J Biol Chem, Vol. 274, Issue 41, 29323-29330, October 8, 1999

Itk/Emt/Tsk Activation in Response to CD3 Cross-linking in Jurkat T Cells Requires ZAP-70 and Lat and Is Independent of Membrane Recruitment^*

Xiaochuan Shan and Ronald L. Wange

From the Laboratory of Biological Chemistry, Gerontology Research Center, NIA, National Institutes of Health, Baltimore, Maryland, 21224-6825

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES

The Tec family tyrosine kinase, Itk has been implicated in T cell antigen receptor (TCR) signaling, yet little is known about Itk regulation. Here, we investigate the role of the tyrosine kinase ZAP-70 in regulating Itk. Whereas Itk was activated in Jurkat T cells in response to CD3 cross-linking, Itk activation was defective in the ZAP-70-deficient P116 Jurkat T cell line. Itk responsiveness to TCR engagement was restored in P116 cells stably transfected with ZAP-70 cDNA. ZAP-70 itself could not directly phosphorylate the Itk kinase domain, indicating an indirect regulation of Itk activity. No role was found for ZAP-70 in regulating Itk recruitment to the plasma membrane, an event that has been suggested to be rate-limiting for the activation of Tec family kinases. Indeed, Itk was found to be constitutively targeted to the membrane fraction in both Jurkat and P116 cells. Lat, a prominent in vivo substrate of ZAP-70 that mediates assembly of multimolecular signaling complexes at the plasma membrane of T cells was also found to be required for TCR-stimulated Itk activation. Itk could not be activated by CD3 cross-linking in a Lat-negative cell line, unless Lat expression was restored. Lat and Itk were observed to co-associate in response to CD3 cross-linking in Jurkat T cells, but not in P116 T cells. The Lat-Itk association correlated with Lat tyrosine phosphorylation, which was deficient in the P116 T cells. These data suggest that ZAP-70 and Lat play important, probably sequential, roles in regulating the activation of Itk following TCR engagement.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES

Engagement of the T cell antigen receptor (TCR)¹ initiates a complex cascade of biochemical events, which, in the context of co-stimulatory signals, culminate in proliferation and acquisition of effector functions by the T cell. The most receptor-proximal events initiated upon TCR stimulation are the activation of several protein tyrosine kinases (PTKs) of the Src, Syk, and Tec families (1, 2). TCR signaling is initiated when the Src family PTKs Lck and Fyn phosphorylate specific tyrosine residues within the immunoreceptor tyrosine-based activation motifs of the CD3 (, , and ) and TCR subunits of the TCR/CD3 complex (3, 4). The doubly phosphorylated immunoreceptor tyrosine-based activation motifs recruit the Syk family PTK ZAP-70 to the TCR via interaction with the tandem Src homology 2 domains of ZAP-70 (5). Recruitment to the TCR is required for the subsequent tyrosine phosphorylation and activation of ZAP-70 (6, 7).

ZAP-70 lies at a key point in the TCR signaling pathway, being required for the activation of calcium mobilization and Erk activation pathways, which are in turn required for interleukin-2 production and T cell proliferation (7, 8). This function is accomplished, in part, by regulating the formation of certain key multimolecular signaling complexes involved in these pathways, by catalyzing the phosphorylation of the hematopoietic-specific proteins SLP-76 and Lat (9-11). Both of these proteins have been shown to be in vivo substrates of ZAP-70 (12-14). SLP-76, when tyrosine-phosphorylated, binds to another hematopoietic-specific protein, Vav, and has been implicated in playing an important, as yet undefined role in both Ca²⁺ mobilization as well as Ras activation in T cells (12, 15, 16). Tyrosine-phosphorylated Lat, which is primarily resident in the glycolipid-enriched membrane (GEM) fraction of the plasma membrane, recruits PLC1, Grb2, and PI3-K to the GEMs (14, 17). The GEM membrane lipid rafts have been proposed to serve as platforms for signal transduction upon TCR engagement (18, 19). Localization of PLC1 and PI3-K to the GEMs positions these enzymes near their shared substrate, phosphatidylinositol 4,5-biphosphate, which is enriched in the GEM fraction, and may facilitate the tyrosine phosphorylation of PLC1. Grb2/SOS recruitment to the GEMs also serves to position SOS in the vicinity of its substrate, Ras, which is constitutively targeted to the GEMs (20).

The Tec family tyrosine kinases have been implicated in antigen receptor signaling in a variety of hematopoietic cell types. Btk, a Tec family member primarily expressed in B cells and mast cells, is involved in B cell antigen receptor signaling and was found to be defective in the human and murine immunodeficiencies, X-linked agammaglobulinemia, and X-linked immunodeficiency, respectively (21-23). Itk, also known as Emt or Tsk, is expressed in T cells and NK cells (24-26); is tyrosine-phosphorylated in response to cross-linking of TCR, CD28, or CD2 (27-29); and has been implicated in thymocyte development and the activation of T cells through TCR and CD28 engagement. Mice engineered with a null mutation within the Itk gene have decreased numbers of mature thymocytes. Furthermore, T cells isolated from these mice are compromised in their proliferative response to allogeneic MHC stimulation, and to anti-TCR/CD3 cross-linking (30). These T cells also exhibit defective PLC1 tyrosine phosphorylation, inositol trisphosphate production, and Ca²⁺ influx in response to TCR cross-linking (31).

How Itk activity is regulated in response to TCR engagement is still poorly understood. Structural studies have shown that Itk forms intramolecular interactions between its Src homology 3 domain and the proline-rich region of its Tec homology domain, and it has been proposed that these associations may be involved in regulating its activity (32). Experiments in Jurkat T cells lacking Lck have demonstrated a requirement for Lck in Itk activation in response to TCR engagement (28). Lck has also been shown to phosphorylate the critical activation loop tyrosine of Itk in vitro (33). The activation of Itk by G protein subunits has also been reported (34), but it remains unclear whether this plays any role in TCR-stimulated Itk activation. Experiments involving overexpression of Itk and c-Src in COS-7 cells demonstrated a requirement for PI3-K in recruiting Itk to the plasma membrane, where it became phosphorylated and activated by c-Src (35). Likewise, in A20 B cells, overexpression of the catalytic subunit of PI3-K was demonstrated to synergize with Btk, Itk, or Tec overexpression in stimulating IP3 production and the rise of [Ca²⁺]_i in response to B cell antigen receptor cross-linking (36).

These latter studies by August et al. (35) and Scharenberg et al. (36) are consistent with the recently proposed model that Tec family kinases possessing pleckstrin homology domains (Itk/Tsk/Emt, Btk, and Tec) are regulated by changes in their subcellular localization. This model holds that receptor engagement activates PI3-K, which raises the plasma membrane concentration of phosphatidylinositol 3,4,5-trisphosphate, which is a high affinity binding site for the Tec family pleckstrin homology domain (37). This results in recruitment of Tec family kinases to the plasma membrane, where they are phosphorylated and activated by membrane-resident Src family kinases. Most of the evidence supporting this model has been gathered by study of Btk regulation (36, 38-40). Testing this model in Jurkat T cells, we found no evidence for TCR-stimulated Itk recruitment to the plasma membrane. Indeed, Itk was found to be constitutively present in the membrane fraction. In unstimulated Jurkat T cells, Itk could be recovered from both the GEM and bulk membrane fractions, and neither fraction exhibited any change in the Itk level upon TCR stimulation, arguing against redistribution of Itk to the plasma membrane as a mechanism for activation in Jurkat T cells. We also examined the role of ZAP-70 in Itk activation by measuring Itk activation in the ZAP-70-deficient P116 T cell line. We found that ZAP-70 is required for phosphorylation of Itk and activation of its kinase activity. However, ZAP-70 was unable to directly phosphorylate the kinase domain of Itk, suggesting that the role of ZAP-70 in regulating Itk tyrosine phosphorylation and activation is indirect. We also report the TCR-stimulated association of Itk with Lat, which was only observed in ZAP-70 replete cells. This association occurs with the same kinetics as Itk tyrosine phosphorylation and activation, and seems to be required for Itk activation in response to CD3 cross-linking, because OKT3-induced Itk activation was markedly reduced in Lat-deficient JCaM2.5 cells.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES

Cells and Antibodies-- The ZAP-70 deficient Jurkat T cell line, P116, has been previously described (8, 41). P116 and the parental Jurkat E6 cells were the kind gift of R. Abraham (Duke University, Durham, NC). P.WT.18, a gift of L. Samelson (NCI, National Institutes of Health, Bethesda, MD), is a stable transfectant of P116 that expresses Myc-tagged, wild type ZAP-70 at a level comparable to the parental Jurkat line and has been characterized previously (41, 42). The Lat-deficient Jurkat T cell mutant JCaM2.5 (43) and its Lat-reconstituted cell line, JCaM2.5B3² have been described and are the kind gifts of A. Weiss (University of California, San Francisco, CA) and L. Samelson (NCI, National Institutes of Health, Bethesda, MD). The JCaM2.5 cells show intact early signaling events, including tyrosine phosphorylation of CD3 chains and tyrosine phosphorylation and activation of ZAP-70, indicating that the Src family kinases Lck and Fyn function normally in these cells (43). All cells were maintained in RPMI 1640 medium supplemented with 7.5% fetal bovine serum (Hyclone), 10 µg/ml ciprofloxacin (Bayer), and 2 mM glutamine. The OKT3 monoclonal antibody to human CD3 and polyclonal rabbit antisera specific for human ZAP-70 and Lck have been described (6). The anti-phosphotyrosine monoclonal antibody, 4G10 was from Upstate Biotechnology, Inc. (Lake Placid, NY). A polyclonal rabbit antiserum specific for human Itk was used to immunoprecipitate Itk and was kindly provided by G. Mills (University of Texas M. D. Anderson Cancer Center, Houston, TX). Itk was immunoblotted with the monoclonal antibody 2F12 directed against the N-terminal 26 amino acids of Itk, which was the gift of L. Berg (University of Massachusetts, Worcester, MA). The cytosolic domain of band III protein (cdb3) was prepared as described previously (44). GST fusion proteins containing the kinase domain of either Itk or Lck were the kind gift of J. Watts and R. Aebersold (University of Washington, Seattle, WA) and have been previously described (45).

Cell Stimulation and Lysis-- Cells were harvested by centrifugation, washed once, and resuspended in cold RPMI 1640 medium at a density of 1 × 10⁸ cells/ml. After equilibration to 37 °C for 10 min, the cells were stimulated with OKT3 (1:50 ascites) for the indicated duration. Stimulation was terminated by addition of 5 volumes of 4 °C lysis buffer (20 mM Hepes (pH 7.4), 1% Triton X-100, 50 mM -glycerophosphate, 2 mM EGTA, 10 mM sodium fluoride, 1 mM sodium orthovanadate, 10% glycerol, 10 µg/ml leupeptin, 10 µg/ml aprotinin, 100 µg/ml 4-(2-aminoethyl)-benzenesulfonyl fluoride). After a 30-min incubation on ice, postnuclear lysates were prepared by a 10-min centrifugation at 4 °C, 21,000 × g. The lysates were either directly analyzed by Western blotting or subjected to immunoprecipitation followed by immunoblotting or kinase assay. In some experiments, when indicated, a Brij 96 lysis buffer was used (1% Brij 96, 150 mM NaCl, 25 mM Tris-HCl (pH 7.5), 5 mM EDTA, 10 mM sodium fluoride, 1 mM sodium orthovanadate, 10 µg/ml leupeptin, 10 µg/ml aprotinin, 100 µg/ml 4-(2-aminoethyl)-benzenesulfonyl fluoride).

Immunoprecipitation and Western Blotting-- The postnuclear whole-cell lysates were incubated with protein A-agarose beads and corresponding antibodies for 2-16 h at 4 °C. In some of the anti-Itk immunoprecipitations, ZAP-70, Lck, and other TCR-associated proteins were depleted from the lysates by three rounds of OKT3 and/or anti-Lck immunoprecipitation. Immunoprecipitates that were to be analyzed by immunoblotting were washed three times with the above lysis buffer supplemented with 150 mM NaCl. Whole-cell lysates and immunoprecipitates to be analyzed by Western blotting were denatured by heating to 100 °C in Nu-PAGE sample buffer, electrophoresed on either 4-12% Nu-PAGE gradient or 6% Tris-glycine gels, and transferred to nitrocellulose membrane according to manufacturer's instructions (NOVEX, San Diego, CA). The blots were developed with the ECL system of Amersham Pharmacia Biotech and autoradiographed on BMR film (Eastman Kodak Co.).

In Vitro Itk Kinase Assay-- Itk-associated tyrosine kinase activity was assessed by immune complex kinase assay. Anti-Itk immunoprecipitates from lysates depleted of CD3 (and CD3-associated proteins) and Lck were washed twice with lysis buffer + 150 mM NaCl, twice with 4 °C LiCl wash buffer (100 mM Tris-HCl (pH 7.5), 0.5 M LiCl) and twice with 4 °C dH₂O. To each sample of washed beads 30 µl of kinase reaction mixture (10 mM MgCl₂, 10 mM Hepes (pH 7.0), 2 mM sodium orthovanadate, 5 µCi of [-³²P]ATP, and 5 µg of RR-SRC substrate peptide (Sigma) were added. The reaction was performed at room temperature for 15 min with frequent mixing, then terminated by addition of acetic acid to 30% of the total volume. The reactions were centrifuged briefly and supernatants were spotted onto p81 phosphocellulose discs (Life Technologies, Inc.). After 4-6 washes with 75 mM phosphoric acid, the ³²P incorporation was measured by liquid scintillation. In some assays, the kinase activity was normalized to the relative amount of Itk recovered in the anti-Itk immunoprecipitates. The relative amount of Itk was measured by densitometric analysis of x-ray films using the public domain NIH Image program (developed at the National Institutes of Health).

In Vitro Phosphorylation of Itk Kinase Domain (Itk-KD) and Cdb3-- The kinase reaction was carried out at 30 °C for 10 min in a total volume of 30 µl of reaction buffer (50 mM Tris-HCl (pH 7.5), 10 mM MnCl₂, 50 µM ATP, and 10 µCi of [-³²P]ATP). The reaction was terminated by the addition of 10 µl of 4× reducing sample buffer and heating to 100 °C for 5 min. Phosphorylated proteins were resolved by SDS-PAGE and subjected to autoradiography. The GST fusion proteins containing the kinase domains of either Itk or Lck were purified from recombinant insect cells as described previously (45). The Itk-KD was cleaved away from the GST fusion partner by proteolytic cleavage with thrombin, and the GST fragment removed. Several of the preparations of Itk-KD prepared in this manner were found to have low intrinsic kinase activity, although they migrated normally on the gel. One of these low activity preparations was used as a substrate in the assay. The use of cdb3 as a ZAP-70 substrate has been described (6). The ZAP-70 enzyme used in this assay was immunoprecipitated from activated (3 min at 37 °C with 1:100 OKT3) Jurkat whole-cell lysates that were depleted of Lck by three rounds of preclearing with antiserum recognizing Lck. The Lck enzyme used in the assay was the GST-Lck-KD fusion protein, which has been previously described (45).

Preparation of Cytosolic and Membrane Fractions-- Cells (2.5 × 10⁷) were centrifuged quickly in cold phosphate-buffered saline after OKT3 stimulation and resuspended in 1.5 ml of cold hypotonic lysis solution (20 mM Hepes (pH 7.6), 5 mM sodium pyrophosphate, 5 mM EGTA, 1 mM MgCl₂, 10 µg/ml aprotinin, 1 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride), 1 mM sodium orthovanadate). The cell suspension sat on ice for 30 min, followed by cellular disruption with 10 passes of a Dounce homogenizer. After centrifugation at 100,000 × g at 4 °C for 1 h, the supernatant was collected as the cytosolic fraction. The pellet was rinsed with cold hypotonic lysis buffer and solubilized in 1.5 ml of the membrane solubilization solution (1% Triton X-100, 20 mM Hepes (pH 7.4), 150 mM NaCl, 1 mM MgCl₂, 1 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride), 1 mM sodium orthovanadate) on ice for 30 min, followed by centrifugation at 100,000 × g at 4 °C for 1 h. This supernatant was taken as the membrane fraction.

Purification of GEM Fractions-- Cells (1 × 10⁸) were centrifuged quickly in cold phosphate-buffered saline after OKT3 stimulation. The pellets were then lysed on ice in 1 ml of 1% Triton X-100 in TNEV buffer (10 mM Tris-HCl (pH 7.5), 150 mM NaCl, 5 mM EDTA, 1 mM Na₃VO₄), with 15 strokes of a Dounce homogenizer, and mixed with 1 ml of 80% sucrose made with TNEV buffer. After transfer of the lysate to the centrifuge tube, 2 ml of 30% sucrose in TNEV buffer was overlaid, and then 1 ml of 5% sucrose in TNEV was overlaid. After centrifugation for 17 h at 200,000 × g in a Beckman SW55Ti, 0.4-ml gradient fractions were collected from the top of the gradient, in which the third fraction contained GEMs.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES

ZAP-70 Is Required for Tyrosine Phosphorylation and Activation of Itk in Response to CD3 Cross-linking-- To study the importance of ZAP-70 in Itk activation during TCR signaling, the human T cell line Jurkat and its ZAP-70-deficient mutant P116 were stimulated by CD3 cross-linking. As has been shown previously (28, 33) and in the top panel of Fig. 1A, Itk was rapidly tyrosine-phosphorylated in Jurkat T cells upon OKT3 stimulation. Tyrosine phosphorylation could be detected within 45 s and returned to basal level by 15 min. In contrast, there was comparatively little increase in tyrosine phosphorylation of Itk in ZAP-70-negative cells receiving the same stimuli. The difference in Itk tyrosine phosphorylation in the two cell lines was not due to differences in Itk recovery, as the immunoprecipitates contained comparable amounts of Itk (Fig. 1A, bottom panel).

View larger version (31K): [in this window] [in a new window]   Fig. 1.   ZAP-70 is required for tyrosine phosphorylation and activation of Itk during TCR signaling. Jurkat and P116 T cells were stimulated with OKT3 ascites at 37 °C for the times indicated, and Itk immunoprecipitates were prepared from 107 cell equivalents of CD3/Lck-cleared postnuclear lysates and subjected either to immunoblotting (A) or to an in vitro kinase assay (B). A, portions of the Itk immunoprecipitates were resolved on a 6% Tris-glycine gel, transferred to nitrocellulose and immunoblotted for phosphotyrosine (4G10), stripped, and blotted for Itk (2F12). B, kinase activity was determined by measuring the tyrosine phosphorylation of an exogenous substrate, RR-SRC (as described under "Experimental Procedures"). S.E. for triplicate samples is shown as error bars along the y axis. Where error bars are not apparent, the symbol is larger than the error bars.

In the same experiment, we also examined whether or not ZAP-70 could regulate increased Itk kinase activity in response to TCR cross-linking. We compared the Itk kinase activity in both Jurkat and P116 cells, as measured by ³²P incorporation into the substrate RR-SRC in an in vitro, immune-complex kinase assay (Fig. 1B). OKT3 stimulation induced a rapid increase in the kinase activity recovered from Itk immunoprecipitates from Jurkat cells, but not in Itk immunoprecipitates from similarly treated ZAP-70-deficient P116 cells. Equal amounts of Itk were detected in the immunoprecipitates (Fig. 1A, bottom panel). Although one group has reported that RR-SRC is not a good substrate for recombinant Itk (33), others have reported the peptide to be a suitable substrate (28), and we find RR-SRC to be a good substrate for both immunoprecipitated and recombinant (not shown) Itk. In particular, it is unlikely that the kinase activity measured in this assay is subject to interference from ZAP-70 or Lck, because RR-SRC is not a ZAP-70 substrate under the assay conditions used, and depletion of Lck from the lysates had no effect upon the OKT3-mediated activation of kinase activity recovered in Itk immunoprecipitations (not shown).

To test whether the failure of tyrosine phosphorylation and subsequent kinase activation of Itk in P116 cells following TCR stimulation is due to the absence of ZAP-70, we examined the phosphorylation and activation of Itk in P.WT18 cells. This stable transfectant of P116 has been characterized previously, and expresses ZAP-70 at levels comparable to those observed in Jurkat (Fig. 2B, top panel, and Refs. 41 and 42). As seen in Fig. 2A, upon stimulation with OKT3, Itk kinase activity in the P.WT18 cells increased 2.7-fold, exhibiting a similar fold increase in activity as that elicited by similarly stimulated parental Jurkat T cells (2.9-fold). P116 cells, on the other hand, had poor Itk activation in response to CD3 cross-linking, exhibiting only a 1.1-fold increase in the kinase activity. No kinase activity was observed in the control normal rabbit serum immunoprecipitations. Equal amounts of Itk were used in the kinase assay for all three cell lines (not shown). The restoration of the responsiveness of Itk kinase activity to CD3 cross-linking correlated with restoration of OKT3-stimulated tyrosine phosphorylation of Itk (Fig. 2B, bottom panel). These results indicate that the defect in Itk activation in P116 cells can be restored by expression of ZAP-70.

View larger version (40K): [in this window] [in a new window]   Fig. 2.   Expression of wild type ZAP-70 in P116 cells restores the phosphorylation and activation of Itk induced by TCR stimulation. A, Jurkat, P116, and P.WT18 were either unstimulated (Unstim.) or stimulated (OKT3) with OKT3 for 45 s at 37 °C. 107 cell equivalents of Itk immunoprecipitates or normal rabbit serum (NRS) immunoprecipitates from CD3-depleted lysates were subjected to an in vitro, kinase assay as in Fig. 1. S.E. for triplicate samples is indicated as error bars along the y axis. B, portions of the Itk immunoprecipitates were resolved on a 6% Tris-glycine gel, transferred to nitrocellulose, and immunoblotted for phosphotyrosine (bottom panel). Whole-cell lysates were resolved on a 4-12% Nu-PAGE gel in MOPS buffer, transferred to nitrocellulose, and immunoblotted for ZAP-70 (top panel).

ZAP-70 PTK Acts Indirectly in Regulating Itk Tyrosine Phosphorylation-- Most protein kinases undergo activation in response to phosphorylation of key residues present within the activation loop of the kinase domain. Itk is no exception, and it has been shown to be positively regulated in response to phosphorylation of tyrosine 511 within its activation loop (33). That this is the principal site of tyrosine phosphorylation-dependent positive regulation of Itk kinase activity has been demonstrated by expression of wild type and Y511F mutant Itk in insect cells. Insect cells expressing wild type Itk show extensive tyrosine phosphorylation of multiple cellular substrates, whereas cells expressing Itk carrying the Y511F mutation do not (33). It is likely therefore that, if the positive regulatory effect of ZAP-70 upon Itk activity were due to direct phosphorylation, Tyr-511 would be the site of phosphorylation in Itk. We therefore tested the ability of the KD of Itk to serve as a substrate for ZAP-70 (Fig. 3). The Itk kinase domain used in this experiment had low intrinsic autocatalytic activity under the conditions used in the assay, allowing us to study the ability of either recombinant Lck or immunoprecipitated ZAP-70 to phosphorylate the Itk kinase domain. Although ZAP-70 could phosphorylate cdb3, a protein previously demonstrated to be a good substrate for ZAP-70, it did not phosphorylate the Itk-KD. Lck, on the other hand, was able to phosphorylate the Itk kinase domain, but not cdb3. Lck also exhibited considerable autophosphorylation. In the absence of added kinase, cdb3 was not phosphorylated in the assay. Under the conditions of the assay, ZAP-70 showed little autophosphorylation. The absence of recognizable ZAP-70 phosphorylation sites within Itk, and the inability of ZAP-70 to phosphorylate the Itk kinase domain in vitro suggest that ZAP-70 acts indirectly to stimulate Itk tyrosine phosphorylation and activation. However, our data do not allow us to rule out the possibility of Itk being phosphorylated by ZAP-70 on tyrosine residues outside its kinase domain, nor that ZAP-70 could phosphorylate Itk in vivo.

View larger version (83K): [in this window] [in a new window]   Fig. 3.   The Itk kinase domain is not an in vitro substrate of ZAP-70. An in vitro kinase assay was carried out with the combination of kinases and substrates indicated, as described under "Experimental Procedures." The kinases were either ZAP-70, immunoprecipitated from 107 cell equivalents of OKT3-stimulated Jurkat T cells, or purified recombinant Lck (1 µg). The substrates were either cdb3 (5 µg) or Itk kinase domain (5 µg). The products of the reaction were resolved on a 4-12% Nu-PAGE gel and transferred to nitrocellulose, and 32P incorporation was detected by autoradiography.

Itk Is Constitutively Located in the Glycolipid-enriched Region of the Plasma Membrane-- Finding no evidence that ZAP-70 was directly phosphorylating and activating Itk, we examined whether or not ZAP-70 was regulating the subcellular distribution of Itk, because it had been previously proposed that Tec family tyrosine kinases can be activated in response to recruitment to the plasma membrane (35, 36, 38-40, 46). Jurkat and P116 T cells were either left unstimulated or stimulated by 45 s of CD3 cross-linking, conditions that give maximal Itk tyrosine phosphorylation and activation. Triton X-100 extracts of cytosolic and membrane fractions were immunoblotted for Itk (Fig. 4A, top panel). A portion of the lysates were subjected to Itk immunoprecipitation and also blotted for Itk (Fig. 4A, bottom panel). Itk was constitutively present in both the cytosolic and membrane fractions, with approximately 50% of the Itk being present in the membrane. This is in stark contrast to the distribution pattern that has been reported for Btk in bone marrow-derived mast cells, in which it was found that greater than 95% of the Btk is present in the cytosolic fraction (38). Because the Itk recovered from unstimulated Jurkat T cells has minimal activity, yet roughly half of the total Itk is constitutively present in the membrane fraction, distribution into the membrane fraction does not appear to be the rate-limiting event in Itk activation in Jurkat T cells. This is in contrast to what has been reported for Btk activation in B cells and mast cells. Furthermore, we could detect no measurable net redistribution between the two pools upon OKT3 stimulation. A similar pattern was also observed in unstimulated and OKT3-stimulated P116 cells.

View larger version (47K): [in this window] [in a new window]   Fig. 4.   Itk is constitutively located in GEMs and exhibits no redistribution with TCR stimulation. Jurkat and P116 T cells were stimulated with OKT3 for 45 s at 37 °C. A, cytosolic (C) and membrane (M) fractions, prepared as described under "Experimental Procedures," were either resolved on a 4-12% Nu-PAGE gel (top panel) or subjected to immunoprecipitation for Itk and analyzed on a 6% Tris-glycine gel (bottom panel), transferred to nitrocellulose, and immunoblotted for Itk (2F12). B, GEMs were prepared by sucrose gradient ultracentrifugation of 1% Triton X-100 cell homogenates (see under "Experimental Procedures"). Fractions 1, 3, and 10 were resolved on a 4-12% Nu-PAGE gel, transferred to nitrocellulose, and immunoblotted for Itk (top panel) and Lat (bottom panel).

It remained possible that although CD3 cross-linking in Jurkat T cells did not result in a gross translocation of Itk from the cytosol to the plasma membrane, perhaps CD3 cross-linking caused a redistribution of Itk between compartments within the plasma membrane itself. It has recently been recognized that the plasma membrane is not uniform, but rather contains lateral domains that differ in composition from bulk plasma membrane. These domains are typified by high concentrations of certain lipids, including cholesterol, glycosphingolipids, sphingomyelin, and phosphatidylinositol 4,5-bisphosphate, although being relatively poor in other phospholipids. These domains are resistant to disruption in nonionic detergents at 4 °C, and have been referred to as lipid rafts, detergent-insoluble membranes (DIGs) or glycolipid-enriched membranes (GEMs) (47, 48). These domains also contain a high concentration of glycosyl phosphatidylinositol-linked proteins and several key signaling molecules, and it has been suggested that these membrane regions serve as signaling rafts or nodes. In T cells, Fyn, Lck, Lat, c-Cbl, CD4, and Ras have been reported to be constitutively targeted to the GEMs, whereas TCR chains, ZAP-70, Shc, PLC1, and Vav can be recovered from the GEMs following TCR stimulation (17, 18, 20). To test whether TCR cross-linking induces a shift of Itk into the GEM fraction of the plasma membrane, GEM preparations were made by sucrose gradient ultracentrifugation as previously reported (49), and the Itk present in fractions 1 (top of gradient), 3 (GEMs), and 10 (Triton X-100 soluble fraction) from the sucrose gradient were assessed by immunoblotting. Itk was present in both the GEMs and the bulk membrane fractions, and the net distribution of Itk between these fractions did not change upon CD3 cross-linking (Fig. 4B, top panel). This same pattern was observed in both Jurkat and the ZAP-70-negative P116 cells. This argues against the regulatory role of ZAP-70 in Itk activation being one of regulating recruitment to the GEMs, because Itk is constitutively present in the GEMs and shows no net recruitment to the GEMs under conditions that lead to Itk activation. Furthermore, there is no difference between Jurkat and P116 with regard to the Itk distribution to the GEMs, whereas there are clear differences with regard to Itk activation in response to OKT3 in these two cell lines. That comparable cell equivalents were loaded from the unstimulated and OKT3-stimulated samples was determined by immunoblotting the same membrane for Lat, the distribution of which between bulk and GEM membrane fraction is not believed to change during the course of the 45-s incubation with OKT3 (Fig. 4B, bottom panel).

Lat Is Required for OKT3-stimulated Itk Activation-- Lat is a recently described T cell- and NK cell-specific transmembrane phosphoprotein that is thought to play a key role as a transmembrane linker protein involved in TCR signaling. Lat has also been reported to be an in vivo substrate for ZAP-70 and rapidly becomes tyrosine-phosphorylated in response to TCR engagement, whereupon it acts as a docking site for key signaling proteins, such as Grb2, Grap, PLC1, PI3-K, Vav, SLP-76, and c-Cbl, recruiting these molecules to the GEM fraction of the plasma membrane (14, 17). In addition, Lat-deficient Jurkat T cells (JCaM2.5) have been reported to be defective in TCR-mediated signaling, including a failure to induce tyrosine phosphorylation of PLC1 or Ca²⁺ mobilization (43). Given that both Lat and Itk are involved in regulating TCR-induced Ca²⁺ influx, and that Lat is a known substrate of ZAP-70, we investigated whether or not Lat is required for Itk activation in response to CD3 cross-linking. This was approached by using the JCaM2.5 Jurkat T cell line and the JCaM2.5B3, Lat-reconstituted line (43). However, this analysis was complicated by the observation that both of these cell lines exhibit deficient Itk expression as compared with Jurkat cells, as can be seen in Itk blots of whole cell lysates and Itk immunoprecipitates (Fig. 5A). There is approximately 8 times less Itk in JCaM2.5 and 2 times less Itk in JCaM2.5B3 compared with Jurkat. Therefore, the number of cell equivalents used in subsequent experiments examining Itk tyrosine phosphorylation and kinase activity were adjusted accordingly to ensure recovery of comparable levels of Itk in the immunoprecipitates. As described previously, Lat expression is very low in JCaM2.5 and is restored by stable expression of an epitopically tagged Lat in JCaM2.5B3, as shown in Lat blots of whole cell lysates (Fig. 5A). Jurkat, JCaM2.5, and JCaM2.5B3 cells were stimulated for 0, 1, or 10 min by CD3-cross-linking with OKT3. The fold increase in the kinase activity of Itk was measured in Itk immunoprecipitates and normalized for the amount of Itk recovered in the immunoprecipitations (Fig. 5B). CD3 cross-linking induced strong activation of Itk kinase activity in Jurkat and JCaM2.5B3 cells but only weakly activated Itk in JCaM2.5 cells, consistent with Lat playing an important role in Itk activation. The tyrosine phosphorylation status of the immunoprecipitated Itk correlated well with the OKT3-induced kinase activity (Fig. 5C, top panel). The amount of Itk recovered in the Itk immunoprecipitates is shown in the bottom panel of Fig. 5C.

View larger version (33K): [in this window] [in a new window]   Fig. 5.   Lat is required for Itk kinase activity and phosphorylation. A, 2 × 105 cell equivalents of the whole-cell lysates or 5 × 106 cell equivalents of the immunoprecipitates of Itk from unstimulated Jurkat, JCaM2.5, and JCaM2.5.B3 cells were resolved on a 4-12% Nu-PAGE gel and blotted for Itk (top panel) and Lat (bottom panel). B, the three cell lines were stimulated with OKT3 at 37 °C for the times indicated. Itk was immunoprecipitated from the Lck and CD3 predepleted whole-cell lysates. 3 µl of the antiserum was used for 1 × 107 Jurkat, 4 × 107 JCaM2.5, 2 × 107 JCaM2.5.B3 cells. 5 × 106, 4 × 107, and 1 × 107 cell equivalents of Jurkat, JCaM2.5, and JCaM2.5.B3, respectively, were used in the in vitro Itk kinase assay (see under "Experimental Procedures"). Each sample was assayed in triplicate, and the average cpm was normalized to the amount of Itk protein. The Itk kinase activity was plotted as fold increase upon OKT3 stimulation over unstimulated. C, the same samples of the immunoprecipitates as in B were resolved on a 6% Tris-glycine gel (top panel) and a 4-12% Nu-PAGE gel (bottom panel) and blotted for phospho-tyrosine (4G10) and Itk (2F12), respectively.

CD3 Cross-linking Induces an Association between Itk and Tyrosine-phosphorylated Lat-- Given an apparent role for Lat in regulating Itk activation, and the established role of Lat in forming multimolecular signaling complexes, we assessed whether or not we could measure a TCR stimulation-induced association between Itk and Lat in Jurkat T cells. Examining by anti-Itk immunoblot the presence of Itk in anti-Lat immunoprecipitations from Brij96 whole-cell lysates of OKT3-stimulated or unstimulated Jurkat and P116 T cells, we found that CD3 cross-linking induced a rapid association between Lat and Itk in Jurkat T cells, but not in the ZAP-70-negative P116 T cells (Fig. 6A, top panel). This observation is also consistent with the report of Zhang et al. (14) who, in addition to noting the OKT3-stimulated association of tyrosine-phosphorylated PLC1, c-Cbl, Vav, and SLP-76 with Lat, also found an unidentified tyrosine-phosphorylated protein of 70 kDa. These researchers found that this protein was not ZAP-70 nor SAM-68. Our results suggest that this protein is likely to be Itk. A longer exposure (not shown) of the membrane in the top panel of Fig. 6A could detect some basal association between Lat and Itk in unstimulated Jurkat T cells but still revealed no association between these two proteins in OKT3-stimulated P116 cells. Equal or greater amounts of Lat were recovered in the anti-Lat immunoprecipitations from the P116 cells, as compared with immunoprecipitations from Jurkat cells (Fig. 6A, bottom panel), and the anti-Lat antiserum was found to be depleting under the conditions used in this experiment (not shown). In a similar experiment, we examined the tyrosine phosphorylation status of Lat in OKT3-stimulated Jurkat and P116 cells. As reported previously Lat is rapidly tyrosine-phosphorylated in Jurkat T cells following CD3 cross-linking (14); however, Lat tyrosine phosphorylation was deficient in P116 cells receiving the same stimulus (Fig. 6B). The amount of the Lat present in the immunoprecipitates is shown in the bottom panel of Fig. 6B.

View larger version (54K): [in this window] [in a new window]   Fig. 6.   CD3 cross-linking induces an association between Itk and tyrosine-phosphorylated Lat. Jurkat and P116 T cells were stimulated with OKT3 for the times indicated at 37 °C. Lat was immunoprecipitated with a rabbit polyclonal antiserum (3023) from 1% Brij 96 lysate (see under "Experimental Procedures"). A, 2 × 105 cell equivalents of whole cell lysates (WCL) or 1 × 107 cell equivalents of Lat immunoprecipitates were resolved on a 6% Tris-glycine gel, transferred onto nitrocellulose, and immunoblotted for Itk with mouse mAb 2F12 (top panel). The anti-Lat immunoprecipitated samples are over-exposed relative to the whole cell lysate samples. The same samples were resolved on a 4-12% Nu-PAGE gel and immunoblotted for Lat (3023) (bottom panel). The whole cell lysate samples are overexposed relative to the anti-Lat immunoprecipitates. B, 5 × 106 cell equivalents of Lat immunoprecipitate was resolved on a 4-12% Nu-PAGE gel, transferred to nitrocellulose, and blotted for phospho-tyrosine (4G10) (top panel) and then stripped and blotted for Lat (3023) (bottom panel).

It should be noted that it has been shown that the mutual co-localization of proteins to the Triton-insoluble GEMs can permit co-precipitation of proteins even when there is no direct interaction between the proteins. Indeed, glycosyl phosphatidylinositol-anchored proteins are co-precipitated with Fyn and Lck in Triton X-100 lysates of T cells, despite the fact that glycosyl phosphatidylinositol-linked proteins only attach to the outer leaflet of the plasma membrane and Fyn and Lck only associate with the inner leaflet (50-52). However, it seems unlikely that the co-immunoprecipitation of Itk with Lat from Brij96 lyastes of OKT3-stumlated Jurkat that we report here results from such a mechanism. First of all, we also observe the anti-CD3-induced association of Itk with Lat in Octylglucoside extracted GEM fractions (not shown), under conditions demonstrated to disrupt greater than 90% of the GEMs (49, 52). Second, we could detect no change in the amount of Itk or Lat present in the GEMs upon OKT3 stimulation nor any significant differences between the two cell lines in the amount of Itk or Lat in the GEM fractions. This being so, GEM-mediated co-immunoprecipitation would be predicted to give the same amount of co-immunoprecipitation from both cell lines, regardless of OKT3 stimulation, which clearly is not what was observed.

Relevant to our observation of constitutive localization of Itk to the membrane fraction in Jurkat T cells, a GFP-Itk fusion protein expressed in Jurkat TAg T cells was also found to be constitutively associated with the plasma membrane and exhibited no detectable redistribution upon TCR stimulation.³ Furthermore, two recent studies examining the effect of knocking out the  isoform of the p85 (and the p55 and p50 splice variants) regulatory subunit of PI3-K on lymphocyte development and responsiveness to antigen found that the p85 knockout mice had impaired B cell development but perfectly normal T cell development (53, 54). Although the ability of Itk within the T cells isolated from these mice to undergo activation in response to TCR engagement has not been reported, it is presumably normal, because these T cells demonstrate normal responsiveness to TCR stimulation (53). This argues against pleckstrin homology domain-mediated recruitment of Itk to the plasma membrane in T cells in response to PI3-K activation as playing a role in T cell development and function. However, it remains possible that the  isoform of PI3-K p85 may be more important in T cell function and that this is compensating for the loss of p85 in T cells isolated from these mice.

It is possible that the phenomenon of constitutive Itk association with the membrane may be a peculiarity of the transformed phenotype of Jurkat T cells, and it remains to be established whether the same pattern is observed in normal T cells. Interestingly, one of the enzymes that terminates the signal initiated by PI3-K is PTEN, a tumor suppressor gene, that dephosphorylates the D-3 position of phosphatidylinositol 3,4,5-trisphosphate. Deficits in either expression or function of PTEN have been reported in a number of hematological malignancies (55). PTEN function in Jurkat has not yet been assessed, but if it is abnormal, then the unopposed basal PI3-K activity could cause recruitment of Itk to the plasma membrane in the absence of TCR stimulation. Interestingly, Mills and co-workers (56) have recently reported that the PI3-K inhibitors wortmannin and LY294002 do not block Itk activation in response to CD3 cross-linking; however, they did not address the effects of these agents on subcellular distribution of Itk. It is also possible that the constitutive association of Itk with the plasma membrane is mediated by the high affinity binding of its pleckstrin homology domain with PIP₂ (57). PIP₂ is constitutively present within the GEM fraction of the plasma membrane. Regardless of the mechanism for the constitutive association of Itk with the plasma membrane, recruitment to the plasma membrane is clearly not sufficient to give Itk activation in Jurkat T cells, because roughly 50% of Itk is recovered from the membrane fraction of unstimulated cells, in which Itk kinase activity is negligible, and further recruitment of Itk to the membrane could not be detected under conditions that cause increased Itk activity.

We propose a working model for Itk activation wherein ZAP-70-initiated tyrosine phosphorylation events are required for the recruitment of additional signaling proteins into the GEMs, in particular proteins that are capable of competitively disrupting the intramolecular association of the Tec homology and Src homology 3 domains of Itk, stimulating Itk activation (32). Proteins likely to be important for this process would be Lat and SLP-76, both of which are involved in forming multimolecular complexes in response to becoming tyrosine-phosphorylated. Our observations that Itk undergoes a TCR engagement-inducible association with Lat and that Itk activation requires Lat expression suggest that Lat might be acting in a ZAP-70-dependent manner to co-localize Itk and its activation partners within the GEM fraction of the plasma membrane.

In conclusion, this represents the first report that the ZAP-70 PTK can regulate the activity of another protein tyrosine kinase. The regulation of Itk by ZAP-70 does not appear to require direct phosphorylation of Itk by ZAP-70. Rather, ZAP-70 indirectly regulates Itk activation, possibly by phosphorylating Lat. We also demonstrate that Lat is required for TCR-stimulated Itk activation and demonstrate that Lat and Itk undergo an inducible association in response to CD3 cross-linking. The facts that this association 1) is induced upon CD3 cross-linking, 2) occurs with kinetics similar to those of Itk tyrosine phosphorylation and activation, and 3) requires ZAP-70 are consistent with this association being required for Itk activation and with this being the site of action of ZAP-70 in the signaling pathway leading to Itk activation. Additional studies will be needed to more fully address the requirement for ZAP-70 and Lat in Itk activation and the nature of the association between Lat and Itk. Additionally, we report finding no evidence to support the plasma membrane recruitment model for the activation of Itk in Jurkat T cells in response to TCR engagement. It will first be necessary to confirm these results in normal T cells before we can assess whether or not this finding is peculiar to the Jurkat T cell line, or whether this reflects the possibility that different Tec family proteins are regulated by different mechanisms in different cell types.

	ACKNOWLEDGEMENTS

We are grateful to Drs. R. Abraham, R. Aebersold, L. Berg, G. Mills, L. Samelson, J. Watts, and A. Weiss for their gifts of reagents. We also thank Drs. S. Bunnell, J. van Leeuwen, D. McVicar, and W. Zhang for many useful conversations and their critical review of the manuscript.

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: NIA, National Institutes of Health, IRP, Gerontology Research Center, Box 12, 5600 Nathan Shock Dr., Baltimore, MD 21224-6825. Tel.: 410-558-8054; Fax: 410-558-8107; E-mail: wanger@grc.nia.nih.gov.

² W. Zhang, R. Tribble, and L. E. Samelson, unpublished results.

³ M. J. Czar and P. L. Schwartzberg, personal communication.

	ABBREVIATIONS

The abbreviations used are: TCR, T cell antigen receptor; Itk, interleukin-2-inducible T cell kinase; Btk, Bruton's tyrosine kinase; Lat, linker for activation of T cells; PI3-K, phosphatidylinositol 3-kinase; PLC1, phospholipase C; GEM, glycolipid-enriched membrane; PTK, protein tyrosine kinase; GST, glutathione S-transferase; PAGE, polyacrylamide gel electrophoresis; KD, kinase domain; MOPS, 4-morpholinepropanesulfonic acid.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES