß1 treatment for the indicated periods at the end of the 20-h incubation. Images were taken after the incubation on Fn. Quantitation of spread cells (cells with the longest distance from one end to the other end at least twice longer than the shortest distance) were counted from five isolated images of each experimental condition, and average values were graphed (mean ± standard deviation). A P value less than 0.05 from paired Student's t tests was considered significant. (D) Phosphorylation of the FA molecules was abolished by suppression of PKC{delta} protein. Cells were first transfected with siRNA to down-regulate PKC{delta} (PKC{delta} siRNA) or a control siRNA (Cont siRNA). Twenty-four hours after the transfection, cells were manipulated to be replated on Fn for the indicated experimental conditions. Cell lysates were then prepared and used for Western blots. Data shown represent three independent experiments. (E) Cell spreading was abolished by suppression of PKC{delta} protein. Cells were manipulated as explained for panel D, except that a part of the cells cotransfected with pcDNA3-GFP plus control siRNA or PKC{delta} siRNAwere replated on Fn-precoated cover glasses. Cells on cover glasses were processed for actin staining using phalloidin-conjugated TRITC. The other part of the cells was harvested after the 20-h incubation, and the lysates were immunoblotted using the indicated antibodies. Data shown were representative of three isolated experiments. (F) PKC{alpha} appeared not to be involved in the TGFß1 effects. (Left) No enhancement of phosphorylation of the FA molecules upon PKC{alpha} overexpression. Cells were infected with either control virus (Ad-LacZ) or PKC{alpha} adenovirus (Ad-PKC{alpha}). One day after, cells were replated on Fn for the 20-h incubation under the indicated conditions. Lysates were prepared after the incubation and used in the immunoblots for the indicated molecules. (Right) Cells were infected with dominant-negative PKC{alpha} adenovirus (Ad-DN PKC{alpha}). Twenty-four hours later, cells were replated on Fn for the 20-h incubation in the presence of TGFß1 treatment. After the incubation, the cell image was taken and then lysates were prepared for the immunoblots for the indicated molecules. Data shown were representative of 3 independent experiments.

 
Immunofluorescence microscopy. Cells were first cotransfected with pcDNA3-GFP with either FAK-related nonkinase (FRNK, pBS-FRNK; a kind gift from Juliano L. Rudy, University of North Carolina, Chapel Hill, NC), pKH3-Y416F c-Src (18a), or pCMV-Y31/118/157F paxillin (31). Twenty-four hours later, cells were replated on 10 µg/ml Fn-precoated glass coverslips and incubated for 20 h at 37°C. After incubation, cells were fixed with 3.7% formaldehyde in phosphate-buffered saline (PBS), permeabilized with 0.5% Triton X-100 in PBS at room temperature, and washed three times with PBS. The cells were then incubated with anti-phospho-Y397FAK antibody for 1 h and washed three times with PBS (3 times for 10 min). Cells were then incubated with anti-rabbit immunoglobulin G (IgG)-conjugated TRITC (Chemicon) in a dark and humidified chamber for 1 h. In the case of actin staining, cells were cotransfected with pSF2-integrin {alpha}2, {alpha}3, or pcDM{alpha}5 (23), control siRNA (see above), or PKC{delta} siRNA and pcDNA3-GFP constructs, replated on Fn, fixed, and permeabilized as explained above. Cells were then stained with phalloidin-conjugated TRITC (Molecular Probes, Eugene, OR) for 1 h before washing three times with PBS and mounting with a mounting solution (DakoCytomation, Germany). Mounted samples were visualized by a fluorescent microscope.

Immunoprecipitation. Cells were replated on Fn under diverse conditions as explained above. After the 20-h incubation, cells were washed with cold PBS and immediately lysed in an immunoprecipitation buffer (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.5 mM EGTA, 0.2 mM Na3VO4, 0.2 mM phenylmethylsulfonyl fluoride, 1% Triton X-100, and 0.5% NP-40) on ice. The lysates were cleared by a centrifugation at 13,000 rpm for 30 min at 4°C. An equal amount of anti-FAK or Nck antibody was added directly to the cell extracts with an equal amount of proteins and incubated for 2 h or overnight at 4°C with rotation. After incubation, 30 µl of 50% slurry of protein A/G Sepharose beads (Upstate, Waltham, MA) was added to each sample, and incubation for an additional 2 h at 4°C with rotation was done. Immunoprecipitates were collected by a centrifugation at 13,000 rpm for 3 min at 4°C and washed twice with ice-cold lysis buffer and twice with cold PBS. The immunoprecipitates were then eluted with 2x SDS-polyacrylamide gel electrophoresis (PAGE) sample buffer, and proteins were separated by SDS-PAGE and probed via standard Western blotting.

Flow cytometry. Flow cytometric measurements of integrin subtypes on cells were performed as described previously (23). To study the TGFß1 effects on integrin expression levels in a time-dependent manner, one set of cells was replated on Fn and concomitantly untreated or treated with 5 ng/ml TGFß1 for 0, 8, 12, or 20 h at the end of a total of 20 h incubation, before harvesting for the measurements. To study the effects of PKC inhibition on integrin expression, cells were replated on Fn in the absence or presence of 5 ng/ml TGFß1 treatment without or with pretreatment of 12.5 µM GF-109203X. The raw data were analyzed by using a software program (WinMDI 2.7; Scripps Institute, San Diego, CA).

Wound-healing assay. Normal cells or PKC{delta} wild type (WT)-expressing adenovirus (Ad-PKC{delta})-infected cells in the serum-free replating media were seeded at a high density on 60-mm culture dishes precoated with Fn (10 µg/ml). Twelve hours later, wounds were made by scraping through the cell monolayer with a pipette tip. Cells were washed twice with RPMI 1640 and then treated with 5 ng/ml TGFß1 in the absence or presence of PKC inhibitors (GF-109203X or rottlerin). After 36 h of incubation at 37°C, several images around wounds in each condition were taken.

Invasion assay. A thick layer of matrigel (90 µl of 2.84 mg/ml per well of a 24-well transwell chamber) (BD Biosciences, Oxford, United Kingdom) was prepared on an upper chamber 6 h prior to cell replating. Routinely, the thickness of the layer was 500 mm. Normal or Ad-PKC{delta} WT-infected cells in serum-free RPMI containing 1% BSA were then replated on top of the matrigel. The lower chamber was filled with RPMI 1640 containing 10% fetal bovine serum or 1% BSA. After incubation for 72 h, cells inside of the upper chamber were mopped up. Cells beneath the membrane filter were fixed with 3.7% formaldehyde in PBS and stained with crystal violet, and images were taken with a phase-contrast microscope.

Statistical analysis. Paired Student's t tests were performed for comparisons of mean values to see if the difference is significant. P values of ≤0.05 were considered significant.


   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
TGFß1-, integrin-, and PKC{delta}-mediated spreading of gastric carcinoma cells. We have interests in studying the roles of collaborative signaling of integrins with the TGFß1 pathway in regulation of cellular behaviors. Specifically, we observed that TGFß1 treatment of normally round-shape SNU16mAd gastric carcinoma cells on Fn caused spreading (Fig. 1A). In order to determine if this TGFß1-mediated cell spreading requires integrin-mediated engagements with ECMs, cells were replated on PL with a concomitant 5 ng/ml TGFß1 treatment for 20 h. However, TGFß1 treatment did not cause spreading in cells replated on PL (Fig. 1A). Furthermore, TGFß1-induced cell spreading appeared to depend on Smad pathways, since cells infected with adenovirus encoding for the Smad7, an inhibitory Smad, blocked the spreading, whereas cells infected with adenovirus for ß-galactosidase (Ad-LacZ) maintained spread (Fig. 1B). These data suggest that both integrin engagement to ECM and Smad-dependent TGFß1 signaling are required for spreading of the gastric carcinoma cells.



View larger version (121K):
[in this window]
[in a new window]
  		FIG. 1. Integrin- and TGFß1-mediated spreading of gastric carcinoma cells on fibronectin. (A) TGFß1 and integrin signaling are required for spreading of gastric SNU16mAd carcinoma cells on fibronectin. The cells were trypsinized, collected, washed with RPMI 1640 containing 1% BSA, kept in suspension for 1 h at 37°C with rolling over, replated on Fn (10 µg/ml)- or PL (10 µg/ml)-precoated dishes, and incubated without or with 5 ng/ml TGFß1 treatment for 20 h. Phase-contrast images were taken after the incubation. (B) Integrin- and TGFß1-mediated cell spreading was blocked by Smad7 overexpression. Cells were infected with adenovirus encoding for ß-galactosidase (LacZ) or Smad7, an inhibitory Smad. Twenty-four hours later, infected cells were manipulated, as explained above.

 
We next examined the effects of various pharmacological inhibitor treatments to determine which intracellular signaling molecules were responsible for the integrin- and TGFß1-mediated spreading. In these tests, we found that the cell spreading was blocked by PKC inhibition using GF-109203X (a general inhibitor of PKCs) or rottlerin (Rot, an inhibitor of PKC{delta}) (Fig. 2A). These results indicate that SNU16mAd cell spreading depends on integrin, TGFß1, and PKC signal transduction. To determine which PKC isoform(s) is involved in the cell spreading, we attempted to correlate phosphorylation of the isoforms with the spreading behaviors. Among the isoforms, Ser643 phosphorylation of PKC{delta} correlated closely with the spreading behaviors; cell spreading and Ser643 phosphorylation of PKC{delta} were minimal in the absence of TGFß1 treatment on Fn but were induced by TGFß1 in a GF treatment-dependent manner (Fig. 2B). However, PKC{alpha}/ßII, PKC{varepsilon}, PKC{zeta}/{lambda}, and PKC{theta} appeared not to be involved in the spreading, since phosphorylation of PKC{alpha}/ßII at Thr638/641 did not correlate with the spreading behaviors (Fig. 2B) and PKC{varepsilon} phosphorylation by using anti-phospho-pan PKC antibody and PKC{zeta}/{lambda} at Thr410/403 did not either (data not shown). PKC{theta} was not expressed in the cells (data not shown). Taken together, these observations suggest that integrin- and TGFß1-mediated SNU16mAd cell spreading may involve PKC{delta}.



View larger version (56K):
[in this window]
[in a new window]
  		FIG. 2. PKC pathway correlated with the integrin- and TGFß1-mediated cell spreading. (A) Integrin- and TGFß1-mediated cell spreading was abolished by PKC inhibition using GF-109203X (GF) and rottlerin (Rot). The cells were pretreated with GF-109203X (12.5 µM) or rottlerin (10 µM) 30 min prior to replating on Fn and TGFß1 treatment for 20 h. (B) Phosphorylation of PKC{delta} serine 643 was increased by TGFß1 treatment for 20 h and correlated with the integrin- and TGFß1-mediated cell spreading. The cells were manipulated as in panel A. GF-109203X was treated as explained above. After the incubation for 20 h, cell lysates were prepared as described in Materials and Methods. Data shown are representative of several independent experiments.

 
The cell spreading requires activation of focal adhesion molecules. Because integrin-mediated cell adhesion and spreading activates focal adhesion (FA) molecules including FAK, paxillin, and c-Src, we analyzed phosphorylation of these molecules when TGFß1 was used to treat cells on Fn for various periods (Fig. 3A). It was observed that the longer TGFß1 was treated at the end of the total 20-h incubation, the higher the phosphorylations of FAK Tyr397, paxillin Tyr118, c-Src Tyr416, and PKC{delta} Ser643 (Fig. 3A). Interestingly, in addition to Ser643 phosphorylation, TGFß1 treatment for longer than 4 h at the end of the 20-h incubation (e.g., 8 h) also enhanced PKC{delta} expression (Fig. 3A). In addition, blocking of PKC{delta} activity by pharmacological inhibitors (GF-109203X or rottlerin) decreased phosphorylation of the FA molecules (Fig. 3B). These data demonstrated that TGFß1-, integrin-, and PKC-mediated cell spreading correlated with phosphorylation of the FA molecules. In order to verify the correlation of cell spreading with phosphorylation of the FA molecules, we examined the spreading of cells transiently cotransfected with the expression vectors of green fluorescent protein (GFP) and either dominant-negative or inactive forms of the FA molecules. Transfection of GFP alone did not block the TGFß1-mediated spreading (see Fig. 6E). Cells transfected with FRNK (dominant negative), dominant-negative (Y31/118/157F) paxillin, or inactive (Y416F) c-Src did not spread by TGFß1 treatment on Fn, whereas the surrounding untransfected cells did (Fig. 3C). Among GFP-positive cells, 90% (± 3.0%) of FRNK-transfected cells, 85% (± 4.5%) of Y31/118/157F paxillin-transfected cells, and 92% (± 3.2%) of Y416F c-Src-transfected cells showed round shapes [i.e., (the longest distance from one end to the other end of a cell)/(the shortest distance) < 2.0]. Previously it was shown that a complex formation of FAK with other adapter proteins including p130Cas was involved in cell spreading (5). In this study, formation of a protein complex including FAK, p130Cas, and Nck correlated with the cell spreading behaviors (Fig. 3D), supporting a previous suggestion that the complex might stabilize the active multiprotein complex at FAs (6). Therefore, these data suggest that the TGFß1-, integrin-, and PKC-mediated cell spreading requires activation of the FA molecules and also involves formation of stable protein complexes at FAs.

The cell spreading requires new synthesis of PKC{delta} and integrins {alpha}2 and {alpha}3. To determine whether the cell spreading depends on new protein synthesis, we examined the effects of cycloheximide treatment on the cell spreading. Inhibition of protein synthesis abolished phosphorylation of the FA molecules, expression and Ser643 phosphorylation of PKC{delta} (Fig. 4A, left), and cell spreading (Fig. 4A, right) by TGFß1 treatment. In addition to increased expression of PKC{delta} by TGFß1 (Fig. 3A and 4A), the cell spreading also correlated with increased expression of integrins {alpha}2 and {alpha}3, but not integrins {alpha}4 or {alpha}5, ß1-conjugating integrins, {alpha}1(I) or {alpha}2(I) collagen I chains (a major integrin {alpha}2 binding partner), or {alpha}3 chain of laminin 5 (a major integrin {alpha}3 binding partner), in a cycloheximide treatment-dependent manner (Fig. 4B and data not shown). Flow cytometric analysis also revealed that TGFß1 treatments with cells on Fn increased the expression of integrins {alpha}2 and {alpha}3 on the cell surface in a time-dependent manner (Fig. 4C); the expression was inhibited by PKC inhibition (Fig. 4D) or Smad7 overexpression (Fig. 4E). Taken together, these data suggest that the cell spreading mediated by TGFß1, integrin, and PKC pathways involves Smad-dependent increases in PKC{delta} expression and Ser643 phosphorylation and expression of integrins {alpha}2 and {alpha}3.

Next, we investigated the significance of increased integrin expression with regard to the cell spreading. Cells were premixed with functional blocking integrin antibodies to preoccupy the integrins on the cell surface and then replated. Preincubation with anti-integrin {alpha}2 (clone P1E6) or {alpha}3 (clone P1B5), but not {alpha}5 (clone P1D6), antibody inhibited the cell spreading (Fig. 5A) and phosphorylation of the FA molecules (Fig. 5B). Interestingly, phosphorylation of the FA molecules was more effectively reduced by integrin {alpha}3 blockage than integrin {alpha}2 blockage, presumably indicating a specificity of signal transduction through integrin subtypes. However, the integrin blocking study did not reduce PKC{delta} expression or Ser643 phosphorylation, indicating that PKC{delta} acts upstream of the integrins (Fig. 5B). Meanwhile, the PKC{alpha} level was not changed by TGFß1 treatment (Fig. 5B), indicating again that the TGFß1-mediated effects did not involve PKC{alpha} (also as indicated in Fig. 2B and 6F). Therefore, we suggest that the cell spreading requires increased expression and activation of integrins {alpha}2 and {alpha}3, which function downstream of PKC{delta}. However, overexpression of human integrin {alpha}2 or {alpha}3 did not lead to cell spreading when TGFß1 was not treated (Fig. 5C), indicating that additional TGFß1-mediated signaling activity in addition to integrin expression is necessary for the cell spreading.



View larger version (86K):
[in this window]
[in a new window]
  		FIG. 5. The TGFß1 effects depend on expression and activation of integrins {alpha}2 or {alpha}3. The cells were pretreated with 10 µg/ml anti-integrin {alpha}2 (P1E6), {alpha}3 (P1B5), or {alpha}5 (P1D6) antibodies, 30 min prior to the replating on Fn and a concomitant TGFß1 treatment for 20 h. After the incubation, cell images were taken, and lysates were prepared and used in immunoblots using antibodies against the indicated molecules. Data shown are representative of two isolated experiments. (A) Functional blocking anti-integrin {alpha}2 or {alpha}3, but not {alpha}5, antibodies abolished the cell spreading. (B) Functional blocking of the integrins inhibited phosphorylation of the FA molecules. (C) Cells were transiently cotransfected with a control plasmid (C) or pSF2-integrin {alpha}2 or {alpha}3 with pCDNA3-GFP. One day after the transfection, cells were replated onto Fn (10 µg/ml)-precoated cover glasses in the absence of TGFß1 treatment. After incubation for 20 h at 37°C and 5% CO2, cells were processed for actin staining with phalloidin-conjugated TRITC. Another set of cells in 60-mm culture dishes were harvested for lysates prior to performing immunoblotting using the indicated antibodies. BLK mAb, functional blocking monoclonal antibody.

 
Regulation of PKC{delta} expression and activity affects the cell spreading. Because PKC{delta} Ser643 phosphorylation correlated with the cell spreading during the inhibitor experiments, we investigated the significance of PKC{delta} in promoting cell spreading through regulation of PKC{delta} expression and phosphorylation. When cells were infected with various amounts of PKC{delta} WT-expressing adenovirus (Ad-PKC{delta}), Ser643 phosphorylation and the expression level of PKC{delta} dramatically increased (Fig. 6A). Thus, we could use the Ad-PKC{delta} to enable PKC{delta} overexpression and Ser643 phosphorylation. We next investigated whether PKC{delta} overexpression (and thus Ser643 overphosphorylation) could cause enhanced activation of the spreading-related FA molecules. PKC{delta} overexpression enhanced Ser643 phosphorylation of PKC{delta}, expression of integrin {alpha}2, and TGFß1-mediated activation of the FA molecules, but GF-109203X or rottlerin treatment abolished the enhancements (Fig. 6B). Moreover, compared to cells infected with the control adenovirus (Ad-LacZ), spreading of PKC{delta}-overexpressing cells was minor in the absence of TGFß1 and was much more enhanced by TGFß treatment (Fig. 6C). Next, we tested if PKC{delta} down-regulation through its siRNA transfection affected the spreading. PKC{delta} suppression attenuated activation of the FA molecules and the integrin {alpha}2 expression level (Fig. 6D). PKC{delta}-suppressed cells did not spread, whereas normal neighbor cells spread (Fig. 6E). These data indicate that PKC{delta} indeed mediates TGFß1-induced phosphorylation of the FA molecules, expression of the integrins, and cell spreading. Meanwhile, overexpression of PKC{alpha} using adenovirus with its cDNA did not cause additional enhancement of FA molecules phosphorylation, and dominant-negative PKC{alpha} could not abolish the TGFß1-mediated spreading (Fig. 6F). These observations suggest that the cell spreading involves PKC{delta}, but not PKC{alpha}, signaling.

The more sufficiently the integrin-related signaling is activated, the better the cell spreading. In this study, PKC{delta} overexpression alone did not cause significant activation of the FA molecules (Fig. 6B, left, lanes 1 and 4) and led to a minor spreading (Fig. 6C), although additional TGFß1 treatment caused complete and accelerated spreading (Fig. 6C). Therefore, we hypothesized that a slight increase in either integrin signaling (Fig. 7A, B, and C) or TGFß1 signaling (Fig. 7D and E) might facilitate the cell spreading. To test this possibility, we first examined the effects of cell replating on various ECM-precoated dishes on activation of the FA molecules and the cell spreading. Activation of the FA molecules was increased by TGFß1 on the tested ECMs but not significantly on poly-L-lysine (Fig. 7A). Interestingly, basal (without TGFß1 treatment) phosphorylation of the FA molecules was higher on collagen I than any of the other tested ECMs (Fig. 7A, lane 5). We thus examined the spreading behavior of cells overexpressing PKC{delta} on collagen I. PKC{delta} overexpression induced cell spreading (Fig. 7B) and consistently increased basal phosphorylation of the FA molecules and integrin {alpha}2 expression, compared to those in Ad-LacZ-infected cells (Fig. 7C, lanes 1 and 4) even without TGFß1 treatment. It appeared that higher basal integrin signaling activity could cause the cell spreading even without TGFß1 treatment. Furthermore, compared to the Ad-LacZ-infected cells, quantitatively (i.e., higher spreading rate) and qualitatively (i.e., wider spreading) enhanced cell spreading by TGFß1 treatment was observed on collagen I (Fig. 7B). Interestingly, being consistent with the quantitatively and qualitatively enhanced spreading, the TGFß1-increased phosphorylation of the FA molecules and expression level of integrin {alpha}2 were observed in PKC{delta}-overexpressing cells (Fig. 7B and C, lanes 2 and 5). In addition, this spreading was blocked by PKC inhibition (Fig. 7B), as decreased phosphorylation of the FA molecules and suppressed expression of integrins by PKC inhibition (Fig. 7C). We next investigated if an increased TGFß1 signaling presumably through Smad2 or Smad3 overexpression would affect the spreading. The TGFß1 effects were investigated using cells infected with adenovirus encoding for either ß-galactosidase or FLAG-tagged Smad2 or Smad3 that still require TGFß1 treatment for their activation (30). Interestingly, overexpression of Smad3, but not Smad2, enhanced the TGFß1-mediated activation of the FA molecules, expression and Ser643 phosphorylation of PKC{delta}, expression of integrin {alpha}2 (Fig. 7D), and cell spreading (Fig. 7E) in a PKC activity-dependent manner. Therefore, these observations suggest the significance of the Smad3-dependent TGFß1 signaling in expression and activation of PKC{delta} and integrins, activation of the FA molecules, and the spreading.






View larger version (330K):
[in this window]
[in a new window]
  		FIG. 7. Cells under more efficient integrin-related signaling activities lead to enhanced cell spreading. (A) Phosphorylation of the FA molecules in cells replated on poly-L-lysine or diverse ECM proteins. Cells were replated on culture dishes precoated with various ECMs (10 µg/ml) (Coll, collagen I; Ln, laminin I; Vn, vitronectin), in the absence or presence of TGFß1 treatment for 20 h. Cell lysates were then prepared and used for Western blots with antibodies against the indicated molecules. Data shown represent three independent experiments. (B) Spreading of PKC{delta} WT-overexpressing cells on collagen I even without TGFß1 treatment. Twenty-four hours after infection with Ad-LacZ or Ad-PKC{delta}, cells were manipulated to be replated on collagen I-precoated dishes. Certain cells were pretreated with GF-109203X (GF), as described above. Twenty hours after the replating, images were taken. The rate of spread cells in percent graphed was determined as explained for Fig. 6C. A P value of ≤0.05 was considered significant. (C) Cells infected with Ad-PKC{delta} enhanced basal and TGFß1-mediated expression of integrin {alpha}2 and phosphorylation of the FA molecules on collagen I. Cell manipulation and cell lysate preparation were performed, as explained above. Data shown were representative of three independent experiments. (D, E) Smad3 overexpression enhanced TGFß1-mediated phosphorylation of the FA molecules (D) and cell spreading (E). Cells were infected with adenovirus encoding for either control (Ad-LacZ), Flag-Smad2 (Ad-Smad2), or Flag-Smad3 (Ad-Smad3). Twenty-four hours later, infected cells were replated on Fn. In certain cases, GF-109203X or Rottlerin (Rot) pretreatment was done as explained above. After 20 h of incubation at 37°C, cell lysates were prepared for Western blots, or cell images were taken. Quantitative determinations of the spread cells were done as explained for Fig. 6C. A P value of ≤0.05 was considered significant. Data shown were representative of three isolated experiments.

 
The cell spreading-related signaling network also increased migration and invasion through matrigel. Cell spreading may enable a cell to respond to extracellular cues during migration and invasion. To investigate whether cell spreading and spreading-related biochemical activities correlate with cell motility and invasiveness, we performed wound-healing and invasion assays. Compared with the untreated cells, wound healing was significant in the TGFß1-treated cells on Fn, in a PKC{delta} activity-dependent manner. This increased wound healing was not due to alteration in cell proliferation or apoptosis (data not shown) but presumably was due to migration of cells towards the wounds. When cells overexpress PKC{delta}, TGFß1-induced wound healing was further enhanced, but slightly higher than that of TGFß1-treated normal control cells (Fig. 8A). This increased motility by TGFß1 treatment and PKC{delta} activity in cells on Fn indicate that the motility appears to correlate with the spreading behaviors. In addition, the TGFß1-treated cells showed increased invasion through matrigel dependent on PKC activity, compared with the untreated cells. Furthermore, cells overexpressing PKC{delta} also showed more enhanced invasion, also depending on PKC activity (Fig. 8B). Therefore, the signaling network required for the cell spreading was also involved in the motility and invasion, indicating that the motility and invasion correlate with the spreading behaviors.



View larger version (97K):
[in this window]
[in a new window]
  		FIG.8. Wound healing and invasion also depend on TGFß1, PKC{delta}, and integrins. (A) TGFß1- and PKC{delta} activity-mediated wound healing on Fn. Prior to making wounds, normal or Ad-PKC{delta}-infected cells were replated onto Fn-precoated six-well culture plates. Twelve hours later, wounds (marked as dotted lines) were created by scraping through the monolayer. Cells were then incubated in the absence or presence of TGFß1 treatment without or with PKC inhibition (GF-109203X [GF] or rottlerin [Rot]) for 36 h at 37°C, prior to taking images. Note that TGFß1-mediated healing in PKC{delta}-overexpressing cells is slightly enhanced, compared to that in normal control cells (i.e., compare healing around the vertical arrowheads). (B) TGFß1- and PKC{delta} activity-mediated invasion of the cells through matrigel. Normal or Ad-PKC{delta}-infected cells in RPMI 1640 containing 1% BSA were replated onto matrigel, prepared as explained in Materials and Methods, and concomitantly treated with or without 5 ng/ml TGFß1. Certain cells were pretreated with 12.5 µM GF-109203X, 30 min prior to the TGFß1 treatment. Then the upper chambers were placed into 24-well culture dishes containing 0.6 ml of RPMI 1640 with 10% fetal bovine serum (FBS) or with 1% BSA. After incubation at 37°C for 72 h, invasive cells were fixed and stained with crystal violet, and then images were taken, as explained in Materials and Methods. The images shown are representative of three isolated experiments. Stained cells in three independent images for each condition were counted for the graphic presentation (mean ± standard deviation).

 

   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
In this study, we explored the signaling mechanisms underlying the cell spreading process alone, separated from adhesion, using gastric carcinoma SNU16mAd cells. These cells are normally round but spread polygonally by TGFß1 treatment in the absence or presence of serum. The TGFß1-mediated spreading depends on integrin interaction with ECMs and signal transduction involving PKC{delta}. We found that this spreading of gastric carcinoma cells involved induction and Ser643 phosphorylation of PKC{delta}, expression and activation of integrins ({alpha}2 and {alpha}3), and activation of the FA molecules. Furthermore, the spreading-related signaling activities were involved in the wound healing and invasion. Observations from this study suggest that the signaling network involving TGFß1, PKC{delta}, and integrins underlies spreading and migration and invasion of an adherent gastric carcinoma variant cell line, SNU16mAd.

SNU16mAd cells used in this study were obtained from subsequent cultures to collect adherent cells among mostly anchorage-independent SNU16{alpha}5 cells (18). A long period was required for adherence when the cells were replated on Fn-precoated dishes, and the cell shape appeared round even when fully adhered to the substrate. When cells replated on Fn were treated with TGFß1 for a long period (e.g., 20 h), the cells became spread. Therefore, this cell line is a good model system to study cell spreading events by specific stimuli, separate from the adhesion process. This may be an important distinction, as many normally adherent cell types spread gradually and spontaneously after being replated.

So far, signaling networks consisting of integrins, TGFß1, and PKC (especially PKC{delta}) have not been thoroughly investigated, especially for cell spreading and invasiveness, although a previous report showed that general PKC activity preceded integrin-mediated cell adhesion on fibronectin (41). In this study, we demonstrated a complicated signaling network underlying a specific cellular behavior (i.e., cell spreading). Smad-dependent TGFß1 signaling led to increased expression and Ser643 phosphorylation of PKC{delta}, which correlated with induction and activation of integrins, activation and stable complex formation of the FA molecules, and cell spreading. Although the effects of TGFß1 and integrins on metastasis have been previously reported (11, 20, 42), the positive involvement of PKC{delta} in the migration and invasion has not been fully elucidated. The observations from this study suggest that the spreading correlates with increased motility and invasion, since TGFß1-mediated wound healing and invasion also depended on PKC{delta} expression and activation and integrin-related signaling activation. The TGFß1-mediated wound healing on Fn in PKC{delta}-overexpressing cells could be slightly enhanced, compared to that of TGFß1-treated normal control cells, probably because this cell line is much less motile in the absence of serum. This observation was consistent with a slight (but statistically significant) increase in quantitative spreading rate that was accompanied by a qualitatively wider spread (Fig. 6C).

In addition to the PKC inhibitor studies, overexpression by PKC{delta} adenovirus and down-regulation by PKC{delta} siRNA showed the significance of PKC{delta} in the cell spreading, wound healing, and invasion. Among other PKC isoforms, PKC{alpha}/ßII, PKC{zeta}/{lambda}, and -{varepsilon} appeared not to be involved in the system because their phosphorylation status did not correlate well with the cell spreading patterns under the experimental conditions, and PKC{theta} was not expressed in our cells. In addition, PKC{alpha} and PKC{theta} overexpression did not result in spreading on fibronectin. More conclusively, cells infected with adenovirus with dominant-negative PKC{alpha} could still spread on fibronectin upon TGFß1 treatment. Therefore, it is likely that the TGFß1-mediated spreading involves PKC{delta} at least, but not PKC{alpha} and -{theta}, and that PKC{delta} is a mediator for TGFß1 to integrin signaling pathways and acts upstream of integrins and the FA molecules (Fig. 9).



View larger version (33K):
[in this window]
[in a new window]
  		FIG. 9. Schematic working model of the TGFß1-, PKC{delta}-, and integrin-mediated cellular functions in gastric carcinoma cells. TGFß1 treatments of cells on ECMs lead to increased expression and phosphorylation of PKC{delta}, cell surface levels of integrins {alpha}2 or {alpha}3, and activation of the focal adhesion molecules. This signaling network results in spreading, migration, and invasion of the SNU16mAd gastric carcinoma cell variant. In addition to a linear connection, presumably additional bypassing connections may contribute to the TGFß1 effects.

 
On the other hand, the signaling network of TGFß1, PKC{delta}, integrins, and integrin-related signaling molecules appeared to be complicated rather than in order. First, overexpression of PKC{delta} in the absence of TGFß1 treatment did not result in a complete spreading, although TGFß1 treatment facilitated the spreading of PKC{delta}-overexpressing cells, compared to control cells. Second, overexpression of integrins {alpha}2 or {alpha}3 did not cause spreading in the absence of TGFß1. These observations indicate that just a linear connection from the Smad-dependent TGFß1 pathway to integrin induction and activation via PKC{delta} induction and phosphorylation is not sufficient for the spreading and that probably different TGFß1-mediated biochemical actions function to activate and/or assemble downstream effectors (complexes), such as the FA molecules (Fig. 9). Recently there has been diverse evidence that TGFß1 activates diverse intracellular signaling molecules that are also regulated by integrin-mediated cell adhesion (7, 28).

In this study, incubation of the cells with TGFß1 for 20 h in the absence of serum increased Ser643 phosphorylation as well as expression of PKC{delta}. However, it is currently a controversial assumption that Ser643 phosphorylation affects kinase activity of PKC{delta}. One previous study reported that a Ser643 to alanine mutation had no effect on the kinase activity of PKC{delta} (39), whereas another showed that Ser643 of PKC{delta} is an important autophosphorylation site for its enzymatic activity (26). Furthermore, we observed that TGFß1 treatment for only 6 or ~8 h caused Ser643 phosphorylation of PKC{delta} but not induction of PKC{delta} and integrins, phosphorylation of the FA molecules, and cell spreading (data not shown). We also found that 15-h TGFß1-free incubation even after 5-h TGFß1 treatment caused the TGFß1 effects (data not shown), indicating that the TGFß1 treatment alone for such a short period (e.g., for 6 or ~8 h) was not enough to cause the TGF effects. Therefore, it is likely that the Ser643 phosphorylation of PKC{delta} is not critical for the cell spreading, although TGFß1 treatment for 20 h resulted in the phosphorylation correlated with the spreading. Although the significance of the Ser643 phosphorylation for PKC{delta} activity is controversial, pharmacological inhibition of PKC{delta} activity abolished the spreading in this study. Therefore, it appears that the cell spreading importantly involves induction and activation of PKC{delta} and integrins and activation of integrin-related signaling.

There have been evidences for effects of TGFß1 on integrins and/or ECMs, and vice versa (36). Although the increased integrins {alpha}2 or {alpha}3 bind collagen I or laminin 5, respectively (33), neither collagen I ({alpha}2 and {alpha}1 chains) nor laminin 5 ({alpha}3 chain) expression levels of the SNU16mAd cells were changed by TGFß1 treatment upon immunoblotting with commercial antibodies against them (Fig. 4B). Although the TGFß1 effect was shown in all ECMs we tested, we performed most experiments on fibronectin, just because the TGFß1-mediated effect was more obvious with no basal signaling activity on fibronectin, compared to on collagen I or laminin 1, and because these integrins {alpha}2 or {alpha}3 also bind to fibronectin (32, 40). We cannot rule out the possibility that collagen I and/or laminin I might support the integrins {alpha}2 or {alpha}3 enhanced by TGFß1, since the cells still expressed them, although their expressions were not enhanced by the TGFß1 treatment. We did not see TGFß1-mediated spreading on fibronectin when integrin {alpha}5 (a typical Fn receptor) was ectopically overexpressed (data not shown), being consistent with no change in integrin {alpha}5 by TGFß1. Furthermore, the cell spreading was abolished by functional blocking of integrins {alpha}2 or {alpha}3, but not {alpha}5, using their inhibitory monoclonal antibodies. Therefore, we can conclude that the TGFß1-, PKC{delta}-, and integrin-mediated spreading involves increases in specific integrin {alpha}2 or {alpha}3 expression, presumably, but not increases in ECM production. On the other hand, it may be likely that integrins {alpha}2 or {alpha}3 have specific and exclusive linkage(s) to the FA molecules, although both were shown to be involved in this spreading system. We observed that integrin {alpha}3 expression increased suddenly at a specific time point presumably after signal accumulations by TGFß1 treatment surpassed a threshold (Fig. 4C). However, integrin {alpha}2 expression is increased gradually in a time-dependent manner after TGFß1 treatment. This narrower window of TGFß1-mediated integrin {alpha}3 increase rendered the changes in the integrin {alpha}3 expression level much more difficult to detect. Furthermore, phosphorylation of the pY118Paxillin was abrogated by integrin {alpha}3 blocking, using anti-integrin {alpha}3 monoclonal antibody, but not significantly by integrin {alpha}2 blocking, whereas pY397FAK was abrogated by both inhibitory antibodies. These observations may suggest a specific role(s) of each integrin subtype for the TGFß1-mediated FA molecule activation.

We showed in Fig. 7A that basal (without TGFß1 treatment) activation of the FA molecules was more prominent on collagen I and was much more enhanced by TGFß1 treatment, compared to on Fn. This higher basal activation of the FA molecules correlated with cell spreading on collagen I with PKC{delta} overexpression alone even in the absence of TGFß1 treatment, and TGFß1-mediated activity might correlate with a wider spreading (Fig. 7B). Therefore, depending on the ECM, spreading of SNU16mAd cells may require different signaling activities for complete and sufficient spreading. On collagen I, integrin and PKC{delta} signal pathways were enough for the spreading, and additional TGFß1 signaling activity correlated with (qualitatively) wider spreading. Meanwhile, the spreading on Fn appears to additionally require TGFß1 signaling probably to support the PKC{delta} and integrin signaling, and the combined signaling output through complicated signal connections may lead to complete cell spreading (Fig. 9).

All together, in this study, the positive roles of PKC{delta} in the TGFß1 and integrin effects on cellular functions were clearly suggested by the observations that PKC{delta} is required to mediate TGFß1 treatment for integrin expression and activation, leading to spreading, migration, and invasion of human gastric carcinoma SNU16mAd cells.


   ACKNOWLEDGMENTS
 
We deeply thank Jang-Soo Chun (Gwangju Institute of Science and Technology, Gwangju, Korea) for adenovirus encoding for PKC{delta} or PKC{alpha} and Rudy L. Juliano (University of North Carolina at Chapel Hill, Chapel Hill, NC) for expression vectors such as pBS-FRNK and etc, respectively.

This work was supported by a Korea Research Foundation Grant (KRF-2004-015-C00445) to J. W. Lee.


   FOOTNOTES
 
* Corresponding author. Mailing address: Cancer Research Institute, College of Medicine, Seoul National University, 28, Yongon-dong, Chongno-gu, Seoul 110-799, Republic of Korea. Phone: 82-2-3668-7030. Fax: 82-2-766-4487. E-mail: jwl{at}snu.ac.kr. Back

{dagger} Present address: Laboratory of Clinical Biology and Pharmacogenomics, Department of Preventive Medicine, School of Oriental Medicine, Kyunghee University, Seoul 130-701, Republic of Korea. Back


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 
1. Akhurst, R. J., and R. Derynck. 2001. TGF-ß signaling in cancer—a double-edged sword. Trends Cell Biol. 11:S44-S51.[Medline]

2. Alahari, S. K., P. J. Reddig, and R. L. Juliano. 2002. Biological aspects of signal transduction by cell adhesion receptors. Int. Rev. Cytol. 220:145-184.[Medline]

3. Arthur, W. T., and K. Burridge. 2001. RhoA inactivation by p190RhoGAP regulates cell spreading and migration by promoting membrane protrusion and polarity. Mol. Biol. Cell 12:2711-2720.[Abstract/Free Full Text]

4. Bhowmick, N. A., R. Zent, M. Ghiassi, M. McDonnell, and H. L. Moses. 2001. Integrin ß1 signaling is necessary for transforming growth factor-ß activation of p38MAPK and epithelial plasticity. J. Biol. Chem. 276:46707-46713.[Abstract/Free Full Text]

5. Bruce-Staskal, P. J., and A. H. Bouton. 2001. PKC-dependent activation of FAK and Src induces tyrosine phosphorylation of Cas and formation of Cas-Crk complexes. Exp. Cell Res. 264:296-306.[CrossRef][Medline]

6. Buday, L., L. Wunderlich, and P. Tamas. 2002. The Nck family of adapter proteins: regulators of actin cytoskeleton. Cell. Signal. 14:723-731.[CrossRef][Medline]

7. Derynck, R., and Y. E. Zhang. 2003. Smad-dependent and Smad-independent pathways in TGF-ß family signalling. Nature 425:577-584.[CrossRef][Medline]

8. Eliceiri, B. P. 2001. Integrin and growth factor receptor crosstalk. Circ. Res. 89:1104-1110.[Abstract/Free Full Text]

9. Engel, M. E., M. A. McDonnell, B. K. Law, and H. L. Moses. 1999. Interdependent SMAD and JNK signaling in transforming growth factor-ß-mediated transcription. J. Biol. Chem. 274:37413-37420.[Abstract/Free Full Text]

10. Gschwendt, M., H. J. Muller, K. Kielbassa, R. Zang, W. Kittstein, G. Rincke, and F. Marks. 1994. Rottlerin, a novel protein kinase inhibitor. Biochem. Biophys. Res. Commun. 199:93-98.[CrossRef][Medline]

11. Guo, W., and F. G. Giancotti. 2004. Integrin signalling during tumour progression. Nat. Rev. Mol. Cell Biol. 5:816-826.[CrossRef][Medline]

12. Hartsough, M. T., and K. M. Mulder. 1995. Transforming growth factor ß activation of p44MAPK in proliferating cultures of epithelial cells. J. Biol. Chem. 270:7117-7124.[Abstract/Free Full Text]

13. Hayashida, T., and H. W. Schnaper. 2004. High ambient glucose enhances sensitivity to TGF-ß1 via extracellular signal-regulated kinase and protein kinase C{delta} activities in human mesangial cells. J. Am. Soc. Nephrol. 15:2032-2041.[Abstract/Free Full Text]

14. Hocevar, B. A., T. L. Brown, and P. H. Howe. 1999. TGF-ß induces fibronectin synthesis through a c-Jun N-terminal kinase-dependent, Smad4-independent pathway. EMBO J. 18:1345-1356.[CrossRef][Medline]

15. Hynes, R. O. 2002. Integrins: bidirectional, allosteric signaling machines. Cell 110:673-687.[CrossRef][Medline]

16. Ignotz, R. A., and J. Massague. 1987. Cell adhesion protein receptors as targets for transforming growth factor-ß action. Cell 51:189-197.[CrossRef][Medline]

17. Juliano, R. L. 2002. Signal transduction by cell adhesion receptors and the cytoskeleton: functions of integrins, cadherins, selectins, and immunoglobulin-superfamily members. Annu. Rev. Pharmacol. Toxicol. 42:283-323.[CrossRef][Medline]

18. Kim, H. P., M. S. Lee, J. Yu, J. A. Park, H. S. Jong, T. Y. Kim, J. W. Lee, and Y. J. Bang. 2004. TGF-ß1 (transforming growth factor-ß1)-mediated adhesion of gastric carcinoma cells involves a decrease in Ras/ERKs (extracellular-signal-regulated kinases) cascade activity dependent on c-Src activity. Biochem. J. 379:141-150.[CrossRef][Medline]

18. Kim, H. P., T. Y. Kim, M. S Lee, H. S Jong, T. Y. Kim, J. W. Lee, and Y. J. Bang. 2005. TGF-ß1-mediated activations of c-Src and Rac1 modulate levels of cyclins and p27Kip1 CDK inhibitor in hepatoma cells replated on fibronectin. Biochim. Biophys. Acta 1743:151-161.[Medline]

19. Kim, M. G., G. Lee, S.-K. Lee, M. Lolkema, J. Yim, S. H. Hong, and R. H. Schwartz. 2000. Epithelial cell-specific laminin 5 is required for survival of early thymocytes. J. Immunol. 165:192-201.[Abstract/Free Full Text]

20. Koivunen, J., V. Aaltonen, S. Koskela, P. Lehenkari, M. Laato, and J. Peltonen. 2004. Protein kinase C {alpha}/ß inhibitor Gö6976 promotes formation of cell junctions and inhibits invasion of urinary bladder carcinoma cells. Cancer Res. 64:5693-5701.[Abstract/Free Full Text]

21. Kumar, N. M., S. L. Sigurdson, D. Sheppard, and J. S. Lwebuga-Mukasa. 1995. Differential modulation of integrin receptors and extracellular matrix laminin by transforming growth factor-ß1 in rat alveolar epithelial cells. Exp. Cell Res. 221:385-394.[CrossRef][Medline]

22. Lee, J. W., and R. Juliano. 2004. Mitogenic signal transduction by integrin- and growth factor receptor-mediated pathways. Mol. Cells 17:188-202.[Medline]

23. Lee, J. W., and R. L. Juliano. 2000. {alpha}5ß1 integrin protects intestinal epithelial cells from apoptosis through a phosphatidylinositol 3-kinase and protein kinase B-dependent pathway. Mol. Biol. Cell 11:1973-1987.[Abstract/Free Full Text]

24. Lee, J. W., and R. L. Juliano. 2002. The {alpha}5ß1 integrin selectively enhances epidermal growth factor signaling to the phosphatidylinositol-3-kinase/Akt pathway in intestinal epithelial cells. Biochim. Biophys. Acta 1542:23-31.[Medline]

25. Lee, M. S., S. G. Ko, H. P. Kim, Y. B. Kim, S. Y. Lee, S. G. Kim, H. S. Jong, T. Y. Kim, J. W. Lee, and Y. J. Bang. 2004. Smad2 mediates Erk1/2 activation by TGF-ß1 in suspended, but not in adherent, gastric carcinoma cells. Int. J. Oncol. 24:1229-1234.[Medline]

26. Li, W., J. Zhang, D. P. Bottaro, and J. H. Pierce. 1997. Identification of serine 643 of protein kinase C-{delta} as an important autophosphorylation site for its enzymatic activity. J. Biol. Chem. 272:24550-24555.[Abstract/Free Full Text]

27. Mainiero, F., A. Gismondi, R. Strippoli, J. Jacobelli, A. Soriani, S. Morrone, and A. Santoni. 2000. Integrin-mediated regulation of cytokine and chemokine production by human natural killer cells. Eur. Cytokine Netw. 11:493-494.[Medline]

28. Massague, J. 2000. How cells read TGF-ß signals. Nat. Rev. Mol. Cell Biol. 1:169-178.[CrossRef][Medline]

29. Massague, J. 1998. TGF-ß signal transduction. Annu. Rev. Biochem. 67:753-791.[CrossRef][Medline]

30. Nakao, A., T. Imamura, S. Souchelnytskyi, M. Kawabata, A. Ishisaki, E. Oeda, K. Tamaki, J. Hanai, C. H. Heldin, K. Miyazono, and P. ten Dijke. 1997. TGF-ß receptor-mediated signalling through Smad2, Smad3 and Smad4. EMBO J. 16:5353-5362.[CrossRef][Medline]

31. Oktay, M., K. K. Wary, M. Dans, R. B. Birge, and F. G. Giancotti. 1999. Integrin-mediated activation of focal adhesion kinase is required for signaling to Jun NH2-terminal kinase and progression through the G1 phase of the cell cycle. J. Cell Biol. 145:1461-1469.[Abstract/Free Full Text]

32. Plow, E. F., T. A. Haas, L. Zhang, J. Loftus, and J. W. Smith. 2000. Ligand binding to integrins. J. Biol. Chem. 275:21785-21788.[Free Full Text]

33. Reddy, K. V., and S. S. Mangale. 2003. Integrin receptors: the dynamic modulators of endometrial function. Tissue Cell 35:260-273.[CrossRef][Medline]

34. Rodriguez-Barbero, A., J. Obreo, L. Yuste, J. C. Montero, A. Rodriguez-Pena, A. Pandiella, C. Bernabeu, and J. M. Lopez-Novoa. 2002. Transforming growth factor-ß1 induces collagen synthesis and accumulation via p38 mitogen-activated protein kinase (MAPK) pathway in cultured L6E9 myoblasts. FEBS Lett. 513:282-288.[CrossRef][Medline]

35. Runyan, C. E., H. W. S