Protein complex immunological separation assay (ProCISA): a technique for investigating single protein properties.

Analysis of the posttranslational modification of proteins, such as phosphorylation, might yield misleading results due to the presence of other proteins with similar electrophoretic properties that coimmunoprecipitate with the target protein. The aim of the present work was to develop a reliable, easy and economical technique to completely isolate a protein from its complex. Here we present a new assay developed to fully isolate proteins from macromolecular complexes that consists of an initial SDS/PAGE (under reducing conditions), which isolates the target protein, followed by transfer of the proteins to a buffer, from which the target protein is recaptured by conventional immunoprecipitation. This technique, that we have termed "Protein Complex Immunological Separation Assay" (ProCISA), successfully separated proteins of different sizes, such as pp60Src and the IP3 receptor (IP3R), from their complexes. We show that ProCISA allows the investigation of the tyrosine phosphorylation state of isolated proteins. This technique could also be used to study other posttranslational modifications without risk of misleading results resulting from contamination with other proteins of similar electrophoretic mobility which complex with the protein of interest.

Expression, purification, and bioactivity of GST-fused v-Src from a bacterial expression system.

v-Src is a non-receptor protein tyrosine kinase involved in many signal transduction pathways and closely related to the activation and development of cancers. We present here the expression, purification, and bioactivity of a GST (glutathione S-transferase)-fused v-Src from a bacterial expression system. Different culture conditions were examined in an isopropyl beta-D-thiogalactopyranoside (IPTG)-regulated expression, and the fused protein was purified using GSH (glutathione) affinity chromatography. ELISA (enzyme-linked immunosorbent assay) was employed to determine the phosphorylation kinase activity of the GST-fused v-Src. This strategy seems to be more promising than the insect cell system or other eukaryotic systems employed in earlier Src expression.

Assaying protein kinase activity.

Protein kinases, encoded by approx 2% of eukaryotic genes, represent one of the major classes of cell-regulatory molecules. Assessment of the catalytic activity of a specific protein kinase can be an important step in elucidating signal-transduction pathways that affect cell behavior. As an example of approaches taken to measure protein kinase activity, this chapter presents methods useful for determination of the activity of the oncogenic protein-tyrosine kinase v-Src. Included are protocols for heterologous expression of the kinase in yeast Saccharomyces cerevisiae, immunoaffinity purification from yeast cell lysates, kinase reactions using incorporation of 32P into peptide substrates, and quantifying protein kinase activity. The Notes section discusses alternative approaches for assaying the activity of Src recovered from vertebrate cells and it gives recommendations for assaying the activity of the other protein kinases with respect to the substrate specifity and the composition of kinase reaction buffer.

Expression of protein tyrosine kinase pp60v-src variants in Dictyostelium discoideum.

We achieved production of v-Src of the low-oncogenic PRC and its variant proviral structure H19 in Dictyostelium discoideum, an emerging host system suitable for synthesis of heterologous proteins. To accomplish their expression, the first six codons of the N-terminus of v-src had to be changed according to the D. discoideum codon preference. Alternatively, N-terminal fusions of 6xHis-tag or GFP were sufficient to overcome the incompatibility in codon usage. D. discoideum-expressed v-Src kinases of the expected molecular weight were recognized by Src-specific antibodies; GFP-PRC was distributed uniformly in the cytosol. In contrast to other lower eukaryotes, where the accumulation of v-Src leads to growth inhibition, D. discoideum cells silenced the kinase activity of PRC-derived v-Src and showed no developmental or growth defects.

Capillary electrochromatography/nanoelectrospray mass spectrometry for attomole characterization of peptides.

The successful coupling of capillary electrochromatography (CEC) to an ion trap mass spectrometer via a nanoelectrospray interface (nESI) is described. Using a conductively coated tip butted to the end of a CEC column, it was possible to obtain a stable spray without any sheath liquid being employed. Selected small peptides were separated with CEC columns (100 microm i.d./25 cm long) packed with 3 microm Hypersil C8 or C18 bonded silica particles with an eluent composed of ammonium acetate/acetonitrile. Peptide mixtures of desmopressin, peptide A, oxytocin, carbetocin and [Met(5)]-enkephalin were detected in the mid-attomole range, which is the lowest amount analyzed using CEC combined with MS detection. It was also observed that sensitivity can be compromised at higher separation voltages. We demonstrate that CEC/nESI-MS, at the current stage of development, represents one of the most sensitive systems for peptide analysis.

pp60v-src phosphorylates and activates low molecular weight phosphotyrosine-protein phosphatase.

Low M(r) phosphotyrosine-protein phosphatase belongs to the non-receptor cytosolic phosphotyrosine-protein phosphatase subfamily. It has been demonstrated that this enzyme dephosphorylates receptor tyrosine kinases, namely the epidermal growth factor receptor in vitro and the platelet-derived growth factor receptor in vivo. Low M(r) phosphotyrosine-protein phosphatase is constitutively tyrosine-phosphorylated in NIH/3T3 cells transformed by pp60v-src. The same tyrosine kinase, previously immunoprecipitated, phosphorylates this enzyme in vitro as well. Phosphorylation is enhanced using phosphatase inhibitors and phenylarsine oxide-inactivated phosphatase, consistently with the existence of an auto-dephosphorylation process. Intermolecular dephosphorylation is demonstrated adding the active enzyme in a solution containing the inactivated and previously phosphorylated one. This tyrosine phosphorylation correlates with an increase in catalytic activity. Our results provide evidence of a physiological mechanism of low M(r) phosphotyrosine-protein phosphatase activity regulation.

p80/85 cortactin associates with the Src SH2 domain and colocalizes with v-Src in transformed cells.

Expression of oncogenic variants of pp60src leads to dramatic changes in cytoskeletal organization characteristic of transformation. Activated Src associates with the cytoskeletal matrix, resulting in tyrosine phosphorylation of specific cytoskeletal substrates. We have previously shown that stable association of Src with the cytoskeletal matrix is mediated by the Src SH2 domain in a phosphotyrosine-dependent interaction. In this report, we demonstrate that one of the cytoskeletal binding partners of Src is p80/85 cortactin. The association was observed in lysates of transformed cells but was not seen in normal fibroblasts. The interaction could be reconstituted in vitro using transformed cell extracts and a glutathione S-transferase (GST) fusion protein containing the Src SH2 domain but not with GST-Src SH3 or with GST-Src SH2 containing a point mutation in the FLVRES sequence. Confocal microscopy revealed that cortactin redistributed and colocalized with v-Src and a Src SH3 deletion mutant in transformed cells. However, in cells expressing a Src SH2 deletion mutant, the redistribution of cortactin and colocalization with Src did not occur. Furthermore, biochemical fractionation of transformed cells indicated that a significant increase in cortactin distribution to the cytoskeletal fraction occurred, which correlated with a shift in the tyrosine-phosphorylated form of the protein. Cortactin fractionated from cells expressing kinase-defective or myristylation-defective Src mutants did not exhibit this shift. These data suggest a molecular mechanism by which tyrosine phosphorylation of cortactin and association with the Src SH2 domain influence the cytoskeletal reorganization induced in Src-transformed cells.

The E-cadherin complex contains the src substrate p120.

Using normal MDCK cells, and MDCK cells stably transfected with a temperature-sensitive viral src allele (pp60 ts-v-src), we have examined the composition and tyrosine phosphorylation of the E-cadherin complex. E-cadherin is a transmembrane calcium-dependent cell-cell adhesion molecule that is complexed with cytoplasmic proteins including alpha-catenin, beta-catenin, plakoglobin (gamma-catenin), and actin. We have identified two heterodimeric complexes which demonstrate that alpha-catenin interacts directly with beta-catenin, or with plakoglobin, in the absence of E-cadherin. beta-Catenin has previously been shown to bind directly to E-cadherin. We propose that E-cadherin associates with alpha-catenin, and thereby the actin cytoskeleton, via either beta-catenin or plakoglobin. We have further identified three new but related protein components of the E-cadherin complex, which are each cross-reactive by Western blot analysis to antibodies directed against p120, a phosphotyrosine substrate of src, and a phosphotyrosine, phosphoserine, and phosphothreonine substrate of growth factor-stimulated signaling pathways. Greater quantities of the p120-related proteins were found present in the E-cadherin immunoprecipitates of ts-src MDCK cells compared to normal MDCK cells, while two of the p120 cross-reactive species were significantly tyrosine phosphorylated in both normal and ts-src MDCK cells. The association of p120-related species with the E-cadherin complex adds them to our consideration of possible modulators of cadherin function. Likewise, the newly identified alpha-catenin-beta-catenin and alpha-catenin-plakoglobin dimers may have interesting biological properties, conceivably including the titration of catenins between cadherin and APC complexes.

Microtubule reorganization and lysosome redistribution by a viral v-src oncogene, in mouse Balb/3T3 cells expressing human EGF receptor.

The epidermal growth factor (EGF)-induced endocytosis of its receptor is an obligatory pathway for the cellular regulation of the EGF-specific receptor (EGF-R). BNER4 is a mouse Balb/3T3 cell line transfected with human EGF-R complementary DNA (cDNA). B4/src-13 and B4/src-24 are BNER4 cells transfected with a viral oncogene v-src. Indirect immunofluorescence study demonstrated that EGF-R was mostly localized at the perinuclear region in BNER4 cells at 60 min after EGF addition, whereas it was diffusely distributed throughout the cytoplasm in its v-src transfectants. Double indirect immunofluorescence study further confirmed that EGF-R was localized in lysosomes in BNER4 and B4/src-13 cells at 60 min after EGF addition. Intracellular distribution of the Golgi apparatus, clathrin-coated vesicles and early endosomes were similar in all cell lines. However, the lysosomes detected by anti-lysosomal membrane protein (LGP85) antibodies were diffusely distributed throughout the cytoplasm in the v-src transfectacts. By contrast, in the parental BNER4 cells, the lysosomes were mostly localized in the perinuclear region. The organization of microtubules, but not of actin, was markedly different between BNER4 cells and its v-src transfectants. Nocodazole, which depolymerizes microtubules, altered the distribution of the lysosomes and EGF-R in BNER4 cells. Both intracellular lysosome distribution and microtubule organization in nocodazole-treated BNER4 cells were found to be similar to those in its v-src transfectants without nocodazole treatment. These findings support the notion that changes in lysosome distribution may be correlated with microtubule reorganization by v-src in mouse Balb/3T3 cells.

Phosphopeptide occupancy and photoaffinity cross-linking of the v-Src SH2 domain attenuates tyrosine kinase activity.

Phosphorylation of c-Src at carboxyl-terminal Tyr-527 suppresses tyrosine kinase activity and transforming potential, presumably by facilitating the intramolecular interaction of the C terminus of Src with its SH2 domain. In addition, it has been shown previously that occupancy of the c-Src SH2 domain with a phosphopeptide stimulates c-Src kinase catalytic activity. We have performed analogous studies with v-Src, the transforming protein from Rous sarcoma virus, which has extensive homology with c-Src. v-Src lacks an autoregulatory phosphorylation site, and its kinase domain is constitutively active. Phosphopeptides corresponding to the sequences surrounding c-Src Tyr-527 and a Tyr-Glu-Glu-Ile motif from the hamster polyoma virus middle T antigen inhibit tyrosine kinase activity of baculovirus-expressed v-Src 2- and 4-fold, respectively. To determine the mechanism of this regulation, the Tyr-527 phosphopeptide was substituted with the photoactive amino acid p-benzoylphenylalanine at the adjacent positions (N- and C-terminal) to phosphotyrosine. These peptides photoinactivate the v-Src tyrosine kinase 5-fold in a time- and concentration-dependent manner. Furthermore, the peptides cross-link an isolated Src SH2 domain with similar rates and specificity. These data indicate that occupancy of the v-Src SH2 domain induces a conformational change that is transmitted to the kinase domain and attenuates tyrosine kinase activity.

The marine natural product, halistanol trisulfate, inhibits pp60v-src protein tyrosine kinase activity.

Halistanol trisulfate, a sulfated steroid derivative, was isolated from the extracts of two different marine sponges (genus Topsentia) by bioassay-guided fractionation. It exhibited an IC50 of approximately 4 microM against pp60v-src, the oncogenic protein tyrosine kinase encoded by Rous sarcoma virus. Removing the sulfate groups by acid hydrolysis produced the inactive tris-alcohol, halistanol. The kinetics of inhibition were examined and revealed that halistanol trisulfate is a competitive inhibitor with respect to the peptide substrate, [val5]-angiotensin II, and a mixed inhibitor with respect to ATP. A number of monosulfated steroids were studied for protein tyrosine kinase inhibitory activity, but were found to be inactive.

Functional role of GTPase-activating protein in cell transformation by pp60v-src.

Morphological transformation of NIH 3T3 cells was observed following coexpression of a portion of the ras GTPase-activating protein (GAP) comprising the amino terminus (GAP-N) and a mutant of v-src (MDSRC) lacking the membrane-localizing sequence. Cells expressing either of these genes alone remained nontransformed. Coexpression of GAP-N with MDSRC did not alter the subcellular localization, kinase activity, or pattern of cellular substrates phosphorylated by the MDSRC product. In contrast to SHC, phospholipase C-gamma 1, and the p85 alpha phosphatidylinositol 3'-kinase subunit, the endogenous GAP product (p120GAP) was highly tyrosine-phosphorylated only in cells transformed by wild-type v-src. Furthermore, for transformation induced by wild-type v-src as well as by coexpression of MDSRC and GAP-N, a strict correlation was observed between cell transformation, elevated tyrosine phosphorylation of p62, p190, and a novel protein of 150 kDa, and complex formation between these proteins and p120GAP. As with cells transformed by wild-type v-src, the MDSRC plus GAP-N transformants remained dependent on endogenous Ras. The results suggest that tyrosine phosphorylation and complex formation involving p120GAP represent critical elements of cell transformation by v-src and that complementation of the cytosolic v-src mutant by GAP-N results, at least in part, from the formation of these complexes.

Deletion of the SH3 domain of Src interferes with regulation by the phosphorylated carboxyl-terminal tyrosine.

A current model for the regulation of the Src protein-tyrosine kinase proposes that the COOH-terminal phosphotyrosine, Tyr-527, binds to the Src homology 2 (SH2) region in an intramolecular interaction that represses the kinase domain. This model is consistent with the activation of Src by mutations in the SH2 domain or COOH terminus. Mutations in the SH3 domain also activate Src, although this region is not thought to bind phosphotyrosine. Seidel-Dugan et al. (Seidel-Dugan, C., Meyer, B. E., Thomas, S. M., and Brugge, J. S. (1992) Mol. Cell. Biol. 12, 1835-1845) have shown that Src mutants with deletions in the SH2 or SH3 domain transform chicken embryo fibroblasts and have increased kinase activity. These mutant proteins are underphosphorylated at Tyr-527, a change that could in itself activate the mutants. Therefore, it is not possible to distinguish whether the SH2 and SH3 domains are needed for phosphorylation of Tyr-527 or for Src to adopt or maintain the repressed state. We have artificially increased the level of Tyr-527 phosphorylation of SH2 and SH3 deletion mutants by coexpressing them with the Tyr-527 kinase, Csk, in yeast cells. We find that both the SH2 and SH3 domains are needed for inhibition of Src by Csk. The SH2 domain is needed for efficient phosphorylation by Csk, both in yeast cells and in vitro. The SH3 domain is needed for Src to be inhibited when Tyr-527 is phosphorylated by Csk. This suggests that the SH3 domain cooperates with the SH2 domain and phosphorylated Tyr-527 to inhibit the kinase domain. Dephosphorylation of SH3 domain mutants at Tyr-527 in fibroblasts could be a consequence of a failure of the proposed SH2/phosphotyrosine interaction.

Detection of Src homology 3-binding proteins, including paxillin, in normal and v-Src-transformed Balb/c 3T3 cells.

The Src homology 3 (SH3) domain, located in the amino-terminal, noncatalytic half of pp60src, is highly conserved among members of the Src family of tyrosine kinases. SH3 domains have also been identified in a variety of proteins otherwise unrelated to protein-tyrosine kinases. The presence of SH3 domains in proteins with diverse functions suggests this domain may be important for directing protein-protein interactions necessary for protein function or cellular localization. To explore possible interactions between the SH3 domain and cellular proteins, we have established conditions for the isolation of proteins that bind in solution to the Src SH3 domain. A 67-amino acid fragment of c-Src containing either the entire glutathione S-transferase-SH3 domain (GST-SH3) or the SH3 domain from the neuronal form of c-Src (GST-SH3+) was expressed as a glutathione S-transferase fusion protein. The GST fusion proteins were incubated with lysates from [35S]methionine-labeled Balb/c 3T3 cells or v-Src-transformed Balb/c 3T3 cells. We found that GST-SH3, but not wild-type GST, specifically interacted with multiple cellular proteins, whereas GST-SH3+ only weakly associated with a small subset of these proteins. The majority of the SH3-binding proteins were found in particulate and detergent-insoluble cell fractions. Anti-phosphotyrosine immunoblots of the SH3-binding proteins revealed that several of the SH3-binding proteins are phosphorylated on tyrosine in v-Src-transformed cells. In addition, a number of the SH3-binding proteins were phosphorylated on serine and/or threonine in in vitro kinase assays, suggesting that one or more of the SH3-binding proteins has kinase activity. We identified paxillin, a vinculin-binding protein, as one of the Src SH3-binding proteins. This finding strongly supports the hypothesis that SH3 domains may be involved in subcellular localization of proteins to cytoskeleton and/or cellular membranes.

Phospholipase C-gamma 1 associates with viral and cellular src kinases.

A tyrosine kinase inhibitor, genistein, caused the subcellular translocation of phosphoinositide-specific phospholipase C (PLC) activity from membrane fractions to cytosolic fractions in rat 3Y1 fibroblasts and their transformants by Rous sarcoma virus, SR-3Y1. The ratio of PLC activities associated with the membrane fractions to those of the homogenate fractions was greater in SR-3Y1 (32.6%) than in 3Y1 (20.8%) whereas membrane-associated PLC activities were strikingly reduced to the same levels in both cells by treatment with genistein. Moreover, it was found by immunoblotting analyses of membrane fractions that the amounts of PLC-gamma 1 isozyme were reduced to 20.4% of initial level in SR-3Y1 and to 30.2% of that in 3Y1 cells. While the levels of PLC-delta, another detectable PLC isozyme, were not altered by genistein suggesting that tyrosine kinase plays an important role in the association of PLC-gamma 1 with membranes. PLC-gamma 1 molecules were detected in anti-p60arc antibody immunoprecipitates of both 3Y1 and SR-3Y1 cells. The amounts of PLC-gamma 1 co-immunoprecipitating with src kinases were higher in SR-3Y1 than 3Y1 cells and were reduced in both cell types by treatment with genistein. In addition, it was confirmed that PLC-gamma 1 purified from rat liver was phosphorylated at a tyrosine residue and associated with viral src kinase and that src kinases associated with the recombinant SH2 region of PLC-gamma 1, expressed in Escherichia coli, depending upon phosphorylation of tyrosine residues. These findings suggest that both viral and cellular src kinases associate with PLC-gamma 1 and may mediate cellular signaling in normal and transformant cell growth.

Evidence for autoinhibitory regulation of the c-src gene product. A possible interaction between the src homology 2 domain and autophosphorylation site.

In the previous study (Sato, K., Miki, S., Tachibana, H., Hayashi, F., Akiyama, T., and Fukami, Y. (1990) Biochem. Biophys. Res. Commun. 171, 1152-1159), we found a synthetic peptide, termed peptide A, that inhibited the kinase activity of p60v-src. The peptide A sequence corresponds to residues 137 to 157 of p60v-src which are included in the amino-terminal portion of the src homology 2 domain. In this study, we attempted to specify the inhibitory sequence in this domain and to identify its target site. The most potent peptide A derivative was one that corresponds to residues 140 through 157. The target site of peptide A was assumed to reside in the autophosphorylation site of p60v-src, since synthetic peptides containing the sequence Phe424-Pro-Ile-Lys-Trp428 which is present downstream of the autophosphorylated Tyr416 partially counteracted the inhibitory effect of peptide A. An antibody was prepared against one of such target peptides, termed pepY. Cross-linking experiments showed that 125I-labeled peptide A could bind to p60v-src blotted on a membrane, and the binding was blocked by the anti-pepY antibody but not by other anti-p60v-src antibodies. Conversely, immunoblotting of p60v-src with anti-pepY antibody was blocked by the cross-linking of peptide A to p60v-src. To our surprise, anti-pepY antibody did not affect the p60v-src activity. Furthermore, p60c-src was activated 2- to 6-fold by this antibody. These results suggest that the pepY region in the catalytic domain of p60v-src or of p60c-src is not essential for the catalytic activity but rather is involved in the negative regulation of the kinase activity of p60c-src.

Comparison of src-family cDNAs reveals distinct mechanisms underlying focus formation in transfected fibroblasts.

Despite the intensive study of both cellular transformation and src-family protein-tyrosine kinases, there have been no direct comparisons of transforming potency for normal members of this gene-family. In this study, the focus-forming activity of normal c-src, fyn, and lck cDNAs were compared in NIH 3T3 cell transfection assays. Focus formation was studied quantitatively, and individual foci were analyzed for phosphotyrosine content and expression of appropriate translational products. Each foci arising from c-src transfectants had a marked increase in phosphotyrosine content, and the majority of these foci expressed a c-src protein with an aberrant carboxyl terminus. Foci derived from lck transfectants also had a marked increase in phosphotyrosine content, and some foci expressed a lck protein with an aberrant carboxyl terminus. In contrast, foci from fyn-transfected cells were not distinguished from G418-selected mass cultures in terms of total phosphotyrosine content or expression of p59fyn. These studies support the previously published concept that overexpression of the normal fyn protein contributes to focus formation in transfected NIH 3T3 cells but suggest that the focus-forming activity observed after c-src or lck transfections is frequently attributable to mutational events. Because lck mutations have not been previously described in transformed foci, we characterized the lck transcript expressed in two foci and identified a novel point mutation that encodes a lck protein with increased in vivo kinase and focus-forming activity.

Transforming properties and substrate specificities of the protein tyrosine kinase oncogenes ros and src and their recombinants.

To determine the sequences of the oncogenes src (encoded by Rous sarcoma virus [RSV]) and ros (encoded by UR2) that are responsible for causing different transformation phenotypes and to correlate those sequences with differences in substrate recognition, we constructed recombinants of the two transforming protein tyrosine kinases (PTKs) and studied their biological and biochemical properties. A recombinant with a 5' end from src and a 3' end from ros, called SRC x ROS, transformed chicken embryo fibroblasts (CEF) to a spindle shape morphology, mimicking that of UR2. Neither of the two reverse constructs, ROS x SRC I and ROS x SRC II, could transform CEF. However, a transforming variant of ROS x SRC II appeared during passages of the transfected cells and was called ROS x SRC (R). ROS x SRC (R) contains a 16-amino-acid deletion that includes the 3' half of the transmembrane domain of ros. Unlike RSV, ROS x SRC (R) also transformed CEF to an elongated shape similar to that of UR2. We conclude that distinct phenotypic changes of RSV- and UR2-infected cells do not depend solely on the kinase domains of their oncogenes. We next examined cellular proteins phosphorylated by the tyrosine kinases of UR2, RSV, and their recombinants as well as a number of other avian sarcoma viruses including Fujinami sarcoma virus Y73, and some ros-derived variants. Our results indicate that the UR2-encoded receptorlike PTK P68gag-ros and its derivatives have a very restricted substrate specificity in comparison with the nonreceptor PTKs encoded by the rest of the avian sarcoma viruses. Data from ros and src recombinants indicate that sequences both inside and outside the catalytic domains of ros and src exert a significant effect on the substrate specificity of the two recombinant proteins. Phosphorylation of most of the proteins in the 100- to 200-kDa range correlated with the presence of the 5' src domain, including the SH2 region, but not with the kinase domain in the recombinants. This corroborates the conclusion given above that the kinase domain of src or ros per se is not sufficient to dictate the transforming morphology of these two oncogenes. High-level tyrosyl phosphorylation of most of the prominent substrates of src is not sufficient to cause a round-shape transformation morphology.

An Epstein-Barr virus transformation-associated membrane protein interacts with src family tyrosine kinases.

In latently infected growth-transformed human lymphocytes, Epstein-Barr virus (EBV) encodes two integral plasma membrane proteins: LMP1, which constitutively induces B-lymphocyte activation and intercellular adhesion, and LMP2A, which associates with LMP1 and is a tyrosine kinase substrate. We now demonstrate that LMP2A associates with src family protein tyrosine kinases, particularly lyn kinase, in nonionic detergent extracts of transfected B lymphoma cells or in extracts of EBV-transformed B lymphocytes. The LMP2A and tyrosine kinase association is stable in nonionic detergents and includes a 70-kDa cell protein which is also an in vitro or in vivo kinase substrate. This LMP2A association with B-lymphocyte src family tyrosine kinases is likely to be an important pathway in EBV's effects on cell growth.