Journal of Cell Science 113, 215-226 (2000) 215 Printed in Great Britain © The Company of Biologists Limited 2000 JCS0951 c-Cbl localizes to actin lamellae and regulates lamellipodia formation and cell morphology

Robin M. Scaife* and Wallace Y. Langdon Department of Pathology, University of Western Australia, QE II Medical Centre, Nedlands WA 6907, Australia *Author for correspondence (e-mail: [email protected])

Accepted 4 November 1999; published on WWW 13 January 2000

SUMMARY

Adhesive and locomotive properties of cells have key roles lamellae, lamellipodia and membrane ruffles. The in normal physiology and disease. Cell motility and induction of lamellipodia and membrane ruffles are also adhesion require the assembly and organization of actin inhibited during cell spreading and migration, conditions microfilaments into stress fibers, lamellipodia and when these structures are normally most prominent. The filopodia, and the formation of these structures is mediated inhibitory effect of truncated c-Cbl expression on by signalling through Rho GTPases. Here we identify c-Cbl lamellipodia formation can be reversed by mutational (a multi-adaptor proto-oncogene product involved in inactivation of its divergent SH2 domain, by the co- protein tyrosine kinase signalling) as an important expression of constitutively active Rac or by the regulator of the actin cytoskeleton. By immunofluorescence overexpression of c-Cbl. This study therefore identifies a microscopy we have determined that c-Cbl co-localizes cytoskeletal role for c-Cbl which may involve the regulation with the adaptor protein Crk to submembranous actin of Crk and Rac, and which is dependent on targeting of c- lamellae in NIH 3T3 fibroblasts and that c-Cbl’s actin Cbl to actin lamellae and the ability to recruit signalling localization requires specific SH3-binding sequences. protein(s) associated with its divergent SH2 domain. Further, we have found that truncation of this SH3-binding domain in c-Cbl profoundly alters the morphology of NIH Key words: Cbl, Actin cytoskeleton, src-homology, Signal 3T3 fibroblasts by inhibiting the formation of actin transduction, Cell shape

INTRODUCTION pathways. For example, developmental defects resulting from partial loss-of-function mutations of the Caenorhabditis The cytoskeleton is comprised primarily of microfilaments, elegans Let-23 receptor tyrosine kinase are reversed by microtubules and intermediate filaments, and is required for a mutation of the c-Cbl ortholog Sli-1 (Jongeward et al., 1995; plethora of cellular events. Among these are cell motility and Yoon et al., 1995). Also, engagement of a variety of plasma cell adhesion, which involve the organization of actin membrane receptors, including receptor tyrosine kinases, microfilaments into stress fibers, lamellipodia and filopodia. antigen receptors and integrin receptors results in tyrosine Stress fibers play important roles in the maintenance of cell phosphorylation of c-Cbl and its association with numerous shape and adhesion as a result of their contractile nature and cytoplasmic signalling proteins (Liu and Altman, 1998; their ability to interact with components of focal adhesion Miyake et al., 1998; Ojaniemi et al., 1997; Smit and Borst, complexes (Ben-Ze’ev, 1997; Hall, 1998). Lamellipodia and 1997; Thien and Langdon, 1998). membrane ruffles (sheet-like extensions of the plasma The c-Cbl protein contains many structural domains membrane that contain a meshwork of F-actin) and filopodia involved in protein interactions (see Fig. 1A). Firstly the N- (pointed actin-rich membrane protrusions) play important roles terminal v-Cbl region, which is highly conserved among all in cell spreading and motility (Lauffenberger and Horwitz, Cbl/Sli proteins, has been found to interact directly with 1996; Mitchison and Cramer, 1996). The critical importance of activated protein tyrosine kinases such as the EGF and PDGF the actin cytoskeleton in normal physiology and disease is receptors, ZAP-70 and Syk (Bonita et al., 1997; Bowtell and hence largely due to its central role in cell motility and Langdon, 1995; Galisteo et al., 1995; Lupher et al., 1996, 1998; adhesion. The assembly and organization of the actin Thien and Langdon, 1997). This interaction is dependent on cytoskeleton is subject to input from numerous signal the tyrosine phosphorylation of these kinases and a divergent transduction pathways involving Rho GTPases (Hall, 1998). Src homology (SH) 2 domain (designated as SH2*) in the N- A common component of many signalling pathways is the terminal region of c-Cbl (Meng et al., 1999). The maintenance multi-adaptor proto-oncogene product c-Cbl. Genetic and of SH2* binding to phosphotyrosine residues in protein biochemical studies have implicated c-Cbl in the attenuation tyrosine kinases is essential for fibroblast transformation and of receptor-coupled tyrosine kinase mediated signalling transcriptional activation in T cells by oncogenic forms of Cbl 216 R. M. Scaife and W. Y. Langdon

(Bonita et al., 1997; Thien and Langdon, 1997; van Leeuwen Cells were grown to confluency in DME + 10% FCS prior to et al., 1999; Zheng et al., 1998). wounding of the monolayers with a plastic pipette tip. Migration of c-Cbl also contains a highly conserved Ring finger motif of cells into the monolayer wound was monitored using an inverted unknown function while the C-terminal portion contains phase contrast microscope (Leitz, ×20 objective lens) equipped for multiple SH3 domain-binding proline motif sequences as well capture of digital images (Optimas software). as several tyrosine residues which, when phosphorylated, form Immunofluorescence microscopy SH2 domain-binding sites (Andoniou et al., 1996; Blake et al., Cells were seeded at low density onto coverslips coated with 1991; Feshchenko et al., 1998). These Src homology domain- polylysine (0.1 mg/ml) and cultured at least 24-48 hours prior to binding sequences serve as association sites for specific signal fixation in 4% p-formaldehyde/PBS. The fixed cells were then transducing proteins such as Grb2, phosphatidylinositol (PI) 3- permeabilized for 2 minutes with 0.2% Triton X-100 in PBS kinase, CAP, Nck, Vav and Crk, however, it is not known how containing 2.5 mg/ml BSA. Coverslips were rinsed with PBS and these associations mediate, or affect, the biological activity of incubated for 60 minutes with either 0.5 µg/ml TRITC-phalloidin c-Cbl. In hematopoietic cells, attenuation of receptor induced (Sigma) or primary antibodies (2.5 µg/ml anti-c-Cbl mAb, signal transduction by c-Cbl appears to involve effects on Transduction Laboratories catalog number C40320; 100 ng/ml 3F10 anti-HA mAb, Boehringer; 8 µg/ml affinity-purified anti-c-Cbl R2 specific intracellular kinases such as Syk and ZAP-70 (Lupher µ et al., 1998; Murphy et al., 1998; Naramura et al., 1998; Ota polyclonal antibodies (Blake et al., 1993) or 2 g/ml anti-CrkII mAb, Transduction Laboratories catalog number C12620) at 37¡C for 90 and Samelson, 1997; Thien et al., 1999) and/or modulation of minutes in PBS containing 2.5 mg/ml BSA. Following a PBS wash the activity of low-molecular mass GTPases such as Rap1 the coverslips were incubated with either 5 µg/ml biotin-SP-goat anti- (Boussiotis et al., 1997). The attenuation of receptor signalling mouse antibodies (Jackson Laboratories, catalog number 115-065- by c-Cbl also involves the enhanced ubiquitination and 003), 1 µg/ml biotin-SP-goat anti-rat antibodies (Jackson degradation of the EGF, PDGF and CSF-1 receptors (Lee et Laboratories, catalog number 112-065-003) or TRITC-conjugated al., 1999; Levkowitz et al., 1998; Miyake et al., 1998, 1999). swine anti-rabbit antibodies (Dako, catalog number R1560) at 37¡C Recently, c-Cbl has also been implicated in PI 3-kinase for 60 minutes in PBS containing 2.5 mg/ml BSA. Biotin-labeled dependent macrophage spreading and migration (Meng and antigen-antibody complexes were then visualized by incubation for µ Lowell, 1998). Since PI 3-kinase and Crk are involved in Rac- 60 minutes with PBS containing 2.5 mg/ml BSA and 2 g/ml Alexa dependent effects on the actin cytoskeleton (Kiyokawa et al., 488-conjugated Streptavidin (Molecular Probes). Midbodies were visualized by staining with 7.5 µg/ml anti-Tubulin antibodies (Sigma 1998a; Rodriguez-Viciana et al., 1997) these results suggest mAb, catalog number T-4026) followed by FITC-conjugated anti- that c-Cbl, by binding to these proteins, could play a role in mouse secondary antibody (Silenus) diluted in PBS containing 2.5 cytoskeletal assembly and organization. In order to assay for mg/ml BSA (60 minutes at 37¡C). Following a PBS rinse coverslips c-Cbl involvement in modulation of the cytoskeleton we have were mounted with SlowFade Light Antifade reagent (Molecular expressed full-length and truncated versions of c-Cbl in NIH Probes). Images of representative fields were obtained with Comos 3T3 fibroblasts. We report here that c-Cbl is targeted with Crk and Confocal Assistant software (Bio-Rad) following capture on a to membrane ruffle-associated actin lamellae. Deletion of c- Nikon Diaphot 300 microscope equipped for UV laser scanning Cbl SH3 domain-binding sequences prevents its localization confocal microscopy (Bio-Rad MRC 1000/1024). The 543 nm to these structures and as a result profoundly alters the excitation signal from TRITC and the 488 nm excitation signals from organization of the actin cytoskeleton and cell morphology. EGFP and Alexa 488 were collected sequentially with 580/32 and 522/35 nm emission filters, respectively. For each Cbl construct, Our results therefore suggest that by influencing components visualisation of the actin cytoskeleton by immunofluorescence of the cytoskeleton c-Cbl may regulate a number of microscopy (×400 magnification) permitted determination of the fundamental cellular properties. number of cycling cells that contained visually distinct lamellipodia. Triplicate counts of 50 cells each were scored for the presence of strong linear actin staining at the cell periphery. The short actin fragments (see Fig. 2A and B, 3) associated with focal adhesion sites MATERIALS AND METHODS (identified independently by anti-Vinculin immunofluorescence staining, data not shown) in 480-Cbl expressing cells were not counted Cell culture, transfections and monolayer wound healing as actin lamellae. Similarly, staining of the actin cytoskeleton NIH 3T3 cells (passage number 127) were obtained from ATCC and permitted determination of the number of spreading (freshly plated) cultured (up to passage number 150) in DMEM (Trace Biochemicals) cells that formed actin lamellae. Triplicate counts of 50 cells each containing 10% FCS (Gibco/BRL) and 2 mM L-glutamine (Trace were scored for the presence of a continuous arrangement of actin Biochemicals) at 37¡C and 5% CO2. lamellae around the cell periphery. CrkII was expressed by transient transfection of column purified (Bresatec) pUCCAGGS cDNA-plasmid constructs (Klemke et al., Cell lysis, immunoprecipitation and western blotting 1998) using LipofectAmine Plus reagent (Gibco/BRL), as described by Cells grown on plastic culture dishes (Falcon) were lysed in ice-cold the manufacturer. Constitutively active Rac was expressed by transient 20 mM Tris-HCl buffer (pH 7.4) containing 150 mM NaCl, 0.2% co-transfection of column purified pRK5 plasmid containing the cDNA Triton X-100, 1 mM Na3VO4 and protease inhibitors. Anti-Crk and coding for L61 Rac1 (Lamarche et al., 1996) with column purified anti-PI 3-kinase immunoprecipitations were carried out by incubation enhanced green fluorescent protein expressing plasmid pEGFP of clarified lysates (1,500 g, 3 minutes) with either 4 µg/ml anti-Crk (Clontech). pJZenNeo vectors encoding HA epitope tagged c-Cbl mAb (Transduction Laboratories), or 4 µl anti-PI 3-kinase p85 (UBI, cDNAs have been described previously (Andoniou et al., 1994). catalog number 06-195) and BSA-blocked (5 mg/ml) Protein A- Constructs were either electroporated into Ψ2 packaging cells to Sepharose (Pharmacia) for 4 hours at 4¡C, followed by three washes generate virus particles for infection of NIH 3T3 cells which were with lysis buffer. Aliquots of cell lysates and immunoprecipitates were selected with 400 µg/ml active G418 (Gibco/BRL), or expressed subjected to SDS-PAGE and transferred onto nitrocellulose transiently by transfection of the cDNAs (in pSRαneo) using Fugene membranes (Amersham). Membranes were blocked with 10% non-fat transfection reagent (Boehringer) as recommended by the manufacturer. milk (Carnation) in TBS containing 0.5% Tween-20, and probed with Regulation of cell shape and the cytoskeleton by Cbl 217

Fig. 1. Deletion of the C-terminal sequences of c-Cbl alters the morphology of NIH 3T3 cells. (A) Linear representation of c-Cbl. Structural motifs are shown in addition to the positions of C-terminal truncations for 655-Cbl, 480-Cbl and v-Cbl (truncation at residue 357). (B) Anti-HA western blot of lysates derived from cells stably expressing HA tagged full-length c-Cbl, 480-Cbl and v-Cbl. The relative mobilities of molecular mass markers are indicated. (C) Wild-type NIH 3T3 cells (1) and cells stably expressing full-length HA-c-Cbl (2), HA-480-Cbl (3) or HA-v-Cbl (4) were observed by phase contrast microscopy. Bar, 200 µm. (D) High magnification phase contrast microscopy images of wild- type NIH 3T3 cells (1) and cells stably expressing HA-480-Cbl (2). Bar, 50 µm. antibodies directed against either c-Cbl (Transduction Laboratories), to v-Cbl (Andoniou et al., 1994). The expression of HA-tagged HA tag (mAb 3F10), Crk (mAb Transduction Laboratories) or PI 3- c-Cbl, 480-Cbl and v-Cbl in NIH 3T3 cells is shown in Fig. kinase p85 (rabbit polyclonal serum generously provided by Dr L. 1B by anti-HA blotting. By densitometric scanning of anti-HA Varticovski), followed by HRP-conjugated secondary antibody and anti-Cbl blots we determined that these HA-Cbl constructs (Zymed). Antigens were then visualized by chemiluminescence are approximately 30-fold more abundant than endogenous c- (ECL, Amersham) using Hyperfilm MP (Amersham). Electrophoresis reagents were purchased from Bio-Rad. All other Cbl. reagents were obtained from Sigma Chemical company, unless stated We found that the overexpression of c-Cbl had no otherwise. discernible effect on cell shape or cell-cell contacts. Both wild- type NIH 3T3 cells and c-Cbl overexpressing cells, viewed by phase contrast microscopy, grew as clusters of elongated overlapping cells (Fig. 1C, 1 and 2). 480-Cbl expressing cells, RESULTS by contrast, grew in a strikingly different pattern. The 480-Cbl expressing cells grew in an ordered and uniform pattern on the Deletion of the C-terminal sequences of c-Cbl alters culture substrate and exhibited greatly reduced cell-cell the morphology of NIH 3T3 cells contacts (Fig. 1C, 3). These cells were also markedly altered The c-Cbl proto-oncogene contains multiple protein-protein in that they appeared more symmetrical, and failed to become interaction sites (shown in Fig. 1A). The N-terminal SH2* highly elongated (Fig. 1D). As we previously reported domain is highly conserved in all Cbl homologues and its (Andoniou et al., 1994), the 480-Cbl expressing cells are not ability to bind protein tyrosine kinases appears essential for all transformed and their morphology and growth pattern is also Cbl functions. In order to determine the biological activities of markedly distinct from those expressing oncogenic v-Cbl (Fig. c-Cbl’s C-terminal protein interaction sequences we expressed 1C, 4). a markedly truncated form of c-Cbl (480-Cbl) that has lost all SH3-binding proline motifs (between amino acid residues 480 Alteration of the actin cytoskeleton by 480-Cbl and 655) in addition to the three major SH2-binding Since cell morphology and adhesion are dependent on the phosphotyrosine residues (Y700, Y731, Y774). This form of cytoskeleton (Hall, 1998), we examined the effects of 480-Cbl Cbl protein is not oncogenic since further truncation by expression on the organization of cytoskeletal filaments. We deletion of sequences encompassing the Ring finger domain found that although 480-Cbl expression did not affect the are necessary to generate a transforming oncogene equivalent microtubule array (data not shown), the organization of the 218 R. M. Scaife and W. Y. Langdon

Fig. 2. (A,B) Alteration of the actin cytoskeleton by 480- Cbl. Wild-type NIH 3T3 cells (1), and cells stably expressing full-length HA-c-Cbl (2), or HA-480-Cbl (3) were grown in DMEM containing 10% FCS. Cells were fixed and stained with TRITC phalloidin. Confocal immunofluorescence microscopy Z-series projection images are shown in A while a single 1.0 µm Z-series apical section of the cells is shown in B. Wide arrowheads indicate membrane lamellipodia and narrow arrowheads indicate actin lamellae. Bars, 30 µm. actin cytoskeleton appeared significantly perturbed. expressing cells only revealed small areas of dense F-actin Immunofluorescent staining with fluor-tagged phalloidin staining at the cell periphery, and an absence of continuous revealed that subconfluent c-Cbl overexpressing NIH 3T3 cells actin lamellae (Fig. 2B, compare 1 and 2 with 3). resembled wild-type NIH 3T3 cells in that they also frequently displayed one or more lamellipodial protrusions along the cell periphery (indicated by arrowheads in Fig. 2A, 1 and 2). Wild- type and c-Cbl overexpressing NIH 3T3 cells also exhibit numerous membrane ruffle-associated actin lamellae (dense and highly organized F-actin bands). These actin lamellae were readily visualized by individual confocal microscopy Z-series sections (indicated by arrowheads in Fig. 2B, 1 and 2). Hence, c-Cbl overexpression did not appear to significantly alter the organization of the actin cytoskeleton. By contrast, we found that lamellipodia and actin lamellae formation were greatly reduced by expression of 480-Cbl. The periphery of these cells contained few lamellipodial protrusions (Fig. 2A, 3). Further, individual cross sections (confocal Z-series) of 480-Cbl

Fig. 3. Morphological and cytoskeletal effects of truncated Cbl require deletion of the carboxy-terminal SH3-binding proline motifs. (A) NIH 3T3 cells stably expressing HA-655-Cbl (1), HA-563-Cbl (2) or HA-528-Cbl (3) were grown in DMEM containing 10% FCS. Cells were fixed and stained with TRITC phalloidin. Confocal immunofluorescence microscopy Z-series projection images are shown in A. A prominent actin-rich lamellipodium is indicated by the arrowhead.Bar, 30 µm. (B) Wild-type NIH 3T3 cells and cells stably expressing full-length HA-c-Cbl, HA-655-Cbl, HA-563-Cbl, HA-528-Cbl or HA-480-Cbl were grown in DMEM containing 10% FCS. Cells were fixed and stained with TRITC phalloidin and the percentage of lamellipodia containing cells was determined following observation by immunofluorescence microscopy. Average values of triplicate counts of 50 cell are shown ± s.e.m. Regulation of cell shape and the cytoskeleton by Cbl 219

Fig. 4. Attenuation of motility- and adhesion-induced actin lamellae formation during by 480-Cbl expression. Wild-type NIH 3T3 cells (A,C,E), and cells stably expressing HA-480-Cbl (B,D,F) were cultured in DMEM containing 10% FCS. Double-label images of cytokinetic cells (A and B) were obtained by confocal immunofluorescence microscopy of fixed cell following staining with anti-Tubulin (green) and TRITC phalloidin (red). Anti-tubulin stained midbodies are evident for both pairs of cytokinetic cells. Cytokinetic wild-type cells contain numerous continuous bands of actin lamellae (indicated by an elongated arrowhead) while 480-Cbl expressing cells lack these structures and only contain small areas of dense F-actin staining at the periphery. Images of migrating cells (C and D) were obtained by phase contrast microscopy following monolayer wound healing (the wound edge is indicated by a dotted line). The wide arrowhead indicates a membrane lamellipodium. Images of spreading cells (E and F) were obtained by fixation 60 minutes after plating on polylysine coated glass coverslips. TRITC phalloidin staining was visualized as single Z-series confocal immunofluorescence microscopy sections. Wild-type cells formed extensive lamellipodia and actin-lamellae (indicated by an elongated arrowhead) while freshly plated 480-Cbl expressing cells primarily formed filopodia. Bars: 30 µm (A,B,E,F); 50 µm (C and D).

cytoskeleton of 655-Cbl overexpressing NIH 3T3 cells resembled those of wild-type NIH 3T3 cells in that the cells were elongated and they frequently displayed one or more lamellipodial protrusions along the cell periphery (Fig. 3A, 1). By contrast, we found that further deletion to amino acid residue 563, or 528, resulted in morphological and cytoskeletal changes characteristic of 480-Cbl cells (Fig. 3A, 2 and 3). For each Cbl construct the percentage of cells that have visually distinct lamellipodia was determined by immunofluorescence microscopy of cells stained with fluor- tagged phalloidin. These results, shown in Fig. 3B, revealed a marked decrease in the percentage of cells with lamellipodia that occurred upon extension of the C-terminal truncation from amino acid residue 655 to residue 563 of the Cbl construct. The c-Cbl sequence between residues 655 and 563 contains four potential SH3 binding sites and the deletion of one or more of these sites is therefore necessary for the inhibition of lamellipodia formation. Interestingly this sequence does not encompass the known Grb2 SH3 binding sites which are located between amino acids 494-499 and 532-550 (Donovan et al., 1996). The formation of motility- and adhesion-induced actin lamellae is attenuated by 480-Cbl expression Lamellipodia and submembranous actin lamellae are associated with the motile leading-edge of cells, and their formation can be greatly enhanced by induction of cell motility and adhesion. We therefore assayed the effects of 480-Cbl Morphological and cytoskeletal effects of truncated expression on lamellipodia and actin lamellae formation Cbl require deletion of the carboxy-terminal SH3- induced by daughter-cell separation following cytokinesis and binding proline motifs by cell migration in response to monolayer wound healing. Truncation of Cbl at amino acid residue 480 results in a loss Cells in the final stages of cytokinesis were identified by of all SH3-binding proline motifs (between residues 480 and their characteristic midbody staining. Double labeling with 635) as well as SH2-binding phosphotyrosine residues 700, fluor-tagged phalloidin demonstrated that daughter-cell 731 and 774. We therefore expressed Cbl truncated at amino separation induced extensive membrane ruffle formation in the acid residue 655 (655-Cbl) to determine whether loss of Cbl majority (greater than 80%) of wild-type (Fig. 4A) and c-Cbl SH2 binding sequences could result in morphological and/or overexpressing NIH 3T3 cells (data not shown). On the other cytoskeletal changes. Immunofluorescent staining with fluor- hand, we found that formation of membrane ruffle-associated tagged phalloidin revealed that the morphology and actin actin lamellae was suppressed in cytokinetic 480-Cbl 220 R. M. Scaife and W. Y. Langdon

Fig. 5. Constitutively active Rac reverses the effect of 480- Cbl on lamellipodia formation. Wild-type (A and C) and HA- 480-Cbl expressing NIH 3T3 cells (B and D) were transiently transfected with plasmids coding for L61 Rac1 and pEGFP. The cells were then fixed, stained with TRITC-phalloidin and observed as single Z-series confocal immunofluorescence microscopy sections. The TRITC channel is shown in A and B while the EGFP channel is shown in C and D. Two transfected wild-type and 480-Cbl expressing cells are evident from the EGFP signal. Bar, 30 µm. expressing cells (Fig. 4B). We found that actin lamellae were actin cytoskeleton was visualized by staining with TRITC- few or altogether absent in the majority (greater than 65%) phalloidin. We found that, similar to transfection of wild-type of cytokinetic 480-Cbl expressing cells. Suppression of NIH 3T3 cells (Fig. 5A), transfection of 480-Cbl expressing lamellipodia and actin-lamellae formation by 480-Cbl cells with the L61 Rac1 cDNA resulted in extensive expression was also evident in migrating cells. We found that lamellipodia formation around the entire cell periphery (Fig. motile wild-type NIH 3T3 cells at a monolayer wound edge 5B). Constitutively active Rac protein could therefore reverse form extensive lamellipodia (Fig. 4C) while lamellipodia are the inhibitory effect of 480-Cbl on lamellipodia formation, mostly absent at a wound edge of 480-Cbl expressing cells suggesting that this effect is due to signalling events ‘upstream’ (Fig. 4D). of Rac. Cell substrate adhesion generally induces extensive rearrangement of the actin cytoskeleton. Hence, we examined Alteration of the actin cytoskeleton and cell whether 480-Cbl expression can affect adhesion-induced morphology by 480-Cbl requires a functional SH2* lamellipodia and filopodia formation. We found that while domain spreading (i.e. freshly plated) wild-type NIH 3T3 cells contain We wished to determine whether the effects of 480-Cbl on cell abundant actin lamellae (extensive actin lamellae formation in morphology and the actin cytoskeleton could result from greater than 80% of cells) (Fig. 4E), formation of these competition with endogenous c-Cbl for targets of the adhesion induced actin-based structures was suppressed in phosphotyrosine binding SH2* domain. To assay this we 480-Cbl expressing cells (substantial actin lamellae formation introduced the G306E null-mutation into 480-Cbl. This in fewer than 10% of cells) (Fig. 4F). Rather, the periphery of mutation was originally identified as a loss-of-function freshly plated 480-Cbl expressing cells had numerous actin- mutation in the C. elegans ortholog Sli-1 (Yoon et al., 1995) containing filopodial spikes, suggesting selective impairment and abolishes the phosphotyrosine binding activity of the c-Cbl of adhesion induced signalling to the actin cytoskeleton by SH2* domain (Lupher et al., 1997; Meng et al., 1999; Thien expression of 480-Cbl. and Langdon, 1997). Thus, if 480-Cbl is functioning in a dominant negative manner by competing with c-Cbl for SH2* Constitutively active Rac reverses the effect of 480- targets then introduction of the G306E mutation into 480-Cbl Cbl on lamellipodia formation ought to abolish this effect and revert the cells to a wild-type Organization of the actin cytoskeleton is regulated by the low- phenotype. We found that the expression of G306E-480-Cbl molecular mass Rho GTPases. Specifically, the GTPases (Fig. 6A) did indeed inactivate the effects of 480-Cbl since the Cdc42 and TC10 have been implicated in filopodia formation cells showed a normal NIH 3T3 growth pattern and (Neudauer et al., 1998; Nobes et al., 1995) while Rac has been morphology (Fig. 6B). Further, unlike 480-Cbl expression, shown to play a central role in lamellipodia and membrane G306E-480-Cbl did not alter the actin cytoskeleton in terms of ruffle formation (Ridley and Hall, 1992). Hence we wished to lamellipodia (Fig. 6C) and actin lamellae formation (Fig. 6D). determine whether the inhibitory influence of 480-Cbl on A functional SH2* domain is therefore necessary for the lamellipodia formation might be due to modulation of Rac modulation of cell morphology and the actin cytoskeleton by mediated signal transduction. To assess whether 480-Cbl can 480-Cbl. modulate Rac mediated signalling, we examined whether constitutively active Rac could revert the effect of 480-Cbl c-Cbl overexpression induces actin lamellae and expression on the actin cytoskeleton. Wild-type and 480-Cbl lamellipodia in 480-Cbl cells expressing NIH 3T3 cells were cotransfected with Requirement of a functional SH2* domain for modulation of constitutively active Rac (L61 Rac1) as well as the transfection cell morphology and the actin cytoskeleton by 480-Cbl marker enhanced green fluorescent protein (EGFP), and the indicates these effects could be the result of competition Regulation of cell shape and the cytoskeleton by Cbl 221

Fig. 6. Alteration of the actin cytoskeleton and cell morphology by 480-Cbl requires a functional SH2* domain. Cell lysates from NIH 3T3 cells stably expressing HA-480-Cbl or HA-G306E-480-Cbl were subjected to SDS-PAGE and probed by western blotting using anti-HA antibodies (A). Cells expressing HA-G306E-480-Cbl were observed either directly by phase contrast microscopy (B) (bar, 200 µm) or by confocal immunofluorescence microscopy following staining with TRITC-phalloidin as a Z-series projection (C) and as a single 1.0 µm Z-series apical section of the cells (D). The wide arrowhead indicates a membrane lamellipodium and the narrow arrowhead indicates an actin lamellum. Bar, 30 µm.

submembranous actin lamellae (Fig. 8C, 1 and 2). The staining of the actin lamellae with anti-HA antibodies was specific for the expressed antigen since no anti-HA staining of submembranous actin lamellae was detectable in wild-type NIH 3T3 cells (Fig. 8B and C, 3). Specific SH3-binding proline motifs target c-Cbl to actin lamellae The SH2* domain, SH3 binding proline motifs and SH2 binding C-terminal tyrosine residues of c-Cbl permit direct interactions with a multitude of proteins. In order to determine which of these distinct regions of c-Cbl is responsible for its association with the actin cytoskeleton, we examined the subcellular distribution of N-terminally mutated and C- terminally truncated c-Cbl constructs by immunofluorescence microscopy. Although all previously identified in vivo functional properties of c-Cbl and oncogenic Cbl proteins are abolished by the G306E mutation, we found that full-length c- Cbl containing the G306E mutation in the SH2* domain still between c-Cbl and the truncated constructs for an SH2* domain binding partner. To determine whether the action of 480-Cbl may indeed be due to such competition we overexpressed full-length c-Cbl in 480-Cbl expressing cells. We found that transient overexpression of c-Cbl reversed the 480-Cbl induced morphological and cytoskeletal alterations. 480-Cbl cells transiently transfected with an EGFP control plasmid retained their characteristic morphological and cytoskeletal properties (Fig. 7A and C), whereas overexpression of c-Cbl induced the formation of lamellipodia and actin lamellae (Fig. 7B and D). c-Cbl localizes to submembranous actin lamellae Further evidence for c-Cbl involvement in modulating the actin cytoskeleton was obtained by examining the subcellular distribution of HA tagged c-Cbl. The subcellular distribution of c-Cbl was determined by immunofluorescence microscopy using two monoclonal antibodies, one directed against the HA tag and the other against c-Cbl. These two antibodies Fig. 7. c-Cbl overexpression induces actin lamellae and lamellipodia in 480-Cbl are highly specific for HA-c-Cbl since western blots cells. NIH 3T3 cells expressing 480-Cbl were transiently transfected with of cell lysates probed with these antibodies contained plasmids coding for EGFP (A and C) or HA-c-Cbl (B and D). The cells were fixed 24 hours post transfection, and observed by confocal immunofluorescence a single immunoreactive band corresponding to the microscopy following staining with TRITC-phalloidin (B and D) or anti-HA antigen (Figs 1B and 8A). In addition to diffuse antibodies, followed by biotinylated secondary antibodies, and Alexa 488- perinuclear staining (Blake et al., 1993; Levkowitz et conjugated Streptavidin (C). EGFP fluorescence is shown in A. The prominent al., 1998), both antibodies stained peripheral filaments actin-rich lamellipodia in the c-Cbl transfected cells are indicted by arrow heads. (Fig. 8B, 1 and 2) which co-localized precisely with Bar, 30 µm. 222 R. M. Scaife and W. Y. Langdon

Fig. 8. c-Cbl localizes to submembranous actin lamellae. (A) Lysates from NIH 3T3 cells stably expressing full-length HA-c- Cbl, and wild-type NIH 3T3 cells were subjected to SDS- PAGE and probed by western blotting using anti-c-Cbl monoclonal antibodies. The relative mobilities of molecular mass markers are indicated. (B,C) Cells stably expressing full-length HA-c- Cbl (1 and 2), and wild-type NIH 3T3 cells (3) were grown in DMEM containing 10% FCS. Cells were fixed, stained with antibodies directed against c-Cbl (1) or HA-tag (2 and 3) followed by biotinylated secondary antibodies, Alexa 488- conjugated Streptavidin (B) and TRITC phalloidin (C) and viewed as single 1.0 µm confocal immunofluorescence microscopy Z-series sections. Arrowheads indicate positions of actin lamellae. Bars, 30 µm. localized to actin lamellae of NIH 3T3 cells (Fig. 9A, 1). containing 10% FCS). HA-480-Cbl cells were used as co- Similarly, we found that a Cbl construct from which the SH2 immunoprecipitation controls since this form of Cbl is unable binding phosphotyrosine residues have been removed by to interact with either Crk or p85 (Liu and Altman, 1998; truncation at amino acid 655 also localized to these actin Miyake et al., 1998; Thien and Langdon, 1998). To determine structures (Fig. 9A, 2). However, we found that further whether the association of c-Cbl with Crk could be of relevance truncation of the HA-Cbl constructs (563-Cbl, 528-Cbl), to cytoskeletal organization, we compared the intracellular resulting in progressive removal of potential SH3-binding distributions of both Crk and c-Cbl following their proline motifs, reduced anti-HA staining of actin lamellae to overexpression in NIH 3T3 cells. Indeed, we found Crk to be background levels (Fig. 9, 3 and 4). These results therefore abundantly localized to membrane ruffle-associated actin suggest that the targeting of c-Cbl to the cytoskeleton requires lamellae of NIH 3T3 cells (Fig. 10C, 1 and 2) in a pattern an interaction between its proline motif sequences and an actin- which is identical to that of c-Cbl (Fig. 10C, 3 and 4). associated SH3 domain protein. Therefore interaction with c-Cbl may prove to be one of several ways to localize Crk to actin lamellae (Klemke et al., 1998). Interaction and co-localization of Crk with c-Cbl Rac dependent lamellipodia and membrane ruffle formation can be initiated by signalling pathways involving Crk and PI DISCUSSION 3-kinase (Dolfi et al., 1998; Rodriguez-Viciana et al., 1997). In cells where c-Cbl is tyrosine phosphorylated both of these Attenuation of tyrosine kinase signalling by c-Cbl has signalling molecules have been shown to associate with c-Cbl previously been suggested on the basis of both genetic and in an SH2 dependent manner. Since 480-Cbl has lost c-Cbl’s biochemical evidence. It has been determined that partial loss- known SH2-binding sites, we wished to address the possibility of-function mutations in the EGF receptor ortholog of C. that c-Cbl’s interaction with Crk or the p85 regulatory subunit elegans (Let-23) could be reversed by mutation of the Cbl/Sli- of PI 3-kinase might be relevant for the role of c-Cbl in the 1 gene (Jongeward et al., 1995). Thus, inhibition of Sli-1 regulation of the actin cytoskeleton. Hence, we initially probed permits sufficient signal transduction from the partial loss-of- for a physical association between Crk and p85 with c-Cbl in function Let-23 for normal vulval induction to take place. This cycling NIH 3T3 cells. Lysates were immunoprecipitated with suggests that wild-type Sli-1 attenuates signalling from the Let- either anti-Crk (Fig. 10A) or anti-p85 antibodies (Fig. 10B) and 23 receptor. Since Sli-1 has extensive sequence homology to then probed for co-immunoprecipitated c-Cbl. We found that c-Cbl, a similar tyrosine kinase signal attenuating activity has c-Cbl could associate with Crk (Fig. 10A), and to a lesser been proposed for c-Cbl (Yoon et al., 1995). However, due to extent with p85 (Fig. 10B), in lysates derived from adherent the smaller size of Sli-1, this sequence homology is mainly cells propagated under growth conditions that were used restricted to the N-terminal portion (i.e. the SH2* domain and throughout this study (i.e. subconfluent cycling cells in media Ring finger). The presence in c-Cbl of more extensive C- Regulation of cell shape and the cytoskeleton by Cbl 223

Fig. 9. SH3-binding proline motifs target c-Cbl to actin lamellae. (A,B) Cells stably expressing HA-G306E-c-Cbl (1), or transiently expressing HA-655-Cbl (2), HA-563-Cbl (3) or HA-528-Cbl (4) were grown in DMEM containing 10% FCS. Cells were fixed, stained with antibodies against HA-tag followed by biotinylated secondary antibodies, Alexa 488-conjugated Streptavidin (A) and TRITC phalloidin (B). Arrowheads indicate positions of actin lamellae. Bar, 30 µm. terminal sequences containing multiple SH2 and SH3 domain- structures. Indeed we found by actin staining and confocal binding sites indicates that mammalian c-Cbl may have microscopy that expression of 480-Cbl can profoundly inhibit additional functions in addition to acting like Sli-1 in the lamellipodia formation. This is most likely the reason why attenuation of receptor tyrosine kinase signalling. Indeed, these cells display an altered morphology and lack the induction of c-Cbl phosphorylation on C-terminal tyrosine overlapping highly elongated protrusions that are evident in residues following activation of cell surface receptors and the clusters of normal NIH 3T3 cells. We found that these abundance of proline residues between amino acids 480 and cytoskeletal alterations are also evident in actively spreading 655, where there are at least 8 potential SH3-binding motifs cells that have been freshly plated or in the process of (Liu and Altman, 1998; Miyake et al., 1998; Smit and Borst, cytokinesis. These cells normally display an excess of 1997; Thien and Langdon, 1998), supports the notion that Sli- lamellipodia and membrane ruffles but the formation of these 1-like activity may only be one of multiple functions for c-Cbl structures is attenuated when 480-Cbl is expressed. Since in signal transduction. lamellipodia formation is also regulated by Rac signalling our By expressing C-terminally truncated c-Cbl constructs and findings suggest that expression of truncated c-Cbl constructs assaying for their involvement in modulation of cellular may affect the actin cytoskeleton by selective modulation of properties we have discovered a novel regulatory function for signalling by this member of the Rho GTPase family. c-Cbl. Unlike the truncated v-Cbl protein whose expression Although it is not clear how 480-Cbl expression may result results in transformation of NIH 3T3 fibroblasts, we find that in attenuation of Rho GTPases, mutational inactivation of the a non-oncogenic form of Cbl, which retains the Ring finger but SH2* domain by the G306E substitution reverses the has lost SH3 and SH2 binding sequences, can significantly morphological and cytoskeletal effects of 480-Cbl. Clearly, the alter the shape of cells, the growth pattern of cells on culture G306E mutation would cancel this inhibition, resulting in full dishes and the organization of the actin cytoskeleton. Since it restoration of the activity of endogenous c-Cbl. Inhibition of is the induction of lamellipodia protrusions during the process endogenous c-Cbl by 480-Cbl, due to competition for SH2* of cell spreading that gives fibroblasts their distinctive shape domain ligands, is further supported by our finding that we hypothesized that 480-Cbl may alter cell shape and growth co-overexpression of c-Cbl reverses the effects of 480-Cbl patterns by directly affecting the formation of these actin-based on morphology and the cytoskeleton. Interestingly, our 224 R. M. Scaife and W. Y. Langdon

Fig. 10. Interaction and co- localization of Crk with c-Cbl. Clarified lysates from NIH 3T3 cells stably expressing HA-c-Cbl and HA- 480-Cbl were immunoprecipitated with either anti-Crk (A) or anti-PI 3- kinase p85 subunit antibodies (B), subjected to SDS-PAGE and probed by western blotting using either anti- Cbl, anti-Crk, anti-HA or anti PI 3- kinase p85 subunit antibodies as specified. Arrowheads indicate the gel positions of the corresponding antigens. (C) HA-c-Cbl expressing NIH 3T3 cells were transiently transfected with Crk. The cells were fixed, stained with anti-Crk antibodies, biotinylated secondary antibodies, and Alexa 488-conjugated Streptavidin (1 and 3), followed either by TRITC phalloidin (2) or affinity purified R2 anti-c-Cbl antibodies and TRITC-conjugated anti-rabbit secondary antibodies (4) and viewed as single 1.0 µm confocal immunofluorescence microscopy Z- series sections. Arrowheads indicate the position of actin lamellae. Bar, 30 µm. demonstration of c-Cbl localization to membrane ruffle- This SH2*-binding protein is represented as a tyrosine kinase associated actin lamellae indicates that normal regulation of the in Fig. 11 since all c-Cbl SH2* domain-binding proteins actin cytoskeleton requires targeting of c-Cbl to these actin identified to date are tyrosine kinases. Further, as demonstrated structures. This is inferred by the perturbed formation of these structures in 480-Cbl expressing cells and the inability of 480- Cbl to associate with actin lamellae. It appears therefore that in addition to attenuation of signalling from tyrosine kinase linked receptors to MAP kinases, c-Cbl is also implicated in signalling pathways affecting the actin cytoskeleton. We therefore propose that F-actin associated c-Cbl functions as a protein scaffold, permitting efficient signal transduction from ‘upstream’ signalling proteins such as Crk and PI 3-kinase to ‘downstream’ signalling components such as Rho GTP-ases (see Fig. 11). We also propose that the recruitment to actin filaments of an unidentified SH2* domain-binding protein is crucial for the normal function of this protein complex to form lamellipodia.

Fig. 11. A model for the role of c-Cbl in the regulation of the cytoskeleton and cell morphology. Activation of tyrosine kinase linked membrane receptors (e.g. RTKs, integrins) results in tyrosine phosphorylation of c-Cbl, permitting direct SH2 domain mediated association with the Rac activators Crk and PI 3-kinase. Rac is primarily localized to membrane ruffle-associated actin lamellae (Neudauer et al., 1998) and it is proposed that the targeting of c-Cbl to actin by an SH3 domain protein regulates the activity of Rac through an SH2*-bound tyrosine kinase (dashed arrow). Competition for the SH2*-bound kinase by 480-Cbl (broad arrow) would prevent its localization to actin lamellae, thus altering the activity of Rac, resulting in attenuated lamellipodia formation and perturbed cell morphology. Regulation of cell shape and the cytoskeleton by Cbl 225 for the Crk binding protein DOCK180, there is precedent for phosphorylation of c-Cbl facilitates adhesion and spreading while the activation of Rac by the tyrosine phosphorylation of suppressing anchorage-independent growth of v-Abl transformed NIH 3T3 ‘upstream’ activators (Kiyokawa et al., 1998b). fibroblasts. Oncogene 18, 3703-3715. Galisteo, M. L., Dikic, I., Batzer, A. G., Langdon, W. Y. and Schlessinger, It appears that regulation of lamellipodia and membrane J. (1995). Tyrosine phosphorylation of the c-cbl proto-oncogene product and ruffle formation by c-Cbl is not restricted to fibroblasts since association with epidermal growth factor (EGF) receptor upon EGF similar effects on the actin cytoskeleton have recently been stimulation. J. Biol. Chem. 270, 20241-20245. reported in osteoclasts following microinjection of truncated Hall, A. (1998). Rho GTPases and the cytoskeleton. Science 279, 509-514. c-Cbl fusion proteins (Ejiri et al., 1998). It will be of interest Jongeward, G. D., Clandinin, T. R. and Sternberg, P. W. (1995). sli-1, a negative regulator of let-23-mediated signaling in C. elegans. Genetics 139, to determine whether the role of c-Cbl in bone resorption 1553-1566. (Tanaka et al., 1996) may in fact reflect its involvement in the Kiyokawa, E., Hashimoto, Y., Kobayashi, S., Sugimura, H., Kurata, T. and regulation of the morphology and actin cytoskeleton of Matsuda, M. (1998a). Activation of Rac1 by a Crk SH3-binding protein, osteoclasts. Additionally, c-Cbl appears to be involved in DOCK180. Genes Dev. 12, 3331-3336. Kiyokawa, E., Hashimoto, Y., Kurata, T., Sugimura, H. and Matsuda, M. modulation of actin stress-fiber assembly in adipocytes and (1998b). Evidence that DOCK180 up-regulates signals from the CrkII- Abl-transformed fibroblasts (Feshchenko et al., 1999; Ribon p130(Cas) complex. J. Biol. Chem. 273, 24479-24484. et al., 1998), suggesting a prominent involvement of c-Cbl in Klemke, R. L., Leng, J., Molander, R., Brooks, P. C., Vuori, K. and the regulation of a variety of cytoskeleton dependent cellular Cheresh, D. A. (1998). Cas/Crk coupling serves as a ‘molecular switch’ for properties. induction of cell migration. J. Cell. Biol. 140, 961-972. Lamarche, N., Tapon, N., Stowers, L., Burbelo, P., Aspenstrom, P., Bridges, T., Chant, J. and Hall, A. (1996). Rac and Cdc42 induce actin We thank Dr Kristiina Vuori, Dr Christine Thien and Ms Tamaris polymerization and G1 cell cycle progression independently of p65PAK and Morshead for many helpful suggestions and assistance, Professor the JNK/SAPK MAP kinase cascade. Cell 87, 519-529. Alan Hall for kindly providing the Rac cDNA constructs, Dr Lauffenberger, D. and Horwitz, A. F. (1996). Cell migration: a physically Michiyaki Matsuda for providing the Crk constructs, and Dr Lyuba integrated molecular process. Cell 84, 359-369. Varticovski for anti-p85 serum. We also thank Dr Paul Rigby for Lee, P., Wang, Y., Dominguez, M., Yeung, Y.-G., Murphy, M., Bowtell, D. assistance with confocal microscopy. This work was funded by grants and Stanley, R. (1999). The Cbl protooncogene stimulates CSF-1 receptor from the Medical Research Fund of Western Australia and from the multiubiquitination and endocytosis, and attenuates macrophage National Health and Medical Research Council (Canberra). proliferation. EMBO J. 18, 3616-3628. Levkowitz, G., Waterman, H., Zamir, E., Kam, Z., Oved, S., Langdon, W. Y., Beguinot, L., Geiger, B. and Yarden, Y. (1998). c-Cbl/Sli-1 regulates REFERENCES endocytic sorting and ubiquitination of the epidermal growth factor receptor. Genes Dev. 12, 3663-3674. Andoniou, C. E., Thien, C. B. F. and Langdon, W. Y. (1994). Tumour Liu, Y.-C. and Altman, A. (1998). Cbl: complex formation and functional induction by activated abl involves tyrosine phosphorylation of the product implications. Cell. Signalling 10, 377-385. of the cbl oncogene. EMBO J. 13, 4515-4523. Lupher, M. L., Jr, Reedquist, K. A., Miyake, S., Langdon, W. Y. and Band, Andoniou, C. E., Thien, C. B. F. and Langdon, W. Y. (1996). The two major H. (1996). A novel PTB domain in the N-terminal transforming region of sites of cbl tyrosine phosphorylation in abl-transformed cells select the crkL Cbl interacts directly and selectively wth ZAP-70 in T cells. J. Biol. Chem. SH2 domain. Oncogene 12, 1981-1989. 271, 24063-24068. Ben-Ze’ev, A. (1997). Cytoskeletal and adhesion proteins as tumor Lupher, M. L., Songyang, Z., Shoelson, S. E., Cantley, L. C. and Band, H. suppressors. Curr. Opin. Cell Biol. 9, 99-108. (1997). The Cbl phosphotyrosine-binding domain selects a D(N/D)XpY Blake, T. J., Shapiro, M., Morse, H. C., III and Langdon, W. Y. (1991). motif and binds to the TyrP292 negative regulatory phosphorylation site of The sequences of the human and mouse c-cbl proto-oncogenes show v-cbl ZAP-70. J. Biol. Chem. 272, 33140-33144. was generated by a large truncation encompassing a proline-rich domain and Lupher, M. L., Rao, N., Lill, N., Andoniou, C. E., Miyake, S., Clark, E., a leucine zipper-like motif. Oncogene 6, 653-657. Druker, B. and Band, H. (1998). Cbl-mediated negative regulation of the Blake, T. J., Heath, K. G. and Langdon, W. Y. (1993). The truncation that Syk tyrosine kinase. J. Biol. Chem. 273, 35273-35281. generated the v-cbl oncogene reveals an ability for nuclear transport, DNA Meng, F. and Lowell, C. A. (1998). A β1 integrin signaling pathway involving binding and acute transformation. EMBO J. 12, 2017-2026. Src-family kinases, Cbl and PI 3-kinase is required for macrophage Bonita, D. P., Miyake, S., Lupher Jr, M. L., Langdon, W. Y. and Band, H. spreading and migration. EMBO J. 17, 4391-4403. (1997). Phosphotyrosine binding domain-dependent upregulation of the Meng, W., Sawasdikosol, S., Burakoff, S. and Eck, M. (1999). Structure of platelet-derived growth factor receptor a signaling cascade by transforming the amino-terminal domain of Cbl complexed to its binding site on ZAP-70 mutants of Cbl: implications for Cbl’s function and oncogenicity. Mol. Cell. kinase. Nature 398, 84-90. Biol. 17, 4597-4610. Mitchison, T. J. and Cramer, L. P. (1996). Actin-based cell motility and cell Boussiotis, V. A., Freeman, G. J., Berezovskaya, A., Barber, D. and Nadler, locomotion. Cell 84, 371-379. L. M. (1997). Maintenance of human T-cell anergy: blocking of IL-2 gene Miyake, S., Luper, M. L., Andoniou, C., Lill, N. L., Ota, S., Douillard, P., transcription by activated rap1. Science 278, 124-128. Rao, N. and Band, H. (1998). The cbl proto-oncogene product: from Bowtell, D. D. L. and Langdon, W. Y. (1995). The protein product of the c- enigmatic oncogene to central stage of signal transduction. Crit. Rev. cbl oncogene rapidly complexes with the EGF receptor and is tyrosine Oncogen. 8, 189-218. phosphorylated following EGF stimulation. Oncogene 11, 1561-1567. Miyake, S., Mullane-Robinson, K., Lill, N., Douillard, P. and Band, Dolfi, F., Garcia-Guzman, M., Ojaniemi, M., Nakamura, H., Matsuda, M. H. (1999). Cbl-mediated negative regulation of platelet-derived growth and Vuori, K. (1998). The adaptor protein Crk connects multiple cellular factor receptor-dependent cell proliferation. J. Biol. Chem. 274, 16619- stimuli to the JNK signaling pathway. Proc. Nat. Acad. Sci. USA 95, 15394- 16628. 15399. Murphy, M., Schnall, R., Venter, D., Bertoncello, I., Thien, C. B. F., Donovan, J. A., Ota, Y., Langdon, W. Y. and Samelson, L. E. (1996). Langdon, W. Y. and Bowtell, D. L. (1998). Tissue hyperplasia and Regulation of the association of p120cbl with Grb2 in Jurkat T cells. J. Biol. enhanced T cell signalling via ZAP-70 in c-Cbl deficient mice. Mol. Cell. Chem. 271, 26369-26374. Biol. 18, 4872-4882. Ejiri, S., Sahni, M., Bartiewcz, M., Neff, L., Sabatakos, G., Levy, J., Naramura, M., Kole, H., Hu, R.-J. and Gu, H. (1998). Altered thymic Ozawa, H. and Baron, R. (1998). ASBMR-IBMS, a1068. positive selection and intracellular signals in Cbl-deficient mice. Proc. Nat. Feshchenko, E., Langdon, W. Y. and Tsygankov, A. (1998). Fyn, Yes, and Acad. Sci. USA 95, 15547-15552. Syk phosphorylation sites in c-Cbl map to the same tyrosine residues that Neudauer, C. L., Joberty, G., Tatsis, N. and Macara, I. G. (1998). Distinct become phosphorylated in activated T cells. J. Biol. Chem. 273, 8323- cellular effects and interactions of the Rho-family GTPase TC10. Curr. Biol. 8331. 8, 1151-1160. Feshchenko, E., Shore, S. and Tsygankov, A. (1999). Tyrosine Nobes, C. D., Hawkins, P., Stephens, L. and Hall, A. (1995). Activation of 226 R. M. Scaife and W. Y. Langdon

the small GTP-binding proteins rho and rac by growth factor receptors. J. and Baron, R. (1996). c-Cbl is downstream of c-Src in a signalling pathway Cell Sci. 108, 225-233. necessary for bone resorption. Nature 383, 528-531. Ojaniemi, M., Martin, S. S., Dolfi, F., Olefsky, J. M. and Vuori, K. (1997). Thien, C. B. F. and Langdon, W. Y. (1997). EGF receptor binding and The proto-oncogene product p120cbl links c-Src and phosphatidylinositol 3′- transformation by v-cbl is ablated by the introduction of a loss-of-function kinase to the integrin signaling pathway. J. Biol. Chem. 272, 3780-3787. mutation from the Caenorhabditis elegans sli-1 gene. Oncogene 14, 2239- Ota, Y. and Samelson, L. E. (1997). The product of the proto-oncogene c- 2249. Cbl Ð a negative regulator of the Syk tyrosine kinase. Science 276, 418- Thien, C. B. F. and Langdon, W. Y. (1998). c-Cbl: a regulator of T cell 420. receptor-mediated signalling. Immunol. Cell Biol. 76, 473-482. Ribon, V., Herrera, R., Kay, B. K. and Saltiel, A. R. (1998). A role for CAP, Thien, C. B. F., Bowtell, D. and Langdon, W. Y. (1999). Perturbed regulation a novel, multifunctional src homology 3 domain-containing protein in of ZAP-70 and sustained tyrosine phosphorylatoin of LAT and SLP-76 in formation of actin stress fibers and focal adhesions. J. Biol. Chem. 273, c-Cbl-deficient thymocytes. J. Immunol. 162, 7133-7139. 4073-4080. van Leeuwen, J., Paik, P. and Samelson, L. (1999). Activation of nuclear Ridley, A. and Hall, A. (1992). The small GTP-binding protein rho regulates factor of activated T cells-(NFAT) and activating protein 1 (AP-1) by the assembly of focal adhesions and actin stress fibers in response to growth oncogenic 70Z Cbl requires an intact phosphotyrosine binding domain but factors. Cell 70, 389-399. not Crk(L) or p85 phosphatidylinositol 3-kinase association. J. Biol. Chem. Rodriguez-Viciana, P., Warne, P. H., Kwaja, A., Marte, B. M., Pappin, D., 274, 5153-5162. Das, P., Waterfield, M., Ridley, A. and Downward, J. (1997). Role of Yoon, C. H., Lee, J., Jongeward, G. D. and Sternberg, P. W. (1995). phosphoinositide 3-OH kinase in cell transformation and control of the actin Similarity of sli-1, a regulator of vulval development in C. elegans, to the cytoskeleton by ras. Cell 89, 457-467. mammalian proto-oncogene c-cbl. Science 269, 1102-1105. Smit, L. and Borst, J. (1997). The Cbl family of signal transduction Zheng, Z., Elly, C., Altman, A. and Liu, Y.-C. (1998). Dual regulation of T molecules. Crit. Rev. Oncogen. 8, 359-379. cell receptor-mediated signaling by oncogenic Cbl mutant 70Z. J. Biol. Tanaka, S., Amling, M., Neff, L., Peyman, A., Uhlmann, E., Levy, J. B. Chem. 274, 4883-4889.