T-cell receptor–induced phosphorylation of the ζ chain is efficiently promoted by ZAP-70 but not Syk

T-cell immune responses are coupled to activation of the T-cell receptor (TCR). Stimulation of this receptor results in the activation of signal transduction pathways, culminating in expression of cytokines and cellular proliferation.¹  Current models of antigen/major histocompatibility complex (MHC)–induced T-cell activation are presented as ordered events with a sequential interaction of Src and ZAP-70/Syk protein tyrosine kinases (PTKs) with the TCR/CD3/ζ complex. Specifically, TCR engagement activates the Src family PTKs Lck/Fyn, which phosphorylate the tyrosines present in the immunoreceptor tyrosine activation motif (ITAM)^2-4  conserved in the CD3 and ζ subunits of the TCR complex.^5,6  The ZAP-70/Syk PTKs then bind to the phosphorylated ITAMs via their respective SH2 domains, allowing their activation.^2,7,8  Once activated, ZAP-70/Syk kinases phosphorylate downstream, signaling intermediates such as Vav, Lat, and SLP-76 that are required for appropriate recruitment of downstream signaling cascades.

ZAP-70 and Syk are structurally homologous; both proteins are composed of 2 tandemly arranged Src homology 2 (SH2) domains and a carboxy-terminal kinase domain.^7,9  Overall, these 2 kinases share more than 50% sequence identity with conserved tyrosine and serine phosphorylation sites. Although these 2 PTKs have some overlapping functions,¹  it has been hypothesized that in vivo, they are nonredundant because of their distinct expression profiles. ZAP-70 was initially reported to be expressed exclusively in thymocytes, T cells, and natural killer (NK) cells, whereas Syk was described as being expressed in a wide variety of hematopoietic cells but in only low levels in peripheral T cells.¹⁰  Nevertheless, it has since been shown that Syk is expressed at high levels in human CD4⁺ effector T cells and a subpopulation of TCR-stimulated αβ T cells.^1,11  Moreover, the expression of Syk is not limited to hematopoietic cells, as it has been detected in vascular endothelial cells,¹²  in epithelial cells, and in breast tissue.¹³  With regard to ZAP-70, this PTK now has been shown to be expressed throughout normal B-cell development, at least in mice.¹⁴  In humans, ZAP-70 has recently been detected in a subset of chronic lymphocytic leukemia (CLL) B cells, and importantly, its expression has been correlated with a poor prognosis.^15-17  The increased phosphorylation of Syk in ZAP-70–expressing CLL cells was the first indication that ZAP-70 may modulate the activity of Syk^18,19  rather than vice versa. Indeed, until these recent data were reported, Syk was assumed to be “superior” to ZAP-70, as its kinase activity is 100-fold higher than that of ZAP-70,²⁰  it can be activated in an Lck-independent fashion,^21-25  and TCR-induced calcium flux is higher in the presence of Syk than ZAP-70.^26,27 

It has previously been shown that both Syk and ZAP-70 bind to Lck^28-30  and that in vitro, the 2 kinases bind to ζ chain ITAMs with similar affinities.^31-33  Indeed, it has been hypothesized that ZAP-70 may promote ζ ITAM phosphorylation by bringing Lck into the TCR-ζ chain complex.^28,34  In view of these data, it has been perplexing that ζ chain phosphorylation is compromised in ZAP-70–deficient thymocytes and T cells expressing high levels of Syk.^27,35,36  Interestingly, though, in multiple ZAP-deficient T-cell models, the presence of Syk is sufficient to trigger at least a subset of signaling cascades.^27,35-38  These data are supported by previous work indicating that the coupling of TCR signals to some downstream events can proceed independently of the 21- or 23-kDa phosphorylated forms of the ζ chain.^39,40 

The observation that ζ chain phosphorylation is severely attenuated in ZAP-70–deficient T cells expressing Syk indicated that Syk's interaction with this subunit may differ from that of ZAP-70 in vivo. Moreover, the report that ZAP-70 enhances Syk phosphorylation in B-CLL cells¹⁹  suggested that ZAP-70 also may positively modulate Syk activity in T cells. Here, we have investigated the specific contributions of ZAP-70 and Syk in the initial steps of TCR phosphorylation in primary T cells as well as in Jurkat T-cell lines engineered to stably express equivalent levels of ZAP-70 and Syk. We find that the association of ZAP-70 with TCR-ζ was 5- to 10-fold higher than that of Syk following TCR activation. Appropriate tyrosine phosphorylation of TCR-ζ resulted in a notable increase in Lck binding to the receptor in cells expressing ZAP-70 as compared to Syk. Furthermore, the presence of ZAP-70 in Syk-expressing cells was associated with a significantly enhanced tyrosine phosphorylation of this latter kinase. Thus, the propagation of TCR signaling cascades in T cells expressing either Syk or ZAP-70 already differs at the level of the TCR itself.

Human Jurkat clones expressing ZAP-70 and Syk (77-6.8, kindly provided by Dr Kendall Smith, Cornell University, NY), ZAP-70 alone (E6.1), and neither ZAP-70 nor Syk (p116)³⁶  were cultivated in RPMI with 10% fetal calf serum (FCS). p116 cells stably expressing a vesicular stomatitis virus (VSV) epitope-tagged wild-type (WT) or mutant ZAP-70 (Y319F or D461N [kinase dead]) have been previously described.^41,42  CD4⁺ T cells from a previously described ZAP-70–deficient patient²⁷  as well as control CD4⁺ T cells were cultured in Yssel medium⁴³  supplemented with 1% human AB⁺ serum and recombinant human interleukin-2 (IL-2) (100 U/mL). Cells were stimulated every other week with phytohemagglutinin (PHA) (0.5 μg/mL) (Murex, Dartford, England) and irradiated feeder cells consisting of peripheral blood mononuclear cells (PBMCs) and Epstein-Barr virus (EBV)–transformed JY cells as described.⁴⁴  Primary T-lymphocyte experiments were performed on “resting phase” cells that had not been stimulated with irradiated feeder cells during the 10 preceding days. Prior to activation, cells were rested overnight in the indicated medium without FCS or IL-2.

The αZAP-70, αζ-chain, and αSyk Abs were generous gifts of Art Weiss (UCSF, CA). The phosphospecific (Y319/Y352) ZAP/Syk Ab and pAb recognizing the phosphorylated T183/Y185 form of Erk1/Erk2 were from Cell Signaling Technologies (Beverly, MA). The polyclonal αLck Ab was kindly provided by B. Sefton (La Jolla, CA). The αErk2 mAb was from Transduction Laboratories (Lexington, KY), and the 4G10 αphosphotyrosine mAb was from Upstate Biotechnology (UBI, Lake Placid, NY). The αCD3 UCHT1 and αCD4 ST4 mAbs were kindly provided by Doreen Cantrell (Cancer Research UK, London, England) and Sanofi (Montpellier, France), respectively. The αCD69-PE mAb and αmouse F(ab′)₂ fragment were from Immunotech (Marseille, France). Recombinant human ZAP-70 protein was purchased from UBI, and recombinant GST-Syk protein was generously provided by Peter Coopman (University Montpellier, France).

The human wild-type ZAP-70 was expressed from a previously described Moloney murine leukemia virus (MLV)–based retroviral vector, LZRS-ZAP-70/EGFP,⁴⁵  wherein enhanced green fluorescence protein (eGFP) is downstream of an encephalomyocarditis-derived internal ribosomal entry site.^46,47  Virions harboring the LZRS-ZAP-70/EGFP vector were produced in the PG13 cell line expressing the Gibbon ape leukemia virus (GALV) envelope.^45,48  The human WT Syk cDNA¹³  was cloned into the BamHI and XhoI sites of LZRS. The LZRS-Syk/EGFP retroviral vector was packaged in the 293T-based amphotropic Phoenix cell line.⁴⁷  Briefly, a pool of Phoenix/LZRS-Syk/EGFP cells was obtained by 3 sequential sorts of high-EGFP expressing cells on a FACSVantage Flow Cytometer (Becton Dickinson, San Jose, CA), followed by limiting dilution culture. All vector-containing retroviral supernatants were harvested after a 24-hour incubation of confluent cells at 32° C. The collected culture medium was filtered through 0.45-μm filters and stored at –80° C for further use.

ZAP-70–deficient CD4⁺ primary T cells and Jurkat T-cell clones were transduced with the indicated retroviral vectors on fibronectin-coated plates (8 μg/cm,²  kindly provided by Takara Shuzo, Shiga, Japan) essentially as reported.^45,48  After a 6-hour exposure to retrovirus, primary T cells or Jurkat cell lines were incubated in fresh medium with recombinant IL-2 or FCS, respectively. After an overnight incubation, the transduction procedure was repeated. Transduced cells were sorted on a FACSVantage Flow Cytometer on the basis of EGFP expression. Levels of EGFP in sorted cell lines were verified prior to all experiments and were more than 90%.

Cells (2 × 10⁷ cells/mL) were stimulated with an αCD3 mAb or αCD3/αCD4 mAbs (2.5 μg/mL) followed by cross-linking with an αmouse F(ab′)₂ fragment (5 μg/mL) or with pervanadate (0.1 mM Na₃VO₄ and 0.3 mM H₂O₂). After activation, cells were lysed in a buffer containing 1% NP40 and 60 mM n-octyl-B-D-glucoside together with protease/phosphatase inhibitors. Postnuclear supernatants were immunoprecipitated for 1 hour at 4° C with the indicated Ab followed by collection on protein A Sepharose beads (Pharmacia, Uppsala, Sweden).⁴⁹  Immunoprecipitates or whole-cell lysates were boiled, resolved on sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) gels, and transferred to nitrocellulose membranes (Schleicher & Schuell, Dassel, Germany). Membranes were blocked for 1 hour in Tris (tris(hydroxymethyl)aminomethane)–buffered saline (TBS) (150 mM NaCl, 20 mM Tris, pH 7.5) containing 5% milk and 0.1% Tween 20, incubated with the indicated primary Ab for 1 hour at room temperature, incubated with horseradish peroxidase–conjugated goat–αrabbit or αmouse secondary Abs (Amersham, Arlington Heights, IL), and visualized using enhanced chemiluminescence (ECL). For reblotting, filters were stripped as reported.⁴⁹  Band intensities were quantified using NIHimage software (Research Services Branch, National Institute of Mental Health, Bethesda, MD).

T-cell receptor signaling in the absence of the ZAP-70/Syk protein tyrosine kinases is defective, with markedly decreased activation of downstream effector molecules. Furthermore, CD3-mediated phosphorylation of the TCR ζ chain is markedly diminished in primary CD4⁺ T cells isolated from ZAP-70–deficient patients^{35,37,38,50,51}  (Figure 1A) as well as in a ZAP-70–deficient Jurkat T-cell line (p116).³⁶  Notably, this defect is observed even in the presence of high levels of endogenous or even ectopic Syk.^27,35,36,38 

We previously characterized primary T cells from ZAP-70–deficient patients with significant expression of endogenous Syk (herein referred to as ZAP^–/Syk⁺).²⁷  These cells were used to determine whether defective ζ phosphorylation is related to the presence of Syk or, alternatively, is related to the absence of ZAP-70. To this end, ZAP-70 was introduced into these primary T cells by means of a MuLV-based retroviral vector harboring the wild-type ZAP-70 gene fused to a carboxy-terminal VSV-G tag as previously described.⁴⁵  Importantly, both basal and CD3-induced ζ chain phosphorylation were reconstituted in these Syk⁺ T cells expressing ectopic ZAP-70 (herein referred to as ZAP^E/Syk⁺) (Figure 1A). Of note, these ZAP^E/Syk⁺ cells express levels of Syk similar to the parental ZAP-70–deficient T cells, indicating that the defect in ζ chain phosphorylation was not due to Syk expression per se but to the absence of ZAP-70 (Figure 1A). Additionally, the level of phosphorylation of the ITAMs within the ζ chain is obviously dependent on the level of the ζ chain itself as well as the quantity of Lck associated with CD4. ZAP^–/Syk⁺ and ZAP^E/Syk⁺ cells expressed equivalent levels of ζ chain, surface CD4, and Lck-associated CD4 (Figure 1A; data not shown). Thus, these data demonstrate that the defective ζ chain phosphorylation observed in primary ZAP-70–deficient T cells expressing Syk was not the result of abnormalities in other TCR effector molecules.

In view of the many similarities between ZAP-70 and Syk,¹  it was curious that ζ phosphorylation differed to such a great extent in the presence of these 2 related proteins. As such, we assessed the association of both kinases with TCR-ζ. As observed in Figure 1B, the association of Syk and ZAP-70 with TCR-ζ was marginal in unstimulated cells, but the binding of both proteins to this receptor subunit increased following receptor cross-linking with αCD3/αCD4 antibodies. Thus, both Syk and ZAP-70 are capable of associating with TCR-ζ. However, tyrosine phosphorylation of associated Syk molecules was significantly less than that of associated ZAP-70 molecules, as detected with an anti-phosphotyrosine monoclonal antibody (Figure 1B). Indeed, phosphorylation of TCR-associated Syk was not always detectable (data not shown). Nevertheless, one difficulty in interpreting these experiments stems from the fact that 2 different antibodies were used to detect ZAP-70 and Syk, and as a consequence it was impossible to directly compare the relative levels of ZAP-70 and Syk in these cells. Of note, Ashe et al have found that in thymocytes, the extent of ζ phosphorylation is related to the level of ZAP-70 expression.³⁴  As such, we could not exclude the possibility that the defective ζ phosphorylation observed in these primary T cells was due to a level of Syk lower than that of endogenous ZAP-70 in WT T cells.

To circumvent this problem and rigorously determine whether ZAP-70 and Syk, when expressed at equivalent levels, differentially mediate ζ chain phosphorylation, we attempted to introduce ectopic ZAP-70 and Syk into the p116 Jurkat T-cell line that does not express either protein. Notably though, stable introduction of high levels of Syk into the p116 cell line has not previously been achieved, due to significant toxicity.³⁶  As such, we chose to introduce Syk and ZAP-70 via retroviral vectors where their expression is driven from the long terminal repeat (LTR), as we have determined that their expression from this vector is 1- to 3-fold that of the endogenous proteins.⁴⁵  In this vector, either ZAP-70 or Syk is expressed concomitantly with EGFP, which is encoded downstream of an internal ribosome entry site. p116 Jurkat cells were transduced with virions harboring one of these 2 vectors, and transduced cells were sorted by FACS on the basis of EGFP expression. In an attempt to generate cell lines expressing equivalent levels of Syk and ZAP-70, multiple pools of cells with similar EGFP fluorescence were FACS sorted, and pools that were subsequently determined to express similar levels of Syk and ZAP-70 are shown in Figure 2A. It should be noted that this approach was chosen rather than the more conventional approach of tagging the 2 kinases with the same peptide because the presence of a tag has been reported to reduce the biologic activity of Syk.²⁵  Importantly, p116 cell lines stably expressing ZAP-70 as well as Syk could be established using this protocol.

To determine whether these 2 cell lines did indeed express equivalent levels of ZAP-70 and Syk, herein referred to as ZAP^E/Syk^– and ZAP^–/Syk^E, respectively, protein levels were monitored in comparison with known amounts of recombinant ZAP-70 and Syk proteins. We were thereby able to determine that these 2 cell lines expressed approximately 20 pg of either Syk or ZAP-70 (Figure 2B). This ectopic level of ZAP-70 is similar to that of the endogenous ZAP-70 in ZAP⁺/Syk⁺ Jurkat cells, and the level of ectopic Syk is approximately 3-fold higher than that of the endogenous Syk in these ZAP⁺/Syk⁺ Jurkat cells (Figure 2C). The nomenclature of these various cell lines and their respective CD3 expression levels are noted in Table 1.

We then compared the abilities of ZAP-70 and Syk to promote ζ chain phosphorylation in these p116-derived cell lines expressing equivalent levels of these 2 kinases. In parental ZAP^–/Syk^– p116 cells, basal ζ chain phosphorylation was difficult to detect and upon CD3 cross-linking, only the partially phosphorylated p21 ζ isoform was evident. Introduction of ZAP-70 into these cells resulted in increased basal phosphorylation of ζ as well as induction of both the p21 and p23 phosphorylated ζ isoforms following CD3 cross-linking. In contrast, introduction of Syk did not increase either basal or CD3-stimulated ζ phosphorylation (Figure 3). These data are similar to those obtained in primary T cells, demonstrating the ineffectiveness of Syk in promoting ζ chain phosphorylation. It is important to note that the introduced Syk was functional, as full phosphorylation of the ζ chain was detected in ZAP^–/Syk^E but not in ZAP^–/Syk^– cells following activation with pervanadate, an agent that indiscriminately activates intracellular kinases independently of the TCR.⁵² 

Both ZAP-70 and Syk were recruited to the ζ chain, with a 2- to 4-fold augmentation following CD3 engagement. As the antibodies used to detect these 2 proteins are not the same, it was not possible to compare the actual quantity of ζ-associated Syk with that of ZAP-70. However, tyrosine phosphorylation of the associated ZAP-70 protein increased following CD3 ligation, whereas tyrosine phosphorylation of ζ-associated Syk was either not detectable or not significantly higher than basal levels (Figure 3). Although the observed low level of Syk phosphorylation may at least in part be due to its weak association with ζ, our data indicate that it is also related to particularities of TCR-mediated signaling vis-a-vis Syk. Specifically, pervanadate treatment resulted in an increased phosphorylation of ζ-associated Syk, without altering the overall level of Syk association with ζ (Figure 3).

In the experiments presented in Figure 3 as well as those performed in primary T cells, binding of Syk to the TCR may have been impeded by the defective phosphorylation state of ζ. Therefore, we assessed the relative associations of ZAP-70 and Syk in a Jurkat clone (77-6.8) where both proteins are endogenously expressed (ZAP^+/Syk⁺; Table 1). In these latter cells, the ζ chain is fully phosphorylated following CD3 engagement (Figure 4A).

The amounts of ZAP-70 and Syk associated with ζ were determined by quantifying their relative levels in ζ-immunoprecipitates and total cell lysates. In this manner, we were able to determine that under conditions of full ζ phosphorylation, a significantly higher percentage of ZAP-70 associates with ζ. Specifically, following CD3 ligation, 16% of ZAP-70 but only 3% of Syk associated with ζ in a representative experiment. The level of binding of both proteins was further augmented following stimulation with pervanadate, but the relative increase in ZAP-70 association was significantly higher: 70% of ZAP-70 and only 7% of Syk (Figure 4B-C). Thus, in activated T cells expressing ZAP-70 together with Syk, the association of ZAP-70 with ζ is 5- to 10-fold higher than that of Syk, indicating that these proteins are differentially coupled to the TCR.

The catalytic activity of Lck is required for the appropriate phosphorylation of the ITAMs within the TCR ζ chain as well as for ZAP-70 itself.²  Furthermore, the association of the CD4/Lck complex with the ζ subunit²⁸  is crucial for subsequent activation steps. As such, we assessed whether Lck association with ζ was differentially modulated in the presence of the ZAP-70 and Syk kinases. In ZAP^–/Syk^– cells, there is a slight increase in the association of p56 and p59 Lck isoforms with ζ upon receptor cross-linking, and although this response is not significantly augmented by the introduction of Syk, it is clearly enhanced by ZAP-70 (Figure 5A). Importantly, this increased association of Lck with the TCR in the presence of ZAP-70, as compared to Syk, also was observed in primary T cells (data not shown).

It previously has been proposed that Lck binding to ZAP-70 phosphorylated on tyr319 stabilizes the interactions of Lck with the TCR.³⁰  It was therefore important to assess the level of tyrosine 319 and 352 phosphorylation on ζ-associated ZAP-70 and Syk, respectively. In TCR-ζ immunoprecipitates, there was a significant increase in tyrosine 319 phosphorylation of ZAP-70 following CD3 engagement, whereas no phosphorylation of the homologous tyrosine 352 residue of Syk was detected (Figure 5B). Phosphorylation of this tyrosine was clearly visible in pervanadate-treated cells and was not a function of increased Syk association. Indeed, equivalent levels of Syk were associated with the ζ chain in CD3- and pervanadate-stimulated cells (Figure 5B). Thus, the defective Syk phosphorylation of tyrosine 352 is specific to signaling mediated via the TCR.

The data presented in Figure 5A and 5B indicated that increased ζ-Lck binding is associated with phosphorylation of ZAP-70 on tyr319. Nevertheless, it was not clear whether the Y319 residue of ZAP-70 played a positive regulatory role in ζ phosphorylation. We therefore assessed ζ phosphorylation in p116 cells stably expressing a Y319F mutant of ZAP-70. CD3-induced ζ phosphorylation was not significantly modulated, although basal ζ chain phosphorylation was notably decreased (Figure 5C). Moreover, neither basal nor CD3-induced ζ chain phosphorylation required ZAP-70 kinase activity, as previously reported.^34,36,41  Rather, introduction of a kinase-dead (KD) ZAP-70 mutant into p116 cells appeared to augment ζ phosphorylation (Figure 5C). While this result was somewhat surprising, the KD mutant had the expected dominant-negative effect on downstream signaling with a block in CD3-induced Erk phosphorylation.⁵³  Altogether, these data show that, at least in Jurkat cells, the dependence of Y319 phosphorylation on ζ chain phosphorylation differs under constitutive and TCR-induced conditions.

In T cells expressing both ZAP-70 and Syk, we discerned that although ZAP-70 was always the main phosphorylated kinase associated with the TCR, Syk phosphorylation appeared to be enhanced as compared with cells expressing Syk alone. As ζ-associated Syk accounts for less than 10% of total Syk, it was important to determine whether the global activation state of total cellular Syk (of which the majority was not associated with ζ) is enhanced by ZAP-70. To respond to this question, ZAP-70 and Syk were immunoprecipitated from the Jurkat T-cell lines engineered to express one or both of these kinases, and their relative tyrosine phosphorylation levels were assessed. As expected, tyrosine phosphorylation of ZAP-70 was stimulated by ligation of CD3 but surprisingly, CD3-induced Syk phosphorylation was only minimal in the absence of ZAP-70 (Figure 6). Importantly, Syk phosphorylation was significantly increased in the presence of ZAP-70, but the converse was not observed. The enhancing effect of ZAP-70 on Syk tyrosine phosphorylation was specific to CD3 ligation, as Syk is highly phosphorylated in the absence of ZAP-70 following pervanadate treatment, and its pervanadate-induced phosphorylation was not increased by ZAP-70 (Figure 6).

The ensemble of data presented here indicated that ζ phosphorylation and Lck association is decreased in Syk-expressing cells, in the absence of ZAP-70. Nevertheless, we and others have found that the presence of Syk is sufficient to trigger at least a subset of signaling cascades in ZAP-70–deficient T cells.^27,35-38  It was therefore important to study downstream signaling in these cells. As such, we assessed the kinetics of Erk phosphorylation and found that although Erk phosphorylation was observed in all cell lines 3′ following CD3 engagement, the signal was significantly lower in both the ZAP^–/Syk^– and ZAP^–/Syk^E cells as compared to ZAP^E/Syk^– cells at 5′ and 10′ after activation (Figure 7A). The distinct kinetics of Erk phosphorylation in the absence and presence of ZAP-70 previously has been reported,⁵³  and we now show that Syk is not equivalent to ZAP-70 with regard to this response. However, both ZAP-70 and Syk were associated with increased constitutive surface levels of the CD69 activation marker, although CD3-induced CD69 surface expression was higher in ZAP-expressing cells (Figure 7B). Altogether, these data point to differences between ZAP-70 and Syk in modulating constitutive and CD3-induced signaling responses.

The data presented here demonstrate that ZAP-70, but not the related Syk protein tyrosine kinase, promotes phosphorylation of the TCR-ζ chain ITAMs in primary human CD4⁺ T cells as well as in transformed Jurkat T leukemia cells. The vast majority of previous work has focused on the role of the Lck/Fyn kinases in mediating ITAM phosphorylation.^2-4,54  However, Ashe et al have reported that ZAP-70 promotes TCR ITAM phosphorylation in immature double-positive thymocytes.³⁴  The role of ZAP-70 in promoting ζ ITAM phosphorylation does not appear to be dependent on its proper kinase activity,^34,36,41  but rather ZAP-70 is hypothesized to act as a “signal amplifier” by bringing Lck into the TCR-ζ chain complex.²⁸  ZAP-70 also has been proposed to enhance TCR-ζ chain phosphorylation by protecting ITAMs from dephosphorylation.²  However, this does not appear to be its major role. As in murine thymocytes,³⁴  we obtained similar ITAM phosphorylation data in the absence or presence of phosphatase inhibitors (not shown).

The association of ZAP-70 with the TCR-ζ chain was 5- to 10-fold higher than that of Syk. Moreover, even in pervanadatetreated cells where TCR-ζ was fully phosphorylated, fewer than 10% of Syk was associated with the TCR. Indeed, irrespective of the activation conditions, ZAP-70 was always the dominant kinase associated with the TCR. These data may appear to be somewhat in contradiction with previous studies reporting similar affinities of the ZAP-70 and Syk proteins for the various TCR ITAMs.^31-33  Nevertheless, the aforementioned studies were performed in vitro with peptide ITAMs and not in a cellular context where many other variables come into play.

In addition to an inferior recruitment of Syk to the TCR-ζ chain following CD3 engagement, it is also important to note that ζ-associated Syk was poorly tyrosine phosphorylated. In accord with these data, phospho-Syk/TCR complexes have been difficult or impossible to detect in various T-cell models, even under conditions where both TCR-ζ and Syk were highly phosphorylated.^25,55  This does not appear to be due to a more efficient internalization of the TCR in the presence of Syk as compared to ZAP-70. Dumont et al previously reported that at low TCR occupancy, down-regulation of the TCR is significantly augmented by the presence of ZAP-70.⁴¹  Moreover, we have found that the down-regulation of CD3 induced by its own engagement is lower in Syk-expressing T cells than in ZAP-70–expressing T cells at early time points (60 minutes), but this difference is not significant by 120 minutes after activation (data not shown).

With regard to the domain(s) of ZAP-70 responsible for promoting ζ phosphorylation, it was previously shown that its kinase activity is dispensable.^34,36,41  Intriguingly, we report here that Y319 phosphorylation of kinase-dead ZAP-70 is increased as compared to the WT protein, and this phenomenon is associated with augmented ζ-chain phosphorylation. Although elimination of the entire interdomain of ZAP-70 (including regulatory tyrosines Y292, Y315, and Y319) has not been found to modulate ζ phosphorylation,⁵⁶  the importance of an Lck/ZAP-70 complex, mediated via Y319 of ZAP-70, is underlined by the finding that inhibiting this interaction results in reduced IL-2 secretion following TCR activation.^42,57  Moreover, the individual mutation of Y315 and Y292 of ZAP-70 has been shown to result in attenuated and augmented ζ chain phosphorylation, respectively.^58,59  We observed defective basal, but not CD3-induced, ζ chain phosphorylation in p116 cells into which a Y319F ZAP-70 mutant was introduced. Altogether, these data point to a complex interplay between the interdomain tyrosines in modulating both basal and CD3-induced ζ chain phosphorylation.

In our studies, the introduction of ZAP-70 in primary ZAP-70–deficient CD4⁺ T cells resulted in significantly enhanced TCR-ζ phosphorylation. It is interesting to note that the converse, that is, augmented TCR-ζ phosphorylation in the presence of Syk, also has been shown to be the case in cell models where Lck is either absent or not recruited due to the absence of CD4/CD8.^24,25  The distinct effects of ZAP-70 and Syk in these models may be due to the fact that the coupling of CD4 to the formation of a TCR signaling complex is not equivalent in all T cells. Indeed, in Th1 cells, the MHC/peptide-mediated recruitment of TCR complexes into raft structures is regulated by CD4, but this is not the case for Th2 cells.⁶⁰  Thus, the importance of stabilizing a CD4/Lck/ZAP-70/CD3-ζ complex is likely to differ in Th1 and Th2 cells. Under conditions where the complexing of CD4/Lck to the TCR is expendable, Syk may play a more significant role in positively regulating downstream TCR signaling cascades.

It is notable that some signaling cascades, such as the Ras-MAPK pathway, can be activated by CD3 engagement in T cells with altered TCR-ζ chain phosphorylation and defective Lck/ZAP-70 activation, albeit with distinct kinetics (Figure 7A).^27,53,61-63  These T cells are not “null” with regard to signal transduction but are clearly modulated. Indeed, we previously reported that in ZAP-70–/Syk⁺ primary T cells with defective TCR-ζ chain phosphorylation, at least a subset of proximal signaling intermediates such as linker for activation of T cells (LAT), phospholipase C γ1 (PLCγ1), and SH2-containing leukocyte protein 76 (SLP-76) are activated, but IL-2 secretion and proliferation are diminished.²⁷  While the decreased TCR responsiveness of Syk-expressing T cells may seem somewhat surprising given the higher kinase activity of Syk,²⁰  it is in agreement with a recent report demonstrating significantly augmented TCR-induced up-regulation of CD69 and IL-2 secretion in cells expressing a ζ/ZAP-70 chimera as compared to a ζ/Syk chimera.⁶⁴  Importantly, TCR-induced signaling pathways also may be regulated by changes in the level of the ζ chain itself. In T cells isolated from some patients with malignancies, autoimmune diseases, viral and bacterial infections, as well as in vitro–activated CD4⁺ effector T cells, ζ expression is significantly decreased.^11,65-71  Thus, alterations in ζ chain phosphorylation as well as expression are likely to add to the versatility of TCR signaling via a differential recruitment of downstream signaling cascades.

As per the physiological consequence of concomitant ZAP-70/Syk expression, it now appears that the 2 protein tyrosine kinases are simultaneously expressed under both physiological and pathological conditions: both kinases have been reported to be expressed in immature thymocytes, naive transgenic CD8⁺ T cells, CD4⁺ effector T cells, pre–B cells, and a subset of poor-prognosis B-CLL cells.^{10,11,14-17,72}  Importantly, the presence of ZAP-70 in Syk-expressing Jurkat T cells resulted in increased global levels of phosphorylated Syk. This was a particularity of ZAP-70 as the converse, augmentation of ZAP-70 phosphorylation in the presence of Syk, was not observed. The ensemble of the data presented here as well as that reported in a T-cell hybridoma²⁵  and human B-CLL cells¹⁸  strongly suggests that the interplay between ZAP-70 and Syk, rather than increased expression of one or the other kinase, expands the responsiveness of T and B lymphocytes to receptor stimulation.

We are grateful to G. Bismuth, P. Coopman, S. Latour, J. Madrenas, and R. Wange for helpful input at various stages of this work, and to C. Dupperay for expert assistance with FACS sorting. We are indebted to P. Coopman, A. Weiss, B. Sefton, and D. Cantrell for generously providing reagents and antibodies, and to Ikunoshin Kato and Setsuko Yoshimura of Takara Shuzo for providing the recombinant fibronectin fragment. We appreciate the continuous input of V. Dardalhon, S. Jaleco, S. Kinet, and C. Mongellaz, and critical comments from R. Hipskind, M. Villalba, and M. Sitbon.