Association of tyrosine phosphatase SHP-2 with F-actin at low cell densities.

SHP-2 is an intracellular SH2 domain-containing protein-tyrosine phosphatase with an essential role in cell signaling. Here we demonstrate that localization of SHP-2 is regulated by cell density in a cell adhesion-dependent manner. When cells were plated at low densities, SHP-2 was distributed in Triton X-100-insoluble fractions, whereas it was totally soluble when cells were plated at high densities or when low density cells approached confluency. In all cases, the total protein level of SHP-2 was not changed. Fluorescent cell staining revealed that SHP-2 was co-localized with actin stress fibers to the cell peripheral at low cell densities but was diffused in the entire cytoplasm at high cell densities. Transient transfection of cells with truncated forms of SHP-2 demonstrated that the catalytic domain of the enzyme was responsible for the density-regulated distribution of SHP-2, but the catalytic activity was not required. An in vitro co-sedimentation study demonstrated direct binding of full-length and SH2 domain-truncated forms of SHP-2 to F-actin. The data indicate that SHP-2 is regulated by cell density and that it may have a role in assembling and disassembling of the actin network.

SHP-2 is an intracellular SH2 domain-containing protein-tyrosine phosphatase with an essential role in cell signaling. Here we demonstrate that localization of SHP-2 is regulated by cell density in a cell adhesion-dependent manner. When cells were plated at low densities, SHP-2 was distributed in Triton X-100-insoluble fractions, whereas it was totally soluble when cells were plated at high densities or when low density cells approached confluency. In all cases, the total protein level of SHP-2 was not changed. Fluorescent cell staining revealed that SHP-2 was co-localized with actin stress fibers to the cell peripheral at low cell densities but was diffused in the entire cytoplasm at high cell densities. Transient transfection of cells with truncated forms of SHP-2 demonstrated that the catalytic domain of the enzyme was responsible for the density-regulated distribution of SHP-2, but the catalytic activity was not required. An in vitro co-sedimentation study demonstrated direct binding of full-length and SH2 domaintruncated forms of SHP-2 to F-actin. The data indicate that SHP-2 is regulated by cell density and that it may have a role in assembling and disassembling of the actin network.
SHP-2 is a widely distributed protein-tyrosine phosphatase (PTP) 1 that contains tandem SH2 domains (1)(2)(3)(4). By structural nature, SHP-2 is an intracellular enzyme. However, it is recruited to the plasma membrane by binding to tyrosine phosphorylated molecules, including growth and cytokine receptors, the T and B cell receptors (1)(2)(3)(4), and other cell surface anchor proteins including SHPS-1 (5, 6), PECAM-1 (7,8), and PZR (9). Overwhelming studies have shown that SHP-2 is most often a positive signal transducer. For example, it plays a positive role in activation of the mitogen-activated protein kinase pathway induced by platelet-derived growth factor, epidermal growth factor, insulin, insulin growth factor-1, and fibroblast growth factor (10 -17). Furthermore, disruption of the SHP-2 gene in mice caused death of mouse embryos at mid-gestation (18). Recent studies also suggested a role of SHP-2 in cell adhesion and migration. Overexpression of a catalytically inactive form of SHP-2 blocked insulin growth factor-1-stimulated chemo-taxis (19). Platelet-derived growth factor-induced membrane edge ruffling and chemotaxis were suppressed in cells expressing the Tyr 1009 to Phe mutant form of platelet-derived growth factor-receptor ␤, which lacks SHP-2 binding ability (20). Fibroblast cells derived from SHP-2-deficient mice exhibited impairment in cell migration and spreading and in integrininduced activation of Src family tyrosine kinases and mitogenactivated protein kinase and tyrosine phosphorylation of FAK, paxillin, and p130 cas (21,22). Furthermore, upon integrin engagement, a fraction of SHP-2 moves to focal contacts where it binds to tyrosine phosphorylated SHPS-1, a transmembrane glycoprotein with characteristics of adhesion molecules (22). These data indicate an essential role of SHP-2 in integrinmediated cell signaling. SHP-2 also plays an important role in the control of cell shape by contributing to cytoskeletal organization. It has been shown that overexpression of a catalytically inactive mutant form of SHP-2 in Chinese hamster ovary cells or Rat-1 fibroblasts induced a marked change in cell morphology accompanied by substantial increases in the numbers of actin stress fibers and focal adhesion contacts (23). In Madin-Darby canine kidney cells, expression of the same SHP-2 mutant markedly increased the formation of stress fibers and focal adhesions and inhibited scattering of the cells (24). Finally, SHP-2 was found to be necessary for morphological transformation by v-Src because v-Src-induced reorganization of the actin cytoskeleton and the formation of podosomes were compromised in SHP-2-deficient cells (25).
Cell density has major effects on cell activities including cell morphology, motility, and cytokinesis. In this study, we have investigated the effects of cell density on intracellular distribution of SHP-2. We found that cell density regulates localization of SHP-2. At low cell densities, SHP-2 was co-localized with F-actin. To our knowledge, this is the first example that a tyrosine phosphatase constitutes a part of actin stress fibers.

EXPERIMENTAL PROCEDURES
Materials-Polyclonal anti-SHP-2 antibody (C-18) was purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Rhodamineconjugated phalloidin, fluorescein isothiocyanate-labeled goat anti-rabbit IgG, and cytochalasin D were from Sigma. Horseradish peroxidaseconjugated anti-rabbit immunoglobulin was from Amersham Pharmacia Biotech. ⌬SHP-2 is an SH2 domain-truncated form of SHP-2 (amino acid residues 200 -593), whereas ⌬SHP-2M represents its catalytically inactive Cys to Ser mutant form. Both constructs were built with the pRC/CMV vector (Invitrogen) as described (26). Purified SHP-2 and ⌬SHP-2 enzymes were obtained from recombinant Escherichia coli cells as described previously (27). The full-length SHP-2 was also expressed as a glutathione S-transferase (GST) fusion protein in E. coli cells by using the pGex-2T vector. The fusion protein designated GST-SHP-2 was purified on a glutathione-Sepharose column. GST alone was purified following the same procedure and was used as a control.
Cell Culture and Transfection-HeLa, HepG2, NIH-3T3, and HT-1080 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum and antibiotics at 37°C and 5% CO 2 . Confluent cells were trypsinized, washed with the culture me-dium, and replated in 10-cm plastic tissue culture plates at densities of 0.2-6 ϫ 10 6 cells/plate. Transfection of HT-1080 cells with pRC/CMV-⌬SHP-2 and pRC/CMV-⌬SHP-2M was carried out by using the Fu-GENE6 cell transfection system (Roche Molecular Biochemicals). To culture HeLa cells in suspension, Cytostir suspension culture flasks (Knotes) were used, and the cells were grown at 0.1-10 ϫ 10 5 /ml under constant stirring at 30 rpm.
Cell Extraction and Western Blotting-Cells were washed twice with ice-cold phosphate-buffered saline (PBS) and then lysed on ice in 0.5-1 ml of Buffer A containing 25 mM ␤-glycerolphosphate (pH 7.3), 5 mM EDTA, 2 mM EGTA, 0.1 M NaCl, 0.25% Triton X-100, 10 mM ␤-mercaptoethanol, 0.2 mM Na 3 VO 4 , 1 mM benzamidine, 0.1 mM phenylmethylsulfonyl fluoride, 2 g/ml leupeptin, 1 M pepstatin A, and 1 g/ml aprotinin. Unless indicated otherwise, lysis of attached cells was performed directly on the plates. The lysates were centrifuged at 15,000 ϫ g for 15 min at 4°C, and the supernatant was referred to as the Triton X-100-soluble fractions. The pellets were washed once with Buffer A and then extracted with SDS gel sample buffer containing 1% SDS followed by boiling and was referred to as Triton X-100-insoluble fractions. For Western blotting, cell extracts (usually containing 10 g of proteins) were resolved by 10% SDS-polyacrylamide gel electrophoresis and electrophoretically transferred to polyvinylidene difluoride membranes. The membranes were probed with various primary antibodies and then with appropriate horseradish peroxidase-conjugated secondary antibodies. Detection was made by employing a ECL Western blotting reagent (Amersham Pharmacia Biotech).
Immunofluorescent Cell Staining-The cells were grown overnight on glass coverslips (Becton Dickinson Labware) in complete cell culture medium. After washing twice with PBS, the cells were fixed in 3.7% formaldehyde for 20 min followed by further wash with PBS. The fixed cells were then permeabilized with 0.2% Triton X-100 in PBS for 5 min and then blocked by 10 mM Tris-HCl (pH 7.5) followed by 5% goat serum in PBS. For detection of actin filaments, namely, F-actin, the treated cells were stained with rhodamine-conjugated phalloidin (5 g/ml in PBS). For indirect immunofluorescent staining of SHP-2, the cells were incubated with 1:100 diluted anti-SHP-2 antibody for 2 h at room temperature and then with fluorescein isothiocyanate-labeled goat anti-rabbit antibody at a concentration recommended by the manufacturer. Coverslips were mounted onto glass slides and visualized by using a Zeiss Axiophot microscope with a 100ϫ oil immersion objective. Cell images were captured with a cooled CCD digital camera (Princeton Instruments, Inc.) and processed by using the Adobe Photoshop software.
In Vitro Binding of SHP-2 with F-actin-G-actin was purified from rabbit skeletal muscle following the procedure of Pardee and Spudich (28). The purified G-actin showed a homogenous band on SDS gels. G-actin was diluted to 2 mg/ml in the depolymerization buffer containing 2 mM Tris-HCl (pH 8.0), 0.2 mM ATP, 1 mM ␤-mercaptoethanol, and 0.2 mM CaCl 2 . GST-SHP-2, GST, SHP-2, and ⌬SHP-2 were diluted to 0.1 mg/ml in a buffer containing 20 mM Tris-HCl (pH 8.0), 1 mM glutathione, 1 mM dithiothreitol, and 1.0 mg/ml bovine serum albumin. The solutions were centrifuged at 100,000 ϫ g for 45 min to remove any possible precipitation, and the supernatants were used for binding assays. Equal volumes of G-actin and GST-SHP-2, GST, SHP-2, or ⌬SHP-2 were mixed. Polymerization of actin was triggered by addition of 2 M KCl, 20 mM ATP, 1 M MgCl 2 , and 0.1 M EGTA to final concentrations of 0.1 M KCl, 1 mM ATP, 5 mM MgCl 2 , and 0.5 mM EGTA. The samples were incubated on ice for 2 h and then centrifuged at 100,000 ϫ g for 45 min. The supernatants were removed and used to make SDS gel samples. The pellets were washed with the same polymerization buffer and centrifuged as above. The final pellets were dissolved in 1ϫ SDS sample buffer. GST and GST-SHP-2 were analyzed by Western blotting with an anti-GST antibody, whereas SHP-2 and ⌬SHP-2 were detected by using anti-SHP-2 antibody.

RESULTS
Cell Density Regulates the Level of Triton X-100-soluble SHP-2-As an important transducer of cell signaling, SHP-2 has a major implication in cell migration (3). Because cell migration is regulated by cell density, SHP-2 may play a role in densitydependent cellular processes. We plated four different cell lines, including HeLa, HepG2, NIH 3T3, and HT-1080 in plastic cell culture dishes at densities of 0.2-6 ϫ 10 6 /10-cm plate. After overnight culturing, cells were lysed on the plate with a buffer containing 0.25% Triton X-100. The protein level of SHP-2 was determined by Western blotting with an anti-SHP-2 polyclonal antibody. As shown in the top panel of Fig. 1, cell density had a profound effect on the protein level of SHP-2 in the Triton X-100-soluble cell fraction with all the cell lines tested. At low cell densities, SHP-2 was hardly detected. As the cell density increased, a marked elevation of SHP-2 level was observed. As a control, the protein level of focal adhesion kinase FAK was not affected by the change in cell density (Fig. 1, middle panel). In fact, our recent studies showed that tyrosine phosphorylation of FAK decreased as the cell density increased (29). We used HeLa cells to investigate further the effects of cell density on the SHP-2 protein level. The cells were plated at a low density and cultured for 6 days. As shown in the bottom panel of Fig. 1, the level of SHP-2 was lower at day 1 but gradually increased as the cells approached confluency. We also investigated the change of SHP-2 level during a shorter time course (Fig. 1, bottom panel, right). Trypsinized HeLa cells were plated at low, medium, and high densities (0.2, 1, and 6 ϫ 10 6 /plate) and cultured cells for 3, 6, and 24 h. Before plating, the trypsinized suspension cells displayed a high level of soluble SHP-2. A near total depletion of soluble SHP-2 was seen after 3 h of plating at the low cell density (0.2 ϫ 10 6 /plate), and a slight gradual increase was observed by 24 h. When cells were plated at higher densities (1 and 6 ϫ 10 6 /plate), the level of soluble SHP-2 was constantly high. These data thus suggest that cell density modulated SHP-2 protein level in the Triton X-100-soluble fraction.
Regulation of Soluble SHP-2 Level by Cell Density Is Dependent on Cell Adhesion-As noted above, the level of soluble SHP-2 in trypsinized suspension cells was high, and thus the decrease must be related to cell attachment to the plates. To further investigate the effects of cell adhesion on SHP-2 solubility, we cultured HeLa cells in suspension as described under "Experimental Procedures." HeLa cells grew in suspension medium at a rate similar to that observed with adherent cell culture. Western blot analyses demonstrated a constantly high level of SHP-2 in HeLa cells cultured in suspension despite cell FIG. 1. The protein level of SHP-2 in the Triton X-100-soluble cell extract is regulated by cell density. HeLa, HepG2, NIH-3T3, and HT-1080 cells were plated at the indicated densities (0.2-6 ϫ 10 6 /10-cm plates) and cultured overnight (top and middle panels). HeLa cells were plated at a low density (0.2 ϫ 10 6 /10-cm plate) for 1-6 days as indicated or at the indicated densities and cultured for 0, 3, 6, and 24 h (bottom panel). The Cells were lysed at 0°C on the plates with Buffer A. Cell extracts containing 10 g of total proteins were separated on 10% SDS gel and then transferred to polyvinylidene difluoride membranes. Western blotting was performed with the anti-SHP-2 and anti-FAK antibodies. densities and the time the cells were cultured in suspension (Fig. 2, upper panel). However, once the cells were transferred to cell culture plates and allowed to attach to the plates at a low density, the level of soluble SHP-2 decreased dramatically. As expected, when the attached cells approached confluence on the plates, the level of soluble SHP-2 was recovered. To further the study, we detached the cells plated at the low density (0.2 ϫ 10 6 /plate) by incubating them in PBS at 37°C for 0.5 h. Interestingly, this caused a total recovery of SHP-2 in the soluble fraction (Fig. 2, lower panel). PBS treatment of the cells plated at higher densities exhibited no effects on SHP-2, which remained in the soluble fraction. These results indicate that the protein level of SHP-2 in the Triton X-100-soluble fraction is associated with both cell density and cell attachment.
SHP-2 Is Partitioned in the Triton X-100-insoluble Fraction at Low Cell Densities-The protein level of SHP-2 in the Triton X-100-soluble fraction may not necessarily reflect the total protein level of SHP-2 in the cells. To clarify this, we extracted cells first with Buffer A and then re-extracted the pellets with the SDS sample buffer containing 1% SDS. As shown in Fig. 3, a significant portion of SHP-2 was found in the Triton X-100insoluble pellets when cells were grown at a low density. As the density increased, the SHP-2 level in the Triton X-100-soluble fraction increased, but that in the Triton X-100-insoluble fraction displayed a concurrent decrease. As expected, when the cells were directly extracted with SDS gel sample buffer containing 1% SDS, the total level of SHP-2 showed no change at different cell densities. Similar results were obtained with HT-1080 cells (Fig. 3, bottom panel). In this case, a nonspecific band below SHP-2 served as an internal control to show a protein whose protein level was not affected by cell density. Taken together, the data suggest that cell density regulates cellular distribution of SHP-2.
SHP-2 Is Co-localized with Actin Stress Fibers at Low Cell Densities-Cytoskeletal proteins constitute a major part of the Triton X-100-insoluble cell pellets. Partition of SHP-2 in the Triton X-100-insoluble fraction implies that it may be associated with cytoskeleton. To find the exact intracellular localization of SHP-2, we performed immunofluorescent cell staining with anti-SHP-2 antibody. The results are shown in Fig. 4. At low cell densities, SHP-2 was condensed at the periphery of cells with a fiber-like structure. At high densities, however, SHP-2 was diffused in the entire cytosol. We then co-stained the cells with rhodamine-conjugated phalloidin that detects F-actin. In low density cells, F-actin staining essentially super-imposed the SHP-2 staining, indicating that SHP-2 is co-localized with the actin stress fibers formed by F-actin. At high cell densities, SHP-2 was dissociated from the actin stress fibers. It should be noted that the distribution of the actin stress fibers at low cell densities was different from that at high cell densities. At low cell densities, actin stress fibers are distributed around the periphery of cells, whereas they were dispersed over the ventral surface in high-density cells. In addition, at lower density, the cells appeared larger because of extensive spreading. To further verify the co-localization of SHP-2 with F-actin, we treated cells with cytochalasin D to break actin stress fibers. As shown in Fig. 4, cytochalasin D caused total disruption of the fiber-like structure of SHP-2 together with the actin stress fibers. Interestingly, SHP-2 remained co-localized with F-actin with a punctuated pattern in the cytochalasin D-treated cells.
The Catalytic Domain of SHP-2 Is Responsible for Its Density-regulated Cellular Distribution, but the Catalytic Activity Is Not Required-To dissect which part of the SHP-2 molecule is responsible for the distribution of SHP-2 at low cell density and whether catalytic activity is required, we expressed an SH2 domain-truncated form of the enzyme and the correspondent catalytically inactive mutant. We chose HT-1080 cells for transient expression of the truncated enzymes. As shown in Fig. 5, both ⌬SHP-2 and ⌬SHP-2M displayed a density-regulated distribution pattern similar to that of the endogenous full-length SHP-2. These results suggest that neither the SH2 domains nor the catalytic activity of SHP-2 is required for the binding of SHP-2 to actin fibers. Therefore, the interaction may be independent of tyrosine phosphorylation.
SHP-2 Binds F-actin Directly in Vitro-Co-localization of SHP-2 with actin fiber suggests an interaction of SHP-2 with F-actin. The interaction can be direct or can be mediated by other proteins. To examine the possible direct interaction, we performed in vitro binding assays by using purified GST-SHP-2, SHP-2, and ⌬SHP-2. GST was used as a control. The data shown in Fig. 6 indicate co-sedimentation of SHP-2, SHP-2, and ⌬SHP-2 with F-actin. As a negative control, no co-sedimentation of GST with F-actin was observed. These To examine the effects of cell detachment on soluble SHP-2 level, the cells were plated at the indicated densities (0.2-6 ϫ 10 6 /10-cm plate) and cultured overnight. The cells were either collected by incubating in PBS at 37°C before lysis or directly lysed on the plate (lower panel). Cell lysis and Western blotting with anti-SHP-2 were performed as described in Fig. 1.   FIG. 3. SHP-2 is partitioned in the Triton X-100-insoluble fraction at low cell densities. HeLa and HT-1080 cells were plated at the indicated cell densities and cultured overnight. The cells were extracted on plates with Buffer A, and the lysates were cleared by centrifugation at 15,000 ϫ g for 15 min at 4°C. The supernatants were collected to make SDS samples, and the pellets were washed once with Buffer A and then extracted with an equal volume of 1ϫ SDS gel sample buffer containing 1% SDS. Equivalent amounts of Triton X-100-soluble and insoluble cell extracts were separated on 10% SDS gel followed by Western blotting with anti-SHP-2. NSB denotes nonspecific protein bands.
results suggest that SHP-2 directly bind F-actin and that the interaction is mediated by the catalytic domain. However, considering the relative amounts of actin and SHP-2 used in the binding assays and the fraction of SHP-2 co-sedimented with F-actin, the binding is far from stoichiometric. There might be other factors facilitating a stronger binding in vivo.

DISCUSSION
In the present study, we have demonstrated that cell density regulates intracellular distribution of SHP-2 and that SHP-2 is co-localized with the entire actin stress fibers at low cell densities. To our knowledge, this is the first example that a tyrosine phosphatase is co-localized with F-actin fibers. By support-ing the plasma membrane of eukaryotic cells, the actin cytoskeleton plays a critical role in a number of cellular processes including cell shape, motility, chemotaxis, endocytosis, exocytosis, and cell division. Change in the rigidity of the cortical actin network constitutes an important process in the cellular response to receptor activation. Being a major component of the actin fibers when cells are at low densities and are poised to expand, SHP-2 must have a crucial role. Our data thus provided a novel mechanism by which SHP-2 regulates cell signaling and controls cell activities. The data also provided an explanation for the changes in morphology and cytoskeletal organization of cells with defective SHP-2 expression (23-25).
The actin network is regulated by a variety of actin-binding proteins. The important role of the small molecular weight FIG. 4. SHP-2 is co-localized with actin stress fibers at low cell densities. HeLa cells were plated on glass coverslips at low (0.2 ϫ 10 6 /10-cm plate) or high (6 ϫ 10 6 /10-cm plate) densities as indicated and cultured overnight. The cells were left untreated (top and middle panels) or treated with 10 g/ml cytochalasin D for 30 min at 37°C (bottom panels). Cell staining with rhodamineconjugated phalloidin (for F-actin) or anti-SHP-2 followed by fluorescein isothiocyanate-labeled secondary antibody was performed as described under "Experimental Procedures." FIG. 5. The catalytic domain of SHP-2 is responsible for its density-regulated cellular distribution but the catalytic activity is not required. HT-1080 cells (50 -60% confluency) were transfected with SH2 domain truncated ⌬SHP-2 or its Cys to Ser mutant form ⌬SHP-2M for 24 h. The cells were then trypsinized and plated at the indicated cell densities. After overnight culture, the cells were extracted with Buffer A and SDS sample buffer as described in Fig. 3. Equivalent amounts of Triton X-100-soluble and insoluble cell extracts were separated on 10% SDS gel followed by Western blotting with anti-SHP-2. The bottom panel shows the schematic structures of the SHP-2 constructs. NSP denotes nonspecific protein bands. GTPases of the Rho family has been well accepted (30,31). Involvement of tyrosine kinases in cytoskeletal control has also been well documented. Src family members regulate actin assembly and cell shape by inducing the tyrosine phosphorylation of a diversity of cytoskeletal-associated proteins including cortactin, talin, paxillin, p130 cas , WASP, and FAK (32,33). In particular, actin cross-linking activity of cortactin and its calpain-mediated proteolysis are regulated by tyrosine phosphorylation (34,35). FAK is another major tyrosine kinase involved in cytoskeletal regulation (36). It regulates cell adhesion that mediates cell and extracellular matrix interactions. FAK phosphorylates p130 cas and paxillin in response to integrinmediated adhesion, and the phosphorylation is also regulated by Rho (37). The third major tyrosine kinase that regulates cytoskeletal organization is Abl (38). Abl has a C-terminal actin binding motif, and it induces cytoskeletal abnormalities when expressed as an oncogenic BCR-Abl (38 -41). As counterparts of tyrosine kinases, PTPs have also been implicated in cytoskeletal regulation. Treatment of cells with PTP inhibitors such as phenylarsine oxide and pervanadate resulted in profound changes in cell morphology and actin distribution (42)(43)(44). Tyrosine phosphatase activity was markedly increased when actin stress fibers were disassembled by cell detachment or by cytochalasin D, a drug known to rapidly disrupt the actin cytoskeleton. When PTP inhibitors such as vanadate and phenylarsine oxide are added in combination with cytochalasin D, focal adhesions and actin stress fibers are preserved from the cytochalasin D-mediated disruption (45,46). These data suggest that PTPs play an important role in assembly of actin cytoskeleton. Many PTPs have been shown to be involved in cell adhesion and integrin signaling. These include SHP-2 (21,22), PTP1B (47), PTP-PEST (48), PTEN (49,50), PTP36 (51), SHP-1 (52), CD45 (53), LAR (54,55), PTP␣ (56,57), and PTP␤ (58). In particular, SHP-2 Ϫ/Ϫ cells displayed increased tyrosine phosphorylation of FAK accompanied by enhanced focal adhesions and impaired cell migration (21). Furthermore, we have recently shown that FAK is hyper-phosphorylated on tyrosine at low cell density and is dephosphorylated at high cell density (29). The cell density-regulated distribution of SHP-2 described in the current study may play a role in regulating the phosphorylation of FAK. Above all, co-localization of SHP-2 with entire actin stress fibers makes SHP-2 a unique player in regulating the actin network.
Our study not only revealed co-localization of SHP-2 with actin stress fiber but also demonstrated that this is regulated by cell density. Cell density has a major role in regulation of cellular activities such as cell migration and proliferation. Translocation of SHP-2 is one way that the regulation is carried out. Regulation of PTPs by cell density has been extensively studied. It has been shown that PTP activity in the membrane fractions of contact-inhibited Swiss 3T3 cells was much higher compared with proliferating cells and that cells overcome density-dependent growth inhibition in the presence of vanadate, a nonspecific PTP inhibitor (59). DEP-1, a receptor PTP, is dramatically induced when cells reach high densities (60). RPTP and RPTP, which are localized to cell-cell contacts, are also up-regulated by increasing cell density (61)(62)(63)(64). Both enzymes can mediate homophilic interaction that results in accumulation of the enzymes at cell-cell contracts. In addition, RPTP is associated with ␤-catenin and ␥-catenin/plakoglobin (62), whereas RPTP binds directly to the intracellular domain of E-cadherin and regulates its tyrosine phosphorylation (63,64). This contact-induced clustering of RPTPs presumably causes dephosphorylation of intracellular substrates at cell-cell contact regions. Up-regulated by increase in cell density, all of these PTPs may play a negative role in cell prolifer-ation. In contrast, as an intracellular enzyme, SHP-2 is unique by interacting with the entire actin stress fibers inside of the cells when cells are at low densities poised to proliferate. Association of SHP-2 with F-actin may regulate its enzymatic activity, thereby affecting cell growth. When cells reach a high density, SHP-2 becomes diffused in the cytosol where it stays at the inactive state as defined by the purified enzyme (27).
In vitro binding of SHP-2 to F-actin suggests that co-localization of SHP-2 with actin stress fiber may be mediated by direct interaction of these proteins. Because the association is found in vivo only when cells are at low densities, other proteins and factors may be involved. The presence of SHP-2 on the entire actin fiber suggests the general importance of the enzyme in regulating the actin network. The physiological substrates of SHP-2 are still to be defined. Some of the possible targets include c-Src, cortactin, and c-Abl, which are associated with actin fibers. In addition, actin itself may also be a target of SHP-2. In fact, actin is phosphorylated on tyrosine in Dictyostelium, and the phosphorylation is associated with changes of cell shapes (65). Furthermore, PTP1 in Dictyostelium cells appears to control the phosphorylation (66). Tyrosine phosphorylation of actin in mammalian cells has not been reported. This may be due to the transient nature of tyrosine phosphorylation that turns over rapidly. After all, protein tyrosine phosphorylation and dephosphorylation exist in a dynamic balance, and likewise, the cytoskeletal structure is under constant reorganization inside cells.