Identification and characterization of an interleukin-3 receptor-associated 110-kDa serine/threonine kinase.

We recently reported that interleukin-3 (IL-3) stimulation of the murine IL-3-responsive cell line, B6SUtA, results in the rapid phosphorylation of the β subunit of the IL-3 receptor (IL-3R), not only on tyrosine residues but on serine/threonine (Ser/Thr) residues as well. Since this occurred even at 4°C, it suggested that a Ser/Thr-specific kinase might be closely associated with the IL-3R. To test this possibility, IL-3R complexes were isolated with anti-IL-3R (αIL-3R) antibodies, and in vitro phosphorylation studies were undertaken. These revealed the presence of a 110-kDa protein that was heavily phosphorylated in vitro on serine and threonine residues and that bound selectively to -ATP-Sepharose beads. Moreover, this protein, which was not the 110-kDa subunit of phosphatidylinositol 3-kinase, was tyrosine phosphorylated in response to IL-3 and was specifically labeled in vitro with azido-[P]ATP. These data, together with in vitro kinase inhibitor studies, suggest that an as yet uncharacterized H7- and staurosporine-sensitive 110-kDa Ser/Thr kinase may be constitutively associated with the IL-3R and activated following IL-3 stimulation. A comparison of IL-3R and erythropoietin receptor complexes suggests that this 110-kDa protein may be preferentially associated with the IL-3R.

The identification of receptor-associated proteins is an essential step in elucidating downstream signals initiated by the binding of growth factors to their receptors. In the case of interleukin-3 (IL-3), 1 a cytokine that stimulates the proliferation and differentiation of various hemopoietic cell lineages and activates the functions of mature macrophages, eosinophils, and mast cells (1), its high affinity receptor consists of two subunits, designated ␣ and ␤ (1). While the 70-kDa ␣ chain is specific for IL-3 (2), the 140-kDa ␤ subunit (␤ c ) is shared by IL-3, granulocyte-macrophage colony-stimulating factor, and IL-5 (3). In the mouse, there is a second ␤ subunit called ␤ IL-3 , which shares 91% amino acid identity with ␤ c and is specific for IL-3 (4). Although none of these receptor subunits possess intrinsic kinase activity, IL-3 binding rapidly induces the tyrosine phosphorylation of a group of cellular proteins, including its own ␤ subunits (5)(6)(7)(8)(9)(10)(11)(12)(13). Concomitant with tyrosine phosphorylation, IL-3 receptors (IL-3Rs) also become phosphorylated on serine and threonine (Ser/Thr) residues (8,9). Since many IL-3-induced phosphorylations occur very rapidly, even at 4°C, it has been suggested that the protein kinases responsible for these phosphorylations and many of their protein substrates may be associated with the IL-3R prior to ligand binding (1,14).
In early studies to identify components of the IL-3R, 125 I-IL-3 cross-linking studies revealed the presence of 140-and 70-kDa proteins (15)(16)(17)(18)(19)(20) that were subsequently shown to correspond to the IL-3R ␤ and ␣ (plus a cleavage product of the ␤) chains (2,4,8,21), respectively. More recently, the search for IL-3Rassociated proteins has revealed that Jak2, a member of the Jak family of tyrosine kinases (22), binds to the membrane proximal region of the ␤ subunit and becomes activated following IL-3 binding (23). This tyrosine kinase is most likely responsible, at least in part, for the rapid tyrosine phosphorylations that occur following IL-3 stimulation. In this regard, another protein, which has been shown to bind to the IL-3R following IL-3 stimulation, is hematopoietic cell phosphatase (24,25). Association of this tyrosine-specific phosphatase with the IL-3R ␤ subunit appears to lead to dephosphorylation of the IL-3R and Jak2 and down-regulation of the response of hemopoietic cells to IL-3 (24,25). The kinase(s) responsible for the rapid Ser/Thr phosphorylation of the IL-3R, however, has not as yet been identified, but the fact that these phosphorylations have been observed at 4°C suggests that it might be receptor associated as well.
In this study, we have investigated whether a Ser/Thr kinase is associated with the IL-3R using both anti-IL-3R (␣IL-3R) antibody and biotinylated IL-3/streptavidin agarose-based approaches. Our results suggest that an H7-and staurosporinesensitive 110-kDa Ser/Thr kinase is constitutively associated with the IL-3R, becomes tyrosine phosphorylated in vivo in response to IL-3, and may be more strongly associated with the IL-3R than the EpR.

MATERIALS AND METHODS
Reagents-Pure Cos cell-derived recombinant murine IL-3 was prepared and biotinylated as previously described (26). Recombinant human Ep was purified from culture supernatants of baby hamster cells expressing an Ep cDNA and biotinylated as previously described (27). The ␣-phosphotyrosine (␣PY) monoclonal antibody, 4G10, ␣p85 of PI 3K, and ␥-phosphate linked ATP-Sepharose were obtained from UBI (Lake Placid, NY). Rabbit antiserum to the human IL-3R ␤ c subunit was generated by immunizing animals with a glutathione S-transferase (GST) fusion protein containing the intracellular domain of the hIL-3R ␤ c subunit, and the resulting ␣IL-3R antibody was affinity purified using immobilized antigen. This antibody was highly effective at immunoprecipitating the murine IL-3R ␤ subunits (9). ␣Jak2 antibody * This work was supported in part by the National Cancer Institute of Canada and the Medical Research Council of Canada with core support from the British Columbia Cancer Foundation and the British Columbia Cancer Agency (to G. K.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ Cells-The murine IL-3-dependent cell line, B6SUtA 1 , which expresses high levels of murine IL-3Rs (15), was propagated in RPMI 1640 containing 10% fetal calf serum (FCS) and 5% pokeweed mitogenstimulated spleen cell-conditioned medium. The murine IL-3-dependent cell line, Ba/F3, was retrovirally infected with a JZenTKneo vector containing the wild type murine EpR cDNA as previously described (28), and one high EpR-expressing clone, designated as BA-ER, was used in this study and routinely maintained in RPMI 1640 containing 10% FCS and 0.5 units/ml Ep.
Labeling of Cells with [ 32 P]Orthophosphate-B6SUtA 1 cells were washed twice with phosphate-free RPMI and resuspended in phosphate-free RPMI plus 2% FCS for 2 h at 37°C to deprive the cells of growth factors and inorganic phosphate. They were then resuspended in phosphate-free RPMI supplemented with 2% FCS and 0.25 mCi/ml carrier-free [ 32 P]orthophosphate for 90 min at 37°C. The labeled cells were washed once with phosphate-free RPMI, resuspended in 4°C medium for 10 min, and then IL-3 or control buffer was added for a further 10 min at 4°C. The cells were then solubilized with 0.5% Nonidet P-40 in PSB and immunoprecipitated as previously described (29).
In Vitro Phosphorylations-Immunoprecipitates were washed once with kinase reaction buffer (50 mM NaCl, 5 mM MnCl 2 , 5 mM MgCl 2 , 10 mM Hepes, pH 7.4) and gently rotated in 25 l of kinase reaction buffer containing 20 Ci of [␥-32 P]ATP for 30 min at 23°C. The beads were then washed three times and boiled in SDS-sample buffer.
Generation of Phosphorylated Biotin IL-3⅐IL-3R Complexes-Biotinylated IL-3 (bIL-3) and biotinylated Ep were used to isolate their receptor complexes as described by Yoshimura and Lodish (30). Briefly, 5 ϫ 10 7 cells were incubated with biotinylated ligand with or without excess underivatized ligand for 10 min at 37°C, washed with PBS, and crosslinked for 30 min at 4°C with 0.5 mM dithiobissuccinimidyl propionate (DSP) in PBS containing 0.1 mM sodium vanadate. After washing with PBS containing 20 mM Tris-Cl (pH 7.5), cells were lysed (30), incubated with streptavidin agarose beads for 2 h at 4°C, and the beads were washed three times with lysis buffer (30) containing 0.5% Nonidet P-40. The beads were then washed with kinase reaction buffer and gently rotated for 30 min in 25 l of kinase reaction buffer containing 20 Ci of [␥-32 P]ATP. The beads were washed three times, boiled in 20 l of 2% SDS, and re-extracted in 50 l of boiling water. The supernatants were combined, diluted to 1 ml with lysis buffer containing 0.5% Nonidet P-40, and immunoprecipitated with ␣IL-3R or ␣PY antibodies.
GST-IL-3R ␤ Subunit Affinity Chromatography-A GST-IL-3R fusion protein consisting of the 27-kDa amino-terminal of GST linked to the entire intracellular domain of the IL-3R ␤ c subunit was expressed in a pGEX-2T plasmid and affinity purified from Escherichia coli cell lysates with glutathione (GSH)-agarose, as previously described (31). The fusion protein was specifically eluted from the beads with 20 mM GSH, 75 mM Hepes, pH 7.4, 150 mM NaCl, 5 mM dithiothreitol, and 0.1% Triton X-100, dialyzed overnight at 4°C against PBS to remove GSH, and rebound to GSH-agarose. The beads were then incubated with B6SUtA 1 cell lysates, washed extensively with PSB, 0.5% Nonidet P-40, and subjected to in vitro phosphorylation with [␥-32 P]ATP; the proteins were boiled off in SDS-sample buffer, separated by SDS-PAGE, and detected by autoradiography. Parallel experiments were carried out with GST-agarose beads to assess nonspecific sticking.
␥-Phosphate-linked ATP-Sepharose Chromatography-This was carried out essentially as described by Haystead et al. (32). Briefly, B6SUtA 1 cells, incubated with or without IL-3, were lysed, and IL-3R complexes were immunoprecipitated with ␣IL-3R antibodies. The complexes were freeze-thawed six times, and the supernatants containing the dissociated proteins were diluted with 1 ml of PSB containing 1% Nonidet P-40, 50 mM ␤-glycerophosphate and incubated with ␥-ATP-Sepharose for 45 min at 4°C. The beads were washed to remove unbound proteins, subjected to in vitro phosphorylation with [␥-32 P]ATP, and boiled with SDS-sample buffer or incubated with 1 mM ATP to elute specifically bound proteins. The samples were then subjected to SDS-PAGE and autoradiography.
Azido-[ 32 P]ATP Labeling-IL-3R complexes were isolated using bIL-3/streptavidin agarose beads as described above but without crosslinker. The beads were suspended in 50 l of kinase reaction buffer containing 2 l of 8-azido-[␣-32 P]ATP, in the presence or absence of 5 mM unlabeled ATP, and then exposed to ultraviolet light ( 254 ) for 5 min at 23°C. The beads were washed with lysis buffer, and the proteins were boiled off with SDS-sample buffer and resolved using SDS-PAGE.
Phosphoamino Acid Analysis-The gel slice containing the p110 protein was excised from SDS-gels, rehydrated by rocking for 15 min in 50 mM NH 4 HCO 3 , homogenized with a small glass pestle in 0.8 ml of 50 mM NH 4 HCO 3 , and rotated at 23°C for 48 h. The extracted protein was precipitated with 15% trichloroacetic acid in the presence of 40 mg of bovine serum albumin for 1 h at 4°C and washed three times with cold acetone. The pellets were air dried, resuspended in 0.2 ml of 6 M HCl, and incubated for 60 min at 110°C. The HCl was then removed by lyophilization, and the samples were resuspended in 1:1 ethanol:water, mixed with non-radioactive phosphoamino acid standards, and electrophoresed on cellulose thin layer plates (8). The plates were stained with 0.5% ninhydrin, dried, and subjected to autoradiography.

RESULTS
Identification of IL-3R-associated Proteins-As a first step toward determining whether a Ser/Thr kinase was associated with the IL-3R, we generated affinity-purified polyclonal antibodies to the ␤ subunit of the IL-3R using, as immunogen, a GST fusion protein containing the entire intracellular region of this receptor subunit. These antibodies were incubated with lysates from B6SUtA 1 cells, which are IL-3-dependent murine myeloid cells expressing approximately 100,000 cell surface IL-3Rs (15). Following immunoprecipitation with protein A-Sepharose beads and in vitro phosphorylation with [␥-32 P]ATP, the immunoprecipitates were subjected to SDS-PAGE and autoradiography. As can be seen in Fig. 1A (middle lane), several phosphorylated protein bands were observed, and these bands were not present when the same cell lysates were immunoprecipitated with either ␣GST antibodies or with ␣IL-3R antibodies in the presence of competitive antigen. The same set of protein bands was also observed using biotinylated-IL-3/DSP cross-linking/streptavidin agarose or GST-IL-3R ␤ subunit affinity chromatography (see "Materials and Methods"), indicating that these proteins were present in our ␣IL-3R immunoprecipitates because of their association with the IL-3R rather than because they were cross-reacting directly with the ␣IL-3R antibody (data not shown).
To compare the intensities of the in vitro phosphorylated protein bands in our ␣IL-3R immunoprecipitates with their protein levels, these precipitates were subjected to SDS-PAGE and silver staining (Fig. 1B). As expected, the IL-3R ␤ subunit was the most intensely staining band (apart from the IgG heavy chain). Interestingly, though, the heaviest 32 P-labeled band, migrating with an apparent molecular mass of 110 kDa, appeared to be only a minor component within the IL-3R complex, as assessed by silver staining.
To determine if these IL-3R-associated proteins were present within the IL-3R complex prior to ligand binding, ␣IL-3R immunoprecipitates from IL-3-stimulated and -unstimulated cells were subjected to in vitro phosphorylation with [ 32 P]ATP and analyzed using SDS-PAGE and autoradiography. As can be seen in Fig. 2A (left panel), all of the proteins detected following IL-3 stimulation were present before stimulation. However, on occasion, an increase in the intensity of some of the bands was observed following exposure to IL-3. This preassociation is consistent with previous 4°C results (9,14).
The IL-3R-associated p110 Binds Specifically to ␥-Phosphate-linked ATP-Sepharose and Is Phosphorylated in Vitro on Serine and Threonine Residues-To determine if one or more of the IL-3R-associated, in vitro phosphorylated proteins was a kinase, we incubated the proteins present in ␣IL-3R immunoprecipitates with ␥-phosphate-linked ATP-Sepharose, since these beads recently have been shown to be highly effective in binding protein kinases (32). Specifically, B6SUtA 1 cells, incubated with or without IL-3, were lysed, and the IL-3R complexes were immunoprecipitated with ␣IL-3R antibodies and then dissociated by freeze-thawing. The dissociated proteins were then incubated with ATP-Sepharose, and, following several washes of the beads, the tightly bound proteins were subjected to phosphorylation with [␥-32 P]ATP. SDS-PAGE and autoradiography of the ATP-Sepharose bound material revealed that p110 was greatly enriched by this process (Fig. 2A,  middle panel). Moreover, we found that p110 could be eluted from these beads with 1 mM ATP, suggesting that p110 bound specifically to the ATP moiety of ATP-Sepharose ( Fig. 2A, right  panel). Interestingly, this 110-kDa protein always migrated slightly more slowly following IL-3 stimulation, suggesting that it might be phosphorylated in vivo in response to IL-3. Phosphoamino acid analysis of ATP-Sepharose bound, in vitro phosphorylated p110 demonstrated that phosphorylation occurred primarily on serine and threonine residues (Fig. 2B). These observations proved that a Ser/Thr kinase was present within the IL-3R complex, since p110 was phosphorylated in vitro on Ser/Thr residues, and that this kinase might be p110.
The IL-3R-associated p110 Is Specifically Labeled with Azido-[ 32 P]ATP-Since not all kinases autophosphorylate (33), it was possible that our ATP-Sepharose beads bound a nonautophosphorylating kinase in addition to p110 and that it was actually this other protein that was the Ser/Thr kinase responsible for in vitro phosphorylating p110. To test this, we silver stained SDS-gels of ATP-Sepharose bound material. Although the only band visible was a very faint 110-kDa protein band (data not shown), indicating a substantial enrichment of this protein compared to its relative level in ␣IL-3R immunoprecipitates (see Fig. 1B), we could not rule out the presence of extremely low levels of a non-autophosphorylating Ser/Thr kinase in the ATP-Sepharose purified preparation. We therefore set out to determine whether the IL-3R-associated p110 was a Ser/Thr kinase by incubating isolated IL-3R complexes with azido-[ 32 P]ATP. As can be seen in Fig. 3, a 110-kDa protein was labeled by this procedure and was specifically competed by excess unlabeled ATP. Since no other proteins were labeled using this kinase-specific probe, it strongly suggested that p110 was indeed an IL-3R-associated kinase.
Inhibition of the p110 Protein Kinase with Staurosporine and H7 and Lack of Identity with the p110 Subunit of PI 3K-To characterize the IL-3R-associated p110 protein kinase, various kinase inhibitors were tested in in vitro kinase assays. Specifically, IL-3R complexes were isolated from B6SUtA 1 cells by immunoprecipitation with ␣IL-3R antibodies, and the protein kinase was purified using ATP-Sepharose. The washed, p110bound ATP-Sepharose beads were then incubated with [␥-32 P]ATP in the presence or absence of the protein kinase C-specific inhibitor, compound 3, the Ser/Thr kinase inhibitors staurosporine and H7, and the tyrosine kinase inhibitors genistein and herbimycin A. As shown in Fig. 4A, staurosporine at 1 M and H7 at 100 M were effective inhibitors of p110 phosphorylation, while compound 3, genistein, and herbimycin A had little effect at concentrations as high as 10 M, 200 g/ml, and 8 M, respectively. The protein kinase A-specific inhibitor, H89, the protein kinase C inhibitor, chelerythrine, and the tyrosine kinase inhibitors, tyrphostin B42 and B46, were also without effect. Dose response studies with staurosporine and H7 revealed that as little as 50 M H7 and 0.1 M staurosporine completely inhibited the in vitro phosphorylation of p110 (Fig. 4B).
Since it was recently shown that the 110-kDa subunit of PI 3K phosphorylates not only phospholipids but proteins as well (34,35), and since we and others have reported that PI 3K associates with certain cytokine receptors (31,36,37), we investigated whether the IL-3R-associated 110-kDa protein might be the p110 subunit of this enzyme. To test this, B6SUtA 1 cells, incubated with or without IL-3, were lysed and immunoprecipitated with ␣IL-3R or ␣p85 (of PI 3K) antibodies, and the precipitates were subjected to in vitro phosphorylation in the presence and absence of Wortmannin, a specific inhibitor of PI 3K (38). As can be seen in Fig. 4C, Wortmannin dramatically inhibited the phosphorylation of the p110 subunit of PI 3K but not the p110 present in the IL-3R complex, strongly suggesting that the heavily phosphorylated 110-kDa protein in the IL-3R complex was not the 110-kDa subunit of PI 3K. This gel also revealed that the p110 subunit of PI 3K migrated slightly more slowly than the 110-kDa IL-3R-associated protein, further proving their non-identity.

IL-3 Stimulates the Tyrosine Phosphorylation of p110 in Vivo and Increases the Ser/Thr Phosphorylation of the IL-3R-To
determine if the IL-3R-associated p110 was tyrosine phosphorylated in response to IL-3, B6SUtA 1 cells, treated with or without IL-3 for 5 min at 37°C, were lysed, proteins were immunoprecipitated with ␣IL-3R antibodies, and Western blot analysis was carried out with the ␣PY antibody, 4G10. These studies revealed the presence of several faint but consistently observed tyrosine-phosphorylated proteins, including one minor band of 110 kDa (Fig. 5A, left panel). Reprobing this blot with ␣IL-3R ␤ subunit antibodies demonstrated equal loading of IL-3Rs in control and IL-3-stimulated samples (Fig. 5A, right  panel). To test whether the p110 Ser/Thr kinase was tyrosine phosphorylated in response to IL-3 in a more rigorous fashion, isolated IL-3R complexes were labeled with azido-[ 32 P]ATP and then dissociated by boiling in SDS and immunoprecipitated with ␣PY antibodies. As can be seen in Fig. 5B, a 110-kDa protein band was clearly present in SDS-gels and was specifically inhibited by excess unlabeled ATP.
To investigate whether IL-3 activated the IL-3R-associated p110 Ser/Thr kinase, B6SUtA 1 cells were labeled with [ 32 P]orthophosphate, cooled to 4°C, and incubated for 10 min at 4°C with or without IL-3. The cells were then lysed, and proteins were immunoprecipitated with ␣IL-3R antibodies and subjected to SDS-PAGE and autoradiography. As can be seen in Fig. 6A, several IL-3R-associated proteins showed increased phosphorylation, even at 4°C, in response to IL-3. Phosphoamino acid analysis of one of these proteins (i.e. the IL-3R ␤ subunit itself) revealed that this increase was due to increases in both tyrosine and Ser/Thr phosphorylation (Fig.  6B). This finding, together with our other results to this point, suggested that p110 becomes activated following IL-3 stimulation.
Comparison of IL-3R and EpR Complexes Suggests p110 Is Preferentially Associated with the IL-3R-While considerable overlap appears to exist between the signaling pathways uti- lized by IL-3 and Ep (28), recent data suggest that stimulation of EpRs and IL-3Rs leads to different cellular events in some cell types (39 -42). For example, Ep, but not IL-3, has been shown to induce EpR-expressing Ba/F3 cells (i.e. BA-ER cells) to accumulate ␤-globin mRNA (39,40) and to express cell surface glycophorin (41). To determine whether the IL-3Rassociated 110-kDa protein also interacts with the EpR, BA-ER cells were stimulated with IL-3 or Ep and lysed. The lysates were then incubated with either ␣IL-3R or ␣EpR antibodies, and the immunoprecipitated complexes were subjected to in vitro phosphorylation, SDS-PAGE, and autoradiography. As shown in Fig. 7A, p110 was clearly observed in IL-3R immunoprecipitates but barely detected in EpR immunoprecipitates. On the other hand, a 130-kDa band, which was established by immunoblotting studies to be Jak2 (data not shown), appeared to be present in EpR complexes at a much higher level than in the IL-3R complexes in these cells, and this level appeared to be unaffected by Ep binding, in keeping with previously published findings (43). Alternatively, even though there are more EpRs (approximately 10,000) than IL-3Rs (approximately 3000) on the surface of these BA-ER cells (28), it was possible the antibodies used to immunoprecipitate these complexes (both generated against the entire intracellular domains of the receptors) sterically hindered the binding of p110 on the one hand and Jak2 on the other. To test this, bIL-3 and biotinylated Ep, together with streptavidin-agarose and DSP cross-linking, were employed to isolate the receptor complexes. In vitro phosphorylation of these complexes revealed, once again, that the 110-kDa protein was clearly present within the IL-3R complex in these cells but either not present or not accessible for in vitro phosphorylation within the EpR complex (Fig. 7B). DISCUSSION In the present study, we have demonstrated that there is a Ser/Thr kinase constitutively associated with the IL-3R (as evidenced by the in vitro phosphorylation of p110 on serines FIG. 5. The 110-kDa serine/threonine kinase is tyrosine phosphorylated following IL-3 stimulation. A, B6SUtA 1 cells, treated with or without IL-3 for 5 min at 37°C, were lysed, and proteins were immunoprecipitated with ␣IL-3R antibodies. SDS-PAGE and Western analysis was then carried out with the ␣PY, 4G10 (left panel). The blot was stripped and reprobed with ␣IL-3R antibodies to demonstrate equal loading (right panel). B, B6SUtA 1 cells were treated with bIL-3, lysed without prior DSP cross-linking, and incubated with streptavidin beads. The beads were washed extensively, treated with azido-[ 32 P]ATP in the presence and absence of excess ATP, and then boiled in 50 l of 1% SDS. The eluted proteins were diluted to 1 ml with 0.5% Nonidet P-40 lysis buffer and subjected to immunoprecipitation with ␣PY beads. The ␣PY-immunoprecipitated proteins were eluted by boiling in SDSsample buffer for SDS-PAGE and autoradiography.
FIG. 6. The IL-3R ␤ subunit is phosphorylated on Ser/Thr residues at 4°C in response to IL-3. A, B6SUtA 1 cells, labeled with [ 32 P]orthophosphate, were treated with or without IL-3 for 10 min at 4°C and lysed; the proteins were subjected to immunoprecipitation with ␣IL-3R antibodies. The immunoprecipitated proteins were then electrophoresed on SDS-gels and subjected to autoradiography. B, the 140-kDa IL-3R ␤ subunit band from the gel was eluted and subjected to phosphoamino acid analysis.

FIG. 7.
The 110-kDa protein kinase is associated more with the IL-3R than with the EpR. A, B6SUtA 1 and BA-ER cells were stimulated with IL-3 or Ep and lysed; proteins were immunoprecipitated with ␣IL-3R or ␣EpR antibodies, subjected to in vitro phosphorylation, SDS-PAGE, and autoradiography. B, BA-ER cells were stimulated with bIL-3 or biotinylated Ep, cross-linked with DSP, and lysed; the proteins were exposed to streptavidin-agarose beads. Bound proteins were labeled with [ 32 P]ATP, boiled off with SDS, diluted, and subjected to immunoprecipitation with ␣PY beads, SDS-PAGE, and autoradiography. and threonines (Fig. 2, A and B)) and that this kinase is most likely the 110-kDa IL-3R-associated protein. Preliminary evidence for the latter comes from our in vitro phosphorylation studies with ␣IL-3R immunoprecipitates, which showed that this minor protein, as assessed by silver staining, was the most heavily labeled. Since many protein kinases have been shown to be heavily phosphorylated in vitro, this prompted us to investigate this possibility further. Further evidence for this p110 being a kinase was obtained from ␥-ATP-Sepharose binding studies (which demonstrated a marked enrichment of the 110-kDa protein from disrupted IL-3R complexes) and specific labeling of p110 with azido-[ 32 P]ATP. Our results are consistent with an earlier report by Schreurs et al. (44) who demonstrated that a Ser/Thr kinase remains associated with the IL-3R following bIL-3/streptavidin agarose enrichment of the latter. We also attempted to demonstrate that this protein was a kinase by SDS-PAGE renaturation assays (45), but, while this procedure successfully identified kinases present in total cell lysates, it did not renature the kinase(s) present in our ␣IL-3R immunoprecipitates. Characterization of this 110-kDa protein kinase with various kinase inhibitors revealed that it was markedly inhibited by the Ser/Thr kinase inhibitors, staurosporine and H7, but not by the protein kinase C-specific inhibitors compound 3 or chelerythrine, nor the protein kinase A-specific kinase inhibitor H89, nor the tyrosine kinase inhibitors genistein, herbimycin A, and tyrphostins B42 and 46. Interestingly, in this regard, when we added either H7 or staurosporine to ␣IL-3R immunoprecipitates to inhibit the in vitro phosphorylation of p110, we inhibited the Ser/Thr phosphorylation of all the other proteins in the complex, consistent with all the phosphorylations being carried out by one Ser/Thr kinase. Based on its molecular mass and its sensitivity to these kinase inhibitors, p110 does not appear to be either one of the known protein kinase C isoforms or protein kinase A and may therefore be an as yet uncharacterized Ser/Thr kinase. In this regard, IL-3 has been shown to modulate the activity of several Ser/Thr kinases, including Raf-1 (46,47), protein kinase C (48), and interferon-inducible double-stranded RNA-dependent kinase (49), but none of these kinases have been shown to associate with the IL-3R and none possess a molecular mass close to 110 kDa.
To determine whether the p110 reported here was the Ep and IL-3-induced major 90 -100-kDa tyrosine-phosphorylated protein previously reported (10,50,51), we co-electrophoresed ␣PY immunoprecipitates from Ep and IL-3-stimulated BA-ER cell lysates with our in vitro 32 P-labeled ␣IL-3R immunoprecipitates on two-dimensional O'Farrell gels and found that the p110 detected by autoradiography was distinct from the 90 -100-kDa protein detected by anti-PY Western blots (data not shown).
In summary, we have demonstrated that a group of proteins specifically and constitutively associates with the IL-3R and that one of these proteins is an as yet unidentified 110-kDa serine/threonine kinase. Comparison of IL-3R and the EpR immune complexes suggests that this kinase has a higher affinity for the IL-3R than for the EpR, and this may provide a clue to the differences seen when these two receptors are triggered in Ba/F3 cells.