Increased Expression of a Novel Early Activation Surface Membrane Receptor in Cutaneous T Cell Lymphoma Cells Maria Nikolova,*1 Abdul Tawab,*1 Anne Marie-Cardine,* Martine Bagot,*² Laurence Boumsell,* and Armand Bensussan* *INSERM 448 and ²Service de Dermatologie, HoÃpital Henri Mondor, CreÂteil, France
cytes in Sezary syndrome patients was signi®cantly increased in comparison with controls (p < 0.01) and correlated with the morphologically detected percentage of Sezary syndrome cells in peripheral blood (p < 0.001). In one patient we clearly demonstrated that the circulating malignant T cells coexpress SC5 molecules. Importantly, ligation of SC5 receptor in a cutaneous T cell lymphoma cell line profoundly inhibited the anti-CD3-induced proliferation. Consequently, the expression of SC5 receptor in the peripheral blood of Sezary syndrome patients may serve not only to detect the presence of circulating malignant CD4+ cells but also as a target for immunotherapy. Key words: CTCL/ lymphocyte antigen/ lymphocyte proliferation inhibition/ Sezary syndrome. J Invest Dermatol 116:731±738, 2001
Using a newly generated monoclonal antibody we identi®ed the 96 kDa transmembrane receptor SC5 expressed simultaneously on a human Sezary cell line and a minor T cell subset in normal individuals. SC5 antigen was detected mostly on CD45RO+ lymphocytes from both CD4+ and CD8+ subsets as well as on natural killer and B lineage cells. SC5 surface expression increased very early after polyclonal stimulation of CD3+ cells due to the transfer of intracellular SC5 molecules to the cell membrane. Engagement of SC5 receptor by its monoclonal antibody inhibited the anti-CD3-induced proliferation and cytokine secretion of peripheral blood T cells and cell clones, whereas SC5 monoclonal antibody did not affect the cytotoxic activity of CD8+ T cell clones. Extensive phenotypic analysis revealed that the percentage of SC5+ CD4+ circulating lympho-
T
1999; Hutloff et al, 1999; Tamada et al, 2000). Negative regulation of T lymphocyte responses can be mediated by CD152-induced signals (Thompson and Allison, 1997), by apoptosis through members of the tumor necrosis factor (TNF) receptor superfamily (Lenardo et al, 1999), or by the inhibitory receptors recently discovered on subsets of natural killer cells and CD8+ T lymphocytes (Mingari et al, 1995, 1998; D'Andrea et al, 1996). In this report, we describe a novel 96 kDa transmembrane receptor, SC5, which delineates a minor subset of peripheral blood lymphocytes (PBL) in normal individuals. We found that SC5 was predominantly located in the intracellular compartment of resting T lymphocytes and its surface membrane expression increased rapidly after cell activation. SC5 molecule engagement inhibited the anti-CD3 monoclonal antibody (MoAb) induced proliferation of resting T lymphocytes or T cell clones. Further, we observed that the percentage of SC5+ CD4+ circulating lymphocytes was signi®cantly increased in Sezary syndrome (SS) patients compared to PBL of normal individuals. Importantly, we were able to demonstrate that Sezary cells express SC5 molecules and that their triggering resulted in a strong inhibition of a cutaneous T cell lymphoma (CTCL) cell line proliferation induced by anti-CD3 MoAbs. Thus, SC5 molecule expression may serve to detect the presence of circulating malignant CD4+ cells in SS patients. Moreover, the identi®cation of a cell surface molecule providing negative signals upon ligation constitutes a new tool to investigate the mechanisms of pathologic T cell growth and could be used to prevent it.
lymphocyte immune responses are regulated by functional cell surface molecules providing positive signals that lead to the expansion of antigen-speci®c clones, and negative signals that prevent excessive stimulation or responsiveness to self-antigens. The signal provided by the engagement of T cell receptors (TCR) must be accompanied by a second positive signal in order to result in optimal T cell response. Recent studies demonstrated that CD3/TCR stimulation leads to a redistribution of the detergent-insoluble glycolipid-enriched membrane fraction (DIG) or raft, which results in the aggregation of TCR/CD3 and DIG-associated signal-transducing molecules (Montixi et al, 1998). CD28, which is considered as the major T cell costimulatory receptor (Watts and DeBenedette, 1999; Yumi et al, 2000), enhances raft redistribution to the site of TCR engagement (Viola et al, 1999). Further, a series of other molecules may provide costimulatory signals to T cells, acting at different time points, affecting different subsets, or promoting distinct effector functions (Shuford et al, 1997; Agrawal et al, Manuscript received October 17, 2000; revised December 8, 2000; accepted for publication December 29, 2000. Reprint requests to: Dr. Armand Bensussan, INSERM 448, Faculte de MeÂdecine de CreÂteil, 8 rue du GeÂneÂral Sarrail, 94010 CreÂteil, France. Email:
[email protected] Abbreviations: DIG, detergent-insoluble glycolipid-enriched fraction; KIR, killer cell immunoglobulin receptor; MF, mycosis fungoides; PBL, peripheral blood lymphocytes; PBMC, peripheral blood mononuclear cells; SS, Sezary syndrome. 1 The ®rst two authors contributed equally to this work. 0022-202X/01/$15.00
´ Copyright # 2001 by The Society for Investigative Dermatology, Inc. 731
732
NIKOLOVA ET AL
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
Table I. Expression of SC5 in peripheral blood subpopulations SC5+ cells Cell population
N-tested
mean (%)
SD
Total Ly CD3+ Ly CD4+ Ly CD8+ Ly CD56+ Ly CD19+ Ly Total Mo Total Gr
15 10 8 8 10 5 5 5
14.1 10.5 9.4 14.8 19.4 12.5 100 45
6.8 5.6 5.1 5.1 11.4 5.7 ± 5.2
MATERIALS AND METHODS Production of anti-SC5 MoAb SC5 MoAb was obtained by immunizing 6-wk-old BALB/C mice with the natural killer cell line Ytindi, using an established protocol (David et al, 1990). Hybridoma supernatants were screened for selective reactivity with the immunizing cell line and a CTCL cell line termed Pno (Poszepczynska et al, 2000) by indirect immuno¯uorescence and ¯ow cytometry. The reactive supernatants were further tested on PBL. A hybridoma supernatant reacting with both tumoral cell lines and with a minor PBL subpopulation was selected and cloned twice, and cloned hybridomas were passaged into pristane-primed BALB/C mice to produce ascites. The antibody was termed anti-SC5 MoAb, and its isotype was determined as IgM. The ascites was dialyzed against phosphate-buffered saline (PBS), sterilized by ultra®ltration, and further utilized at a ®nal dilution of 1:200. Patients After informed consent and approval by an ethics committee (CCPPRB, HoÃpital Henri Mondor, Creteil), we obtained blood samples from 12 patients with CTCL. Nine patients had an SS with 10%±80% circulating CD3+, CD4+ Sezary cells, whereas three patients had a transformed mycosis fungoides (MF) and presented with disseminated skin tumors with a CD3+, CD4+, CD8± phenotype. The patients had not been previously treated with chemotherapy. Cells and cell lines Peripheral blood mononuclear cells (PBMC) were isolated by the technique of Ficoll-Isopaque (Pharmacia, Piscataway, NJ) density gradient centrifugation. Human T cell clones GDS.3 (CD3+ TCRab+ CD4+ CD8±), DS6 (CD3+ TCRgd+ CD4± CD8±), LSO (CD3+ TCRgd+ CD4± CD8±), and JF1 (CD3+ TCRab+ CD4± CD8+), described elsewhere (David et al, 1987, 1988; Vilmer et al, 1988), were fed each 8 d with irradiated allogenic PBMC in culture medium consisting of RPMI 1640 (Gibco, Paisley, U.K.), 2 mmol per l L-glutamine, penicillin (100 U per ml), streptomycin (100 mg per ml), 10% heat-inactivated human serum, 50 IU per ml recombinant interleukin-2 (rIL-2) (kindly provided by Sano® SyntheÂlabo, LabeÁge, France), and 1 mg per ml phytohemagglutinin (PHA) (Wellcome, Beckenham, U.K.). Functional studies were performed at day 7 after the feeding. Standard human leukemic cell lines were mycoplasma-free and maintained in logarithmic growth in complete RPMI medium supplemented with 10% fetal bovine serum and antibiotics. The human CTCL cell clone, Pno, was established from peripheral blood of a CTCL patient and maintained in culture as previously described (Poszepczynska et al, 2000). The Pno cell line has a CD3+ Vb22+ CD4+ CD8aa+ CD25± phenotype and proliferates in response to IL-7 and to anti-CD3 MoAb stimulation. Monoclonal antibodies and ¯ow cytometry studies Indirect immuno¯uorescent staining was performed with hybridoma supernatants or ascites ¯uid using ¯uorescein isothiocyanate (FITC) conjugated goat antimouse Ig from Caltag Laboratories (San Francisco, CA). For twocolor analysis the immuno¯uorescent staining was performed by incubating 3 3 105 cells with anti-SC5 MoAb for 30 min at 4°C. Cells were then washed and incubated with FITC-conjugated goat antimouse IgM (Caltag Laboratories) followed by a second PE-, ECD- or TRIconjugated speci®c MoAb of IgG isotype. Stained cells were analyzed using a single argon ¯ow cytometer analyzer (Epics XL, BeckmanCoulter, Miami, FL) as described previously (Schiavon et al, 1999). For intracellular labeling, cells were ®xed in PBS 4% p-formaldehyde for
20 min at 4°C, washed, and then permeabilized with staining buffer containing 0.1% saponin. Conjugated anti-CD3, anti-CD4, anti-CD8, anti-CD45RO, anti-CD69, and anti-TCRVb22 MoAbs were purchased from Immunotech (Marseille, France). Other MoAbs were locally produced or obtained through the exchanges of the Vth International Workshop on white cell differentiation antigens (Boston, MA, November 1994). Immunoprecipitation of biotinylated surface molecules Two 3 107 cells were washed twice in PBS and surface-labeled with SulfoNHS-LC biotin (Pierce Europe, Interchim, MontlucËon, France), 2 mg per ml in PBS for 20 min at 4°C. After quenching for 20 min with RPMI 1640, cells were washed twice in PBS and resuspended in lysis buffer (20 mM Tris-HCl pH 7.5, 150 mM NaCl, 1% Triton X-100, 1 mM Na vanadate, 10 mM NaF, 1 mM phenylmethylsulfonyl ¯uoride, 1 mg per ml aprotinin, and 1 mg per ml leupeptin) for 1 h at 4°C. Postnuclear supernatant was then incubated for 2 h at 4°C in a 96-well plate (Maxisorp Nunc immunoplate) precoated with goat antimouse IgG + IgM (Caltag Laboratories) followed by anti-SC5 MoAb or antiTCRb (C305 isotype matched MoAb, kindly provided by Dr. A. Weiss, University of California, San Francisco, CA). Immunoprecipitates were washed four times with washing buffer (20 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.1% Triton X-100, 1 mM Na vanadate, 10 mM NaF, 1 mM phenylmethylsulfonyl ¯uoride) and precipitated proteins were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Western blot analysis was performed using streptavidinperoxidase (Immunotech) and the ECL detection system according to the manufacturer's recommendations (Amersham Pharmacia, Orsay, France). Proliferation assays For lymphocyte activation, PBMC were cultured in culture medium containing 15% human serum in the presence of 1 mg per ml PHA (Wellcome). For proliferation assays, 5 3 104 cells were cultured in triplicates in 96-well round-bottomed plates (Greiner, NuÈrtingen, Germany) in a ®nal volume of 0.2 ml culture medium. When needed, the plates were precoated with anti-CD3 MoAb in the indicated dilutions as previously described (Poszcepczynska et al, 2000). For stimulation of the Pno cell line, precoated anti-CD3 MoAb (2.5 mg per well) was used in combination with 2 ng per ml phorbol 12bmyristate 13a-acetate (PMA, Sigma Biochemicals). Cells were cultured for 4 d and were pulsed with 1 mCi of 3H[TdR] during the last 8±16 h of culture. 3H[TdR] incorporation was measured in a liquid scintillation counter (Topcount; Packard Instrument, Meriden, CT) For the determination of IL-2 production, 105 PBL were stimulated with immobilized anti-CD3 MoAb in the presence of anti-SC5 or control MoAb. After 32 h of culture, 100 ml of supernatant was added to 105 cells from an Il-2-dependent T cell clone. CD3-induced redirected cytotoxicity assays The redirected cytotoxicity assay was carried out as already described (Le Cleach et al, 2000). The target murine mastocytoma P815 tumor cells were loaded with 100 ml 51Cr (2.5 mCi per ml) for 90 min at 37°C, washed, and incubated with anti-CD3 for 15 min at room temperature. Effector cells were likewise preincubated with anti-SC5 MoAb (1:200 ®nal dilution of ascites) and added to target cells in the ratio 5:1, in a ®nal volume of 150 ml in 96-well V-bottomed microtiter plates for 4 h. After centrifugation 100 ml aliquots were counted in a gamma-counter to determine 51Cr release. The spontaneous release was always less than 20% of the maximum release (target cells with 1% Nonidet P-40). The percentage of speci®c 51Cr release was calculated as previously described (David et al, 1987). Statistical analysis Statistical analysis of the results was performed with Statistica software version 5.0 (StatSoft, Los Angeles, CA). Signi®cant differences between SC5 expression in patients and the control group were evaluated by the Mann±Whitney U test, and data were presented using boxplots. Correlation between the percentage of SC5+ CD4+ cells and the percentage of malignant cells in patients' peripheral blood was evaluated by the Spearman rank order correlations test.
RESULTS Expression of SC5 molecules on normal resting and activated peripheral blood cells Monoclonal antibodies raised against the functional tumor cell line Ytindi were analyzed for their simultaneous reactivity with the immunizing cells and the CTCL cell line Pno (Poszcepczynska et al, 2000). The SC5-speci®c MoAb, of IgM isotype, stained both the tumor cells and a minor subset of PBL. The SC5 MoAb reactive molecule was expressed by
VOL. 116, NO. 5 MAY 2001
NOVEL CELL MEMBRANE MARKER FOR CTCL CELLS
733
Figure 1. Cell membrane expression of the SC5 molecule on PBL from normal individuals. PBMC from normal individuals were stained with anti-SC5 MoAb (ascites 1:200) followed by FITC-conjugated isotype-speci®c goat antimouse second reagent and one of the following PE-conjugated MoAbs: anti-CD3, and anti-CD45RO. The percentage of double-stained lymphocyte-gated cells is shown in the upper right quadrant. This experiment is representative for ®ve donors studied.
Figure 2. Cell membrane expression of the SC5 molecule during activation of peripheral blood T cells. PBMC were stimulated with 1 mg per ml PHA and the kinetics of SC5 and CD69 expression was studied in parallel on the CD3+ cells.
a variable percentage of CD3+ T cells, never exceeding 20% (mean 10.5%, SD 65.6). Both CD4+ and CD8+ lymphocytes were stained by anti-SC5 MoAb. A subpopulation of peripheral blood CD56+ natural killer cells was also reactive with anti-SC5 MoAb, though with more important interindividual variations (mean 19.4%, SD 611.4) (Table I). It should be noted that most SC5+ lymphocytes belonged to the activation/memory pool, over 80% of the SC5-positive cells coexpressing CD45RO (Fig 1). Table I further indicates that anti-SC5 MoAb stained a subpopulation of B cells, practically all monocytes, and 30%±50% of granulocytes. As SC5 molecule expression was detected in all IL-2-dependent T lymphocyte clones tested (data not shown), we veri®ed whether its self-surface expression was induced during T lymphocyte activation. Stimulation of PBL with PHA promptly increased the expression of SC5 on CD3+ lymphocytes. Both the percentage of SC5+ T cells and the level of SC5 expression rose very rapidly after stimulation (Fig 2 and data not shown). As early as 4 h after stimulation with PHA, the percentage of SC5+ CD3+ cells increased 2- to 3-fold and reached a maximum after 24±48 h. In the course of 5±7 d of in vitro stimulation SC5 expression returned to the basal level. In parallel we studied the expression of the very early activation antigen CD69 and we observed similar kinetic pro®les of CD69 and SC5 on CD3+ lymphocytes. It should be noted that unlike CD69 molecules low levels of SC5 were already present on circulating PBL before stimulation (Figs 1, 2). Activation-induced antigen expression results either from de novo protein synthesis or cell surface export of preexisting intracellular
Figure 3. Intracellular localization of the SC5 molecule in PBL. Anti-SC5, anti-CD3z chain MoAb, and anti-BY55 isotype-matched MoAb were used for staining permeabilized (right) and nonpermeabilized (left) gated lymphocytes from a normal donor. The shaded histograms represent the labeling obtained with the indicated MoAb compared to an irrelevant control MoAb.
molecules. Therefore, we studied the expression of SC5 in nonactivated permeabilized PBL. As shown in Fig 3, anti-SC5 MoAb stained over 90% of permeabilized lymphocytes versus 16% of the same nonpermeabilized cells. The speci®city of anti-SC5 MoAb staining was con®rmed by using the isotype-matched MoAb BY55, which detects a GPI-anchored cell surface structure expressed by a subset of circulating lymphocytes (Agrawal et al, 1999). A comparable percentage of BY55+ cells was found in permeabilized and nonpermeabilized lymphocytes. As a positive control for the detection of an intracellular epitope, we used an anti-CD3z chain MoAb that only labeled permeabilized T
734
NIKOLOVA ET AL
lymphocytes. Thus, the surface expression of SC5 is dynamically regulated by the activation state of cells and its induction during T cell activation is due to the enhanced transport of the molecule from an intracellular pool to the surface membrane. Biochemical characterization of SC5 antigen In order to characterize the molecular weight of the structure identi®ed by anti-SC5 MoAb, an IL-2-dependent TCRgd+ clone, termed DS6 (David et al, 1988), was surface labeled with biotin and the cell lysates were immunoprecipitated with either anti-SC5 MoAb or an isotype-matched anti-TCRb MoAb (as negative control). AntiSC5 MoAb recognized a single molecule with apparent molecular
Figure 4. Biochemical analysis of SC5 molecule. DS6 T cell clone was surface labeled with biotin and 1% Triton X-100 lysates were immunoprecipitated using anti-SC5 MoAb. The samples were analyzed by SDS-10%-PAGE under reducing conditions. Anti-SC5 MoAb was used in two concentrations: 1:200 (lane 2) and 1:500 (lane 3). The negative control samples were precipitated with goat antimouse alone (lane 1) and isotype-matched irrelevant MoAb (lane 4). The molecular weight markers (kDa) are indicated on the left.
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
weight 96 kDa under reducing conditions (Fig 4). As tumor cell lines (the immunizing Ytindi cells) are often characterized by aberrant expression of carbohydrate epitopes (Derappe et al, 1996; Schiavon et al, 1999), we studied the reactivity of anti-SC5 MoAb with the SC5+ natural killer cell line NK3.3 after treatment with neuraminidase or sodium periodate in concentrations destroying known sialylated epitopes (digestion of CD75 epitope on Raji cells was used as a positive control). Expression of SC5 was not affected by neuraminidase or sodium periodate treatment (data not shown). Thus, we excluded a possible carbohydrate nature of the epitope. Inhibition of the anti-CD3 MoAb-induced T lymphocyte proliferation by anti-SC5 MoAb In order to examine the function of the SC5 molecule in normal T cells, we studied the effect of anti-SC5 MoAb alone or during the proliferative responses induced by immobilized anti-CD3 MoAb. We found that anti-SC5 MoAb alone or in combination with PMA did not induce the proliferation of peripheral blood T lymphocytes (data not shown). Interestingly, when soluble anti-SC5 MoAb was present together with immobilized anti-CD3 MoAb, a signi®cant inhibition of lymphocyte proliferation rate was observed. The inhibition varied between 33% and 66% at day 4, depending on the donor, compared to the effect of an isotype-matched irrelevant control MoAb (data not shown). As SC5 molecules are expressed on monocytes, it was important to determine whether the inhibitory effect was monocyte independent. We studied the proliferation of two T cell clones, GDS.3 and LSO, stimulated by immobilized anti-CD3 MoAb in the presence of anti-SC5 MoAb or an isotype control MoAb. The results obtained with both T cell clones indicated that engagement of SC5 molecules resulted in an inhibition of the proliferation to anti-CD3 MoAb (Fig 5A). It should be noted that the inhibitory effect of anti-SC5 MoAb on T cell proliferation depended on the concentration of the agonistic anti-CD3 MoAb. The inhibition obtained with anti-SC5 MoAb was signi®cant only when T cells were stimulated with optimally
Figure 5. Anti-SC5 MoAb modulates the anti-CD3-induced proliferation and cytokine secretion of normal T lymphocytes without affecting their cytotoxic activity. (A) The T cell clones LSO and GDS.3 were stimulated with immobilized anti-CD3 MoAb (1 mg of MoAb coated per well) or with IL-2 (50 IU per ml) in the presence of anti-SC5 MoAb (1:200 ®nal dilution of ascites) or an isotype-matched irrelevant MoAb (anti-CD34). Results shown are representative of at least three separate experiments and are expressed as mean cpm 6 SD of triplicate wells. (B) GDS.3 CD4+ T cell clone was stimulated with the indicated concentrations of precoated anti-CD3 MoAb in the presence of anti-SC5 MoAb or isotype-matched irrelevant MoAb. (C) IL-2 secretion in the supernatant of anti-CD3-stimulated PBL, in the presence of anti-SC5 or control MoAb, was assessed by measuring the proliferation of an IL-2-dependent T cell clone in the presence of: medium alone (a), supernatant from PBL/anti-CD3+ control MoAb (b), supernatant from PBL/anti-CD3+ anti-SC5 MoAb (c). The data represent the mean values 6 SD of triplicate determinations of one out of three representative experiments. (D) JF1, a cytotoxic CD8+ cell clone, was incubated with anti-SC5 MoAb (1:200 ®nal dilution of ascites) or with an isotype-matched control MoAb (anti-CD34) before coculture with the FcgR+ murine tumor cells P815 at an E:T ratio of 5:1. Target cells were preincubated with the indicated ®nal concentrations of anti-CD3 MoAb. Results are expressed as the mean of triplicate wells.
VOL. 116, NO. 5 MAY 2001
diluted anti-CD3 MoAb (Fig 5B). Such inhibitory behavior was also reported for anti-killer cell immunoglobulin receptor (antiKIR) MoAbs (Cambiaggi et al, 1999). The decrease of T cell proliferation following SC5 engagement could be due to a direct induction of cell death, a perturbation of the IL-2R expression, or the inhibition of cytokine synthesis. We found that anti-SC5 MoAb did not inhibit the IL-2-dependent proliferation of T cell clones (Fig 5A). Furthermore, the addition of rIL-2 to T cells preincubated with anti-SC5 MoAb (ascites diluted to 1:200) in combination with anti-CD3 for 48 h restored their proliferation rate. This would not be the case if the speci®c MoAb inhibited IL-2R expression or caused cell death (data not shown). Finally, in order to prove that SC5 triggering may inhibit cytokine production, we compared the amount of IL-2 secreted by anti-CD3-activated PBL in the presence or absence of anti-SC5 MoAb, using an IL-2-dependent T cell line. The quantity of IL-2 produced in the presence of anti-SC5 MoAb represented 40%±65% of the control values obtained with an isotype-matched antibody (Fig 5C). Thus, the simultaneous engagement of SC5 and CD3 molecules in T cells results in the decrease in cytokine production. We next studied the effect of anti-SC5 MoAb on T cell effector cytotoxic activity using an anti-CD3 MoAb redirected killing assay against the FcgR+ murine cell line P815. Figure 5(D) shows a representative experiment performed with a cytotoxic T cell clone at an E:T ratio of 5:1. The CD8+ T cell clone JF1, expressing a high amount of SC5 molecules, was induced to kill murine target cells following anti-CD3 MoAb stimulation. The simultaneous targeting of SC5 by its speci®c MoAb did not in¯uence JF1 cytotoxic activity, regardless of the E:T ratio (data not shown) or the concentration of stimulating anti-CD3 MoAb. Expression of SC5 molecules on malignant lymphocytes obtained from patients with SS As we initially screened antiSC5 MoAb for its reactivity with the Pno CTCL cell line, we further studied the expression of SC5 molecules in PBL isolated from nine patients with SS. Two-color ¯ow cytometry analysis showed that expression of SC5 was increased in the CD4+ subset (see representative results obtained with one patient in Fig 6A). The percentage of CD4+ SC5+ lymphocytes in SS patients' blood was signi®cantly higher compared to healthy controls (p = 0.0015). As shown in Fig 6(B) practically no overlap existed between the values observed for the SC5/CD4 population in the patients' group (median 32%, range 16%±83%) and in the control group (median 11%, range 3%±12%). In the nine SS patients studied, the expression of SC5 on CD4+ cells correlated with the percentage of circulating malignant cells detected by cytomorphology (r = 0.92, p = 0.0012) (Fig 6C). We also studied SC5 molecule expression in peripheral blood samples from three MF patients without blood involvement. No signi®cant difference in the levels of SC5 expression on peripheral blood CD4+ cells was observed between MF patients and healthy donors. In one SS patient we had previously demonstrated that the circulating malignant cells were clonal lymphocytes with a CD4+ TCRVb22+ phenotype (Poszepczynska et al, 2000). Here we demonstrated that SC5 was coexpressed by the TCRVb22+ cells, i.e., by the malignant cell population (Fig 7A). It should be noted that the pro®le of SC5 MoAb binding to TCRVb22+ cells corresponded to a weak homogeneous expression by the whole malignant clonal cell population (Fig 7B). Both the circulating malignant cells and the derived IL-7-dependent TCRVb22 Pno cell line exhibited a similar pro®le upon staining with anti-SC5 MoAb (data not shown). Interestingly, the malignant TCRVb22+ cells did not coexpress common activation markers such as CD69 (Fig 7C), CD25, or CD30 (Poszepczynska et al, 2000). Thus, the increased levels of SC5 in SS PBL were mostly due to the expression of the antigen on circulating malignant T cells. Inhibition of the anti-CD3 MoAb-induced CTCL cell line proliferation by anti-SC5 MoAb We next investigated whether SC5 molecules were able to provide a negative signal in malignant CTCL cells. We used the long-term cultured Pno
NOVEL CELL MEMBRANE MARKER FOR CTCL CELLS
735
CTCL tumor T cell line, which strongly proliferates in the presence of IL-7. As previously reported, the Pno cell line expresses functional T cell receptors, as immobilized anti-CD3 MoAb or the combination of soluble anti-CD3 MoAb and PMA signi®cantly induced its proliferation (Poszepczynska et al, 2000). We examined the effect of SC5 molecule engagement on the proliferation of Pno cells. Figure 8 shows that anti-SC5 MoAb strongly inhibited the anti-CD3 MoAb-induced proliferation of the CTCL cell line. In contrast, PMA alone (2 ng per ml) produced a low level of proliferation, which was not signi®cantly affected by anti-SC5. Notably, anti-SC5 MoAb had no effect on the IL-7-dependent proliferation, as Pno cells incubated with IL-7 in the presence of anti-SC5 MoAb proliferated as vigorously as in the presence of the isotype control MoAb. Thus, we concluded that SC5 antigen is functional in a CTCL malignant cell line and may speci®cally and profoundly inhibit its proliferation triggered through the TCR/ CD3 signaling pathway. DISCUSSION A number of receptors with dynamic surface expression can modulate T cell immune responses. In this study, we identi®ed a novel 96 kDa functional molecule expressed by a minor subset of PBL whose surface expression is induced rapidly during in vitro T cell activation. SC5 is distinct from the established molecules with similar molecular weight and/or T cell modulatory functions. First, SC5 surface expression was induced after 4 h of T cell activation. A similar kinetics has been observed for the very early activation antigen CD69 (Resti et al, 1994). Most established T-cell-speci®c inducible receptors such as 4-1BB (Shuford et al, 1997), OX-40 (Watts et al, 1999), LIGHT (Tamada et al, 2000), or CD152 (CTLA-4) (Thompson and Allison, 1997) appear on the T cell surface between 10 and 24 h following stimulation. Further, SC5 expression is not T cell restricted. The antigen was detected on subpopulations of natural killer and B lymphocytes, monocytes, and granulocytes. Such a wide distribution pattern is typical of certain KIR (Cantoni et al, 1999; Falco et al, 1999) and immunoglobulinlike transcript inhibitory receptors (Colonna et al, 1998; LopezBotet and Bellon, 1999). Most of these receptors do not depend on the activation status of cells, however. The prompt induction of SC5 molecules at the surface of activated T cells followed by their downregulation may result from traf®cking of intracellular molecules to the cell membrane and vice versa, as we detected SC5 in most nonactivated T lymphocytes after cell permeabilization. Such a cellular localization, which is modi®ed upon activation, has already been reported for CD152 (Linsley et al, 1996) and a heterogeneous family of lysosome associated membrane proteins (Goyert, 1997). Additional studies on the intracellular localization and traf®cking of SC5 would de®ne its proper role in T cell activation and functions. Our studies revealed that triggering of SC5 by its speci®c MoAb inhibited the anti-CD3 MoAb-induced T cell proliferation. This inhibitory effect required an optimal anti-CD3 MoAb stimulation, which is necessary to induce maximal SC5 surface expression in PBL. At the same time, the inhibitory threshold posed by SC5 engagement could be bypassed by excessive anti-CD3 MoAb stimulation. Recently, the negative regulation of immune responses has been extensively investigated, in particular in T lymphocytes and natural killer cells. The best studied T-cell-speci®c inhibitory receptor, CD152 (CTLA-4), is a typical activation-induced antigen, functionally coupled to CD28 costimulatory receptor. Indeed, CD152 may overcome CD28-positive signals at low antigenic concentrations and prevent immune response. Alternatively, it may downregulate and terminate T cell proliferation induced through CD28 by affecting IL-2R expression and IL-2 synthesis (Krummel and Allison, 1996; Thompson and Allison, 1997). Similarly to CTLA-4, SC5 is an intracellular molecule whose surface expression is regulated by the activation status of cells. Both receptors mediate their effects by affecting IL-2 synthesis. The identi®cation of SC5-speci®c ligands and signaling
736
NIKOLOVA ET AL
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
Figure 6. Cell membrane expression of the SC5 molecule on PBL from CTCL patients. Cells from peripheral blood of nine patients with SS and nine healthy donors were analyzed by two-color ¯uorescence for the expression of SC5 molecule by CD4+ lymphocytes. (A) Coexpression of SC5 and CD4 molecules in PBL from a patient with SS in comparison to a normal individual. (B) Plots representing the expression of SC5 on CD4+ cells in the two indicated groups. The ``box'' indicates the 25th to 75th percentile range, the median value is designated inside, and the whiskers specify the min and max values observed. (C) Correlation between the percentage of SC5+ CD4+ cells and the percentage of malignant cells in SS patients' peripheral blood.
Figure 7. SC5 surface membrane expression on malignant cells from an SS patient. (A) Two-color immuno¯uorescence analysis of PBMC from an SS patient with a circulating TCRVb22+ malignant T cell clone (Poszcepczynska et al, 2000). Cells were stained with anti-SC5 MoAb followed by PE-conjugated isotype-speci®c goat antimouse and FITCconjugated anti-TCRVb22 MoAb. (B) Anti-SC5 MoAb weakly labeled (shaded histogram) the TCRVb22-positive malignant cells. (C) Cells from the same patient stained with PE-conjugated anti-CD69 MoAb and FITC-conjugated antiTCRVb22 MoAb.
pathways will de®ne the temporal and spatial correlation between these negative T cell regulators. In several aspects, SC5 is comparable to KIR, equally expressed by T cells and other lymphocyte subpopulations.These receptors were shown to inhibit the antigen-speci®c and superantigen-driven activation of T cells, including cytotoxicity, proliferation, and cytokine secretion (Mingari et al, 1995; 1998; D'Andrea et al, 1996). SC5 is de®nitely distinct from KIR, however, as we demonstrated that it selectively inhibited the proliferation of antiCD3 MoAb-stimulated T cells without affecting their effector cytotoxic functions. An important characteristic of the SC5 molecule is its signi®cantly increased expression in PBL of patients with SS in comparison to healthy subjects. SS is characterized by the generalization of a primary epidermotropic malignant T cell clone with mature phenotype (Bagot et al, 1998). Actually, none
of the established T cell functional receptors conclusively identi®es peripheral blood SS tumor cells. MF/SS has been considered as the expansion of a normally existing CD4+ CD7± T cell memory subset (Reinhold et al, 1993; Reinhold and Abken, 1997). Interestingly, signi®cantly higher CD4+ CD7± cell counts were reported in patients with advanced CTCL compared to patients with early CTCL, and these counts correlated with the detection of clonality (Laetsch et al, 2000). Another study demonstrated that the CD4+ CD7± population does not directly represent the dominant T cell clone in SS patients, however, as clonal cells can be detected in the CD4+ CD7+ population as well (Dummer et al, 1999). Established activation antigens such as CD69, CD25, and CD30 are not consistently detected on Sezary cells, as con®rmed by our own investigations. Expression of CD25 correlates instead with the progression of the disease and the transformation of MF/SS cells into a large cell variant in a minority of cases (Diamandidou et al,
VOL. 116, NO. 5 MAY 2001
Figure 8. Modulation by anti-SC5 MoAb of anti-CD3 MoAbinduced proliferation of CTCL malignant cell line. Pno cell line was stimulated with IL-7 (10 ng per ml), PMA alone (2 ng per ml), immobilized anti-CD3 MoAb, or a combination of PMA/anti-CD3 MoAb. SC5 MoAb (1:200 ®nal dilution of ascites) or an isotypematched irrelevant MoAb, anti-CD34, were added at the start of cultures. Results shown are representative of four separate experiments and are expressed as mean cpm 6 SD of triplicate wells.
1998; Stefanato et al, 1998). An increased rate of expression of T cell activation molecules such as HLA-DR, CD25, or CD71 in SS has been primarily associated with reactive tumor-in®ltrating T cells (Wood et al, 1982; Wieczorek et al, 1986; Asadullah et al, 1997). A 78 kDa very late activation antigen, BE2, was reported to be speci®cally increased in SS peripheral blood, but not detected on normal T cells (Berger et al, 1982). In fact, BE2 was not consistently detected in all SS patients studied, and BE2 was equally detected in patients without clinical evidence of blood involvement. In contrast, SC5 was detected in nine of nine SS cases studied. Although the percentage of malignant cells determined by cytomorphology varied considerably from patient to patient (10%±80%), the expression of SC5 on CD4+ cells correlated with this percentage. Finally, no elevated expression of SC5 was detected on circulating CD4+ cells from MF patients without blood involvement or from atopic dermatitis patients (data not shown). Further, we demonstrated that Sezary cells identi®ed by the expression of the clonotypic TCRVb chain coexpressed SC5 molecules, con®rming that increased levels of SC5 in the peripheral blood of SS patients were due to its expression on the malignant T cell clone. A very important observation was that SC5 expression on Sezary cells induced a downmodulation of the CD3-mediated proliferation of tumor cells, analogous to the inhibition observed in normal T cells. It has already been demonstrated that malignant CTCL lymphocytes preserve certain functional properties as they may be activated through basic signaling pathways, although producing aberrant pro®les of cytokines (Bagot et al, 1998; Poszepczynska et al, 2000). We now demonstrate that an inhibitory pathway is preserved in CTCL tumor cells that may allow their proliferation to be counterbalanced. Notably, the inhibitory effect observed after anti-SC5 MoAb binding on anti-CD3 MoAb-stimulated CTCL Pno cells was much more important than in normal T cells. AntiSC5 MoAb does not seem to directly mediate cell death of either normal or Sezary T cells, as it does not modulate the rIL-2-induced proliferation. Moreover, T cells cultivated in the presence of antiSC5 MoAb can be restimulated by addition of rIL-2. At the same time, we observed a diminished IL-2 production by anti-SC5 MoAb-treated T cells, which should account at least partially for the inhibition of their proliferation. Leukocyte functional receptors with inhibitory effects on malignant cells have already been described. Recently, Moretta et al reported inhibition of the proliferation of normal and leukemic myeloid cells mediated by the CD33 myeloid-speci®c antigen and the homologous natural killer cell sialoadhesin receptor AIRMI (Vitale et al, 1999). Unlike SC5, these receptors inhibited proliferation in the presence of growth factors and most probably
NOVEL CELL MEMBRANE MARKER FOR CTCL CELLS
737
by induction of apoptosis. Interestingly, another T cell activation antigen, CD30, expressed by a subset of normal CD45RO+ T cells, is likely to mediate the regression of primary CTCL other than MF/SS. This effect is related to the natural ligand of CD30 (CD30L) that is colocated within the cytoplasm of malignant cells and may induce cytolytic cell death independently of the Fas/FasL system (Mori et al, 1999). In this aspect, the identi®cation of SC5 ligand will enlighten the mechanism of its own inhibitory action. In conclusion, we have identi®ed an early activation antigen that is distinct, by phenotypic, biochemical, and functional criteria, from the established T lymphocyte receptors. As has been emphasized, the SC5 molecule is expressed by a subpopulation of the postactivation/memory peripheral blood T cell subset. Likewise, continuous expression of KIR is detected on minor T cell clones, and this expression seems to depend on TCR occupancy by speci®c antigens. It was recently proposed that such clones might be autoreactive and maintained tolerant by KIR signaling (Huard and Karlsson, 2000). We may speculate that SC5 surface expression is also preserved on auto-reactive T cells, suggesting that SS tumor cells might originate from such clones. Importantly, the expression of SC5 on CTCL malignant cells will serve to better de®ne their origin and the nature of the signal driving mature T cells to uncontrolled proliferation. In addition, the inhibitory effect of anti-SC5 MoAb on tumor T cells could constitute a novel therapeutic approach for patients with advanced CTCL. NOTE ADDED IN PROOF While this manuscript was submitted, Saverino et al, J Immunol 165:3742±3755, 2000, revealed that the inhibitory receptor CD85/ ILT2 was present in the cytoplasm of all human T lymphocytes and its surface expression was increased upon activation. We performed phenotypic and biochemical comparison between SC5 and CD85/ ILT2 molecules using speci®c MoAbs. The results clearly demonstrated that these molecules are different. This work was supported partially by grants from INSERM, Paris, France; Association de la Recherche contre le Cancer (grant numbers 9669 and 5427), Villejuif, France; SocieÂte FrancËaise de Dermatologie, Paris, France; Laboratoires La Roche Posay, La Roche-Posay, France; and AcadeÂmie de MeÂdecine, Paris, France. We would like to thank Pr. H. Taskov, Central Laboratory of Immunology, National Center of Infectious Diseases, So®a, for his expert help in statistics.
REFERENCES Agrawal S, Marquet J, Freeman GJ, et al: Cutting edge. MHC class I triggering by a novel cell surface ligand costimulates proliferation of activated human T cells. J Immunol 162:1223±1226, 1999 Asadullah K, Friedrich M, Docke WD, Jahn S, Volk HD, Sterry W: Enhanced expression of T-cell activation and natural killer cell antigens indicates antitumor response in early cutaneous T cell lymphoma. J Invest Dermatol 108:743± 747, 1997 Bagot M, Echchakir H, Mami-Chouaib F, et al: Isolation of tumor-speci®c cytotoxic CD4+ and CD4+CD8dim+ T-cell clones in®ltrating a cutaneous T cell lymphoma. Blood 91:4331±4341, 1998 Berger CL, Morrison S, Chu A, et al: Diagnosis of cutaneous T cell lymphoma by use of monoclonal antibodies reactive with tumor-associated antigens. J Clin Invest 70:1205±1215, 1982 Cambiaggi A, Darche S, Guia S, Kourilski P, Abastado JP, Vivier E: Modulation of T cell functions in KIR3DL3 (CD158b) transgenic mice. Blood 94:2396±2402, 1999 Cantoni C, Bottino C, Augugliaro R, et al: Molecular and functional characterization of IRp60, a member of the immunoglobulin superfamily that functions as an inhibitory receptor in human NK cells. Eur J Immunol 29:3148±3159, 1999 Colonna M, Samaridis J, Cella M, et al: Human myelomonocytic cells express an inhibitory receptor for classical and nonclassical MHC class I molecules. J Immunol 160:3096±3100, 1998 D'Andrea A, Chang C, Phillips JH, Lanier LL: Regulation of T cell lymphokine production by killer cell inhibitory receptor recognition of self HLA class I allels. J Exp Med 184:789±794, 1996 David V, Bourge J-F, Gugliemi P, Mathieu-Mahul D, Degos L, Bensussan A: Human T cell clones use a CD3-associated surface antigen recognition
738
NIKOLOVA ET AL
structure to exhibit both NK-line and allogeneic cytotoxic reactivity. J Immunol 138:2831±2836, 1987 David V, Leca E, Vilmer E, et al: Expression of T cell gamma gene by a functionally de®ned T cell clone. Scand J Immunol 27:473±483, 1988 David V, Bachelez E, Leca G, Degos L, Boumsell L, Bensussan A: Identi®cation of a novel 110 kD structure on a subset of T cell receptor-gd-bearing cloned lymphocytes. J Immunol 144:1±6, 1990 Derappe C, Haentjens G, Lemaire S, et al: Circulating malignant lymphocytes from Sezary syndrome express high level of glycoproteins carrying b (1±6) Nacetylglucosanine-branched N-linked oligosaccharides. Leukemia 10:138±141, 1996 Diamandidou E, Colome-Grimmer M, Fayad L, Duvic M, Kurzrock R: Transformation of mycosis fungoides/Sezary syndrome: clinical characteristics and prognosis. Blood 92:1150±1159, 1998 Dummer R, Nestle FO, Niedere E, et al: Genotypic phenotypic and functional analysis of CD4+CD7± T lymphocyte subsets in Sezary syndrome. Arch Dermatol Res 291:307±311, 1999 Falco M, Biassoni R, Bottino C, et al: Identi®cation and molecular cloning of p75/ AIRM1, a novel member of the sialoadhesin family that functions as an inhibitory receptor in human natural killer cells. J Exp Med 190:793±801, 1999 Goyert S: CD68 Workshop panel report. In: Kishimoto T, von Kikutani H, dem Borne AEG, et al (eds). Leukocyte Typing VI. New York: Garland Publishing, 1987:pp 1015±1016 Huard B, Karlsson L: KIR expression on self-reactive CD8+ T cells is controlled by T-cell receptor engagement. Nature 403:325±328, 2000 Hutloff A, Dittrich A, Beier KC, Eljaschewitsch B, Kraft R, Anagnostopoulos I, Kroczek R: ICOS is an inducible T-cell co-stimulator structurally and functionally related to CD28. Nature 397:263±266, 1999 Krummel M, Allison J: CTLA-4 engagement inhibits IL-2 accumulation and cell cycle progression upon activation of resting T cells. J Exp Med 183:2533±2540, 1996 Laetsch B, HaÈffner A, DoÈbbeling U, Seifert B, Ludwig E, Burg G, Dummer R: CD4+/CD7+ T cell frequency and polymerase chain reaction-based clonality assay correlate with stage in cutaneous T cell lymphomas. J Invest Dermatol 114:107±111, 2000 Le Cleach L, Delaire S, Boumsell L, Bagot M, Bourgault-Villada I, Bensussan A, Roujeau JC: Blister ¯uid T lymphocytes dring toxic epidermal necrolysis are functional cytotoxic cells which express human natural killer (NK) inhibitory receptors. Clin Exp Immunol 119:225±230, 2000 Lenardo M, Chan FKM, Hornung F, McFarland H, Siegel R, Wang J, Zheng L: Mature T lymphocyte apoptosis-immune regulation in a dynamic and unpredictable antigenic environment. Annu Rev Immunol 17:221±253, 1999 Linsley P, Bradshaw J, Greene J, Peach R, Bennett K, Mittler R: Intracellular traf®cking of CTLA-4 and focal colocalization towards sites of TCR engagement. Immunity 4:535±543, 1996 Lopez-Botet M, Bellon T: Natural killer cell activation and inhibition by receptors for MHC class I. Curr Opin Immunol 11:301±307, 1999 Mingari MC, Vitale C, Cambiaggi A, Sciavetti F, Melioli G, Ferrini S, Poggi A: Cytolytic T lymphocytes displaying natural killer (NK) -like activity: expression of NK-related functional receptors for HLA class I molecules (p58 and CD94) and inhibitory effect on the TCR-mediated target cell lysis or lymphokine production. Int Immunol 7:697±703, 1995 Mingari MC, Moretta A, Moretta L: Regulation of KIR expression in human cells: a safety mechanism that may impair protective T-cell responses. Immunol Today 19:155±157, 1998
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
Montixi C, Langlet C, Bernard AM, et al: Engagement of T cell receptor triggers its recruitment to low-density detergent-insoluble membrane domains. EMBO J 17:5334±5348, 1998 Mori M, Manuelli C, Pimpinelli N, et al: CD30±CD30 ligand interaction in primary cutaneous CD30+ T-cell lymphomas: a clue to pathophysiology of clinical regression. Blood 94:3077±3083, 1999 Poszepczynska E, Bagot M, Echchakir H, et al: Functional characterization of an IL7-dependent CD4+CD8aa+ Th3-type malignant cell line derived from a patient with a cutaneous T-cell lymphoma. Blood 96:1056±1063, 2000 Reinhold U, Abken H: CD4+CD7± T cells: a separate subpopulation of memory T cells? J Clin Immunol 17:265±267, 1997 Reinhold U, Abken H, Kukel S, Moll M, Muller R, Oltermann I, Kreysel HW: CD7± T cells represent a subset of normal human blood lymphocytes. J Immunol 150:2081±2089, 1993 Resti R, D'Ambrosio D, De Maria R, Santoni A: The CD69 receptor: a multipurpose cell-surface trigger for hematopoetic cells. Immunol Today 15:479±483, 1994 Schiavon V, Elhabazi A, Agrawal S, Tawab A, Nikolova M, Boumsell L, Bensussan A: G10.3 monoclonal antibody identi®es novel functional cell surface structures expressed by normal B lymphocytes and various malignant cell lines. Tissue Antigens 53:23±32, 1999 Shuford W, Klussman K, Trichler DD, et al: 4-1BB costimulatory signals preferentially induce CD8+ T cell proliferation and lead to ampli®cation in vivo of cytotoxic T cell responses. J Exp Med 186:47±55, 1997 Stefanato CM, Tallini G, Crotty PL: Histologic and immunophenotypic features prior to transformation in patients with transformed cutaneous T-cell lymphoma: is CD25 expression in skin biopsy predictive of large cell transformation in cutaneous T-cell lymphoma? Am J Dermatopathol 20:1±6, 1998 Tamada K, Shimozaki K, Chapoval AI, et al: Modulation of T-cell-mediated immunity in tumor and graft-versus-host disease models through the LIGHT co-stimulatory pathway. Nature Med 6:283±289, 2000 Thompson K, Allison J: The emerging role of CTLA-4 as an immune attenuator. Immunity 7:445±450, 1997 Vilmer E, Triebel F, David V, et al: Prominent expansion of circulating lymphocytes bearing g T-cell receptors, with preferential expression of variable g genes after allogeneic bone marrow transplantation. Blood 72:841±849, 1988 Viola A, Schroeder S, Sakakibara Y, Lanzavecchia A: T lymphocyte co-stimulation mediated by reorganization of membrane microdomains. Science 283:680±682, 1999 Vitale C, Falco M, Ponte M, et al: Engagement of p75/AIRM1 or CD33 inhibits the proliferation of normal and leukemic myeloid cells. Proc Natl Acad Sci USA 96:15091±15096, 1999 Watts T, DeBenedette M: T cell co-stimulatory receptors other than CD28. Curr Opin Immunol 11:286±293, 1999 Wieczorek R, Suhrland M, Ramasay D, Reed ML, Knowles DM,II: LeuM1 antigen expression in advanced (tumor) stage mycosis fungoides. Am J Clin Pathol 86:25±32, 1986 Wood GS, Deneau DG, Miller RA, Levy R, Hoppe RT, Warnke RA: Subtypes of cutaneous T-cell lymphoma de®ned by expression of Leu-1 and Ia. Blood 59:876±882, 1982 Yumi Y-O, Zhou X-Y, Toyo-oka K, et al: Non-CD28 costimulatory molecules present in T cell rafts induce T cell costimulation by enhancing the association of TCR with rafts. J Immunol 164:1251±1259, 2000