Oncogene (2001) 20, 1816 ± 1824 ã 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00 www.nature.com/onc ORIGINAL PAPERS The small-GTPase RalA activates transcription of the urokinase plasminogen activator receptor (uPAR) gene via an AP1-dependent mechanism
Emel Okan1, Victoria Drewett1, Peter E Shaw1 and Peter Jones*,1
1School of Biomedical Sciences, Queen's Medical Centre, University of Nottingham, Nottingham, NG7 2UH, UK
The urokinase plasminogen activator receptor (uPAR) and beta-family integrins (Xue et al., 1997; Kindzelskii focuses extracellular protease activity to the cell surface, et al., 1996; Wei et al., 1996). It also has a role in signal modulates cell adhesion and activates intracellular signal transduction, possibly through interactions with trans- transduction pathways. In a range of cancers uPAR membrane molecules such as gp130 or integrins that expression often has a negative correlation with function as conduits of uPAR-mediated signal trans- prognosis. Here we show that uPAR transcription is duction (Koshelnick et al., 1997, 1999). That uPAR stimulated by V12 H-Ras, the eector loop mutant V12 plays an important role in determining malignancy of H-Ras G37 and constitutively-active RalA 72L. RalA- most human tumours is based on a large number of dependent transcription required the presence of the experimental studies of both human cancers and model ATF2-like AP1-site at 770 bp and the c-Jun binding systems and appears to be a negative prognostic motif at 7184 bp in the uPAR promoter. Consistent marker in a number of cancers (Andreasen et al., with this, both Gal4-c-Jun- and Gal4-ATF2-fusion 1997, 2000). proteins were activated by RalA signalling through As the level of uPAR expression critically aects the phosphorylation of their activation domains at Ser63 invasive capacity of a cell, its expression is subject to and Ser73 of c-Jun or Thr69 and Thr71 of ATF2. A complex regulation at the levels of translation and transdominant inhibitory mutant of c-Jun N-terminal transcription (Shetty et al., 1997; Aguirre-Ghiso et al., kinase (JNK) failed to inhibit uPAR transcription 1999a). Many of the growth factor- and cytokine- demonstrating that JNK activation is not a prerequisite activated signalling pathways that control uPAR for RalA-dependent uPAR transcription. A dominant transcription converge on the small GTP-binding negative inhibitor of c-Src eectively inhibited RalA- protein Ras (Aguirre-Ghiso et al., 1999a). The ras dependent uPAR transcription identifying it as a down- genes (H-ras,N-ras and K-ras) frequently acquire stream eector in the RalA signalling pathway. These activating mutations in human cancers (Bos, 1998; data provide evidence for the existence of a novel Hunter, 1997; Yamamoto et al., 1999) and at least H- signalling pathway that links RalA to the activation of Ras and K-Ras have been demonstrated to activate uPAR transcription via a c-Src intermediate and uPAR gene transcription (Aguirre-Ghiso et al., 1999a; activation of AP1. Oncogene (2001) 20, 1816 ± 1824. Allgayer et al., 1999; Muller et al., 2000). This suggests a mechanism whereby the activity of oncogenic Ras Keywords: RalA; c-Src; c-Jun; ATF2; uPAR; Ras can be linked to the increased invasiveness of malignant cells through the uPA/uPAR proteolytic system. Introduction Recent studies have demonstrated that Ras stimu- lates a number of dierent downstream eectors whose The urokinase plasminogen activator receptor (uPAR) activities are essential for its transforming properties is a glycosylphosphatidylinositol (GPI) membrane- (Shields et al., 2000; Yamamoto et al., 1999). The use anchored receptor that binds the serine protease of multiple eector pathways by Ras has been urokinase plasminogen activator (uPA). The receptor demonstrated through the use of constitutively-acti- anchors uPA to the leading edge of migrating cells vated mutants of Ras (V12 H-Ras) that also bear where the complex stimulates migration by concentrat- mutations in the eector loop, amino acids 32 ± 40 ing the proteolytic activity of uPA to the surface of the (White et al., 1995). These mutants interact with at invading cell. Cell adhesion is also aected by uPAR least three dierent eector molecules; c-Raf, Phos- through its interactions with the extracellular matrix phoinositide 3-kinase (PI3K) and RalGDS. The serine/ protein vitronectin (Wei et al., 1994; Waltz et al., 1997) threonine kinases c-Raf, activates the ERK (Extra- cellular Signal-Regulated Kinase) mitogen-activated protein kinase (MAPK) activity that regulates a diverse range of biological events especially gene expression. *Correspondence: P Jones Received 21 November 2000; revised 22 December 2000; accepted Phosphoinositide 3-kinase (PI3K) has an emerging role 15 January 2001 in promoting cell survival (Rodriguez-Viciana et al., RalA activation of uPAR transcription E Okan et al 1817 1997; Kaumann-Zeh et al., 1997). The third and most mutants containing 5'-deletions of increasing size. We recently established eector, is the Ral guanine found that proximal region of the uPAR promoter nucleotide exchange factor, RalGDS. RalGDS is a between 7139 bp and the major transcription start site guanine nucleotide exchange factor (GEF) for the (+1) was the minimum sequence required to retain the small GTPase Ral. Currently, ®ve Ral GEFs have been stimulation of transcription by the transforming hu- identi®ed; RalGDS, Rgl, Rlf, Rgr and RalGEF2 that man H-Ras (V12 H-Ras) (data not shown). Potential are direct targets for Ras (Bos, 1998; de Bruyn et al., binding sites for transcription factors in this region are 2000). They provide a mechanism for Ral activation by an AP1/GRE site, a binding site for NF-kB and a extracellular signals via a variety of receptors, includ- number of binding sites for Sp1 (Figure 1a). To ing G-protein-coupled receptors and receptor tyrosine determine if Ras utilized a preferred eector to activate kinases (Hofer et al., 1998; Wolthuis et al., 1998a,b). uPAR transcription from this region of the promoter, RalA is one of two highly similar (85% identity) we used luciferase reporter assays to compare the GTPases that constitute a distinct family of Ras-related eects of constitutively activated V12 H-Ras and the proteins (Bos, 1997). The demonstration that domi- eector-loop mutants RasV12/C40, RasV12/G37 and nant-negative mutants of Ras inhibit insulin- and RasV12/S35 on transcription in the cell line NIH3T3 epidermal growth factor-induced activation of Ral (Figure 1b). Transcription from the minimal uPAR has established that Ral de®nes a downstream promoter of 139 bp was stimulated approximately signalling pathway from Ras (Wolthuis et al., 1998a). ®vefold by V12 H-Ras and over twofold by Ras 35S However, Ral can also be activated by Ras-indepen- and Ras 40C, that utilize Raf-1 and PI3K as their dent mechanisms (Hofer et al., 1998; Wolthuis et al., downstream eectors respectively (White et al., 1995). 1998b). The most eective stimulation of transcription by a The role played by Ral-family proteins in the cell is unclear but they appear to be linked predominantly with cell proliferation and dierentiation (Urano et al., 1996; Wolthuis et al., 1997). These properties are proposed to result from of the ability of Ral to couple signals from Ras to the induction of gene transcription or to the inhibition of transcription factors like AFX (Murai et al., 1997; Okazaki et al., 1997; Kops et al., 1999; Medema et al., 2000). The signalling pathway that links Ral activity to transcription is currently unknown. Our previous work to identify the signalling pathway utilized by oncogenic Ras in activating uPAR gene transcription suggested that RalGDS could have some role in this process (Muller et al., 2000). In this study, we demonstrate that a constitutively activated RalA mutant, RalA 72L, stimulates transcription from the uPAR promoter and that this requires the presence of at least one of the AP1 sites located 7184 or 770 bp upstream of the major start point of transcription. Activation of transcription is inhibited by a dominant- negative mutant of c-Src, indicating that c-Src is a downstream eector of RalA. We also show that RalA activates the Jun-family members c-Jun and ATF2 in a manner dependent on the phosphorylation sites Ser63 and Ser73 of c-Jun and Thr69 and Thr71 of ATF2. The uPAR is the ®rst example of a gene that can be up-regulated by RalA through an AP1 dependent mechanism. Figure 1 (a) Diagram showing the location of potential transcription factor binding sites in the 1.55 kb upstream sequence of the uPAR promoter. The region between the SmaI site at 7139 bp and the major start point of transcription is Results expanded. (b) Eects of ras eector-loop mutants on the activation of transcription from the proximal region of the uPAR Differential stimulation of uPAR transcription by Ras promoter. NIH3T3 cells were co-transfected with the luciferase effector loop mutants through DNA response elements reporter construct and either V12 H-Ras or V12 H-Ras contain- ing a mutation in the eector loop. The downstream eectors of located within 139 bp of the transcription start site Ras 37G, Ras 35S and Ras 40C are RalGDS, c-Raf and PI3K To locate the Ras-response elements within the respectively. The RLA represents the relative luciferase activity after normalization for dierences in transfection eciencies. The promoter, we used luciferase reporter assays to data points represent the mean of three independent experiments measure expression from a series of uPAR promoter with standard deviations indicated by error bars
Oncogene RalA activation of uPAR transcription E Okan et al 1818 Ras-eector loop mutant was obtained with Ras G37 potential binding sites for three classes of transcription that utilizes RalGDS as an eector. Stimulation of factor: NF-kB, SP1 and AP1 at an AP1-like site transcription by this mutant was similar to that designated AP1/GRE (Figure 1a). As Ras-mediated obtained using V12 H-Ras. The similar levels stimu- signal transduction has been demonstrated to stimulate lated by V12 H-Ras and Ras G37 indicate that Ras the phosphorylation and transcriptional activity of c- promotes expression from this region of the uPAR Jun (Smeal et al., 1991; Johnson et al., 1996), we promoter largely through RalGDS as its immediate reasoned that the AP1 binding site at 770 represented eector. the best candidate for a DNA response element that could be involved in H-Ras-dependent, and therefore possibly RalA-dependent, signal transduction. In the A constitutively-activated mutant of Ral A stimulates context of the uPAR promoter however, an AP1 site at transcription from the uPAR promoter 7184 bp could also contribute to stimulation of The Ras eector RalGDS, is a guanine nucleotide transcription by Ras or RalA (Figure 3a). To exchange factor for the Ras-related GTPase RalA. To determine if either AP1 site is utilized in RalA- determine whether RalA is capable of activating uPAR mediated activation of transcription, DNA probes gene transcription, NIH3T3 cells were co-transfected containing normal or mutated AP1 sites at 770 and with a luciferase reporter construct containing 1.5 kb 7184 were incubated with nuclear extracts from of 5' uPAR promoter sequence and V12 H-Ras or HEK293 cells. The protein-DNA complexes were RalA 72L, a constitutively-activated mutant of RalA. analysed by EMSA (Figure 3b). Complexes were V12 H-Ras stimulated uPAR transcription ®vefold formed in extracts from both untransfected cells and while RalA 72L resulted in a twofold increase (Figure from cells transfected with RalA 72L. As before, the 2a). When the same experiment was performed using a mobility of the complex obtained with nuclear extracts luciferase reporter construct that contained only the from cells expressing RalA 72L was retarded. The 139 bp proximal promoter sequence, the RalA 72L intensity of this band was greatest for the probe that mutant was more ecient than V12 H-Ras in contained two functional AP1 sites and absent with the stimulating transcription (Figure 2b). Although the probe in which both AP1 sites had been mutated. In level of transcription from the 139 bp promoter comparison, mutation of just the AP1 site at 770 bp sequence was lower than that of the larger 1.55 kb resulted in a small loss of complex formation while uPAR promoter, transcription of the uPAR gene is mutation of the site at 7184 bp produced a much subject to regulation by a RalA-dependent signal greater reduction. These observations indicate that transduction pathway in each case. The lower levels RalA-mediated activation of uPAR gene transcription of transcription from the 139 bp promoter may result involves the binding of AP1-transcription factors to from the loss of sites further upstream in the uPAR both AP1 sites. The marked decrease in complex promoter necessary for the assembly or stabilization of formation as a result of mutation of the AP1 at basal transcription factors on this TATA-less promoter 7184 bp suggests that this site is likely to make the (Soravia et al., 1995). greatest contribution to the activation of uPAR The ability of H-Ras- or RalA-mediated signalling to transcription. in¯uence transcription factors recruited to the proximal To clarify whether RalA 72L-mediated signal region of the promoter was supported by electro- transduction can utilize either of the two AP1 binding phoretic mobility shift assay (EMSA) (Figure 2c). sites in the uPAR promoter to activate transcription, Nuclear extracts from HEK293 cells generated a single we used PCR-based mutagenesis to inactivate either or complex with a probe spanning +1 to 7210 bp of the both AP1 sites. Luciferase-reporter constructs contain- uPAR promoter, which is likely to result from the ing the uPAR promoter region between 750 and binding of components of the basal transcription 7190 bp with combinations of active and inactive AP1 machinery. When nuclear extracts from V12 H-Ras- sites were co-transfected with RalA 72L or V12 H-Ras or RalA 72L-transfected cells were used, the complex into HEK293 cells and the levels of luciferase activity was lost and a new complex formed. This suggests that determined (Figure 4). Promoters that contained both both H-Ras and RalA cause the additional binding, or AP1 sequences (770/7184) were inducible by Ras and exchange, of transcription factors within this region of RalA 72L. These results were consistent with those the uPAR promoter. The data are consistent with those determined earlier (Figure 2b). Mutation of the 7184 obtained from luciferase reporter assays and clearly AP1 site (7184 m) reduced the basal transcriptional indicate that both H-Ras and RalA activate signal activity of the promoter, however both V12 H-Ras and induced changes that stimulate transcription of the RalA 72L were still able to increase the levels of uPAR gene from response elements within 200 bp of transcription approximately sixfold, demonstrating that the transcription start site. the AP1 site at 770 can mediate Ras- and RalA- induced transcription. The low levels of basal tran- scription that result from loss of the 7184 AP1 site RalA activation of uPAR transcription is mediated via may be a consequence of its importance in the basal AP1 sites at 7184 and 770 transcriptional activity of the TATA-less uPAR The DNA sequence between the major start point of promoter. Mutation of the AP1/GRE site at 770 transcription of the uPAR gene and 7139 bp contains had little eect on basal transcription levels and was
Oncogene RalA activation of uPAR transcription E Okan et al 1819
Figure 2 (a) Eects of constitutively-activated mutants of Ras and RalA on the activation of transcription from the 1.55 kb or 139 bp regions of the uPAR promoter. HEK 293 cells were transfected with the luciferase reporter construct containing either the upstream 1.55 kb of promoter sequence (a) or 139 bp reporter fragment (b) and V12 H-ras or RalA 72L. Transfection with the plasmid pMT3 was a control to demonstrate that the background vector had no eect on transcription. The data points represent the mean of three independent experiments with standard deviation indicated by error bars. (c) Eect of V12 H-Ras or RalA 72L expression on the electrophorectic mobility shift pattern of a DNA probe derived from the sequence from bp +1 to 7210 of the uPAR promoter. Nuclear extracts (15 mg) prepared from untransfected HEK293 cells or cells transfected with RalA 72L or V12 H- Ras and incubated with 50 fmoles of DNA probe for 15 min at room temperature. Arrowheads indicate the positions of the transcription complexes formed stimulated by RalA 72L and V12 H-Ras to a similar sequences. Inactivation of both AP1 sites totally level as the promoter that contained both correct AP1 abolished any detectable response from RalA 72L or
Oncogene RalA activation of uPAR transcription E Okan et al 1820
Figure 4 Eects of mutations in the AP1 and AP1/GRE sites on the activation of transcription by RalA 72L. HEK293 cells were transfected with luciferase reporter constructs that contained un- mutated AP1 sites (770/7184); two mutated AP1 sites (770 m/ 7184 m), or alternately mutated AP1/GRE (770 m) and AP1 (7184 m) sites cloned upstream of the luciferase gene. The promoter region used corresponds to the shaded sequence in Figure 3a. The RLA represents the relative luciferase activity after normalization for dierences in transfection eciencies. The data points represent the mean of three independent experiments with standard deviation indicated by error bars
mediate activation of uPAR transcription by RalA, while the 7184 site also in¯uences basal levels of transcription.
c-Jun and ATF2 are targets of the RalA pathway AP1 is a collective term referring to dimeric transcrip- tion factors composed of Jun Fos or ATF proteins that bind to related DNA elements. The sequences of the AP1 binding sites at 7184 and 770 bp dier and they Figure 3 (a) The DNA sequence of the proximal region of the resemble those that would be bound preferentially by uPAR promoter showing the location of the AP1 sites at c-Jun-containing and ATF-containing AP1 dimers 7184 bp and the AP1/GRE site at 770 bp (underlined). The respectively (Hai and Curran, 1991). As RalA sites were changed using PCR-based mutagenesis to the sequences stimulates transcription through both AP1 sites in the shown in bold (lower case). The sequence shaded shows the region of the promoter used as a probe in electrophoretic mobility uPAR promoter, we tested the ability of RalA to shift assays. (b) Eect of mutating the AP1 sites of the uPAR activate c-Jun- and ATF2-dependent transcription. promoter on the electrophoretic mobility shift pattern. Nuclear HEK293 cells were co-transfected with plasmid vectors extracts (15 mg) were incubated with probes generated from the expressing RalA 72L, hybrid proteins consisting of the sequence shown in A that contained; un-mutated AP1 sites (770/ transactivation domains of either c-Jun or ATF2 fused 7184); two mutated AP1 sites (770 m/7184 m), or alternately mutated AP1/GRE (770 m) and AP1 (7184 m) sites. Arrow- to the DNA-binding domain of the yeast transcription heads indicate the position of the complexes formed factor GAL4 and a GAL4-luciferase reporter. RalA 72L activated both GAL-ATF2 and GAL-c-Jun dependent transcription but showed a greater propen- V12 H-Ras. The loss of the AP1/GRE site (770 m) sity for the activation of c-Jun (Figure 5). Activation of had least eect on inhibiting uPAR transcription. c-Jun-GAL4 and ATF2-GAL4 required the presence of These results indicate that both the AP1 sites can their MAP-kinase phosphorylation sites in their
Oncogene RalA activation of uPAR transcription E Okan et al 1821 transactivation domain as mutation of Ser63 and Ser73 this tyrosine kinase in the signalling pathway between of c-Jun or Thr69 and Thr71 of ATF2 severely RalA and the activation of uPAR transcription. impaired their ability to respond to RalA signalling. Thus, RalA 72L results in the activation of both c-Jun and ATF2 and these factors are likely to stimulate Discussion uPAR transcription from the c-Jun-like binding site at 7184 bp and the ATF2-like binding site at 770 bp. Increased extracellular protease activity is one of the distinguishing features of metastatic cells and is crucial for the degradation of basement membranes and The tyrosine kinase c-Src functions downstream of RalA stromal extracellular matrix that facilitates the invasion in the activation of uPAR transcription of malignant cells. The urokinase plasminogen acti- Having established that c-Jun and ATF2 are activated vator receptor (uPAR) contributes to cell invasion by by RalA, we next tried to identify intermediate eectors focusing the serine protease activity of its ligand uPA, in the RalA signalling pathway. As c-Jun responded to the cell surface. It may also give additional more eectively to RalA 72L stimulation, we attempted advantages to tumour cells through activating Src- to inhibit RalA-induced transcription from the uPAR family tyrosine kinase, JAK/STAT or MAPK path- promoter with a trans-dominant inhibitory mutant of ways (Tang et al., 1998; Dumler et al., 1998; Nguyen et the c-Jun N-terminal kinase (SAPKbK4A). However, al., 1999). It is therefore important to identify the co-expression of SAPKbK4A failed to inhibit uPAR biochemical and molecular mechanisms that regulate transcription, demonstrating that it is not the MAPK the expression of uPAR in normal cells and how these that activates c-Jun or ATF2 (Figure 6). Similarly, are perturbed in transformed cells. We focused on the dominant-negative mutants of the extracellular signal- activation of uPAR transcription by Ras as it is a point regulated kinases ERK1 and ERK2 also failed to of convergence for diverse extracellular signal-stimu- inhibit transcription from the uPAR promoter. Indeed, lated pathways and frequently suers activating RalA-induced uPAR reporter expression was elevated mutations in a wide range of cancers. In addition, somewhat in their presence. These results imply that the introduction of oncogenic Ras into cells increases neither ERKs nor SAPKs act downstream of RalA in extracellular protease activity in general and the the activation of c-Jun and ATF2. expression of uPAR in particular (Westermark and As recent evidence suggests that the tyrosine kinase KaÈ haÈ ri, 1999; Lengyel et al., 1995; Jankun et al., 1991; c-Src functions downstream of RalA in the phosphor- Allgayer et al., 1999; Janulis et al., 1999; Muller et al., ylation of STAT 3 (Goi et al., 2000) we also explored 2000). The work of White et al. (1995) established that whether c-Src functions downstream of RalA in the the activation of multiple-ras eector pathways is activation of uPAR transcription. A dominant-negative necessary to induce cell transformation. The best- mutant of c-Src was a potent inhibitor of RalA characterized eectors of Ras are the serine kinase activation of uPAR transcription (Figure 6), placing
Figure 5 Activation of c-Jun and ATF2 by constitutively- activated RalA 72L. HEK 293 cells were transfected with the luciferase/reporter Gal-luc3 (containing ®ve Gal4-binding sites), Figure 6 HEK 293 cells were transfected with the 1.55 kb uPAR expression vectors for Gal-c-Jun (GJ), Gal-JunD 63/73 (GJm), luciferase reporter alone (7) or with an expression vector for Gal-ATF2 (GA) or Gal-ATF2D 69/71 (GAm) and either control RalA 72L (+) and vectors for trans-dominant inhibitory versions vector (7) or an expression vector for Ral A72L (+). Error bars of ERK1, ERK2, SAPKB or c-Src. Error bars show the standard show the standard deviations from two experiments each deviations from two experiments each performed with duplicate performed with duplicate samples for each point samples for each point
Oncogene RalA activation of uPAR transcription E Okan et al 1822 Raf-1, phosphoinositide 3 kinase (PI3K) and members of dominant-negative mutants. Dominant-negative of the Ral GDS family. Through these eectors, Ras mutants of ERK1, ERK2 and SAPKb failed to inhibit contributes to the transformed phenotype by promot- RalA activation of uPAR transcription. This ruled out ing anchorage-independent growth and by regulating the involvement of members of the Ras-ERK pathway apoptosis (Kauman-Zeh et al., 1997; Rodriguez- in RalA signalling. It also demonstrated that although Viciana et al., 1997; Webb et al., 1998; Yang et al., RalA signalling can activate c-Jun and ATF2, the 1998). Here we provide evidence that V12 H-Ras or failure of the trans-dominant inhibitor SAPKb to Ras 37G, the eector-loop mutant that activates Ral inhibit uPAR transcription indicates that JNK is not GDS, and constitutively-activated RalA can activate a prerequisite for the activation of RalA-dependent transcription from the proximal sequence of the uPAR uPAR transcription. At this stage the identity of the promoter. The recruitment of transcription factors to MAPK involved is unclear, however a possible this region of the uPAR promoter in response to H- candidate is the MAPK p38. ATF2 is a substrate for Ras or RalA-dependent signal transduction was also p38 and its activation may stimulate uPAR transcrip- demonstrated by electrophoretic mobility shift analysis. tion in the presence of the JNK inhibitor SAPKb.We RalA 72L consistently stimulated higher levels of are currently investigating this possibility. transcription from the 139 bp uPAR-promoter frag- The recent ®nding that c-Src functions downstream ment than from the larger 1.55 kb promoter fragment. of RalA in a signalling pathway that results in the This suggests RalA-mediated uPAR transcription may phosphorylation of the transcription factor STAT3 be subject to modulation by additional transcription (Goi et al., 2000) prompted us to investigate whether a factors in vivo. The presence or activity of these may be dominant-negative mutant of c-Src would also inhibit regulated through the function of other signal transdu- RalA-activated uPAR transcription. We have shown cing pathways. Interestingly, the uPA gene also that a dominant-negative mutant of c-Src totally appears to be regulated by a RalA-dependent mechan- inhibited RalA-dependent uPAR transcription, thereby ism (Aguirre-Ghisho et al., 1999b) raising the possibi- ®rmly placing c-Src in the signalling pathway between lity that RalA activity may co-ordinate the expression RalA and the activation of transcription. Whether the of uPA with its receptor uPAR. RalA-dependent Src tyrosine kinase is the direct eector of RalA is stimulation of uPAR transcription required the pre- currently unknown, however the recent ®nding that Src sence of at least one of the AP-1 sites at 770 or 7184 tyrosine kinase activity is directly stimulated through in the uPAR promoter, indicating that transcription is interacting with the GasorGai subunits of hetero- stimulated by a member of the Jun-family and that acts trimeric G-proteins makes this a distinct possibility downstream of RalA. The binding site at 7184 (Ma et al., 2000). (TGAGTCA) resembles that bound preferentially by While this paper was under review a paper describing Jun-Jun and Jun-Fos dimers; while the putative the activation of c-Jun by Ral appeared in print (de combined glucocorticoid response/AP-1-like element Ruiter et al., 2000). The majority of our ®ndings are in at 770 (TGACTCGC), resembles that preferentially agreement with theirs. We dier however in that we have bound by Jun-ATF dimers or ATF homodimers (Hai provided evidence in this paper that RalA signalling can and Curran, 1991; Karin et al., 1997). This indicates activate both c-Jun and ATF2, while de Ruiter et al. that RalA-dependent signalling could activate AP1 (2000) conclude that Ral signalling leads to the dimers containing combinations of c-Jun or ATF2 activation of c-Jun only. A possible explanation for proteins. Using hybrid proteins consisting of the these dierent conclusions may be due to dierences in activation domains of c-Jun or ATF2 fused to the the methods employed for the activation of Ral. More DNA-binding domain of GAL4, we showed that both importantly their criterion for the lack of activation of ATF2 and c-Jun are indeed activated by RalA signal ATF2 by Ral-mediated signal transduction is based on transduction. This was consistent with the ability of the failure of a dominant-negative RalB-N28 mutant to RalA 72L to stimulate transcription from a uPAR inhibit ATF2 phosphorylation. In our experiments the promoter fragment that contained alternate functional RalA signalling pathway was speci®cally activated using and mutated AP1 sites and with transcription from the a constitutively-activated RalA mutant. In comparison, proximal 139 bp uPAR promoter fragment that the activation of Ral by de Ruiter et al. (2000) was contains only the AP1/GRE site at 770 bp. Which initiated upstream of Ral through the treatment of cells AP1 site is used in vivo may be determined by the with insulin. If insulin activates a signalling pathway that abundance of speci®c AP-1 proteins. ATF2 expression results in the phosphorylation of ATF2 by a Ral- is constitutive unlike that of Jun and Fos whose levels independent mechanism, this pathway would not be are subject to changes in expression of their genes and inhibited by the dominant-negative RalB-N28 mutant. changes in their stability (Karin, 1995). Alternatively, This would lead to the conclusion made by de Ruiter et the AP1 site utilized may be dictated by what other al. (2000) that Ral-mediated signal transduction does not transcription factors occupy the promoter (at any activate ATF2. particular time). The identi®cation of a signalling pathway directed Having identi®ed the nature of the transcription towards the activation of uPAR transcription is factors activated by RalA-dependent signal transduc- signi®cant because numerous studies have demonstrated tion, we next tried to identify some of the downstream that the downregulation of uPAR is able to switch some components of the signalling pathway through the use human carcinomas into a protracted state of dormancy,
Oncogene RalA activation of uPAR transcription E Okan et al 1823 or to inhibit the metastasis of others (Aguirre-Ghiso et centrifugation at 1850 g for 10 min at 48C. The pellet was al., 1999c; Andreasen et al., 1997, 2000). The elucidation washed with PBS and rapidly resuspended in ®ve packed cell of the complete RalA-dependent signalling pathway may volumes (PCV) of hypotonic buer (10 mM HEPES pH 7.9 therefore enable the identi®cation of speci®c targets for at 48C, 1.5 mM MgCl2,10mM KCl). Cells were pelleted by therapeutic intervention that result in the downregula- centrifugation at 1850 g for 5 min and supernatant was discarded. The cells were then resuspended in three times the tion of uPAR expression in tumours. original PCV and incubated on ice for 10 min for swelling to occur before transferring them to a glass homogenizer (Dounce with type B tight pestle). Homogenization was Materials and methods carried out by 10 ± 20 gentle strokes. The nuclei were then pelleted by centrifugation at 3300 g for 15 min. The nuclei Plasmids were suspended in half the packed nuclear volume (PNV) of low salt buer (10 mM HEPES pH 7.9 at 48C, 1.5 mM The construction of the luciferase reporter pGL3-uPAR MgCl2, 0.02 M KCl and a half PNV of high salt buer with containing the uPAR-promoter region +9 to 71553 has 0.8, 1.2, 1.6 M KCl. were added drop wise. The suspensions been described previously (Muller et al., 2000). The plasmids were homogenized with a further two strokes in the Dounce containing constitutively activated mutants V12H-Ras con- homogenizer and the nuclear proteins were extracted for struction and use of V12 H-Ras and the partial eector 30 min by gentle mixing. Nuclear membranes were precipi- mutants V12/G37, V12/C40 and V12/S35 in the vector pSG5 tated by centrifugation at 25 000 g for 30 min and the (Stratagene) have also been described (Rodriguez-Viciana et supernatant was dialysed against (20 mM HEPES, 100 mM al., 1997) and were generously provided by J Downward KCl, 20% (v/v) Glycerol, 0.2 mM EDTA) until the (ICRF, London). The Gal4-luciferase reporter vector was conductivity of the sample reached equilibrium that of constructed by cloning 56Gal4 binding sites and the E4-tata 100 mM KCl. The extracts were removed from the dialysis promoter upstream of the luciferase gene in the vector pGL 3 bag and centrifuged at 25 000 g for 20 min to remove any (Promaga). The c-Jun-Gal4 and ATF2-Gal4 fusion proteins precipitate formed during dialysis. The protein concentration were constructed by cloning the activation domains of c-Jun was determined by colourimetric assay using Bio-RadTM DC (amino acids 1 ± 166) and ATF2 (amino acids 19 ± 96) protein assay kit and stored at 7808C downstream of the Gal4 DNA binding domain (amino acids 1 ± 147) in the vector pSG424 (Sadowski and Ptashne, 1989). The cDNAs encoding c-Jun and ATF2 were a gift from Dr Electrophoretic mobility shift assay (EMSA) Peter Angel. The transdominant JNK/SAPK inhibitor Binding reactions were carried out in a reaction volume of SAPKb has been described previously (Frost et al., 1997) 20 ml. Each reaction contained 50 mM KCl, 2 mM MgCl2, 15 mM HEPES pH 7.9, 2.5 mM EDTA, 2 mM Spermidine, 7% Cell culture, transfection and measurement of promoter activity glycerol, 2 mg of BSA, 2.5 mg of herring-sperm DNA, 1 mgof the oligonucleotide poly dI-dC, and between 10 and 30 mgof The human embryonic kidney cell line HEK293 were nuclear-protein extract. Reactions were preincubated for maintained in DMEM (Gibco/BRL) supplemented with 15 min at room temperature before the addition of 50 ± 10% FCS. The cells were cultured in a humidi®ed atmo- 100 fmoles (20 000 ± 50 000 c.p.m.) of double-stranded DNA sphere of 5% CO2. Cells were transfected when they had probe. The samples were incubated for a further 10 min at obtained approximately 70% con¯uency. For each transfec- room temperature before analysis on a 5% polyacrylamide gel tion, 2 ± 4 mg of luciferase reporter construct or plasmids in TBE buer. Polyacrylamide gels were dried and autoradio- encoding constitutively-activate alleles of small GTP-binding- graphed. EMSAs were analysed using a Fuji FLA-2000 proteins was used. The ®nal amount of DNA used in each PhosphorImager and Aida 211 data image-analysis software. transfection of cells in a 90 mm dish was adjusted to a total of 10 mg using pBSII (Bluescript) plasmid (Stratagene). Transfected cells were lysed 48 h post-transfection in Generation of DNA probes 250 mM KCl, 50 mM HEPES pH 7.5, 0.1% NP40, 10% Double-stranded synthetic probes for EMSA were prepared glycerol and luciferase activity was determined. Transfection by the polymerase chain reaction (PCR). Each 100 ml eciencies were determined by measuring the b galactosidase reaction contained 20 pmole each of forward and reverse activity resulting from co-transfection of 1 mg of pCH110 primers, 10 ng of template DNA, 200 mMol of each dNTP (Amersham Pharmacia Biotec) per plate. When necessary, the and 2.5 mM Mg2+ ions and 2.5 U of Taq DNA polymerase. amount of luciferase activity in each transfection experiment The ampli®cation reactions were performed for 30 cycles of was normalized to the relative transfection eciency of each 948C (1 min), 688C (1 min) 728C (1 min) followed by a experimental point. further 10 min extension reaction at 728C for 10 min. The PCR products were puri®ed from agarose gels using a TM Preparation of nuclear extracts Qiagen agarose gel extraction kit. Puri®ed PCR products were 5'-end labeled using T4-polynucleotide kinase and [a-32P] Nuclear extracts were prepared essentially as described by a dCTP as described previously (Sambrook et al., 1989). modi®ed procedure of Dignam et al. (1983). All solutions were used at 48C and contained 0.2 mM PMSF and 0.5 mM DTT, these were added immediately before use. Nuclear extracts were obtained from 30 ± 50, 90 mm culture dishes of Acknowledgments con¯uent HEK 293 cells. The extracts were prepared from The authors would like to thank Dr J Downward, Dr H cells that had been transfected with plasmids that constitu- Gille and Dr P Angel for their donations of plasmid tively expressed activated mutants of ras or ralA. Nuclear constructs. The work was funded in part by grants from extracts from untransfected cells were used as negative the European Social Fund to P Jones and the Wellcome control. Cells were harvested in PBS and pelleted by Trust UK to PE Shaw.
Oncogene RalA activation of uPAR transcription E Okan et al 1824 References
Aguirre-Ghiso JA, Alonso DF, Farias EF, Gomez DE and Medema RH, Kops GJ, Bos JL and Burgering BM. (2000). de Kier Joe' EB. (1999a). Eur. J. Biochem., 263, 295 ± Nature, 404, 782 ± 787. 304. Murai H, Ikeda M, Kishida S, Ishida O, Okazaki-Kishida M, Aguirre-Ghiso JA, Frankel P, Farias EF, Lu Z, Jiang H, Matsuura Y and Kikuchi A. (1997). J. Biol. Chem., 272, Olsen A, Feig LA, de Kier Joe' EB and Foster DA. 10483 ± 10490. (1999b). Oncogene, 18, 4718 ± 4725. Muller SM, Okan E and Jones P. (2000). Biochem. Biophys. Aguirre-Ghiso JA, Kovalski K and Ossowski L. (1999c). J. Res. Com., 270, 892 ± 898. Cell Biol., 147, 89 ± 104. Nguyen DH, Catling AD, Webb DJ, Sankovic M, Walker Allgayer H, Wang H, Shirasawa S, Sasazuki T and Boyd D. LA, Somylo AV, Weber MJ and Gonias SL. (1999). J. Cell (1999). Br.J.Cancer,80, 1884 ± 1891. Biol., 146, 149 ± 164. Andreasen PA, Kjoller L, Christensen L and Duy MJ. OkazakiM,KishidaS,HinoiT,HasegawaT,TamadaM, (1997). Int. J. Cancer, 72, 1 ± 22. Kataoka T and Kikuchi A. (1997). Oncogene, 14, 515 ± Andreasen PA, Egulend R and Petersen HH. (2000). Cell 521. Mol. Life Sci., 57, 25 ± 40. Rodriguez-Viciana P, Warne PH, Khwaja A, Marte BM, Bos JL. (1989). Cancer Res., 49, 4682 ± 4689. Pappin D, Das P, Water®eld MD, Ridley A and Down- Bos JL. (1997). Biochem et Biophys Acta., 1333, M19 ± M31. ward J. (1997). Cell, 89, 457 ± 467. Bos JL. (1998). EMBO J., 17, 6776 ± 6782. Sadowski I and Ptashne M. (1989). Nucl. Acid Res., 17, 7539. de Bruyn KMT, de Rooij J, Wolthuis RMF, Rehmann H, Sambrook J, Fritsch EF and Maniatis T. (1989). Molecular Wesenbeek J, Cool RH, Wittinghofer AH and Bos JL. Cloning: A Laboratory Manual, 2nd edn., Cold Spring (2000). J. Biol. Chem., 275, 29761 ± 29766. Harbour Press: Cold Spring Harbour, N.Y. de Ruiter ND, Wolthuis RMF, van Dam H, Burgering BMT Shetty S, Kumar A and Idell S. (1997). Mol. Cell Biol., 17, and Bos JL. (2000). Mol. Cell Biol., 20, 8480 ± 8488. 1075 ± 1083. Dignam JD, Lebovitz RM and Roeder RG. (1983). Nucl. Shields JM, Pruitt K, McFall A, Shaub A and Der CJ. Acids Res., 11, 1475 ± 1489. (2000). Trends Cell. Biol., 10, 147 ± 154. Dumler I, Weis A, Mayboroda OA, Maasch C, Jerke U, Smeal T, Binetruy B, Mercola DA, Birrer M and Karin M. Haller H and Gulba DC. (1998). J. Biol. Chem., 273, 315 ± (1991). Nature, 354, 494 ± 496. 321. Soravia E, Grebe A, De luca P, Helin K, Suh TT, Degan JL Frost JA, Steen H, Shapiro P, Lewis T, Ahn N, Shaw PE and and Blasi F. (1995). Blood, 86, 624 ± 635. Cobb MH. (1997). EMBO J., 16, 6426 ± 6438. Tang H, Kerins DM, Hao Q, Inagami T and Vaughan DE. Goi T, Shipitsin M, Lu Z, Foster DA, Klinz SG and Feig LA. (1998). J. Biol. Chem., 273, 18268 ± 18272. (2000). EMBO J., 19, 623 ± 630. Urano T, Emkey R and Feig LA. (1996). EMBO J., 15, 810 ± Hai T and Curran T. (1991). Proc. Natl. Acad. Sci. USA, 88, 816. 3720 ± 3724. Waltz DA, Natkin LR, Fijita RM, Wei Y and Chapman HA. Hofer F, Berdeaux R and Martin GS. (1998). Curr. Biol., 8, (1997). J. Clin Invest., 100, 58 ± 67. 839 ± 842. Webb CP, Van Aelst L, Wigler MH and Vande Woude GF. Hunter T. (1997). Cell, 88, 333 ± 346. (1998). Proc. Natl. Acad. Sci. USA, 95, 8773 ± 8778. Jankun J, Maher VM and McCormick JJ. (1991). Cancer Westermarck J and KaÈ haÈ ri VM. (1999). FASEB J., 8, 781 ± Res., 51, 1221 ± 1226. 792. Janulis M, Silberman S, Ambegaokar A, Gutkind JS and Wei Y, Waltz DA, Rao N, Drummond RJ, Rosenburg S and Shultz RM. (1999). J. Biol. Chem., 274, 801 ± 813. Chapman HA. (1994). J. Biol. Chem., 269, 32380 ± 32388. Johnson RS, Spiegelman BM, Hanahan D and Wisdom R. WeiY,LukashevM,SimonDI,BodarySC,RosenbergS, (1996). Mol. Cell Biol., 16, 4504 ± 4511. Doyle MV and Chapman HA. (1996). Science, 273, 1551 ± KarinM,LiuZ-GandZandiE.(1997).Curr. Opin. Cell 1555. Biol., 9, 240 ± 245. Webb CP, Van Aelst L, Wigler MH and Vande Woude GF. Karin M. (1995). J. Biol. Chem., 270, 16483 ± 16486. (1998). Proc. Natl. Acad. Sci. USA, 95, 8773 ± 8778. Kaumann-Zeh A, Rodriguez-Viciana P, Ulrich E, Gilbert White MA, Nicolette C, Minden A, Polverino A, Van Aelst C, Coer P, Downward J and Evan G. (1997). Nature, 385, L, Karin M and Wigler MH. (1995). Cell, 80, 533 ± 541. 544 ± 548. Wolthuis RM, de Ruiter ND, Cool RH and Bos JL. (1997). Kindzelskii AL, Laska ZO, Todd RF and Petty HR. (1996). EMBO J., 16, 6748 ± 6761. J. Immunol., 156, 297 ± 309. Wolthius RMF, Zwartkruis F, Moen TC and Bos JL. Kindzelskii AL, Eszes MM, Todd RF and Petty HR. (1997). (1998a). Curr. Biol., 8, 471 ± 474. Biophys. J., 73, 1777 ± 1784. Wolthuis RMF, Franke B, van Triest M, Bauer B, Cool RH, Kops GJ, de Ruiter ND, Vries-Smits AM, Powell DR, Bos Camonis JH, Akkerman JW and Bos JL. (1998b). Mol. JL and Burgering BM. (1999). Nature, 398, 630 ± 634. Cell Biol., 18, 2486 ± 2491. Koshelnick Y, Ehart M, Hufnagl P, Heinrich PC and Binder Xue W, Mizukami I, Todd III RF and Petty HR. (1997). BR. (1997). J. Biol. Chem., 272, 28563 ± 28567. Cancer Res., 57, 1682 ± 1689. Koshelnick Y, Ehart M, Stockinger H and Binder BR. Yamamoto T, Taya S and Kaibuchi K. (1999). J. Biochem., (1999). Thromb. Haemostas., 82, 305 ± 311. 126, 799 ± 803. Lengyel E, Stepp E, Gum R and Boyd D. (1995). J. Biol. Yang J-J, Kang J-S and Krauss RS. (1998). Mol. Cell Biol., Chem., 270, 23007 ± 23012. 18, 2586 ± 2595. Ma Y-C, Huang J, Ali S, Lowry W and Huang X-Y. (2000). Cell, 102, 635 ± 646.
Oncogene