Vav3 mediates receptor protein tyrosine kinase signaling, regulates GTPase activity, modulates cell morphology, and induces cell transformation

Abstract

A recently reported new member of the Vav family proteins, Vav3 has been identified as a Ros receptor protein tyrosine kinase (RPTK) interacting protein by yeast two-hybrid screening. Northern analysis shows that Vav3 has a broad tissue expression profile that is distinct from those of Vav and Vav2. Two species of Vav3 transcripts, 3.4 and 5.4 kb, were detected with a differential expression pattern in various tissues. Transient expression of Vav in 293T and NIH 3T3 cells demonstrated that ligand stimulation of several RPTKs (epidermal growth factor receptor [EGFR], Ros, insulin receptor [IR], and insulin-like growth factor I receptor [IGFR]) led to tyrosine phosphorylation of Vav3 and its association with the receptors as well as their downstream signaling molecules, including Shc, Grb2, phospholipase C (PLC-gamma), and phosphatidylinositol 3 kinase. In vitro binding assays using glutathione S-transferase-fusion polypeptides containing the GTPase-binding domains of Rok-alpha, Pak, or Ack revealed that overexpression of Vav3 in NIH 3T3 cells resulted in the activation of Rac-1 and Cdc42 whereas a deletion mutant lacking the N-terminal calponin homology and acidic region domains activated RhoA and Rac-1 but lost the ability to activate Cdc42. Vav3 induced marked membrane ruffles and microspikes in NIH 3T3 cells, while the N-terminal truncation mutants of Vav3 significantly enhanced membrane ruffle formation but had a reduced ability to induce microspikes. Activation of IR further enhanced the ability of Vav3 to induce membrane ruffles, but IGFR activation specifically promoted Vav3-mediated microspike formation. N-terminal truncation of Vav3 activated its transforming potential, as measured by focus-formation assays. We conclude that Vav3 mediates RPTK signaling and regulates GTPase activity, its native and mutant forms are able to modulate cell morphology, and it has the potential to induce cell transformation.

FIG. 1

Schematic diagram of the Vav3 constructs used in biochemical and biological assays.

FIG. 2

Northern blot analysis of Vav3 transcripts. A PstI-PstI fragment of Vav3 which included the N-terminal SH3 and SH2 domains was labeled with [α-³²P]ATP and used as the probe. Human Multiple Tissue Northern (MTN) Blots (Clontech) (samples 7759-1 and 7760-1) were prehybridized and hybridized in Express Hyb Hybridization Solution at 68°C as described in the manufacturer's manual. Human β-actin cDNA (Clontech) was used as a control probe for monitoring the quantities of mRNAs.

FIG. 3

Vav3 is tyrosine phosphorylated by Ros and EGFR and interacts with these receptors. (A) Samples of 100 ng of ER2 and 10 μg of Vav3SH were cotransfected into 293T cells as indicated. After serum starvation and treatment for 15 min with 100 ng of EGF/ml, cells were lysed and subjected to immunoprecipitations (IP) and Western blotting (IB). EB69 cells (B) and 293T cells (C) were transfected with either empty vector pHEF neo or Vav3F. Vav3 was immunoprecipitated with anti-Ros or anti-EGFR antibodies. The tyrosine phosphorylation of Vav3 was visualized by immunoprecipitation with anti-Vav3 antibody and Western blotting with anti-phosphotyrosine antibody.

FIG. 4

Vav3 is tyrosine phosphorylated and associates with the IR upon its activation. 293T cells were transfected with 10 μg of Vav3SH (A) or 10 μg of Vav3F (B) in the absence or presence of exogenously transfected IR (100 ng). A sample of 1 mg of total-cell lysate was used in the immunoprecipitations (IP) and Western blottings (IB) as indicated.

FIG. 5

Vav3 becomes tyrosine phosphorylated and interacts with the IGF-1 receptor upon IGF-1 stimulation. Samples of 100 ng of pHEF IGFR and 10 μg of Vav3SH (A) or 10 μg of Vav3F, Vav3 5-10, or Vav3 6-10 (B) were used in cotransfections of 293T cells. Vav3 and IGFR were immunoprecipitated (IP) from 1 mg of total-cell lysate and subjected to Western blotting (IB) as indicated.

FIG. 6

Vav3 associates with PLC-γ and PI3 kinase p85 in ER2-expressing NIH 3T3 cells (EB69). EB69 cells were transfected with either Vav3F (10 μg) or pHEF neo (10 μg) as a control. At 48 h posttransfection, cells were stimulated with EGF (100 ng/ml) for 15 min and lysed. A sample of 1 mg of total-cell lysate was used in immunoprecipitations (IP) and Western blottings as indicated.

FIG. 7

Vav3 interacts with Shc and Grb-2 in IR-overexpressing 293T cells. Cotransfections were performed with 293T cells as indicated. A sample of 1 mg of lysate with or without insulin stimulation was immunoprecipitated (IP) with anti-Grb2 or anti-Shc antibody, followed by Western blotting (IB) with anti-Vav3 antibody. The tyrosine phosphorylation level of IR and Vav3 as well as the protein amounts of IR, Vav3, Shc, and Grb2 were determined by immunoprecipitation and Western blotting with antibodies against phosphotyrosine and these proteins, respectively.

FIG. 8

Activation of GTPases by Vav3 and its mutant 6-10. Vav3F (1.5 μg) and pHEF 6-10 (1.5 μg) were cotransfected with HA-tagged wild-type RhoA (A), Rac-1 (B), or Cdc42 (C) (0.5 μg each) into NIH 3T3 cells. Then, 36 h later, cells were serum starved overnight and extracted in NP-40 lysis buffer. A sample of 500 μg of total-cell lysate was used in Rok-, Pak-, and Ack-GST binding assays as described in Materials and Methods. The data from three independent experiments were normalized for protein expression. The average fold activation (Fold Act) of the respective GTPase and the standard error (SE) were calculated and are presented along with a representative blot. Ca, constitutively activated.

FIG. 9

Insulin enhances Vav3-mediated Rac-1 activation. Empty vector or Vav3F was cotransfected with HA-tagged wild-type RhoA (A), Rac-1 (B), or Cdc42 (C) into NIH 3T3 cells overexpressing IR. After cells were treated with 50 nM insulin for 15 min, they were extracted using NP-40 buffer and subjected to Rok-, Pak-, and Ack-GST pull-down assays. The data from two independent experiments were normalized for protein expression. The average fold activation (Fold ACT) of the respective GTPase was calculated and is presented along with a representative blot.

FIG. 10

IGF-1 enhances Vav3-mediated Cdc42 activation. Empty vector or Vav3F was cotransfected with HA-tagged wild-type RhoA (A), Rac-1 (B), or Cdc42 (C) into NIH 3T3 cells overexpressing IGFR. Cells were treated with 100 ng of IGF-1/ml for 15 min and extracted using NP-40 buffer followed by Rok-, Pak-, and Ack-GST pull down assays. The data from two independent experiments were normalized for protein expression. The average fold activation (Fold Act) of the respective GTPase was calculated and is presented along with a representative blot.

FIG. 11

Effect of Vav3 and its mutants on the morphology of NIH 3T3 cells. NIH 3T3 cells transfected with 2 μg of pEGFP C1 (clontech) (A), Vav3SH (B and C), Vav3F (D to F), pHEF5-10 (G to I), or pHEF6-10 (J to L) were fixed and visualized by staining with anti-Vav3 antibody and subsequently anti-rabbit secondary antibody coupled to fluorescein isothiocyanate (FITC). The proportion of positive cells with distinct morphology in the absence of serum is under the corresponding panels.

FIG. 12

N-terminal deletion mutants of Vav3-induced foci in NIH 3T3 cells. (A) The morphology of foci induced by Vav3F (III), pHEF 5-10 (IV), and pHEF 6-10 (V). NIH 3T3 cells transfected with empty vector (I), Vav3F (II), or RasV12 (VI) were included as controls and for comparison of transforming morphology. (B) The number of foci induced by full-length Vav3 and its mutants, 5-10 and 6-10.