The Lck SH3 Domain Is Required for Activation of the Mitogen-activated Protein Kinase Pathway but Not the Initiation of T-cell Antigen Receptor Signaling*

Initiation of T-cell antigen receptor (TCR) signaling is dependent upon the activity of protein tyrosine kinases. The Src family kinase Lck is required for the initial events in TCR signaling, such as the phosphorylation of the TCR complex and the activation of ZAP-70, but little is known of its role in downstream signaling. Expression of a mutated form of Lck lacking SH3 domain function (LckW97A) in the Lck-deficient T-cell line JCaM1 revealed a requirement for Lck beyond the initiation of TCR signaling. In cells expressing LckW97A, stimulation of the TCR failed to activate the mitogen-activated protein kinase (MAPK) pathway, despite normal TCR z chain phosphorylation, ZAP-70 recruitment, and ZAP-70 activation. Activation of extracellular signal-regulated kinase (ERK) and MAPK kinase (MEK), as well as the induction of CD69 expression, was greatly impaired in JCaM1/LckW97A cells. In contrast, the phosphorylation of phospholipase C g 1 (PLC g 1) and corresponding elevations in intracellular calcium concentration ([Ca 2 1 ] i ) were intact. Thus, cells expressing LckW97A exhibit a selective defect in the activation of the MAPK pathway. These results demonstrate that Lck has a role in the activation of signaling pathways beyond the initiation of TCR signaling and suggest that the MAPK pathway may be selectively controlled by regulating the function of Lck.


Initiation of T-cell antigen receptor (TCR) signaling is
Activation of T-lymphocytes is initiated by engagement of the T-cell antigen receptor (TCR) 1 (1)(2)(3). Signaling through the TCR is dependent upon the activity of protein-tyrosine kinases which induce the phosphorylation of a number of proteins, including the subunits of the TCR complex itself. Tyrosine phosphorylation of the TCR complex occurs within amino acid sequences known as immunoreceptor-based tyrosine activation motifs (ITAMs) and is required for receptor signaling function (4 -6). Activation of downstream signaling pathways, including the Ras/mitogen-activated protein kinase (MAPK) pathway, the phosphatidylinositol (PIP 2 ) pathway, and the phosphatidylinositol 3-kinase (PI-3-K) pathway, is dependent upon proteintyrosine kinase function (7)(8)(9)(10). However, the mechanisms by which the induction of tyrosine kinase activity initiates these diverse signaling processes is unknown.
Unlike receptor tyrosine kinases, the cytoplasmic portion of the TCR lacks intrinsic catalytic activity. Instead, the induction of tyrosine phosphorylation following engagement of the TCR requires the expression of non-receptor kinases. Both the Src family and the Syk/ZAP-70 family of tyrosine kinases are required for normal TCR signal transduction (11)(12)(13)(14)(15)(16)(17)(18). A role for Lck has been identified in the initiation of TCR signaling; T-cells lacking functional Lck fail to initiate ITAM phosphorylation or induce ZAP-70 recruitment and activation (12,17,19). ZAP-70 has been implicated in downstream signaling events such as activation of the PIP 2 and the MAPK pathways (18, 20 -22). A current model for the initiation of TCR signaling proposes that Lck and ZAP-70 function sequentially (1)(2)(3). Heterologous cell systems expressing ZAP-70 and the cytoplasmic domain of the TCR chain displayed enhanced ITAM phosphorylation and ZAP-70 activation when Lck is co-transfected with ZAP-70 (19,23,24), suggesting Lck mediates both the initial ITAM phosphorylation and the subsequent phosphorylation and activation of ZAP-70. However, this does not preclude a role for Lck in TCR signaling which is independent of the activation of ZAP-70 (25).
Like other Src family kinases, Lck contains a C-terminal kinase domain, a single Src homology 2 (SH2) and SH3 domain, and an N-terminal region that is distinct from other family members (26). SH3 domains are able to mediate protein interactions by binding certain proline-rich amino acid sequences (27,28), and a number of proteins have been reported to bind the Lck SH3 domain including c-Cbl, PI-3-K, Ras-GAP, HS1, and CD2 (29 -33). The SH3 domain of Lck has also been proposed to stabilize the formation of Lck homodimers which may potentiate TCR signaling following co-ligation of the receptor and CD4 (34,35). Previous work indicated that deletion of the Lck SH3 domain interfered with the ability of an oncogenic form of Lck to enhance interleukin-2 production, supporting a role for the Lck SH3 domain in regulating T-cell activation (36).
In the present study, we have examined the involvement of the Lck SH3 domain in TCR signaling processes by expressing an altered form of Lck which contained an inactive SH3 domain in the Lck-deficient T-cell line JCaM1 (12). TCR signaling pathways displayed a differential sensitivity to loss of Lck SH3 * 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  domain function. The induction of tyrosine phosphorylation and activation of the PIP 2 pathway was independent of the Lck SH3 domain, whereas activation of the MAPK pathway was strictly SH3 domain-dependent. The inability of this altered form of Lck to activate the MAPK pathway despite ZAP-70 activation suggests that Lck participates directly in the stimulation of downstream signaling pathways following TCR ligation.

EXPERIMENTAL PROCEDURES
Cells and Plasmids-Derivatives of the Lck-deficient Jurkat cell line JCaM1 were maintained at 37°C and 5% CO 2 in RPMI 1640 supplemented with 10% fetal bovine serum, glutamine, penicillin, and streptomycin. The LckW97A mutant cDNA was generated from wild type human Lck cDNA using base pair mismatched primers and polymerase chain reaction amplification and was confirmed by sequencing. The Lck cDNA was subcloned into pBP1, a plasmid derived from pUCH-13 (37) which contains a cytomegalovirus promoter sequence regulated by tetracycline operators, and a gene conferring resistance to the antibiotic hygromycin. Clones that express wild type Lck or LckW97A were generated by electroporation of a G418-resistant JCaM1 derivative, which expresses a VP16-tetracycline repressor fusion protein (37), and were isolated by plating at limiting dilution in the presence of hygromycin and G418. Clones that expressed TCR and wild type Lck, or LckW97A, at levels equivalent to the parental Jurkat cell line were maintained for further analysis.
Analysis of TCR and CD69 Expression-TCR expression was measured by staining cells with a mouse antibody recognizing CD3 (Leu-4), followed by a FITC-conjugated goat anti-mouse secondary antibody, and analyzed by fluorescence flow cytometry. To assess the induction of CD69, cells expressing wild type Lck or LckW97A, or Lck-deficient control cells, were incubated with media alone, PHA (0.3 g/ml), or PMA (50 ng/ml) for 16 -20 h at 37°C. Cells were stained with a mouse anti-CD69 antibody (PharMingen) and a FITC-conjugated goat antimouse secondary antibody then evaluated by flow cytometry.
In 15 Ci of [␥-32 P]ATP. The reactions were terminated by addition of 2ϫ SDS sample buffer.
Following SDS-PAGE and staining with Coomassie Blue, the gels were dried to filter paper and exposed to x-ray-sensitive film (X-OMat, Eastman Kodak Co.). The regions corresponding to phospho-GST-and phospho-Lck or phospho-GST-Band III were excised, and the incorporation of 32 P was measured by counting Cerenkov radiation.
Calcium Measurement-Cells were loaded with the fluorescent calcium indicator indo-1, washed extensively in Hepes-buffered saline (HeBS, pH 7.4) solution containing (in mM) 25 Hepes, 125 NaCl, 5 KCl, 1 CaCl 2 , 0.5 MgCl 2 , 1 Na 2 HPO 4 , 0.1% (w/v) bovine serum albumin and 0.1% (w/v) D-glucose, and kept on ice. Prior to use, the aliquots were warmed to 37°C for 10 min and then placed in a spectrofluorometer equipped with a thermally jacketed cuvette holder maintained at 37°C. The fluorescence intensity was monitored continuously at the emission wavelength of 400 nm following excitation at 334 nm. Fluorescence intensity values were corrected for cell autofluorescence and then converted to [Ca 2ϩ ] i using a K d value of indo-1 for Ca 2ϩ of 250 nM (39).

RESULTS
To examine the role of Lck in TCR signaling, we have utilized the Lck-deficient JCaM1 T-cell line, derived from the Jurkat cell line (12,40). The advantage of using JCaM1 is that TCR signaling can be studied in clones expressing an altered form of Lck independent of the contributions of endogenous wild type Lck (41). As such, the importance of the SH3 domain of Lck in TCR signaling was examined by transfecting JCaM1 cells with an Lck cDNA that encoded a protein containing an inactive SH3 domain. A conserved tryptophan in SH3 domains stabilizes their association with proline-rich target peptides (28); mutation of this tryptophan residue abolishes ligand binding (42,43). In this study, the SH3 domain of Lck was inactivated by mutating this essential tryptophan, located at position 97, to alanine using site-directed mutagenesis (LckW97A). In preliminary experiments, we observed that the SH3 domain of LckW97A, when expressed as a fusion protein with glutathione S-transferase (GST), was unable to bind the known substrate c-Cbl (not shown). Thus, the LckW97A mutation represents a loss of function mutation for the Lck SH3 domain. Stable transfection of JCaM1 cells generated several independent clones expressing similar levels of TCR, and levels of wild type Lck or LckW97A, equivalent to those in the parental Jurkat cell line (Fig. 1).
Lck kinase activity is essential for the initiation of TCR signaling, and it is possible that disruption of the SH3 domain of Lck affects its catalytic activity. To determine if the SH3 domain of Lck influenced its catalytic activity in vitro, we incubated immunoprecipitates of LckW97A or wild type Lck with [␥-32 P]ATP and an exogenous substrate consisting of the cytosolic domain of the chain of the TCR fused to GST ( Fig.  2A). Both wild type Lck and LckW97A displayed autophosphorylation and GST-phosphorylation, indicating that the SH3 domain of Lck was not required for kinase activity. In fact, LckW97A displayed enhanced catalytic activity relative to wild type Lck, consistent with an autoinhibitory role for the Lck SH3 domain (Fig. 2B). Thus, any disturbances in TCR signaling in cells expressing LckW97A cannot be attributed to a general loss of catalytic activity.
Although LckW97A was active in vitro, mutation of the SH3 domain may disrupt TCR-induced substrate phosphorylation within the cell, possibly by altering Lck distribution or by interfering with the binding of potential substrates. Therefore, we examined the levels of cellular tyrosine phosphoproteins in unstimulated and TCR-stimulated JCaM1 transfectants expressing wild type Lck, or LckW97A, by immunoblotting lysates with an anti-phosphotyrosine antibody (Fig. 3A). The profiles of tyrosine phosphoproteins in unstimulated and TCRstimulated cells expressing LckW97A closely resembled those of cells expressing wild type Lck, whereas TCR stimulation of plasmid vector-transfected JCaM1 cells failed to induce tyrosine phosphorylation (not shown). Although cells expressing LckW97A displayed a slight, but reproducible, elevation in the basal level of tyrosine phosphorylation, it is apparent that mutation of SH3 domain of Lck did not substantially alter the tyrosine phosphorylation of cellular phosphoproteins prior to or following TCR stimulation. As such, LckW97A was capable of inducing tyrosine phosphorylation upon TCR stimulation.
Although the induction of tyrosine phosphorylation appeared to be intact, we examined in greater detail the effects of mutation of the Lck SH3 domain on the initiation of TCR signaling. Specifically, we focused on the induction of ITAM and ZAP-70 phosphorylation in cells expressing LckW97A since these FIG. 2. LckW97A displayed both autophosphorylation and TCR chain phosphorylation in vitro. A, in vitro kinase assay of Lck immunoprecipitates from JCaM1 cells expressing either wild type Lck or LckW97A. Immunoprecipitates were incubated at 37°C with [␥-32 P]ATP and the exogenous substrate GST-. Aliquots were removed at 1, 3, and 5 min after the start of the assay as indicated and analyzed by SDS-PAGE and autoradiography. Each lane contained equivalent amounts of GSTas determined by staining with Coomassie Blue (not shown). A fraction of each immunoprecipitate was immunoblotted with an anti-Lck antibody to confirm that equivalent amounts of wild type Lck or LckW97A were recovered. Results are representative of three separate experiments. B, the kinase activity of wild type Lck and LckW97A was determined by excising the band corresponding to GSTas shown in A and measuring 32 P content by counting Cerenkov radiation. 32 P incorporation remained linear over the time course of the assay. The catalytic activity of LckW97A is expressed relative to that of wild type Lck in three independent experiments (mean Ϯ S.E.). Similar LckW97A catalytic activity was observed in a second clone. events are mediated directly by Lck. Immunoprecipitates of TCR chain were immunoblotted for phosphotyrosine to evaluate ITAM phosphorylation and associated ZAP-70 phosphorylation (Fig. 3B). TCR stimulation of cells expressing either wild type Lck or LckW97A induced the phosphorylation of TCR chain and association of tyrosine-phosphorylated ZAP-70 (Fig. 3B), but vector-transfected control cells failed to induce detectable levels of TCR chain or ZAP-70 phosphorylation (not shown). To confirm that TCR stimulation of cells expressing LckW97A elicited ZAP-70 activation, we examined ZAP-70 phosphorylation in the JCaM1 transfectants and evaluated ZAP-70 kinase activity in vitro. Anti-phosphotyrosine blotting of ZAP-70 immunoprecipitates revealed that TCR stimulation induced similar levels of ZAP-70 phosphorylation in cells expressing either LckW97A or wild type Lck but not vectortransfected cells (Fig. 4A). The induction of ZAP-70 phosphorylation in TCR-stimulated cells corresponded to an elevation in ZAP-70 kinase activity in vitro, demonstrating that the SH3 domain of Lck is not required for the activation of ZAP-70 (Fig.  4B). Although a slight reduction in TCR-induced ZAP-70 activation was observed in JCaM1 cells expressing LckW97A, this difference was not statistically significant (p Ͼ 0.05). Thus, the previously identified roles for Lck in the initiation of TCR signaling, namely ITAM phosphorylation and the subsequent phosphorylation and activation of ZAP-70, can occur independently of Lck SH3 domain function.
Since mutation of the SH3 domain did not alter the initial TCR signaling events that are known to require Lck, we investigated the possibility that the SH3 domain of Lck participates in the activation of downstream signaling events. Specifically, we examined the ability of this altered form of Lck to activate the PIP 2 and the MAPK signaling pathways following TCR stimulation. Activation of the PIP 2 pathway upon TCR stimulation was assessed by measuring elevations in [Ca 2ϩ ] i and the induction of PLC␥1 phosphorylation. Elevations in [Ca 2ϩ ] i elicited by TCR stimulation were examined using cells loaded with the Ca 2ϩ -sensitive fluorescent dye indo-1. We analyzed a range of anti-TCR antibody concentrations since subtle disturbances in the activation of the PIP 2 pathway would be more apparent at submaximal levels of TCR stimulation. A deficit in the activation of the PIP 2 pathway would be manifested as a shift in the concentration-response curve of TCR-stimulated elevations in [Ca 2ϩ ] i . However, TCR stimulation of cells expressing wild type Lck or LckW97A induced similar elevations in [Ca 2ϩ ] i at every concentration of anti-TCR antibody tested (Fig. 5A). Similarly, the kinetics of these TCR-stimulated elevations in [Ca 2ϩ ] i were identical in cells expressing wild type Lck or LckW97A (Fig. 5B). The time-to-onset, peak elevation, decline phase, and plateau were all unaltered by inactivation of the Lck SH3 domain. We confirmed that the Lck SH3 domain was not required for the activation of the PIP 2 pathway by measuring PLC␥1 phosphorylation in cells expressing LckW97A. Anti-phosphotyrosine immunoblots of PLC␥1 immunoprecipitates from cells expressing LckW97A showed that PLC␥1 tyrosine phosphorylation was induced upon TCR stimulation (Fig.  5C). Although lower levels of inducible PLC␥1 phosphorylation were observed in JCaM1/LckW97A cells, it did not alter the ability of these cells to elicit elevations in [Ca 2ϩ ] i following TCR stimulation. Thus, the activation of the PIP 2 pathway following TCR stimulation is not dependent upon the SH3 domain of Lck.
We also examined the activation of the MAPK pathway following TCR stimulation of cells expressing LckW97A. In contrast to the initiation of TCR signaling and the activation of the PIP 2 pathway, the activation of the MAPK pathway was strictly dependent upon the SH3 domain of Lck. MAPK pathway activity was assessed initially by immunoblotting cell lysates with antibodies that specifically recognize the activated forms of the MAP kinases ERK1 and ERK2 and the MAPK kinases MEK1 and MEK2. TCR stimulation of cells expressing LckW97A failed to induce ERK or MEK phosphorylation, whereas stimulation of cells expressing wild type Lck elicited substantial ERK and MEK phosphorylation (Fig. 6A). A kinetic analysis demonstrated that extending the period of TCR stimulation as much as 20 min did not enhance ERK phosphorylation in JCaM1 cells expressing LckW97A (not shown). TCR-independent activation of the MAPK pathway using PMA resulted in phosphorylation of both ERK and MEK irrespective of Lck expression (Fig. 6A). The selective disruption of the MAPK pathway in cells expressing LckW97A demonstrates that an additional requirement for Lck exists in the activation of downstream signaling events, which is distinct from its previously identified role in mediating the initiation of TCR signaling.
Engagement of the TCR initiates intracellular signaling processes that culminate in characteristic changes in gene expression, including expression of the cell-surface marker CD69 (44). In light of the pronounced deficits in MAPK pathway activation, we suspected that TCR stimulation of JCaM1 cells expressing LckW97A would be incapable of inducing CD69 expression (45). Treatment of cells expressing LckW97A with phytohemagglutinin (PHA) resulted in little or no induction of CD69 expression, whereas PHA treatment of cells expressing wild type Lck elicited a 10-fold elevation in median fluorescence intensity (Fig. 6B). TCR-independent stimulation with PMA elicited similar levels of CD69 expression in all cell lines, regardless of Lck expression. Thus, the SH3 domain of Lck is required for TCR-dependent CD69 expression. The inability of LckW97A to activate the MAPK pathway or induce CD69 ex- FIG. 4. LckW97A induced ZAP-70 activation following TCR stimulation. A, anti-phosphotyrosine blot of ZAP-70 immunoprecipitates from unstimulated and stimulated cells expressing wild type Lck, or LckW97A, or Lck-deficient JCaM1 cells. The position of phosphorylated ZAP-70 is indicated. A fraction of each ZAP-70 immunoprecipitate was immunoblotted with an anti-ZAP-70 antibody to confirm that equivalent amounts were recovered. Results are representative of three experiments using two JCaM1 clones expressing LckW97A. B, ZAP-70 kinase activity was measured in vitro using immunoprecipitates from unstimulated or TCR-stimulated JCaM1 cells expressing wild type Lck, or LckW97A, or vector-transfected controls. Immunoprecipitates were incubated with [␥-32 P]ATP and the exogenous substrate GST-Band III. Reactions were analyzed by SDS-PAGE and autoradiography. Coomassie Blue staining confirmed that each lane contained equivalent amounts of GST-Band III (not shown). 32 P incorporation into GST-Band III was measured by counting Cerenkov radiation, and ZAP-70 activation was expressed as the fold induction in kinase activity upon TCR stimulation (n ϭ 4, mean Ϯ S.E.). Statistical analysis was performed by one-way analysis of variance and Dunnett's test (p Ͻ 0.05, two-tailed). Values that differ significantly from wild type Lck are indicated by an asterisk. ZAP-70 activation induced by LckW97A did not differ significantly from wild type Lck. Similar results were obtained using a second LckW97A clone. pression demonstrates a critical role for the SH3 domain of Lck in TCR signaling and T-cell activation. DISCUSSION We have identified a crucial role for the SH3 domain of Lck in activation of the MAPK pathway following TCR stimulation. The Lck-deficient cell line JCaM1 was transfected with an altered form of Lck in which a conserved tryptophan located within the SH3 domain was mutated to alanine, thereby abolishing Lck SH3 domain function. Substantial biochemical and genetic evidence has demonstrated that this conserved tryptophan is essential for ligand binding (28,42,43,46); thus, it is very unlikely that LckW97A has any residual SH3 domain function. JCaM1 clones expressing LckW97A failed to induce CD69 expression and MAPK activation upon TCR stimulation, revealing a defect in TCR signaling. Since LckW97A possessed kinase activity in vitro, the block in T-cell activation in cells expressing LckW97A could not be attributed to loss of catalytic function. Furthermore, cells expressing LckW97A had only minor alterations in the tyrosine phosphorylation of cellular proteins and exhibited inducible TCR chain phosphorylation and ZAP-70 activation upon TCR stimulation. Thus, the Lck SH3 domain is not required for the initiation of TCR signaling but is essential for events that are independent of, or subsequent to, the recruitment and activation of ZAP-70. Additional analysis of downstream signaling pathways demonstrated that cells expressing LckW97A were capable of activating the PIP 2 pathway following TCR stimulation; both PLC␥1 phosphorylation and subsequent elevations in [Ca 2ϩ ] i were independent of Lck SH3 domain function. However, unlike the PIP 2 pathway, TCR stimulation of cells expressing LckW97A failed to induce activation of ERK or MEK, indicating that there was a selective loss of this TCR signaling pathway. Thus, the Lck SH3 domain is required specifically for the activation of the MAPK pathway but not the initiation of TCR signaling or the activation of the PIP 2 pathway.
A current model of TCR signal transduction proposes that Lck and ZAP-70 function sequentially to initiate signaling (1)(2)(3). Lck phosphorylates the TCR which allows ZAP-70 recruitment, followed by ZAP-70 phosphorylation and activation, which is also mediated by Lck. The subsequent stimulation of the PIP 2 and MAPK pathways is dependent upon the activation of ZAP-70. In the simplest model, the sole role of Lck in TCR signaling would be to recruit and activate ZAP-70. However, our data demonstrate that the involvement of Lck in TCR signaling is more complex and establish that a selective requirement for Lck exists in the activation of downstream signaling pathways. TCR stimulation of JCaM1/LckW97A cells failed to activate the MAPK pathway, despite activation of ZAP-70 and the PIP 2 pathway. Recent studies with a constitutively active form of ZAP-70 support the notion that Lck is required for downstream signaling (25). Constitutively active ZAP-70 is able to elicit interleukin-2 gene expression independently of TCR stimulation but only when co-expressed with Lck. Taken together, these studies indicate an additional role for Lck in TCR signaling beyond the recruitment and activation of ZAP-70. Several models could explain the requirement for both activated ZAP-70 and Lck in the stimulation of the MAPK pathway. Lck and activated ZAP-70 may provide separate sig-FIG. 6. Activation of the MAPK pathway following TCR stimulation was defective in JCaM1 cells expressing LckW97A. A, JCaM1 cells expressing wild type Lck, or LckW97A, or vectortransfected controls were incubated with either anti-TCR antibody for 2 min, or PMA (50 ng/ml) for 5 min, or left unstimulated. Top panel, ERK phosphorylation was evaluated by immunoblotting Nonidet P-40-soluble cell lysates with an antiphosphorylated ERK antibody. Incubating the JCaM1/LckW97A cells with anti-TCR antibody for up to 20 min did not enhance ERK phosphorylation (not shown). Results are representative of three separate experiments using two JCaM1/LckW97A clones. Lower panel, MEK1/2 phosphorylation was evaluated by immunoblotting lysates with an antiphosphorylated MEK antibody. Results are representative of three experiments. B, JCaM1 cells expressing either wild type Lck or LckW97A were incubated overnight with media alone (unstim.), PHA (0.3 g/ml), or PMA (50 ng/ml). Following stimulation, the cells were stained with an anti-CD69 antibody, followed by a FITC-conjugated secondary antibody, and analyzed for the expression of CD69 by flow cytometry. Results are representative of three separate experiments using three clones expressing LckW97A. nals, both of which are needed for activation of the downstream pathway. Alternatively, ZAP-70 and Lck may act together on a common substrate that is essential for the stimulation of the MAPK pathway.
The Lck SH3 domain has been shown to bind several signaling molecules implicated in the regulation of the MAPK pathway. The binding of these target proteins by the Lck SH3 domain may modify their activity as a result of conformational changes, subsequent tyrosine phosphorylation, or recruitment to the membrane. PI-3-K binds to the Lck SH3 domain (30) and is capable of activating the MAPK pathway in some cell types (47). However, it is unlikely that the loss of MAPK activation in JCaM1/LckW97A cells is due to alterations in the function of PI-3-K since we did not observe any alteration in PI-3-K localization or activity following TCR stimulation of JCaM1 cells expressing either wild type Lck or LckW97A (not shown). The proto-oncogene product c-Cbl also associates with the SH3 domain of Lck (29). c-Cbl acts as a negative regulator of MAPK activation (48), although this appears to be the result of a block in the activation of Syk/ZAP-70 making it unlikely that c-Cbl would selectively regulate the MAPK pathway (49). Lck has also been shown to interact directly with the kinases comprising the MAPK signaling pathway, including Raf-1, MEK, and MAPK (50 -53) as well as the inhibitor of Ras activity, Ras-GAP (31). Future studies will be required to understand the specific function of the Lck SH3 domain in the activation of the MAPK pathway following stimulation of the TCR.
The SH3 domains of Src family kinases have been proposed to be autoinhibitory. X-ray crystallographic studies examining the structure of Src and Hck revealed that the SH3 domain mediates an intramolecular interaction with an atypical binding site located in the region linking the SH2 and kinase domains (54,55). Inactivating the SH3 domain of Src induces an 8 -10-fold elevation in its kinase activity (42). However, in the present study, mutation of the SH3 domain of Lck caused a comparatively modest 2-fold elevation in its kinase activity. Moreover, deleting the SH3 domain of Lyn reduces its kinase activity (56). These findings suggest that despite their significant structural similarity, the SH3 domains of Src family kinases may not be functionally identical, although it is possible that other regulatory mechanisms may compensate for loss of SH3 domain function in Lck and Lyn.
In summary, we have identified an additional role for Lck in TCR signaling which is distinct from its ability to mediate ITAM phosphorylation and ZAP-70 activation. Our results demonstrate that TCR-induced activation of the MAPK pathway, but not the PIP 2 pathway, requires the SH3 domain of Lck. The selective requirement for Lck SH3 domain function in the MAPK pathway suggests that the activation of specific downstream signaling events could be controlled by regulating the function of Lck. Such a mechanism may be involved in thymocyte development where activation of the MAPK pathway is required for positive selection but not negative selection (57,58). Future studies will address the molecular basis of the role of the Lck SH3 domain in activation of the MAPK pathway.