Skip Navigation


Human Molecular Genetics Advance Access originally published online on July 1, 2003
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
12/16/1981    most recent
ddg209v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (17)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Hou, P.
Right arrow Articles by Gorski, J. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hou, P.
Right arrow Articles by Gorski, J. L.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Human Molecular Genetics, 2003, Vol. 12, No. 16 1981-1993
DOI: 10.1093/hmg/ddg209
© 2003 Oxford University Press

Fgd1, the Cdc42 GEF responsible for Faciogenital Dysplasia, directly interacts with cortactin and mAbp1 to modulate cell shape

Peng Hou1, Lourdes Estrada1, Andrew W. Kinley4, J. Thomas Parsons4, Anne B. Vojtek3 and Jerome L. Gorski1,2,*

1Department of Pediatrics and Communicable Diseases, 2Department of Human Genetics, University of Michigan School of Medicine, Ann Arbor, MI 48109, USA, 3Department of Biological Chemistry, University of Michigan School of Medicine, Ann Arbor, MI 48109, USA and 4Department of Microbiology and Cancer Center, University of Virginia Health Science Center, Charlottesville, VA 22908, USA

Received April 2, 2003; Revised June 2, 2003; Accepted June 16, 2003


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
FGD1 mutations result in Faciogenital Dysplasia (FGDY), an X-linked human disease that affects skeletal formation and embryonic morphogenesis. FGD1 and Fgd1, the mouse FGD1 ortholog, encode guanine nucleotide exchange factors (GEF) that specifically activate Cdc42, a Rho GTPase that controls the organization of the actin cytoskeleton. To further understand FGD1/Fgd1 signaling and begin to elucidate the molecular pathophysiology of FGDY, we demonstrate that Fgd1 directly interacts with cortactin and mouse actin-binding protein 1 (mAbp1), actin-binding proteins that regulate actin polymerization through the Arp2/3 complex. In yeast two-hybrid studies, cortactin and mAbp1 Src homology 3 (SH3) domains interact with a single Fgd1 SH3-binding domain (SH3-BD), and biochemical studies show that the Fgd1 SH3-BD directly binds to cortactin and mAbp1 in vitro. Immunoprecipitation studies show that Fgd1 interacts with cortactin and mAbp1 in vivo and that Fgd1 SH3-BD mutations disrupt binding. Immunocytochemical studies show that Fgd1 colocalizes with cortactin and mAbp1 in lamellipodia and membrane ruffles, and that Fgd1 subcellular targeting is dynamic. By using truncated cortactin proteins, immunocytochemical studies show that the cortactin SH3 domain targets Fgd1 to the subcortical actin cytoskeleton, and that abnormal Fgd1 localization results in actin cytoskeletal abnormalities and significant changes in cell shape and viability. Thus, this study provides novel in vitro and in vivo evidence that Fgd1 specifically and directly interacts with cortactin and mAbp1, and that these interactions play an important role in regulating the actin cytoskeleton and, subsequently, cell shape.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
FGD1 is implicated as a critical participant in mammalian development because mutations in this gene result in the disease Faciogenital Dysplasia (FGDY; Aarskog syndrome), a skeletal dysplasia and multiple congenital anomaly syndrome (1,2). Most FGD1 mutations are predicted null alleles; thus the X-linked recessive phenotype appears to be due to the absence of the gene product in affected males (2). FGD1 encodes a guanine nucleotide exchange factor (GEF) that specifically activates the Rho (Ras homology) GTPase Cdc42 (3,4). Cdc42 plays a critical role in regulating actin polymerization, vesicular transport, the cell cycle and cell polarity (5). By activating Cdc42, FGD1 potently stimulates reorganization of the actin cytoskeleton and the formation of filopodia, long cross-linked parallel actin filaments (3). Through Cdc42, FGD1 also stimulates the passage of cells through the G1 phase of the cell cycle (6), promotes cell transformation (7) and activates c-Jun N-terminal kinase (3,4). During embryogenesis, the mouse FGD1 ortholog, Fgd1 (960 residues, >95% identical to human FGD1), exhibits a limited pattern of expression that is largely restricted to regions of incipient and active ossification including skeletal elements involved in the FGDY phenotype (8). Together, these data indicate that Fgd1/Cdc42 signaling plays an important role in mammalian embryonic development and skeletal formation (2).

Rho guanine nucleotide exchange factor (RhoGEF) proteins play critical roles in regulating Rho GTPase activity (5). RhoGEFs activate Rho GTPases by catalyzing the exchange of bound GDP for GTP, thereby inducing a conformational change in the GTP-bound GTPase that allows it to interact with downstream effector proteins (9). Data indicate that RhoGEF activation is a multi-step process that proceeds sequentially from RhoGEF recruitment to a specific subcellular region to RhoGEF activation (10). In addition to a GEF (or Dbl homology, DH) domain, most RhoGEF proteins contain a variety of different signaling domains (9). Mouse (and human) Fgd1 is composed of several different signaling domains including a proline-rich N-terminal region, adjacent DH and pleckstrin homology (PH) domains, a FYVE domain, and a second C-terminal PH (PH2) domain (1,11). Thus, like other RhoGEFs, Fgd1 contains several putative signaling domains (2). Biochemical and immunocytochemical studies localize endogenous and overexpressed Fgd1 protein to the cell cortex, cytoplasm, and Golgi complex (12). Data also indicate that, as part of their targeted activation, RhoGEFs interact with different proteins to form signaling complexes (13). Thus, these data suggest that Fgd1 signaling domains might be involved in targeting Fgd1 to different subcellular regions to regulate Cdc42 activation.

To further understand Fgd1 signaling and begin to elucidate the molecular pathophysiology of Faciogenital Dysplasia, we have identified two distinct but related Fgd1-interacting proteins, cortactin and mouse actin-binding protein (mAbp1). Yeast two-hybrid analyses and binding studies show that a single Fgd1 SH3-binding domain (SH3-BD) directly interacts with cortactin and mAbp1 SH3 domains. Immunocytochemical studies show that Fgd1 colocalizes with both cortactin and mAbp1 in the subcortical actin cytoskeleton. Studies also indicate that the cortactin SH3 domain targets Fgd1 to the cell cortex, and that abnormal Fgd1 targeting induces changes in cell shape and viability. These data strongly suggest that Fgd1 plays a key role in regulating the actin cytoskeleton: by binding to cortactin and mAbp1, Fgd1 is targeted to the dynamic cortical actin cytoskeleton; by activating Cdc42, Fgd1 indirectly activates N-WASP to stimulate Arp2/3-directed actin polymerization and filopodia formation.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Fgd1 interacts with cortactin and mAbp1 in yeasttwo-hybrid studies
Yeast two-hybrid analyses were performed to identify Fgd1-interacting proteins. For this purpose, L-F1, a LexA-Fgd1 fusion bait plasmid encoding full-length Fgd1, was used to screen a mouse osteoblast two-hybrid cDNA library. Studies showed that 22 independently isolated cDNA clones displayed L-F1-dependent yeast reporter gene transactivation. Sequence analyses showed that these cDNA clones fell into two distinct classes: (i) 12 clones encoded mouse cortactin (14), and (ii) 10 clones encoded mouse actin-binding protein (mAbp1; SH3P7) (15,16). Analyses showed that cortactin and mAbp1 shared similar structural domains (Fig. 1A and B); both proteins contained (in order) N-terminal actin-binding domains, {alpha}-helical regions, proline–serine–threonine (PST)-rich regions, and C-terminal Src homology 3 (SH3) domains (16,17). Thus, these data showed that, in the yeast two-hybrid system, Fgd1 interacts with two different but related actin-binding proteins.



View larger version (23K):
[in this window]
[in a new window]
  		Figure 1. Fgd1 interacts with cortactin and mAbp1 in the yeast two-hybrid system. Schematic diagram illustrates composition of isolated cortactin and mAbp1 cDNAs. The predicted structural organizations of (A) cortactin (17) and the (B) mAbp1 (16) cDNAs are illustrated in the top two panels. Cortactin (546 residues) and mAbp1 (433 residues) both encode N-terminal actin-binding domains: (A) cortactin contains an Arp2/3 complex-binding domain (Arp2/3-BD; gray box) and a series of actin-binding domain (actin-BD) repeats; (B) mAbp1 contains an N-terminal actin-binding ADF/cofilin domain. Cortactin and mAbp1 both contain {alpha} helical domains (hatched boxes), proline-serine-threonine (PST)-rich regions (white boxes) and C-terminal SH3 domains (black boxes). Numbers denote the residues present at the beginning of each of the domains. Bars below the top illustrations indicate the portions of the cortactin (A) and mAbp1 (B) transcripts contained in isolated yeast two-hybrid cDNA clones. Amino acid residues (aa) encoded in each of the cDNAs are indicated to the right of each bar; the number of independently isolated clones is indicated to the far right in parentheses. (C) Amino acid sequence alignment of cortactin and mAbp1 SH3 domains. Identical residues are shaded; similar residues are shown in bold.

 
Since all isolated cDNAs encoded a SH3 domain (Fig. 1A and B), data suggested that cortactin and mAbp1 interacted with a Fgd1 SH3 binding domain (SH3-BD). Sequence analyses showed that the cortactin and mAbp1 SH3 domains were highly related with 82% similarity and 67% identity (Fig. 1C); cortactin and mAbp1 also bind to identical SH3-BD ligands (18,19). As shown in Figure 2A and B, L-F4, a bait plasmid containing part of the Fgd1 N-terminal proline-rich domain (PRD) interacted with cortactin and mAbp1 SH3 domains in the yeast two-hybrid system to transactivate yeast HIS3 and LacZ reporter genes. By contrast, Fgd1 bait plasmids containing the DH and PH domains (L-F2) and the FYVE and PH2 domains (L-F3) failed to interact. Synthetic peptides corresponding to individual Fgd1 N-terminal polyproline sequences were used to generate bait plasmids to define the specific Fgd1 SH3-BD region. Inspection of the Fgd1 PRD showed that it contained at least two putative SH3-BDs: codons 158–168 encoded a PKPQVPPKP sequence that was similar to a predicted cortactin SH3-BD consensus sequence (Fig. 2C); residues 178–187 encoded a EPIPPPPSRP sequence that was similar to the predicted c-Abl SH3-BD consensus sequence (1). L-F6, a bait plasmid containing the first putative Fgd1 SH3-BD, interacted with both cortactin and mAbp1 SH3 domains (Fig. 2A and B), but a bait plasmid containing the second putative SH3-BD (L-F5) failed to interact. In addition, bait plasmids containing mutations predicted to disrupt the first SH3-BD (L-F7 and L-F8; Fig. 2A and B) failed to interact. By contrast, bait plasmids containing missense mutations flanking the critical SH3-BD residues interacted with the cortactin and mAbp1 SH3 domains (data not shown). These data suggested that a single Fgd1 SH3-BD interacted with the cortactin and mAbp1 SH3 domains.



View larger version (21K):
[in this window]
[in a new window]
  		Figure 2. The Fgd1 SH3-BD interacts with cortactin and mAbp1 SH3 domains in the yeast two-hybrid system. (A) A schematic diagram showing bait LexA-Fgd1 plasmid content. The top diagram illustrates the domains that comprise the Fgd1 protein including the N-terminal proline-rich domain (PRD; white bar), the Dbl-homology (DH) and pleckstrin homology (PH) domains (DH/PH; black bar), and the FYVE and C-terminal PH (PH2) domains (FYVE/PH2; gray bars). Numbers denote the residue present at the beginning of a domain. Figures below the top bar show the composition of LexA-Fgd1 fusion constructs. Plasmid L-F1 (codons 1–960) expresses full-length Fgd1 protein; L-F2 encodes the DH and PH domains (codons 335–707); L-F3 encodes the FYVE and PH2 domains (codons 714–922); and L-F4 encodes part of the PRD (codons 131–195). L-F5 and L-F6 encode PRD residues 170–190 and 147–169, respectively. L-F7 encodes residues 147–169 with a K164E mutation; L-F8 encodes residues 147–169 with P159A and P162A mutations; asterisks indicate mutated residues. (B) Analyses of yeast two-hybrid interactions between Fgd1, cortactin and mAbp1. Yeast L40 cells were transformed with various pairs of plasmids. Transactivation of the HIS3 reporter genes was assayed by growth on His- plates: a plus sign (+) indicates growth on His- plates; a minus sign (-) indicates no growth. ß-Galactosidase activity was determined by X-Gal staining (see Materials and Methods). The strength of interaction was indicated by the timing and intensity of blue color: a plus sign (+) indicates blue visible in 1 h; a minus sign (-) indicates a white colony overnight. Control activation domain plasmids (pAD) include plasmids VP16 (20) and pACT2 (Clontech). (C) Alignment of proline-rich sequences contained within cortactin SH3-binding proteins. Proteins demonstrated to bind to the cortactin SH3 domain and their putative or confirmed SH3-BD(s) are aligned with a consensus binding sequence (bottom) obtained by peptide phage display library screening (21). The Fgd1 SH3-BD sequence is identical to that contained in human FGD1 (1,11). Aligned sequences include CortBP-1 (22), Shank-3 (23) and Dynamin (24).

 
The Fgd1 SH3-BD directly binds to cortactin and mAbp1 SH3 domains in vitro
Binding studies were performed to determine if cortactin and mAbp1 SH3 domains bound directly to the Fgd1 SH3-BD (Fig. 3A). A radiolabeled GST-fusion protein containing the mAbp1 SH3 domain (GST-AbpSH3) selectively bound to Fgd1 GST-fusion proteins containing either long (GST-F4) or short (GST-F6) fragments spanning the Fgd1 SH3-BD. By contrast, GST-AbpSH3 failed to bind to GST-F5, a GST-fusion protein containing a second putative Fgd1 SH3-BD, and bound to GST-F8, a GST-fusion protein containing a mutant Fgd1 SH3-BD, with much less affinity. A labeled GST-fusion protein containing the cortactin SH3 domain (GST-CortSH3) yielded a similar binding pattern (Fig. 3A). Labeled mAbp1 and cortactin SH3 GST-fusion proteins failed to bind to control GST protein, and labeled GST protein failed to bind to any of the proteins (Fig. 3A); thus, these analyses indicated that the observed interactions were specific.



View larger version (35K):
[in this window]
[in a new window]
  		Figure 3. Binding assays show direct interactions between the cortactin and mAbp1 SH3 domains and the Fgd1 SH3-BD. (A) Radiolabeled GST-fusion proteins containing either the mAbp1 SH3 domain (GST-AbpSH3; residues 318–433) or the cortactin SH3 domain (GST-CortSH3; residues 490–546) were incubated with identically loaded membranes containing purified GST protein alone, and GST-Fgd1 fusion proteins including GST-F4 (residues 131–195), GST-F5 (residues 170–190), GST-F6 (residues 147–169), and GST-F8 (residues 147–169 with P159A and P162A mutations). Labeled control GST protein was incubated to an identically loaded membrane. (B) A membrane containing SH3 domains derived from mAbp1 (GST-AbpSH3), cortactin (GST-CortSH3) and Grb2 (GST-Grb2) was incubated with a labeled Fgd1 SH3-BD GST-fusion protein, GST-F6. (C) Membranes containing GST-SH3 fusion proteins derived from mAbp1 (GST-AbpSH3) and cortactin (GST-CortSH3) were incubated with radiolabeled Fgd1 SH3-BD GST-fusion protein GST-F6 (3 nM) and competed with varying concentrations (nM) of GST-fusion proteins containing either native Fgd1 SH3-BD (GST-F6) or mutated Fgd1 SH3-BD (GST-F8). Coomassie-stained gel shows that equivalent amounts of GST-fusion protein (1–2 µg/lane) were loaded in each lane of each membrane (A–C); molecular mass markers (in kDa) are indicated to the right of the Coomassie stained gels (A and C).

 
The specificity of these interactions was further analyzed by incubating a labeled GST-fusion protein containing Fgd1 SH3-BD (GST-F6) with membranes containing mAbp1, cortactin and Grb2 SH3 GST-fusion proteins. Labeled GST-F6 protein selectively bound to both mAbp1 and cortactin SH3 GST-fusion proteins but failed to bind to either of the two Grb2 SH3 domains (Fig. 3B). Competition-binding experiments were also performed. In these studies, GST-fusion proteins containing either a wild-type (GST-F6) or mutated (GST-F8) Fgd1 SH3-BD were assayed for the ability to competitively interfere with the binding of labeled GST-F6 to mAbp1 and cortactin SH3 GST-fusion proteins. As shown in Figure 3C, the binding of labeled Fgd1 SH3-BD GST-fusion protein was reduced in a dose-dependent manner with increasing concentrations of non-labeled GST-F6 in the nanomolar range (30–300 nM). In contrast, 300 nM of mutant GST-F8 protein failed to block Fgd1-mAbp1 and Fgd1-cortactin binding. These data indicated that in vitro binding between Fgd1 SH3-BD and cortactin and mAbp1 SH3 domains was direct and specific.

Fgd1 colocalizes with cortactin and mAbp1 in the membrane ruffles of stimulated cells
Since our data showed that Fgd1 physically interacted with cortactin and mAbp1, GFP-tagged Fgd1, and Flag-tagged cortactin and mAbp1 proteins were transiently expressed in COS-7 cells to determine if protein distributions were consistent with direct interactions. We previously showed that, in these cells, transiently overexpressed Fgd1 proteins demonstrated a subcellular distribution indistinguishable from that observed in cells that endogenously express Fgd1 (12). When expressed in quiescent serum-deprived cells, mAbp1 and cortactin proteins were primarily distributed in the cytoplasm; relatively minor amounts of the Flag-tagged proteins were localized at the cell cortex (Fig. 4A). In contrast, upon PDGF stimulation, Flag-mAbp1 and Flag-cortactin were rapidly redistributed to lamellipodia, findings consistent with previous studies (16,25). GFP-tagged Fgd1 displayed a dynamic subcellular distribution similar to that observed with cortactin and mAbp1 (Fig. 4A). In quiescent cells, GFP-Fgd1 was primarily distributed in the cytoplasm and Golgi complex with little protein present at the cell cortex. By contrast, in stimulated cells, a significant amount of GFP-Fgd1 was redistributed to the cell cortex. Cotransfection studies were performed to determine if Fgd1 colocalized with cortactin, and mAbp1 (Fig. 4B). When coexpressed with Flag-mAbp1 in stimulated cells, in addition to being present in the cytoplasm and Golgi complex, like Flag-mAbp1, GFP-Fgd1 was targeted to the leading edge of cells in lamellipodia. Similarly, when GFP-Fgd1 was coexpressed with Flag-cortactin, these proteins were primarily localized in cortical ruffles at the leading edge of motile cells (Fig. 4B). These data showed that the subcellular distribution of Fgd1 was dynamic and that Fgd1 colocalized with both cortactin and mAbp1 in the cell cortex.



View larger version (52K):
[in this window]
[in a new window]
  		Figure 4. The Fgd1 protein colocalizes with mAbp1 and cortactin in the membrane ruffles of stimulated cells. (A) Immunofluorescent localization of Fgd1, mAbp1 and cortactin in quiescent and growth factor-stimulated cells. COS-7 cells were transfected with either Flag-tagged mAbp1 (Flag-Abp; codons 1–433), Flag-tagged cortactin (Flag-Cort; codons 1–546), or GFP-tagged Fgd1 (GFP-F1; codons 1–960); labels indicate whether the cells were quiescent (PDGFß-) or stimulated with PDGF (PDGFß+). Upon stimulation with PDGF, there was a recruitment of mAbp1, cortactin and Fgd1 to large cortical ruffles (arrowheads). (B) Immunofluorescent colocalization of Fgd1 with mAbp1 or cortactin in the membrane ruffles of stimulated cells. COS-7 cells were cotransfected with GFP-F1 and either Flag-Abp or Flag-Cort, and stimulated with PDGFß (see Materials and Methods). Two different cells coexpressing GFP-Fgd1 and Flag-Cort are shown at relatively low (middle panels) and high (bottom panels) magnification. Merged image pairs demonstrate regions of overlap (yellow) between Fgd1 and mAbp1 (top panel), and Fgd1 and cortactin (bottom two panels). Scale bar=10 µm.

 
Fgd1 specifically interacts with overexpressed cortactin and mAbp1 in vivo
Previously, we showed that endogenous Fgd1 is present in low concentrations and prone to rapid proteolytic digestion (12). Thus, to determine whether Fgd1 interacted with mAbp1 and cortactin in vivo, epitope-tagged proteins were expressed in COS-7 cells, and Flag-tagged cortactin or Flag-tagged mAbp1 was immunoprecipitated with anti-Flag antibody. As shown in Figure 5, when Flag-Abp and GFP-tagged Fgd1 (GFP-F1) were coexpressed in cells, anti-Flag antibody coimmunoprecipitated GFP-F1 with Flag-Abp. Control studies showed that immunoprecipitation was dependent upon the expression of both Flag-Abp and GFP-F1. The specificity of this interaction was analyzed by expressing Flag-Abp with a full-length GFP-tagged Fgd1 protein containing P159A and P162A mutations in the SH3-BD (GFP-{Delta}F1); yeast two-hybrid analyses and in vitro binding studies showed that these mutations interfered with cortactin and mAbp1 binding (see above). Despite equivalent expression of GFP-F1 and GFP-{Delta}F1 proteins, when coexpressed with Flag-Abp, anti-Flag antibody failed to coimmunopreciptate GFP-{Delta}F1 protein with Flag-Abp (Fig. 5). Similar results were observed with cortactin; when Flag-Cort and GFP-F1 were coexpressed in cells, anti-Flag antibody coimmunoprecipitated GFP-F1 with Flag-Cort (Fig. 5). In contrast, when coexpressed with Flag-Cort, anti-Flag antibody failed to coimmunopreciptate GFP-{Delta}F1 protein. These data showed that Fgd1 selectively and specifically interacted with mAbp1 and cortactin in vivo.



View larger version (26K):
[in this window]
[in a new window]
  		Figure 5. Fgd1 immunoprecipitates with cortactin and mAbp1. Epitope-tagged proteins were expressed in COS-7 cells, and Flag-tagged mAbp1 (Flag-Abp) and Flag-tagged cortactin (Flag-Cort) proteins were immunoprecipitated with anti-Flag antibodies. GFP-tagged Fgd1 proteins included GFP-F1 (residues 1–960) and GFP-{Delta}F1 (residues 1–960 with mutations P159A and P162A). Epitope-tagged proteins were transiently expressed at low levels in COS-7 cells either alone, or in combination, as indicated. Equal amounts of cell homogenates were subjected to immunoprecipitation (IP) with anti-Flag antibodies, and fractionated proteins were detected by western blot (WB) analysis. The upper two blots (IP: anti-Flag) show the immunoprecipitated proteins detected by using either anti-GFP or anti-Flag antibody; the epitope-tagged proteins detected by the western blot analysis are indicated on the right side of the figure. The lower two blots (Input) show the results of western blot analyses of cell extracts before immunoprecipitation.

 
Truncated cortactin redirects Fgd1 targeting
Our data showed that the distribution of Fgd1 was dynamic, and that Fgd1 colocalized with cortactin at the cell cortex; thus, experiments were performed to examine whether Fgd1-cortactin interactions affected Fgd1 localization. For this purpose, we examined whether the expression of truncated cortactin proteins lacking localization domains altered Fgd1 targeting (Fig. 6); the results of these experiments are summarized in Table 1. Two different truncated myc-tagged cortactin proteins were used in these studies: myc-CterCort (residues 350–546) and myc-CortSH3 (residues 491–546); both tagged cortactin proteins lack the N-terminal actin-binding regions and display a cytoplasmic and perinuclear distribution (17). When coexpressed in cells, in addition to being distributed in the cytoplasm and perinuclear region, GFP-tagged Fgd1 (GFP-F1) and Flag-tagged full-length cortactin (Flag-Cort) colocalized with phalloidin-stained actin filaments at the cell cortex in large lamellipodia in motile cells (Fig. 6A). In contrast, a markedly different protein distribution was observed in cells coexpressing GFP-F1 and the truncated cortactin protein, myc-CterCort (Fig. 6B). Compared to GFP-F1/Flag-Cort-expressing cells, cells coexpressing GFP-F1 and myc-CterCort were significantly less motile. In these cells, GFP-F1 and myc-CterCort proteins were generally absent from the cell cortex; instead, these proteins colocalized in the cytoplasm and perinuclear region with actin aggregates. To confirm these studies, additional experiments were performed with cells coexpressing GFP-F1 and a second truncated cortactin protein, myc-CortSH3 (Fig. 6C). Again, like the GFP-F1/myc-CterCort-expressing cells, cells coexpressing GFP-F1 and myc-CortSH3 displayed significantly reduced motility. In these cells, myc-CortSH3 and GFP-F1 colocalized with the actin aggregates in the perinuclear region but were notably absent from the cell cortex. These cells also displayed a strikingly unusual phenotype with an unusual actin cytoskeleton and large perinuclear F-actin aggregates. Since GFP-F1 colocalized with these unusual actin structures, these data suggested that misdirected Fgd1 played a role in their formation. To address this question, we studied cells coexpressing full-length Flag-Cort and GFP-F{Delta}PRD, a GFP-tagged Fgd1 protein lacking the N-terminal PRD; GFP-F{Delta}PRD is not distributed to the cell cortex (12). When coexpressed in cells, GFP-F{Delta}PRD and Flag-Cort appeared to be significantly toxic, and expression was limited to a small number of non-motile cells that expressed low levels of GFP-F{Delta}PRD (Fig. 6D). In these cells, Flag-Cort was present at the cell cortex; in contrast, GFP-F{Delta}PRD colocalized with F-actin aggregates in the perinuclear region and was notably absent from the cell cortex (Fig. 6D). Together, these data demonstrated that Fgd1 interacted with truncated cortactin proteins in vivo, and that, in the absence of actin-binding domains, these truncated proteins targeted Fgd1 to the cytoplasm and perinuclear region, away from the subcortical actin cytoskeleton. These data also suggested that misdirected Fgd1 resulted in an altered actin cytoskeleton.



View larger version (122K):
[in this window]
[in a new window]
  		Figure 6. Truncated cortactin proteins redirect Fgd1 targeting. (A–D) Immunofluorescent localization of epitope-tagged Fgd1 and cortactin proteins and filamentous actin in COS-7 cells; actin filaments were stained with Alexa Fluor 633-conjugated phalloidin. Photomicrographs show cells coexpressing: (A) GFP-tagged Fgd1 (GFP-F1; residues 1–960) and Flag-tagged cortactin (Flag-Cort; residues 1–546); (B) GFP-F1 and a myc-tagged C-terminal portion of cortactin (myc-CterCort; residues 350–546); (C) GFP-F1 and a myc-tagged cortactin SH3 domain (myc-CortSH3; residues 491–546); and (D) Flag-Cort and a GFP-tagged Fgd1 construct lacking the N-terminal PRD (GFP-F{Delta}PRD; residues 359–922). Epitope-tagged Fgd1 (GFP) and cortactin (Flag/myc) protein distributions are shown in the left two columns in green and red, respectively; phalloidin-stained actin filaments (actin) are shown in purple. These images are superimposed and shown to the right (merge): an overlap in the distribution of GFP and Flag/myc is shown in yellow; an overlap in the distribution of GFP, Flag/myc, and phalloidin-stained molecules is shown in white. Large arrowheads (A) indicate cell ruffles, small arrowheads (B–D) indicate perinuclear actin aggregates, and arrows (C) indicate filopodia. Scale bar=10 µm.

 

View this table:
[in this window]
[in a new window]
  		Table 1. Effects of Fgd1 and cortactin expression in cultured cells
 
Altered Fgd1 targeting affects the actin cytoskeleton
Additional studies were performed to further characterize the consequences of aberrant Fgd1 targeting (Fig. 7). For this purpose, GFP-F{Delta}PRD was expressed to block targeting to the cell cortex and cells were examined shortly after transfection to minimize the effects of cell toxicity; the results of these analyses are summarized in Table 1. COS-7 cells expressing control GFP-F1 protein displayed a normal polarized motile phenotype and a significant proportion of GFP-F1 was present at the cell cortex at the leading edge (Fig. 7A). Similar results were obtained with several different mammalian cells including mouse osteoblast MC3T3-E1 cells and mouse fibroblast Swiss 3T3 and NIH 3T3 cells (data not shown). In contrast, cells expressing GFP-F{Delta}PRD displayed significantly reduced motility and an unusual morphology characterized by an aberrant actin cytoskeleton (Fig. 7B–D). In expressing cells, GFP-F{Delta}PRD was distributed in the cytoplasm and perinuclear region; regardless of the level of expression, GFP-F{Delta}PRD was notably absent from the cell cortex. Phalloidin staining showed that, like GFP-F1/CortSH3-expressing cells (Fig. 6C), GFP-F{Delta}PRD-expressing cells displayed an unusual actin cytoskeleton and contained large perinuclear actin aggregates. Cells expressing relatively small amounts of GFP-F{Delta}PRD typically contained unusual thick cords of actin in the cytoplasm and at the cell cortex (Fig. 7B). Cells expressing relatively more GFP-F{Delta}PRD displayed large cytoplasmic polyhedral rings of F-actin (Fig. 7C and D). These rings appeared to delimit GFP-F{Delta}PRD distribution; GFP-F{Delta}PRD was primarily distributed within the rings and excluded from the cell cortex (Fig. 7C–E). Similar abnormalities were observed in NIH 3T3 and Swiss 3T3 fibroblasts (data not shown). These data showed that aberrant Fgd1 targeting resulted in a complex cellular phenotype that was characterized by decreased cell motility, reduced cell viability, and a markedly altered actin cytoskeleton with perinuclear actin aggregates, unusual actin cords, and cytoplasmic polyhedral actin rings. Together, these data indicated that anomalous Fgd1 targeting results in altered actin cytoskeletal organization and, consequently, an abnormal cell phenotype.



View larger version (95K):
[in this window]
[in a new window]
  		Figure 7. Failure to target Fgd1 to the cell cortex induces significant changes in actin cytoskeletal organization. (A–D) Immunofluorescent localization of GFP-tagged Fgd1 proteins and filamentous actin in quiescent COS-7 cells; actin filaments were stained with TRITC-conjugated phalloidin. Photomicrographs show cells expressing either (A) a moderate amount of full-length GFP-F1, or (B–D) varying amounts of GFP-tagged Fgd1 construct lacking the N-terminal PRD (GFP-F{Delta}PRD) including (B) low, (C) moderate or (D) high levels of GFP-F{Delta}PRD expression. The distribution of Fgd1 protein (GFP) and phalloidin-stained actin filaments (actin) are shown in the left two columns in green and red, respectively. These images are superimposed and shown to the right (merge). Large arrowheads indicate lamellipodia, small arrowheads indicate filopodia, large arrows indicate thick actin cords and small arrows indicate perinuclear polyhedral F-actin rings. Scale bar=10 µm.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
In this study, we demonstrate that Fgd1, a Cdc42-specific GEF involved in mammalian embryonic development, directly interacts with cortactin and mAbp1, actin-binding proteins that regulate the actin polymerization through the Arp2/3 complex. These data show that Fgd1 binds to the SH3 domains of cortactin and mAbp1 by a single N-terminal SH3-BD to target Fgd1 to the subcortical actin cytoskeleton. Several lines of evidence demonstrate that Fgd1 directly interacts with cortactin and mAbp1 in vitro and in vivo. Fgd1 interacts with cortactin and mAbp1 in yeast two-hybrid studies and mutations in the Fgd1 SH3-BD specifically disrupt these interactions. Biochemical analyses show that the Fgd1 SH3-BD binds directly to the cortactin and mAbp1 SH3 domains in vitro. Immunocytochemical studies show that Fgd1 colocalizes with both cortactin and mAbp1 in lamellipodia and membrane ruffles. Although protein instability interfered with our ability to perform endogenous coimmunoprecipitation studies, coimmunoprecipitation analyses performed with expressed proteins indicate that Fgd1 interacts with cortactin and mAbp1 in vivo and that Fgd1 SH3-BD mutations block this binding. Furthermore, immunocytochemical studies with truncated epitope-tagged proteins show that Fgd1 interacts with the cortactin SH3 domain in living cells and that this interaction redirects Fgd1 subcellular targeting. These studies also show that altered Fgd1 localization results in an anomalous actin cytoskeleton and abnormal cell morphology. Thus, this study provides novel in vitro and in vivo evidence that Fgd1 specifically and directly interacts with cortactin and mAbp1, and that Fgd1 appears to act with these proteins to regulate the actin cytoskeleton, and subsequently, cell shape.

Signal transduction relies on specific protein–protein interactions to provide accurate protein recruitment and precise subcellular targeting. Our data indicate that, like cortactin and mAbp1, the subcellular distribution of Fgd1 is dynamic; these data show that Fgd1 is redistributed between the cytoplasm and the cell cortex in response to growth factor stimulation. The observation that constitutively and exogenously expressing cells display the same subcellular distribution of Fgd1 strongly suggests that this localization is accurate and biologically relevant (12). These data also show that cortactin interacts with Fgd1 to precisely target it to particular regions of the subcortical actin cytoskeleton. Although it is formally possible that Fgd1 plays a role in cortactin (and mAbp1) subcellular targeting, the observation that truncated cortactin SH3 protein redirects Fgd1 away from the cell cortex indicates that cortactin plays a dominant role in Fgd1 targeting. Thus, these data strongly suggest that cortactin binds to Fgd1 in the cytoplasm and subsequently brings Fgd1 to the cell cortex. This hypothesis is consistent with the observation that cortactin and mAbp1 contain actin-binding domains that directly link to actin filaments (16,17), motifs not found in Fgd1. Data show that RhoGEF activation is a multi-step process that is typically initiated by protein translocation (10). Cdc42, the Fgd1 substrate, has been localized to several subcellular regions including the plasma membrane (26). Consequently, by being targeted to the cell cortex, Fgd1 is poised to activate Cdc42 signaling at the plasma membrane. Thus, these data strongly suggest that cortactin and mAbp1 play important roles in targeting Fgd1 to the subcortical actin cytoskeleton for activation.

Since these data indicate that mAbp1 and cortactin directly interact with Fgd1, it is reasonable to hypothesize that these proteins might be involved in related cellular processes. Abp1p, the Saccharomyces cerevisiae mAbp1 homolog, is targeted to cortical actin patches, structures that are concentrated at growing yeast cell surfaces (27). In yeast, Abp1p overexpression results in a depolarized distribution of cortical actin and loss of spatially regulated cell growth (27); these data suggest that Abp1p is critical in regulating the actin cytoskeleton and cell polarity. Data also implicates Abp1p as playing a role in endocytosis; for example, Abp1p has a redundant function with Sla2p in endocytic trafficking (28), and Abp1p interacts with proteins that function in endocytosis (29). In mammals, mAbp1 interacts directly with dynamin (19), a GTPase that regulates the formation of endocytic vesicles. Cortactin binds to filamentous actin and is enriched in cell–matrix contact sites and the lamellipodia of cultured cells (17). Since cortactin also binds to large scaffolding proteins, it is implicated in playing a role in assembling large protein complexes to regulate actin cytoskeletal organization (30). Cortactin is also a proto-oncogene and its amplified expression results in increased cell invasiveness (31). Cortactin is overexpressed in a variety of cancers including ovarian, bladder, breast and lung carcinomas and head and neck tumors (32). As cortactin overexpression is associated with a poor prognosis for many of these cancers (33), these data suggest that elevated cortactin expression directly contributes to tumor severity. Together, accumulated data indicates that mAbp1 and cortactin play critical roles in regulating cell polarity, endocytosis and actin cytoskeletal organization. Thus, by inference, these data suggest that Fgd1 might also play an important role in these or related processes. Since these processes are, in part, regulated by the GTPase Cdc42 (5), this contention is consistent with Fgd1 being a Cdc42-specific GEF (3,4,12).

Actin polymerization is a dynamic process in which actin nucleation, polymerization and actin filament stabilization are tightly regulated. The determination that Fgd1 binds directly to cortactin and mAbp1 strongly suggests that, through these interactions, Fgd1 directly or indirectly regulates the actin cytoskeleton. As a Cdc42-specific GEF targeted to the cell cortex, Fgd1 activates Cdc42 to stimulate N-WASP and Arp2/3 complex activation. We previously showed that Fgd1 stimulates Cdc42-dependent filopodia formation (3,12), a process driven by N-WASP activation of the Arp2/3 complex (34). Thus, these data showed that, through Cdc42, Fgd1 regulates N-WASP stimulation and Arp2/3 activation. Our data also indicate that Fgd1 directly binds to cortactin and mAbp1 at the subcortical actin cytoskeleton. Cortactin and yeast Abp1 both enhance the ability of the Arp2/3 complex to form branched actin filaments (35–37). Data also show that cortactin-binding proteins are capable of indirectly stimulating the Arp2/3 complex by binding cortactin in vitro (38). Thus, it is reasonable to hypothesize that Fgd1 might interact with cortactin to regulate Arp2/3 complex activation. It has also been shown that cortactin and N-WASP act synergistically to stimulate the Arp2/3 complex (37). Together, these data strongly suggest that, as a cortactin-binding Cdc42GEF, Fgd1 is poised to modulate Arp2/3 activity by regulating and/or interacting with both N-WASP and cortactin. Conversely, it is possible that cortactin and/or mAbp1 bind to Fgd1 to regulate GEF activity. Additional experiments will be necessary to fully characterize the consequences of these interactions.

Our data also indicate that abnormal Fgd1 targeting results in an aberrant actin cytoskeleton and altered cell shape. Two different approaches were used to misdirect Fgd1 targeting: Fgd1 was drawn away from the cell cortex by binding it to truncated cortactin proteins; and truncated Fgd1 was misdirected from the actin cytoskeleton by the removal of its N-terminal domain. Each experiment yielded similar results: by either means, abnormal Fgd1 targeting resulted in a variety of actin cytoskeletal abnormalities including perinuclear actin aggregates, unusual actin cables and perinuclear actin rings. The observation that either full-length or truncated Fgd1 protein induced similar abnormal cellular phenotypes suggests that actin cytoskeleton aberrations are the result of abnormal Fgd1 targeting, not altered Fgd1 activity. Previous studies have clearly demonstrated that overexpressed and/or misdirected cortactin, mAbp1, and WASP-like proteins all cause defects in actin organization and unusual cell phenotypes (16,27,39). Although it is unclear as to whether this is a general characteristic of all Arp2/3-interacting proteins, these results suggest that the overexpression of these proteins causes mislocalization of the Arp2/3 complex and, consequently, an abnormal actin cytoskeleton (40). Collectively, these results strongly suggest that Fgd1 is an important component in regulating the actin cytoskeleton and cell shape.

Based on these results, a clearer image of the role of Fgd1/FGD1 in mammalian skeletal formation is beginning to emerge. Expressed in osteoblasts and localized to the subcortical actin cytoskeleton, this Cdc42 activator is poised to influence osteoblast cytoskeletal organization, cell polarity, and vesicular transport (12), cellular processes critical to normal cellular function and differentiation. In S. cerevisiae, Schizosaccharomyces pombe, Drosophila melanogaster, and Caenorhabditis elegans, the loss of RhoGEF activity is functionally equivalent to the loss of the target Rho protein (41); thus, it is likely that a loss of FGD1 activity results in disrupted osteoblast Cdc42 signaling. Several lines of evidence indicate that Cdc42 signaling and actin cytoskeletal regulation play critical roles in the molecular biology of skeletogenesis. Data show that, like epithelial cells, osteoblasts display apical-basolateral polarity and secrete extracellular matrix (ECM) molecules in a polarized fashion (42). Transgenic mice (43–45) and in vitro biochemical studies (46,47) show that, in part, the ECM regulates osteoblast growth and function. Rho GTPases play a critical role in mediating signals from the ECM to modify cellular proliferation, cytoskeletal organization and differentiation (5). In particular, Cdc42 signaling plays a critical role in regulating integrin-dependent cellular adhesion and cytoskeletal organization (48). In embryonic chick osteoblasts, an intact actin cytoskeleton is required for the regulated expression of ECM molecules in response to mechanical strain (49). In addition, studies show that the actin cytoskeleton and the Rho signaling cascade are involved in the detection and response to fluid shear induced mechanical signaling in osteoblast-like cells (50). Studies have also demonstrated that the actin cytoskeleton is critical for determining and maintaining bone cell shape (51). Although we cannot rule out the formal possibility the FGD1 may be involved in additional signaling pathways, taken together, these observations suggest that FGD1/Cdc42 signaling plays a role in regulating the osteoblast actin cytoskeleton and that Faciogenital Dysplasia is a disorder of the FGD1/Cdc42 signaling.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
Cell culture and transfections
Mouse osteoblast cell line MC3T3-E1 (52) was provided by Dr Franceschi (University of Michigan) and COS-7 cells were purchased (ATCC, Manassas, VA). Calvarial cells were isolated as described (8). Cells were cultured as described (8,12). For transfections, cells were plated into six-well trays with coverslips. Cells were transfected with 2–5 µg DNA per well using Lipofectamine-PLUS as recommended by the supplier (Life Technologies, Grand Island, NY). Cells were analyzed 12–36 h after transfection. To study the effects of growth factor stimulation, transfected cells were first allowed to recover in Dulbecco's modified Eagle Medium (DMEM) containing 10% fetal calf serum for 3 h, washed four times with DMEM, serum-deprived for 16 h in DMEM containing 0.1% BSA (Sigma Chemical, St Louis, MO), and stimulated with PDGFß 10 ng/ml (Sigma Chemical) in DMEM for 15 min.

cDNA libraries, cDNA clones, and DNA constructs
RNA was isolated from mouse calvarial and MC3T3-E1 cells as described (1). Osteoblast-derived poly(A)+mRNA and random and oligo-dT primers were used to construct a pGAD10 cDNA library (Clontech, Palo Alto, CA) for yeast two-hybrid analysis. PCR primers were used to amplify a full-length mAbp1 cDNA clone; mouse Fgd1 and cortactin cDNA clones were previously described (11,14).

For yeast two-hybrid studies, LexA-Fgd1 fusion protein expression constructs encoding Fgd1 were generated by PCR amplification; these included L-F1 (codons 1–960), L-F2 (codons 335–707), L-F3 (codons 714–922) and L-F4 (codons 131–195). PCR amplification was used to generate GAL4 activation domain (pAD) fusion proteins: full-length cortactin (pAD-Cort) and mAbp1 (pAD-Abp), and C-terminal fusion proteins pAD-AbpSH3 (codons 318–433) and pAD-CortSH3 (codons 490–546). Synthetic oligonucleotides were used to generate LexA-Fgd1 expression constructs encoding short PRD Fgd1 peptides: L-F5 (codons 170–190), L-F6 (codons 147–169), L-F7 (codons 147–169 with a K164E mutation) and L-F8 (codons 147–169 with P159A and P162A mutations); these same synthetic oligonucleotides were cloned into a modified pGEX-2TK (Amersham Pharmacia Biotech, Piscataway, NJ) vector to generate glutathione S-transferase (GST)-Fgd1 fusion protein constructs GST-F4, GST-F5, GST-F6 and GST-F8, respectively. PCR amplification was used to generate GST-protein fusion constructs for cortactin (GST-CortSH3; codons 490–546), mAbp1 (GST-AbpSH3; codons 318–433), and Grb2 (GST-Grb2; codons 2–217). PCR amplification was used to generate green fluorescent protein (GFP)-tagged Fgd1 expression constructs GFP-F1 (codons 1–960) and GFP-F{Delta}PRD (codons 359–922) as described (12). Site-directed mutagenesis was used to generate GFP-{Delta}F1, a GFP-tagged Fgd1 expression construct containing a mutated SH3-BD (codons 1–960 with mutations P159A and P162A). PCR amplification and pCMV-tag2 (Stratagene) vector were used to generate a Flag-tagged mAbp1 construct Flag-Abp (codons 1–433); Flag- and myc-tagged cortactin expression constructs Flag-Cort (codons 1–546), myc-CterCort (codons 350–546), and myc-CortSH3 (codons 491–546), were described (17). All constructs were verified by DNA sequencing. Fusion protein expression was confirmed by western blot analysis.

Yeast two-hybrid analyses
Yeast two-hybrid analyses were performed as described (20). S. cerevisiae reporter strain L40 cells containing plasmid L-F1 were transformed with the osteoblast-derived pGAD10 cDNA library. An estimated 10x106 transformants were grown in synthetic media to select for His+ colonies; His+ colonies were assayed for ß-galactosidase activity by filter assay as described (20). Of 201 independent His+ colonies, 194 were also LacZ+; 48 of these colonies were randomly selected to determine the proportion of LexA-Fgd1-dependent His+ LacZ+ colonies. Mating assays were performed to identify library clones that transactivated the HIS3 and LacZ reporter genes in a bait-specific manner (20); 24 of 48 colonies strongly transactivated reporter genes with LexA-Fgd1 plasmid. Library plasmids were isolated by transforming HB101 Escherichia coli (Life Technologies) with total yeast DNA (20); 22 independent transactivating library plasmids were recovered.

Protein-binding analyses
GST-fusion protein constructs were transformed into E. coli strain BL21 bacteria (Amersham Pharmacia Biotech); fusion protein expression was induced with 0.1–0.4 mM isopropyl ß-D-thiogalactoside (Amersham) for 2–3 h. GST-fusion proteins were purified from cell lysates on glutathione-agarose beads as described (53). Protein expression was verified by immunoblot analysis; protein concentrations were determined by the Cu+/bicinchoninic acid method (Micro-BCA Protein Assay, Pierce) with BSA serving as the standard. GST-fusion proteins were fractionated by SDS–PAGE and transferred to Immobilon-P membranes (Millipore, MA) membranes as described (8). To perform in vitro protein-binding analyses, GST-fusion proteins (100 µg protein/reaction) were radiolabeled using protein kinase A (54). To detect protein–protein interactions, blot overlays were performed essentially as described (55). For competition experiments, in addition to the radiolabeled protein, membranes were also incubated with different concentrations of unlabeled GST-fusion protein (see Results).

Antibodies, immunoblotting and immunoprecipitation
For immunoprecipitation analyses, transfected COS-7 cells were lysed with protease inhibitor cocktail as described (12). Cell homogenates were centrifuged at 10 000g for 30 min and clarified lysates (200 µg) were incubated with 50 µl of FLAG M2 affinity resin (Sigma Chemical) overnight at 4°C. Immune complexes were collected by centrifugation and washed with lysis buffer. Collected complexes were fractionated by SDS-PAGE, transferred to membranes and blotted with anti-Flag and anti-GFP antibodies as described (8). Anti-cortactin antibodies, anti-N-terminal and 4F11, were described (25). Monoclonal anti-myc antibody (9E10) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Goat polyclonal anti-GST antibody was purchased from Amersham Pharmacia Biotech; mouse monoclonal anti-GFP antibody was purchased from Zymed Laboratories (San Francisco, CA); mouse monoclonal anti-Flag antibody (M2) was purchased from Sigma Chemical. Western blots were incubated with secondary HRP-conjugated antibody and detected by using chemiluminescent reagents (ECL, Amersham).

Immunofluorescence and microscopy
Cells were plated on glass coverslips, fixed and permeabilized as described (12). After blocking, cells were incubated with primary antibodies diluted in PBS for 1 h and secondary antibody for 45 min. To detect actin filaments, cells were stained with either 0.1 µg/ml TRITC-conjugated phalloidin (Sigma Chemical) as described (12) or Alexa Fluor 633-conjugated phalloidin per supplier recommendations (Molecular Probes Inc., Oregon, WA). Cells transfected with GFP fusion constructs were fixed 12 h after transfection, rinsed and mounted by inverting coverslips onto 5 µl of Fluoromount G (Electron Microscopy Sciences, Ft Washington, PA). Cells were examined with a LSM 510 laser scanning Zeiss confocal microscope and LSM 510 (version 4.0) digital imaging software (Zeiss, Inc., Thornwood, NY).


   ACKNOWLEDGEMENTS
 
We are grateful to Dr Ben Margolis for the GST-Grb2 expression construct, and Dr Bruce Donohoe and Dr Chris Edwards for technical advice. This work was funded, in part, by the NIH training grant 5T32HD07505-02 (L.E.), American Cancer Society Research Scholar Grant RSG-01-177-01-MGO (A.B.V.), NIH grant GM 38542 (J.A.C.), and by grants from the March of Dimes-Birth Defects Foundation (6-FY99-425) and the National Institutes of Health grant HD34446 (J.L.G.).


   FOOTNOTES
 
* To whom correspondence should be addressed at: Division of Pediatric Genetics, Room 3570 Medical Science Research Building II, Box 0688, University of Michigan Medical School, Ann Arbor, MI 48109-0688, USA. Tel: +1 7346472908; Fax: +1 7347639512; Email: jlgorski{at}med.umich.edu Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  1. Pasteris, N.G., Cadle, A.B., Logie, L.J., Porteous, M.E.M., Schwartz, C.E., Stevenson, R.E., Glover, T.W., Wilroy, R.S. and Gorski, J.L. (1994) Isolation and characterization of the faciogenital dysplasia (Aarskog–Scott syndrome) gene: a putative rho/rac guanine nucleotide exchange factor. Cell, 79, 669–678.[CrossRef][ISI][Medline]

  2. Gorski, J.L. (2001) Aarskog–Scott Syndrome. In Scriver, C.R., Beaudet, A.L., Sly, W.S. and Valle, D. (eds), The Metabolic and Molecular Basis of Inherited Disease, 8th edition. McGraw-Hill, New York, pp. 6153–6165.

  3. Olson, M.F., Pasteris, N.G., Gorski, J.L. and Hall, A. (1996) Faciogenital dysplasia protein (FGD1) and Vav, two related proteins required for normal embryonic development, are upstream regulators of Rho GTPases. Curr. Biol., 6, 1628–1633.[CrossRef][ISI][Medline]

  4. Zheng, Y., Fischer, D.J., Tigyi, G., Pasteris, N.G., Gorski, J.L. and Xu, Y. (1996) The faciogenital dysplasia gene product FGD1 functions as a Cdc42Hs-specific guanine-nucleotide exchange factor. J. Biol. Chem., 271, 33169–33172.[Abstract/Free Full Text]

  5. Van Aelst, L. and D'Souza-Schorey, C. (1997) Rho GTPases and signaling networks. Genes Dev., 11, 2295–2322.[Free Full Text]

  6. Nagata, K., Lamarche, N., Gorski, J.L. and Hall, A. (1998) Activation of G1 progression, JNK MAP kinase and actin filament assembly by the exchange factor FGD1. J. Biol. Chem., 273, 15453–15457.[Abstract/Free Full Text]

  7. Whitehead, I.P., Abe, K., Gorski, J.L. and Der, C.J. (1998) Cdc42 and FGD1 cause distinct signaling and transforming activities. Mol. Cell. Biol., 18, 4689–4697.[Abstract/Free Full Text]

  8. Gorski, J.L., Estrada, L., Hu, C. and Liu, Z. (2000) Skeletal-specific expression of Fgd1 during bone formation and skeletal defects in Faciogenital Dysplasia (FGDY; Aarskog syndrome). Dev. Dynam., 218, 573–586.[CrossRef][ISI][Medline]

  9. Cerione, R.A. and Zheng, Y. (1996) The Dbl family of oncogenes. Curr. Opin. Cell. Biol., 8, 216–222.[CrossRef][ISI][Medline]

  10. Gulli, M.-P. and Peter, M. (2001) Temporal and spatial regulation of Rho-type guanine nucleotide exchange factors: the yeast perspective. Genes Dev., 15, 365–379.[Free Full Text]

  11. Pasteris, N.G., de Gouyon, B., Cadle, A.B., Campbell, K., Herman, G.E. and Gorski, J.L. (1995) Cloning and regional localization of the mouse faciogenital dysplasia (Fgd1) gene. Mamm. Genome, 6, 658–661.[CrossRef][ISI][Medline]

  12. Estrada, L., Caron, E. and Gorski, J.L. (2001) Fgd1, the Cdc42GEF responsible for Faciogenital Dysplasia, is localized to the subcortical actin cytoskeleton and Golgi membrane. Hum. Mol. Genet., 10, 485–495.[Abstract/Free Full Text]

  13. Hoffman, G.R. and Cerione, R.A. (2002) Signaling to the Rho GTPases: networking with the DH domain. FEBS Letters, 513, 85–91.[CrossRef][ISI][Medline]

  14. Wu, H., Reynolds, A.B., Kanner, S.B., Vines, R.R. and Parsons, J.T. (1991) Identification and characterization of a novel cytoskeleton-associated pp60src substrate. Mol. Cell. Biol., 11, 5113–5124.[Abstract/Free Full Text]

  15. Larbolette, O., Wollscheid, B., Schweikert, J., Nielsen, P.J. and Wienands, J. (1999) SH3P7 is a cytoskeleton adapter protein and is coupled to signal transduction from lymphocyte antigen receptors. Mol. Cell. Biol., 19, 1539–1546.[Abstract/Free Full Text]

  16. Kessels, M.M., Engqvist-Goldstein, A.E.Y. and Drubin, D.G. (2000) Association of mouse actin-binding protein 1 (mAbp1/SH3P7), a Src kinase target, with dynamic regions of the cortical actin cytoskeleton in response to Rac1 activation. Mol. Biol. Cell., 11, 393–412.[Abstract/Free Full Text]

  17. Weed, S.A., Karginov, A.V., Schafer, D.A., Weaver, A.M., Kinley, A.W., Cooper, J.A. and Parsons, J.T. (2000) Cortactin localization to sites of actin assembly in lamellipodia requires interactions with F-actin and the Arp2/3 complex. J. Cell. Biol., 151, 29–40.[Abstract/Free Full Text]

  18. McNiven, M.A., Kim, L., Krueger, E.W., Orth, J.D., Cao, H. and Wong, T.W. (2000) Regulated interactions between dynamin and the actin-binding protein cortactin modulate cell shape. J. Cell. Biol., 151, 187–198.[Abstract/Free Full Text]

  19. Kessels, M.M., Engqvist-Goldstein, A.E.Y., Drubin, D.G. and Qualmann, B. (2001) Mammalian Abp1, a signal-responsive F-actin-binding protein, links the actin cytoskeleton to endocytosis via the GTPase dynamin. J. Cell. Biol., 153, 351–366.[Abstract/Free Full Text]

  20. Vojtek, A.B., Cooper, J.A. and Hollenberg, S.M. (1997) Searching for interacting proteins with the two-hybrid system II. In Bartel, P.L.and Fields, S. (eds), The Yeast Two-Hybrid System. Oxford University Press, Oxford, UK, pp. 29–42.

  21. Sparks, A.B., Rider, J.E., Hoffman, N.G., Fowlkes, D.M., Quilliam, L.A. and Kay, B.K. (1996) Distinct ligand preferences of Src homology 3 domains from Src, Yes, Abl, cortactin, p53bp2, PLC{gamma}, Crk, and Grb2. Proc. Natl Acad. Sci. USA, 93, 1540–1544.[Abstract/Free Full Text]

  22. Du, Y., Weed, S.A., Xiong, W.-C., Marshall, T.D. and Parsons, J.T. (1998) Identification of a novel cortactin SH3 domain-binding protein and its localization to growth cones of cultured neurons. Mol. Cell. Biol., 18, 5838–5851.[Abstract/Free Full Text]

  23. Naisbitt, S., Kim, E., Tu, J.C., Xiao, B., Sala, C., Valtschanoff, J., Weinberg, R.J., Worley, P.F. and Sheng, M. (1999) Shank, a novel family of postsynaptic density proteins that binds to the NMDA receptor/PSD-95/GKAP complex and cortactin. Neuron, 23, 569–582.[CrossRef][ISI][Medline]

  24. Cook, T.A., Urrutia, R. and McViven, M.A. (1994) Identification of dynamin 2, an isoform ubiquitously expressed in rat tissues. Proc. Natl Acad. Sci. USA, 91, 644–648.[Abstract/Free Full Text]

  25. Weed, S.A., Du, Y. and Parsons, J.T. (1998) Translocation of cortactin to the cell periphery is mediated by the small GTPase Rac1. J. Cell. Sci., 111, 2433–2444.[Abstract]

  26. Backlund, P.S., Jr (1993) Carboxyl methylation of the low molecular weight GTP-binding protein G25K. Biochem. Biophys. Res. Comm., 196, 534–542.

  27. Drubin, D.G., Miller, K.G. and Botstein, D. (1988) Yeast actin-binding proteins: evidence for a role in morphogenesis. J. Cell. Biol., 107, 2551–2561.[Abstract/Free Full Text]

  28. Holtzman, D.A., Yang, S. and Drubin, D.G. (1993) Synthetic-lethal interactions identify two novel genes, SLA1 and SLA2, that control membrane cytoskeleton assembly in Saccharomyces cerevisiae. J. Cell. Biol., 122, 635–644.[Abstract/Free Full Text]

  29. Lila, T. and Drubin, D.G. (1997) Evidence for physical and functional interactions among two Saccharomyces cerevisiae SH3 domain proteins, an adenylyl cyclase-associated protein and the actin cytoskeleton. Mol. Biol. Cell., 8, 367–385.[Abstract]

  30. Sheng, M. and Kim, E. (2000) The Shank family of scaffold proteins. J. Cell. Sci., 113, 1851–1856.[Abstract]

  31. Huang, C., Liu, J., Hauderschild, C.C. and Zhan, X. (1998) The role of tyrosine phosphorylation of cortactin in the locomotion of endothelial cells. J. Biol. Chem., 273, 25770–25776.[Abstract/Free Full Text]

  32. Schuuring, E. (1995) The involvement of the chromosome 11q13 region in human malignancies: cyclin D1 and EMS1 are two new candidate oncogenes—a review. Gene, 159, 83–96.[CrossRef][ISI][Medline]

  33. Hall, M. and Peters, G. (1996) Genetic alterations of cyclins, cyclin-dependent kinases, and Cdk inhibitors in human cancer. Adv. Cancer Res., 68, 67–108.[ISI][Medline]

  34. Miki, H.T., Sasaki, T., Taki, Y. and Takenawa, T. (1998) Induction of filopodium formation by a WASP-related actin-depolymerizing protein N-WASP. Nature, 391, 93–96.[CrossRef][Medline]

  35. Goode, B.L., Rodal, A.A., Barnes, G. and Drubin, D.G. (2001) Activation of the Arp2/3 complex by the actin filament binding protein Abp1p. J. Cell. Biol., 153, 627–634.[Abstract/Free Full Text]

  36. Uruno, T., Liu, J., Zhang, P., Fan, Y., Egile, C., Li, R., Mueller, S.C. and Zhan, X. (2001) Activation of Arp2/3 complex-mediated actin polymerization by cortactin. Nat. Cell Biol., 3, 259–266.[CrossRef][ISI][Medline]

  37. Weaver, A.M., Karginov, A.V., Kinley, A.W., Weed, S.A., Li, Y., Parsons, J.T. and Cooper, J.A. (2001) Cortactin promotes and stabilizes Arp2/3-induced actin filament network formation. Curr. Biol., 11, 370–374.[CrossRef][ISI][Medline]

  38. Schafer, D.A., Weed, S.A., Binna, D., Karginov, A.V., Parsons, J.T. and Cooper, J.A. (2002) Dynamin2 and cortactin regulate actin assembly and filament organization. Curr. Biol., 12, 1852–1857.[CrossRef][ISI][Medline]

  39. Machesky, L.M. and Insall, R.H. (1998) Scar1 and the related Wiskott–Aldrich syndrome protein, WASP, regulate the actin cytoskeleton through Arp2/3 complex. Curr. Biol., 8, 1347–1356.[CrossRef][ISI][Medline]

  40. Olazabal, I.M. and Machesky, L.M. (2001) Abp1p and cortactin, new ‘hand-holds’ for actin. J. Cell. Biol., 154, 679–682.[Abstract/Free Full Text]

  41. Boguski, M.S. and McCormick, F. (1993) Proteins regulating Ras and its relatives. Nature, 366, 643–653.[CrossRef][Medline]

  42. Ilvesaro, J., Metsikko, K., Vaananen, K. and Tuukkanen, J. (1999) Polarity of osteoblasts and osteoblast-like UMR-108 cells. J. Bone Min. Res., 14, 1338–1344.[CrossRef][ISI][Medline]

  43. Ducy, P., Desbois, C., Boyce, B., Pinero, G., Story, B., Dunstan, C., Smith, E., Bonadio, J., Goldstein, S., Gundberg, C., Bradley, A. and Karsenty, G. (1996) Increased bone formation in osteocalcin-deficient mice. Nature, 382, 448–452.[CrossRef][Medline]

  44. Xu, T., Bianco, P., Fisher, L.W., Longenecker, G., Smith, E., Goldstein, S., Bonadio, J., Boskey, A., Heegaard, A.M., Sommer, B. et al. (1998) Targeted disruption of the biglycan gene leads to an osteoporosis-like phenotype in mice. Nat. Genet., 20, 78–82.[CrossRef][ISI][Medline]

  45. Delany, A.M., Amling, M., Priemel, M., Howe, C., Baron, R. and Canalis, E. (2000) Osteopenia and decreased bone formation in osteonectin-deficient mice. J. Clin. Invest., 105, 915–923.[ISI][Medline]

  46. Xiao, G., Wang, D., Benson, M.D., Karsenty, G. and Franceschi, R.T. (1998) Role of the {alpha}2-integrin in osteoblast-specific gene expression and activation of the Osf2 transcription factor. J. Biol. Chem., 273, 32988–32994.[Abstract/Free Full Text]

  47. Xiao, G., Jiang, D., Thomas, P., Benson, M.D., Guan, K., Karsenty, G. and Franceschi, R.T. (2000) MAPK pathways activate and phosphorylate the osteoblast-specific transcription factor, Cbfa1. J. Biol. Chem., 275, 4453–4459.[Abstract/Free Full Text]

  48. Price, L.S., Leng, J., Schwartz, M.A. and Bokoch, G.M. (1998) Activation of Rac and Cdc42 by integrins mediates cell spreading. Mol. Biol. Cell., 9, 1863–1871.[Abstract/Free Full Text]

  49. Toma, C.D., Ashkar, S., Gray, M.L., Schaffer, J.L. and Gerstenfeld, L.C. (1997) Signal transduction of mechanical stimuli is dependent on microfilament integrity: identification of osteopontin as a mechanically induced gene in osteoblasts. J. Bone Miner. Res., 12, 1626–1636.[CrossRef][ISI][Medline]

  50. Pavalko, F.M., Chen, N.X., Turner, C.H., Burr, D.B., Atkinson, S., Hsieh, Y.-F., Qui, J. and Duncan, R.L. (1998) Fluid shear-induced mechanical signaling in MC3T3-E1 osteoblasts requires cytoskeletal-integrin interactions. Am. J. Physiol., 275, C1591–C1601.

  51. Tanaka-Kamioka, K., Kamioka, H., Ris, H. and Lim, S.-S. (1998) Osteocyte shape is dependent on actin filaments and osteocyte processes are unique actin-rich projections. J. Bone Miner. Res., 13, 1555–1568.[CrossRef][ISI][Medline]

  52. Sudo, H., Kodama, H., Amagai, Y., Yamamoto, S. and Kasai, S. (1983) In vitro differentiation and calcification in a new clonal osteogenic cell line derived from newborn mouse calvaria. J. Cell. Biol., 96, 191–198.[Abstract/Free Full Text]

  53. Smith, D.B. and Johnson, K.S. (1988) Single-step purification of polypeptides expressed in Escherichia coli as fusions with glutathione S-transferase. Gene, 67, 31–40.[CrossRef][ISI][Medline]

  54. Kaelin, W.G., Jr, Krek, W., Sellers, W.R., DeCaprio, J.A., Ajchenbaum, F., Fuchs, C.S., Chittenden, T., Li, Y., Farnham, P.J. and Blanar, M.A