Molecular Vision 1999; 5:18 <http://www.molvis.org/molvis/v5/p18/> Received 24 March 1999 | Accepted 11 August 1999 | Published 20 August 1999

Download Reprint

FGF-2 facilitates binding of SH3 domain of PLC-[gamma]1 to vinculin and SH2 domains to FGF receptor in corneal endothelial cells

Sun Young Park,¹ Ernesto Barron,¹ Pan-Ghil Suh,² Sung Ho Ryu,² EunDuck P. Kay^1,3
 
 

¹Doheny Eye Institute and ³Department of Ophthalmology, University of Southern California School of Medicine, Los Angeles, CA, USA; ²Pohang University of Sciences and Technology, Pohang, Korea

Correspondence to: EunDuck P. Kay, D.D.S., Ph.D., Doheny Eye Institute, 1450 San Pablo Street DVRC #203, Los Angeles, CA, 90033; Phone: (323) 442-6625, FAX: (323) 442-6688; email: ekay@hsc.usc.edu

Abstract

Purpose: To determine the cellular localization of the Src homology (SH)2 and SH3 domains of PLC-[gamma]1 and their cytoplasmic binding partners, living corneal endothelial cells were microinjected with the fusion proteins containing SH domains.

Methods: Fusion proteins were prepared from plasmid vectors, and the fusion proteins containing SH2-SH2 [(SH2)₂], SH2-SH2-SH3 [(SH2)₂-SH3] or SH3 were isolated using affinity chromatography. Following microinjection, immunolocalization was analyzed using confocal laser microscope.

Results: Microinjected SH domains were targeted to the subcellular location following stimulation with FGF-2: the SH3 domain appeared to be targeted to cytoskeleton; the (SH2)₂ domain showed a dual localization in cytoplasm and plasma membrane; the (SH2)₂-SH3 domain was predominantly localized at membrane and perinuclear sites. In the absence of stimulation by FGF-2, the microinjected fusion proteins remained at the injection sites. When cytoplasmic binding partners were determined by double-staining, the SH3 domain demonstrated colocalization with vinculin: the staining profile of the SH3 domain was identical to that of vinculin, which demonstrates characteristic punctated profiles. The punctated staining of SH3 disappears toward the basal membrane, while that of vinculin remains in all confocal optical sections. On the other hand, some fraction of the (SH2)₂ domain was colocalized with FGF receptor at the membrane site. When PLC-[gamma]1 and F-actin were double-stained, the endogenous PLC-[gamma]1 demonstrated a diffuse cytoplasmic staining and/or perinuclear staining, while phalloidin staining demonstrated that all cells have filamentous cytoplasmic distribution of F-actin.

Conclusions: These findings indicate that the SH3 domain directs PLC-[gamma]1 to bind to vinculin and that the SH2 domains may mediate the binding of PLC-[gamma]1 to receptor tyrosine kinase. Furthermore, they suggest that phosphorylation is not required for targeting of PLC-[gamma]1 to membrane or cytoskeleton sites.

Introduction

Corneal endothelium is a monolayer of differentiated cells located in the posterior portion of the cornea. The corneal endothelium is essential for maintaining corneal transparency, but its capacity for regeneration after injury is severely limited in humans, primates, and cats [1]. In response to certain pathological conditions, corneal endothelial cells (CEC) in vivo may respond by converting to fibroblast-like cells. These morphologically modulated cells then resume their proliferation ability and begin to produce fibrillar collagens, leading to the formation of a fibrillar extracellular matrix. A clinical example of this process is the development of a retrocorneal fibrous membrane [2,3], the presence of which blocks vision, thereby causing blindness. In our previous studies [4-6], corneal endothelium modulation factor (CEMF) released by polymorphonuclear leukocytes, fibroblast growth factor-2 (FGF-2) or a combination of the two factors was found to modulate phenotypes of CEC, leading to a modulation similar to that observed in vivo (up-regulation of cell proliferation, cell shape changes and collagen phenotype alteration). We further found that CEMF could induce de novo synthesis of FGF-2 and that the newly produced FGF-2 is the direct mediator for the modulation of CEC [6].

The biological actions of FGF-2 are mediated by binding and activating receptors that possess tyrosine kinase activity [7,8]. Thus, binding of the growth factor induces receptor dimerization followed by activation of the kinase and autophosphorylation. This process creates binding sites for signal transduction molecules with enzymatic activity, such as phospholipase C[gamma]1 (PLC-[gamma]1) phosphatidylinositol 3-kinase (PI3-kinase) and Ras GTPase activating protein, or for adapter molecules, such as Shc and Grb2. Proteins involved in the signaling pathway consist of one or more modular domains that regulate signal transduction through their ability to mediate protein-protein interactions [9-11]. Among the modules, SH2 domains bind to short phosphotyrosine-containing sequences in growth factor receptors and other phosphoproteins, and SH3 domains bind to target proteins through sequences containing proline and hydrophobic amino acids. PLC-[gamma]1 also contains two SH2 domains and one SH3 domain. The microinjected SH3 domains of PLC-[gamma]1 in living cells are targeted to the cytoskeletal network of microfilaments, while the microinjected SH2 domains of PLC-[gamma]1 have a diffuse cytoplasmic distribution in rat embryo fibroblasts [12]. An in vitro study showed that PLC-[gamma]1 binds to actin cytoskeleton by its COOH-terminal SH2 domain but not the SH3 domain [13]. We previously demonstrated that PLC-[gamma]1 associated with cytoskeleton (vinculin and actin) plays a role in mitogenesis mediated by FGF-2 [14]. In the present study, we attempted to resolve the inconsistency of whether the SH3 domain of PLC-[gamma]1 binds to actin cytoskeleton and further to determine the cellular localization of the SH2 and SH3 domains of PLC-[gamma]1 in response to FGF-2 and their cytoplasmic binding partners in living cells.

Methods

Cell Cultures

Isolation and establishment in culture of rabbit CEC were performed as previously described [4]. Briefly, Descemet's membrane-corneal endothelium complex was treated with 0.2% Type 2 collagenase (Worthington Biochemical Co., Lakewood, NJ) and 0.05% hyaluronidase (HSE; Worthington Biochemical Co.) for 90 min at 37°C. Cultures were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum and 50 µg/ml of gentamicin (DMEM-10) in a 5% CO₂ incubator. This procedure has been shown to promote cell proliferation during the early phase of culture and to maintain the culture as a contact-inhibited monolayer when the cells reach confluency. First passaged CEC were used for all experiments. In some experiments, cells were treated with cytoskeleton stabilization buffer (4 M glycerol, 25 mM PIPES, pH 6.9, 1 mM EGTA, 1 mM MgCl₂).

Preparation of glutathione S-transferase-SH fusion proteins

With cDNA of PLC-[gamma]1 (GenBank accession number J03806) [15] as the template, DNA fragments corresponding to the various domains [(SH2)₂, (SH2)₂-SH3, SH3] were synthesized using the polymerase chain reaction. The amplified DNA was cloned into the BamH1 and EcoR1 site of the pGEX4T-1 bacterial expression plasmid (Pharmacia, Piscataway, New Jersey), which was then used to transform E. coli-competent cells. The glutathione S-transferase (GST) fusion proteins were expressed by induction with 0.1 mM isopropyl-ß-D-thiogalacto-pyranoside, and the cells were collected by centrifugation (7700 x g for 15 min). The cells were sonicated in phosphate-buffered saline (PBS), and Triton-X-100 was added to the lysate to bring the final concentration to 1%. After centrifugation at 12000 x g for 15 min, the supernatant was mixed with a 50% slurry of glutathione Sepharose 4B equilibrated with PBS for 30 min at room temperature. After a brief centrifugation, the supernatant was removed. The glutathione Sepharose 4B pellet was washed with 10 bed volumes of PBS, and the proteins were eluted from the Sepharose beads by glutathione elution buffer (10 mM reduced glutathione, 50 mM Tris-HCl, pH 8.0). Following centrifugation, the resultant supernatant was used for microinjection. For preparation of SH-truncated proteins without GST, the glutathione Sepharose 4B pellet was first washed with PBS; thrombin (10 µg/ml) was then added, and the suspension was gently shaken at room temperature for 16 h. The suspension was centrifuged at 500 x g for 5 min and the supernatant was used as the truncated SH peptides. Concentration of the purified proteins GST-SH3, SH3, GST-(SH2)₂, (SH2)₂, GST-(SH2)₂-SH3, and (SH2)₂-SH3 was assessed with a Bio-Rad DC protein assay system (Bio-Rad Laboratories, Hercules, CA).

Microinjection and Immunofluorescent Staining

Cells were plated on a gridded coverslip (22 x 22 mm) at a concentration of 1.3 x 10⁵/ml. When cells reached approximately 80% confluency, the cells were starved of serum for 24 h then microinjected with fusion proteins. Fusion proteins were diluted to a final concentration of 0.5 mg/ml in PBS, and microinjection was performed using an Eppendorf Microinjector 5242 and Micromanipulator 5170 (Bio-Rad Laboratories). Borosilated pipettes (O.D. 1.0 mm, I. D. 0.78 mm; Sutter Instrument Co., Novato, CA) were pulled using a Flaming/Brown micropipette puller model P-87 (Sutter Instrument Co.). Samples were injected into designated cells on a coverslip; the duration of the injection was 200 msec, the holding pressure was 56 kPa, and the injection pressure was 120 kPa. Following microinjection, cells were incubated for 2 h in serum free medium at 37 °C in a CO₂ incubator, then incubated for an additional 4 h in the presence or absence of FGF-2 (10 µg/ml supplemented with 10 µg/ml of heparin). After incubation, the cells were washed with PBS and fixed with 3% paraformaldehyde in PBS. All washes and incubations were carried out in PBS at room temperature. Cells were then permeabilized with 0.5% Triton X-100 in PBS for 5 min and blocked with 2% bovine serum albumin (BSA) for 1 h. Cells were incubated with primary antibody (1:100 dilution) for 2 h, then washed with PBS. Cells were next incubated with the biotinylated secondary antibody (1:200 dilution; Vector Laboratories, Inc., Burlingame, CA) for 1 h followed by an extensive rinse and then incubation with fluorescein conjugated to avidin (1:100 dilution; Vector Laboratories) for 30 min. For the colocalization experiments of the SH3 domain with vinculin and the (SH2)₂ domain with FGF receptor, the experimental procedures were modified because all primary antibodies were produced in mice. Staining for colocalization of the (SH2)₂ domain and FGF receptor was achieved as follows: cells were incubated with anti-GST antibody (1:100 dilution; Upstate Biotechnology, Lake Placid, NY) prepared in 1% BSA and 0.1% Triton-X-100 in PBS at 37 °C for 1 h, then washed in PBS. Cells were next incubated with Texas Red-conjugated anti-mouse IgG (1:100 dilution; Vector Laboratories) for 30 min at room temperature in the dark. After the first monoclonal reactions were completed, the cells were incubated with normal mouse serum to saturate any open antigen binding sites on the first secondary antibody so that it cannot bind the second primary antibody. Any remaining Fc sites were further blocked by incubating the cells with 0.01 mg/ml of goat anti-mouse IgG Fab fragment (ICN, Aurora, Ohio) at 37 °C for 1 h (or at 4 °C over night). After extensive rinsing in PBS, the second monoclonal antibody, anti-human FGF receptor I (1:100 dilution; Upstate Biotechnology), was allowed to react at 37 °C for 1 h, then rinsed in PBS. This incubation was followed by incubation with the corresponding secondary antibody, FITC-conjugated anti-mouse IgG (1:100 dilution; Vector Laboratories), at room temperature for 1 h in the dark. After a thorough rinse, the stained slides were mounted with vectashield anti-bleaching medium (Vector Laboratories). The identical experimental approach was used for the colocalization of the SH3 domain of PLC-[gamma]1 with vinculin. To identify the SH3 domain of PLC-[gamma]1, anti-PLC-[gamma]1 antibody (Upstate Biotechnology) was used instead of anti-GST-antibody. Antibody labeling was examined using a Zeiss LSM 210 laser scanning confocal microscope equipped with a barrier filter for fluorescein and rhodamine epi fluorescence. A plan-neofluar x40 (N.A. 1.3) oil immersion objective lens was used for imaging of fluorescently labeled cells. For the double labeling experiments separate optical images were generated from the same optical plane, one for fluorescein and another for Texas-Red. The captured images were then pseudocolored green or red and digitally overlaid to visualize any colocalization. Regions of colocalization appear yellow in color, reflecting the additive effect of superimposing green and red pixels. Image analysis was performed using the standard system operating software provided with the Zeiss LSM microscope (Version 2.08). Color photomicrographs were taken using a Sony printer connected to the video output of the microscope. Approximately 30% of the microinjected cells showed the immunoreactivity with the fusion proteins.

SDS-Polyacrylamide Gel Electrophoresis and Immunoblotting

The conditions of electrophoresis were as described by Laemmli [16]. Fusion proteins were analyzed on a 10% gel under the reduced condition using a discontinuous Tris-glycine buffer system (pH 8.3). Proteins separated by SDS-PAGE were stained with Coomassie brilliant blue R250 (0.25 g of Coomassie blue in 90 ml of methanol:H₂O and 10 ml of glacial acetic acid) for 5 h at room temperature. The gel was then destained with methanol/acetic acid for 10 h. For immunoblotting analysis, proteins separated by SDS-PAGE were electrophoretically transferred to a polyvinylidene difluoride (PVDF) membrane at 0.22 ampere for 10 h in a semi-dry transfer system (Transfer buffer: 25 mM Tris-HCl, pH 8.3, 190 mM glycine, 20% MeOH). Immunoblot analysis was performed as described previously [6,14], using a commercial ABC Vectastain kit (Vector Laboratories). All washes and incubations were carried out at room temperature in TTBS (0.9% NaCl, 100 mM Tris-HCl, pH 7.5, 0.1% Tween 20). Briefly, the PVDF membrane was immediately placed into the blocking buffer (5% nonfat milk containing TTBS) for 1 h. The incubation with primary antibody (1:5000 dilution) was carried out for 2 h; incubation with biotinylated secondary antibody (1:2500 dilution) for 1 h, and incubation with the ABC reagent for 30 min. The membrane was treated with enhanced chemiluminescence (ECL) reagent (Amersham Life Science, Buckinghamshire, England) for 1 min, and the ECL-treated membrane was exposed to ECL film.

Results

Colocalization of the SH3 domain of PLC-[gamma]1 with vinculin

Various PLC-[gamma]1 truncated proteins were generated (Figure 1) to investigate whether the subcellular localization of PLC-[gamma]1 is regulated by its SH2 and SH3 domains. The construct containing two SH2 and SH3 domains encompasses nucleotides corresponding to amino acids 533-851; the construct containing two SH2 domains encompasses nucleotides corresponding to amino acids 533-756; and the construct containing one SH3 domain encompasses nucleotides corresponding to amino acids 791-851. The GST-fusion proteins containing various SH domains were further examined on SDS-PAGE using Coomassie blue staining and immunoblotting analysis (Figure 2A,B): the fusion protein containing the SH3 domain of PLC-[gamma]1 (GST-SH3) showed a molecular size of approximately 39 kDa; the fusion protein containing the (SH2)₂ domain of PLC-[gamma]1 [GST-(SH2)₂] had a molecular weight of 62 kDa. The fusion protein containing the (SH2)₂-SH3 domain of PLC-[gamma]1 [GST-(SH2)₂-SH3] had a molecular weight of approximately 78 kDa. The assessed molecular weights of the fusion proteins are in agreement with the calculated sizes. The identity of the peptide bands with lower molecular weights observed in GST-(SH2)₂-SH3 by immunoblotting analysis is not known (Figure 2B, lane 3). In order to confirm that anti-PLC-[gamma]1 antibody reacts with SH3 domain and does not react with the (SH2)₂ domain, the (SH2)₂-SH3 domain and the (SH2)₂ domain were immunoblotted with anti-PLC-[gamma]1 antibody. Figure 2D shows that the (SH2)₂-SH3 domain demonstrated a positive reaction with anti-PLC-[gamma]1 antibody, while the antibody did not react with the (SH2)₂ domain. When phosphorylation of tyrosine residues at 771 and 783 in the GST-(SH2)₂-SH3 was examined using anti-phosphotyrosine antibody, the truncated protein demonstrated lack of phosphorylation, which is agreeable because the fusion proteins were cloned in bacterial expression vector (data not shown).

To examine whether the SH3 domain mediates interactions of PLC-[gamma]1 with the cytoskeleton, GST-SH3 or GST-(SH2)₂-SH3 was microinjected into the cytoplasm of CEC in the presence or absence of stimulation by FGF-2. The localization of the truncated protein within the injected cells was determined by immunofluorescence microscopy, using anti-PLC-[gamma]1 antibody. As shown in Figure 3A,C, the fusion proteins were restricted at the injection site in the absence of FGF-2. With FGF-2 stimulation, GST-SH3 showed a characteristic punctated staining profile (Figure 3B), whereas GST-(SH2)₂-SH3 demonstrated prominent plasma membrane staining and perinuclear staining (Figure 3D). Of interest is that the targeting of the microinjected GST-fusion proteins is restricted to those areas adjacent to the injection site during the 4 h stimulation with FGF-2. It is possible that the antibody reacts with the endogenous PLC-[gamma]1; therefore, the cells, which were not used for microinjection in Figure 3 were simultaneously examined (Figure 4). The endogenous PLC-[gamma]1, in the presence of the 4 h stimulation by FGF-2, demonstrated a diffuse cytoplasmic staining and perinuclear staining (Figure 4A), which differs from the staining profiles of the microinjected fusion proteins. In the absence of FGF-2 stimulation, the endogenous PLC-[gamma]1 in the cells that were not used for microinjection markedly loses its staining potential (Figure 4B), whereas the two cells microinjected with GST-SH3 showed the restricted localization similar to that shown in Figure 3A.

Since the staining profile of GST-SH3 in CEC is similar to the profile seen with vinculin, which binds F-actin, cells microinjected with GST-SH3 were simultaneously stained with anti-PLC-[gamma]1 and anti-vinculin antibodies. Because these two monoclonal antibodies were produced in mice, immunofluorescent staining protocols were modified to distinguish the two proteins. Figure 5 shows that most of the microinjected GST-SH3 is colocalized with vinculin (Figure 5C). When the same cell was further analyzed using confocal optical sectioning, GST-SH3 staining was observed toward the apical cell membrane (Figure 5A) and disappeared toward the plane of attachment (Figure 5D). Unlike the specific localization of GST-SH3, vinculin staining was observed at both apical and basal cell membranes (Figure 5B and E). This result further confirms the specific staining potential of anti-PLC-[gamma]1 and anti-vinculin antibodies. The colocalization of PLC-[gamma]1 and vinculin is consistent with our previous finding in which PLC-[gamma]1 is associated with vinculin and actin following stimulation with FGF-2 when determined by immunoprecipitation and immunoblotting [14].

Colocalization of the SH2-SH2 domain of PLC-[gamma]1 with FGF receptor

To confirm that the SH2 domain of PLC-[gamma]1 binds to the activated receptor tyrosine kinases, GST-(SH2)₂ fusion protein was microinjected into the cytoplasm of CEC. Anti-GST antibody was used because the commercially available anti-PLC-[gamma]1 antibody recognizes only the SH3 domain of PLC-[gamma]1 as shown in Figure 2 and the antibodies recognizing the (SH2)₂ domain of PLC-[gamma]1 are not presently available. Figure 6 demonstrates that GST-(SH2)₂ fusion proteins were localized at the injection site in the absence of FGF-2 (Figure 6D). Upon stimulation of FGF receptors (FGFR) with exogenous FGF-2, the microinjected GST-(SH2)₂ protein was predominantly present in cytoplasm, and positive staining was also partly detected both in plasma membrane and in the nuclear membrane (Figure 6A). The positive staining of GST-(SH2)₂at the membrane sites in CEC differs slightly from that in rat embryo fibroblasts (REF) [12], in which a diffuse cytoplasmic staining was dominant in addition to a nuclear staining. To investigate whether the membrane site of GST-(SH2)₂ is coincidental with FGFR, colocalization of the FGFR and the SH2 domains of PLC-[gamma]1 was determined using double-staining; colocalization of these two proteins is observed at some cell membrane and nuclear membrane sites (Figure 6C). In order to confirm that the secondary antibodies bind the correct primary antibodies, anti-FGF receptor I antibody was omitted in some experiments, followed by incubating cells with FITC-conjugated anti-mouse IgG. In this control experiment, FITC staining was not observed (data not shown). The observation that only a minor subset of PLC-[gamma]1 is colocalized with FGFR following the FGF-2 stimulation is consistent with our previous findings in which only a minor fraction of PLC-[gamma]1 is translocated to the membrane fraction following FGF-2 stimulation when determined by immunoprecipitation and immunoblotting [14].

Colocalization of F-actin and PLC-[gamma]1

To determine whether the treatment of cells with cytoskeleton stabilization buffer (CSB) causes differential staining potentials observed in CEC and REF, CEC were treated with CSB to stabilize actin filament. CEC were double-stained with anti-PLC-[gamma]1 antibody and fluorescein-conjugated phalloidin. As seen in Figure 7A, PLC-[gamma]1 is diffusely stained in the cytoplasm, while F-actin demonstrated filamentous structures (Figure 7B). These two proteins are not colocalized in CEC regardless of the treatment of CSB (Figure 7C). The staining pattern of PLC-[gamma]1 differs from the pattern in REF-52 cells, in which the endogenous PLC-[gamma]1 is colocalized with F-actin [17]. It should be noted that the staining profile of PLC-[gamma]1 in CEC shows a variation, as observed in Figure 4A in which some cells demonstrate perinuclear staining as well as the diffuse cytoplasmic staining. In other experiments in which cells were treated with CSB, approximately 5% of cell population show filamentous staining, as shown in REF-52 [17].

Discussion

The SH2 and SH3 domains are small conserved modules that mediate protein-protein interaction in signaling pathways regulated by receptor protein tyrosine kinases. Signaling proteins that contain the SH2 and SH3 domains include PLC-[gamma]1, Ras GTPase-activating protein, and the p85 subunit of PI3-kinase. Numerous studies indicate that the tyrosine autophosphorylation sites of growth factor receptors recruit the SH2 domain containing signaling proteins [9-11,18,19]. Binding of the SH2 domains to their target proteins is totally dependent on tyrosine phosphorylation of the target proteins. Recent studies show that PLC-[gamma]1 directly binds to the FGFR-1 [20] and that the SH2 domains of PLC-[gamma]1 recognize a common motif found in FGF receptors [20] and EGF receptors [21]. While it is now well established that SH2 domains interact with tyrosine-phosphorylated proteins and that one major partner is receptor tyrosine kinase, less information is available regarding the binding nature of the SH3 domain of PLC-[gamma]1. Schlessinger and his coworkers reported that the SH3 domain of PLC-[gamma]1, by itself, can localize to the actin cytoskeleton and that this binding is independent of phosphorylation of tyrosine residues in PLC-[gamma]1 [12]. They further speculated that the SH3 domain may have a role in directing PLC-[gamma]1 to a subcellular localization to regulate actin polymerization. However, an in vitro binding study demonstrates that PLC-[gamma]1 binds the actin cytoskeleton by its COOH-terminal SH2 domain but not the SH3 domain [13]. The inconsistency of these two studies may be the result of the experimental approaches used for analysis: immunolocalization in living cells versus the in vitro binding study.

In the present study we attempt to determine the target proteins of the SH2 and SH3 domains of PLC-[gamma]1 and to investigate their subcellular localization in living cells following the stimulation with FGF-2. Our previous study [14] demonstrates that PLC-[gamma]1 is associated with actin and vinculin and that the PLC-[gamma]1 complex formed with the two proteins is responsible for the stimulatory activity of FGF-2 on cell proliferation of CEC. Cytochalasin B is able to block the complex formation and to further inhibit the cell proliferation mediated by FGF-2 [14]. Thus, a microfilament system is involved in relaying the mitogenic signals of FGF-2 via PLC-[gamma]1. In the present study, we have demonstrated that the SH3 domain of PLC-[gamma]1 is colocalized with vinculin when microinjected into CEC. This observation confirms our previous results on the complex formation of PLC-[gamma]1 with vinculin and actin [14]. Vinculin is known to play a major role in anchoring actin to the membrane. Together, the previous study (REF) and our present study (CEC) suggest that the SH3 domain of PLC-[gamma]1 directly binds to the actin cytoskeleton or binds to the actin cytoskeleton via vinculin. Our finding that PLC-[gamma]1 binds to the actin cytoskeleton via vinculin is more acceptable than the association of PLC-[gamma]1 with F-actin because vinculin, which is composed of a large globular head domain and a rod-like tail domain, has a more flexible structure, with the actin-binding site at the tail and other interacting proteins bound to the head domain [22]. Furthermore, vinculin has a proline-rich motif [23,24]. It is highly likely that the SH3 domain of PLC-[gamma]1 binds the proline-rich sequences of vinculin. The differential staining profiles of the endogenous PLC-[gamma]1 from that of F-actin further demonstrate that PLC-[gamma]1 may not bind directly to F-actin in CEC.

An interesting observation is that the colocalization of the SH3 domain of PLC-[gamma]1 and vinculin is found to be restricted: vinculin is localized at both the basal membrane and the apical membrane, but the SH3 domain of PLC-[gamma]1 is located only in the apical portion of cells. This observation suggests that the interaction of PLC-[gamma]1 with cytoskeleton is restricted to the region of the apical cell membrane and that the growth factor receptor signaling pathway via the PLC-[gamma]1-vinculin complex is not present in the focal adhesion complex. On the other hand, vinculin appears to be involved in focal adhesion via the integrin-signaling pathway as well. The restricted subcellular localization of PLC-[gamma]1 at the apical cell membrane is also observed in 3T3 fibroblasts (Suh et al., unpublished). Another interesting observation is that there is a differential staining pattern of the microinjected GST-(SH2)₂-SH3 and GST-SH3 proteins: the SH3 domain alone is targeted to the actin cytoskeleton through binding to vinculin; however, the additional presence of the SH2 domain further promotes localization in the plasma membrane and perinuclear regions. The additional subcellular localization of the (SH2)₂SH3 domain suggests that the SH2 domain may further promote binding of PLC-[gamma]1 to actin cytoskeleton and plasma membrane. This observation may confirm the findings obtained from the in vitro binding study, which demonstrates binding between the COOH-terminal SH2 domain and actin cytoskeleton [13]. When the cytoplasmic target proteins of the (SH2)₂ domain were determined, some fraction of (SH2)₂ was colocalized with the FGF receptor, while the major portion of (SH2)₂ is localized in cytoplasm, confirming our previous data that only a minor subset of PLC-[gamma]1 is translocated to the membrane fraction following FGF-2 stimulation [14]. Although it appears that such a low level of receptor-associated PLC-[gamma]1 is sufficient for the mitogenic signaling pathway, the present study indicates that the SH2 and SH3 domains of PLC-[gamma]1 are able to target the injected truncated proteins to their natural intracellular sites.

Together, these results may suggest that the function of the SH2 and SH3 domains of PLC-[gamma]1 is to regulate specific protein-protein interactions and cellular localization during signal transduction mediated by FGF-2. While targeting of the SH2 domains to FGF receptors, albeit at a low level, is expected in mitogenic signal transduction, the biological role of the binding of the SH3 domain of PLC-[gamma]1 to vinculin at the apical membrane requires further study. The present study further suggests that there may be three-dimensional structural interaction among the proteins that associate with one another in signal transduction, and that confocal optical sectioning is required to investigate the protein-protein interaction at a three-dimensional level.

Acknowledgements

The authors thank Marthann Salazar for her technical expertise in microinjection. Supported by Grants EY06431 and EY03040 from the National Institutes of Health, Bethesda, Maryland; and by an unrestricted grant from Research to Prevent Blindness, New York, New York.

References

1. Van Horn DL, Sendele DD, Seideman S, Buco PJ. Regenerative capacity of the corneal endothelium in rabbit and cat. Invest Ophthalmol Vis Sci 1977; 16:597-613.

2. Michels RG, Kenyon KR, Maumence AE. Retrocorneal fibrous membrane. Invest Ophthalmol 1972; 11:822-31.

3. Brown SI, Kitano S. Pathogenesis of the retrocorneal membrane. Arch Ophthalmol 1966; 75:518-25.

4. Kay EP, Nimni ME, Smith RE. Modulation of endothelial cell morphology and collagen synthesis by polymorphonuclear leukocytes. Invest Ophthalmol Vis Sci 1984; 25:502-12.

5. Kay EP, Gu X, Ninomiya Y, Smith RE. Corneal endothelial modulation: a factor released by leukocytes induces basic fibroblast growth factor that modulates cell shape and collagen. Invest Ophthalmol Vis Sci 1993; 34:663-72.

6. Kay EP, Gu X, Smith RE. Corneal endothelial modulation: bFGF as direct mediator and corneal endothelium modulation factor as inducer. Invest Ophthalmol Vis Sci 1994; 35:2427-35.

7. Ullrich A, Schlessinger J. Signal transduction by receptors with tyrosine kinase activity. Cell 1990; 61:203-12.

8. Klagsbrun M, Baird A. A dual receptor system is required for basic fibroblast growth factor activity. Cell 1991; 67:229-31.

9. Koch CA, Anderson D, Moran MF, Ellis C, Pawson T. SH2 and SH3 domains: elements that control interactions of cytoplasmic signaling proteins. Science 1991; 252:668-74.

10. Schlessinger J. SH2/SH3 signaling proteins. Curr Opin Genet Dev 1994; 4:25-30.

11. Pawson T. Protein modules and signalling networks. Nature 1995; 373:573-80.

12. Bar-Sagi D, Rotin D, Batzer A, Mandiyan V, Schlessinger J. SH3 domains direct cellular localization of signaling molecules. Cell 1993; 74:83-91.

13. Pei Z, Yang L, Williamson JR. Phospholipase C-gamma 1 binds to actin-cytoskeleton via its C-terminal SH2 domain in vitro. Biochem Biophys Res Commun 1996; 228:802-6.

14. Gu X, Seong GJ, Lee YG, Kay EP. Fibroblast growth factor 2 uses distinct signaling pathways for cell proliferation and cell shape changes in corneal endothelial cells. Invest Ophthalmol Vis Sci 1996; 37:2326-34.

15. Suh PG, Ryu SH, Moon KH, Suh HW, Rhee SG. Cloning and sequence of multiple forms of phospholipase C. Cell 1988; 54:161-9.

16. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970; 227:680-5.

17. McBride K, Rhee SG, Jaken S. Immunocytochemical localization of phospholipase C-gamma in rat embryo fibroblasts. Proc Natl Acad Sci U S A 1991; 88:7111-5.

18. Feng GS, Hui CC, Pawson T. SH2-containing phosphotyrosine phosphatase as a target of protein-tyrosine kinases. Science 1993; 259:1607-11.

19. Cohen GB, Ren R, Baltimore D. Modular binding domains in signal transduction proteins. Cell 1995; 80:237-48.

20. Mohammadi M, Honegger AM, Rotin D, Fischer R, Bellot F, Li W, Dionne CA, Jaye M, Rubinstein M, Schlessinger J. A tyrosine-phosphorylated carboxy-terminal peptide of the fibroblast growth factor receptor (Flg) is a binding site for the SH2 domain of phospholipase C-gamma 1. Mol Cell Biol 1991; 11:5068-78.

21. Rotin D, Margolis B, Mohammadi M, Daly RJ, Daum G, Li N, Fischer EH, Burgess WH, Ullrich A, Schlessinger J. SH2 domains prevent tyrosine dephosphorylation of the EGF receptor: identification of Tyr992 as the high-affinity binding site for SH2 domains of phospholipase C-gamma. EMBO J 1992; 11:559-67.

22. Kioka N, Sakata S, Kawauchi T, Amachi T, Akiyama SK, Okazaki K, Yaen C, Yamada KM, Aota S. Vinexin: a novel vinculin-binding protein with multiple SH3 domains enhances actin cytoskeletal organization. J Cell Biol 1999; 144:59-69.

23. Coutu MD, Craig SW. cDNA-derived sequence of chicken embryo vinculin. Proc Natl Acad Sci U S A 1988; 85:8535-9.

24. Bubeck P, Pistor S, Wehland J, Jockusch BM. Ligand recruitment by vinculin domains in transfected cells. J Cell Sci 1997; 110:1361-71.