Diminished TCR signaling in cutaneous T cell lymphoma is ... - Nature

Leukemia (1997) 11, 1338–1346  1997 Stockton Press All rights reserved 0887-6924/97 $12.00

Diminished TCR signaling in cutaneous T cell lymphoma is associated with decreased activities of Zap70, Syk and membrane-associated Csk M-C Fargnoli1,2, RL Edelson1, CL Berger1, S Chimenti2, C Couture3, T Mustelin3 and R Halaban1 Department of Dermatology, Yale University School of Medicine, New Haven, CT, USA; 2University of L’Aquila, L’Aquila, Italy; and Division of Cell Biology, La Jolla Institute for Allergy and Immunology, La Jolla, CA, USA

1 3

Malignant cells of patients with cutaneous T cell lymphoma (CTCL) are of monoclonal origin and of the CD4+/CD45RO+ subset. Since unlike their normal counterparts, triggering of their TCR/CD3 in vitro elicits only a weak mitogenic response, we set out to determine which of the signal transduction molecules initiated by anti-CD3e antibodies are affected in neoplastic cells. The results obtained from analysis of tumor cells from four patients show a general reduction in basal and induced tyrosine phosphorylation of a wide range of signaling proteins. Furthermore, the function of members from distinct families of protein tyrosine kinases was altered in neoplastic cells. The enzymatic activity of the membrane-bound fraction of Csk was suppressed, and its association with other cellular proteins was altered. There was a decline in the amount and activity of Syk, and a slight decrease in the specific activity of Lck kinases. Zap70 tyrosyl phosphorylation was reduced or undetectable and the kinase associated weakly, or not at all, with the TCR z chain. We propose that dampened TCR-triggered responses in CTCL are caused by suppression of an array of effector molecules required for coupling cell surface receptors to early and late signaling events. Keywords: phosphoproteins; protein-tyrosine kinase; receptor– CD3 complex; CD45RO+; protein tyrosine phosphatase

Introduction Cutaneous T cell lymphoma (CTCL, which includes mycosis fungoides and its leukemic variant, Se´zary syndrome) is a neoplasm of clonal T cells that infiltrates the skin. The cells display a phenotype of CD4+ helper/inducer, CD45RO+ T cells characteristic of ‘memory’ T lymphocytes.1–5 A high percentage of the malignant cells constitutively express BE2, a late activation marker of 78 kDa, that appears on cloned normal T cells only after stimulation with specific antigen, and on freshly isolated normal T cells in response to mitogens,6,7 suggesting that CTCL cells have been activated by their surrounding dermal environment. Since CTCL lymphocytes can proliferate in the cutaneous compartment but are extremely difficult to cultivate in vitro, their neoplastic growth appears to reflect in vivo stimulation, possibly through their TCR, or antigen-independent pathways or cytokines. IL-7 might be one of the required growth factors for CTCL cells in the skin since transgenic mice constitutively producing IL-7 developed extensive dermal infiltration followed by dermotropic lymphoma.8,9 However, although IL-7 is produced by keratinocytes, it is a weak in vitro CTCL mitogen,10 and purified populations of CTCL cells were sustained in culture for only a short period (such as a week) by mixtures of cytokines that include IL-7 and IL-2.11,12 The preferential expression of CD45RO and the clonal TCR rearrangement suggest that CTCL clones have encountered a Correspondence: R Halaban, Department of Dermatology, Yale University School of Medicine, PO Box 20859, New Haven, CT 065208059, USA Received 29 November 1996; accepted 18 April 1997

specific antigen and have undergone antigen-driven activation, even though the nature of the antigen(s) remains to be elucidated.13 Antigen recognition by the TCR can lead to activation, proliferation, differentiation or cell death, depending on the maturation stage of the T cells.14 It is possible that CTCL cells develop a protective mechanism against TCR-mediated apoptotic signals, since triggering of TCR/CD3 in CTCL cells in vitro produces only a weak proliferative response (see for example Ref. 15). Because none of the previous studies determined the cause of this stunted response, we investigated whether immediate TCR signaling is impaired in CTCL cells, and if so, we set out to identify the molecular basis for this deficiency. TCR engagement rapidly induces tyrosine phosphorylation of several cellular substrates in normal T cells (reviewed in Refs 16–19). At least five protein tyrosine kinases (PTKs) belonging to three different families play active roles in TCR/CD3-induced early signal transduction. They include p56lck (Lck) and p59fyn (Fyn) of the Src family, p70zap (Zap70) and p72syk (Syk) of the Syk family, and p50csk (Csk) of the Csk family (reviewed in Refs 17, 18, 20 and 21). The tyrosine kinase activities of Lck and Fyn increase moderately within seconds after TCR cross-linking,22 and their activity is subsequently down-regulated perhaps by phosphorylation of their C-terminal negative regulatory tyrosine residues by Csk (reviewed in Refs 18 and 21). This Csk repressive phosphorylation can be reversed by the protein tyrosine phosphatase (PTPase) CD45 (reviewed in Ref 18). In addition, TCR engagement results in Zap70 activation after its binding to the tyrosylphosphorylated immunoreceptor tyrosine-based activation motifs (ITAMs) of the z chains and the CD3 complex23,24 (for review see Ref. 25) and in rapid tyrosine phosphorylation and activation of Syk.26 The signaling initiated by TCR activation is transmitted from the cell membrane through the cytoplasm to the nucleus and induces quiescent T cells to undergo cell division and maturation, processes critical for clonal expansion and initiation of the immune responses. We selected for investigation several immediate key reactions essential for T cell receptor-mediated signal transduction, proliferation and effector functions. Our results, based on analysis of a limited sample of CTCL cells (from four patients), show nevertheless, common characteristics, ie deficiencies in critical immediate functions associated with TCR activation. These include molecules participating in events proximal and distal to the TCR such as: (1) reduced levels of basal and stimulated tyrosyl-phosphorylated proteins belonging to a wide range of signaling molecules; (2) suppressed specific activity of membrane-bound Csk and Lck; (3) failure to activate membrane-bound PTPase; (4) failure to activate Zap70 and to induce its association with the TCRz chain; and (5) reduced expression and activity of Syk. Limited comparative analysis with the CD4+/CD45RO+ subpopulation indicated that protein levels of several kinases in these cells were intermediate between normal CD4+ T cells and CTCL

Zap70, Syk and Csk in CTCL TCR signaling M-C Fargnoli et al

1339

cells. Our results imply that the reduced responsiveness of CTCL cells to TCR triggering is the result of impairment in a complex array of normal T cell functions. This may be relevant to the proliferation or survival of these neoplastic cells. Material and methods

T cells and TCR stimulation PBMCs were purified from six healthy adult volunteers and four selected patients with erythrodermic CTCL (designated BU, HA, RU and WO). These patients were chosen because they had marked clonal expansion of neoplastic cells representing 80–95% of the mononuclear cells (Table 1). Informed consent was obtained for all specimens in accordance with the policy of Yale University Human Investigation Review Board. The CTCL cells were identified by their reactivity with mouse mAbs recognizing the variable region of the TCR a or b chains (T-Cell Diagnostic, Cambridge, MA, USA). Monthly phenotyping by flow cytometry indicated that the CTCL cells maintained expression of both the CD3 molecule and the T cell receptor (Table 1). As is characteristic in this group of patients, the marked expansion of the clonal malignant T cell population resulted in a reduced representation of normal T and B cells. The B cell population, as identified by the presence of CD19, ranged from 0–1.7%, compared to 10 ± 5% in normal healthy individuals (Table 1). The CD8+ subpopulation of cells was also reduced but the levels of CD3e and the a and b chains of the TCR were similar in the different populations of cells used in our studies, as determined by FACS analysis (Table 1), except for patient HA who had two different populations of tumor cells with high and low CD3 expression (Figure 1). The CTCL cells from this patient also had comparatively low levels of CD45RO+ (Table 1 and Figure 2f). Loss of membrane antigens in particular CD3 on the malignant T cells has been seen in some CTCL patients and may relate to subclone formation and more aggressive disease. However, this particular patient responded Table 1 cells

Immunophenotype analysis of patients’ peripheral blood

Marker/Patient

% Positive Normal

T cells CD3 CD4 CD8 CD4/CD8 ratio B cells CD19 Others BE2 CD45RO CLA Monocyte population MO-1 T cell receptor VB5

RU

72 (±7) 98.2 45 (±10) 97.0 28 (±8) 2.0 0.8–3.5 48.5/1

BU

HA

WO

97.7 97.6 0 97.6/0

95.8 96.2 0 96.2/0

95.0 89.0 5.0 17.8/1

10 ± 5

0.6

1.6

1.7

0

15–40 8–23

0 98.0 1.1

36.1 79.9 0.3

36.9 20.5 0.4

15.0 88.0 2.0

50.9

24.8

80.3

0

94.8

96.4

91.8

87.0

1–5

Values for normal controls express (standard deviation).

range or mean ± s.d.

Figure 1 CTCL cells display CD3 expression similar to that of normal lymphocytes. Indirect immunofluorescence profiles generated by flow cytometry of the staining of peripheral blood lymphocytes with aCD3 monoclonal antibody. The light line is the CD3 positive population. The heavy line is the negative control demonstrated by the binding of murine ascites fluid of the same isotype (IgG1) as the CD3 monoclonal antibody but of irrelevant specificity. There were 3500– 5000 cells in each sample. See Table 1 for the expression of other specific markers.

well to therapy and at present has decreased numbers of circulating malignant cells. Therefore, it is not possible to ascribe clinical significance to the relative reduction in CD3 expression found on one population of cells in this patient. CTCL cells from BU and HA were isolated at the time of diagnosis (before treatment) as well as later when the patients were undergoing photopheresis. 27 The other two patients (RU and WO) were receiving photopheresis throughout the study period. CTCL cells were collected from patients under treatment 1 month after a cycle of photopheresis and before the beginning of the next one. Although we encountered some variability in the expression of specific proteins in cells from the same patient between early and late collections (as discussed below), the general cellular phenotype remained the same. The normal and neoplastic cells isolated from the buffy coat by Ficoll–Histopaque 1077 density gradient centrifugation were washed twice in RPMI 1640 (GIBCO BRL, Grand Island, NY, USA) supplemented with antibiotics and 20% heat-inactivated FCS (GIBCO BRL). Only freshly isolated normal T cells were used in the expreriments. In contrast, the CTCL cells, unless otherwise indicated, were immediately frozen in liquid nitrogen until use. They retained 90% viability upon thawing as determined by trypan blue exclusion test, and remained viable for at least a week in the presence of a mixture of cytokines that included IL-2 and IL-7. However, we were unable to maintain these cells in culture for longer than 10 days using various combinations of cytokines, antibodies or other stimulants described in the literature (see for example Refs 11, 12). Freshly isolated and frozen cells were incubated in RPMI/20% FCS medium overnight at 37°C before any further manipulations. Preliminary studies using freshly isolated and frozen cells from patients RU and BU indicated that storage in liquid nitrogen did not affect the cellular response (as described below). Normal CD4+ and CD4+/CD45RO+ T cells from healthy donors were purified by negative selection as follows. Density gradient purified PBMCs were depleted of CD8+ T and B cells by incubation with magnetic beads precoated with mAb specific for the CD8 and CD19 Ags (Dynal, Oslo, Norway), on a rotator for 1 h at 4°C. After removal of the magnetic beads,


1340

the non-adhering mononuclear cells in the supernatants were washed twice in RPMI 1640/20% FCS (heat inactivated FCS, GIBCO BRL), and incubated overnight at 37°C. The CD4+/CD45RO+ subpopulation was enriched by removal of the CD45RA+ cells from the CD4+ population using antiCD45RA mAb28 (anti-2H4; Coulter, Hialeah, FL, USA) followed by immunomagnetic depletion with sheep anti-mouse IgG1 (Fc) (Dynal). The percentage of CD4+ and CD4+/CD45RO+ T cells in the selected populations was determined by FACS analysis employing anti-CD4 and antiCD45RO mAb (UCHL1, Dako, Carpinteria, CA, USA). Jurkat T leukemia cells were grown in RPMI-1640 supplemented with 5% heat inactivated FCS, L-glutamine and antibiotics. After overnight incubation in RPMI 1640/20% FCS, neoplastic T cells, normal CD4+ T cells and CD4+/CD45RO+ T cells were each washed twice with RPMI 1640 and resuspended at a concentration of 10 × 106 cells/ml. Half of the cells were stimulated with a mAb against the CD3e chain of the TCR/CD3 complex (aCD3e, 20 mg/ml) at 37°C for 2 min. The aCD3e antibody (CD3 (IgG1)) used was from Coulter (Miami, FL, USA), or was purified by us from the conditioned medium of the hybridoma producing cells (OKT3, ATCC CRL 8001). The normal and malignant cells were then processed further as described below.

Cell fractionation

Figure 2 CTCL cells display reduced tyrosyl-phosphorylated proteins and an altered pattern of kinase expression. (a, b, c and e) Antiphosphotyrosine (aPY) Western blots of whole cell lysates prepared from normal CD4+ T cells (NTC), CTCL cells (from patients HA, BU or RU, as indicated), or CD4+/CD45RO+ selected normal T cells (CD45RO). Cells were harvested without any further stimulation (−), or after stimulation with aCD3e mAb (+) for 2 min at 37°C. In (a), freshly prepared CTCL cells were used in the case of patients BU and RU, and frozen cells of patient HA. The numbers on the top of (b) indicate weeks of storage of CTCL cells in liquid nitrogen prior to stimulation. Normal T cells freshly isolated from four different healthy donors are represented in (a–f). Spear-headed arrow points at the complex of phosphorylated proteins with M r 21 to 26 kDa that is likely to represent differentially phosphorylated forms of the z and CD3 chains and arrowhead at the approximate position of Zap 70 as determined by immunoblotting with the respective antibody (not shown). (d) Successive probing of the membrane presented in (c) (with repeated stripping except when indicated) with antibodies against CD45, PLCg, STAT and (without stripping) Vav and MAPK; arrows indicate the position of the signaling proteins as listed. (e) Normal CD4+T (CD4+) and CD4+/CD45RO+-selected (CD45RO) cells were purified from the same donor and proteins blotted with aPY. (f) Successive probing with antibodies against Csk, Fyn, Lck and CD45; numbers under each blot indicate arbitrary units of scanned bands. The panel designated %CD45RO+ indicates the percentage of cells expressing CD45RO of the same population used for Western blotting as measured by FACS. Equal protein loading in all lanes was verified by staining the polyacrylamide gels after electrotransfer with Coomassie brilliant blue. The equal levels of signaling proteins in (d) is also an indication for equal protein loading. Positions of prestained high molecular size markers (Bio-Rad Laboratories) are given on the left in kDa according to the manufacturer’s size determinations.

Cells were suspended in 1 ml hypotonic lysis buffer (10 mM Tris-HCl pH 7.2, 10 mM NaCl, 10 mM b-glycerolphosphate, 2 mM MgCl2, 10 mM KCl, 1 mM EDTA, 1 mM EGTA, 1 mM Na3VO4, 0.4 mM PMSF, 10 mg/ml leupeptin, 10 mg/ml aprotinin) on ice for 30 min. The suspensions were then slightly sonicated for 2 s on ice to ensure disruption (monitored by microscopic observations). Nuclei were removed by low-speed centrifugation and the supernatants were subjected to ultracentrifugation at 100 000 g for 1 h at 4°C. The pellets (particulate fractions) were resuspended in 100 ml hypotonic lysis buffer supplemented with 1% NP-40, sonicated on ice for 2 s at the lowest output and spun at 10 000 r.p.m. for 10 min to remove insoluble material. The high-speed supernatants (cytosolic fractions) were concentrated 10-fold using Centricon 10 (Amicon, Beverly, MA, USA) and supplemented with NP-40 to a final 1% concentration. Protein concentration was determined with the Bio-Rad protein assay reagent (Bio-Rad Laboratories, Hercules, CA, USA).

Western blotting, immunoprecipitation and antibodies For Western blotting of whole cell lysates, the stimulated and unstimulated cells (3 × 106 cells/condition) were centrifuged at 15 000 r.p.m. for 2 min at 4°C, and resuspended directly into boiling sample buffer, kept for an additional 5 min at 100°C, sonicated to disrupt DNA, and centrifuged. Supernatants were subjected to SDS-PAGE and Western blotting using the NOVEX gel apparatus, 10% polyacrylamide gels, PVDF (polyvinylidene difluoride) protein-sequencing membranes (Bio-Rad Laboratories) following the manufacturer’s instructions (NOVEX, San Diego, CA, USA). Antigen-antibody complexes were detected by enhanced chemiluminescence (ECL; Amersham, Arlington Heights, IL, USA), according to the manufacturer’s instructions. For successive probing, the membranes were stripped of the preceding antibodies with a solution containing 7 M guanidine hydrochloride, 50 mM glycine


(pH 10.8), 50 mM EDTA, 100 mM KCl, and 20 mM 2-ME for 10 min at room temperature on a shaking platform. The membranes were then rinsed, non-specific binding sites blocked with 5% BSA for 1 h, and probed with the next antibody. The efficiency of stripping was confirmed by the lack of signal after blotting the membranes with secondary antibody only. For immunoprecipitation, stimulated and unstimulated cells (1 × 108 cells/assay) were lysed in buffer containing 1% NP 40, 10 mM Tris-HCl pH 7.2, 10 mM NaCl, 10 mM b-glycerolphosphate, 2 mM MgCl2, 10 mM KCl, 1 mM EDTA, 1 mM EGTA, 1 mM Na3VO4, 0.4 mM PMSF, 10 mg/ml leupeptin, and 10 mg/ml aprotinin (lysis buffer) and kept on ice for 30 min. Lysates were slightly sonicated and centrifuged at 15 000 r.p.m. for 20 min at 4°C, to remove insoluble material. Lysates (200 mg protein) were incubated with the indicated antibodies for 1 h and then for an additional hour with protein A/G PLUS agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA, USA) on ice, on a shaking platform. To detect Zap70 and its associated proteins, the immobilized anti-Zap70 precipitates (rabbit polyclonal antibodies26) were washed extensively with high salt lysis buffer (20 mM Tris-HCl, pH 7.5, 650 mM NaCl, 5 mM EDTA, 1% NP-40 containing the protease inhibitors), and the eluted proteins were fractionated on a 10% SDSPAGE gel, transferred to nitrocellulose, and immunoblotted as described below. To affinity purify Csk binding proteins, particulate and cytosolic fractions (200 mg proteins each) were incubated with 10 mg of GST-SH2/SH3-Csk expressed as a bacterial fusion protein29 in 1 ml of lysis buffer for 2 h on ice, followed by incubation with preswollen glutathione beads (50 ml) (Sigma, St Louis, MO, USA) for 1 h on ice, on a shaking platform. The beads were then washed three times in lysis buffer and proteins were eluted with 50 ml hot sample buffer and subjected to Western blotting as described above. The following antibodies were used: 4G10 mAb against phosphotyrosine (aPY; Upstate Biotechnology, (UBI), Lake Placid, NY, USA); anti-Fyn mAb (fyn(15) Santa Cruz Biotechnology), or rabbit polyclonal antibodies (FYN3; Santa Cruz), or rabbit anti-p59fyn (UBI); anti-Lck mouse mAb (3A5; Santa Cruz) or rabbit polyclonal antibodies (No. 2102; Santa Cruz); anti-Csk rabbit antiserum (aCSKC),30 or mAb (Transduction Laboratories, Lexington, KY, USA), or rabbit polyclonal antibodies (Santa Cruz); anti-CD45 mAb (Biosource International, Camarillo, CA, USA), or the mAb GAP 8.3 made from the supernatant medium of the hybridoma clone (a gift from Dr D Rothstein, Yale University, New Haven, CT, USA), recognizing all the isoforms of the leukocyte common antigen; anti-MAPK rabbit polyclonal Ab 691;31 anti-PLCg mAb (Transduction Laboratories); anti-Vav mAb (UBI); anti-Syk rabbit antiserum26 or affinity-purified rabbit polyclonal Abs (Santa Cruz); anti-Zap rabbit polyclonal Abs26 or mAb raised against GST-fusion protein corresponding to the two tandem SH2 domains of human Zap70 (UBI); antiGAP rabbit antiserum (UBI); and anti-FAK mAb (Transduction Laboratories). The intensities of the bands were determined by scanning the ECL-X-ray films with the Computing Densitometer.

Immune complex kinase and phosphatase assays All kinase assays were performed on immunoprecipitates in azide-free conditions employing the synthetic peptide Raytide (Oncogene Science, Manhasset, NY, USA) as a substrate since preliminary experiments with poly Glu-Tyr 1:1 and rabbit

muscle enolase (both from Sigma) showed that it was the most efficient target in our system. The A/G PLUS-agarose-bound antibody/antigen complexes were washed twice in lysis buffer, once in kinase buffer (20 mM Hepes pH 7.6, 20 mM MgCl2, 20 mM MnCl2, 20 mM b-glycerophosphate, 20 mM pnitrophenylphosphate, 0.1 mM Na3V04) and then incubated with 10 mg Raytide in kinase buffer supplemented with 5 mM ATP and 5 mCi [g-32P]-ATP (3000 Ci/mmol, NEN Research Products, Boston, MA, USA), for 10 min at 30°C. Preliminary experiments indicated that the kinase reaction was linear up to 20 min. The assays were terminated by spinning the reaction mixtures for 30 s and blotting 10 ml of the supernatants on to 1 × 2 cm phosphocellulose strips (Whatman P81; Maidstone, UK). The strips were then air dried, rinsed four times in 0.5% phosphoric acid, and the radioactivity measured in a liquid scintillation counter. Immunoprecipitation with rabbit IgG was included for each cell type (negative control). The phosphotransferase values of the negative controls were substracted from the experimental values for each condition. PTPase activity was assayed by measuring the release of 32P from the tyrosine-phosphorylated Raytide. Raytide was phosphorylated in vitro by v-abl tyrosine kinase (Oncogene Science) and [g-32P]-ATP in a kinase reaction mixture as described.32 Cells were preincubated with 37 mM phenylarsine oxide (PhAsO; Sigma) for 8 min at 37°C, followed by stimulation of half of the population with aCD3e (2 min). For membrane preparation, cells were lysed in hypotonic buffer with 0.1% 2-ME and processed as above (cell fractionation). The particulate fraction was resuspended in phosphatase buffer (20 mM Hepes pH 7.4, 150 mM KCl, 340 mM Sucrose, 0.4 mM PMSF, 10 mg/ml leupeptin and 10 mg/ml aprotinin). For immune complex phosphatase assays, cells were dissolved in lysis buffer supplemented with 0.1% 2-ME, incubated with aCD45 mAb GAP 8.3, antigen–antibody complexes bound to A/G PLUS-agarose beads washed twice in lysis buffer without 2-ME and once in phosphatase buffer. Membrane fractions (10 mg) or A/G PLUS-agarose beads bound CD45 were incubated with the radiolabeled Raytide (°20 000 c.p.m./assay) for 2 min at 30°C and the reactions were terminated by adding 750 ml of 10% activated charcoal in 0.9 M HCl, 90 mM Na4P2O7, 2 mM NaH2PO4. Aliquots of the supernatant were counted in a scintillation counter. Results

CTCL cells display reduced levels of tyrosylphosphorylated proteins and altered kinase expression The earliest response to TCR cross-linking is an increase in tyrosine phosphorylation of multiple proteins, several of which have been identified. Therefore, we first analyzed the pattern of tyrosine phosphorylation on signaling proteins to evaluate responsiveness of CTCL cells to TCR stimulation. In order to examine three to four different tumor samples in parallel, we had to use frozen material for most of our experiments since patients with high tumor load were only available at monthly intervals and we were unable to maintain the cells in culture for more than a week. We therefore determined first whether storage in liquid nitrogen affected cellular response. Western blotting with aPY antibody showed that basal as well as induced levels of tyrosyl-phosphorylated proteins were markedly reduced in CTCL cells compared with normal T cells (Figure 2a, b and c). In some cases, as in Figure 2a, the tumor cells did not respond at all to aCD3 stimulation whether they

1341


1342

were used immediately after isolation or after a freezing period (Figure 2a, compare responses of stored cells from patient HA to freshly used cells from patients BU and RU). The degree of unresponsiveness might be related to disease state or to the particular treatment the patient received prior to isolation. In addition, the limited responsiveness displayed by CTCL cells from patient BU was not affected by storage in liquid nitrogen for up to 26 weeks (Figure 2b and c). Furthermore, the reduced responsiveness was not due to diminished levels of CD3e, as indicated before (Figure 1), or low expression of several known signaling proteins as shown by the presence of equal levels of CD45, PLCg, Vav, STAT and MAPK (Figure 2d as indicated). It thus appears that CTCL suffer from severe impairments in immediate signal transduction which affected a wide spectrum of molecules, including, in addition to those mentioned above, reduced phosphorylation of Zap, CD3 and z chains of the TCR complex (Figure 2a, b as indicated by the arrowhead and speared arrow, respectively). Since CTCL cells are of memory CD4+/CD45RO+ subtype, we went on to determine if the stunted responsiveness is a general characteristic of this population of memory T cells. As shown in Figure 2e, aCD3 stimulation triggered a similar increase in the intensity of tyrosyl-phosphorylated proteins in CD4+ and CD4+/CD45RO+ selected subtype of normal T cells purified from the same donor. This is in spite of the fact that the population of selected memory T cells had profiles of CD45 expression, as well as expression of several critical kinases known to mediate T cell receptor engagement similar to that of CTCL cells isolated from two patients (Figure 2f as marked). Probing of membranes with antibodies to three kinases showed, that with some variation in the case of patient HA, Csk was consistently more abundant whereas Fyn and Lck were expressed at variable levels (see also below), in the selected CD4+/CD45RO+ memory T and CTCL cells compared to unselected CD4+ T cells.

previous report34 aCD3e caused redistribution of Csk as shown by a loss in kinase activity in the cytosolic fraction and a gain in the membranous fraction (Figure 2b,c, normal T cells and the CTCL of patient BU). The overall activity of Csk in whole lysates of the neoplastic cells was greatly suppressed in comparison to that of the cytosolic fractions and resembled that of the membrane-associated fractions (Figure 3, compare a to b), suggesting inhibition of Csk by a tightly binding factor released by the detergent from the membrane. The normal Csk activity in cytosols of CTCL cells from patient BU is a good indication that the use of frozen cells did not compromise Csk kinase activity in our studies (Figure 3c, compare BU to NTC/CD4+).

Membrane-bound Csk activity is markedly reduced in CTCL cells

Differential association of target proteins to the SH2/SH3 domain of Csk in vitro

The stunted responsiveness of CTCL cells presented in Figure 2 could be due to diminished PTK activity and/or an increase in protein tyrosine phosphatase (PTPase) activity. The kinase activity of Csk was tested first since this PTK occupies a central role in signal transduction. It may down-regulate the T cell receptor (TCR) signaling by phosphorylating the (inhibitory) C-terminal tyrosine of Lck and Fyn and potentially of the CD45 PTPase. The results show that Csk basal and induced kinase activity was three-to four-fold lower in CTCL cells than normal T cells (Figure 3a). Because Csk is found in both cytoplasm and membrane,30,33 the immune complex assay was repeated using soluble and particulate fractions. This set of experiments showed that Csk activity was drastically diminished in the particulate fraction of CTCL cells, with no or very small differences in the cytosol when compared to normal T cells (Figure 3b,c). This suppressed activity could not be attributed to low levels of Csk protein because, except for patient HA in the experiment described here, Csk was more abundant in CTCL compared to normal CD4+ T cells (Figure 3 blots in b and c, see also Figure 2f). The levels of Csk varied in HA cells, probably due to changes in the cell population (as seen in Figure 1). In addition, reduced kinase activity could not be caused by a Csk-associated PTPase, since none was detected in immune complex PTPase assays performed on anti-Csk precipitates (data not shown). In agreement with a

Csk may be a part of a multiprotein complex that could regulate its activity since the SH2 (Src homology 2) domain of Csk associates with several tyrosyl-phosphorylated proteins in Jurkat and normal T cells independently of TCR/CD3 activation.29,35–38 We therefore looked for differences in GSTSH2/SH3-Csk-associated proteins in cell fractions from normal and neoplastic cells. The aPY Western blots of the affinity purified proteins revealed qualitative differences between the two cell types. Normal T cells preferentially displayed phosphorylated SH3/SH2-Csk binding proteins of 140 kDa and 50 kDa in the membrane fraction, and 50 kDa and several other higher molecular weight species in the cytosolic fraction (Figure 4a,b, arrows). However, a major protein of about 120 kDa was apparent only in the membranes prepared from CTCL cells but not from normal T cells (Figure 4a, solid arrowhead). Reprobing the membrane with antibodies against p120FAK or p120GAP, tyrosyl-phosphorylated signaling proteins failed to detect any association (data not shown). The 60 kDa, and 110 kDa Csk-SH2-associated proteins described before29,36 were present at almost equal levels in both normal and malignant cells (Figure 4a,b, marked by heavy arrows), whereas the 90 kDa was less abundant in the malignant cells (Figure 4a, marked by empty arrow). These results reinforced the notion that Csk activity could be regulated by a membrane-resident protein.

Figure 3 Differences in Csk activity between normal and neoplastic cells. Histograms represent in vitro Csk phosphotransferase activity in immune complex kinase assays precipitated from whole cell lysates (a) , or cell fractions (b and c). Cell type and treatments are as described for Figure 2, and as marked in the Figure. The level of Csk in each fraction prior to immunoprecipitation is shown by the Csk Western blots in (b) and (c). These blots indicate that, except for patient HA, Csk levels were higher in the particulate fractions of the neoplastic compared to normal cells but yet exhibited lower total kinase activity, implicating reduced specific activity compared to normal control T cells.


1343

Figure 4 Differences in Csk binding proteins between normal and neoplastic cells. aPY Western blots of affinity-purified proteins eluted from control GST, or GST-SH2/SH3-Csk fusion protein bound to glutathione beads. Precipitations were done with membrane (a) or cytosolic (b) fractions. Small arrows point at GST-SH2/SH3 Csk binding proteins in normal T cells, not detected in CTCL cells; arrowhead in (a) (right-hand side) indicates the °120 kDa aPY reactive GSTSH2/SH3 Csk binding protein present in neoplastic but not in normal T cells. Heavy and empty arrows indicate the position of proteins reported to bind Csk-SH2 in other cellular systems.

Kinase activity of Csk-substrates As mentioned above, Csk substrates include the C-terminal inhibitory tyrosine residues of the Src-family kinases, and possibly Tyr-1193 of CD45 phosphatase (see reviews, Refs 17–20). A decrease in Csk activity might be expected to result in increased specific kinase activity of Lck, Fyn and possibly, a decrease in CD45 PTPase activity in CTCL cells compared to normal CD4+ T lymphocytes. However, repeated immune complex kinase assays showed, that the specific activity of each of the Src-family kinases in the malignant cells was not increased, but rather, in some cases, was reduced by 50– 80% (Figure 5a,b). Likewise, very little impact of suppressed membrane-bound Csk was found on PTPase activity. Basal PTPase activity in the intact membrane fraction or in CD45 immune precipitates derived from CTCL cells was similar (or slightly increased) compared to that of normal T lymphocytes (Figure 5c and d). As expected, there was a two-fold increase in the membranebound phosphatase activity in normal T cells but little or no enhancement in the neoplastic T cells after aCD3e stimulation (Figure 5c).

Dysfunctional Syk family kinases The status of Syk and Zap family of kinases was also examined, since reduced phosphorylation of Zap was suggested in our aPY immunoblotting experiments (Figure 2). Differential activation of Zap70 with aCD3e was clearly demonstrated in the aZap70 immunoprecipitation/immunoblotting experiments. Induction of Zap70 tyrosyl-phosphorylation and association with the phosphorylated TCR z chains was detected in normal and Jurkat T cells (Figure 6a,b, left panels, speared shaped and empty arrows, respectively). In contrast, the weakly tyrosyl-phosphorylated Zap70 precipitated from the neoplastic cells of one patient (HA) was not associated with the phosphorylated z chain (Figure 6a, left panel, HA). In cells from another patient (RU), Zap70 did not become tyrosyl-phosphorylated, and only a faint band of z chain was visible in one experiment (Figure 6a, left panel, RU), but not in the second one (Figure 6b, left panel, RU). Probing the same

Figure 5 Src-family kinases and PTPase activity are not markedly altered in CTCL cells. (a and b) Histogram shows phosphotransferase activity in Fyn and Lck immune complex kinase assays. Blot on the bottom shows the levels of the respective kinases in the whole cell lysates used in these experiments. (c) PTPase activity in membrane fractions. (d) Immune complex CD45 PTPase activity. Unstimulated (−) and aCD3e stimulated (+) normal CD4+ (NTC) and neoplastic cells from patients (as indicated) were used in these experiments.

membranes with aZap70 mAb showed that, in one experiment equal amounts of Zap70 was precipitated from the normal and neoplastic HA, and in the second experiment, from neoplastic RU and Jurkat T cells (Figure 6, right panels, aZap70). In the experiment presented in Figure 6a, for unknown reasons very little Zap70 was precipitated from RU CTCL cell lysate (Figure 6a, right panel). Furthermore, additional experiments also showed that Zap70 protein is expressed at normal levels in the neoplastic cells of HA, RU and BU, but does not become tyrosyl-phosphorylated in response to stimulation with aCD3e (Figure 7, compare pattern of tyrosyl phosphorylation with position of Zap70 protein band in cell lysates). These results indicate that activation of Zap70 is deficient in CTCL cells. Since Syk can compensate for lack of Zap70,20 we determined its expression and activity in the different cell types. Syk PTK positively correlated with its protein levels and was directly proportional to the levels of tyrosyl-phosphorylated proteins (Figure 7a,b,c). Reprobing the same membrane with antibodies against Csk Fyn, and Lck showed, as before, that these enzymes were expressed at higher levels in CTCL cells, with no correlation to the tyrosyl-phosphorylation pattern (data not shown). Thus, Syk could not substitute for Zap70 in CTCL cells. Discussion Our studies examined, for the first time, signals transduced through the stimulation of the TCR in highly purified clonal populations of CTCL cells isolated directly from the patients


1344

Figure 6 Triggering of Zap70 is aborted in CTCL cells. Proteins of normal T cells (CD4+), CTCL (HA and RU), or Jurkat T cells were precipitated (ip) with aZap70 rabbit polyclonal antibodies (a and b, Zap70), or with normal rabbit serum (b, NRS) and subjected to Western blotting (ib) with mAbs to phosphotyrosine (PY), and, after stripping, to Zap70. Protein phosphorylation and Zap70 expression in immunoprecipitates was compared to that of whole cell lysates (WCL). Spear-headed arrows indicate Zap70 and open arrows point at the co-precipitated phosphorylated z chains. The °40 kDa aPY reactive band in the stimulated sample of normal T cells in (a) is not yet identified. The °55 kDa band in (b) is OKT3 reacting with the secondary antibodies in the ECL detection system. Stimulation of cells was as described before.

Figure 7 Aberrant expression and/or activity of Syk family of kinases. Syk expression and activity positively correlate with levels of tyrosyl phosphorylated proteins. (a) aPY Western blot of whole cell lysates as described above. Spear-headed arrows indicate the position of phosphorylated forms of the z and CD3 chains and an arrowhead that of Syk. (b) Expression of Syk. Western blot of the stripped membrane in (a) with aSyk. The histogram shows arbitrary units of scanned bands of the Syk immunoblot (presented on the bottom). (c) Phosphotransferase activity in Syk immune complex kinase assays were performed on the same lysates as presented in (a).

without any adaptation to culture conditions. The information derived from these studies is particularly valuable since these cells cannot be established in culture as immortal cell lines. CTCL cells, from four erythrodermic patients possessing high percentages of circulating neoplastic T cells, did not initiate, or did so to a much lesser degree, the activation cascade characteristic to aCD3 stimulation in normal CD4+ T cells. The stunted response was manifested as reduced levels of tyrosyl-phosphorylated proteins and was not due to changes in the level of expression of components of the TCR. Likewise, this phenotype could not be the result of selective expression of CD45RO in these cells, since in our hands the CD4/CD45RO+ subpopulation of normal T cells responded to TCR ligation in a manner similar to that of unselected CD4+ T cells, and extensive studies by others failed to detect consistent differences in TCR signaling between cells expressing different isoforms of CD45 (see for example review, Refs 39 and 40). In fact, memory T cells proliferate at a faster rate than naive cells when stimulated with anti-CD3 monoclonal antibody (see for example Ref. 41). The block in CTCL cells involved key members of at least two families of kinases, the membrane-bound Csk, Zap70 and Syk. Csk, Zap70 and Syk occupy central roles in the activation of immediate and proximal signals. Suppression of membranebound Csk may explain the failure to activate membranebound PTPase (presumably CD45, the major phosphatase in membranes of T cells, see review, Ref 42), since Csk controls not only the Src-like kinases but possibly also CD45.43 On the other hand, reduction in Zap70 function in CTCL can, by itself, be the cause for the unresponsiveness of CTCL cells. Peripheral CD4+ T cells in patients carrying inactivating mutations in Zap70 are incapable of transducing normal TCRmediated signals manifested as severe combined immunodeficiency (reviewed in Refs 20, 44). Furthermore, altered peptide ligands (APLs) that induce anergy, mediate their signals through an alternative TCR pathway involving the altered phosphorylation pattern of TCR z and failure to activate and recruit Zap70.45,46 The TCR signaling in CTCL cells could not be rescued by Syk, whose expression and activity was also reduced. Membrane-bound Csk was clearly regulated at the activity level, since the protein was, on average, °5-fold more abundant in CTCL cells compared to normal T cells. Csk localization and catalytic activity is regulated by TCR cross-linking, but the exact mechanism of its activation is still unknown since post-translational modifications, such as phosphorylation on Tyr and/or Ser/Thr have not been found.29 The marked reduction in overall Csk activity in detergent-lysed CTCL cells but not in cytosolic fractions, suggests that the membrane fraction contains a Csk-binding inhibitor. Indeed, a °120 kDa tyrosine-phosphorylated protein associated with the GST-SH2/SH3-Csk from the membrane fraction of CTCL cells but not from normal CD4+ T cells. We ruled out the possibility that this protein is GAP or the focal adhesion kinase FAK. Association of the SH2 domain of Csk with a selective set of phosphotyrosine containing substrates, including a 120 kDa protein was recently demonstrated in mouse T cell hybridoma cells expressing constitutively active Lck.38 Further identification of this protein and studies on its effect on Csk activity will determine if indeed it is the putative Csk inhibitor. Reduced activity of membrane-bound Csk did not cause a steady-state activation of the Src-like kinases Lck and Fyn in CTCL cells, as would be expected from the known physiological role of Csk. This is surprising since point mutations in the inhibitory C-terminal tyrosine residues phosphorylated by Csk


induced hyper-reactivity and massive tyrosine phosphorylation responses to TCR/CD3 stimulation.47 Furthermore, cell lines derived from Csk knock-out mouse embryos displayed an order of magnitude increase in the activity of Src and Fyn kinase.48 Therefore, the activity of the Src-like kinases in CTCL cells, might be mostly influenced by the cytosolic enzyme whose activity is not altered in the neoplastic cells. In addition, Csk-related enzymes, such as Lsk or Ctk,49,50 may play a role. The reduced responsiveness to aCD3e stimulation in CTCL might be a result of their malignant phenotype and might confer a growth advantage. Stimulation of the TCR can produce a spectrum of functional responses that include not only increases in the proliferation of reactive clones that mediate immunity and inflammation, but also cell death, in order to limit further inflammation. Whereas naive, resting T cells are resistant to receptor-stimulated suicide, activated T cells of all subsets are sensitive.51,52 Recent experiments also suggested a requirement for PTK activity (possibly that of fyn) in super antigen-induced programmed cell death of peripheral T cells in vivo.53 Therefore, the ability to avoid apoptotic signaling may protect the neoplastic lymphocytes in CTCL and provide a selective growth advantage.

8 9 10

11

12

13

14 15

Acknowledgements We thank Dr David Rothstein for valuable discussion; ShuLing Yan for the flow cytometry analysis; Elaine Cheng and Yuhua Zhang for technical assistance; Jack Schreiber for help in preparing the manuscript; and Inger Christensen for help in collecting the blood from patients and volunteers. This work was supported by UPHS grants CA44542 (RH), CA43058 (RLE) and AR41942 (Yale Skin Diseases Research Center; RE Tigelaar, principal investigator), AI35603 and GM48960 (TM). Dr C Couture is the recipient of a postdoctoral fellowship from ‘Le Fonds de la Recherche´ en Sante´ du Que´bec’.

16

17 18 19 20 21

References 1 Edelson R, Berger C, Raafat J, Warburton D. Karyotype studies of cutaneous T cell lymphoma: evidence for clonal origin. J Invest Dermatol 1979; 73: 548–550. 2 Weiss LM, Hu E, Wood GS, Moulds C, Cleary ML, Warnke R, Sklar J. Clonal rearrangements of T-cell receptor genes in mycosis fungoides and dermatopathic lymphadenopathy. New Engl J Med 1985; 313: 539–544. 3 Berger CL, Eisenberg A, Soper L, Chow J, Simone J, Gapas Y, Cacciapaglia B, Bennett L, Edelson RL, Warburton D, Benn P. Dual genotype in cutaneous T cell lymphoma: immunoglublin gene rearrangement in clonal T cell malignancy. J Invest Dermatol 1988; 90: 73–77. 4 Reinhold U, Pawelec G, Fratila A, Leippold S, Bauer R, Kreysel HW. Phenotypic and functional characterization of tumor infiltrating lymphocytes in mycosis fungoides: continuous growth of CD4+ CD45R + T-cell clones with suppressor-inducer activity. J Invest Dermatol 1990; 94: 304–309. 5 Foss FM, Sausville EA. Prognosis and staging of cutaneous T-cell lymphoma. (Review). Hematol Oncol Clin North Am 1995; 9: 1011–1019. 6 Berger CL, Morrison S, Chu A, Patterson J, Estabrook A, Takezaki S, Sharon J, Warburton D, Irigoyen O, Edelson RL. Diagnosis of cutaneous T cell lymphoma by use of monoclonal antibodies reactive with tumor-associated antigens. J Clin Invest 1982; 70: 1205–1215. 7 Heald PW, Berger C, Yamamura T, Macnamara J, Edelson RL. BE2 antigen: appearance during T cell activation and increase with

22

23

24

25 26

27 28 29

stimulation and growth of T cells. J Invest Dermatol 1990; 94: 452–455. Rich BE, Leder P. Transgenic expression of interleukin 7 restores T cell populations in nude mice. J Exp Med 1995; 181: 1223–1238. Rich BE, Campos-Torres J, Tepper RI, Moreadith RW, Leder P. Cutaneous lymphoproliferation and lymphomas in interleukin 7 transgenic mice. J Exp Med 1993; 177: 305–316. Dalloul A, Laroche L, Bagot M, Mossalayi MD, Fourcade C, Thacker DJ, Hogge DE, Merle-Beral H, Debre P, Schmitt C. Interleukin-7 is a growth factor for Sezary lymphoma cells. J Clin Invest 1992; 90: 1054–1060. Abrams JT, Lessin S, Ghosh SK, Ju W, Vonderheid EC, Nowell P, Murphy G, Elfenbein B, DeFreitas E. A clonal CD4-positive T-cell line established from the blood of a patient with Sezary syndrome. J Invest Dermatol 1991; 96: 31–37. Foss FM, Koc Y, Stetler-Stevenson MA, Nguyen DT, O’Brien MC, Turner R, Sausville EA. Costimulation of cutaneous T-cell lymphoma cells by interleukin-7 and interleukin-2: potential autocrine or paracrine effectors in the Sezary syndrome. J Clin Oncol 1994; 12: 326–335. Heald PW, Yan SL, Edelson RL, Tigelaar R, Picker LJ. Skin-selective lymphocyte homing mechanisms in the pathogenesis of leukemic cutaneous T-cell lymphoma. J Invest Dermatol 1993; 101: 222–226. Janeway C Jr, Bottomly K. Signals and signs for lymphocyte responses. (Review). Cell 1994; 76: 275–285. Hansen ER, Vejlsgaard GL, Cooper KD, Heidenheim M, Larsen JK, Ho VC, Ross CW, Fox DA, Thomsen K, Baadsgaard O. Leukemic T cells from patients with cutaneous T-cell lymphoma demonstrate enhanced activation through CDw60, CD2, and CD28 relative to activation through the T-cell antigen receptor complex. J Invest Dermatol 1993; 100: 667–673. Rudd CE, Janssen O, Cai YC, da Silva AJ, Raab M, Prasad KV. Two-step TCR zeta/CD3-CD4 and CD28 signaling in T cells: SH2/SH3 domains, protein-tyrosine and lipid kinases. (Review). Immunol Today 1994; 15: 225–234. Weiss A, Littman DR. Signal transduction by lymphocyte antigen receptors. (Review). Cell 1994; 76: 263–274. Mustelin T. T Cell antigen receptor signaling: three families of tyrosine kinases and a phosphatase. Immunity 1994; 1: 351–356. Kolanus W, Romeo C, Seed B. T cell activation by clustered tyrosine kinases. Cell 1993; 74: 171–183. van Oers NSC, Weiss A. The Syk/ZAP-70 protein tyrosine kinase connection to antigen receptor signalling processes. Semin Immunol 1995; 7: 227–236. Chow LML, Veillette A. The Src and Csk families of tyrosine protein kinases in hemopoietic cells. Semi Immunol 1995; 7: 207– 226. Burkhardt AL, Stealey B, Rowley RB, Mahajan S, Prendergast M, Fargnoli J, Bolen JB. Temporal regulation of non-transmembrane protein tyrosine kinase enzyme activity following T cell antigen receptor engagement. J Biol Chem 1994; 269: 23642–23647. Wange RL, Kong AN, Samelson LE. A tyrosine-phosphorylated 70kDa protein binds a photoaffinity analogue of ATP and associates with both the zeta chain and CD3 components of the activated T cell antigen receptor. J Biol Chem 1992; 267: 11685–11688. Straus DB, Weiss A. The CD3 chains of the T cell antigen receptor associate with the ZAP-70 tyrosine kinase and are tyrosine phosphorylated after receptor stimulation. J Exp Med 1993; 178: 1523–1530. Cambier JC, Jensen WA. The hetero-oligomeric antigen receptor complex and its coupling to cytoplasmic effectors. (Review). Curr Opin Genet Dev 1994; 4: 55–63. Couture C, Baier G, Altman A, Mustelin T. p56lck- independent activation and tyrosine phosphorylation of p72syk by T-cell antigen receptor/CD3 stimulation. Proc Natl Acad Sci USA 1994; 91: 5301–5305. Edelson RL. Photopheresis: a clinically relevant immunobiologic response modifier. (Review). Ann NY Acad Sci 1991; 636: 154– 164. Morimoto C, Letvin NL, Distaso JA, Aldrich WR, Schlossman SF. The isolation and characterization of the human suppressor inducer T cell subset. J Immunol 1985; 134: 1508–1515. Oetken C, Couture C, Bergman M, Bonnefoy-Berard N, Williams S, Alitalo K, Burn P, Mustelin T. TCR/CD3-triggering causes

1345


1346 30

31 32

33

34 35 36 37

38 39 40 41

increased activity of the p50csk tyrosine kinase and engagement of its SH2 domain. Oncogene 1994; 9: 1625–1631. Bergman M, Mustelin T, Oetken C, Partanen J, Flint NA, Amrein KE, Autero M, Burn P, Alitalo K. The human p50csk tyrosine kinase phosphorylates p56lck at Tyr-505 and down regulates its catalytic activity. EMBO J 1992; 11: 2919–2924. Boulton TG, Cobb MH. Identification of multiple extracellular signal-regulated kinases (ERKs) with antipeptide antibodies. Cell Regul 1991; 2: 357–371. Stover DR, Charbonneau H, Tonks NK, Walsh KA. Protein-tyrosine-phosphatase CD45 is phosphorylated transiently on tyrosine upon activation of Jurkat T cells. Proc Natl Acad Sci USA 1991; 88: 7704–7707. Nada S, Okada M, MacAuley A, Cooper JA, Nakagawa H. Cloning of a complementary DNA for a protein-tyrosine kinase that specifically phosphorylates a negative regulatory site of p60c-src. Nature 1991; 351: 69–72. Howell BW, Cooper JA. Csk suppression of src involves movement of csk to sites of src activity. Mol Cell Biol 1994; 14: 5402–5411. Ford CE, Furlong MT, Geahlen RL, Harrison ML. Signalinginduced association of a tyrosine-phosphorylated 36-kDa protein with p50csk . J Biol Chem 1994; 269: 30378–30385. Catipovic B, Schneck JP, Brummet ME, Marsh DG, Rafnar T. Csk is constitutively associated with a 60-kDa tyrosine-phosphorylated protein in human T cells. J Biol Chem 1996; 271: 9698–9703. Neet K, Hunter T. The nonreceptor protein-tyrosine kinase CSK complexes directly with the GTPase-activating protein-associated p62 protein in cells expressing v-Src or activated c-Src. Mol Cell Biol 1995; 15: 4908–4920. Cloutier JF, Chow LM, Veillette A. Requirement of the SH3 and SH2 domains for the inhibitory function of tyrosine protein kinase p50csk in T lymphocytes. Mol Cell Biol 1995; 15: 5937–5944. Chan AC, Desai DM, Weiss A. The role of protein tyrosine kinases and protein tyrosine phosphatases in T cell antigen receptor signal transduction. (Review). Annu Rev Immunol 1994; 12: 555–592. Yakura H. The role of protein tyrosine phosphatases in lymphocyte activation and differentiation. Crit Rev Immunol 1994; 14: 311– 336. Sanders ME, Makgoba MW, June CH, Young HA, Shaw S. Enhanced responsiveness of human memory T cells to CD2 and

42 43

44 45 46 47 48

49 50 51 52 53

CD3 receptor-mediated activation. Eur J Immunol 1989; 19: 803–808. Fischer EH, Charbonneau H, Tonks NK. Protein tyrosine phosphatases: a diverse family of intracellular and transmembrane enzymes. (Review). Science 1991; 253: 401–406. Autero M, Saharinen J, Pessamorikawa T, Soularothhut M, Oetken C, Gassmann M, Bergman M, Alitalo K, Burn P, Gahmberg CG, Mustelin T. Tyrosine phosphorylation of CD45 phosphotyrosine phosphatase by p50csk kinase creates a binding site for p56lck tyrosine kinase and activates the phosphatase. Mol Cell Biol 1994; 14: 1308–1321. Perlmutter RM. Immunodeficiency. Zapping T-cell responses (news). Nature 1994; 370: 249–250. Madrenas J, Wange RL, Wang JL, Isakov N, Samelson LE, Germain RN. z phosphorylation without ZAP-70 activation induced by TCR antagonists or partial agonists. Science 1995; 267: 515–518. Sloan-Lancaster J, Shaw AS, Rothbard JB, Allen PM. Partial T cell signaling: altered phospho-z and lack of Zap70 recruitment in APL-induced T cell anergy. Cell 1994; 79: 913–922. Abraham N, Miceli MC, Parnes JR, Veillette A. Enhancement of T-cell responsiveness by the lymphocyte-specific tyrosine protein kinase p56lck. Nature 1991; 350: 62–66. Imamoto A, Soriano P. Disruption of the csk gene, encoding a negative regulator of Src family tyrosine kinases, leads to neural tube defects and embryonic lethality in mice. Cell 1993; 73: 1117–1124. Hamaguchi I, Iwama A, Yamaguchi N, Sakano S, Matsuda Y, Suda T. Characterization of mouse non-receptor tyrosine kinase gene, HYL. Oncogene 1994; 9: 3371–3374. Klages S, Adam D, Class K, Fargnoli J, Bolen JB, Penhallow RC. Ctk: a protein-tyrosine kinase related to Csk that defines an enzyme family. Proc Natl Acad Sci USA 1994; 91: 2597–2601. Russell JH, Rush BJ, Abrams SI, Wang R. Sensitivity of T cells to anti-CD3-stimulated suicide is independent of functional phenotype. Eur J Immunol 1992; 22: 1655–1658. Russell JH, Rush B, Weaver C, Wang R. Mature T cells of autoimmune lpr/lpr mice have a defect in antigen-stimulated suicide. Proc Natl Acad Sci USA 1993; 90: 4409–4413. Migita K, Eguchi K, Kawabe Y, Nagataki S. Tyrosine phosphorylation participates in peripheral T-cell activation and programmed cell death in vivo. Immunology 1995; 85: 550–555.