INTRODUCTION

Erythropoietin (Epo),¹ a 34-kDa glycoprotein hormone, uniquely regulates the proliferation and differentiation of cells committed to the erythroid lineage (1). The interaction of Epo with the erythropoietin receptor (EpoR) activates the EpoR itself and several cellular proteins, which lead to a cascade of biochemical events (2-4). Although the EpoR itself does not contain a tyrosine kinase catalytic domain in its cytoplasmic domain, Epo rapidly induces the tyrosine phosphorylation of several cellular proteins such as Janus protein tyrosine kinase (JAK) 2 (5, 6) and Shc (7), as well as its own receptor (8-10). Furthermore, a number of other biochemical events have been associated with activation of the EpoR, including an increase in the activities of phosphatidylinositol (PI) 3-kinase (11, 12), phospholipase C-1 (13), p21^ras (14), Raf-1 (15), and mitogen-activated protein kinase (MAPK) (16). Although the tyrosine phosphorylation of these intracellular proteins correlates with Epo-induced mitogenesis (8), little is known about the sequence of these molecules and their importance to Epo-induced growth and differentiation in the target cells.

The vav proto-oncogene product is specifically expressed in hematopoietic cells and encodes a 95-kDa protein (Vav) that contains multiple structural motifs commonly found in intracellular signaling molecules, including Src homology (SH) 2, SH3, and pleckstrin homology domains as well as a helix-loop-helix domain, a leucine zipper-like domain, and a Rho guanine nucleotide exchange factor (GEF) homology domain (17-21). Vav has been shown to be tyrosine-phosphorylated by cross-linking of the T-cell receptor (21, 22), immunoglobulin (Ig) M antigen receptor (23), and CD19 (24) and in response to a variety of stimuli including interleukin-2 (IL-2) (25), IL-3, granulocyte macrophage-colony stimulating factor (GM-CSF) (26), stem cell factor (27), platelet-derived growth factor (20), and epidermal growth factor (20, 21). Recently, Epo has been demonstrated to induce tyrosine phosphorylation of Vav, and Vav may be involved in growth signaling from the EpoR (28). However, the potential downstream signaling proteins interacting with Vav in Epo signaling pathway remain unclear.

In the present study, we investigated the role of Vav in the Epo signaling pathway by characterizing its interaction with proteins that are known to be involved in Epo signal transduction. We observed a stable association of Vav with the EpoR and the physical interaction between Vav and p85 in response to Epo. We also examined the kinase responsible for the tyrosine phosphorylation of Vav and detected an Epo-induced association of JAK2 with Vav and tyrosine phosphorylation of Vav by JAK2.

EXPERIMENTAL PROCEDURES

F-36P (kindly provided by Dr. S. Chiba, University of Tokyo, Japan), a human IL-3 and GM-CSF-dependent cell line, was maintained in RPMI 1640 medium supplemented with 10% heat-inactivated fetal calf serum and 10 ng/ml IL-3. Anti-JAK2 and EpoR polyclonal antibodies, anti-rat p85 PI3-kinase antiserum, and anti-human Vav monoclonal antibody were purchased from Upstate Biotechnology, Inc. (Lake Placid, NY). Anti-EpoR antiserum was kindly provided by Dr. O. Miura (Tokyo Medical and Dental University, Japan). Anti-phosphotyrosine (anti-Tyr(P)) monoclonal antibody (PY20) was from ICN Biomedicals, Inc. (Costa Mesa, CA). Anti-p85 monoclonal antibody was purchased from Transduction Laboratories (Lexington, KY). The enhanced chemiluminescence detection kit was obtained from Amersham Corp. Recombinant human IL-3 and Epo were kindly provided from Kirin Brewery (Tokyo, Japan). All other reagents were purchased from commercial sources.

Cells (1 × 10⁷) were starved of growth factors for 24 h and then stimulated with or without Epo (50 units/ml) at 37 °C for the indicated periods. The cells were washed and lysed in lysis buffer (1% Nonidet P-40, 20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 2 mM EDTA, 0.1% aprotinin, 1 mM phenylmethylsulfonyl fluoride, and 1 mM Na₃VO₄). After 15 min of incubation at 4 °C, the insoluble materials were removed by centrifugation for 15 min at 15,000 rpm at 4 °C. The supernatants were incubated with the indicated antibodies or antiserum for 1 h to overnight at 4 °C and immunoprecipitated with Protein A-Sepharose. The immunoprecipitates were washed four times with the lysis buffer and eluted by boiling for 5 min in Laemmli's SDS sample buffer for SDS-polyacrylamide gel electrophoresis (SDS-PAGE).

After SDS-PAGE, the proteins were electrophoretically transferred onto a nitrocellulose membrane (Hybond-C super; Amersham Corp., UK) using a semi-dry transfer cell (Bio-Rad). The filter was blocked by incubation in TBS buffer (20 mM Tris-HCl, pH 7.6, 150 mM NaCl) containing 2% bovine serum albumin for 1 h at room temperature or overnight at 4 °C. The blots then were incubated for 1 h with an appropriate concentration of the primary antibody in TBS, washed three times for 5 min each with TBS-T (TBS buffer containing 0.1% Tween 20), and probed with a 1:1,000 dilution of biotinylated anti-mouse or anti-rabbit Ig followed by incubation with a 1:5,000 dilution of horseradish peroxidase-conjugated streptavidin (Caltag, South San Francisco, CA). After washing, the blots were visualized by using an enhanced chemiluminescence Western blotting detection system (Amersham Corp.).

Short-term proliferative responses to Epo were examined by a [³H]thymidine incorporation assay. The cells (1 × 10⁴) were incubated at 37 °C for 24 h in the presence or absence of 20 units/ml Epo before the addition of 0.25 µCi of [³H]thymidine. The cells were pulsed for 4 h and then harvested onto glass fiber filters. The filter strips were dried, and the amount of radioactivity present on each filter was determined.

Agarose-conjugated glutathione S-transferase (GST) -p85 SH2-N (amino acids 333-428) and SH2-C (amino acids 624-718) were purchased from Upstate Biotechnology Inc. For the in vitro binding assays, the GST fusion proteins were incubated in cell lysates at 4 °C for 1 h. The samples were washed four times with lysis buffer, eluted by boiling in Laemmli's SDS sample buffer, separated by SDS-PAGE, and immunoblotted with anti-Vav antibody.

Cells (1 × 10⁷) were stimulated with or without Epo, lysed, and immunoprecipitated with anti-p85, anti-Tyr(P), or anti-Vav antibody as described above. The immunoprecipitates were washed three times with lysis buffer, three times with 50 mM LiCl in 100 mM Tris-HCl, pH 7.4, and twice with TNE buffer (10 mM Tris-HCl, pH 7.4, 100 mM NaCl, 1 mM EDTA) containing 100 µM Na₃VO₄. The immunoprecipitates were resuspended in 30 µl of TNE buffer, followed by the addition of 10 µl of PI (2 mg/ml in 10 mM Tris-HCl, pH 7.4, 1 mM EGTA) and 10 µl of 100 mM MgCl₂. The reactions were initiated by adding 20 µCi of [-³²P]ATP and 2 µl of 100 mM ATP in 20 mM MgCl₂ and then were incubated for 10 min at 25 °C. The reactions were terminated with 100 µl of 1 N HCl and 200 µl of CHCl₃-MeOH (1:1), vortexed briefly, and separated into phases by centrifugation (2,000 rpm) for 10 min. The CHCl₃ phases were spotted onto oxalate-treated thin layer chromatography (TLC) plates and developed using a solvent system of CHCl₃:MeOH:H₂O:ammonium hydroxide (90:70:14.6:5.4). The PI3-phosphate was visualized by autoradiography and quantified by Bio-Imaging analyzer BAS2000 (Fuji Film, Tokyo).

Antisense (AS) and complementary sense (S) ODNs that targeted the translation initiation regions of the regulatory subunit of phosphatidylinositol 3-kinase (p85) and the proto-oncogene vav were synthesized in fully phosphorothioated forms. The sequence of each 20-mer was as follows: AS-p85 (5-TGGTACCCCTCAGCACTCAT-3), S-p85 (5-ATGAGTGCTGAGGGGTACCA-3), AS-vav (5-CATTGGCGCCACAGCTCCAT-3), S-vav (5-ATGGAGCTGTGGCGCCAATG-3). The F-36P cells (1 × 10⁴ cells) were exposed to ODNs at the indicated concentration in the defined medium for 6 h prior to the addition of Epo (50 units/ml). After 18 h of incubation, the cells were labeled for 4 h with 0.25 µCi of [³H]thymidine prior to harvest. For the analysis of the expression of p85 and Vav, the ODN-treated cells were lysed, and total cell lysates were separated by SDS-PAGE, followed by anti-p85 or anti-Vav immunoblotting as described above.

The JAK2 and Vav proteins were immunoprecipitated from serum-starved F-36P cells as described above and resuspended in the kinase buffer (40 mM HEPES, pH 7.4, 10 mM MgCl₂, 3 mM MnCl₂). Sepharose beads-conjugated Vav protein immunoprecipitated from Epo-unstimulated cells were mixed with the immunoprecipitated JAK2 kinases from the cells stimulated with or without Epo. The kinase reactions were initiated by the addition of 200 mM ATP, incubated for 15 min at 25 °C, and terminated by the addition of SDS sample buffer. The reaction mixtures were boiled for 5 min, separated by SDS-PAGE, and transferred to a nitrocellulose membrane, followed by immunoblotting with anti-Tyr(P) as described above.

All of the experiments described under "Results" were carried out at least twice and yielded similar results.

RESULTS

To examine the growth signaling pathway from the EpoR, the human nonlymphoid leukemia cell line F-36P was utilized. F-36P has an absolute dependence on IL-3 or GM-CSF and can be induced to proliferate in response to Epo (data not shown) (29). Although a previous study has shown that Epo induces tyrosine phosphorylation of Vav and suggested that Vav may play an important role in growth signaling from the EpoR (28), little is known about the molecules involved in the Vav signaling pathway. To identify any phosphorylated proteins that might interact with Vav, F-36P cells were stimulated with Epo, and anti-Vav immunoprecipitates were examined for tyrosine-phosphorylated proteins by anti-phosphotyrosine (anti-Tyr(P)) and anti-Vav immunoblotting (Fig. 1). Consistent with the previous study (28), Epo induced the tyrosine phosphorylation of Vav, and identical levels of Vav protein were present in each lane. In addition, a tyrosine-phosphorylated 70-kDa protein (pp70) also was observed after Epo stimulation. It has been shown that Epo induces tyrosine phosphorylation of the EpoR itself which causes a shift in its gel mobility from 66 to 72 kDa (8, 9). To confirm that the co-immunoprecipitated pp70 represents the tyrosine-phosphorylated EpoR and to examine whether Vav binds to the EpoR constitutively or inducibly after Epo stimulation, anti-EpoR immunoprecipitates were subjected to anti-Vav immunoblotting. Fig. 2A shows that Vav was co-immunoprecipitated with anti-EpoR in an Epo-independent manner. Furthermore, Vav co-immunoprecipitated unphosphorylated EpoR with or without Epo stimulation and phosphorylated EpoR after Epo stimulation (Fig. 2B), whereas neither normal rabbit serum nor preimmune mouse serum co-immunoprecipitated Vav or EpoR, respectively. These results indicate that Vav constitutively associates with the EpoR and suggest that this binding may be independent of tyrosine phosphorylation of the EpoR.

Previous studies have shown that the regulatory subunit of phosphatidylinositol 3-kinase (p85) can bind to tyrosine-phosphorylated EpoR (11, 12). Recently, Gobert et al. (30) have reported that a truncated EpoR, which had no tyrosine residues and no longer bound PI3-kinase, can mediate Epo-induced activation of PI3-kinase and suggested an alternative pathway for PI3-kinase activation (30). To test for the substrates associated with p85, anti-p85 immunoprecipitation and anti-phosphotyrosine immunoblotting were performed (Fig. 3). Epo stimulation led to no tyrosine phosphorylation of p85 as previously reported (11, 12). However, some tyrosine-phosphorylated proteins were associated with p85 after Epo stimulation, including two that migrated at approximately 100 (pp100) and 70 kDa (pp70). Consistent with previous reports (11, 12), pp70 was identified as the tyrosine-phosphorylated EpoR, and p85 associates with the EpoR after Epo stimulation (Fig. 4A). We then examined whether the pp100 that binds to p85 corresponds to the tyrosine-phosphorylated Vav. Fig. 4B (left panel) shows that Vav was detected in anti-p85 immunoprecipitates. Furthermore, as shown in Fig. 4B (right panel), anti-Vav immunoprecipitates blotted with p85 revealed an association of Vav with p85. The binding of p85 to the tyrosine-phosphorylated Vav strongly suggests that this binding may be mediated through the SH2 domains of p85. To address this possibility, the cell lysates were incubated with agarose-conjugated GST fusion protein containing either the amino-terminal SH2 domain of p85 (SH2N) or the carboxyl-terminal SH2 domain (SH2C) and analyzed by immunoblotting with anti-Vav. Fig. 5 shows that both the SH2 domains in the cell lysates from the Epo-stimulated cells can bind Vav in vitro, although the carboxyl-terminal SH2 domain binds to Vav with a higher affinity than does the amino-terminal SH2 domain. To better clarify the interaction between Vav and PI3-kinase, in vitro PI3-kinase assays were performed with anti-Tyr(P) and anti-Vav immunoprecipitates. We found that the Epo-induced in vitro PI3-kinase activity was associated with Vav (Fig. 6). Although the PI3-kinase activity of the anti-Vav immunoprecipitates was approximately 1/100 that of the anti-p85 immunoprecipitates after Epo stimulation, the activity of anti-Vav immunoprecipitates resulted in accounting for approximately 80% that of anti-Tyr(P) immunoprecipitates (data not shown). These results suggest that Epo stimulation leads to the association of PI3-kinase with tyrosine-phosphorylated Vav via the SH2 domain(s) of p85.

Previous studies have demonstrated that the membrane proximal region of the EpoR, which is crucial for growth signaling (31, 32), also is required for the activation of JAK2 (5, 6) and the tyrosine phosphorylation of Vav (28). This suggests that Vav plays an important role in growth signaling. On the other hand, p85 has been shown to associate with the carboxyl-terminal region of the EpoR, a domain that may not play a role in growth signaling (12).

To clarify the functional role of Vav and p85 in EpoR signaling pathway, we utilized antisense (AS) and complementary sense (S) oligodeoxynucleotides (ODNs) that targeted the translation initiation regions of the proto-oncogene vav and the regulatory subunit of phosphatidylinositol 3-kinase (p85) in fully phosphorothioated forms. Epo-stimulated F-36P cells were incubated with AS- or S- ODNs for 24 h and analyzed by immunoblotting. Treatment of AS-vav and AS-p85 resulted in significant decreases in the expression of Vav and p85 compared with the sense treatment (Fig. 7, A and C).

If Vav significantly contributes to the proliferation signaling from the EpoR, down-regulation of the expression of Vav should result in the inhibition of cell growth. To test this hypothesis, the effect of AS-vav on cell proliferation was examined by a [³H]thymidine incorporation assay. Fig. 7A shows that AS-vav treatment inhibited the Epo-dependent proliferation of F-36P cells in a dose-dependent manner after co-incubation for 24 h, a time point when cell viability in all cultures was >90% (data not shown), whereas S-vav had no significant effect. This indicates that the observed inhibition of cell growth was due to antisense-mediated loss of Vav and not to growth suppression by ODN degradation products. Since we have observed the physical interaction between Vav and p85, we next examined the effect of AS-vav treatment on Epo-induced activation of PI3-kinase. F-36P cells were incubated in defined medium with Epo and 10 µM AS- or S-vav for 24 h. The cells were lysed, and in vitro PI3-kinase assays were carried out with anti-p85 immunoprecipitates. As shown in Fig. 7B, treatment of cells with AS-vav resulted in a marked inhibition of PI3-kinase activity, suggesting the possible role of Vav on Epo-induced PI3-kinase activity. To confirm the involvement of Vav-p85 pathway in Epo-induced proliferation, we subsequently examined the effect of AS-p85 on cell proliferation. Although the cell viability in all cultures was >90% and no significant differences of the viability were observed with or without AS-ODN treatment (data not shown), AS-p85 treatment significantly suppressed the [³H]thymidine uptake (Fig. 7C), as observed in AS-vav treatment.

It is important to identify the possible tyrosine kinases responsible for the tyrosine phosphorylation of Vav in the Epo signaling pathway. An important advance toward understanding cytokine actions was provided by the recently defined association between cytokine receptors and members of the JAK family (33). Since JAK2 is tyrosine-phosphorylated and associates with the EpoR (5, 6) in response to Epo, we hypothesized that JAK2 may be a tyrosine kinase capable of phosphorylating Vav. To examine whether Vav might be associated with JAK2, anti-JAK2 immunoprecipitates were subjected to anti-Vav immunoblotting. Fig. 8A shows that Epo stimulation resulted in the association of Vav with JAK2.

To test whether Vav can serve as a substrate of JAK2, we examined the in vitro tyrosine phosphorylation of Vav by JAK2. Immunoprecipitated Vav protein was mixed with immunoprecipitated JAK2 kinase, and these mixtures, as well as the immunoprecipitated Vav or JAK2 alone, were resuspended in kinase buffer. ATP then was added, and the mixtures were incubated for 15 min at 25 °C. The reaction mixtures were separated by SDS-PAGE, transferred to a nitrocellulose membrane, and immunoblotted with anti-Tyr(P). As shown in Fig. 8B, activated JAK2 from the Epo-stimulated cells, but not inactivated JAK2, can phosphorylate Vav on tyrosine residues. Tyrosine-phosphorylated JAK2 was co-immunoprecipitated with Vav from Epo-stimulated F-36P cells. This result may be due to the autophosphorylation of JAK2 in vitro, which was co-immunoprecipitated with anti-Vav antibody, by the addition of ATP. These data indicate that JAK2 can interact with Vav and may be responsible for the tyrosine phosphorylation of Vav in the Epo signaling pathway.

DISCUSSION

Vav contains multiple protein-protein interaction motifs including SH2, SH3, leucine zipper, helix-loop-helix, pleckstrin homology, and nuclear translocation domains (17-21) through which Vav has been shown to bind to signaling molecules (34, 35) and translocate into the nucleus (36). The complex structure of Vav can potentially give rise to numerous protein-protein interactions in cells stimulated by Epo. A subset of interacting proteins might be expected to interact stably and to be co-immunoprecipitated with Vav isolated from cell extracts. The present study has shown that Vav is constitutively associated with the EpoR, and Epo stimulation induces tyrosine phosphorylation of Vav and a physical interaction between Vav and p85, suggesting proximal roles of Vav and p85 in EpoR signaling pathway. Previous studies have demonstrated that tyrosine-phosphorylated proteins that migrated at 90-100 kDa could be co-immunoprecipitated with EpoR and p85 (6, 11, 12). Our results indicate that one of these EpoR or p85-bound proteins might be Vav. An association between Vav and growth factor receptors or surface molecules has been reported in the epidermal growth factor, platelet-derived growth factor, CD19, and prolactin signaling pathways (20, 21, 24, 36). Since these associations are dependent on ligand stimulation, they are thought to be mediated through the SH2 domains of Vav. On the other hand, our results indicate a stable association of Vav with the EpoR in an Epo-independent manner, suggesting that its interaction may be independent of tyrosine phosphorylation of the EpoR and may be mediated via other domains, such as an SH3 domain, rather than the SH2 domains. Analysis of the minimal core recognition motifs for the SH3 domains of several proteins have revealed a highly conserved proline-rich motif, namely the PXXP consensus motif (37, 38). Since the EpoR contains the PXXP motifs in its cytoplasmic domain, including its amino-terminal region, these motifs might participate in the interaction between Vav and EpoR. The essential domain through which Vav associates with the EpoR remains to be determined. A physical interaction between Vav and p85 is induced by ligand binding, suggesting that this association might be dependent on tyrosine phosphorylation of Vav and mediated through the SH2 domains of p85. In vitro binding assays (Fig. 5) are consistent with this hypothesis. Based on the use of phosphopeptide libraries to determine the peptide-binding site specificity of SH2 domains, both the amino- and carboxyl-terminal SH2 domains of p85 have been shown to specifically bind to the Tyr(P)-XXM motif in tyrosine kinase receptors and cellular substrates (39, 40). Recently, Damen et al. (41) identified a new recognition motif for the amino- and carboxyl-terminal SH2 domains of p85, i.e. Tyr(P)-VAC, on the basis of the fact that both methionine and cysteine at +3 position can provide a hydrophobic side chain with sulfur atoms positioned in a similar orientation. In this regard, the YDFC sequence of Vav may be a recognition motif for p85 that contributes to the interaction between Vav and p85.

Mechanisms through which Vav may mediate signals have focused on its guanine nucleotide exchange factor (GEF) activity for Ras, leading to the sequential activation of MAPKs. GEF activity for Ras associated with Vav has been demonstrated in T-cell activation and vav-transfected fibroblasts (22, 42), whereas another group has disrupted this activity of Vav and found that Vav can cooperate with normal Ras proteins to transform NIH3T3 cells (43). A recently published report has indicated that the prolactin-induced activation of GEF activity co-immunoprecipitated with an anti-Vav antibody did not require the expression of Vav, suggesting that another protein may serve as the Vav-associated GEF activity (36). In another report, Vav has been suggested to be involved in the activation of the MAPKs that are distinct from the Ras pathway (44). Although we have not assessed the Epo-induced GEF activity associated with Vav, disassociation of tyrosine phosphorylation of Vav with the activation of MAPKs (16) provides the possibility of a signaling pathway of Vav distinct from Ras and MAPKs.

PI3-kinase activity has been shown to interact with a large number of receptor and non-receptor protein-tyrosine kinases (45). In Epo signaling pathway, Epo simulation causes a physical association of p85 with the tyrosine-phosphorylated carboxyl-terminal region of the EpoR (11, 12). Since this region may not play an important role in growth signaling by analyses of truncated mutant EpoRs (31, 32), it has been inferred that PI3-kinase is not required for the Epo-mediated proliferative signal (12). However, since the PI3-kinase catalytic activity in the EpoR mutants was analyzed by using anti-EpoR immunoprecipitates, the possibility that a distinct pathway may regulate this activity cannot be eliminated when p85 associates with other proteins. On the basis of a report that a truncated EpoR, which had no tyrosine residues and no longer bound PI3-kinase, could mediate Epo-induced activation of PI3-kinase and sustain Epo-induced growth efficiently (30), we believe that an alternative pathway to associate with and activate PI3-kinase might be involved. Another group subsequently confirmed this possibility by means of mutant EpoRs bearing tyrosine to phenylalanine substitutions (41). In this study, we have identified a Vav-p85 pathway that mediates the Epo-induced proliferative signal. The Vav-associated PI3-kinase activity was only a part of that associated with p85. This result may indicate the involvement of an alternative pathway(s) in the activation of PI3-kinase. It is also possible that most of the kinase may exist in a free and activated form after Epo stimulation. These possibilities are still unclear and remain to be determined. PI3-kinase consists of two subunits of 85 (p85) and 110 kDa (p110), and it is now believed that p85, lacking intrinsic catalytic activity, plays a regulatory role by binding to the p110 catalytic subunit (45-48). Activation of PI3-kinase activity appears to require conformational changes of the subunits induced by binding of p85 to appropriate ligand sequences including phosphotyrosine-containing sequences (45). In addition to the physical interaction between Vav and p85, we found that Vav-associated PI3-kinase activity accounted for approximately 80% that associated with anti-Tyr(P) and that As-vav treatment inhibited the PI3-kinase activity associated with p85 (Fig. 7B). These results suggest that Vav may play an important role in recruitment of PI3-kinase to an appropriate subcellular compartment and engage in the activation of PI3-kinase. Since p85 possesses some structural signaling motifs such as SH2, SH3, Bcr domains, and internal proline-rich regions (45, 47), p85 is likely to engage in protein-protein interactions with other intracellular signaling molecules in addition to the regulation of p110 catalytic activity. The tyrosine-phosphorylated proteins that we observed which were associated with p85, except for pp70 and pp100, may be involved in the downstream signaling of p85. Further studies are needed to identify the binding site essential for the interaction between Vav and p85 and to investigate the downstream signaling cascades.

The interaction of Vav with the EpoR would make it possible to bring Vav into close proximity with other protein-tyrosine kinases (PTKs) known to associate with the EpoR. Moreover, the Epo-induced tyrosine phosphorylation of Vav suggests the existence of some PTK that is responsible for this reaction. Although Vav has been shown to be a substrate for the p56^lck in TCR-CD3-initiated signal transduction (22), little is known regarding which PTK is responsible for its phosphorylation. Analysis of mutant EpoRs has demonstrated that the tyrosine phosphorylation of Vav requires the membrane proximal region of the EpoR, previously shown to be crucial for the activation of JAK2. Furthermore, the tyrosine phosphorylation of Vav correlates with the activation of JAK2 (28). Recently, it also has been shown that Vav becomes associated with JAK2 in GM-CSF-treated cells and that the tyrosine phosphorylation of Vav is significantly increased in insect cells in which JAK2 and Vav proteins are overexpressed (26). These findings suggest a crucial role of JAK kinases for the tyrosine phosphorylation of Vav. On the other hand, Machide et al. (49) have reported the Epo-induced association of Tec kinase with Vav. Although we have not investigated the Tec kinase, detection of no tyrosine-phosphorylated proteins, except for Vav and JAK2, associated with Vav (Fig. 8B, lane 4) may indicate the less possible involvement of Tec kinase in phosphorylation of Vav in F-36P cells. Since we could utilize neither the purified JAK2 nor the purified Vav, we cannot definitely conclude that JAK2 directly phosphorylates Vav. However, our results may suggest the possible role of JAK2 on tyrosine phosphorylation of Vav. Activation of all known cytokine receptors induces tyrosine phosphorylation and activation of one or more JAK kinases, and JAK kinases have been shown to tyrosine phosphorylate other cellular substrates (33). In this regard, JAK kinases may play a central role for the cellular responses in the cytokine receptor families.

Although recent studies on PTK-mediated relative sequences toward the nucleus have been focused mainly on the JAK- signal transducers and activators of transcription (STAT) pathways (33), the precise roles of STATs in hematopoietic cells remain unclear. The fact that IL-2, IL-3, IL-4, IL-6, and Epo mutant receptors can trigger cell proliferation without STATs activation (50-53) raises the possibility of the presence of an alternative pathway(s) linking it to cell proliferation. In this regard, the Vav-p85 pathway we observed in this study may be a candidate for one of these alternative pathway(s).