The Pro-inflammatory Mediator Leukotriene D4 Induces Phosphatidylinositol 3-Kinase and Rac-dependent Migration of Intestinal Epithelial Cells*

Inflammatory bowel diseases are associated with increased risk of developing colon cancer. A possible role of the pro-inflammatory leukotriene D4 (LTD4) in this process has been implicated by the findings that LTD4 can signal increased proliferation and survival, both hallmarks of a cancer cell, in non-transformed intestinal epithelial cells. Here we make the novel finding that LTD4 can also signal increased motility in these cells. In parallel, we found that LTD4 induced a simultaneous transient 10-fold increase in Rac but not Cdc42 activity. These data were also supported by the ability of LTD4 to activate the Rac GDP/GTP exchange factor Vav2. Further, LTD4 triggered a 3-fold transient increase in phosphatidylinositol 3-kinase (PI3K) phosphorylation, a possible upstream activator of the Vav2/Rac signaling pathway. The activation of Rac was blocked by the PI3K inhibitors LY294002 and wortmannin and by transfection of a kinase-negative mutant of PI3K or a dominant-negative form of Vav2. Furthermore, Rac was found to co-localize with actin in LTD4-generated membrane ruffles that were formed by a PI3K-dependent mechanism. In accordance, the inhibition of the PI3K and Rac signaling pathway also blocked the LTD4-induced migration of the intestinal cells. The present data reveal that an inflammatory mediator such as LTD4 cannot only increase proliferation and survival of non-transformed intestinal epithelial cells but also, via a PI3K/Rac signaling pathway, trigger a motile response in such cells. These data demonstrate the capacity of inflammatory mediators to participate in the process by which inflammatory bowel conditions increase the risk for colon cancer development.

Migration of epithelial cells is essential during the development of the gut and during different pathological situations such as wound healing and tumor metastasis (1). The conversion from a sessile to a migratory phenotype requires an extensive remodeling of the actin cytoskeleton (2). In response to different chemotactic substances, cell migration is initiated by the formation of lamellipodia or membrane ruffles at the leading front of a migrating cell (3,4).
It is well established that epithelial cell migration is stimulated by activation of receptor tyrosine kinases (5,6). The intracellular signaling pathways activated by such receptors and responsible for the coordinated changes in the actin cytoskeleton seen during cell motility generally include the activation of different members of the Rho family of GTPases (3). The Rho family consists of three main members: Rho; Rac; and Cdc42. In most cell types, these proteins execute specific functions, i.e. Rho promotes the formation of stress fibers and focal adhesion complexes and Rac initiates actin polymerization at the cell membrane and is responsible for the generation of lamellipodia and membrane ruffles, whereas Cdc42 promotes the formation of filopodia and microspikes at the cell periphery (7). These monomeric GTPases cycle between a GDP-bound inactive state and a GTP-bound active state. In their inactive state, Rho GTPases are bound to proteins called guanine nucleotide dissociation inhibitors. Guanine nucleotide exchange factors (GEFs), 1 activated by extracellular stimuli, catalyze the exchange of GDP for GTP on Rho GTPases and thereby activate these proteins. In their GTP-bound state, these proteins interact with specific effectors to initiate downstream signals and functions (7). The subsequent hydrolysis of bound GTP to GDP is catalyzed by the family of GTPase-activating proteins (8).
The Dbl family constitutes the main group of GEFs acting on Rho GTPases (9). A hallmark of these exchange factors is that they contain two key conserved domains, a Dbl homology domain that is believed to be responsible for catalyzing GDP/ GTP exchange on Rho GTPases and a pleckstrin homology (PH) domain that is important for cellular localization through its interaction with lipids and/or proteins (10). Notably, the PH domain on GEFs are known to bind phosphatidylinositol 4,5-bisphosphate and phosphatidylinositol 3,4,5trisphosphate, a substrate and a product of the PI3K, respectively. Such a lipid/protein interaction offers a reasonable explanation for how the PI3K can enable activation of Rac GTPase. It has been suggested that the association between GEF and phosphatidylinositol 3,4,5-trisphosphate may cause dissociation of the GEF from its binding to guanine nucleotide dissociation inhibitor (9). Possibly, such a mechanism could explain how agonist-induced activation of PI3K and generation of phosphatidylinositol 3,4,5-trisphosphate could enable a downstream phosphorylation and acti-vation of Vav (11). Src tyrosine kinase, a kinase known to be activated by LTD 4 (12), has been shown to phosphorylate and activate Vav2 and subsequently activate Rac (13). PI3Ks are composed of a catalytic subunit (p110) and a regulatory (p85/ p55) subunit (14). Phosphorylation of the p85 subunit is essential for the translocation and activation of the p110/85 complex from the cytosol to the plasma membrane where it interacts with its effectors (15). Interestingly, the inhibition of PI3K by both molecular and pharmacological approaches blocks membrane ruffling and inhibits cell migration (16).
Leukotrienes belong to an important group of pro-inflammatory mediators derived from arachidonic acid via the 5-lipoxygenase pathway and have been implicated in many inflammatory conditions (17,18). LTD 4 is the most potent of the cysteinyl leukotrienes (LTC 4 , LTD 4 , and LTE 4 ), and it has been implicated in conditions such as asthma and inflammatory bowel diseases (19). LTD 4 mediates its effects by the binding to two different G protein-coupled receptors, CysLT 1 and CysLT 2 (20). The CysLT 1 receptor has a much higher affinity for its ligand than the CysLT 2 -receptor and has been shown to couple to both a pertussis toxin (PTX)-sensitive and a PTX-insensitive G protein, indicating that different signal transduction pathways can be initiated and possibly mediate distinct downstream functions (12,21,22). We have previously shown that prolonged exposure of non-transformed intestinal epithelial cells to LTD 4 resulted in up-regulation of factors related to colon carcinogenesis such as cyclooxygenase-2, ␤-catenin, and Bcl-2 (23). More recently, we have shown that LTD 4 induces prolif- The migration of Int 407 cells was evaluated in a Boyden chamber as described under "Experimental Procedures." A, the cells incubated in the absence or presence of the indicated LTD 4 concentration for 18 h. B, cells were incubated in the absence or presence of 80 nM LTD 4 for indicated periods of time, after which cells were lysed and analyzed with a GST-PAKcrib pull-down assay as described under "Experimental Procedures." Proteins bound to GST-PAKcrib and also the corresponding whole lysates from samples used for the GST-PAKcrib pull-down assays were separated on 12% SDS-PAGE, transferred to polyvinylidene difluoride membranes, and immunoblotted with antibodies specific for Rac or Cdc42 (1:1000). Representative blots of the activity of Rac and Cdc42 are shown together with the accumulated results from the densitometric analysis of the LTD 4 -induced Rac activation. C, cells were incubated in the absence or presence of 80 nM LTD 4 for 15 min or 100 ng/ml EGF for 15 min. The cells were lysed, and Vav2 was immunoprecipitated (IP) from the different samples. The precipitates were immunoblotted subsequently with an anti-phosphotyrosine (PTyr) antibody (1:5000) and then reprobed with an anti-Vav2 antibody (1:1000) as a loading control. Representative blots and the accumulated results of the densitometric analysis of LTD 4 -induced tyrosine phosphorylation of Vav2 are shown. The degree of Vav2 phosphorylation was calculated as the percentage of unstimulated cells and given as the means Ϯ S.E. of four separate experiments. D, this blot shows Rac activity (analyzed as in B) in cells transfected with either empty vector (lanes 1-2) or DN-Vav2 vector (lanes 3-4) and then incubated in the absence or presence of LTD 4 . Representative blots and the accumulated results of the densitometric analysis of Rac activities are shown. The degree of Rac activity was calculated as the percentage of unstimulated cells transfected with an empty vector and given as the means Ϯ S.E. of four separate experiments. All of the statistical significances were evaluated using the unpaired Student's t test. *, p Ͻ 0.05; **, p Ͻ 0.01. erative and survival responses, hallmarks of a cancer cell, in non-transformed intestinal epithelial cells (21,24). Taken together with the fact that ulcerative colitis is associated with an increased incidence of neoplastic transformation (25), the indications showing that there exists a link between inflammation and cancer (26) and earlier studies demonstrating that colon cancer is underrepresented in a population of patients with ulcerative colitis treated with non-steroidal anti-inflammatory drugs (27) suggest that LTD 4 could be an essential factor in mediating the coupling between inflammatory bowel disease and colon cancer.
Transformed cancer cells exhibit distinct properties including increased proliferation, survival, and migration. Interestingly enough, we have previously shown that LTD 4 triggered both a proliferative and a survival response in non-transformed intestinal epithelial cells (27). In this study, we explored whether and how the pro-inflammatory LTD 4 affects migration in these intestinal epithelial cells.

EXPERIMENTAL PROCEDURES
Materials-Antibodies for Rac and p85 subunit of PI3K were purchased from Upstate Biotechnology. Akt antibodies (phospho-Akt, Ser 473 ; total Akt) were from Cell Signaling technology (Beverly, MA). LTD 4 was purchased from Cayman Chemical Company (Ann Arbor, MI). ZM198615 was a gift from (AstraZeneca, R&D Lund). ECL Western blot detection reagents and Hyperfilms were from Amersham Biosciences (Buckinghamshire, United Kingdom). Wortmannin and LY294002 were from Calbiochem. PTX was obtained from Speywood Pharma Ltd. (Maidenhead, United Kingdom). Peroxidase-linked goat anti-rabbit and mouse IgG originated from Dako A/S (Copenhagen, Denmark). Alexa 546 phalloidin is from Molecular Probes Inc. (Eugene, OR). All other chemicals were of analytical grade and obtained from Sigma.
Cell Culture-The Intestine 407 cell line was isolated from the intestine of a human embryo of ϳ2-month gestation and was successfully maintained in culture without any immortalizing transfections (28). These human intestinal epithelial cells (Int 407; Flow Laboratory), which exhibit typical epithelial morphology and growth (28), were cultured as a monolayer to ϳ80% confluence for 5 days. Cell cultures were kept at 37°C in a humidified atmosphere of 5% CO 2 and 95% air in basal medium Earle's supplemented with 15% newborn calf serum, 55 IU/ml penicillin, and 55 g/ml streptomycin. The cells were regularly tested to ensure the absence of mycoplasma contamination.
cDNAs and Transfections-Cells were transfected with dominant negative p85 of PI3K or dominant-negative (DN) form of Vav2, which lacks an active dbl domain (L342R/L343S) for 6 h and were allowed to grow in serum containing medium for another 24 h. The DN-p85 construct was generously provided by Dr. Arthur Mercurio (Beth Israel Deaconess Medical Center, Boston, MA), whereas the DN-Vav2 construct was generously provided by Dr. Christopher Carpenter (Beth Israel Deaconess Medical Center). Control cells were transfected with empty pEGFP-N1 vector (Clontech). Transient transfections of the cells were achieved using 3.5 l of Lipofectamine (Invitrogen) and 1.8 g of plasmid DNA/ml and were performed in serum-free medium, essentially according to the protocol provided by the supplier. In all of the transfection experiments, it was routinely confirmed that the empty vector had no effect.
GST Pull-down Assays-The cDNA clone encoding the GST fusion protein of the PAK1B binding domain of Rac and Cdc-42 (PAKcrib; amino acids 56 -267) was cloned into the bacterial expression vector pGEX-2T and was expressed in Escherichia coli and cultured at 30°C (29). The expression of GST fusion proteins was induced with 1 mM isopropyl 1-thio-␤-D-galactopyranoside, and the E. coli were subsequently collected by centrifugation at 3500 ϫ g for 15 min followed by sonication in phosphate-buffered saline. Triton X-100 (final concentration 1%) was added to lysate, and particulate matter was removed by centrifuging at 5000 ϫ g for 15 min. The cleared lysate was incubated with glutathione-Sepharose beads (Sigma) for 1 h at 4°C, and the beads were subsequently washed three times with ice-cold PBS. Lysates of unstimulated or stimulated cells were prepared in 1.0 ml of the lysis buffer supplemented with 10 mM MgCl 2 . GST fusion protein pre-bound to Sepharose beads was incubated with 1.0 ml of the cell lysates (1 mg/ml total cell protein) for 1 h at 4°C. Thereafter, the beads were washed once with ice-cold lysis buffer supplemented with 0.5 M NaCl and twice with buffer A (50 mM Tris, pH 7.5, 1 mM EDTA, 1 mM EGTA, 1 mM Na 3 VO 4, 1% Triton X-100, 50 mM NaF, 5 mM sodium pyrophosphate, 10 mM sodium glycerophosphate, 4 g/ml leupeptin, and 30 g/ml phenylmethanesulfonyl-fluoride).
FIG. 2. LTD 4 induces phosphorylation of the p85 regulatory subunit of PI3K and Akt in Int 407 cells. Cells were incubated in the absence or presence of 80 nM LTD 4 for indicated periods of time, after which the cells were lysed. A, the cell lysates were immunoprecipitated (IP) with an anti-phosphotyrosine (PTyr) antibody (4 g) and immunoblotted with an anti-p85 subunit antibody (1:1000). B, the cells were immunoprecipitated with an anti-p85 subunit antibody (5 g) and immunoblotted with an anti-phosphotyrosine antibody (1:5000). These blots were then reprobed with an anti-p85 antibody to confirm equal loading. Representative blots and the accumulated results of a densitometric analysis of the LTD 4 -induced phosphorylation of the p85 subunit are shown. The p85 tyrosine phosphorylation values were calculated as percentages of those seen in unstimulated cells and are given as the means Ϯ S.E. of three separate experiments. C, the cells were preincubated in the absence or presence of 50 M LY294002 (LY) for 30 min before LTD 4 stimulation. The cells then were lysed, and the proteins were separated by SDS-PAGE and immunoblotted with an antibody specific for phosphorylated Akt (Ser 473 ). Thereafter, the blots were stripped and reprobed with an antibody for total Akt. The accumulated results of densitometric analysis of the LTD 4 -induced phosphorylation of Akt are shown. The illustrated blots are representative of at least three separate experiments. All of the statistical significances were evaluated using the unpaired Student's t test. *, p Ͻ 0.05; **, p Ͻ 0.01.
Cell Lysis-Cells were preincubated with inhibitors (G i protein inhibitor PTX (500 ng/ml) for 2 h and PI3K inhibitors wortmannin and LY294002 for 30 min), and stimulations (80 nM LTD 4 ) were terminated by adding ice-cold buffer A. Thereafter, the cells were kept on ice for 30 min in the lysis buffer, and remaining cell debris was scraped loose into the buffer. The lysates were homogenized 10 times on ice in a glass tissue grinder (Dounce) and then centrifuged at 10,000 ϫ g for 15 min. The protein content of each supernatant was measured and compensated for to ensure equal protein loading.
Immunoprecipitation and Immunoblotting-For immunoprecipitation, lysates containing 1 mg/ml protein were incubated with 4 g of the anti-p85, anti-Vav2, or anti-phosphotyrosine antibodies at 4°C for 2 h, after which 20 l of protein G plus agarose was added and incubated for an additional 1 h. The beads then were washed three times in the lysis buffer mentioned above, mixed with sample buffer, boiled for 10 min, and centrifuged, and the proteins are separated on SDS-polyacrylamide gels. The separated proteins were electrophoretically transferred to a polyvinylidene difluoride membrane and were blocked for 1 h with 5% nonfat dried milk at room temperature and then incubated with a primary antibody for 1 h at room temperature or overnight at 4°C. Phospho-Akt 1:500 and 1:1000 dilutions were used for all of the other antibodies. Subsequently, all of the membranes were washed extensively and incubated with a horseradish peroxidase-linked goat antirabbit, anti-sheep, or anti-mouse antibody (1:5000) for 1 h at room temperature. Thereafter, the membrane was washed again extensively, incubated with ECL Western blot detection reagents, and finally exposed to hyperfilm-ECL to visualize immunoreactive proteins.
Immunofluorescence-The cells were seeded on glass coverslips and grown for 5 days. Thereafter, the cells were serum-starved for 2 h and stimulated with LTD 4 . The stimulation was terminated by fixing for 10 min at room temperature in a 3.7% paraformaldehyde and PBS solution, after which the cells were permeabilized in a 0.5% Triton X-100 and PBS solution for 15 min. The coverslips were subsequently washed twice in PBS and incubated at room temperature in a 5% goat serum albumin, 0.39 Triton X-100, and PBS solution for 40 min, and the cells were stained for 1 h with an antibody against Rac. The coverslips then were washed six times in PBS and incubated with Alexa Fluor 568 goat anti-rabbit secondary antibody (diluted 1:200 in 1% goat serum and PBS solution for 1 h). The coverslips were washed thereafter six times in PBS, counterstained with Alexa phalloidin 546, washed, and mounted in fluorescent-mounting medium (Dako A/S). Samples were examined and photographed with a Nikon Eclipse 800 microscope using a Plan-Apo ϫ60 objective. Images were recorded with a scientific grade CCD camera (Hamamatsu, Japan) and subsequently analyzed with HazeBuster deconvolution software (Vay Tek, Inc., Fairland, CT).
Migration Assay-Migration of intestinal epithelial cells was assayed using Boyden chamber with polycarbonate filters of 8-m pore size. The LTD 4 was added to the lower well of the chamber along with serum-free medium, and the filter was placed above. 100 l of cell suspension (1 ϫ 10 5 cells/ml) in serum-free basal medium Earle's was added to the top chamber. In the case of inhibitor treatments, the cells were pretreated with the indicated inhibitors and the respective inhibitors were also added at the concentrations indicated to the upper chamber during the process of cell migration. After incubation at 37°C in 5% CO 2 for 18 h, the filters were disassembled and the upper surface of each filter was scraped free of cells by wiping it with a cotton swab. Cells that had migrated to the outer surface of the filter were fixed with 3.7% paraformaldehyde and stained for 20 min with 0.1% crystal violet in 10% methanol. The dye was eluted with 100 ml of 1% SDS. The absorbance was measured in Nunc Polysorp 96-well plates, and the absorbance at 590 nm was measured using a Fluostar plate reader (BMG Labtechnologies GmbH, Offenberg, Germany). Fig.  1A, LTD 4 could induce migration of intestinal epithelial cells in a concentration-dependent manner, although less prominent than that seen in a colon cancer cell line, Caco-2 (30). The Rac and Cdc42 GTPase have both been implicated in the regulation of migration of different cell types (7). To examine the ability of LTD 4 to activate Rac and Cdc42, we measured the activity of Rac and Cdc42 in cells using the affinity precipitation assay in which the GST-PAKcrib fusion protein binds the active GTP- bound form of Rac or Cdc42. We found that the amount of active Rac-GTP increased in a time-dependent manner upon LTD 4 stimulation of Int 407 cells with a maximum after 15 min (Fig. 1B). Densitometric analysis confirmed the transient activation of Rac with a maximum of 942 Ϯ 118%, 15 min after the addition of LTD 4 . Under similar conditions, we were unable to detect any activation of Cdc42 (Fig. 1B). Whole cell lysates were used to ensure that the total amounts of Rac and Cdc42 did not change upon LTD 4 stimulation (Fig. 1B). As a positive control for the activation of Rac and Cdc42, the cells were stimulated with the epidermal growth factor (EGF) (Fig. 1B). Rac is known to be activated by GEFs such as Vav and Sos. Therefore, we investigated whether Vav2 or Sos was activated by LTD 4 . We could not detect any activation of Sos by LTD 4 in these cells (data not shown); however, Vav2 immunoprecipitated from cells stimulated with LTD 4 for 15 min, the time for maximal activation of Rac, exhibited a distinctly increased tyrosine phosphorylation (Fig. 1C). This tyrosine phosphorylation of Vav2 is indicative of its activation (31). We transfected cells thereafter with a previously described (13) DN form of Vav2 (DN-Vav2) and noted that, in such transfected cells, the LTD 4induced activation of Rac was abolished (Fig. 1D). These results clearly show that the LTD 4 -induced activation of Rac is mediated through Vav2.

LTD 4 Induces Cell Migration and Activation of Vav and Rac in the Intestinal Epithelial Cell Line, Int 407-As shown in
To further analyze the LTD 4 -induced Rac signaling pathway, we investigated whether LTD 4 could activate PI3K in these cells. The activity of this enzyme was first assessed by investigating the degree of tyrosine phosphorylation of p85 subunit of PI3K (15). The cells were stimulated with LTD 4 for indicated periods of time, and immunoprecipitates of tyrosine phosphorylated proteins were then blotted with an anti-p85 antibody ( Fig. 2A) or vice versa (Fig. 2B). As seen in Fig. 2, A and B, LTD 4 caused an increased tyrosine phosphorylation of the p85 regulatory subunit of PI3K. The time kinetics of this activation of PI3K was similar to the LTD 4 -induced activation Rac (Fig. 1B). Secondly, we also investigated the ability of LTD 4 to induce activation of the serine/threonine kinase Akt, a well known downstream target of PI3K activity. Cells stimulated with LTD 4 exhibited a clear activation of Akt that was also confirmed to be sensitive to the PI3K inhibitor LY294002 (Fig. 2C). These results clearly argue for the ability of LTD 4 to cause activation of PI3K in intestinal epithelial Int 407 cells. two different PI3K inhibitors, LY294002 (50 M for 30 min) or wortmannin (100 nM for 30 min) (32). The cells then were stimulated with LTD 4 for 15 min, after which they were lysed. Thereafter, the different lysates were analyzed in the GST-PAKcrib binding assay to assess the activity of Rac. We have reported earlier that LTD 4 can signal via a PTX-sensitive G protein (21,24). Consistent with those results, preincubation with PTX significantly inhibited the LTD 4 -mediated activation of Rac (Fig. 3A). We also observed that preincubation with either of the PI3K inhibitors led to an almost complete block of the LTD 4 -mediated activation of Rac, suggesting that a PI3K signals upstream of Rac in these cells (Fig. 3A). This conclusion was confirmed by transfecting the cells with a DN p85 expressing vector (DN-p85). As shown in Fig. 3B, the expression of DN-p85 in these cells also blocked the LTD 4 -induced activation of Rac.

LTD 4 Mediates PI3K-and Rac-dependent Formation of Membrane Ruffles in Intestinal Epithelial Cells-The Rac
GTPase has previously been shown to induce the formation of lamellipodia and membrane ruffles in different cell types (33,34). Here we show that cells stimulated with LTD 4 formed more membrane ruffles when compared with unstimulated control cells (Fig. 4). The formation of these membrane ruffles was blocked effectively by inhibiting the PI3K/Rac signaling pathway with either wortmannin or LY294002 (Fig. 4). We also examined whether and how LTD 4 influenced the distribution of Rac during the formation of membrane ruffles. Unstimulated control cells immunostained with an anti-Rac antibody exhibited a fairly homogenous staining of Rac throughout the cell (Fig. 5, top left). In contrast, LTD 4 stimulation caused a redistribution of Rac to the actin-containing membrane ruffles (Fig.  5, bottom left and middle). Merge pictures more clearly show the LTD 4 -induced co-localization of Rac and actin in the membrane ruffles (Fig. 5, the two panels to the right). To ascertain that the LTD 4 -induced membrane ruffles are indeed mediated via the CysLT 1 receptor, we also preincubated cells with the CysLT 1 receptor antagonist ZM198615. Alone this antagonist did not affect the number of ruffles of unstimulated cells, 17 Ϯ 2 ruffles/cell in unexposed control cells and 20 Ϯ 2 ruffles/cell in cells incubated with ZM198615 alone. The obtained results clearly show that, in the presence of ZM198615, all of the LTD 4 -induced membrane ruffles were abolished (Fig. 5).
The LTD 4 -induced Migration Is Dependent on the PI3K/Rac Signaling Pathway-Based on the finding that LTD 4 generates the formation of membrane ruffles via a PI3K/Rac signaling pathway, we investigated whether the same pathway was also involved in the LTD 4 -induced migration of intestinal epithelial cells. As shown in Fig. 6A, the intestinal epithelial cell motility initiated by LTD 4 was effectively impaired by PTX and both of the PI3K inhibitors, LY294002 and wortmannin. Finally, a simplified schematic model of the signaling pathway participating in the regulation of LTD 4 -induced motility in intestinal epithelial cell is outlined in Fig. 6B. DISCUSSION The CysLT 1 -receptor has been shown previously to activate both a PTX-sensitive and a PTX-insensitive heterotrimeric G protein signaling pathway (12,21,22). This receptor has also previously been shown to mediate the activation of the small RhoA GTPase via a PTX-insensitive signaling pathway (12). In contrast, here we show that LTD 4 -triggered activation of the small GTPase Rac was inhibited by PTX similar to the LTD 4induced increase in survival and proliferation of intestinal epithelial cells (21,24). Investigation of leukotriene-mediated activation of Rac in other cell types supports the present observation. In HL-60 cells, LTB 4 -induced Rac activation and chemotaxis were blocked by pretreatment with PTX (35), as was LTD 4 -mediated Rac activation in THP-1 cells (22). The notion that two different signaling pathways regulate the activity of RhoA and Rac in intestinal epithelial cell probably provides the CysLT 1 -receptor with a more sophisticated control of epithelial cell motility. Subsequent to these findings, we then sought to identify the signaling molecules downstream of CysLT 1 -receptor activation of a PTX-sensitive heterotrimeric G protein that could be implicated in the activation of Rac. One obvious candidate to look for is the PI3K. Indeed, it has been reported that lipids generated by the PI3K can participate in the regulation of small G proteins as well as cell motility (11,36). We not only found that LTD 4 induced tyrosine phosphorylation of the p85 subunit of PI3K, which was used here as an index of its activation, but that the time kinetics of LTD 4induced phosphorylation of the p85 subunit of PI3K was also parallel to the activation of Rac in intestinal epithelial cells. To further investigate the possibility that PI3K has a role in the LTD 4 -induced activation of Rac, we blocked the PI3K activity by different means and tested the effects of such treatments on the Rac activity. We found that pretreatment of Int 407 cells either with wortmannin or LY294002, two structurally unrelated PI3K inhibitors, and also transfection of these cells with a DN p85 construct halted the LTD 4 -induced activation of Rac. The fact that we also found that LTD 4 had the ability to cause activation of Akt, a well known downstream target of PI3K, further supports our conclusion that LTD 4 causes the activation of PI3K in intestinal epithelial Int 407 cells. In keeping with these results, a requirement of PI3K activity for the agonist-induced activation of Rac has also been reported in other cell systems (5). It is likely that the PI3K dependence in these experiments is related to the fact that the PI3K-generated phosphatidylinositol 3,4,5-trisphosphate lipid enables the activation of a number of Rac-specific GEFs (14). Vav2 seems to be the GEF involved in the activation of Rac by LTD 4 , because not only does it contain a PH domain and is therefore a target of the PI3K lipid product, we also found that Vav2 was tyrosinephosphorylated, indicative of its activation. In addition, the LTD 4 -induced activation of Rac was totally blocked in intestinal epithelial 407 cells transfected with a dominant-negative form of Vav2.
In this study, we found that stimulation of Int 407 cells with LTD 4 led to the formation of circular membrane ruffles rather than the classical Rac-induced lamellipodia (37). This observation was somehow surprising, because circular membrane ruffle formation has mainly been described in response to different growth factors and we have previously shown that stimulation of Int 407 cells with LTD 4 does not result in a transactivation of either EGF or platelet-derived growth factor receptors (30). Although we cannot totally rule out the possibility of a crosstalk between the Cyst-LT 1 -receptor and alternative growth factor receptors, we do believe that our present results reflect a direct effect of the Cys-LT 1 -receptor on the cytoskeleton. Several findings support the view that the activation of PI3K and its downstream effector Rac mediate the LTD 4 -stimulated membrane ruffling in Int 407 cells. First, decreased formation of membrane ruffles was observed in LTD 4 -stimulated Int 407 cells pretreated with LY294002 or wortmannin as well as in cells transiently transfected with a dominant-negative (N17) mutant of Rac. Secondly, we observed that Rac redistributed to a membrane-enriched fraction upon stimulation of the cells with LTD 4 . Thirdly, we found that Rac co-localized with filamentous actin in the membrane ruffles of cells stimulated with LTD 4 . Furthermore, our conclusion is also indirectly supported by the observation that expression of constitutively active Rac1 results in the formation of membrane ruffles in colon cancer cells (37). Furthermore, microinjection of dominant-negative Rac1 constructs into epithelial cells (33) or rat embryo fibroblasts (34) has been shown to inhibit both the formation of lamellipodia and migration of cells into the wound. Activation of distinct Rho family proteins constitutes key signaling events that link ligation of membrane receptors to cytoskeletal reorganization. For the first time, we show here that stimulation of human intestinal epithelial cells with LTD 4 leads to a transient and time-dependent activation of Rac but not Cdc42. This finding is similar to the observation in vascular smooth muscle cells where angiotensin and thrombin stimulation have been shown to result in the activation of Rac but not Cdc42 (38). In contrast, the majority of agonists that cause activation of Rac also cause a simultaneous activation of Cdc42 in their target cells (39,40). In the present study, we could clearly show that the LTD 4 -induced migratory response in intestinal epithelial Int 407 cells was significantly impaired by PTX, LY294002, and wortmannin, all of which have been shown to also block the LTD 4 -mediated activation of Rac. Taken together, our results support the notion that LTD 4 generates the formation of circu-lar membrane ruffles via a PI3K and Rac signaling pathway and that the formation of these ruffles serves as an important driving force for the LTD 4 -induced migration of intestinal epithelial cells.
Cell migration is central to several important biological processes including tissue morphogenesis, wound healing, and cancer cell metastasis (41). Our previous results demonstrating that the pro-inflammatory mediator LTD 4 can induce significant increases in proliferation and survival (27) can now be extended by the present finding that it also increases the motility of these cells, all hallmarks of tumor cell behavior. Consequently, the present results add further support for an important role of LTD 4 as a link between inflammatory bowel diseases and subsequent development of colon cancer.