INTRODUCTION

p56^lck, a member of the Src family of non-receptor tyrosine protein kinases (1, 2), is expressed predominantly in T cells. Lck function is critical both for T-cell development in the thymus (3, 4) and activation of mature T cells in the periphery by antigen (5, 6). Lck stably associates with the inner surface of the plasma membrane as a result of myristoylation of Gly-2 and palmitoylation of Ser-3 and Ser-5 (7-10). There it binds to the T-cell receptor-associated glycoproteins CD4 and CD8 as well as other proteins through its unique amino terminus (11-15). The activity of Lck is regulated by phosphorylation of two conserved tyrosine residues. Tyr-505 (equivalent to Tyr-527 in c-Src) is located near the carboxyl terminus of Lck and, when phosphorylated, associates intramolecularly with the SH2 domain in the amino-terminal half of the protein. This helps stabilize Lck in a conformation that, biologically, is relatively inactive (16-20). In the absence of phosphorylation at Tyr-505, intramolecular binding of the carboxyl terminus to the SH2 domain does not occur, and Lck exhibits increased activity in vivo. In contrast, phosphorylation of Tyr-394 (equivalent to Tyr-416 in c-Src) stimulates the catalytic activity of Lck (21-23). Phosphorylation of Tyr-394 allows the formation of hydrogen bonds between the phosphate of Tyr(P)¹-394 and amino acid residues in the catalytic cleft. These interactions allow the enzyme to assume a conformation resembling that of activated cyclic AMP-dependent protein kinase A (19, 20, 24-26).

We have previously demonstrated that hydrogen peroxide, a potent activator of Lck, acts by inducing phosphorylation of Lck on Tyr-394 (22, 37). It is likely that the effects of exposing cells to H₂O₂ are mediated by global inhibition of tyrosine phosphatases (27-31). The increase in phosphorylation of Lck at Tyr-394 that we observe in the presence of H₂O₂ may therefore result largely from reduced dephosphorylation of this site. Our previous work did not address the question of whether or not activation of Lck by H₂O₂-induced phosphorylation of Tyr-394 required dephosphorylation of Tyr-505. Thus, the extent to which the H₂O₂-activated population of Lck molecules was phosphorylated on Tyr-505 was unclear. Here we show that the population of Lck that is phosphorylated on Tyr-394 in response to H₂O₂ exposure is also phosphorylated at Tyr-505. Therefore, dephosphorylation of Tyr-505 and untethering of the SH2 domain is not a prerequisite for either phosphorylation of Lck at Tyr-394 or activation of the kinase by Tyr-394 phosphorylation. These results indicate that phosphorylation of Tyr-394 positively regulates Lck activity and is dominant over any negative regulation induced by phosphorylation of Tyr-505.

EXPERIMENTAL PROCEDURES

Jurkat human leukemic T cells were maintained in RPMI medium (Cellgro) supplemented with 10% fetal calf serum (HyClone). Rat 208F fibroblasts were maintained in Dulbecco-Vogt modified Eagle's medium supplemented with 10% calf serum.

H₂O₂ stimulation was achieved by direct addition of 1 M H₂O₂ to 10⁷ Jurkat cells (10⁶ cells/ml) to a final concentration of 5 mM. After incubation at 37 °C for 15 min, the cells were washed once in ice-cold, isotonic, Tris-buffered saline and then lysed in RIPA buffer (150 mM NaCl, 50 mM Tris-HCl, pH 7.2, 1% w/v sodium deoxycholate, 1% Nonidet P-40, 0.1% SDS, 200 µM Na₃VO₄, 50 mM NaF, 2 mM EDTA, 100 kallikrein-inactivating units/ml aprotinin) for 20 min at 4 °C at a concentration of 10⁷ cells/ml. 30 µl of Staphylococcus aureus cells (Pansorbin cells, Calbiochem) were added, and the lysate was clarified by centrifugation at 35,000 × g at 4 °C for 45 min. The supernatant was removed and kept on ice until use.

Total Lck protein was isolated from clarified cell lysates by immunoprecipitation with a rabbit polyclonal antisera specific for Lck (32) prebound to S. aureus cells. Isolated Lck was either analyzed directly at this point or was subjected to re-immunoprecipitation to isolate the population of Lck phosphorylated at Tyr-394. To isolate Lck phosphorylated at Tyr-394, total Lck bound to S. aureus cells was resuspended in 200 µl of boiling lysis buffer (50 mM Tris, pH 8.0, 0.5% SDS, 1 mM dithiothreitol), and boiled for 3 min. The S. aureus cells were pelleted by centrifugation at 15,000 × g at room temperature (22 °C), and the supernatant containing the Lck protein was removed and added to four volumes (800 µl) of ice-cold RIPA correction buffer (187 mM NaCl, 62.5 mM Tris-HCl, pH 7.2, 1.25% w/v sodium deoxycholate, 1.25% Nonidet P-40, 250 µM Na₃VO₄, 62.5 mM NaF, 2.5 mM EDTA, 112 kallikrein-inactivating units/ml aprotinin, 1 mM dithiothreitol). Lck proteins phosphorylated at Tyr-394 were isolated with a rabbit polyclonal antisera specific for a phosphorylated peptide identical to the sequence flanking Tyr-416 of Src (a generous gift from M. Weber) that was prebound to S. aureus cells. Immune complexes were pelleted by centrifugation and washed three times in RIPA buffer and once in TN buffer (40 mM Tris, pH 7.5, 150 mM NaCl). Remaining Lck proteins not immunoprecipitated by the -Tyr(P)-416 sera were isolated by re-immunoprecipitation of the supernatant with the Lck-specific antisera.

Immunoprecipitated Lck proteins were resolved by gel electrophoresis and transferred to a polyvinylidene difluoride (PVDF, Immobilon-P, Millipore) membrane. Western blotting was carried out with -Tyr(P)-416, -Tyr(P) (33), or -Lck antibodies and ¹²⁵I-protein A (ICN) (33, 34) as described previously. The -Tyr(P)-416 antisera were diluted 1:200 before use.

10⁷ Jurkat cells were washed twice with phosphate-free Dulbecco's modified Eagle's medium and incubated with [³²P]phosphoric acid (H₃³²PO₄; 0.5 mCi/ml; ICN) in 4 ml of medium at 37 °C for 5 h. ³²P-Labeled Lck was isolated by immunoprecipitation and digested with tosylphenylalanyl chloromethyl ketone-treated trypsin as described (35). Two-dimensional tryptic peptide mapping was carried out on cellulose thin layer chromatography plates by electrophoresis at pH 8.9 in the first dimension followed by ascending chromatography in phosphochromatography buffer as described (36). Labeled peptides were visualized by autoradiography. Relative peptide phosphorylation levels were determined with a PhosphorImager (Molecular Dynamics).

RESULTS

To examine whether Lck phosphorylated at Tyr-394 following exposure of cells to H₂O₂ was also phosphorylated at Tyr-505, we used an antibody raised against a phosphorylated peptide from the region in Src containing Tyr-416. This antiserum (-Tyr(P)-416, a generous gift of M. Weber) cross-reacts with Lck phosphorylated at Tyr-394 because the amino acid sequences flanking Tyr-394 in Lck are identical to those flanking Tyr-416 in Src. To verify that this serum specifically recognized Lck that was phosphorylated at Tyr-394, we first tested it in Western blotting. We isolated Lck from Jurkat human T leukemia cells or from rat 208F fibroblast cell lines expressing Lck by immunoprecipitation with -Lck antibodies before and after the cells had been exposed to H₂O₂. 208F fibroblasts express no endogenous Lck. Wild type Lck from either unstimulated T cells or from rat fibroblasts is highly phosphorylated at Tyr-505 and poorly phosphorylated at Tyr-394 (22, 32). In contrast, a constitutively active form of Lck where Tyr-505 is mutated to phenylalanine (Lck^F505) is highly phosphorylated on Tyr-394 in unstimulated cells (16). A double mutant of Lck (Lck^A2F505) that lacks the amino-terminal myristoylation site (Gly-2) and contains the carboxyl-terminal tyrosine to phenylalanine mutation completely lacks tyrosine phosphorylation in unstimulated cells (21, 37). All of these forms of Lck isolated from H₂O₂-stimulated cells are highly phosphorylated on Tyr-394 (22, 37).

When Lck from untreated cells was examined, -Tyr(P)-416 reacted strongly with Lck^F505 but poorly or not at all with wild type Lck and Lck^A2F505 (Fig. 1A). In contrast, -Tyr(P)-416 reacted with all forms of Lck isolated from H₂O₂-treated cells. H₂O₂ did not alter the level of Lck in either T cells or fibroblasts (Fig. 1B). These results rule out the possibility that significant amounts of contaminating anti-peptide reactivity or -Tyr(P) reactivity that recognizes Tyr(P)-505 are present in the -Tyr(P)-416 sera. If present, such antibodies would have recognized Lck from unstimulated cells because the majority of Lck in these cells is not phosphorylated on Tyr-394, but highly phosphorylated on Tyr-505.

Analysis of the specificity of -Tyr(P)-416 antiserum by Western blotting.

We also tested whether the -Tyr(P)-416 sera exhibited specificity for Tyr(P)-394 in immunoprecipitation. Total Lck was isolated by immunoprecipitation from Jurkat cells with -Lck antibodies before and after H₂O₂ stimulation, and the Lck immunoprecipitates were subsequently boiled to both disassociate the complex and inactivate the -Lck immunoglobulin. Lck molecules containing Tyr(P)-394 were then isolated by immunoprecipitation with -Tyr(P)-416 antisera. We analyzed the resulting immune complexes by Western blot with -Lck antibodies (Fig. 2A) and -Tyr(P) antibodies (Fig. 2B).

Analysis of the specificity of -Tyr(P)-416 antiserum by immunoprecipitation.

Lck from unstimulated cells was immunoprecipitated with -Tyr(P)-416 very inefficiently, consistent with the low level of Tyr-394 phosphorylation. In contrast, the -Tyr(P)-416 serum readily immunoprecipitated Lck from Jurkat cells that had been exposed to H₂O₂. Equal amounts of Lck were present in both stimulated and unstimulated samples as determined by -Lck Western blot (Fig. 2A, lanes 3 and 4). To exclude the possibility that any -Lck antibodies from the initial immunoprecipitation renatured following the boiling step, we immunoprecipitated with S. aureus cells alone. No Lck was immunoprecipitated indicating that no functional -Lck antibodies remained after boiling (data not shown). Thus, the -Tyr(P)-416 serum exhibited good specificity for Lck phosphorylated at Tyr-394 in both Western blotting and immunoprecipitation.

The shift in electrophoretic mobility seen in Lck following cellular exposure to H₂O₂ (Figs. 1 and 2) appears to be a result of protein phosphorylation because bacterial alkaline phosphatase is able to collapse the multiple Lck bands to a single 56-kDa band.² In addition, it is likely that serine phosphorylation is more important than tyrosine phosphorylation for H₂O₂-induced Lck mobility shifts because stimulation of cells with 12-O-tetradecanoylphorbol-13-acetate induces identical shifts without inducing tyrosine phosphorylation of Lck or activating Lck.²

To determine whether Lck proteins that are phosphorylated on Tyr-394 following H₂O₂ stimulation are also phosphorylated on Tyr-505, we labeled Jurkat cells biosynthetically with ³²P_i and then exposed the cells to H₂O₂. Total Lck was isolated first by immunoprecipitation with -Lck antibodies, and the immunoprecipitates were then boiled. Lck proteins phosphorylated on Tyr-394 were then isolated by reprecipitation with -Tyr(P)-416 antisera. The isolated Lck was subjected to two-dimensional tryptic peptide analysis (Fig. 3).

Analysis of the total population of Lck from unstimulated T cells or rat fibroblasts, immunoprecipitated with -Lck antibodies, showed that a very small fraction of Lck molecules was phosphorylated on Tyr-394; the Tyr(P)-505:Tyr(P)-394 ratio was 15:1 in T cells and 34:1 in fibroblasts (Fig. 3, A and E). In contrast, the ratio of Tyr(P)-505 to Tyr(P)-394 in Lck immunoprecipitated with -Tyr(P)-416 from unstimulated T cells was 0.35:1 (Fig. 3C). This showed that approximately one-third of the small percentage of Lck molecules from unstimulated T cells that were phosphorylated on Tyr-394 were also phosphorylated on Tyr-505. No Lck molecules phosphorylated on Tyr-394 could be detected by immunoprecipitation with -Tyr(P)-416 from unstimulated fibroblasts (Fig. 3G). Following H₂O₂ stimulation of T cells and rat fibroblasts, both the phosphorylation of Tyr-394 and the amount of Lck precipitable by -Tyr(P)-416 increased dramatically (Fig. 2A, compare lanes 1 and 2; Fig. 3, compare B and D to A and compare F and H to E). We found that the ratio of Tyr(P)-505 to Tyr(P)-394 in Lck isolated with -Tyr(P)-416 from H₂O₂-stimulated T cells and fibroblasts ranged from 0.940:1 to 0.955:1 in three independent experiments. (Fig. 3, D and H). This result indicates that approximately 95% of the population of Lck molecules phosphorylated on Tyr-394 following H₂O₂ stimulation was also phosphorylated on Tyr-505.

DISCUSSION

The catalytic activity of Lck is greatly influenced by the phosphorylation state of Tyr-394 (22, 38, 39). Mutation of Tyr-394 to phenylalanine not only decreases Lck activity in unstimulated cells, but also prevents activation of Lck by oxidative stress. In addition, Lck that has been genetically activated by mutation of Tyr-505 to phenylalanine loses its transforming ability when Tyr-394 is also mutated to phenylalanine (21, 40). We previously showed that the extent of Tyr-505 phosphorylation in Lck from H₂O₂-stimulated cells was at least as great as that of Tyr-394. Thus, H₂O₂ activation appeared not to require Tyr-505 dephosphorylation. These observations suggested that Tyr-394 phosphorylation could override any negative regulation of Lck due to Tyr-505 phosphorylation. However, it was impossible to rule out the possibility that two differentially phosphorylated subpopulations of Lck existed in H₂O₂-stimulated cells, a relatively inactive population phosphorylated only on Tyr-505 and an activated population phosphorylated only on Tyr-394. Thus, it was formally possible that activation of Lck by Tyr-394 phosphorylation occurred only in the absence of phosphorylation of Tyr-505.

Through the use of a phosphorylation state-specific antibody, we have now shown formally that Lck that is phosphorylated on Tyr-505 may be additionally phosphorylated on Tyr-394. If dephosphorylation of Tyr-505 were required for phosphorylation of Tyr-394, we would expect that Lck immunoprecipitated by -Tyr(P)-416 sera would only be phosphorylated on Tyr-394. This was not seen. The observation that the Tyr(P)-505:Tyr(P)-394 ratio is approximately 1:1 suggests that the molecules phosphorylated on Tyr-394 are also phosphorylated on Tyr-505. Thus, the activating effects of Tyr-394 phosphorylation are dominant over the inhibitory effects of Tyr-505 phosphorylation. Apparently, even when Lck is in a "closed" conformation with Tyr(P)-505 bound to the SH2 domain and the SH3 domain bound to the polyproline type II helix in the linker region between the SH2 and catalytic domains (19, 20), Tyr-394 is still accessible as a substrate for phosphorylation. Comparison of the crystal structure of Lck phosphorylated at Tyr-394 (24) with that of Src and Hck lacking phosphorylation at this site (19, 20) suggests that phosphorylation of Tyr-394 allows the formation of hydrogen bonds between Tyr(P)-394 and Arg-387 and Arg-363. These interactions appear to induce the repositioning of Glu-288, Leu-385, and Arg-387 (24), and this in turn allows Lys-273, Glu-288, Asp-364, Asn-369, and Asp-382, residues critical to ATP binding and phosphate transfer, to assume positions characteristic of an active catalytic site (24-26). Our data suggest that such repositioning of residues in the catalytic site of Lck can occur while the SH3 domain is still intramolecularly bound to the SH2-kinase linker region.

Our results agree with data presented by others who showed that the Src tyrosine kinase retains activity when phosphorylated on Tyr-416 and Tyr-527 (41).³ Previous work in our laboratory as well as kinetic data by other groups suggest that the activating phosphorylation of Tyr-394 in Src family members is an intermolecular event rather than intramolecular reaction (22, 42-44).⁴ Intermolecular phosphorylation of Tyr-394 may be carried out by Lck in vivo, but it is quite possible that other Src family members, or non-Src tyrosine kinases, may also act to phosphorylate Tyr-394 and activate Lck. Consistent with this second possibility is our finding that H₂O₂ stimulates the activity of an as yet unidentified tyrosine kinase in Lck-deficient JCaM1.6 cells that can phosphorylate Lck on Tyr-394.²

Our results show that it is possible for the Src family member Lck to be phosphorylated at both the conserved tyrosine in the activation loop and the C-terminal tyrosine simultaneously. Because H₂O₂ activates Lck while inducing an increase of phosphorylation of both Tyr-394 and Tyr-505, these results suggest that Lck phosphorylated on Tyr-394 and Tyr-505 is catalytically activated. Thus, activation of Lck can occur in the absence of Tyr505 dephosphorylation or SH3 domain disassociation from the SH2-kinase linker. It is reasonable to predict that phosphorylation of the absolutely conserved tyrosine in the activation loop of other Src kinases will also activate them in the absence of dephosphorylation of the conserved carboxyl-terminal tyrosine.