Herpesvirus saimiri replaces ZAP-70 for CD3- and CD2-mediated T cell activation.

The protein tyrosine kinase ZAP-70 plays a pivotal role involved in signal transduction through the T cell receptor and CD2. Defects in ZAP-70 result in severe combined immunodeficiency. We report that Herpesvirus saimiri, which does not code for a ZAP-70 homologue, can replace this tyrosine kinase. H. saimiri is an oncogenic virus that transforms human T cells to stable growth based on mutual CD2-mediated activation. Although CD2-mediated proliferation of ZAP-70-deficient uninfected T cells was absent, we could establish H. saimiri-transformed T cell lines from two unrelated patients presenting with ZAP-70 deficiencies. In these cell lines, CD2 and CD3 activation were restored in terms of [Ca(2+)](i), MAPK activation, cytokine production, and proliferation. Activation-induced tyrosine phosphorylation of zeta remained defective. The transformed cells expressed very high levels of the ZAP-70-related kinase Syk. This increased expression was not observed in the primary T cells from the patients and was not due to the transformation by the virus because transformed cell lines established from control T cells did not present this particularity. In conclusion, wild type H. saimiri can restore CD2- and CD3-mediated activation in signaling-deficient human T cells. It extends our understanding of interactions between the oncogenic H. saimiri and the infected host cells.

The tyrosine kinase ZAP-70 is essential for activation of mature T cells via CD3. An autosomal recessive form of severe combined immunodeficiency in humans has been described as resulting from mutations within the gene encoding ZAP-70. This deficiency is characterized by an absence of CD8 ϩ T cells and an increased number of nonfunctional CD4 ϩ T cells with a mature phenotype in the periphery. These CD4 ϩ T cells are unresponsive to either antigenic stimulation in vivo or CD2and CD3-mediated activation in vitro (1)(2)(3).
According to current concepts, binding of antigen to the T cell receptor (TCR) 1 initiates a cascade of early signaling events, which includes activation of the protein tyrosine kinases (PTKs) of the Src family. These PTKs phosphorylate the immune-receptor tyrosine-based activation motifs, which are present in all the chains of the CD3-complex. This allows the recruitment of ZAP-70, which is then phosphorylated and activated and subsequently phosphorylates a number of key substrates including LAT, SLP-76, and Vav. These tyrosine kinase reactions are required for CD3-induced mobilization of intracellular free calcium ([Ca 2ϩ ] i ) and activation of the Ras/mitogen-activated protein kinase (MAPK), leading to cytokine production and proliferation, responses that are all defective in the absence of ZAP-70 (4). ZAP-70 not only plays a crucial role in CD3-mediated T cell activation but also in CD2-mediated activation (5,6). CD2 constitutes the so-called alternative pathway of T cell activation (7); simultaneous triggering of two distinct epitopes on CD2 by two mAbs induces T cells to proliferate and secrete lymphokines in the absence of antigen and antigen-presenting cells.
A mutual activation via CD2 is the basis of the autocrine growth of human T cells transformed by Herpesvirus saimiri (8). H. saimiri is an oncogenic virus that induces leukemia and lymphoma in New World and Old World primates (9). This virus transforms human T cells to stable growth in vitro (10). Human T cells transformed by this virus retain essential properties of native T cells. In particular they display a structurally and functionally intact TCR and show a grossly unaltered sensitivity to different apoptotic pathways. The preservation of an intact TCR distinguishes H. saimiri-transformed T cells from T cells transformed with human T cell leukemia virus-1, which tend to lose their TCRs (11).
An essential difference between native uninfected T cells and H. saimiri-transformed T cells is their differential requirement for CD2-mediated activation. Native uninfected T cells are activated via CD2 only by certain pairs of mAbs but not by the binding of CD2 to its ligand CD58. In contrast, H. saimiritransformed T cells are activated by interaction with CD58bearing cells or, alternatively, by a single mAb to the T11.1 epitope of CD2 provided the mAb is cross-linked (8).
The purpose of this study was to obtain further insights into the mechanisms responsible for the transformation of T cells by H. saimiri. To this end, we studied activation of H. saimiritransformed T cell lines established from primary T cells of two severe combined immunodeficiency patients with ZAP-70 deficiencies. We report herein that H. saimiri transforms ZAP-70deficient T cells to stable growth and can overcome the requirement of ZAP-70 for T cell activation. CD2 and CD3 activation of these T cell lines induces activation of MAPK, increase of [Ca 2ϩ ] i , and cytokine production.

EXPERIMENTAL PROCEDURES
Zap-70-deficient Patients, T Cell Culture, and Transformation with H. saimiri-We studied two unrelated severe combined immunodeficiency patients presenting with defects in ZAP-70 expression. The first patient has been described previously (6) and has a homozygous deletion in the zap-70 gene, which leads to a complete absence of ZAP-70 protein expression. The second patient had clinical phenotype typical for ZAP-70 deficiency (3) and biochemically a complete lack of ZAP-70 protein expression (Fig. 1), but the genetic abnormality is not yet defined. Peripheral blood mononuclear cells from both ZAP-70-defective patients were activated with phorbol 12-myristate 13-acetate ϩ ionomycin and expanded in IL-2 for a few days. Subsequently the cells were infected with H. saimiri strain C488 and cultured as described (11). As controls, we used several H. saimiri-transformed T cell clones and lines (ES-8T, SS-8T, Wi-4T, DT, and HT) established from control donors and the transformed cell line CB15 obtained from cord blood.
CD2-and CD3-mediated Cytokine Production-For CD2-mediated activation, the mAb 39C1.5, which recognizes the T11.1 epitope (Coulter Immunotech, Hamburg, Germany), was applied, and for CD3mediated activation, the mAb OKT-3 was applied. The mAbs were cross-linked with murine A20 cells, which express high amounts of Fc receptors. Alternatively, the stimulating mAbs were cross-linked by absorbing to the plastic plates. As a control a mAb recognizing CD58 (purified from the hybridoma TS2/9.1.4.3, which was obtained from ATCC) was used. The Src family-specific inhibitor PP2 (Calbiochem, Bad Soden, Germany) and the Syk/ZAP-70-specific inhibitor piceatannol (Sigma) were applied. PP2 was dissolved in Me 2 SO at 20 mg/ml (66 mM) and used at a final concentration of 10 M. Piceatannol was dissolved in Me 2 SO at 10 mg/ml (41 mM) and used at a final concentration of 10 g/ml. Both inhibitors were used immediately after dissolving. The T cells were preincubated with the diluted inhibitors in the incubator for 30 min and subsequently stimulated. At the applied concentrations, the inhibitors were not cytotoxic, and the solvent Me 2 SO did not interfere significantly with T cell activation. About 5 ϫ 10 4 T cells were seeded per well in a volume of 200 l of medium without IL-2. All experiments were performed in triplicate. Supernatants were collected 24 h after activation. Production of TNF-␣, interferon-␥, and IL-2 was determined by ELISA.
Proliferation Assay-H. saimiri-transformed T cells were seeded in 96-well flat-bottom plates at a density of 2.5 ϫ 10 4 or 1 ϫ 10 5 cells/well in 200 l of culture medium without IL-2. The anti-CD2 mAb 39C1.5 or control rat Ig was added at 4 g/ml at four time points. Three days later, 0.2 Ci of [ 3 H]thymidine (Amersham Pharmacia Biotech) was added for another 16 h. Cultures were harvested and analyzed with the direct ␤ counter Matrix TM96 (Packard Instrument Co.).
Immunoprecipitation and Western Blot Analysis-H. saimiri-transformed T cells or primary T cells prepared as described (6) were left unactivated or activated for 3 min at 37°C in the presence of the anti-TCR mAb UCHT1 (ascitic fluid at 1/1000) or a combination of the two anti-CD2 mAbs X11 and D66 (kindly given by Dr L. Boumsel, INSERM U448, France). Cells were then lysed in lysis buffer (20 mM Tris-HCl, pH 7.4, 140 mM NaCl, 2 mM EDTA, 50 mM NaF, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 100 M Na 3 VO 4 , and protease inhibitors) for 20 min at 4°C. Nuclei and cell debris were removed by centrifugation. Protein concentrations were determined in the postnuclear lysates using a Bio-Rad kit. For immunoprecipitations, the same amount of lysate was precleared at 4°C by rocking with mouse or rabbit purified IgG for 1 h at 4°C. Then protein G-Sepharose beads were added and the nonspecific immunoprecipitates recovered by centrifugation. After this preclearing, lysates were incubated overnight with anti-Syk Abs (Santa Cruz, sc-1077), with anti-mAb (Santa Cruz, sc-1239), or anti-ZAP-70 Abs. Specific immunoprecipitates were recovered by the addition of protein G-Sepharose beads for 1 h and were washed three times in lysis buffer. Immunoprecipitates or post-nuclear cell lysates were then run on standard SDS-polyacrylamide gel electrophoresis and transferred to a polyvinylidene difluoride membrane (Immobilon-P, Millipore). Nonspecific binding was blocked with 5% bovine serum albumin in phosphate-buffered saline, 0.05% Tween.
The antibody/antigen complexes were visualized by an enhanced chemiluminescence detection system according to the manufacturer's instruction (ECL, Amersham Pharmacia Biotech) using anti-mouse Ig or anti-rabbit Ig Abs coupled to horseradish peroxidase as secondary antibodies.
For quantitative Western blot, an anti-mouse Ig coupled to alkaline phosphatase (Dako, Hamburg) and the substrate CDP-star (Roche Molecular Biochemicals) were applied. The quantification was done with the LumiImager (Roche Molecular Biochemicals). Afterward, Coomassie Blue staining of the blotted membranes assessed equal loading of the different lanes.

Measurement of [Ca 2ϩ ] i -[Ca 2ϩ
] i was determined as described previously (6). The T cells were activated via CD3 with the mAb OKT3 and via CD2 with the mAbs 39C1.5 and 6F10.3. Cross-linking was achieved with goat anti-mouse Ig or goat anti-rat Ig, respectively.  (Fig. 1C).

Growth Transformation of ZAP-70-deficient T Cells by H. saimiri-After infection with
Overexpression of Syk in ZAP-70-deficient Transformed T Cells-We analyzed the expression of the ZAP-70-related PTK Syk in primary and transformed T cells. Syk was not overexpressed in the primary T cells from the two patients (Fig. 2). We noted that H. saimiri-transformed T cell lines and uninfected T cell lines from control donors showed a similar expression level of ZAP-70 and of Syk (Fig. 2).
However, when ZAP-70-deficient transformed T cells were infected with H. saimiri, the outgrowing transformed T cells displayed a strong overexpression of Syk (Fig. 2, A and B). This observation was confirmed using a quantitative Western blot (Fig. 2C). It showed that Syk was expressed at least 20-fold more than in other transformed cell lines obtained from control donors, which exhibit low, barely detectable, and variable levels of Syk. The complete lack of CD2-induced [Ca 2ϩ ] i in primary ZAP-70-deficient T cells has been described previously in detail (6).
Here we analyzed the [Ca 2ϩ ] i in H. saimiri-transformed T cells after CD2 activation. Fig. 3 (compare panels A and B) demonstrates that ZAP-70-deficient T cells are rendered responsive to CD2 activation after transformation with H. saimiri.
It is a unique feature of H. saimiri-transformed T cells to become activated via CD2 with one cross-linked mAb directed to the T11.1 epitope (8). Our study shows that ZAP-70-deficient H. saimiri-transformed T cells respond with an increase of [Ca 2ϩ ] i after cross-linking of the T11.1 epitope (Fig. 3F).
H. saimiri Restores Activation-induced ERK Phosphorylation-CD2-and CD3-mediated phosphorylation of ERK1 and ERK2 was defective in the ZAP-70-defective nontransformed T cells (Fig. 4). By contrast, in ZAP-70-deficient H. saimiri-transformed T cells the MAP kinases ERK1 and ERK2 are activated upon CD2 or CD3 activation as reflected by their phosphorylation (Fig. 4). The activation-induced phosphorylation of ERK1 and ERK2 was comparable in cell lines established from the ZAP-70-deficient patients and from control donors.
Defective Phosphorylation of the TCR-Chain in ZAP-70deficient H. saimiri-transformed T Cells-Because H. saimiri restored MAPK activation and [Ca 2ϩ ] i , we analyzed to what extent the CD2-and CD3-dependent signal cascades were restored. CD2 or CD3 triggering of H. saimiri-transformed T cells induced no phosphorylation of (Fig. 5). By contrast, activation of cell lines established from control donors readily induced a phosphorylation of the chain (Fig. 5).
Activation-mediated Phosphorylation of Syk-Because we found that Syk was heavily overexpressed in ZAP-70-deficient H. saimiri-transformed T cells, we studied whether the Syk expressed in the transformed cell lines could be activated. Triggering of CD2 and CD3 induced tyrosine phosphorylation of Syk in H. saimiri-transformed cell lines established from both control donors and ZAP-70-deficient patients (Fig. 5).
Cytokine Production via CD3 and CD2 Activation-Because primary ZAP-70-deficient T cells do not respond to CD2 and CD3 triggering in terms of cytokine production (6), the ZAP- 70-deficient H. saimiri-transformed T cells ZAP-70 -1T and ZAP-70 -2T derived from both patients were analyzed for CD2and CD3-mediated cytokine production. To this end, a mAb recognizing the T11.1 epitope on CD2 or the CD3-specific mAb OKT-3 was added and cross-linked either with Fc receptorexpressing A20 cells (Fig. 6, A and B) or after binding to the microtiter plate (Fig. 6C). As control, a mAb directed to CD58, which is abundantly expressed on H. saimiri-transformed T cells, was applied. The H. saimiri-transformed T cell lines obtained from both of the ZAP-70-defective patients responded to this activation via CD2 or CD3 in a manner similar to the other ZAP-70-expressing transformed T cells with increased production of TNF-␣ (Fig. 6) and interferon-␥ (data not shown). PP2, an inhibitor of Src family kinases, and piceatannol, an inhibitor of Syk/ZAP-70 kinases, blocked the CD2-induced production of IL-2 and TNF-␣ of both the ZAP-70-deficient and control H. saimiri-transformed T cell lines (Fig. 7).
Autocrine Growth Is Mediated via Mutual CD2-mediated Activation-Because ZAP-70-deficient H. saimiri-transformed T cells were responsive to CD2-and CD3-mediated T cell activation, we analyzed whether their autocrine growth was mediated via CD2. To this end we cultured ZAP-70-deficient H. saimiri-transformed T cells and control T cells in the presence of a mAb that blocks the CD2-CD58 interaction. In the presence of this mAb, the spontaneous proliferation of the ZAP-70expressing and ZAP-70-deficient transformed T cells was reduced to a similar extent (Fig. 8). DISCUSSION We demonstrate that the infection by H. saimiri of T cells from two patients with ZAP-70 deficiencies restores T cell activation by CD2 and CD3. H. saimiri does not encode for a gene that has homology to ZAP-70 (13), but it can replace this tyrosine kinase. Different cellular and biochemical events occurring as a consequence of TCR or CD2 activation were analyzed to gain insight into the mechanisms used by the virus to substitute for ZAP-70.
When comparing a series of wild type H. saimiri-transformed T cells with the patient-derived T cell lines, we found that Syk, which belongs to the same family of PTK as ZAP-70, was heavily overexpressed in the ZAP-70-deficient transformed T cells. In contrast, primary T cells from the same patients expressed low levels of Syk that are comparable to the ones observed in mature T cells (14). These low levels of Syk could not substitute for ZAP-70 for CD2-or CD3-triggered activation (6).
In H. saimiri-transformed cell lines, the enormously overexpressed Syk becomes tyrosine-phosphorylated upon CD2 and CD3 activation. This suggests that Syk overexpression in the ZAP-70-deficient lines may be responsible for the CD2-and CD3-induced activation observed in these cell lines. Syk has indeed been shown to be able to substitute to some extent for ZAP-70 in other models. In ZAP-70Ϫ/Ϫ mice, thymocyte development can be restored by Syk expression (15), indicating that ZAP-70 and Syk have overlapping functions. In the P116 Jurkat clone lacking both Syk and ZAP-70 expression, tyrosine phosphorylation of several proteins and NF-AT activation are restored by Syk expression (16). Moreover, it has been shown in T cell blasts obtained from two siblings presenting with ZAP-70 deficiency and overexpressing Syk after expansion that CD3 activation was restored to some extent (17). The results obtained in our present study on ZAP-70-deficient H. saimiritransformed T cells present some similarities to but also has properties distinct from the ones obtained on blasts from ZAP-70-deficient patients.
In this study, we showed that upon transformation with H. saimiri, ZAP-70-deficient T cells became responsive to CD3 and CD2 activation in terms of an increase in [Ca 2ϩ ] i , MAPK activation, cytokine production, and cell proliferation, whereas these responses were impaired in the primary T cells from the ZAP-70-deficient patients. In contrast, in the study published by Noraz et al. (17), CD3-induced MAPK activation and proliferation were not fully restored in the ZAP-70-deficient T cell blasts expressing high levels of Syk. Moreover, the CD3induced increase in [Ca 2ϩ ] i obtained in the ZAP-70 T cell blasts presented unique features compared with those obtained in normal T cell blasts. Altogether, this suggests that the overexpression of Syk, as we observed in this study, can partially but not completely explain the restoration of CD2-and CD3mediated signaling events after transformation with H. saimiri.
Remarkably, as reported in primary T cells from the ZAP-70-deficient patients (6), CD2 or CD3 activation of the transformed cell lines from the patients did not induce any tyrosine phosphorylation of the chain, showing that Syk does not induce exactly the same signaling events as ZAP-70 in T cells.
These data are consistent with another study showing that phosphorylation is defective in the P116 Jurkat cells transfected with Syk (16). It has been proposed that the Src homology-2 domains of ZAP-70 might protect the phosphorylated immune-receptor tyrosine-based activation motifs of the chain from dephosphorylation (4). Because this defect remains in H. saimiri-transformed T cells overexpressing Syk, this argues that Syk cannot replace ZAP-70 for the protection of from dephosphorylation. FIG. 6. CD2-and CD3-mediated activation of cytokine production intact after transformation. The H. saimiri-transformed T cell lines from the first (A) and the second patient (B and C) were activated with the mAb OKT3 directed to CD3 or the mAb 39C1.5 directed to CD2. Cross-linking of these mAbs was achieved with Fc receptor-bearing A20 cells (A and B) or by coating of the microtiter plate with the mAb (C). The mAb to CD58, rat Ig, and medium were used as controls. Supernatants were collected 24 h after activation, and the amount of TNF-␣ was determined by ELISA.
The mechanisms underlying the overexpression of Syk in ZAP-70-deficient transformed cells are unclear. Infection with H. saimiri and subsequent transformation as such are not responsible for the overexpression of Syk, because in none of the transformed T cells from normal donors was Syk expressed at higher levels than in noninfected T cells. One possible explanation for Syk overexpression could be the following. During the process of growth transformation after infection with H. saimiri, there is a selective advantage for cells that can be activated by CD2 because stable growth after infection with H. saimiri is based on CD2-mediated activation. Indeed, we demonstrate that in the ZAP-70-deficient transformed T cells, as in ZAP-70-expressing T cell lines (8), an autocrine growth was mediated via CD2 activation. This suggests that after H. saimiri infection of ZAP-70-deficient T cells there is a selective growth advantage for T cells expressing high levels of Syk. During T cell ontogeny, Syk expression is down-regulated; however, there is a small population of TCR␣␤ CD4 ϩ T cells that expresses a high level of Syk (14). This population may be the one that is transformed by H. saimiri and has a selective advantage to grow.
We observed that a specific inhibitor of the Syk tyrosine kinase family blocked CD2-induced production of TNF-␣ and IL-2 in H. saimiri-transformed T cells from both normal donor and ZAP-70-deficient patients. These results argue that ZAP-70, in the control H. saimiri T cell lines, and Syk, in the ZAP-70-deficient T cell lines, are implicated in T cell activation. Src tyrosine kinases are also involved in both cases, because PP2 inhibited T cell activation of H. saimiri-transformed T cell lines. These results support the notion that this virus utilizes T cell activation pathways to ensure stable growth of the infected T cells.
As previously discussed overexpression of Syk cannot fully explain the restoration of CD2 and CD3 activation observed in the ZAP-70-deficient T cells transformed by H. saimiri. It is possible that viral proteins cooperate with Syk to restore CD2 and CD3 responsiveness. In human transformed T cells, H. saimiri persists episomally, and only two viral genes are transcribed constitutively in human transformed T cells (18). Both corresponding proteins, tyrosine kinase-interacting protein (Tip) and H. saimiri transformation-associated protein of C strains (STP-C), are absolutely required for growth transformation (19) and interact with signaling proteins. Tip binds to Lck and is phosphorylated by Lck (20). In most (21-23) but not all assay systems (24) Tip activated Lck. Tip also activates STAT-1, STAT-3, and NF-AT-dependent transcription (22,25). A model has been proposed in which Lck phosphorylates Tip and the phosphorylated Tip then recruits STATs (26).
The other viral protein essential for transformation, STP-C, binds to Ras, favoring its active GTP-bound state and stimulating MAPK activity (27). STP-C also binds to TRAFs (tumor necrosis factor receptor-associated factor) leading to NF-B activation (28). Oncogenic Ras can replace STP-C for T cell transformation (29). All of these studies indicate that H. saimiri proteins may contribute to modifications of the signal- ) and two control cell lines was analyzed. A rat-derived mAb to CD2 (39C1.5) or rat Ig was added. One to three independent experiments were performed per cell line, and the mean of these experiments is presented. The proliferation in the absence of added Ig was set at 100%, and the proliferation in the presence of the added Ig was calculated.
ing pathways. Normal T cellular activation pathways are not only preserved during transformation but, as we show here, are even reconstituted. H. saimiri reconstituted the CD2-mediated activation pathway, probably because the virus needs this to ensure the growth of the infected T cells. Remarkably, the CD3-mediated activation pathways were also restored, although CD3-mediated activation is not needed for stable growth because antigen-specific T cell clones can be stably transformed independently of the presence of their antigen (11). The reconstitution of the CD3 pathway along with the CD2 pathway probably reflects the fact that CD2 and CD3 signaling pathways share common features as already described (30,31).
The substitution of ZAP-70 by H. saimiri transformation is a special feature of this virus and not a mere consequence of growth transformation, because ZAP-70-deficient T cells that have been transformed by human T cell leukemia virus-1 essentially continue to display the same activation deficits that are characteristic of native ZAP-70-deficient T cells (5,32). Moreover, in contrast to H. saimiri, human T cell leukemia virus-1 down-regulates the expression of ZAP-70 (12).
In conclusion, our study shows that transformation by H. saimiri replaces ZAP-70 for most CD2-and CD3-mediated activation events. This replacement of ZAP-70 was not achieved with a genetically engineered virus but rather with the wild type virus that does not code for a ZAP-70 homologue. To our knowledge this is the first report showing that a wild type virus can restore CD2-and CD3-mediated activation in signaling-deficient human T cells. Because autocrine activation is dependent on CD2 activation, the virus needs to restore this to ensure growing of the latently infected cells. This study shows that thereby CD3 responsiveness is also restored, suggesting that CD3-mediated proliferation shares the signaling pathways with CD2-mediated proliferation.