Oncogene (2007) 26, 4774–4796 & 2007 Nature Publishing Group All rights reserved 0950-9232/07 $30.00 www.nature.com/onc ORIGINAL ARTICLE Ras effector pathways modulate scatter factor-stimulated NF-jB signaling and protection against DNA damage

S Fan1, Q Meng1, JJ Laterra2 and EM Rosen1

1Department of Oncology, Lombardi Comprehensive Cancer Center, Georgetown University, Washington, DC, USA and 2Department of Neurology, The Kennedy Krieger Institute and Johns Hopkins University School of Medicine, Baltimore, MD, USA

Scatter factor (SF) (hepatocyte growth factor) is a Introduction pleiotrophic cytokine that accumulates within tumors in vivo and protects tumor cells against cytotoxicity and Scatter factor (SF) (also called hepatocyte growth factor apoptosis due to DNA damaging agents in vitro. Previous (HGF)) stimulates invasion, proliferation, differentia- studies have established that SF-mediated cell protection tion, angiogenesis and tumorigenesis in different cell involves antiapoptotic signaling from its receptor (c-Met) types under different circumstances.The actions are to PI3 kinase-c-Akt-Pak1 (p21-activated kinase -1)- mediated through its receptor, the c-Met tyrosine kinase NF-jB (nuclear factor-kappa B). Here, we found that (Bottaro et al., 1991). Although SF can stimulate Ras proteins (H-Ras and R-Ras) enhance SF-mediated apoptosis in hepatocarcinoma cells (Matteucci et al., activationof NF- jB and protection of DU-145 and 2003), in most cell types and contexts, SF is cytopro- MDCK (Madin–Darby canine kidney) cells against the tective because of to its ability to stimulate antiapoptotic topoisomerase IIa inhibitor adriamycin. Studies of Ras signaling and enhance cell survival (Frisch and Francis, effector loop mutants and their downstream effectors 1994; Bardelli et al., 1996). SF was found to protect suggest that Ras/PI3 kinase and Ras/Raf1 pathways epithelial, carcinoma and glioma cells against DNA contribute to SF stimulation of NF-jB signaling and cell damaging agents, including adriamycin (ADR, a DNA protection. Further studies revealed that Raf1 positively topoisomerase IIa inhibitor), ionizing radiation, cis- regulates the ability of SF to stimulate NF-jB activity and platinum (a DNA cross-linking agent), and other agents cell protection. The ability of Raf1 to stimulate NF-jB (Fan et al., 1998, 2000, 2001; Bowers et al., 2000). These activity was not due to the classical Raf1-MEK1/2- findings may be significant because expression of SF and ERK1/2 pathway. However, we found that a MEK3/6- c-Met are upregulated in various tumor types (e.g., p38 pathway contributes to SF-mediated activation of breast cancers, gliomas and bladder cancers) (Jin et al., NF-jB. Incontrast, RalA, a target of the Ras/RalGDS 1997; Rosen et al., 1997). Thus, the accumulation of SF pathway negatively regulated the ability of SF to and c-Met might contribute to chemoresistance and stimulate NF-jB activity and cell protection. Ras, Raf1 radioresistance of the tumors. and RalA modulate SF stimulation of NF-jB activity, Studies of the mechanisms of SF protection against inpart, by regulatingI jB kinase (IKK)-b kinase activity. DNA damaging agents have identified some of the These findings suggest that Ras/Raf1/RalA pathways signal transduction pathway components that mediate may converge to modulate NF-jB activationandSF- cell protection.This pathway involves signaling through mediated survival signaling at the IKK complex and/or a phosphatidylinositol-3-kinase (PI3K) and activation of kinase upstream of this complex. c-Akt, a serine/threonine protein kinase (Bowers et al., Oncogene (2007) 26, 4774–4796; doi:10.1038/sj.onc.1210271; 2000; Fan et al., 2000, 2001). The p21-associated kinase- published online 12 February 2007 1 (Pak1) acts downstream of c-Akt to promote cell survival; and nuclear factor-kappa B (NF-kB) and Keywords: scatter factor; Ras; RalA; Raf1; protection; several of its target genes are downstream of these adriamycin kinases in the SF-protection pathway (Fan et al., 2001, 2005).In addition to these survival-promoting proteins, the multi-substrate adapter Grb2-associated binder-1 (Gab1) and the tumor suppressor phosphatase and tensin homolog (PTEN) function upstream of c-Akt to inhibit cell protection by SF (Fan et al., 2001). Previous studies have identified three effector signal- ing pathways through which oncogenic Ras signaling Correspondence: Dr EM Rosen, Department of Oncology, Lombardi can stimulate tumorigenesis and other activities.These Comprehensive Cancer Center, Georgetown University, 3970 include pathways involving PI3K, Raf1 and RalGDS Reservoir Road, NW, Box 571469, Washington, DC 20057-1469, USA. E-mail: [email protected] (Ral guanine nucleotide dissociation stimulator) (Chung Received 27 April 2006; revised 30 October 2006; accepted 14 November et al., 1993; White et al., 1995; Yang et al., 1998; 2006; published online 12 February 2007 Peyssonnaux et al., 2000; McFarlin et al., 2003). Specific Ras pathways modulate SF protection SFanet al 4775 mutations of the ‘effector loop’ of the Ras protein can dominant-negative (DN) H-Ras (RasN17) and R-Ras preserve some interactions with effector proteins, (R-RasN43) reduced SF-stimulated NF-kB activity to whereas disrupting others (de Vos et al., 1988; Milburn 18–32% of the control values (Po0.001). The stimula- et al., 1990), thus enabling studies of one or another tion of NF-kB activity by wt-H-Ras, wt-R-Ras and signaling pathway downstream of Ras.We have R-RasV38 was much smaller in the absence of SF (1.4- investigated the roles of Ras and several pathways to 1.9-fold). RasN17 and RasN43 reduced the basal NF- downstream of Ras in mediating the ability of SF to kB activity to about 25–50% of control (Po0.001); and stimulate NF-kB transcriptional activity and protection the pcDNA3 and pEXV vectors did not alter NF-kB against ADR.These studies identify a stimulatory role activity.Cell transfection efficiencies were D80%, for Raf1 and an inhibitory role for the small GTPase similar to the efficiency numbers found in previous RalA in modulating SF-stimulated NF-kB activity and transient transfection studies (Fan et al., 2001, 2005). cell protection.They have also identified roles for several mitogen-activated protein (MAP) kinases in Ability of mutant Ras proteins to modulate cell these processes. protection by SF To test the effect of Ras proteins on cell protection, DU- 145 cells were transfected with different Ras mutant Results vectors and then assayed for their ability to protect cells against DNA damage caused by a 2 h exposure to the Stimulation of NF-kB signaling by Ras proteins ADR following treatment7SF (100 ng/ml) for 48 h.Cell We tested the effects of wild-type (wt) and mutant Ras viability was determined using 3-(4,5-dimethylthiazol-2- proteins on NF-kB signaling using an NF-kB-Luc yl)-2,5-diphenyltetrazolium bromide (MTT) assays.In reporter to measure NF-kB activity.DU-145 prostate untransfected or empty vector-transfected cells (Figure cancer or Madin–Darby canine kidney (MDCK) epithe- 2a and g), SF protected the cells at all ADR doses tested lial cells were transfected with the indicated vectors, (P 0.001). V12C40Ras gave the largest increase in cell treated7SF (100 ng/ml  24 h) and harvested for luci- o viability in the presence of SF (Figure 2d), followed by ferase assays.Typically, SF gave a 50- to 60-fold V12Ras (Figure 2b).For comparison, the dashed lines increase in NF-kB-Luc activity; and the empty pcDNA3 show results of empty vector transfections.Both Ras vector had no effect on activity.In combination with vectors also increased cell survival in the absence of SF, SF, an activated H-Ras codon 12 mutant (V12Ras) but the increases were smaller than those owing to SF. expression vector yielded a 6.3- to 8-fold higher NF-kB The farnesylation-defective mutant V12S186Ras yielded activity than SF alone (P 0.001, two-tailed t-test) o little or no changes in cell viability with or without SF (Figure 1a).Without SF, V12Ras gave only about 2.2- (Figure 2f), suggesting that farnesylation is required for to 2.5-fold higher activity, suggesting that the effect of modulation of cell survival.V12S35Ras, V12G37Ras Ras alone is relatively modest.A farnesylation-defective and V12E38Ras enhanced cell survival (Figure 2c, e and mutant V12Ras that is unable to activate downstream h, respectively), but to a smaller degree than V12C40Ras effectors (V12S186Ras (Table 1)) gave little or no or V12Ras.The increases in survival were larger with stimulation of NF-kB activity, in the absence or than without SF.Finally, the DN RasN17 abrogated presence of SF. cell protection (Figure 2i), whereas wt-H-Ras enhanced Effector loop mutations alter the ability of Ras to protection by SF (Figure 2j).These results are consistent activate the Raf1 vs PI3K vs RalGDS signaling path- with the NF-kB assays. ways and preferentially activate one pathway but not others.A mutant that activates PI3K but not Raf1 or RalGDS (V12C40Ras) gave slightly more enhancement Raf1 enhances SF stimulation of NF-kB signaling of SF-stimulated NF-kB activity (8.3- to 9-fold) and cell survival (Po0.001) than V12Ras; whereas a mutant that Activated Ras interacts with the serine/threonine kinase selectively activates RalGDS (V12G37Ras) showed Raf1 and causes activation of MEK1/2 (MKK1/2) and reduced ability to enhance SF-induced NF-kB activity ERK1/2 (Moodie et al., 1993; Vojtek et al., 1993). We (3.9- to 4.4-fold, Po0.001) (Figure 1a). Mutants that tested the effects of wild-type and DN Raf1 vectors on selectively activate Raf1 (V12S35Ras and V12E38Ras) stimulation of NF-kB-Luc activity by SF.Here, wt-Raf1 (McFarlin et al., 2003) enhanced SF-induced NF-kB enhanced SF-stimulated NF-kB-Luc activity by about activity (Po0.001) to a somewhat lesser extent (3.9- to fourfold (Po0.001); whereas in the absence of SF, wt- 6-fold) than V12Ras.All mutants gave less stimulation Raf1 gave a smaller (o2-fold) stimulation of NF-kB of NF-kB in the absence of SF (1.5- to 2.2-fold) than in activity (Figure 3a).DN-Raf1 reduced SF-stimulated its presence. NF-kB activity by >90% (Po0.001). We performed a We further found that both wt-H-Ras and wt-R-Ras similar study in the presence of V12Ras, to maximally (the related Ras viral oncogene homolog) enhanced activate Ras signaling.V12Ras stimulated NF- kB-Luc SF-stimulated NF-kB activity (Po0.001), by 5.5- to activity by 2.30- to 2.5-fold in the absence of SF and 5.8-fold and 3.8- to 5.2-fold, respectively (Figure 1b). enhanced the SF-stimulated NF-kB activity by 4- to 4.5- An activated R-Ras mutant (R-RasV38) gave a similar fold (Po0.001) (Figure 3b). In contrast, DN-Raf1 or slightly higher increase in SF-stimulated NF-kB reduced the SF-stimulated NF-kB activity by over activity (Po0.001) than wt-R-Ras. In contrast, 90% and it reduced (SF þ RasV12)-stimulated NF-kB

Oncogene Ras pathways modulate SF protection SFanet al 4776

Figure 1 Ability of Ras mutants to modulate SF stimulation of NF-kB transcriptional activity.( a) Effect of series of Ras mutant proteins on basal and SF-stimulated NF-kB-Luc activity.Subconfluent proliferating DU-145 cells or MDCK cells in 2 cm 2 wells were co-transfected overnight with the indicated Ras vector and NF-kB-Luc reporter using Lipofectamine (0.25 mg plasmid DNA per vector).The wells were washed, postincubated in fresh culture medium 7SF (100 ng/ml) for 24 h and harvested for luciferase assays. Luciferase values are expressed relative to control cells (no vector, 0 SF) and are means7s.e.m.s of four replicate wells. The data shown are representative of at least two independent experiments.( b) Effect of additional Ras proteins on basal and SF-stimulated NF-kB- Luc activity.Assays were performed exactly as described above for panel a, except using a different set of Ras vectors.

activity by over 95% (Po0.001). The empty pcDNA3 DN-Raf1 proteins in DU-145 cells is shown in vector did not affect NF-kB activity.Finally, wt-Raf1 Figure 3d.These results suggest that Raf1 signaling is enhanced, whereas DN-Raf1 blocked SF-stimulated cell required for SF stimulation of NF-kB activity and survival (Po0.001) (Figure 3c). The effects of wt-Raf1 protection against ADR. on survival were smaller without SF than with SF, indicating that exogenous Raf1 enhances protection by SF.DN-Raf1 had little effect on cell survival in the RalA inhibits SF stimulation of NF-kB signaling absence of SF, but attenuated the SF-mediated protec- and cell survival tion, suggesting that endogenous Raf1 contributes to SF In NF-kB-Luc reporter assays, a wt-RalA vector protection against ADR.Expression of the wt-Raf1 and reduced the basal activity and inhibited SF stimulation

Oncogene Ras pathways modulate SF protection SFanet al 4777 Table 1 Ras, Raf1, Ral, Akt, PTEN and MKK expression vectors utilized in this study Vector Description Source of vector (Reference)

Ras vectors V12Ras Constitutively active codon 12 H-Ras mutant (G12S) in J Downward, ICRF, London, UK (Kraus et al., 1984) pcDNA3 V12S186Ras Farnesylation-defective mutant (C186S) in pcDNA3 vector J Downward, ICRF, London, UK (Fitzgerald, 2000) V12S35Ras Retains partial activation of Raf1-MEK-ERK pathway J Downward, ICRF, London, UK (Webb et al., 1998) (T35S) V12C40Ras Effector loop mutant; preferentially binds PI3K (Y40C) J Downward, ICRF, London, UK (Rodriguez-Viciana et al., 1997) V12G37Ras Effector loop mutant preferentially binds RalGDS (E37G) J Downward, ICRF, London, UK (Murai et al., 1997) V12E38Ras Effector loop mutant; selectively activates Raf1 pathway J Downward, ICRF, London, UK (McFarlin et al., 2003) wt-R-Ras Wild-type related Ras viral oncogene homolog in pEXV A Hall, Institute for Cancer Research, London (Mochizuki vector et al., 2000) R-Ras V38 Constitutively active R-Ras mutant A Hall, Institute for Cancer Research, London, UK R-Ras N43 Dominant-negative R-Ras mutant A Hall, Institute for Cancer Research, London, UK wt-H-Ras Wild-type Harvey rat sarcoma viral oncogene homolog J Downward, ICRF, London, UK (Kraus et al., 1984) RasN17 Dominant-negative H-Ras mutant RG Pestell, Thomas Jefferson University, Philadelphia, PA, USA

Raf1 vectors wtRaf1 Wild-type Raf1 in pcDNA3 vector JH Yoon, Yonsei University, Seoul, Korea (Song et al., 2003) DN-Raf1 Dominant-negative (kinase dead) mutant in pcDNA3 JH Yoon, Yonsei University, Seoul, Korea (Song et al., 2003) vector

Ral vectors wt-RalA Wild-type RalA in pRK5 expression vector M Schmidt, Universitatsklinikum Essen, Germany (Voss et al., 1999) DN-RalA Dominant-negative RalA mutant (RalAG26A) in pRK5 M Schmidt, Universitatsklinikum Essen, Germany (Voss et al., vector 1999) wt-RalB Wild-type RalB in pcDNA3 vector Guthrie cDNA Resource Center, Sayre, PA, USA (Oxford et al., 2005)

Akt vectors DN-Akt Dominant-negative kinase-dead mutant (K179A) in pCIS2 MJ Quon, NHLBI, NIH, Bethesda, MD, USA (Cong et al., vector 1997)

PTEN vectors wt-PTEN Wild-type human PTEN in pFLAG-CMV vector M Georgescu, Rockefeller University, New York (Zhong et al., 2000) mut-PTEN Phosphatase-defective mutant PTEN (C124S) in M Georgescu, Rockefeller University, New York (Zhong et al., pFLAG-CMV 2000)

MKK vectors wt-MKK1 Wild-type human MKK1 (MEK1) expression vector DJ Templeton, University of Virginia, Charlottesville, VA, USA DN-MKK1 Dominant-negative mutant MKK1 (S218/222A) DJ Templeton, University of Virginia, Charlottesville, VA, USA wt-MKK2 Wild-type human MKK2 (MEK2) expression vector R Davis, University of Massachusetts Medical School, Worcester, MA, USA DN-MKK2 Dominant-negative mutant MKK2 expression vector R Davis, University of Massachusetts Medical School, Worcester, MA, USA wtMKK3 Wild-type human MKK3 expression vector R Davis, University of Massachusetts, Worcester, MA, USA (Raingeaud et al., 1996) ca-MKK3 Constitutively active mutant MKK3 (S189E/T193E) R Davis, University of Massachusetts, Worcester, MA, USA (Raingeaud et al., 1996) DN-MKK3 Dominant-negative mutant MKK3 (S189A/T193A) R Davis, University of Massachusetts, Worcester, MA, USA (Raingeaud et al., 1996) wt-MKK4 Wild-type human MKK4 expression vector DJ Templeton, University of Virginia (Yan and Templeton, 1994) DN-MKK4 Dominant-negative mutant MKK4 (S220A/T224L) DJ Templeton, University of Virginia (Yan and Templeton, 1994) wt-MKK5 Wild-type human MKK3 expression vector R Davis, University of Massachusetts Medical School, Worcester, MA, USA DN-MKK5 Dominant-negative mutant MKK5 R Davis, University of Massachusetts Medical School, Worcester, MA, USA wtMKK6 Wild-type human MKK6 expression vector R Davis, University of Massachusetts, Worcester, MA, USA (Raingeaud et al., 1996) ca-MKK6 Constitutively active mutant MKK6 (K82E) R Davis, University of Massachusetts, Worcester, MA USA (Raingeaud et al., 1996) DN-MKK6 Dominant-negative mutant MKK6 (K82A) R Davis, University of Massachusetts, Worcester, MA, USA (Raingeaud et al., 1996)

Abbreviations: MKK, MAP kinase kinase; PTEN, phosphatase and tensin homolog.

Oncogene Ras pathways modulate SF protection SFanet al 4778

Figure 2 Effect of Ras proteins on SF protection against ADR.Proliferating DU-145 cells in six-well dishes were transfected overnight with the indicated Ras vector (5 mg of plasmid DNA per well), washed and allowed to recover for several hours.The cells were then harvested, plated into 96-well dishes, allowed to attach overnight, incubated7SF (100 ng/ml  48 h), exposed to different doses of ADR (2 h at 371C) and postincubated for 48 h in fresh drug-free medium.The cells were then analysed for cell viability using MTT assays.Cell viability values are expressed relative to non-ADR-treated control cells and represent means 7s.e.m.s of 10-replicate wells.The results shown are representative of at least two independent experiments.The Ras vectors tested were: empty pcDNA3 vector (a), V12Ras (b), V12S35Ras (c), V12C40Ras (d), V12G37Ras (e), V12S186Ras (f), empty pSVE vector (g), V12E38Ras (h), RasN17 (i) and wt-H-Ras (j).The dashed lines represent empty vector-transfected cells.

of NF-kB activity by about two-thirds (Po0.001) immunoprecipitated (IP’d).Western blotting showed an (Figure 4a).Conversely, a DN-RalA- enhanced SF- increase of tyrosine-phosphorylated c-Met following SF stimulated NF-kB activity (Po0.001), but had a modest treatment (Figure 4d).In addition, RalA, Raf1 and Ras effect on the basal NF-kB activity.RasV12 stimulated were found in the c-Met immunoprecipitations (IPs), the SF-induced NF-kB activity by 3.5- to 4-fold; and wt- with no obvious differences owing to SF treatment (but RalA attenuated the (RasV12 þ SF)-stimulated NF-kB see below).Finally, we tested the ability of wt-RalB to activity by >5-fold (Po0.001). In contrast, DN-RalA modulate the SF stimulation of NF-kB activity.wt-RalB enhanced the (RasV12 þ SF)-stimulated NF-kB activity caused a reduction in SF stimulated NF-kB activity, but to by nearly twofold (Po0.001). Without SF, the empty a slightly lesser extent than RalA (Figure 4e).These pcDNA3 vector and various expression vectors yielded findings suggest that the Ral GTPases inhibit SF-mediated modest or no changes in NF-kB activity.Consistent NF-kB signaling and cell protection. with its ability to inhibit NF-kB signaling, wt-RalA abrogated the SF protection of DU-145 cells against ADR (Figure 4b).The expression of wt-RalA and DN- Ability of Ras, Raf1 and RalA to modulate SF stimulation RalA was confirmed by Western blotting (Figure 4c). of IKK-b kinase activity To test if RalA associates with the SF receptor, DU- NF-kB activity is inhibited by IkB proteins which 145 cells were treated7SF for 20 min; and c-Met was maintain p65RelA within the cytoplasm.These proteins

Oncogene Ras pathways modulate SF protection SFanet al 4779

Figure 3 Raf1 enhances SF-mediated stimulation of NF-kB transcriptional activity and cell survival.( a) Effect of wtRaf1 and DN- Raf1 on NF-kB-Luc activity in DU-145 and MDCK cells.NF- kB activity was determined as above (Figure 1a), except using Raf1 rather than Ras expression vectors.( b) Effect of wtRaf1 and DN-Raf1 on (SF þ RasV12) stimulation of NF-kB-Luc activity.Assays were performed as described above for panel a.( c) Raf1 enhances protection against ADR by SF.Cell protection assays were performed as described in Figure 2 legend, except using Raf1 instead of Ras vectors.The dashed lines represent empty pcDNA3 vector- transfected cells.( d) Western blot showing expression of wtRaf1 and DN-Raf1.DU-145 cells were transfected overnight with the indicated vector (15 mg plasmid DNA per 100 mm dish), washed and postincubated for 24 h to allow gene expression.The cells were then harvested and Western blotted to detect Raf1 or actin (control for loading and transfer).

(IkB-a and IkB-b) are regulated by serine phosphoryla- (Figure 5a).The combination (V12Ras þ SF) caused tions, which mark them for ubiquitin-mediated degra- loss of detectable IkB-a protein, suggesting that V12Ras dation.A key mediator of I kB phosphorylation is enhances the ability of SF to induce IkB-a degradation. the IkB kinase (IKK) complex (Woronicz et al., 1997; To test the effect of Ras mutants on IKK phosphory- Hu and Wang, 1998; May et al., 2000), which contains lation, DU-145 cells were transfected with different IKK-a, IKK-b and IKK-g (NEMO).One measure of Ras vectors, treated with SF for 20 min, and Western IKK activation is phosphorylation on serine of the blotted for phosphorylated and total IKK proteins. IKK-a/b proteins.First, we tested the effect of SF Each mutant Ras protein was expressed (Figure 5b).SF and V12Ras on IkB-a degradation.In both DU-145 and increased the phosphorylation on IKK-a/b; and with SF MDCK cells, SF treatment for 20 min caused a large present, V12Ras and the effector loop mutants (but not reduction of IkB-a protein; whereas V12Ras alone the farnesylation-defective mutant V12S186Ras) further caused only a modest reduction of IkB-a protein enhanced the phosphorylation of IKK-a/b, compared

Oncogene Ras pathways modulate SF protection SFanet al 4780 with the pcDNA3 vector control (Figure 5b).Several IKK-b or actin protein levels.In the absence of SF, experiments gave similar results, with some experiments there was no detectable phosphorylation of IKK-a/b in showing no basal phosphorylation of IKK-a/b in control or V12S186 Ras-transfected cells (Figure 5c). control cells in the absence of SF.As controls, there Phosphorylation on IKK-a/b was detected in the were no consistent differences in total IKK-a, total other V12Ras-transfected cells, suggesting that the

Oncogene Ras pathways modulate SF protection SFanet al 4781 phosphorylation can be mediated by activated Ras MKK4 is a dual-specificity kinase that activates the proteins in the absence of SF. stress-activated kinases c-jun NH2-terminal kinase We tested the effect of Ras, Raf1 and RalA proteins and p38 (Hog1), but not ERK1/2 and is unresponsive on IKK-b kinase activity in DU-145 cells using an assay to Raf1 (Lin et al., 1995). wt-MKK4 caused a modest kit that measures the ability of the sample to phos- increase in SF-stimulated NF-kB activity (D20%); phorylate a biotinylated IkB-a substrate peptide (see whereas DN-MKK4 caused a modest decrease Materials and methods).Here, SF-stimulated kinase in activity (D20%) (data not shown).MKK5 is an activity by about fivefold; and the activated Ras proteins upstream activator of ERK5 (MAPK7) (Mody et al., V12Ras, V12C40Ras, V12S35Ras, V12G37Ras and 2003).Neither wt-MKK5 nor DN-MKK5 altered V12E38Ras- enhanced SF-stimulated kinase activity the basal or SF-stimulated NF-kB activity (data not (Po0.001) by five- to sevenfold (Figure 5d). shown). V12S186Ras and the empty pcDNA3 vector did not In contrast, MKK3 and MKK6 (both of which alter IKK activity.In the absence of SF, the active activate p38 (Derijard et al., 1995; Wang et al., 1997)) RasV12 mutants showed modest (p1.6-fold) stimula- strongly regulated NF-kB activity (Figure 6a and b). tion of IKK-b kinase activity.wt-R-Ras, the activated The wt and ca mutant MKK3 and MKK6 proteins mutant R-RasV38 and wt-H-Ras- enhanced SF-stimu- stimulated basal NF-kB-Luc activity by about three- to lated IKK-b activity by five- to sixfold (Po0.001) fourfold (Po0.001) and enhanced the SF-stimulated (Figure 5e).In contrast, the DN R-RasN43 and RasN17 NF-kB-Luc activity by 3.5- to 5-fold (Po0.001). proteins blocked SF-stimulated IKK-b activity Conversely, DN-MKK3 and DN-MKK6 reduced SF- (Po0.001), suggesting that endogenous Ras contributes stimulated NF-kB activity (Po0.001) to about 15% of to SF stimulation of IKK-b activity. control.A selective inhibitor of p38 kinase activity, In further studies, wt-Raf1 enhanced the SF-stimu- SB202190 (10 mM), reduced SF-stimulated NF-kB activ- lated IKK-b kinase activity by 4.8-fold (Po0.001); but ity by 50–70% and abrogated the ability of MKK3 and it had a modest effect on the kinase activity in the MKK6 to stimulate NF-kB activity in the presence or absence of SF (about twofold stimulation) (Figure 5f). absence of SF (Figure 6c and d).To confirm a role for In contrast, DN-Raf1 abrogated SF stimulation of the endogenous MKK3 and MKK6 proteins in NF-kB IKK-b kinase activity (Po0.001). Finally, wt-RalA signaling, DU-145 cells were treated with MKK3 small inhibited the SF-stimulated kinase activity (Po0.001); interfering RNA (siRNA), MKK6-siRNA or control- whereas DN-RalA enhanced SF-stimulated kinase siRNA (50 nM Â 48 h) before testing for SF-stimulated activity by another fourfold (Po0.001) (Figure 5g). NF-kB activity.The MKK3 and MKK6 siRNAs These findings suggest that Ras and Raf1 positively reduced the SF-stimulated NF-kB-Luc activity to regulate, whereas RalA negatively regulates the ability D30–40% of that of cells treated with control-siRNA of SF to stimulate IKK-b kinase activity.Thus, the or no siRNA (Po0.01) (Figure 6e). These findings results of the IKK-b activity assays correlated well with suggest that in DU-145 cells, NF-kB activity is regulated the NF-kB-transcriptional assays. by an MKK3/6-p38 pathway, but not by the Raf1- MEK1/2-ERK1/2 pathway. We tested the role of p38 in mediating the enhanced Role of MAP kinases in SF stimulation of NF-kB activity SF-stimulated IKK-b kinase activity because of V12Ras We tested the role of various MAP kinase pathways in in DU-145 cells, using the p38 inhibitor SB202190. the SF stimulation of NF-kB activity, using expression Here, we used a higher concentration of SB202190 vectors for wt, constitutively active (ca) and DN (50 mM) than in the NF-kB-Luc assay (10 mM), to ensure MAP kinase kinase proteins.MKK1 and MKK2 maximal inhibition of p38.As before, SF caused a activate the classic extracellular signal-regulated kinases fivefold increase in IKK-b activity; and (SF þ V12Ras) (ERK1 and ERK2) (Zheng and Guan, 1993).Neither caused another fivefold increase beyond that owing to wt-MKK1 nor DN-MKK1 nor wt-MKK2 nor DN- SF alone or (SF þ pcDNA3) (Figure 6f).SB202190 MKK2 had any significant effect on NF-kB-Luc activity abrogated the increase in IKK-b activity owing to SF in DU-145 cells without or with SF present, even though alone and the further increase in IKK-b activity owing the proteins were well expressed (data not shown). to V12Ras.Thus, IKK- b activity for (SF þ V12

Figure 4 RalA inhibits SF-mediated stimulation of NF-kB signaling and cell survival.( a) Effect of wtRalA on NF-kB-Luc activity. Proliferating cells in 2 cm2 wells were transiently co-transfected with indicated vectors overnight, washed, incubated in fresh medium without or with SF (100 ng/ml) for 24 h and harvested for luciferase assays.The luciferase values are expressed relative to control-treated cells (no vector, 0 SF) and are means7s.e.m.s of four replicate wells. (b) RalA inhibits SF protection against ADR.Cell protection assays were performed as described in Figure 2 legend, except that RalA vectors were used instead of Ras vectors.The dashed lines represent empty pcDNA3 vector- transfected cells.( c) Western blot showing expression of RalA and DN-RalA.Proliferating DU-145 cells were transfected overnight with the indicated vector (15 mg plasmid DNA per 100 mm dish), washed and postincubated for 24 h to allow gene expression.The cells were then harvested and Western blotted for RalA or actin.( d) RalA interacts with c-Met receptor in a SF-independent manner.DU-145 cells in 100 mm dishes were incubated in serum-free medium7SF (100 ng/ml) for 20 min; and whole-cell lysates were subjected to IP using an anti-c-Met antibody (see Materials and methods section).Precipitated proteins were analysed by SDS–PAGE and blotted using antibodies for total c-Met, phospho-c-Met (tyrosine-1349), RalA, Raf1 and Ras.( e) Effect of RalB on SF stimulation of NF-kB-Luc activity.Assays were performed as described for panel a, except that a wild-type RalB expression vector was used instead of RalA.

Oncogene Ras pathways modulate SF protection SFanet al 4782 Ras þ SB202190) was slightly less than that for SF with different mutants or with V12Ras itself and tested alone.The ability of the siRNAs to knock down MKK3 for association of the Ras proteins with p110a (catalytic and MKK6 proteins is shown in Figure 6g.These subunit of PI3K), Raf1 and RalGDS.For this purpose, findings suggest that the ability of V12Ras to enhance we constructed FLAG-tagged Ras mutant expression the SF-stimulated IKK-b activity in DU-145 cells is vectors to facilitate IP of the Ras proteins.By IP- dependent upon p38. Western blotting, V12Ras associated with all three effector proteins; whereas V12C40Ras associated only with p110a.V12S35Ras and V12E38Ras interacted only Ras mutants stimulate c-Akt activity with Raf1; and V12G37Ras interacted only with We showed that c-Akt signaling is required for SF RalGDS (Figure 7c).Thus, the mutant Ras proteins stimulation of NF-kB activity and protection against behave as expected. ADR (Fan et al., 2001, 2005). Here, we tested the ability Finally, we studied the role of c-Akt in regulating the of Ras mutant proteins to activate c-Akt, indicated ability of V12Ras to enhance SF-stimulated NF-kB by phosphorylation on serine-473 and threonine-309.A activity.In the absence of V12Ras, a kinase-dead 20 min exposure to SF-stimulated Akt phosphorylation DN-Akt vector blocked the SF-stimulated NF-kB on both residues in empty vector-transfected cells (Figure 7a).V12Ras and all V12Ras effector loop mutants (but not the farnesylation-defective mutant V12S186Ras) enhanced Akt phosphorylation in SF- stimulated cells; whereas there were no significant changes in total Akt or actin protein.V12Ras and the effector loop mutants (but not V12S186Ras) also induced Akt phosphorylation in the absence of SF (Figure 7b).There were no consistently observed differences among these mutants (except V12S186Ras) in the ability to activate Akt. Thus, V12Ras mutant proteins that selectively acti- vate three effector pathways (PI3K, Raf1 or RalGDS) each enhanced SF-stimulated Akt phosphorylation, IKK-b kinase activity, NF-kB activity and cell survival. The finding that a Ras mutant that selectively activates RalGDS (V12G37Ras) has these activities was surpris- ing, because RalA and RalB, two major targets of RalGDS, inhibited the same SF-stimulated activities. We tested our Ras effector loop mutants to confirm that they behave as expected.DU-145 cells were transfected

Figure 5 Impact of Ras, Raf1 and RalA on SF-mediated stimulation of IKK-b kinase activity.( a) Ability of SF and V12Ras to induce degradation of IkB-a.Proliferating cells in 100 mm dishes were transfected with V12Ras (15 mg per dish) or sham-treated for 24 h.The cells were then treated 7SF (100 ng/ml) in serum-free DMEM for 20 min, harvested and subjected to Western blotting for IkB-a or actin.( b and c) Ability of Ras mutants to modulate phosphorylation of IKK complex.Proliferat- ing DU-145 cells were transfected with the indicated Ras vectors and then treated with SF (100 ng/ml) (b) or without SF (c) for 20 min.The cells were harvested and subjected to Western blotting for Ras, phospho-IKK-a/b, total IKK-a, total IKK-b or actin.( d) Ability of Ras mutant proteins to modulate SF stimulation of IKK- b activity.DU-145 cells were transfected with different Ras vectors and treated7SF, as described above.Cell lysates were prepared; and IKK-b kinase activity was determined using a commercial assay kit (see Materials and methods section).The IKK- b kinase activity values are means7s.e.m.s of five replicate wells expressed as a percentage of the control value (no SF, mock transfection).( e) Ability of additional Ras mutant proteins to modulate SF stimulation of IKK-b activity.Assays were performed as described in panel d.( f) Effect of wtRaf1 and DN-Raf1 on IKK-b kinase activity in DU-145 cells.Assays were performed as described in panel d, except that Raf1 vectors were used instead of Ras vectors. (g) Effect of wtRalA and DN-RalA on IKK-b kinase activity in DU-145 cells.Assays were performed as described in panel d, except that RalA vectors were used instead of Ras vectors.

Oncogene Ras pathways modulate SF protection SFanet al 4783

Figure 5 Continued.

Oncogene Ras pathways modulate SF protection SFanet al 4784

Figure 6 Role of MAP kinase pathways in SF stimulation of NF-kB transcriptional activity.DU-145 cells were co-transfected overnight with the NF-kB-Luc reporter and the indicated expression vectors for wild-type or mutant MKK (mitogen-activated protein kinase kinase).The cells were then washed, treated7SF (100 ng/ml  24 h) and assayed for luciferase activity.Data are shown for MKK3 ( a) and MKK6 (b).( c and d) Assays were carried out as above using MKK3 (c) and MKK6 (d) vectors in the absence or presence of the selective p38 inhibitor SB202190 (10 mM).Note: wt, wild-type; DN, dominant negative; ca, constitutively active.( e) the cells were pretreated with MKK3-siRNA or MKK6-siRNA, control-siRNA (50 nM  48 h) or no siRNA; transfected with the NF-kB-Luc reporter; treated7SF (100 ng/ml  24 h); and assayed for luciferase activity.( f) DU- 145 cells were transfected with V12Ras vs empty pcDNA3 vector, pre-incubated7SB202190 (50 mM), treated7SF (100 ng/ml  20 min) and then assayed for IKK-b kinase activity.( g) Ability of the MKK3 and MKK6 siRNAs to knock down the endogenous MKK3 and MKK6 proteins, respectively, after a 48 h treatment with 50 nM of gene-specific or control siRNAs.

Oncogene Ras pathways modulate SF protection SFanet al 4785

Figure 6 Continued. activity (Po0.001) (Figure 7d). As before, V12Ras regulator of c-Akt activity), but not a mutant phospha- further enhanced the SF-stimulated NF-kB activity by tase-defective PTEN, inhibited basal, SF-stimulated and five- to sixfold.However, DN-Akt nearly abrogated the (SF þ V12Ras)-stimulated NF-kB activity (Po0.001) (SF þ V12Ras)-stimulated NF-kB activity (Po0.001). (Figure 7e).These findings suggest that c-Akt is required DN-Akt also inhibited the basal NF-kB activity in the for basal and SF/Ras-stimulated NF-kB activity in absence of SF (Po0.001); and wt-PTEN (a negative DU-145 and MDCK cells.

Oncogene Ras pathways modulate SF protection SFanet al 4786 Effects of Raf1 and RalA knock down on NF-kB and V12Ras-transfected cells compared to pcDNA3-trans- IKK-b activity fected cells (Figure 9a).SF increased the phospho-Raf1 Consistent with DN-Raf1 experiments, treatment of (serine-43) levels in both pcDNA3 and V12Ras-trans- DU-145 and MDCK cells with Raf1-siRNA (but not a fected cells; but the SF-induced increase in phospho- control-siRNA) reduced the basal and SF-stimulated Raf1 was greater in V12Ras-transfected cells.V12Ras NF-kB-Luc activity (Po0.001) (Figure 8a). And like DN-RalA, RalA-siRNA increased the basal and SF- stimulated NF-kB-Luc activity cell types (Po0.001) (Figure 8b).The extent of the increase owing to RalA- siRNA was greater in the presence (>3-fold) than in the absence (twofold) of SF.Similar to the transcriptional assays, Raf1-siRNA (but not control-siRNA) strongly reduced the SF-stimulated (and basal) IKK-b kinase activity (Po0.001) (Figure 8c); and RalA-siRNA enhanced basal and SF-stimulated IKK-b kinase activ- ity (Po0.001) (Figure 8d). Finally, we tested the effects of Raf1 or RalA on IKK-b kinase activity in cells treated with (SF þ V12Ras).SF alone gave a fivefold increase in kinase activity and (SF þ V12Ras) gave a further fivefold increase beyond that of SF alone (Figure 8e).DN-Raf1 abrogated the increased IKK- b kinase activity owing to (SF þ V12Ras); whereas wt- RalA reduced the kinase activity in the presence of (SF þ V12Ras) to 55% of that found in the absence of wt-RalA (Po0.001). The ability of the siRNAs to knock down Raf1 and RalA protein is shown in Figure 8f. Thus, DN-Raf1 was more effective than wt-RalA in modulating the stimulatory effect of the combination of (SF þ V12Ras).

Raf1 and IKK phosphorylation in the c-Met complex We tested the effect of SF and V12Ras on the phosphorylation (a measure of activation) of Raf1. DU-145 cells were transfected with V12Ras vs pcDNA3 vector; treated7SF (20 min) and subjected to IP- Western blotting using an antiRaf1 antibody.The total levels of Raf1 and c-Met protein associated with Raf1 were unaffected by these treatments; but the phospho- rylated c-Met associated with Raf1 was increased in

Figure 7 Ability of Ras mutants to stimulate c-Akt activity.( a and b) Ability of mutant Ras proteins to stimulate Akt phosphorylation. DU-145 cells were transfected with the indicated vectors and then treated with SF (100 ng/ml) (a) or without SF (b) in serum-free DMEM for 20 min.Western blotting of whole-cell lysates was carried out using antibodies against phospho-Akt (serine-473 or threonine-309), total Akt or actin.( c) Interaction of effector loop mutant Ras proteins with effector proteins.DU-145 cells were transfected with the indicated pCMV-Tag2 Ras mutant expression vector, washed, postincubated for 24 h to allow gene expression and harvested.Aliquots of cell lysates (800 mg of total cell protein) were IP’d using an anti-FLAG antibody or non-immune IgG (negative control) (see Materials and methods section).The IPs were Western blotted using an anti-FLAG antibody (to detect transfected Ras) or antibodies against the Ras effectors p110a, Raf1 and RalGDS.A lane corresponding to unprecipitated cell lysate (50 mg) from untransfected control cells is also shown.( d) DN-Akt blocks (SF þ RasV12)-mediated stimulation of NF-kB transcriptional activity.Assays were performed as described in Figure 1a legend. (e) PTEN blocks (SF þ V12Ras)-mediated stimulation of NF-kB activity.Assays were performed as described in Figure 1a legend.

Oncogene Ras pathways modulate SF protection SFanet al 4787

Figure 7 Continued. had little effect on Raf1 phosphorylation without SF; and c-Met.Phospho-IKK- a/b associated with Raf1 and but with SF present, V12Ras increased the amount c-Met in a SF- and V12Ras-dependent manner: SF of phospho-Met associated with Raf1.In a similar increased the amount of phospho-IKK-a/b in Raf1 and experiment using an anti-c-Met antibody for IP, SF c-Met IPs; whereas V12Ras further increased the increased the levels of phospho-Met; and V12Ras phospho-IKK-a/b levels in these IPs in SF-treated cells. further enhanced SF-induced c-Met phosphorylation Finally, we tested the effect of Raf1-siRNA (Figure 9c) (Figure 9b).There were no changes in the total Raf1, or RalA-siRNA (Figure 9d) on the ability of SF but there was more phospho-Raf1 associated with c-Met to induce IKK-a/b phosphorylation.As expected, in cells treated with (SF þ V12Ras) as compared with SF stimulated IKK-a/b phosphorylation, but did not (SF þ pcDNA3) (Figure 9b).These findings suggest that significantly alter total IKK-a, IKK-b or actin levels. activated c-Met may mediate SF-induced Raf1 phos- Knock down of Raf1 blocked the SF-induced IKK-a/b phorylation and that V12Ras enhances SF-induced phosphorylation; whereas knock down of RalA en- Raf1 phosphorylation in the signaling complex. hanced the SF-stimulated phosphorylation.These In the same experiments, RalA associated with both findings suggest that endogenous Raf1 and RalA, each Raf1 and c-Met, in a SF- and V12Ras-independent of which associates with c-Met, modulate the ability of manner.IKK- a and IKK-b also associated with Raf1 SF to induce IKK-a/b phosphorylation.

Oncogene Ras pathways modulate SF protection SFanet al 4788 V12Ras enhances SF-induced p65 nuclear translocation exposure to SF, by immunofluorescence microscopy. and NF-kB target gene expression In basal conditions, cells transfected with V12Ras or We had shown that SF causes rapid translocation of the pcDNA3 vector showed a predominantly cytoplasmic p65 subunit of NF-kB from cytoplasm to nucleus (Fan distribution of p65 (Figure 10a).In response to SF, et al., 2005). Here, we tested the effect of V12Ras on p65 become mostly localized in the nucleus.Consistent nuclear translocation of p65 after a brief (20 min) with the ability of V12Ras to stimulate NF-kB

Figure 8 Effects of knock down of Raf1 and RalA on NF-kB and IKK-b activity.( a) Effect of Raf1-siRNA on NF-kB-transcriptional activity.Cells were pretreated with Raf1-siRNA vs control-siRNA (50 n M) vs no siRNA (vehicle only) for 48 h.The cells were then transfected with the NF-kB-Luc reporter, treated7SF (100 ng/ml) for 24 h and harvested for luciferase assays.The relative luciferase values are means7s.e.m.s of four replicate wells. (b) Effect of RalA-siRNA on NF-kB activity.Experiments were performed as described above, except using RalA-siRNA rather than Raf1-siRNA.( c) Effect of Raf1-siRNA on IKK-b kinase activity.DU-145 cells were pre-treated Raf1-siRNA vs control-siRNA (50 nM) vs no siRNA (vehicle only) for 48 h, then treated7SF (100 ng/ml) for 20 min and harvested for IKK-b kinase activity assays.The values shown are means 7s.e.m.s of five-replicate wells, expressed relative to untreated control cells ( ¼ 100%).( d) Effect of RalA- siRNA on IKK-b kinase activity.Experiments were performed as described in panel c, except using RalA-siRNA rather than Raf1-siRNA. (e) Effects of Raf1 and RalA on IKK-b kinase activity stimulated by (SF þ V12 Ras).DU-145 cells were transfected with the indicated expression vectors; treated7SF (100 ng/ml) for 20 min and assayed for IKK-b kinase activity as above.( f) Knock down of Raf1 and RalA protein levels by the siRNAs.Cells were treated with the indicated siRNAs (50 n M Â 48 h) followed by Western blotting.

Oncogene Ras pathways modulate SF protection SFanet al 4789

Figure 8 Continued. transcriptional activity, V12Ras-transfected cells showed Discussion higher levels of immunoreactive p65 in the nucleus than pcDNA3-transfected cells.The ability of SF to cause These studies identify a role for p21Ras and several ‘scattering’ (cell dispersion and a fibroblastic appear- of its downstream effectors (Raf1 and RalA) in ance) of DU-145 cells is illustrated in the phase contrast modulating the ability of SF to stimulate IKK-b kinase micrographs in Figure 10a.We have also shown that SF activity, NF-kB transcriptional activity and protection stimulates the expression of several known NF-kB against ADR-induced DNA damage.In particular, Ras target genes (cIAP1, cIAP2 and TRAF2) in an NF- and Raf1 signaling enhanced SF stimulation of IKK-b kB-dependent manner (Fan et al., 2005). Here, we tested activity, NF-kB activity and cell protection, whereas the whether V12Ras could enhance the expression of several RalA inhibited these activities.The ability of activated of these genes in SF-treated cells.The ability of SF to Ras (RasV12) to enhance SF-stimulated NF-kB signal- induce increased protein levels of cIAP2 and TRAF2 ing was dependent upon signaling downstream of c-Akt was greater in V12Ras-transfected cells than in and Raf1; and it was negatively regulated by RalA.In pcDNA3-transfected cells (Figure 10b).As expected, addition, SF-stimulated NF-kB signaling was dependent the total Ras levels were significantly greater in the upon endogenous Ras, Raf1 and c-Akt and was V12Ras-transfected cells.In contrast, there were no negatively regulated by endogenous RalA.Finally, differences in the levels of c-Met or actin.These findings although Raf1 was required for SF-stimulated NF-kB are consistent with the above studies showing that activation (and cell protection), the Raf1-MEK1/2- V12Ras enhances the SF-stimulated NF-kB transcrip- ERK1/2 pathway was dispensable.We do not yet know tional activity. the mechanism through which Raf1 mediates the

Oncogene Ras pathways modulate SF protection SFanet al 4790 than V12Ras in enhancing the SF-stimulated IKK-b kinase activity, NF-kB activity and protection against ADR.This finding is consistent with our previous studies indicating a role for PI3K/c-Akt signaling in SF-mediated cell protection (Bowers et al., 2000; Fan et al., 2000, 2001, 2005), and it suggests that Ras acts, in part, upstream of PI3K in the cell protection pathway.The increased activity of V12C40Ras relative to V12Ras is consistent with an inhibitory component of Ras signaling owing to another pathway downstream of Ras (e.g., RalA) that is active in V12Ras but not in V12C40Ras.Alternatively, the V12C40Ras pathway may sequester signaling proteins from another SF signaling pathway that negatively regulates cell protection (Fan et al., 2001). Two Ras effector mutants that selectively bind Raf1 (V12S35 Ras and V12E38Ras) enhanced SF-stimulated IKK-b activity, NF-kB activity and cell survival.These findings, coupled to the finding that wtRaf1 enhanced the kinase, NF-kB and survival-stimulating activity of SF, suggest that Ras/Raf1 signaling contributes to these activities.The ability of DN-Raf1 or Raf1-siRNA to these activities suggests that Raf1 is required for the SF survival pathway in the cell types studied.And the ability of DN-Raf1 to block (SF þ V12Ras)-stimulated NF-kB activity is consistent with the idea that Raf1 acts downstream of Ras in this pathway.Although Raf1 contributes to SF-stimulated NF-kB activity, neither wtMEK1/2 nor DN-MEK1/2 mutants altered the SF- stimulated NF-kB activity, suggesting that the ability of Raf1 to enhance the SF-stimulated NF-kB activity is not Figure 9 Effects of SF and V12Ras on Raf1 phosphorylation. owing to the Raf1/MEK/ERK pathway.In contrast, the (a and b) Raf1 and c-Met IPs.DU-145 cells were transfected with ability of DN-Raf1 to block and wt-Raf1 to enhance V12Ras or empty pcDNA3 vector (15 mg of plasmid DNA per SF-stimulated IKK-b kinase activity suggest that Raf1 100 mm dish), postincubated for 24 h to allow gene expression and treated7SF (100 ng/ml) for 20 min.The cells were harvested; and targets the IKK complex or a kinase upstream of IKK. equal aliquots of total cell protein (500 mg) were IP’d using Our previous studies indicate that c-Akt signaling is antibodies against total Raf1 (a) or total c-Met (b) The IPs were required for SF-mediated NF-kB activation and cell Western blotted for total c-Met, phospho-c-Met (tyrosine-1349), protection (Bowers et al., 2000; Fan et al., 2000, 2001, total Raf1, phospho-Raf1 (serine-43), RalA, phospho-IKK-a/b 2005).The present study revealed that DN-Akt also and total IKK-a/b.(c and d) Effects of Raf1 (c) and RalA (d) siRNAs on IKK phosphorylation.DU-145 cells were transfected blocks the ability of RasV12 to enhance the SF- with the indicated siRNA (50 nM  48 h), treated7SF (100 ng/ stimulated NF-kB activity, suggesting that both Raf1 ml  48 h) and then Western blotted to detect Raf1 or RalA, and c-Akt act downstream of Ras in the SF survival phospho-IKK-a/b, total IKK-a/b and actin. pathway.The findings are consistent with a model in which Raf1 and c-Akt act in a direct pathway leading to NF-kB activation, although these proteins are not activation of IKK-b and NF-kB.However, in a different known to participate in the same phosphorylation context, it was observed that Raf1 can stimulate IKK-b cascade.Alternatively, they also support a model in activity by a mechanism that involves the kinase which Raf1 and c-Akt signaling impinge on the NF-kB MEKK1 (MAP3K1), but is independent of the ERK1/ pathway (e.g., at the level of IKK, a kinase upstream of 2 pathway (Baumann et al., 2000). IKK, and/or NF-kB itself) to synergistically activate this Consistent with the ability of Ras to enhance SF- pathway.Here, inactivation of either Raf1 or c-Akt is stimulated IKK-b kinase and NF-kB activity, the sufficient to block SF-induced NF-kB activation and cell activated mutant V12Ras also enhanced the nuclear survival.The finding that wtRaf1 enhances SF-stimulated translocation of the p65RelA subunit of NF-kB caused IKK-b activity and NF-kB activity by three- to fivefold is by SF as well as the SF-induced increases in the consistent with this model, because it shows that the expression of two known NF-kB target genes, cIAP2 pathway through which Raf1 stimulates SF-inducible and TRAF2.Previously, we showed that SF induces NF-kB activity can be further activated.Similarly, expression of these genes in an NF-kB-dependent we previously reported that exogenous c-Akt enhances manner and that they contribute to the SF-mediated SF-stimulated NF-kB activity by 2.5- to 3-fold, suggesting protection against ADR (Fan et al., 2005). that the pathway through which endogenous c-Akt The V12C40Ras mutant, which selectively binds p110a enhances SF-stimulated NF-kB activity is not maximally (Rodriguez-Viciana et al., 1997) was slightly more active activated and can be further stimulated.

Oncogene Ras pathways modulate SF protection SFanet al 4791

Figure 10 Effect of V12Ras on SF-induced p65 translocation and NF-kB target gene expression.( a) Effect of V12Ras on SF-induced nuclear translocation of p65.DU-145 cells cultured on glass coverslips in six-well dishes were transfected with V12Ras or empty pcDNA3 vector, washed and allowed to recover from transfection.The cells were then treated 7SF (100 ng/ml) for 20 min and processed for immunofluorescence microscopy to detect the p65 subunit of NF-kB (see Materials and methods).( b) Effect of V12Ras on SF-induced NF-kB target gene expression.DU-145 cells were transfected with V12Ras or empty pcDNA3 vector, washed, allowed to recover from transfection and treated7SF (100 ng/ml) for 24 h.The cells were then harvested; and whole-vcell lysates were Western blotted to detect cIAP2, TRAF2 and actin.

V12G37Ras, a mutant that selectively activates the phosphorylation in the absence of SF.The finding that a guanine nucleotide dissociation stimulator RalGDS farnesylation-defective mutant (V12S186Ras) did not (Albright et al., 1993; Hofer et al., 1994), enhanced stimulate IKK-b activity, NF-kB activity or cell protec- SF-stimulated IKK-b activity and NF-kB activity, but tion suggests that this alternative NF-kB stimulatory to a lesser extent than V12Ras.Studies utilizing DN- pathway requires farnesylation of the Ras protein. RalA and RalA-siRNA suggest that RalA negatively Ras can stimulate other effectors than p110a, Raf1 regulates SF stimulation of IKK-b and NF-kB activity and RalGDS, including RIN1, AF6, NORE1, PLC-e and SF protection against ADR and inhibits V12Ras- and other RalGDS family members (RGL, RLF mediated enhancement of the SF stimulation of IKK-b and RGL2) (Shao et al., 1999; Herrmann, 2003; and NF-kB activity.Our findings suggest that the Wohlgemuth et al., 2005). Whether these or other endogenous protein RalA inhibits the SF survival potential Ras effectors contribute to regulation of pathway at the level of the IKK complex and NF-kB NF-kB signaling downstream of Ras remains to be transcriptional activity.In addition, RalB, another discovered. target of RalGDS, inhibited the SF-stimulated NF-kB A role for RalA in an NF-kB-survival pathway is activity.We verified that all Ras effector loop mutant novel, as the role of RalA in cell survival or apoptosis is proteins tested were expressed and behaved as expected not well characterized.The pathways downstream of (e.g., V12G37Ras interacted with RalGDS but not RalA that mediate inhibition of NF-kB signaling and p110a or Raf1).The finding that V12G37Ras enhances cell protection by SF remain to be identified.RalA SF-mediated NF-kB survival signaling deserves com- signaling is not understood in great depth, but one ment, as two major targets of RalGDS, the RalA and target of RalA is phospholipase D1 (PLD1), a protein RalB GTPases, blocked SF-stimulated NF-kB signaling. implicated in vesicle trafficking (Luo et al., 1998). Two possible explanations are that: (1) the Emerging evidence suggests that the exocyst, a protein V12G37Ras/RalGDS pathway activates other downstream complex involved in the interaction of secretory vesicles components besides RalA and RalB; or (2) V12G37Ras with the plasma membrane, is a target for activated (and other Ras effector mutants) activate an NF-kB RalA and that RalA stimulates exocytosis (Camonis stimulatory pathway independent of RalGDS, Raf1 and and White, 2005).PLD1 may contribute to RalA- p110a/PI3K.Possibility (2) is consistent with the finding mediated exocytosis (Vitale et al., 2005). Whether PLD1 that each V12Ras effector mutant enhanced the SF- or exocyst proteins also participate in the regulation of stimulated c-Akt phosphorylation and caused c-Akt NF-kB signaling is conjectural at this time.RalBP1

Oncogene Ras pathways modulate SF protection SFanet al 4792 (RLIP76) is a RalA-interacting protein implicated in the mediate cell protection in carcinoma and epithelial cell transport of glutathione conjugates and in mediating lines, the stimulatory effects of Ras and Raf1 are drug resistance (Cantor et al., 1995; Stuckler et al., significantly greater in the presence of SF than in its 2005), but the role of RalA in these processes is unclear. absence.These findings suggest the possibility that the Although MAP kinase kinases 1–5 had little or no SF and Ras/Raf1 are acting through parallel pathways effect on NF-kB signaling, we found that DN-MKK3 that interact (cross-talk) with each other.SF may and DN-MKK6 inhibited SF stimulation of NF-kB activate more steps than Ras/Raf1, accounting for the activity and wild-type or active mutant MKK3 and greater effects of SF alone than of Ras/Raf1 alone. MKK6 enhanced SF-stimulated NF-kB activity.Knock down of MKKs 3 or 6 blocked SF-induced NF-kB activity, suggesting roles for endogenous MKK3/6 in Materials and methods SF-mediated NF-kB signaling, although we did not tested whether these kinases contribute to cell survival. Cell lines and culture The stress-activated kinase p38 is a substrate for MKK3 DU-145 human prostate cancer cells and MDCK epithelial cells were obtained from the American Type Culture Col- and MKK6; and several studies suggest that p38 can lection (Manassas, VA, USA).The cells were cultured in stimulate NF-kB signaling (Nick et al., 1999; Shuto Dulbecco’s modified Eagle’s medium (DMEM) supplemented et al., 2001; Costantini et al., 2005). Here, we found that with 5% fetal calf serum, non-essential amino acids (100 mM), a selective inhibitor of p38 (SB202190) inhibited SF-, L-glutamine (5 mM), streptomycin (100 mg/ml) and penicillin MKK3/6- and (SF þ V12Ras)-stimulated NF-kB acti- (100 U/ml) (all purchased from BioWhittaker, Walkersville, vity, suggesting that an MKK3/6-p38 pathway con- MD, USA).The cells were grown at 37 1C in a humidified tributes to SF-stimulated NF-kB activity.Interestingly, atmosphere of 95% air and 5% CO2 and were subcultured in addition to regulating IKK-b kinase activity (demon- weekly, using trypsin. strated herein), p38 has been reported to mediate the phosphorylation and activation the p65RelA subunit of Reagents NF-kB in another context, whether directly or indirectly Recombinant human two-chain SF was a gift from Dr Ralph (Rahman et al., 2004). Schwall, Department of Endocrine Research, Genentech Inc. Using IP-Western blotting, we showed that Ras, (South San Francisco, CA, USA).ADR (doxorubicin hydro- Raf1, RalA, IKK-a and IKK-b each associated with the chloride) and MTT were obtained from the Sigma Chemical liganded or unliganded c-Met receptor.We further Co.(St Louis, MO, USA).The p38 inhibitor SB202190 was demonstrated an association of c-Met, RalA, IKK-a purchased from CalBiochem (La Jolla, CA, USA).The and IKK-b with Raf1.Within these signaling com- antibodies used in this study are described below. plexes, SF induced and V12Ras further enhanced the phosphorylation (activation) of Raf1 and IKK-a/b.In Expression vectors studies using siRNAs to knock down endogenous The expression vectors utilized in this study are described in protein levels, we found that Raf1 is required for the Table 1. SF-induced phosphorylation of IKK-a/b and that endogenous RalA suppresses IKK-a/b phosphorylation. Taken together, these findings suggest a physiological Transient transfections Subconfluent proliferating cells were transfected overnight connection between Raf1, RalA and the IKK complex with the vector of interest or with the empty pcDNA3 vector that can modulate the activation of IKK-a/b and NF-kB (Invitrogen, Carlsbad, CA, USA) (15 mg of plasmid DNA per signaling.Although we did not test the activation state 100 mm dish or 5 mg per well in six-well dishes) using of the Ras and RalA GTPases, it was previously Lipofectamine (Life Technologies, Gaithersburg, MD, USA) demonstrated that ligation of c-Met is required for and washed to remove the excess vector and Lipofectamine. Ras activation (conversion to the GTP-bound state) To determine the transfection efficiency, cultures were co- (Graziani et al., 1993); and it is, thus, likely that c-Met transfected with plasmid pRSV-b-gal (Promega Corporation, ligation is also required for RalA activation within the Madison, WI, USA) to allow staining with 5-bromo- signaling complex.Whether c-Met can transmit a 4-choloro-3indolyl-b-D-galactoside reagent and visualization survival signal in the absence of SF is unclear; but it is of transfected (blue-staining) cells. interesting that the multisubstrate docking site of the c-Met receptor (1349YVHVxxxYVNV) – which mediates siRNA treatments the association of c-Met with multiple signaling proteins Validated siRNAs for human Raf1 (ID no.42858), RalA (ID (e.g., Grb2, Shc, Gab1 and p110) and is required for no.120429) and a negative control siRNA were purchased much of its signaling activity – is not required for from Ambion (Austin, TX, USA) (see Ambion siRNA catalog transduction of a Ras signal or cell scattering (Tulasne for details).MKK3-siRNA (catalog no.PA-11386) and et al., 1999). It was suggested that GDP-bound Ras may MKK6-siRNA (catalog no.PA-11388) were purchased from have a role in signal transduction (Stewart and Guan, Dharmacon Inc.(Lafayette, CO, USA).For siRNA transfec- tions, proliferating cells at about 30–50% of confluence were 2000).These unconventional receptor-dependent Ras treated with the indicated siRNA using OligofectAMINE signaling mechanisms are poorly understood. (Invitrogen, Carlsbad, CA, USA) according to manufacturer’s Interestingly, although the Ras/Raf1 signaling path- instructions.Cells were treated with 50 n M of siRNA for 48 h; ways stimulate IKK-b activity, NF-kB activity and cell and the efficacy of the siRNA treatment was verified by survival both in the presence and absence of SF, and to Western blotting.

Oncogene Ras pathways modulate SF protection SFanet al 4793 Measurement of NF-kB transcriptional activity cDNAs were cloned into the pCMV-Tag2B mammalian NF-kB activity was measured as described earlier (Fan et al., expression vector (Invitrogen) (to allow expression of the 2005).The NF- kB-Luc reporter plasmid (NF-kB-Luc) protein with an N-terminal FLAG epitope tag) using the (Stratagene, La Jolla, CA, USA) is composed of five copies BamH1 and XhoI cloning sites and sequenced.DU-145 cells of the NF-kB enhancer element upstream of a TATA box and were transfected overnight with the different FLAG-Ras the luciferase gene.Briefly, subconfluent proliferating cells in vectors (10 mg plasmid DNA per 100 mm dish), washed and 2cm2 wells were transfected overnight with 0.25 mg of NF-kB- postincubated for 24 h to allow gene expression.The cells were Luc reporter and 0.25 mg of each indicated vector, using then harvested by scraping; and total cell lysates were prepared Lipofectamine.The cells were then washed to remove in Tris-buffered saline (TBS) buffer (20 mM Tris–HCl pH 7.5, Lipofectamine and excess vectors, allowed to recover for 137 mM NaCl and 0.5% NP-40) containing Complete Protease several hours, treated7SF (100 ng/ml) for 24 h and harvested Inhibitor Cocktail (Roche, Indianapolis, IN, USA).The for measurement of luciferase activity, using the Dual lysates were passed through a 22 G needle to shear the DNA Luciferase Reporter Assay System (Promega, Piscataway, and centrifuged at 100 000 g at 41C for 30 min.The super- NJ, USA).Transfection efficiency was determined using the natant protein was quantified; and aliquots of protein (700 mg) Galacto-Star Mammalian Reporter Gene Assay System were incubated with anti-FLAG monoclonal M2 (2.5 mg) (Applied Biosystems, Foster City, CA, USA).Luciferase agarose beads (Sigma, St Louis, MO, USA) at 41C for 4 h.A values were means7s.e.m.s of quadruplicate wells. Each control IP was performed using the same quantity of non- experiment was performed at least twice to assure that the immune mouse IgG (catalog no.7056, Cell Signaling findings were reproducible. Technology).The beads were then washed three times with TBS buffer.The precipitated proteins were solubilized in 2% ADR treatment sodium dodecyl sulfate (SDS) buffer at 451C for 1 h and Subconfluent proliferating cells in six-well dishes were analysed by Western blotting. transfected overnight with the indicated vector(s) (5 mgof plasmid DNA per well), washed and allowed to recover for several hours in fresh culture medium.The cells were then Raf1 and c-Met Ips Following transfection with V12Ras or harvested, inoculated into 96-well dishes in complete culture empty pcDNA3 vector and treatment with SF (as described in medium, allowed to attach overnight, washed, incubated7SF the figure legends), DU-145 cells were scraped into 1 ml of lysis (100 ng/ml  48 h) in serum-free medium, exposed to different buffer (20 mM Tris–HCl (pH 7.5), 150 mM NaCl, 10% glycerol, doses of ADR (2 h at 371C) and postincubated for 48 h in fresh 1% NP-40 and protease inhibitor cocktail set I (Calbiochem, drug-free complete culture medium (7SF).The cells were then San Diego, CA, USA)) per 100 mm dish.The lysed cells were analysed for cell viability using MTT assays.For each assay incubated for 30 min at 01C, centrifuged at 17 000g for 15 min condition, cell viability was determined in 10 replicate wells.At at 41C, and the supernatants were collected and protein least two independent experiments were performed for each concentrations determined with the Bio-Rad protein assay, vector tested. using bovine serum albumin (BSA) as a standard.The IP antibodies were as follows: anti-Raf1 rabbit polyclonal anti- body (Ab32025) or anti-c-Met rabbit polyclonal antibody MTT cell viability assay (Ab14570) from Abcam (Cambridge, MA, USA).For IPs, This assay is based on the ability of viable mitochondria to 500 mg of protein was pre-cleared for 1 h by addition of Bio- convert MTT, a soluble tetrazolium salt, into an insoluble Mag beads (Qiagen, Valencia, CA, USA).The pre-cleared formazan precipitate, which is dissolved in dimethylsulfoxide lysates were incubated with IP antibody (5 mg) plus Bio-Mag et al and quantitated by spectrophotometry (Alley ., 1988). beads at 41C overnight.The beads were washed four times After ADR treatment, cells in 96-well dishes were tested for with 0.5 ml of lysis buffer; and 1  loading buffer (25 mM Tris– MTT dye conversion.The cell viability was calculated as the HCl (pH 6.5), 5% glycerol, 1% SDS, 1% 2-mercaptoethanol amount of MTT dye conversion relative to sham-treated and 0.05% bromphenol blue) was added. The samples were control cells. boiled for 3 min, analysed by sodium dodecyl sulfate– polyacrylamide gel electrophoresis (SDS–PAGE)); and Measurement of IKK-b kinase activity Western blotted as described below. After cell transfections and SF treatment of DU-145 cells as described in the figure legends, IKK-b kinase activity was measured using the HTScan IKK-b Kinase Assay Kit (Cell Western blotting Signaling Technology Inc., Danvers, MA, USA). This assay For straight Western blotting, cells were collected using measures the ability of the cell lysate being tested to trypsin, washed twice with phosphate-buffered saline (PBS) phosphorylate a biotinylated IkB-a (Ser32) substrate peptide. and pelleted by centrifugation.The pellet was resuspended in A phospho-IkB-a (Ser32/36)- specific mouse monoclonal radioimmunoprecipitation assay buffer (1  PBS, 1% NP-40, antibody is used to detect the phosphorylated form of the 0.5% sodium desoxycholate and 0.1% SDS), placed on ice for substrate peptide.As a positive control and for standardiza- 30 min and spun for 15 min at 14 000 r.p.m. at 41C to remove tion of the assay, the kit provides an active IKK-b kinase in the the insoluble material.Western blotting was performed as form of a glutathione S-transferase fusion protein.Each described earlier (Fan et al., 2005). Briefly, aliquots of whole experimental condition was tested in five replicate wells, and cell protein (50 mg per lane) or IP’d proteins (see above) were the values of IKK-b kinase activity were expressed relative to electrophoresed in 4–20% SDS–polyacrylamide gradient gels the untransfected, nonSF treated control.Each experiment and transferred to nitrocellulose membranes.The membranes was performed at least twice to assure reproducibility of the were blotted using the primary antibodies listed below and results. then blotted with the appropriate secondary antibodies (Santa Cruz Biotech, Santa Cruz, CA, USA, 1:1000 dilution).The Immunoprecipitation blotted proteins were visualized using the enhanced chemi- Ras IP To facilitate this experiment, the V12Ras, luminescence system (Amersham Pharmacia Biotech, Pis- V12C40Ras, V12S35Ras, V12G37Ras and V12E38 Ras cataway, NY, USA), with colored markers (Bio-Rad

Oncogene Ras pathways modulate SF protection SFanet al 4794 Laboratoriess, Hercules, CA, USA) as molecular size for 10 min at 251C.The cells were then permeabilized using standards. Tween-20 (in PBS) for 10 min at 251C; incubated in PBS The primary antibodies were: phospho-Akt (serine-473) (np. containing 1% BSA for 1 h at 251C to block nonspecific 9271S, Cell Signaling Technology Inc., Beverly, MA, USA; staining; and incubated overnight at 41C with rabbit anti- 1:500); phospho-Akt (threonine-309) (monoclonal 244F9, Cell human p65 polyclonal antibody (1:100 dilution) (Santa Cruz). Signaling Technology); total Akt (no.9271, Cell Signaling The cells were then washed and incubated with Cy5- Technology, 1:500); c-Met (H-190, Santa Cruz, 1:1000); conjugated goat anti-rabbit IgG antibody (Sigma) (1:100 in phospho-c-Met (tyrosine-1349) (Biosource International, Ca- blocking buffer) for 1 h at 371C and washed five times with marillo, CA, USA); FLAG (rabbit anti-FLAG, Sigma, PBS.The coverslips were mounted on glass slides with Mowiol 1:2000); IkB-a (C-15, Santa Cruz, 1:500); IKK-a (rabbit anti- 4-88 (Hoechst Celanese, Somerville, NJ, USA).A laser with a human IKK-a, Sigma, 1:1000); IKK-b (monoclonal 62AT216, wavelength of 638 nm was utilized for excitation of Cy5 Abgent Inc., San Diego, CA, USA; 1:500); phospho-IKK-a/b fluorescence.Fluorescence images were captured with a Nikon (sc-21661, Santa Cruz, 1:500); p110a (sc-8010, Santa Cruz, Eclipse E800 confocal laser fluorescence microscope with an 1:300); MKK3 (no.9232, rabbit polyclonal, Cell Signaling objective lens ( Â 40) (Diagnostic Instruments, Sterling Technology, 1:1000); MKK6 (no.9264, rabbit polyclonal, Cell Heights, MI, USA).The images were collected through the Signaling Technology, 1:1000); Raf1 (C-20, Santa Cruz, 1:1000 specimens every 3 s in the vertical plane and overlaid to or no. 1560–1, Epitomics Inc., Burlingame, CA, USA, 1:500); generate focus composite images.The images were exported phospho-Raf1 (pS259, Epitomics, 1:500); RalA (mouse and displayed images were exported to Adobe Photoshop monoclonal, BD Transduction Laboratories, San Diego, CA, (Adobe, San Jose, CA, USA). USA, 1:100 or R23520, rabbit polyclonal, 1:400, Signal Transduction Laboratories, Lexington, KY, USA); RalGDS Statistical analyses (C-19, Santa Cruz, 1:500); Ras (R4025, Sigma, 1:1000); and Where appropriate, statistical comparisons were made using actin (goat polyclonal, Santa Cruz, 1:1000). the two-tailed Student’s t-test. Immunofluorescence microscopy to detect p65 (RelA) subcellular translocation Acknowledgements After transfection and treatment with SF (see text and figure legends), DU-145 cells cultured on glass coverslips in six-well Sources of support: This work is supported, in part, by United plastic dishes were fixed with 4% paraformaldehyde (Electron States Public Health Service Grants RO1-ES09169 (EMR) and Microscopy Sciences, Fort Washington, PA, USA) in PBS RO1-NS43987 (JJL/EMR).

References

Albright CF, Giddings BW, Liu J, Vito M, Weinberg RA. Chung HH, Benson DR, Schultz PG.(1993).Probing the (1993).Characterization of a guanine nucleotide dissoci- structure and mechanism of Ras protein with an expanded ation stimulator for a ras-related GTPase. EMBO J 12: genetic code. Science 259: 806–809. 339–347. Cong LN, Chen H, Li Y, Zhou L, McGibbon MA, Taylor SI Alley MC, Scudiero DA, Monks A, Hursey ML, Czerwinski MJ, et al. (1997).Physiological role of Akt in insulin-stimulated Fine DL et al. (1988).Feasibility of drug screening with translocation of GLUT4 in transfected rat adipose cells. Mol panels of human tumor cell lines using a microculture Endocrinol 11: 1881–1890. tetrazolium assay. Cancer Res 48: 589–601. Costantini C, Rossi F, Formaggio E, Bernardoni R, Cecconi D, Bardelli A, Longatti P, Albero D, Goruppi S, Schneider C, Della-Bianca V.(2005).Characterization of the signaling Ponzetto C et al. (1996).HGF receptor associates with the pathway downstream p75 neurotrophin receptor involved in anti-apoptotic protein BAG-1 and prevents cell death. beta-amyloid peptide-dependent cell death. J Mol Neurosci EMBO J 15: 6205–6212. 25: 141–156. Baumann B, Weber CK, Troppmair J, Whiteside S, Israel A, de Vos AM, Tong L, Milburn MV, Matias PM, Jancarik J, Rapp UR et al. (2000).Raf induces NF-kappaB by Noguchi S et al. (1988).Three-dimensional structure of an membrane shuttle kinase MEKK1, a signaling pathway oncogene protein: catalytic domain of human c-H-ras p21. critical for transformation. Proc Natl Acad Sci USA 97: Science 239: 888–893. 4615–4620. Derijard B, Raingeaud J, Barrett T, Wu I-H, Han J, Ulevitch RJ Bottaro DP, Rubin JS, Faletto DL, Chan AM-L, Kmiecik TE, et al. (1995).Independent human MAP-kinase signal Vande Woude GF et al. (1991).Identification of the transduction pathways defined by MEK and MKK iso- hepatocyte growth factor receptor as the c-met proto- forms. Science 267: 682–685. oncogene product. Science 251: 802–804. Fan S, Gao M, Meng Q, Laterra JJ, Symons MH, Coniglio S Bowers DC, Fan S, Walter K, Abounader R, Williams JA, et al. (2005).Role of NF-kappaB signaling in hepatocyte Rosen EM et al. (2000).Scatter factor/hepatocyte growth growth factor/scatter factor-mediated cell protection. Onco- factor protects against cytotoxic death in human glioblasto- gene 24: 1749–1766. ma via phosphatidylinositol 3-kinase- and AKT-dependent Fan S, Ma YX, Gao M, Yuan RQ, Meng Q, Goldberg ID pathways. Cancer Res 60: 4277–4283. et al. (2001).The multisubstrate adapter Gab1 regulates Camonis JH, White MA.(2005).Ral GTPases: corrupting the hepatocyte growth factor (scatter factor)-c-Met signaling for exocyst in cancer cells. Trends Cell Biol 15: 327–332. cell survival and DNA repair. Mol Cell Biol 21: 4968–4984. (Review). Fan S, Ma YX, Wang J-A, Yuan R-Q, Meng Q, Laterra JJ Cantor SB, Urano T, Feig LA.(1995).Identification and et al. (2000).The cytokine hepatocyte growth factor/scatter characterization of Ral-binding protein 1, a potential factor inhibits apoptosis and enhances DNA repair by a downstream target of Ral GTPases. Mol Cell Biol 15: common mechanism involving signaling through phospha- 4578–4584. tidyl inositol 30 kinase. Oncogene 19: 2212–2223.

Oncogene Ras pathways modulate SF protection SFanet al 4795 Fan S, Wang J-A, Yuan R-Q, Rockwell S, Andres J, Murai H, Ikeda M, Kishida S, Ishida O, Okazaki-Kishida M, Zlatapolskiy A et al. (1998).Scatter factor protects epithelial Matsuura Y et al. (1997).Characterization of Ral GDP and carcinoma cells against apoptosis induced by DNA- dissociation stimulator-like (RGL) activities to regulate c-fos damaging agents. Oncogene 17: 131–141. promoter and the GDP/GTP exchange of Ral. J Biol Chem Fitzgerald EM.(2000).Regulation of voltage-dependent 272: 10483–10490. calcium channels in rat sensory neurones involves a Ras- Nick JA, Avdi NJ, Young SK, Lehman LA, McDonald PP, mitogen-activated protein kinase pathway. J Physiol 527: Frasch SC et al. (1999).Selective activation and functional 433–444. significance of p38alpha mitogen-activated protein kinase in Frisch SM, Francis H.(1994).Disruption of epithelial lipopolysaccharide-stimulated neutrophils. J Clin Invest 103: cell–matrix interactions induces apoptosis. J Cell Biol 124: 851–858. 619–626. Oxford G, Owens CR, Titus BJ, Foreman TL, Herlevsen MC, Graziani A, Gramaglia D, dalla Zonca P, Comoglio PM. Smith SC et al. (2005).RalA and RalB: antagonistic relatives (1993).Hepatocyte growth factor/scatter factor stimulates in cancer cell migration. Cancer Res 65: 7111–7120. the Ras-guanine nucleotide exchanger. J Biol Chem 268: Peyssonnaux C, Provot S, Felder-Schmittbuhl MP, Calothy G, 9165–9168. Eychene A.(2000).Induction of postmitotic neuroretina cell Herrmann C.(2003).Ras-effector interactions: after one proliferation by distinct Ras downstream signaling path- decade. Curr Opin Struct Biol 13: 122–129.(Review). ways. Mol Cell Biol 20: 7068–7079. Hofer F, Fields S, Schneider C, Martin GS.(1994).Activated Rahman A, Anwar KN, Minhajuddin M, Bijli KM, Javaid K, Ras interacts with the Ral guanine nucleotide dissociation True AL et al. (2004).cAMP targeting of p38 MAP kinase stimulator. Proc Natl Acad Sci USA 91: 11089–11093. inhibits thrombin-induced NF-kappaB activation and Hu MC, Wang Y.(1998).IkappaB kinase-alpha and -beta ICAM-1 expression in endothelial cells. Am J Physiol Lung genes are coexpressed in adult and embryonic tissues Cell Mol Physiol 287: L1017–L1024. but localized to different human chromosomes. Gene 222: Raingeaud J, Whitmarsh AJ, Barrett T, Derijard B, Davis RJ. 31–40. (1996).MKK3- and MKK6-regulated gene expression is Jin L, Fuchs A, Schnitt SJ, Yao Y, Joseph A, Lamszus K et al. mediated by the p38 mitogen-activated protein kinase signal (1997).Expression of scatter factor and c-met receptor in transduction pathway. Mol Cell Biol 16: 1247–1255. benign and malignant breast tissue. Cancer 79: 749–760. Rodriguez-Viciana P, Warne PH, Khwaja A, Marte BM, Kraus MH, Yuasa Y, Aaronson SA.(1984).A position 12- Pappin D, Das P et al. (1997).Role of phosphoinositide activated H-ras oncogene in all HS578T mammary carcino- 3-OH kinase in cell transformation and control of the actin sarcoma cells but not normal mammary cells of the same cytoskeleton by Ras. Cell 89: 457–467. patient. Proc Natl Acad Sci USA 81: 5384–5388. Rosen EM, Lamszus K, Laterra J, Polverini PJ, Rubin JS, Lin A, Minden A, Martinetto H, Claret F-X, Lange-Carter C, Goldberg ID.(1997).HGF/SF in angiogenesis. Ciba Found Mercurio F et al. (1995).Identification of a dual specificity Symp 212: 215–226. kinase that activates the Jun kinases and p38-Mpk2. Science Shao H, Kadono-Okuda K, Finlin BS, Andres DA.(1999). 268: 286–290. Biochemical characterization of the Ras-related GTPases Luo JQ, Liu X, Frankel P, Rotunda T, Ramos M, Flom J et al. Rit and Rin. Arch Biochem Biophys 371: 207–219. (1998).Functional association between Arf and RalA in Shuto T, Xu H, Wang B, Han J, Kai H, Gu XX et al. (2001). active phospholipase D complex. Proc Natl Acad Sci USA Activation of NF-kappa B by nontypeable Hemophilus 95: 3632–3637. influenzae is mediated by toll-like receptor 2-TAK1- Matteucci E, Modora S, Simone M, Desiderio MA.(2003). dependent NIK-IKK alpha/beta-I kappa B alpha and Hepatocyte growth factor induces apoptosis through MKK3/6-p38 MAP kinase signaling pathways in epithelial the extrinsic pathway in hepatoma cells: favouring role cells. Proc Natl Acad Sci USA 98: 8774–8779. of hypoxia-inducible factor-1 deficiency. Oncogene 22: Song KS, Seong JK, Chung KC, Lee WJ, Kim CH, Cho KN 4062–4073. et al. (2003).Induction of MUC8 gene expression by May MJ, D’Acquisto F, Madge LA, Glockner J, Pober JS, interleukin-1 beta is mediated by a sequential ERK Ghosh S.(2000).Selective inhibition of NF-kappaB activa- MAPK/RSK1/CREB cascade pathway in human airway tion by a peptide that blocks the interaction of NEMO with epithelial cells. J Biol Chem 278: 34890–34896. the IkappaB kinase complex. Science 289: 1550–1554. Stewart S, Guan KL.(2000).The dominant-negative Ras McFarlin DR, Lindstrom MJ, Gould MN.(2003).Affinity mutant, N17Ras, can inhibit signaling independently of with Raf is sufficient for Ras to efficiently induce rat blocking Ras activation. J Biol Chem 275: 8854–8862. mammary carcinomas. Carcinogenesis 24: 99–105. Stuckler D, Singhal J, Singhal SS, Yadav S, Awasthi YC, Milburn MV, Tong L, deVos AM, Brunger A, Yamaizumi Z, Awasthi S.(2005).RLIP76 transports vinorelbine and Nishimura S et al. (1990).Molecular switch for signal mediates drug resistance in non-small cell lung cancer. transduction: structural differences between active and Cancer Res 65: 991–998. inactive forms of protooncogenic ras proteins. Science 247: Tulasne D, Paumelle R, Weidner KM, Vandenbunder B, 939–945. Fafeur V.(1999).The multisubstrate docking site of Mochizuki N, Ohba Y, Kobayashi S, Otsuka N, Graybiel AM, the MET receptor is dispensable for MET-mediated Tanaka S et al. (2000).Crk activation of JNK via C3G and RAS signaling and cell scattering. Mol Biol Cell 10: R-Ras. J Biol Chem 275: 12667–12671. 551–565. Mody N, Campbell DG, Morrice N, Peggie M, Cohen P. Vitale N, Mawet J, Camonis J, Regazzi R, Bader MF, (2003).An analysis of the phosphorylation and activation of Chasserot-Golaz S.(2005).The small GTPase RalA controls extracellular-signal-regulated protein kinase 5 (ERK5) by exocytosis of large dense core secretory granules by mitogen-activated protein kinase kinase 5 (MKK5) in vitro. interacting with ARF6-dependent phospholipase D1. J Biol Biochem J 372: 567–575. Chem 280: 29921–29928. Moodie SA, Willumsen BM, Weber MJ, Wolfman A.(1993). Vojtek AB, Hollenberg SM, Cooper JA.(1993).Mammalian Complexes of Ras.GTP with Raf-1 and mitogen-activated Ras interacts directly with the serine/threonine kinase Raf. protein kinase kinase. Science 260: 1658–1661. Cell 74: 205–214.

Oncogene Ras pathways modulate SF protection SFanet al 4796 Voss M, Weernink PA, Haupenthal S, Moller U, Cool RH, formation with IkappaB kinase-alpha and NIK. Science 278: Bauer B et al. (1999).Phospholipase D stimulation by 866–869. receptor tyrosine kinases mediated by protein kinase C and a Yan M, Templeton DJ.(1994).Identification of 2 serine Ras/Ral signaling cascade. J Biol Chem 274: 34691–34698. residues of MEK-1 that are differentially phosphorylated Wang XS, Diener K, Manthey CL, Wang S, Rosenzweig B, during activation by raf and MEK kinase. J Biol Chem 269: Bray J et al. (1997).Molecular cloning and characterization 19067–19073. of a novel p38 mitogen-activated protein kinase. J Biol Chem Yang JJ, Kang JS, Krauss RS.(1998).Ras signals to 272: 23668–23674. the cell cycle machinery via multiple pathways to induce Webb CP, Van Aelst L, Wigler MH, Woude GF.(1998). anchorage-independent growth. Mol Cell Biol 18: Signaling pathways in Ras-mediated tumorigenicity and 2586–2595. metastasis. Proc Natl Acad Sci USA 95: 8773–8778. Zheng CF, Guan KL.(1993).Properties of MEKs, the kinases White MA, Nicolette C, Minden A, Polverino A, Van Aelst L, that phosphorylate and activate the extracellular signal- Karin M et al. (1995).Multiple Ras functions can contribute regulated kinases. J Biol Chem 268: 23933–23939. to mammalian cell transformation. Cell 80: 533–541. Zhong H, Chiles K, Feldser D, Laughner E, Hanrahan C, Wohlgemuth S, Kiel C, Kramer A, Serrano L, Wittinghofer F, Georgescu MM et al. (2000).Modulation of hypoxia- Herrmann C.(2005).Recognizing and defining true Ras inducible factor 1alpha expression by the epidermal binding domains I: biochemical analysis. J Mol Biol 348: growth factor/phosphatidylinositol 3-kinase/PTEN/AKT/ 741–758. FRAP pathway in human prostate cancer cells: implications Woronicz JD, Gao X, Cao Z, Rothe M, Goeddel DV.(1997). for tumor angiogenesis and therapeutics. Cancer Res 60: IkappaB kinase-beta: NF-kappaB activation and complex 1541–1545.

Oncogene