Reduced focal adhesion kinase and paxillin phosphorylation in BCR-ABL-transfected cells




BCR-ABL formation is critical to oncogenic transformation in chronic myelogenous leukemia and has been implicated as a key event leading to alterations in cytoskeletal structures and adhesion in the leukemic cells. The authors therefore investigated the effect of p210BCR-ABL on actin polymerization as well as on the expression and phosphorylation state of the adhesion proteins paxillin and focal adhesion kinase (FAK).


Transfection with BCR-ABL constructs abrogated the ability of NIH 3T3 fibroblasts to adhere and the cells underwent striking morphologic changes.


Scanning electron microscopy revealed that the cells lost their elongated appearance and became rounded. This alteration was associated with significantly reduced actin polymerization. In addition, steady-state levels of paxillin and FAK protein were increased. However, while the overall level of phosphotyrosines was also increased, the amount of tyrosine phosphorylated paxillin and FAK was reduced in the BCR-ABL-transfected cells as compared to the parental cells. Culture on extracellular fibronectin matrix partially reversed the morphologic changes and resulted in a return, albeit incomplete, of filamentous actin in BCR-ABL-transfected 3T3 fibroblasts. In addition, phosphorylation of paxillin and FAK in the BCR-ABL-transfected NIH 3T3 cells was restored.


The authors conclude that, in the current system, transfection of BCR-ABL attenuates FAK and paxillin phosphorylation and reduces actin polymerization, events accompanied by significant alterations in cellular morphology. The observation that exposure of the cells to fibronectin partially reverses all these changes suggests that the focal adhesion proteins and actin structures nevertheless remain responsive to singnaling from the outside. Cancer 2002;95:440–50. © 2002 American Cancer Society.

DOI 10.1002/cncr.10670

Focal adhesion is now regarded as an important route of signal transduction in cell growth and migration.1–8 In fibroblasts and most adherent cells, focal adhesion involves interactions between adhesion molecules (integrins, focal adhesion proteins like focal adhesion kinase [FAK] paxillin, talin, vinculin, tensin, and α-actinin) and cytoskeletonal proteins such as actin.9–18 This process is influenced by extracellular matrix proteins including collagen, fibronectin, and laminin.19–24 After focal adhesion through binding to ligands such as fibronectin or cross-linking to activating antibodies of adhesion molecules such as integrin α5, β1, one of the most apparent events is the phosphorylation of adhesion proteins, especially paxillin (a 68 kD protein) and FAK (a 125 kD protein).25–30 Phosphorylation of paxillin and FAK is also accompanied by cytoskeletal changes, such as actin polymerization.13

When fibroblasts are treated with cytochalasin D, which disrupts actin filaments, the cells are less adherent, the phosphorylation of paxillin and FAK is decreased, and there is less extracellular matrix production.21 Therefore, the cytoskeleton is also important in focal adhesion and signal transduction. In tumor cells, oncogenic inside-out signals can supplant the outside-in adhesion signals to maintain cell growth.2 In addition, many tumor cells produce decreased amounts of extracellular matrix,2 and their expression of adhesion molecules and proteins differs from normal cells.28, 31–35

The BCR-ABL oncogene is formed by translocation between chromosomes 9 and 22 (t(9;22)(q34;q11)). This oncogene product has a higher level of phosphotyrosine kinase activity than normal c-Abl protein and is found in most patients with chronic myelogenous leukemia (CML) and in patients with acute lymphoblastic leukemia.36–38 Progenitor CML cells show normal expression of α5 and β1 integrin but decreased adhesion to stroma in long term bone marrow culture.39, 40 This has been postulated to be due to deficient expression of the lymphocyte function antigen 3 adhesion molecule or to a defect in integrin β1 function.39, 41 Treatment of CML progenitors with interferon-α, as well as exposure of these cells to fibronectin or integrin α5- or β1-activating antibodies, can restore expression and function of these molecules, and the CML cells become more adherent and less proliferative.42, 43 Thus it has been postulated that an adhesion defect may promote premature release of immature CML cells from the bone marrow.39, 40 However this area remains controversial, since recent studies studies by Bazzoni et al. have shown increased adhesion of CML progenitors to fibronectin matrix.44 Furthermore, these investigators showed that Bcr-Abl can stimulate integrin-mediated cell adhesion by a mechanism involving increased ligand-binding.

In the current report, we show that transfection of BCR-ABL into murine fibroblasts reduces adhesion and converts filamentous actin to the punctate form. Steady state paxillin and FAK protein levels were increased in the transfected cells, but the phosphorylation of these proteins was markedly decreased. Growth of the BCR-ABL-transfected fibroblasts on a fibronectin matrix resulted in partial reversal of the morphologic changes, accompanied by an increase in actin polymerization and restoration of FAK and paxillin phosphorylation.


Cell Lines and Cell Culture

The NIH 3T3 cell line (Lewis clone 7, murine fibroblast, 4A2+) and 3T3 cells stably transfected with replication-competent Moloney viral DNA pZAP (pZAP) or with a human p210-encoding BCR-ABL-pZAP construct (Bcr-AblT1) were provided by Jean Y. J. Wang (University of California at San Diego, La Jolla, California). These cell lines were maintained in Dulbecco's modified eagle's medium (DMEM; Gibco, Grand Land, NY) with 10% defined supplemented fetal calf serum (HyClone, Logan, UT). The cell lines were incubated at 37°C in a 5% CO2 atmosphere.


Mouse monoclonal and rabbit polyclonal anti-phosphotyrosine antibodies, mouse monoclonal anti-focal adhesion kinase antibody, and mouse monoclonal anti-paxillin antibody were purchased from Transduction Laboratories (Lexington, KY). Mouse monoclonal and rabbit polyclonal anti-abl antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Rabbit preimmune serum and mouse isotypic antibodies (DAKO Corporation, Carpinteria, CA) were used as controls. Horseradish peroxidase-conjugated goat anti-mouse immunoglobin and horseradish peroxidase-conjugated donkey anti-rabbit immunoglobin were purchased from Amersham Life Science Inc. (Cleveland, OH).

Extracellular Matrix Coating

Experiments were performed to determine the effects of various extracellular matrices on cellular morphology and on actin, paxillin, and FAK expression. Human plasma fibronectin (Beckon Dickinson, Bedford, MA) was diluted in DMEM and a chamber slide or dish was coated at room temperature at a concentration of 40 μg/mL for 1 hour and then rinsed with deionized water. Cells in whole media were added and were cultured for 24–48 hours. Similar experiments were conducted with Engebreth-Holm-Swarm mouse tumor-produced laminin and matrigel (Beckon Dickinson) coatings. Polylysine (Sigma) coated slides were used as a control for the effects of nonspecific adhesion.20

Scanning Electron Microscope

Cells were washed with phosphate buffered saline (PBS) at 37°C and fixed at this temperature with modified Karnovsky's fixative (Ted Pella, Inc., Redding, CA), pH 7.5, for at least 30 minutes. The cells were then rinsed in 0.125 M sodium cacodylate buffer (Ted Pella, Inc.), pH 7.3, 310 mOsm, for three 5 minute periods and then post-fixed in 2% OsO4 in the same buffer for 30 minutes at room temperature. After three more rinses in cacodylate buffer, the cells were dehydrated in a graded series of ethanol, transferred to Peldri II (Ted Pella, Inc.) for dehydration, and placed in a vacuum desiccator for 24 hours. After sublimation of the fluorocarbon, the cells were mounted onto stubs, sputter-coated with 200 Å gold-palladium, 80:20, in a Hummer VI (Technics, Springfield, VA), and examined in a Hitachi Model S520 scanning electron microscope.

Actin Filament Staining

Cells were washed with Dulbecco's PBS after 24–48 hours of cell culture with or without extracellular matrices and fixed with 2% formaldehyde in PBS at room temperature for 30 minutes. After two washes with PBS, the cells were permeabilized with 0.5% NP-40 in PBS for 30 minutes and washed twice with PBS. This was followed by incubation with phalloidin-rhodamine (50 U/mL, Molecular Probes, Inc., Eugene, OR) for 30 minutes. The cells were then covered with 10 μL mounting medium (0.2 M Tris-HCL, pH 8.1, 2% N-propyl-gallate, 60% glycerol in water) and observed under an Olympus AH-2 microscope (Olympus Optical Co., LTD, Tokyo, Japan) with a ×100 fluorescence-free objective.

Immunoprecipitation and Western Blotting

Cells (2 × 106) were lysed with 1 mL immunoprecipitation buffer (10 mM Tris-HCL, pH 7.4, 150 mM NaCl, 1 mM ethylenediamine-tetraacetic acid, 1 mM ethylene glycol di(aminoethyl)-tetraacetic acid, 1% Triton X-100, 0.5% NP–40, 0.2 mM phenylmethyl sulfonyl fluoride, 0.5 μg/mL aprotinin, 0.2 mM sodium orthovanadate) at 4°C for 30 minutes. Lysates were centrifuged at 12,000 rpm at 4°C for 20 minutes and incubated with appropriate antibodies overnight at 4°C. Protein G PLUS/Protein A agarose beads (Oncogene Science, Manhasset, NY) were added and immunocomplexes were collected 2 hours later.

Western blotting was performed as previously described45 with some modifications. Briefly, total cell lysates or immunoprecipitates were separated by 7.5% sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis. Proteins were electrophoretically transblotted onto Immunobilon-P membrane (Millipore Co., Bedford, MA). The blot was blocked by 5% bovine serum albumin in TBST (10 mM Tris-HCL pH 7.4, 100 mM NaCl, 0.1% Tween 20) for 2 hours at room temperature and incubated with the appropriate primary antibody for 1 hour at room temperature or overnight at 4 °C. After being washed with TBST, the membranes were incubated with the appropriate second antibody for 1 hour at room temperature and visualized with an ECL chemiluminescent kit (Amersham Life Science Inc.). For reprobing the same membrane with a different antibody, the membrane was incubated in 62.5 mM Tris-HCL (pH 6.7) containing 2% SDS and 100 MM 2-mercaptoethanol at 50°C for 30 minutes to strip the previous antibodies and then probed with another appropriate antibody.


Morphologic Changes Associated with BCR-ABL Transfection are Partially Reversed by Exposure to Extracellular Matrix

The 3T3 fibroblasts transfected with BCR-ABL had significantly decreased ability to adhere as compared to their parental cells. Viewed under a scanning electron microscope, 3T3 fibroblasts transfected with BCR-ABL appeared rounded (Fig. 1C), whereas parental cells were elongated (Fig. 1A). The cell membrane also had fewer protrusions than the parental 3T3 cells. However, exposure of the BCR-ABL-transfected cells to a fibronectin-coated surface resulted in partial restoration of adherence as well as of partial morphologic changes (Fig. 1F). The cells became flattened and acquired protrusions like those of the parental cell phenotype. Laminin and matrigel matrices had effects similar to those of the fibronectin matrix, albeit not as dramatic (data not shown). The transfection procedure itself caused no changes in morphology, as shown by the observations on the 3T3 fibroblasts transfected with pZAP (Fig. 1B).

Figure 1.

Scanning electron microscopy of the parental and BCR-ABL transfected 3T3 cells. A and D), The parental cell line 4A2+ (murine 3T3 fibroblast). B and E) Line 4A2+ transfected with just pZAP construct (the replication-competent Moloney viral DNA). C and F) Bcr-AblT1, 4A2+ transfected BCR-ABL-pZAP construct. D–F) The cells were grown under identical conditions to those in A–C, except that the plates were coated with fibronectin.

Alteration in Actin Structure Associated with BCR-ABL Transfection is Partially Reversed by Exposure to Extracellular Matrix

In parental 3T3 fibroblasts, phalloidin staining of actin showed extensive actin filaments (Fig. 2A and B). However, in 3T3 cells transfected with BCR-ABL, the predominant actin pattern was of punctate structures (Fig. 2C). After culture on a fibronectin-coated surface, the transfected cells showed an increase in actin filaments (Fig. 2F). However, these actin filaments appeared to be less organized than in the parental cells. Matrigel and laminin caused similar reversion to a filamentous actin structure but to a lesser extent than fibronectin (data not shown). Growth of the cells on polylysine-coated slides (used as a control for the effects of nonspecific adherences20) had no effect on cell morphology or actin polymerization. In addition, the transfection procedure itself had no effect on actin polymerization, as shown by the observations of 3T3 fibroblasts transfected with the control pZAP alone (Fig. 2B)

Figure 2.

Actin filament staining of the parental 3T3 cells and their BCR-ABL-transfected counterparts. Cells were washed with phosphate buffered saline (PBS) after 24–48 hour cell culture with or without fibronectin coating and fixed with 2% formaldehyde in PBS at room temperature. After two washes with PBS the cells were permeabilized with 0.5% NP-40 in PBS for 30 minutes and washed twice with PBS. This was followed by incubation with phalloidin-rhodamine (50 U/mL) for 30 minutes. The cells were then covered with mounting media and observed under an Olympus AH-2 microscope with a ×100 fluorescence free objective. A–C) Actin staining of the parental 3T3 cells (4A2+) and their pZAP-transfected and BCR-ABL-pZAP-transfected conterparts. D–F) Actin staining of the parental 3T3 cells and their pZAP or BCR-ABL-pZAP transfected cells after growing in identical conditions to those in A–C except that the chamber slides were coated with fibronectin.

BCR-ABL Transfection of 3T3 Cells is Accompanied by a Decrease in Phosphorylation of FAK and Paxillin

Immunoprecipitation and Western blotting showed that total steady state protein levels of FAK and paxillin were greater in 3T3 cells transfected with BCR-ABL compared to the parental cells (Fig. 3A and B). However, reprobing these blots with antiphosphotyrosine antibodies (after washing) revealed that tyrosine phosphorylated FAK and paxillin were decreased in the transfected cells (Fig. 3C and D) despite an overall increase in tyrosine phosphorylated proteins (Fig. 3E). This was noted regardless of whether immunoprecipitation was performed with anti-FAK or anti-paxillin antibodies followed by probing the blot with anti-phosphotyrosine antibodies (Fig. 3C and D) or whether immunoprecipitation was performed with anti-phosphotyrosine antibodies followed by probing the blot with anti-FAK or anti-paxillin antibodies (Fig. 4A and B). Results from the 3T3 fibrolasts transfected with pZAP alone showed that transfection itself had no effect on steady state levels or phosphorylated levels of FAK and paxillin (Fig 3A–E).

Figure 3.

Immunoprecipitation following by Western blotting showing total focal adhesion kinase (FAK) total paxillin, phosphorylated FAK and paxillin, and total tyrosine phosphorylated proteins in parental 3T3 cells (4A2+) and their pZAP and BCR-ABL-pZAP transfected counterparts (designated pZAP and Bcr-AblT1). A) Immunoprecipitation followed by Western blotting with monoclonal anti-FAK antibody showing total FAK expression. B) Immunoprecipitation followed by Western blotting with monoclonal anti-paxillin antibody showing total paxillin expression. C and D) Reprobing the blots in A and B, respectively, after washing, with polyclonal antiphosphotyrosine antibodies to show C) phosphorylated FAK and D) phosphoryltated paxillin. Results were identical when antiphosphotyrosine antibodies were used for immunoprecipitation and probing of the blots was done with anti-FAK and anti-paxillin antibodies (see Fig. 4). Reprobing the blots with rabbit preimmune sera yielded no bands (results not shown). E) Immunoprecipitation with polyclonal anti-phosphotyrosine antibody followed by Western blotting with monoclonal anti-phosphotyrosine antibody showing total phosphotyrosine proteins.

Figure 4.

Immunoprecipitation and Western blotting showing phosphorylated paxillin and focal adhesion kinase (FAK) expression in parental 3T3 cells (4A2+) and their pZAP and BCR-ABL-pZAP-transfected counterparts (designated pZAP and Bcr-AblT1) with and without fibronectin. Immunoprecipitation with polyclonal antiphosphotyrosine antibodies followed by Western blotting with A) monoclonal anti-paxillin or C) monoclonal anti-FAK antibodies with and without fibronectin. B and D) Total paxillin and FAK levels without fibronectin (lanes 1–3) and after growing the cells on fibronectin (lanes 4–6). The results for total paxillin and total FAK levels were obtained by immunoprecipitation followed by Western blotting with monoclonal anti-paxillin or anti-FAK antibody, respectively. Similar to the results in Figure 3, phosphorylated paxillin and FAK levels were decreased in the Bcr-AblT1 transfected cells despite an increase in total FAK and paxillin; similarly, after exposure to fibronectin, the Bcr-AblT1-transfected cells also had increased levels of total paxillin and FAK levels. However, the phosphorylated levels of paxillin and FAK were considerably higher in the Bcr-AblT1 transfected cells exposed to fibronectin than in the Bcr-AblT1 transfected cells not exposed to fibronectin and approached the levels seen in the absence of Bcr-AblT1. These results suggest that exposure to fibronectin partially reverses suppressed phosphorylation of paxillin and FAK in Bcr-AblT1 transfected cells.

Fibronectin Causes Partial Reversal of Decreased FAK and Paxillin Phosphorylation in 3T3 Cells Transfected with BCR-ABL

To determine if reversal of the decreased FAK and paxillin phosphorylation accompanies the partial reversal of morphologic changes associated with growth on extracellular matrices, we examined the phosphorylation state of FAK and paxillin after growing the cells on a fibronectin-coated surface for 24–48 hours. On exposure to fibronectin, the phosphorylation of FAK (Fig. 4A and C) and paxillin (Fig. 4B and D) in the transfected cells was restored to levels similar to those seen in untransfected cells.


Focal adhesion has been studied extensively in recent years. Adhesion varies between cell types,2 and several growth factors, mitogens, and drugs can also cause changes in focal adhesion and its related proteins.46–54 In normal cells such as fibroblasts, cell-surface integrin receptor occupancy and clustering through ligands or antibodies trigger a synergistic response that includes the reorganization of cytoskeletal and associated cytoplasmic plaque proteins and the activation of local signaling pathways.55 These pathways involve many kinases such as FAK, Src family kinases, Rho family GTPases, PI-kinases, and biochemical processes, including phosphorylation of paxillin, tensin, and inositol-lipid metabolism.55 Adhesion of fibroblasts to fibronectin also leads to the activation of the Ras/MAP kinase pathway, which may then activate FAK through the GRB2 adapter protein and SOS protein.55 Tumor gene expressing cells have adhesion properties and related adhesion proteins that differ from normal cells, and tumor gene products can replace the outside-in cell adhesion signals to maintain cell growth.3–8, 44–49, 56–58

We studied the effect of BCR-ABL transfection on multiple adhesion-related properties. The parental cells grown for 24–48 hours without an extracellular matrix coating were flattened, and actin polymerization was apparent (Fig. 1A and B, and Fig. 2A and B). However, these cells lost their adhesive properties after transfection with BCR-ABL, consistent with the results by Renshaw et al.45 The transfected cells also became rounded, and the morphology of the surface of the transfected cells was different from that of the parental cells (Fig. 1A–C). In addition, actin filaments in the transfected cells dramatically decreased compared with those in the parental cells (Fig. 2A–C). These results are consistent with the observations of Mcwhirter and Wang,59 suggesting that Bcr-Abl protein binds actin and causes actin depolymerization. After the parental and transfected cells were grown on a fibronectin-coated surface for 24–48 hours, the transfected cells became flattened, with an increase in actin filaments (Fig 2F). Thus the cells transfected with BCR-ABL still responded to extracellular outside-in signals, perhaps because β1 integrin expression is not decreased in the BCR-ABL-transfected cells (data not shown). (Fibroblasts and myeloid progenitors express the heterodimeric adhesion molecule integrin α5β1, which is the receptor for the extracellular matrix protein fibronectin.1, 11, 30 This ligand binding signal may still trigger normal signal transduction pathways, causing changes in the cytoskeleton and morphology, which may partially reverse the reduced actin polymerization caused by BCR-ABL expression. Interestingly, in SRC-transfected chicken embryo fibroblasts, there is a decrease in fibronectin deposition outside of the cells, accompanied by abrogation of cell adhesion.2 We are currently investigating whether a similar reduction of fibronectin deposition occurs in BCR-ABL-transfected cells, which could explain the partial restoration of adhesion that is seen when fibronectin is added to culture conditions.

We also found that two commonly studied focal adhesion proteins, FAK and paxillin, were phosphorylated in our NIH 3T3 cells. However, in the cells transfected with BCR-ABL, FAK and paxillin phosphorylation were decreased, although total amounts of phosphotyrosine proteins and total FAK and paxillin protein levels were higher than in the parental cells (Fig. 3A–E). The decreased phosphorylation of FAK and paxillin may be related to the expression of Bcr-Abl phosphotyrosine kinase and its resultant actin depolymerization.59 Indeed, it has been previously shown that disruption of actin filaments by cytochalasin D results in decreased phosphorylation of FAK and paxillin.21 However, in contrast to these results, in chicken embryo fibroblasts transfected with pp60SRC, actin depolymerization is accompanied by phosphorylation of FAK and paxillin.10, 60 In addition, investigators have shown that in the Bcr-Abl-transfected myeloid cell line 32D, Bcr-Abl may link paxillin through other adaptor proteins such as Crkl, and this causes paxillin phosphorylation.57, 61, 62 Therefore, the interaction between actin, adhesion molecules, and phosphorylated oncoproteins may be distinct in different systems or under variable conditions.

As mentioned above, growth of the Bcr-Abl transfected NIH 3T3 fibroblasts on fibronectin matrix caused partial restoration of morphology (Fig. 1), adhesion, and actin repolymerization (Fig. 2). These changes were accompanied by restoration of FAK and paxillin phosphorylation (Fig. 4). Since exposure of normal fibroblasts to fibronectin has previously been shown to phosphorylate FAK and paxillin,12, 18 our results indicate that this outside-in signalling pathway remains intact in BCR-ABL-transfected cells.

In summary, in NIH 3T3 fibroblasts, Bcr-Abl expression was associated with a reduction in cell adhesion and conversion of filamentous actin to the punctate form. These alterations were accompanied by striking morphologic changes, i.e., the elongated cells become rounded. Additionally, an increase in levels of total phosphotyrosines and in FAK and paxillin steady state protein levels, but a decrease in FAK and paxillin phosphorylation, was observed. The latter alterations may be secondary to actin depolymerization and the lack of adherence after Bcr-Abl expression.45, 59 Analogous results have recently been reported by other investigators, who showed that C3G (a guanine nucleotide exchange factor) is phosphorylated in response to cell adhesion to fibronectin and that this pathway can be disrupted by Bcr-Abl.63 Furthermore, the current results are consistent with prior data showing that abnormal circulation and unregulated proliferation of CML progenitors is related, at least in part, to Bcr-Abl-induced abnormalities in association with beta1 integrin receptors, which results in cytoskeletal abnormalities.64, 65 However, our experiments also showed that fibronectin exposure partially reversed the Bcr-Abl-related morphologic changes, promoted actin repolymerization, and restored FAK and paxillin phosphorylation in BCR-ABL-transfected fibroblasts. Taken together, these observations indicate that Bcr-Abl induces marked alterations in normal focal adhesion/actin-based pathways, but that these pathways remain at least partially responsive to outside-in signals.