Using CD45-deficient clones from the immature B cell line, WEHI-231, we previously demonstrated that CD45 selectively dephosphorylates the Src-family protein tyrosine kinase Lyn and inhibits its kinase activity. To further define the mechanisms of CD45 action on Lyn, we metabolically labeled Lyn from CD45-positive and -negative WEHI-231 cells and analyzed cyanogen bromide fragments by SDS-PAGE analysis. Phosphoamino acid analysis confirmed that Lyn is tyrosine phosphorylated with little serine or threonine phosphorylation. In CD45-negative cells, two bands at 8.2 and 4.1 kDa were phosphorylated in the absence of B cell Ag receptor (BCR) ligation. The 8.2-kDa band corresponded to a fragment containing the positive regulatory site (Tyr397), as assessed by its size and its phosphorylation in an in vitro kinase assay. The 4.1-kDa band was phosphorylated by COOH-terminal Src kinase, suggesting that it contains the COOH-terminal negative regulatory site (Tyr508). CD45 was also shown to dephosphorylate autophosphorylated Lyn in vitro. Thus, CD45 dephosphorylates not only the negative but also the positive regulatory tyrosine residues of Lyn. Furthermore, coimmunoprecipitations using anti-Igα Ab demonstrated that Lyn associated with the resting BCR was constitutively phosphorylated and activated in CD45-negative cells. In the parental cells, both regulatory sites were phosphorylated on BCR ligation. Taken collectively, these results suggest that CD45 keeps both BCR-associated and total cytoplasmic pools of Lyn in an inactive state, and a mechanism by which Lyn is activated by relative reduction of CD45 effect may be operative on BCR ligation.
Engagement of B cell Ag receptor (BCR)5 by anti-IgM Ab or multivalent Ags induces a rapid increase in tyrosine phosphorylation of a number of cellular proteins (1, 2). Signals are propagated downstream and ultimately control gene activation leading to activation, cell death, or anergy in B cells, depending on the differentiation stage, the nature of Ags, and the presence or absence of costimulation (3, 4). Regardless of the context in which stimulation occurs and the final outcome, BCR signaling is dependent on the proximal activation and recruitment of protein tyrosine kinases (PTKs): Src-family PTKs (Lyn, Blk, Lck, and Fyn); and Syk (5, 6, 7).
Extensive analyses have demonstrated that one of the crucial enhancers of BCR signaling is the receptor-type protein tyrosine phosphatase, CD45 (8, 9, 10, 11, 12). CD45 has been implicated in both T and B cell activation based on experiments using CD45-deficient cells (13, 14, 15, 16, 17) and cells from CD45 gene-targeted mice (18, 19). It has been shown that CD45 in T cells dephosphorylates Lck and Fyn at the COOH-terminal negative regulatory sites, Tyr505 (20, 21, 22, 23, 24) and Tyr531 (25, 26, 27), respectively. From these findings, a view has emerged that the binding of the phosphorylated COOH-terminal tyrosine residue of a Src-family PTK to its own SH2 domain causes the PTK to assume an inactive conformation. CD45 activates the kinase by dephosphorylating the COOH-terminal tyrosine, thereby releasing the inhibitory intramolecular conformation. However, there are also discrepant results in which Lck and Fyn are hyperphosphorylated and activated in CD45-deficient YAC-1 T cell clones. In these experiments, the sites of phosphorylation include both the negative COOH-terminal tyrosine (Tyr505) and the positive autophosphorylation site (Tyr394) of Lck (28, 29).
We have previously demonstrated that in CD45-deficient clones generated from immature WEHI-231 cells, Lyn is selectively hyperphosphorylated and activated in the absence of BCR ligation. Furthermore, receptor stimulation did not significantly enhance phosphorylation and activation of Lyn (30). BCR-induced Ca2+ mobilization, growth arrest, and apoptosis were also negatively regulated by CD45 in WEHI-231 cells (16). In contrast, CD45 exerted positive, crucial effects on BCR-induced growth arrest and, to a lesser extent, BCR-induced tyrosine phosphorylation in mature BAL-17 B cells (17). Recent studies on two different B cell lines revealed somewhat different mode of action of CD45. In CD45 gene-targeted clones from the chicken DT40 B cell line, Lyn was shown to be hyperphosphorylated on BCR ligation at the COOH-terminal negative regulatory residue as well as the positive regulatory tyrosine. However, the kinase activity of Lyn was reduced as compared with the activity of wild-type cells (31). In CD45-negative, BCR-transfected J558Lμm3 plasmacytoma cells, only the COOH-terminal tyrosine of Lyn was constitutively hyperphosphorylated, and the enzymatic activity of Lyn was significantly lower than its counterpart isolated from CD45-positive cells (32). The reasons for these differences have not yet been well defined.
This study was initiated to elucidate the mechanisms whereby CD45 regulates Lyn tyrosine kinase in immature B cells. Studies on Lyn in immunoprecipitates from 32P-labeled WEHI-231 CD45-deficient cells with anti-Igα Ab demonstrated that Lyn, total cellular as well as BCR-associated, was constitutively tyrosine phosphorylated and activated. Cyanogen bromide (CNBr) cleavage mapping of Lyn clearly demonstrated that both the COOH-terminal negative regulatory residue (Tyr508) and the positive regulatory residue (Tyr397) were phosphorylated before BCR stimulation in CD45-deficient cells. In the parental cells, by contrast, phosphorylation of not only Tyr397 but also Tyr508 was induced by BCR ligation. Thus, in WEHI-231 cells, the main function of CD45 is to dephosphorylate two major regulatory residues, Tyr508 and Tyr397, and inactivate Lyn kinase in both the receptor-associated and total cellular pools before BCR ligation. These results also suggest a mechanism by which BCR ligation reduces the negative effect of CD45, thereby inducing the phosphorylation and activation of Lyn kinase.
Materials and Methods
Abs and reagents
Goat (Fab′)2 fragments of anti-mouse IgM Ab and intact anti-mouse IgM Ab were purchased from Cappel, Organon Teknika (Durham, NC). Polyclonal Ab against Lyn was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-phosphotyrosine (PY) mAb (4G10) was purchased from United Biotechnology (Lake Placid, NY). Anti-Igα Ab was raised by immunizing rabbits with GST fusion proteins containing the cytoplasmic tails of the protein. Abs were then purified using the immunogen. Alkaline phosphatase (AP)-conjugated goat anti-mouse IgG and mouse anti-rabbit IgG were obtained from Bio-Rad (Richmond, CA), and Jackson Immunoresearch Laboratories (West Grove, PA), respectively.
Enolase was purchased from Sigma (St. Louis, MO).
Cell stimulation, immunoprecipitation, and Western blot analysis
Cells were harvested from log phase cultures, resuspended in fresh prewarmed RPMI 1640 containing 10% FBS supplemented with 20 mM HEPES, and incubated for 3 h at 37°C. The cells were then stimulated with 25 μg/ml F(ab′)2 fragments of anti-IgM Ab for 1 min, and the reactions were terminated with ice-cold PBS containing 2 mM Na3VO4 and 2 mM EDTA (PBS-VE). The cells were centrifuged and solubilized in TNE lysis buffer (1% Nonidet P-40, 20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 2 mM Na3VO4, 2 mM EDTA) or digitonin lysis buffer (1% digitonin, 20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 2 mM Na3VO4, 2 mM EDTA) supplemented with Protease Inhibitor Mixture (Boehringer Mannheim GmbH, Mannheim, Germany). The lysates were centrifuged at 10,000 × g at 4°C for 30 min, and the supernatants were subjected to further analysis.
Immunoprecipitation and Western blot analyses were performed as previously described (30). Each sample was immunoprecipitated with protein G-Sepharose (Pharmacia Biotech, Uppsala, Sweden) coupled with Ab against Lyn or Igα. Anti-Lyn immunoprecipitates were boiled with SDS sample buffer under reducing conditions and subjected to 10% SDS-PAGE. Separated proteins were blotted onto nitrocellulose membranes, and the membranes were incubated overnight with anti-PY mAb and anti-Lyn Ab, followed by AP-conjugated goat anti-mouse IgG and mouse anti-rabbit IgG, respectively. The blots were visualized by developing them with an AP Conjugate Substrate kit (Bio-Rad, Hercules, CA). Anti-Igα immunoprecipitates were washed with 0.5% digitonin lysis buffer, and dissolved in 1 μl of 1% Nonidet P-40, 0.1%SDS lysis buffer. The supernatants were immunoprecipitated with anti-Lyn Ab- or normal rabbit IgG-coated protein G beads. The precipitates were washed with TNE and then subjected to SDS-PAGE or to in vitro kinase assays. The intensity of each band was measured with a Bio-Rad densitometer.
In vitro kinase assay
In vitro kinase assay was performed as previously described (30). Cells were solubilized in TNE or digitonin lysis buffer, and the supernatants were immunoprecipitated with anti-Lyn Ab or anti-Igα and anti-Lyn Abs. The immunoprecipitates were washed with lysis buffer and then with kinase buffer (20 mM HEPES, pH 7.5, 150 mM NaCl, 10 mM magnesium acetate, 20 mM MnCl2). For in vitro kinase assays, 0.35 MBq [γ-33P]ATP (37–110 TBq/mmol, Amersham, Arlington Heights, IL) in kinase buffer containing 10 mM cold ATP was added to the anti-Lyn immunoprecipitates. For anti-Igα and anti-Lyn immunoprecipitates from digitonin lysates, reactions were performed in the same buffer without cold ATP together with exogenous substrate enolase. To assess phosphorylation of Tyr508 of Lyn, 1 μl of recombinant COOH-terminal Src kinase (Csk) (33) was added to the anti-Lyn immunoprecipitates during the in vitro kinase assay. The reactions were terminated by adding SDS sample buffer, and the samples were subjected to 10% SDS-PAGE analysis. The resulting gels were treated with 1 N KOH at 60°C for 90 min to hydrolyze phosphoserine and phosphothreonine, dried, and analyzed with a BAS 2000 Bio-Imaging Analyzer (Fuji Photo Film, Tokyo, Japan).
Metabolic labeling and CNBr digestion
Cells were washed with phosphate-free RPMI 1640 and cultured at 4 × 106 cells/ml for 3 h at 37°C in phosphate-free medium supplemented with 10% dialyzed FBS. The cells were harvested and labeled with 370 MBq of [32P]orthophosphate in 2 ml of medium for 1 h at room temperature. After addition of 8 ml of phosphate-free medium to the culture, cells were incubated for 3 h at 37°C. Labeling was stopped by adding ice-cold PBS-VE, and the labeled cells were lysed and immunoprecipitated with anti-Lyn Ab as described above. Bound proteins were eluted in SDS sample buffer, applied to 10% SDS-PAGE, and transferred onto a nitrocellulose membrane. The Lyn bands (p53 and p56), which were identified by autoradiography and subsequent immunoblotting, were excised and eluted with 250 μl of 150 μg/ml CNBr in 70% formic acid at room temperature for 2 h. The eluted proteins were dried in a vacuum concentrator (Tomy Seiko, Tokyo, Japan) to remove formic acid, and the final pellets were dissolved in SDS sample buffer and applied to SDS-PAGE in 15–25% gradient gel. The separated proteins were transferred to a polyvinylidine difluoride (PVDF) membrane and subjected to BAS 2000 analysis.
Phosphoamino acid analysis
Phosphoamino acid analysis was performed as described earlier (34). 32P-labeled proteins were excised from a PVDF membrane and hydrolyzed with 6 N HCl at 110°C for 2 h. The samples were dried, dissolved in 5 μl of distilled water, and spotted onto TLC plates. Electrophoresis was conducted in pH 1.9 buffer for 20 min at 1.5 kV. After drying, the plates were subjected to second dimension electrophoresis in pH 3.5 buffer for 16 min at 1.3 kV. Separated phosphoamino acids were analyzed with a BAS 2000.
In vitro dephosphorylation assay
WEHI-231 cells stimulated with or without anti-IgM Ab were lysed with TNE, and Lyn was immunoprecipitated and subjected to in vitro kinase assays to phosphorylate the autophosphorylation site. CD45 was immunoprecipitated from WEHI-231 cells by using anti-CD45 mAb. The immunoprecipitates were washed with TNE, and CD45 was eluted with 50 μl 0.17 M glycine-HCl buffer, pH 2.0. The eluted supernatants were neutralized immediately by adding 0.45 ml 1 M Tris-HCl buffer, pH 9.0, and dialyzed to PBS. A phosphatase assay was performed by incubating autophosphorylated Lyn with 0.5 ml purified CD45 for 60 min at 37°C. After incubation, the precipitates were washed with TNE and subjected to SDS-PAGE. The results were visualized by autoradiography.
CD45 dephosphorylates and inactivates Lyn tyrosine kinase
In the immature B cell line WEHI-231, absence of CD45 led to constitutive hyperphosphorylation and activation of Lyn, but not other PTKs including Lck, Blk, Btk, and Syk (30) (T.K. and H.Y., unpublished observations). Representative results from anti-PY blot analyses and in vitro kinase assays of Lyn in WEHI-231 and its CD45-deficient clone (10-5) are shown in Fig. 1 A. Anti-IgM stimulation of the parent cells induced a 7.7-fold increase in tyrosine phosphorylation, but in the CD45-deficient clone, tyrosine phosphorylation was constitutively elevated 3.6-fold and was not enhanced further by BCR ligation. The kinase activity of Lyn was induced only after BCR ligation in the parent, whereas Lyn was constitutively activated (3.8-fold) in CD45-deficient cells, and the activity was slightly increased (to 5.5-fold) on BCR ligation.
Phosphoamino acid analysis was also performed on Lyn in WEHI-231 and 10-5 cells. Cells metabolically labeled with [32P]orthophosphate were cultured with or without 25 μg/ml anti-IgM Ab for 1 min, immunoprecipitated with anti-Lyn Ab, and subjected to SDS-PAGE analysis. Phosphoamino acid analysis on 32P-labeled Lyn demonstrated that Lyn was phosphorylated exclusively on tyrosine residues irrespective of BCR ligation, and in 10-5 cells, tyrosine phosphorylation of Lyn was enhanced 6.4-fold even in the absence of BCR ligation (Fig. 1 B). These results suggest that CD45 selectively dephosphorylates and inactivates Lyn in WEHI-231 cells. Negative regulation of a Src-family kinase by CD45 is in apparent contrast to the prevailing view that CD45 dephosphorylates the COOH-terminal negative regulatory tyrosine residue, thereby activating the kinase activity (8, 11).
Both negative and positive regulatory tyrosine residues are dephosphorylated by CD45
To address the mechanisms by which CD45 regulates Lyn activity, the dephosphorylation sites were determined by the CNBr cleavage method. CD45-positive and -negative WEHI-231 cells were metabolically labeled with [32P]orthophosphate, and Lyn immunoprecipitations were treated with CNBr. The resulting fragments were resolved by SDS-PAGE. CNBr cleavage of Lyn is expected to yield a fragment of 4.1 kDa containing the COOH-terminal negative regulatory tyrosine residue (Tyr508) and a fragment of 8.2 kDa containing the autophosphorylation site (Tyr397), among others (Fig. 2,A). In the parental cells, BCR ligation induced phosphorylation of 8.2- and 4.1-kDa fragments by 2.7- and 3.5-fold, respectively (Fig. 2,B). In CD45-deficient cells, however, phosphorylation of the 8.2-kDa fragment was ∼2-fold greater than control even in the absence of anti-IgM stimulation, and the 4.1-kDa fragment was even more strongly phosphorylated (4.1-fold). The phosphorylation levels of the respective fragments were not significantly enhanced by BCR stimulation in the CD45-deficient clone, 10-5. The 8.2-kDa fragment corresponded to a band containing the positive regulatory tyrosine residue Tyr397, as assessed by its size and phosphorylation in an in vitro kinase assay (Fig. 2C). The 4.1-kDa fragments from both WEHI-231 and 10-5 were phosphorylated by Csk (Fig. 2 D), suggesting that it contained the negative regulatory residue Tyr508. These results suggest that BCR ligation induces phosphorylation of Lyn not only at the autophosphorylation site but also at the negative regulatory tyrosine residue. The net result of phosphorylation at both sites is an activation of the kinase. Furthermore, in the absence of CD45, both the positive and negative regulatory tyrosine residues of Lyn were constitutively phosphorylated; nevertheless, the kinase activity was increased.
To firmly establish that CD45 dephosphorylates the positive regulatory tyrosine residue of Lyn, we examined whether CD45 directly dephosphorylates Tyr397 of Lyn in vitro. Lyn was immunoprecipitated from WEHI-231 cells and phosphorylated by in vitro kinase assays. As shown in Fig. 2,C, Lyn prepared by this method was phosphorylated predominantly at Tyr397 but not at Tyr508. When incubated with autophosphorylated Lyn for 60 min, CD45 strongly dephosphorylated Lyn (Fig. 3, lane 2 vs lane 4). Thus, CD45 dephosphorylates not only the negative but also the positive regulatory tyrosine residues both in vitro and in vivo.
Lyn in BCR complex is constitutively phosphorylated and activated in a CD45-deficient clone
Previous reports suggested that CD45 exerts its regulatory effects differently on the total pool and the TCR- or CD4-associated pools of Src-family PTKs (27, 35). It is therefore possible that the observed regulation of CD45 in WEHI-231 applied only to the total pool of Lyn and not to the more physiologically relevant subpool of Lyn associated with the BCR. To exclude this possibility, we examined the phosphorylation state and the kinase activity of Lyn in the BCR complex. WEHI-231 and its CD45-deficient clone 10-5 were labeled with [32P]orthophosphate. After incubating with or without 25 μg/ml anti-IgM Ab for 1 min, cells were lysed with digitonin and immunoprecipitated with anti-Igα Ab. The immunoprecipitates were dissolved in 1% Nonidet P-40, immunoprecipitated with anti-Lyn Ab or control IgG, and subjected to SDS-PAGE analysis. The results revealed that on BCR ligation, p53Lyn and p56Lyn were phosphorylated in the parental cells (Fig. 4,A, lane 2) but that in the CD45-deficient cells, Igα-associated Lyn was hyperphosphorylated even in the absence of BCR stimulation (Fig. 4,A, lane 3). Control IgG did not precipitate Lyn (Fig. 4 A, lanes 5 and 6). Similar results were obtained with anti-Igβ immunoprecipitation (data not shown).
The enzymatic activity of Lyn in BCR complexes was examined after digitonin lysis and immunoprecipitation with anti-Igα Ab. Compared with the parental cells, Lyn activity, both autophosphorylation and phosphorylation of enolase was significantly increased even before BCR ligation in the CD45-deficient clone 10-5 (Fig. 5 A, lanes 1, 2 vs lanes 3, 4). These results suggest that both receptor-associated and not associated Lyn pools were constitutively hyperphosphorylated and activated in the absence of CD45. Thus, CD45 dephosphorylates both the positive and negative regulatory tyrosine residues of Lyn, and the net effect is inactivation of the kinase.
Our previous studies showed that CD45 inhibits Lyn kinase activity in WEHI-231 cells and negatively regulates BCR-initiated effector phenomena such as growth arrest and apoptosis (16, 30). To further characterize the mechanisms by which CD45 regulates a Src-family PTK, we examined the relationship between the phosphorylation state of regulatory tyrosine residues of Lyn and the enzymatic activity. The results demonstrate that in the absence of CD45, Lyn is activated and hyperphosphorylated exclusively at tyrosine residues (Fig. 1). Phosphorylation was observed in the absence of BCR stimulation and occurred on both the negative regulatory Tyr508 and the positive regulatory Tyr397 (Fig. 2). This negative regulation was observed not only in the total cellular Lyn pool but also in the more physiologically important pool of kinases associated with the BCR (Figs. 4 and 5). Thus, CD45 dephosphorylates the two major regulatory tyrosine residues and inactivates Lyn kinase in WEHI-231 cells.
Accumulating evidence suggests that Src-family PTKs are regulated in part by tyrosine phosphorylation of the inhibitory site in the COOH-terminal tail and the stimulatory site in the kinase domain (36). The different phosphorylation states stabilize a repressed or an activated conformation respectively. Csk phosphorylates the COOH-terminal tyrosine, keeping Src-family PTKs in a repressed conformation (33). Therefore, one way to activate Src-family PTKs is by dephosphorylating the COOH-terminal tyrosine. Mutations in the kinase and SH2 and SH3 domains led to activation of Src, suggesting that conformational changes in these regions are also important factors for activating Src-family PTKs (36). Our results showing that a Src-family PTK can be active even if its negative regulatory COOH-terminal tyrosine is phosphorylated appear to contradict the conventional model of regulation. However, there are several explanations for our findings. One is that the phosphorylation state of the positive regulatory tyrosine is dominant over that of the negative regulatory tyrosine as has been suggested previously (29). An alternative possibility is that in Src-family PTKs that have phosphorylated COOH-terminal tyrosines, activation can occur by a phosphorylation-independent mechanism. It is also possible that the observed phosphorylation at both sites could be caused by the presence of two populations of Lyn, each phosphorylated at different residues. In this case, the enhanced kinase activity is a reflection of the dominant presence of Lyn phosphorylated at the stimulatory Tyr397. This possibility has not been examined in this or any other studies. Given that phosphorylation at the inhibitory Tyr508 was 2.2-fold higher than that at the stimulatory Tyr397 in activated Lyn (Fig. 2 B), we think this possibility is unlikely.
It has been argued that the main function of CD45 is to activate Src-family PTKs. A consensus model is that Csk phosphorylates the COOH-terminal negative regulatory tyrosine residue of Src-family PTKs, keeping PTKs in an inactive conformation, and that on Ag receptor ligation CD45 dephosphorylates the negative regulatory tyrosine, activating their enzymatic activity. This was based on the initial descriptions of CD45-negative T cell lines in which the Src-family PTKs were found to be hyperphosphorylated at their COOH-terminal tyrosines and inactive (20, 21, 22, 23, 24, 25, 26, 27, 37). However, subsequent studies revealed that the function of CD45 might be more complex. In CD45-negative human HPB-ALL T cells, anti-CD4-induced activation and tyrosine phosphorylation of Lck, particularly the CD4-associated pool, was higher than those of CD45-positive cells (35). Studies on three T cell lines, YAC-1, SAKRTLS, and HPB-ALL, demonstrated that Lck and Fyn are constitutively hyperphosphorylated and paradoxically activated in the CD45-negative clones and that in vitro exposure of CD45 to Lck leads to decreases, rather than increases, in the kinase activity (28). CNBr cleavage mapping showed that Lck in the CD45-negative clones is hyperphosphorylated at both the negative regulatory tyrosine residue (Tyr505) and the autophosphorylation site (Tyr394) (28). Phosphoamino acid analysis confirmed that the increased phosphorylation in CD45-negative YAC-1 cells is restricted to tyrosine residues, Tyr505 and, to a lesser extent, Tyr394 (29). Further mutational analysis revealed that mutation of Tyr505 to phenylalanine results in increased kinase activity whereas mutation of Tyr394 to phenylalanine decreases enzymatic activity. The double mutation, Tyr505-Phe and Tyr394-Phe, led to inactivation of the kinase (29). These results suggest that without CD45, phosphorylation at Tyr394 induces activation of Lck despite hyperphosphorylation at Tyr505 and that the phosphorylation state of Tyr394 may have a dominant role in the regulation of Lck.
In B cells, there have been discrepant reports as well. Lyn from CD45-negative, BCR-reconstituted J558Lμm3 plasmacytoma cells was shown to be hyperphosphorylated at the COOH-terminal tyrosine residue. The kinase was neither activated nor recruited to the BCR complex (32). A study on the chicken DT40 B cell line showed that CD45 dephosphorylates both the positive and the negative regulatory tyrosine residues of Lyn but the activity of Lyn is enhanced in the presence of CD45 and that BCR-induced tyrosine phosphorylation and Ca2+ mobilization are severely impaired in the absence of CD45 (31). Although the sites dephosphorylated by CD45 are different, both studies underscore the importance of dephosphorylation of the COOH-terminal tyrosine of Lyn in certain processes of BCR-mediated activation. In contrast, our previous reports (16, 30), together with the present study, propose a different mode of CD45 action, in which CD45 dephosphorylates both Tyr508 and Tyr397 of Lyn and keeps the kinase in an inactive state, exerting negative modulatory effects on BCR-induced Ca2+ mobilization, growth arrest, and apoptosis. It is also important that BCR ligation induces phosphorylation of not only Tyr397 but also Tyr508, yet activates the kinase. All these results suggest that there must be a mechanism whereby BCR ligation negates the effect of CD45, thus inducing phosphorylation of two regulatory sites and activation of the kinase.
What accounts for all the phenotypic differences is not clearly elucidated at present. Recent crystal structural and functional analyses on the inactive form of Src-family PTKs, c-Src and Hck (38, 39, 40), may give us an important hint as to the role of CD45 in the regulation of this family of PTKs. These studies demonstrated that although the SH2 domain binds to the phosphorylated COOH-terminal tail, this interaction does not block the catalytic site. Rather, the kinase is probably inactive because the linker sequence between the catalytic and SH2 domains binds the SH3 domain (38, 39, 41). Furthermore, the PTK that is phosphorylated at the COOH-terminal tyrosine was shown to be activated by disturbing these intramolecular interactions with exogenous SH2 and SH3 ligands (40). Thus, regulation of Src-family PTKs seems to be very complex such that the state of phosphorylation at the two major regulatory tyrosine residues is but one determiner of activation. Various factors, including unknown SH2 and SH3 ligands, contribute to the net enzymatic activity. It may also be possible that tyrosine phosphorylation at the two sites is a consequence of activation or inactivation induced by conformational alterations that dictate the accessibility of Src-family PTK phosphorylation sites to other PTKs and protein tyrosine phosphatases (36). In light of these structure-function relationships, our findings are not necessarily idiosyncratic but reflect a common regulatory activity of CD45. This is supported by observations in other CD45-negative B cell clones in which BCR-induced tyrosine phosphorylation of total cellular proteins is slightly reduced, but phosphorylation of a few species of proteins is almost completely defective (17). Differences observed in the PTK activity in a variety of CD45-deficient cells may be explained partly by the different availability of molecules within the cell capable of affecting the conformation of the Src-family PTKs (42).
In summary, CD45 has an inhibitory effect on Lyn by dephosphorylating the autophosphorylation site as well as the COOH-terminal regulatory tyrosine in immature WEHI-231 cells. This regulation exerted on not only total cellular Lyn but also the BCR-associated pool of Lyn. Furthermore, the fact that both regulatory tyrosines are phosphorylated and activated by BCR ligation suggests that BCR signaling machinery may have a mechanism by which the negative effect of CD45 is somehow inhibited.
We thank Drs. James Clements and Gary Koretzky (University of Iowa, Iowa City, IA) for advice on phosphoamino acid analysis and Dr. Philip Cohen (University of Dundee, Dundee, U.K.) for critical comments on CNBr mapping data.
This work was supported in part by grants-in-aid for Scientific Research and for International Scientific Research from the Ministry of Education, Science, Sports and Culture.
Abbreviations used in this paper: BCR, B cell Ag receptor; PTK, protein tyrosine kinase; PY, anti-phosphotyrosine; AP, alkaline phosphatase; PVDF, polyvinylidine difluoride.