Transmembrane Phosphoprotein Cbp Positively Regulates the Activity of the Carboxyl-terminal Src Kinase, Csk*

Csk (carboxyl-terminalSrc kinase) is a cytoplasmic tyrosine kinase that phosphorylates a critical tyrosine residue in each of the Src family kinases (SFKs) to inhibit their activities. Recently, we identified a transmembrane protein, Cbp (Csk-binding protein), that, when phosphorylated, can recruit Csk to the membrane where the SFKs are located. The Cbp-mediated relocation of Csk to the membrane may play a role in turning off the signaling events initiated by SFKs. To further characterize the Csk-Cbp interaction, we have generated a reconstituted system using soluble, highly purified proteins. Csk and phosphorylated Cbp were co-purified as a large protein complex consisting of at least four Csk·Cbp units. The addition of the phosphorylated, but not nonphosphorylated, Cbp to an in vitro assay stimulated Csk activity toward Src. Csk was also activated by a phosphopeptide containing the tyrosine in Cbp that binds to Csk (Tyr-314). Kinetic analysis revealed that Cbp or the phosphopeptide induced up to a 6-fold reduction in the K m for Src, indicating that the Csk·Cbp complex has a greater affinity for Src than free Csk. These findings suggest that Cbp is involved in the regulation of SFKs not only by relocating Csk to the membrane but also by directly activating Csk.

The Src family kinases (SFKs) 1 are nonreceptor protein tyrosine kinases (PTKs) that are associated with the inner surface of plasma membrane through their fatty-acylated amino termini (1). SFKs are known to act as molecular switches that regulate a variety of cellular events, including cell growth and division, cell attachment and movement, differentiation, survival, or death (2). SFKs are ordinarily present in an inactive state in which the phosphorylated carboxyl-terminal regulatory tyrosine binds to its own SH2 domain (3). In response to an external stimulus, an SFK is activated through dephosphorylation of the carboxyl-terminal tyrosine or through binding to another protein that displaces the intramolecular interaction. The phosphorylation of the regulatory tyrosine of SFK is known to be catalyzed by another PTK, Csk (4,5). In contrast, the phosphatases that activate SFKs have not yet been positively identified, although some candidate molecules have been proposed (6,7). To understand the regulation of SFKs, it is essential to clarify the regulation mechanism controlling the phosphorylation and dephosphorylation of the critical carboxylterminal tyrosine.
Csk is a cytoplasmic PTK consisting of an SH3, an SH2, and a kinase domain. Because it lacks an amino-terminal acylation signal and a carboxyl-terminal tyrosine, the regulatory mechanisms of Csk itself have remained unknown. A line of evidence has suggested that the SH2 and/or SH3 domain of Csk is essential for SFK regulation (8,9). The relocation of Csk to the membrane, specifically to regions where SFKs are active, was also observed (10). In addition, a membrane-targeted form of Csk, containing the myristoylation signal of Src, more actively suppressed SFK functions (11). These facts suggested the possible existence of a membrane factor that can recruit Csk to the membrane where SFKs are active. The importance of the SH2 domain of Csk further suggested that such a membrane factor might be a tyrosine-phosphorylated protein.
To test the hypothesis presented above, we searched for phosphoproteins that can bind tightly to the SH2 domain of Csk and identified a transmembrane phosphoprotein, Cbp (Csk-binding protein) (12). Cbp is involved in the membrane localization of Csk as well as in the Csk-mediated inhibition of Src. When phosphorylated on Tyr-314, Cbp can bind to Csk. Within the plasma membrane, Cbp is exclusively localized in the GM1 ganglioside-enriched detergent-insoluble membrane (DIM) domain, which is thought to play an important role in receptor-mediated signaling and where the majority of SFKs are localized (13)(14)(15)(16). These findings suggested that Cbp is a novel component of the regulatory mechanism controlling the activity of SFKs. To further evaluate the role of Cbp in the Csk-mediated regulation of SFKs, we examined the effect of Cbp binding on the kinase activity of Csk. The binding of phosphorylated Cbp or a phosphopeptide containing Tyr-314 could substantially elevate the affinity of Csk for Src.

EXPERIMENTAL PROCEDURES
Protein Expression-Rat Csk cDNA was inserted into pBlueBac4.5 vector and a recombinant baculovirus was generated according to the manufacturer's instruction manual (Invitrogen). The cDNAs encoding the cytoplasmic domains of Cbp (Cbp⌬N, amino acids 53-424) and mouse neuron type c-Src (Src, amino acids 9 -539) were inserted into the pFastBac HTa vector to generate His-tagged Cbp⌬N and His-tagged Src, respectively. The full-length Src cDNA was inserted into the pFast-Bac1 vector. Production of the recombinant viruses and the protein expression in Sf9 cells was carried out according to the instruction manual for Bac-to-Bac baculovirus expression system (Life Technologies, Inc.). Infections were done as follows. To prepare nonphosphorylated Cbp⌬N, the His-tagged Cbp⌬N virus was used alone. To prepare phosphorylated Cbp⌬N, the His-tagged Cbp⌬N virus was co-infected with full-length, untagged, Src virus. To prepare His-tagged Src as a substrate for Csk in vitro kinase assays, His-tagged Src virus was used * This work was supported in part by grants-in-aid from the Ministry of Education, Science, Sports and Culture of Japan and by a grant from the Program for Promotion of Fundamental Studies in Health Sciences of the Organization for Pharmaceutical Safety and Research. 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.
Protein Purification-To purify untagged Csk, Sf9 cells were lysed in a buffer consisting of 50 mM Tris-HCl, pH 7.4, 1 mM EDTA, 2% Nonidet P-40, 5 mM ␤-mercaptoethanol, 0.15 M NaCl, 10 g/ml aprotinin, 10 g/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride (PMSF). The lysate was centrifuged at 20,000 ϫ g for 30 min, and the 2-fold diluted supernatant was applied onto a Q-Sepharose FF (Amersham Pharmacia Biotech) column equilibrated with 50 mM Tris-HCl, pH 8.0, containing 1 mM EDTA, 5 mM ␤-mercaptoethanol, 0.01% Nonidet P-40, and 5% glycerol. After washing with the same buffer, a linear gradient of 0.075-0.35 M NaCl was applied. Active fractions were diluted 5-fold and applied onto an SP-Sepharose FF column followed by elution with a linear gradient of 0.05-0.25 M NaCl. The active fractions were concentrated on a HiTrapQ HT column and processed on a Superdex 200 pg column. For the phosphorylated form of His-tagged Cbp⌬N and the complex of Csk and His-tagged Cbp⌬N, lysates were made from coinfected cells in 20 mM Tris-HCl, pH 8.0, containing 1 mM EGTA, 10 mM MgCl 2 , 1% Nonidet P-40, 5 mM ␤-mercaptoethanol, 0.15 M NaCl, 1 mM Na 3 VO 4 , 10 g/ml aprotinin, 10 g/ml leupeptin, and 1 mM PMSF. The cleared lysate was applied onto a Q-Sepharose FF column equilibrated with Buffer A consisting of 20 mM Tris-HCl, pH 8.0, 5 mM ␤-mercaptoethanol, 0.01% Nonidet P40 and 5% glycerol, and the materials were eluted with a linear gradient of 0.15-0.45 M NaCl. Active fractions were directly applied onto a His-NTA column (Qiagen) equilibrated with 20 mM Tris-HCl, pH 8.0, containing 0.3 M NaCl, 20 mM imidazole, 0.01% Nonidet P40, and 5% glycerol. After washing, materials were eluted with 250 mM imidazole in the same buffer. The concentrated eluate was then separated on a Superdex 200 column equilibrated with Buffer A containing 0.5 mM Na 3 VO 4 and 0.4 M NaCl. For nonphosphorylated His-tagged Cbp⌬N and His-tagged Src, purification was carried out as described for phosphorylated Cbp⌬N except that MgCl 2 and Na 3 VO 4 were omitted from the buffers.
Immunoprecipitation and Pull-down Assays-The DIM fractions were prepared from neonatal rat brains as described previously (12) and solubilized in ODG buffer consisting of 50 mM Tris-HCl, pH 7.4, 1 mM EDTA, 1% Nonidet P40, 2% octyl-D-glucoside, 5 mM ␤-mercaptoethanol, 0.25 M NaCl, 1 mM Na 3 VO 4 , 10 mM NaF, 10 g/ml aprotinin, 10 g/ml leupeptin, and 1 mM PMSF. The cleared lysate was incubated with antibodies or GST-protein fused to the SH2 domain of Csk (GST-CskSH2) coupled with protein G-Sepharose or glutathione-Sepharose, respectively. The washed beads were then subjected to immunoblot analysis with anti-phosphotyrosine (pY) and anti-Cbp antibodies. Production of GST-CskSH2 and immunoblot analysis were done as described previously (12). In vitro phosphorylation of Cbp in the DIM fraction was carried out as described (12). The phosphopeptides were synthesized with ACT44omega (Advanced ChemTech) and purified by reverse-phase high pressure liquid chromatography.
Csk Assay-Csk or Csk⅐Cbp complex was incubated for 10 min at 30°C in a reaction mixture (20 l) consisting of 50 mM Tris-HCl, pH 7.4, 10 mM MgCl 2 , 10 M [␥-32 P]ATP (2 Ci), 0.5 mM Na 3 VO 4 , 0.25 M NaCl, and His-tagged Src in the presence or absence of effector. The reaction was terminated by the addition of 20 l of 2ϫ SDS sample buffer and separated by SDS-PAGE. The activity was visualized and quantified using a BAS300 Bioimage Analyzer (Fuji).

RESULTS AND DISCUSSION
Purification of the Csk⅐Cbp Complex-To characterize the biochemical features of the Csk-Cbp interaction, we employed a baculovirus expression system. The following recombinant baculoviruses were generated: Cbp lacking the amino-terminal membrane association domain and carrying a His tag (Cbp⌬N); untagged Csk; intact, untagged Src (as a donor of Cbp phosphorylation); and a truncated Src lacking its amino-terminal membrane association domain and carrying a His tag (His-Src, as a substrate for Csk). The phosphorylated Cbp⌬N (pCbp⌬N), nonphosphorylated Cbp⌬N (nCbp⌬N), His-Src, Csk⅐Cbp⌬N complex, and Csk were purified from infected insect cells through sequential column chromatographies as described under "Experimental Procedures" (Fig. 1A). pCbp⌬N migrated slowly on the SDS-polyacrylamide gel as an 85-kDa protein, whereas nCbp⌬N migrated as a 65-kDa protein. Because the molecular size of the His-tagged Cbp⌬N predicted from the cDNA sequence is about 43 kDa, the composition of amino acids and the phosphorylation states might greatly affect the mobility of the proteins on the gel (12). Co-expression of Cbp⌬N, Csk, and Src successfully induced an efficient phosphorylation of Cbp⌬N to generate a substantial amount of the Csk⅐Cbp⌬N complex applicable to biochemical analysis. As described previously (12), the interaction was so stable that it was resistant against high concentrations of salt (up to 1 M NaCl). During purification, the Src protein was completely eliminated from the complex, confirming that there is no close association between Cbp and Src. Upon gel filtration chromatography, the complex was eluted as a large protein complex with a molecular mass of ϳ440 kDa (Fig. 1B, upper panel). When analyzed for pCbp⌬N alone, it gave an apparent molecular mass of ϳ200 kDa (Fig. 1B, lower panel), suggesting at least that cytoplasmic domain of Cbp is capable of being oligomerized even without Csk association. From these observations, together with the calculated molecular masses of Csk (50 kDa) and His-tagged Cbp⌬N (43 kDa), it is predicted that the complex consists of at least four sets of Csk⅐Cbp units. Although further analysis should be undertaken to determine the actual composition of the complex, it seems likely that Csk is concentrated on the surface of the oligomerized Cbp. Indeed, gel filtration analysis of the DIM fraction from neonatal rat brain revealed that the native form of Csk⅐Cbp complex also behaved as a huge protein complex greater than 500 kDa (data not shown). This suggests that the Csk⅐Cbp complex is oligomerized in vivo as well.
Competition of Csk-Cbp Interaction with Phosphopeptide-To further define the condition of the Csk-Cbp interaction, we generated a set of phosphopeptides containing phosphorylation sites of Cbp ( Fig. 2A). Mutational analysis had shown that Tyr-314 of Cbp is a critical site for the binding to Csk (12). Other sites (Tyr-165, -183, -224, -381, and -409) were shown to be phosphorylated in brain extracts by peptide mass finger printing in a mass spectrometer (data not shown). When the phosphorylated Cbp was coimmunoprecipitated with Csk from the DIM fraction, a phosphopeptide containing Tyr-314 (Cbp-4) displaced the interaction under high salt condition (0.25 M NaCl) (Fig. 2B). Likewise, only Cbp-4 was capable of competing with the binding of Cbp to the SH2 domain of Csk (Fig. 2B). Cbp-4 could also interfere with the association of the His-tagged pCbp⌬N with Csk (data not shown). When these assays were carried out in the presence of lower concentrations of salt (0.15 M NaCl), Cbp-3 (Tyr-224) could also partially compete with the interaction (data not shown). These findings reveal that Tyr-314 of Cbp is indeed the critical binding site to the SH2 domain of Csk and that the surrounding 4 -5 amino acids are sufficient to create the binding site.

Activation of Csk by Phosphorylated Cbp or
Phosphopeptide-To examine the effect of Cbp binding on the kinase activity of Csk, pCbp⌬N protein or Cbp-4 phosphopeptide was added to the in vitro kinase assay for Csk. As shown in Fig. 3, A and B, the levels of Src phosphorylation were increased up to about 4-fold depending on the amount of pCbp⌬N but not nCbp⌬N. Although the extent of activation was less, the Cbp-4 phosphopeptide could also activate Csk with the activation being saturated at around 3 g/ml (Fig. 3, A and B). The phosphopeptide (Cbp-2), which could not compete with the Csk-Cbp interaction, had no effect on Csk activity. A similar activation was observed when an artificial substrate, a random polymer of glutamate and tyrosine (polyEY), was used as substrate (data not shown). When native forms of phosphorylated  ). B, quantitative analysis of the activation of Csk by pCbp⌬N and Cbp-4. Radioactivity of the band corresponding to His-Src was quantified and plotted versus concentrations of the effectors. C, effect of brain Cbp on the kinase activity of Csk. Nonphosphorylated Cbp (nCbp) was immunoprecipitated from the DIM fraction of brain prepared in the absence of Na 3 VO 4 . Phosphorylated Cbp (pCbp) was prepared from the DIM fraction of brain that had been treated with ATP in vitro (12). Cbp and its tyrosine phosphorylation were detected by immunoblot (IB) analysis (left two panels). The Cbp immunoprecipitates (IP) were added to the Csk assay with increasing amounts of Csk (0, 375, and 750 ng/ml) (right two panels). Activation of Csk was evident when a low concentration of Csk was used. A slight phosphorylation of nCbp was observed in this assay, probably due to a trace of remaining pCbp. and nonphosphorylated Cbp prepared from the DIM fraction of brain were used as effectors, activation of Csk was also observed (Fig. 3C). These findings suggest that the complex formation with phosphorylated Cbp may modulate the activity of Csk in vivo.
In these assays, we also observed phosphate incorporation into phosphorylated but not nonphsphorylated Cbp (Fig. 3, A  and C). This suggests that Csk cannot phosphorylate nonphosphorylated Cbp but that the tight interaction with phosphorylated Cbp enables Csk to phosphorylate Cbp. Previously, we examined the contribution of SFKs to Cbp phosphorylation in the DIM fraction. The compound PP2, a potent inhibitor of SFK (18), could completely inhibit the Cbp phosphorylation, whereas an overdose of Csk gave a partial inhibition, presumably by inhibiting SFKs (12). This left open the possibility that Csk may contribute to the phosphorylation of Cbp. Further study is needed to determine whether the phosphorylation of Cbp by Csk detected in vitro is physiological.

Mode of Csk Activation by Cbp Binding-To examine how
Cbp activates Csk, we performed a kinetic analysis using the purified Csk⅐Cbp⌬N complex. A substrate saturation curve of the complex was compared with that of floating Csk (Fig. 4A). From the Lineweaver-Burk plots (Fig. 4B), the K m value of the complex for Src was calculated to be ϳ0.21 M, which was about 6 times less than that of free Csk (1.27 M). Binding to Cbp-4 could also reduce the K m value for Src (0.59 M). These observations demonstrate that Cbp binding could modulate the conformation of Csk to increase the affinity for protein (Src) or peptide (polyEY) substrates. It was shown previously by mutational analysis that the SH2 and SH3 domains of Csk are important for optimal kinase activity but apparently play no direct or specific role in substrate recognition (19). In this study, however, binding to phosphorylated Cbp or phosphopeptide could modulate the affinity of Csk for the native substrate Src. Structural analysis should help elucidate the molecular basis of these phenomena.
In this study, we show that 1) Csk and the cytoplasmic domain of phosphorylated Cbp formed a stable complex that could be purified to near homogeneity, 2) the Csk⅐Cbp complex was oligomerized to be a large molecular complex, and 3) the complex formation could enhance the affinity of Csk toward Src. Taken together, these findings predict that Cbp acts as a positive regulator for Csk by recruiting Csk to the membrane where SFKs are present, by concentrating Csk on the oligomerized Cbp, and by directly activating Csk. By these combined mechanisms, the efficient regulation of SFKs may be achieved in vivo. To apply this hypothesis to an in vivo condition, the timing and mechanism of Cbp phosphorylation needs to be investigated. In this context, the roles of potentially responsible kinases, SFKs, and counteracting protein tyrosine phosphatases are now under investigation.