The B-chain of mistletoe lectin I efficiently stimulates calcium signaling in human Jurkat T-cells

Immunology Letters 78 (2001) 57 – 66 www.elsevier.com/locate/

The B-chain of mistletoe lectin I efficiently stimulates calcium signaling in human Jurkat T-cells Hermann Walzel a,*, Matthias Blach a, Peter Neels a, Ulrich Schulz b, Karin Wollenhaupt c, Josef Brock a a

Institute of Medical Biochemistry and Molecular Biology, Uni6ersity of Rostock, Schillingalle 70, D-18057, Rostock, Germany b Institute of Immunology, Uni6ersity of Rostock, Schillingalle 70, D-18057 Rostock, Germany c Unit of Reproducti6e Biology of Farm Animals, Wilhelm Stahl Allee 2, D-18196 Dummerstorf, Germany Received 17 January 2001; received in revised form 10 May 2001; accepted 14 May 2001

Abstract Mistletoe lectin I (ML I), a heterodimeric disulfide-linked type II ribosome inactivating protein, exhibits immunomodulatory potency in stimulating the cytokine release in vitro and in vivo. However, data concerning early activation events in T-cells induced by ML I and its A and B chain preceding cytokine secretion and the receptors involved are of limited availability. Here we show by flow cytometric measurements that human T-lymphoblastoid Jurkat cells express surface glycoprotein receptors for ML I. One of which is shown to be the CD2 antigen involved in a variety of T-cell signaling events. The lectin induces in Jurkat T-cells an increase of the cytosolic calcium concentration ([Ca2 + ]i) consisting of both, the transient release of Ca2 + from internal stores and a sustained influx of extracellular Ca2 + . Studies with isolated A- and B-chains provided evidence that the lectin-induced increase in [Ca2 + ]i is mediated by ML IB. The ML I and ML IB stimulated cellular calcium responses are inhibited by saccharidic competitors. In transiently transfected E6.1 cells ML IB stimulated the expression of the luciferase reporter construct pNFAT-TA-Luc that is activated through the nuclear factor of activated T-cells (NFAT). The ML IB stimulated expression of the reporter luciferase (Luc) is completely inhibited by cyclosporin A (0.2 mM) and by FK 506 at 0.05 mM. Pretreatment of Jurkat E6.1 cells with 1-deoxymannojirimycin (dMJ), an inhibitor of cis-Golgi a-mannosidase I, strongly reduced cell binding of ML IB-FITC and the ML IB induced calcium response. Benzyl-a-GalNAc, an inhibitor of O-linked glycosylation, has slightly decreasing effects in ML IB-FITC binding and was without effects on the lectin stimulated increase in [Ca2 + ]i. Inhibition of the lectin induced calcium responses by cholera toxin and by inhibitors of protein kinases as well as the absence of calcium responses in CD3− and CD45− Jurkat T-cell clones suggest that ML IB has the potency to induce early T-cell activation events. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Jurkat T-cell; Mistletoe lectin I; Calcium signaling

1. Introduction Mistletoe lectin-I from Viscum album belongs to the group of type II ribosome-inactivating proteins (RIPs)

consisting of both a toxic A-chain and a carbohydratebinding B-chain [1,2]. The structure of ML I is comprised of two non-covalently associated pairs of AB dimers that are held together by disulfide bonds. Like

Abbre6iations: Benzyl-a-GalNAc, benzyl-2-acetamido-2-deoxy-a-D-galactopyranoside; BSA, bovine serum albumin; CsA, cyclosporin A; DTT, dithiothreitol; EDTA, ethylene diaminetetraacetic acid; EGTA, ethylene glycol-bis(b-aminoethyl ether)-N,N,N%,N%-tetraacetic acid; [Ca2 + ]i, concentration of intracytoplasmic free calcium; dMJ, 1-deoxymannojirimycin; FITC, fluorescein isothiocyanate; Hepes, N-2-hydroxyethylpiperazine-N-2-ethane sulfonic acid; HRP, horseradish peroxidase; HIC, hydrophobic interaction chromatography; InsP3, inositol-1,4,5-trisphosphate; mAb, monoclonal antibody; Luc, luciferase; MES, 2-(N-morpholino) ethane sulfonic acid; ML I, mistletoe lectin I; ML IA, A-chain of ML I; ML IB, B-chain of ML I; NP 40, Nonidet P 40; NFAT, nuclear factor of activated T-cells; PAGE, polyacrylamide gel electrophoresis; PBS, phosphate-buffered saline; PMSF, phenylmethyl sulfonyl fluoride; RIP, ribosome-inactivating protein; SDS, sodium dodecyl sulfate; TBS, Tris-buffered saline; Tw, Tween 20. * Corresponding author. Tel. + 49-381-494-5759; fax: + 49-381-494-5752. 0165-2478/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 2 4 7 8 ( 0 1 ) 0 0 2 3 8 - 3

H. Walzel et al. / Immunology Letters 78 (2001) 57–66

58

other type II RIP family members ML I is a potent inhibitor of eukaryotic protein synthesis at the ribosomal level by A-chain mediated enzymatic depurination of an adenosine residue in a highly conserved loop of the 28S ribosomal RNA [3]. Because of its very efficient cell toxicity, ML IA-chains have been used for the preparation of immunotoxins [4,5], reagents to be applied for the depletion of tumor cells and immunocompetent cells in autoimmune diseases and transplantation [1]. Studies of the binding specificity revealed that ML I has a broad range of affinity for Gala,b linked sugar sequences with minor specificity for the preterminal sugar and its linkage type to Gal [6,7]. ML IB mediates toxin binding to the cell membrane and is required for the translocation of the enzymatically active A-chain into the cytosol [8]. Beside of its high cell toxicity, interest in ML I has largely increased because of the ability to activate non-specific defense mechanisms. Non-toxic subcutaneous injections of ML I stimulate cytokine secretion in serum, significant increases in NK cell number, phagocytic activity of granulocytes and the number of large granular lymphocytes [9– 12]. The activation pathways of these non-specific defense mechanisms are still unknown, but specific B-chain mediated interactions with cellular galactoglycoconjugates appear to be involved in initiating such an immune response [13]. Therefore, ML IB chain preparations offer the advantage of a controlled clinically beneficial immunomodulation in cancer treatment [9]. Accordingly, to understand the roles of ML IB in immunomodulation it is necessary to identify its interacting glycoconjugates and to study the signaling events. In this article we have identified the CD2 antigen as a recognition molecule for ML I. The lectin as a whole as well as the B-subunit stimulate in human Jurkat T-lymphocytes cell calcium signaling. Reporter gene assays provided evidence that the activation of the transcription factor NFAT is mediated solely by ML IB.

leaves by affinity chromatography on lactosyl Sepharose as previously described [5]. For the preparation of ML IA and ML IB the lectin (40 mg) was immobilized on lactosyl Sepharose 4B. ML IA was obtained by reductive elution with PBS, pH 7.4, supplemented with 5% (v/v) 2-mercaptoethanol. After concentration by ultrafiltration on PM 10 membranes (Amicon), ML IA was passed through a lactosyl Sepharose column. For the preparation of ML IB, the column was eluted with 100 mM lactose in PBS, pH 7.4, containing 5% (v/v) 2-mercaptoethanol. The eluate was concentrated and separated on a Sephadex G 100 column (2.5× 95 cm) equilibrated with PBS, pH 7.4, 10 mM lactose and 50 mM 2-mercaptoethanol. ML IB containing fractions were identified by SDS-PAGE analysis [15]. ML IB was further purified by cation exchange chromatography. After loading ML IB (5.2 mg/ml in 20 mM MESbuffer, pH 5.0) to a Resource S column, gradient elution was performed with 20 mM MES-buffer containing 0.5 M NaCl applying the FPLC-System (Pharmacia). Three fractions of ML IB were obtained by increasing the elution buffer to 17% (fraction I, 4.6 mg), to 23% (fraction II, 0.26 mg), and to 31% (fraction III, 0.22 mg) in binding buffer. ML IB (fraction I, 2.0 mg) was transferred to binding buffer (50 mM Na2HPO4, pH 7.0, 1.5 M (NH4)2SO4) and separated by hydrophobic interaction chromatography (HIC). From a HICisopropyl column (1 ml) ML IB was eluted at 31% elution buffer (50 mM Na2HPO4, pH 7.0) in binding buffer as one peak. Finally, ML IB was separated on a HiLoad 16/60 Superdex™ 75 column equilibrated with 50 mM lactose in PBS, pH 7.4. N-terminal sequence analysis of ML I B was performed with the gas-phase sequencer model 477A in combination with the PTHanalyzer model 120A (Applied Biosystems, Foster City, USA).

2. Materials and methods

Jurkat E6.1 cells (2.5× 107) were cell surface biotinylated with 0.5 mg of the membrane-impermeant derivative sulfosuccinimidobiotin in 2.0 ml PBS, pH 8.0, for 20 min at room temperature [14]. Membrane fractions were prepared by sonicating the cells in a hypotonic lysis buffer (25 mM Tris–HCl, pH 7.5, 25 mM sucrose, 0.1 mM EDTA, 5 mM MgCl2, 5 mM DTT, 1 mM PMSF, 10 mg/ml leupeptin, 10 mg/ml aprotinin), removing the nuclei by low speed centrifugation and sedimenting the membranes at 100 000× g for 60 min at 4 °C. The pellet obtained from 2×108 cells was lysed in 1 ml PBS, pH 7.4, supplemented with 1% (w/v) NP 40, 10 mg/ml aprotinin, 2 mg/ml pepstatin, 2 mg/ml antipain, 2 mg/ml leupeptin, and 1 mM PMSF on ice for 30 min. Then the lysate was centrifuged at 13 000× g for 10 min at 4 °C. The supernatant was incubated

2.1. Cells The human leukemic T-cell line Jurkat (clone E6.1, European Collection of Animal Cell Cultures, Salisbury, UK), the Jurkat T-cell clone J 45.01 (CD45−, a gift from G.A. Koretzky, University of Iowa), and the Jurkat 31-13 cell clone (CD2+CD3−, kindly given by A. Alcover, Institut Pasteur, Paris, France) were grown in RPMI 1640 medium supplemented with 10% (v/v) fetal calf serum and 10 mg/ml kanamycin.

2.2. Protein isolation ML I was separated from extracts of fresh mistletoe

2.3. Separation of Jurkat T-cell membrane lysates on ML I agarose

H. Walzel et al. / Immunology Letters 78 (2001) 57–66

59

end over end for 1 h at 4 °C with 200 ml ML I agarose prepared by coupling 7.5 mg ML I to 1 ml Affi Gel 10 according to the manufacturer’s instructions. Thereafter, the gel suspension was transferred to a 0.22 mm centrifuge filter unit (Costar) and the gel was washed ten times with membrane lysis buffer. The ML I agarose bound membrane fraction was eluted with 0.2 M lactose in membrane lysis buffer. For SDS-PAGE, the ML I agarose bound and the non-bound fraction were treated with an equal volume of double concentrated Laemmli sample buffer [15] for 5 min at 100 °C.

ml 0.1 M NaHCO3 at pH 9.0 for 30 min at room temperature. Separation of the conjugate from nonbound FITC and buffer change were performed on a Bio-Gel P6 column equilibrated with PBS, pH 7.4. For ML IB-FITC binding the cells (1× 106/ml PBS, pH 7.4) were incubated with 1 mg of the conjugate on ice for 5 min. After washing twice, ML IB-FITC binding to E6.1 cells was analyzed by a FACScan (Becton–Dickinson).

2.4. Immunoprecipitation of CD2 from cell lysates

The cells were cultured in RPMI 1640 medium supplemented with 10% (v/v) fetal calf serum and 10 mg/ml kanamycin in the presence of 4 mM benzyl-a-GalNAc for 70 h and with 2.8 mM 1-deoxymannojirimycin (dMJ) for 48 h. Viability of the cells as detected by Trypan blue exclusion was higher than 97%.

Jurkat E6.1 cells (5.1×108) were sonicated in 2.3 ml cell lysis buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 1% (w/v) NP 40, 2 mM EDTA, 1 mM PMSF, aprotinin, leupeptin, pepstatin, antipain (each at 10 mg/ml) and treated on ice for 1 h. The supernatant obtained by centrifugation at 13 000× g for 10 min at 4 °C was used for immunoprecipitation and for the separation on ML I agarose (1.8 ml corresponding to 4×108 cell equivalents). For immunoprecipitation of CD2, the supernatant (0.5 ml corresponding to 1.1× 108 cell equivalents) was incubated with 40 mg anti-CD2 mAb (clone 39C1.5 Coulter-Immunotech) for 1 h at 4 °C. After the addition of 70 ml protein G Sepharose and incubation for 1 h at 4 °C, the beads were collected by centrifugation and washed several times with cell lysis buffer. Then the beads were treated with 70 ml double concentrated electrophoresis sample buffer for 5 min at 100 °C.

2.5. Analysis of the ML I agarose-bound fraction by immunoblotting The ML I agarose-bound glycoproteins were separated by SDS-PAGE and transferred to nitrocellulose membranes. The membranes were blocked in 20 mM Tris–HCl, pH 7.2, 1 M NaCl, 1% (w/v) Tween 20. Then the blots were probed with 1 mg/ml of a biotinylated anti-CD2 mAb (clone 39C1.5) in TBS/Tw (20 mM Tris – HCl, pH 7.2, 1 M NaCl, 0.05% (w/v) Tween 20) for 16 h at 4 °C. After washing with TBS/Tw, the blots were incubated with a streptavidin-HRP conjugate (1:1000 dilution) for 1 h at room temperature. The blots were washed four times in TBS/Tw and once with 0.05 M acetate buffer, pH 5.0, for 1 min. Staining was performed with a mixture of 10 mg 3-amino-9-ethylcarbazole in 1 ml acetone, 25 ml 0.05 M acetate buffer, pH 5.0, and 15 ml H2O2 (30%).

2.6. Flow cytometric measurement of ML IB-FITC binding to E6.1 cells ML IB (300 mg) was labeled with 30 mg FITC in 250

2.7. Treatment of Jurkat E6.1 cells with inhibitors of glycosylation

2.8. Measurement of intracellular calcium Intracytoplasmic free Ca2 + -levels were measured with fura-2 AM. The cells at 2× 106/ml RPMI 1640 medium were loaded with 5 mM of fura-2 AM for 30 min at 37 °C. After addition of 3 ml RPMI 1640 medium, the cells were reincubated for 20 min at 37 °C. Then fura-2 loaded cells were washed three times with 10 mM Na-Hepes buffer, pH 7.4, supplemented with 137 mM NaCl, 5 mM KCl, 0.5 mM MgCl2, 1 mM CaCl2, 5 mM glucose, 1 mM Na2HPO4 and 1 g/l BSA and adjusted to 1× 106 cells/ml. Cell viability was higher than 95%. The fluorescence of the cellular suspension at 37 °C was monitored with a Shimadzu RF-5001 PC spectrofluorimeter by exciting alternatively at 339 and 380 nm and by measuring the fluorescence at 490 nm. Graphic representations of [Ca2 + ]i were calculated by converting the ratio of 339/ 380 to [Ca2 + ]i using a Kd of 224 nM according to Grynkiewicz [16]. Rmax and Rmin were evaluated in 1 mM Ca2 + -containing medium by lysing the cells with 0.5% (w/v) Triton X100 for Rmax, followed by the addition of an excess of EGTA for Rmin.

2.9. Transient transfection and NFAT-reporter gene assay The pNFAT-TA-Luc cis-reporter vector and the corresponding pTAL-Luc (Clontech) as a negative control are designed for monitoring the induction of NFATmediated signaling events by assaying for luciferase activity. pNFAT-TA-Luc contains three tandem copies of the NFAT-consensus sequence upstream of the minimal TA promotor. After binding of endogenous NFAT to the cis-acting enhancer element, transcription is induced and the reporter gene (Luc) is activated. Jurkat E6.1 cells (1×107/0.8 ml RPMI 1640 medium)

60

H. Walzel et al. / Immunology Letters 78 (2001) 57–66

mixed with 25 mg pNFAT-TA-Luc or with 25 mg pTAL-Luc were transferred to a 0.4 cm electroporation cuvette. Cells were electroporated applying a Bio-Rad gene pulser at 350 V, 900 mF and cultured at 5× 106/4 ml supplemented cell culture medim for 5 h at 37 °C in the presence of ML I, ML IA, ML IB, and CD2 mAbs (clone 39C1.5, clone 6F10.3). For inhibition of ML IB induced induction of the reporter, transfected cells were pretreated with cyclosporin A (CsA) at 0.2 mM or with 0.05 mM FK506 for 30 min. Cell lysis and the measurement of luciferase activity were accomplished applying the luciferase assay system (Promega).

3. Results

3.1. SDS-PAGE analysis of ML I, ML IA, and ML IB Under reducing conditions ML I shows one B-chain at 35 kDa and two A-chains around the 30 kDa molecular mass range (Fig. 1, lane 2). ML IA after re-chromatography on lactosyl Sepharose is shown in lane 4. ML IB was resolved by cation exchange chro-

Fig. 2. Fractionation of ML IB by cation exchange chromatography on a 1 ml Resource S column. ML IB (5.2 mg) was loaded on the column equilibrated with 20 mM MES-buffer, pH 5.0. Elution of the ML IB fractions (I, II, III) was performed with a gradient of 0 –0.25 M NaCl in MES-buffer (---) at a flow rate of 1 ml/min.

matography into three lactosyl Sepharose binding fractions (Fig. 2). Fraction I was further purified to homogeneity by hydrophobic interaction chromatography and by separation on a Superdex™ 75 column (Fig. 1, lane 3) and used for the following experiments. Amino-terminal sequence analysis of ML I B was consistent with the sequence data (DDVTCSASEPTVRI) published recently [17].

3.2. Analysis of ML I agarose-bound membrane glycoproteins by immunoblotting

Fig. 1. SDS-PAGE analysis of ML I, ML IA, and ML IB on a 10% separation gel under reducing conditions. ML I was analyzed after affinity chromatography on lactosyl Sepharose followed by gel filtration on Sephadex G100 (lane 2, 5.1 mg per lane) and ML IA after reductive elution of the A-subunits from lactosyl Sepharose bound ML I and rechromatography on a lactosyl Sepharose column (lane 4, 1.2 mg per lane). Lane 3 shows ML IB (fraction I from a Resource S column) after separation on a HIC-isopropyl and a Superdex™ 75 column (lane 3, 4.1 mg per lane). Protein bands were visualized by Coomassie Brillant Blue R 250 staining.

Membrane preparations from surface-biotinylated cells were lysed with the non-ionic detergent NP 40 and separated on ML I agarose. Affinity elution of the ML I agarose bound membrane fraction was performed with 0.2 M lactose. The ML I agarose non-binding as well as the binding material was separated by SDSPAGE and transferred to nitrocellulose membranes. The staining pattern generated after incubation the blots with a streptavidin-HRP conjugate is shown in Fig. 3, lanes 1 and 2. Among several ML I agarosebinding glycoproteins there is a strong band in the 50 kDa molecular mass range (lane 2) as indicated by the arrow. This band was also generated when the ML I

H. Walzel et al. / Immunology Letters 78 (2001) 57–66

61

agarose bound fraction derived from non-biotinylated cell lysates (lane 3) and CD2 immunoprecipitates of Jurkat E6.1 cell lysates (lane 4) were probed with a biotinylated CD2 mAb.

3.3. Calcium mobilization in Jurkat T-cells induced by ML I and ML IB The effects of ML I, ML IB, and ML IA on changes in [Ca2 + ]i in CD2+CD3+ Jurkat E6.1 and in CD2+ CD3− Jurkat 31-13 T-cells are shown in Fig. 4. In the presence of external Ca2 + the addition of ML I at 1.5 mg/ml induced in E6.1 cells a response of about 7509 123 nM, which then decreased to a sustained level of around 600 nM after about 4 min (Fig. 4A, a). ML IB stimulates a comparable Ca2 + -mobilization kinetics (b), but ML I A did not induce changes of the basal [Ca2 + ]i-values (c). However, in contrast to E6.1 cells ML I (d) and ML I B (not shown) have no effects on [Ca2 + ]i in CD3-deficient Jurkat 31-13 cells. Furthermore, stimulation of fura-2 loaded CD45-deficient

Fig. 4. Effects of ML I, ML IB, and ML IA on [Ca2 + ]i in CD2+ CD3+ Jurkat E6.1 cells and in CD2+CD3− 31-13 Jurkat T-cells. A: [Ca2 + ]i levels were measured in a suspension of fura-2 loaded cells in Hepes buffer containing 1 mM CaCl2. ML I at 1.5 mg/ml was added to E6.1 cells (a) and to 31-13 T-cells (d) as well as ML IB at 1.4 mg/ml (b) and ML IA at 3.3 mg/ml (c) were added to E6.1 cells as indicated by the arrow. B: [Ca2 + ]i-response of E6.1 cells induced by 1.5 mg/ml ML I (a) and by ML IB at 1.5 mg/ml (b) in a calcium free medium. EGTA (1.5 mM) was added at time 0. The addition of CaCl2 (1.5 mM) is indicated by the arrow.

Fig. 3. Separation of membrane glycoproteins on ML I agarose and immunodetection of CD2 on blots. Membrane lysates prepared from cell surface biotinylated Jurkat T cells were separated on ML I agarose. The ML I agarose non-bound fraction (lane 1, 5 ×105 cell equivalents) and the bound fraction released by 0.2 M lactose (lane 2, 2× 107 cell equivalents) were separated on a 10% SDS-PAGE gel and transferred to nitrocellulose membranes. The ML I agarose bound fraction (lane 3, 1.8 ×107 cell equivalents) and the CD2 immunoprecipitate (lane 4, 1.6 × 107 cell equivalents) separated from lysates of non-biotinylated cells were incubated with a biotinylated CD2 mAb at 1 mg/ml for 16 h at 4 °C. After incubation with a streptavidin-peroxidase conjugate (1:1000 dilution) for 1 h (lanes 1 –4), the blots were stained applying 3-amino-9-ethylcarbazole.

J45.01 Jurkat T-cells with CD2 mAbs, CD3 mAbs, and with ML I did not elicit an elevation in [Ca2 + ]i (not shown). As illustrated in Fig. 4B, ML I (a) and ML IB (b) induce in Jurkat E6.1 cells in calcium free medium a transient calcium signal with peak values of approximately twice of the basal levels originated by the release of Ca2 + from intracellular stores. Repletion of the medium with CaCl2 results in a second phase that is sustained by an influx of extracellular calcium.

3.4. Effects of protein kinase inhibitors and of cholera toxin on increases in [Ca 2 + ]i The ML I induced response in [Ca2 + ]i was sensitive to protein kinase inhibitors such as herbimycin A and staurosporine as well as to cholera toxin. Preincubation of fura-2 loaded cells with the cell-permeable broad spectrum inhibitor of protein kinases staurosporine (1

62

H. Walzel et al. / Immunology Letters 78 (2001) 57–66

mM, 3 min) completely inhibited the subsequent ML I stimulated increase in [Ca2 + ]i as shown in Fig. 5d and the responses generated by CD2 and CD3 mAbs (data not shown). For comparative studies, Jurkat E6.1 cells were cultured with the tyrosine kinase inhibitor herbimycin A or with cholera toxin under conditions that abrogate the calcium response in the presence of external calcium stimulated by CD2 and CD3 mAbs [18]. As shown in Fig. 5a, the elevation of the intracytoplasmic calcium concentration induced by ML I at 2.0 mg/ml to 940 nM, which represent a 11.5-fold increase over the basal [Ca2 + ]i levels was found to be reduced to 220 nM after herbimycin A treatment (c) and to 344 nM after preincubation the cells with cholera toxin (b).

3.5. Inhibition of ML I and ML IB induced rise in [Ca 2 + ]i by saccharidic competitors To study the structural requirements of the ML I and ML IB induced cellular response in [Ca2 + ]i the effects of different saccharidic competitors were tested. As shown in Fig. 6A, asialofetuin at 30 mM (c) was more active in inhibiting the Ca2 + -response induced by ML I than D-Gal at 30 mM (b) indicating that the lectin prefers triantennary Galb1-4GlcNAc structures [19,20] for recognition. Comparable inhibition kinetics were recorded for the ML IB stimulated increase in [Ca2 + ]i in the presence of 30 mM D-Gal (Fig. 6B, b) and 30 mM asialofetuin (c). The degree of various sugars at 30 mM, of fetuin, and of asialofetuin at 30 mM to inhibit calcium mobilization in Jurkat E6.1 cells induced by ML I and ML IB is listed in Table 1 and expressed

Fig. 5. Increase in [Ca2 + ]i induced by 2.0 mg/ml ML I in non-treated (a), cholera toxin treated (b), herbimycin A treated (c), and in staurosporine treated Jurkat E6.1 cells (d). Fura-2 loaded cells were incubated at 37 °C with 1 mM staurosporine for 3 min. The cells were cultured at 37 °C with 3 mM herbimycin A (5 × 105 cells/ml) for 21 h, with 0.5 mg/ml cholera toxin (1.5 ×106 cells/ml) for 2 h in supplemented cell culture medium, and then loaded with fura-2 AM.

Fig. 6. Inhibition kinetics of MLI and MLI B-stimulated rise in [Ca2 + ]i in Jurkat E6.1 cells in the presence of galactose and asialofetuin. (A) Increase in [Ca2 + ]i induced by ML I at 2.0 mg/ml without (a), in the presence of 30 mM galactose (b), and of asialofetuin at 30 mM (c). (B) Increase in [Ca2 + ]i stimulated by ML IB at 2.0 mg/ml without (a), in the presence of 30 mM galactose (b), and of asialofetuin at 30 mM (c). After addition of the competitors at time 0, the fura-2 loaded cells were equilibrated for 100 s at 37 °C. Then ML I and ML IB were added as indicated by the arrow.

relative to the [Ca2 + ]i-values stimulated by ML I and ML IB in the absence of a competitor. Among the competitors tested the following sequence of increasing inhibitory activity D-FucB D-GalB Galb1-4Glc B Galb1-4Fru  asialofetuin was obtained for ML I and ML IB. When compared with their inhibitory effects on ML I, D-fucose was about 2.8-fold, D-galactose 3.0fold, lactose 6.3-fold, and lactulose 6.8 times more active in inhibiting the ML IB induced Ca2 + -mobilization. D-fucose was less active than D-galactose, showing that the OH-group at C6 participates in binding. L-fucose was less active in inhibiting ML I and failed to show any inhibition of the ML IB stimulated increase in [Ca2 + ]i. Cellobiose (Glcb1-4Glc) which is identical to lactose (Galb1-4Glc) except for the equatorial orientation of the 4-hydroxyl group of the non-reducing Glc was nearly inactive. Asialofetuin containing three triantennary N-linked oligosaccharide chains with terminal Galb1-4 GlcNAc sequences (74%), a isomer with a Galb1-3 linkage in the outer Mana1-6 arm (9%), one

H. Walzel et al. / Immunology Letters 78 (2001) 57–66

63

biantennary chain with terminal Galb1-4GlcNAc (17%) [19], and three O-linked Galb1-3GalNAca-residues [20] as well as fetuin inhibited at three orders of magnitude lower concentrations (30 mM) the ML I and ML IB induced calcium response with a comparable efficiency.

3.6. Effects of glycosylation inhibitors on ML IB-FITC binding and Ca 2 + -mobilization We next examined the effects of 1-dMJ, an inhibitor of the Golgi-located mannosidase I [21], and of benzyla-GalNAc, an inhibitor of the UDP-Gal:GalNAc-b1,3-galactosyltransferase [22], on ML IB-FITC binding and Ca2 + -mobilization in Jurkat E6.1 cells. The ML IB-FITC binding as analyzed by flow cytometry is summarized in Fig. 7. The fluorescence histogram clearly shows that E6.1 cells express appropriately glycosylated cell surface glycoconjugates for ML IB recognition as shown in Fig. 7A, peak 3. In the presence of 30 mM lactose as a competitor the binding of ML IB was strongly, but not completely, inhibited (peak 2) when compared with non-labeled cells (peak 1). A slightly decreased fluorescence intensity was recorded (Fig. 7B, peak 1) after pretreatment of the cells with benzyl-a-GalNAc at concentrations that inhibit the Olinked glycosylation [22]. When the processing of mannose-rich N-linked glycans was blocked by pretreatment with 1-dMJ at 2.8 mM, cell binding of ML IB-FITC was found to be strongly reduced (Fig. 7C, peak 1). Compared with non-treated cells benzyl-aGalNAc pretreatment was without effects on the increase in [Ca2 + ]i stimulated by ML IB (data not shown). However, elevation of the intracytoplasmic calcium concentration induced by ML IB at 2 mg/ml to 595 nM (Fig. 8a) was found to be reduced to 274 nM after dMJ-pretreatment of the cells (Fig. 8b). By conTable 1 Inhibition of ML I and ML IB-induced rise of cytosolic [Ca2+]i in Jurkat E6.1 cells by different saccharides, by fetuin, and by asialofetuin Saccharides

0 mM 30 mM D-Gal 30 mM D-Fuc 30 mM L-Fuc 30 mM Galb1-4Glc 30 mM Galb1-4Fru 30 mM Glcb1-4Glc 30 mM fetuin 30 mM asialofetuin

Relative [Ca

2+

]i-inducing activity

ML I (2.0 mg/ml)

ML IB (2.3 mg/ml)

1.0 0.55 0.65 0.81 0.51 0.41 0.94 0.20 0.13

1.0 0.18 0.23 0.97 0.08 0.06 0.95 0.09 0.18

The inhibitory activity of each compound is expressed relative to the [Ca2+]i-values induced by ML I and ML IB in the absence of a competitor.

Fig. 7. Flow cytometric analysis of ML IB-FITC binding to nontreated, benzyl-a-GalNAc-, and to dMJ-pretreated Jurkat E6.1 cells. The cells were cultured at 1 ×106 in 7 ml supplemented cell culture medium with 4 mM benzyl-a-GalNAc for 70 h and at 1 ×106 cells in 4 ml with 2.8 mM 1-dMJ for 48 h at 37 °C. For ML IB-FITC binding the cells (1 × 106/ml PBS, pH 7.4) were incubated with 1 mg of the conjugate on ice for 5 min. (A) ML IB-FITC binding to non-treated E6.1 cells in the absence (peak 3) and in the presence of lactose at 30 mM (peak 2). Peak 1 represents the fluorescence intensity of non-labeled cells. (B) Fluorescence histogram from nontreated (peak 2) and benzyl-a-GalNAc treated cells (peak 1), (C) Fluorescence histogram obtained from ML IB-FITC binding to non-treated (peak 2) and to 1-dMJ pretreated E6.1 cells (peak 1).

trast with ML IB, practically identical pattern of Ca2 + mobilization were recorded after stimulation of non-treated and dMJ-pretreated E6.1 cells with a CD3 mAb (not shown).

64

H. Walzel et al. / Immunology Letters 78 (2001) 57–66

Fig. 8. [Ca2 + ]i response of non-treated (trace a) and 1-dMJ treated Jurkat E6.1 cells (trace b) following ML IB stimulation in the presence of external calcium. The cells (2 × 106) were pretreated at 37 °C with 2.8 mM 1-dMJ in 7 ml supplemented cell culture medium for 48 h. The addition of ML IB (2 mg/ml) at 100 s is indicated by the arrow.

3.7. ML IB stimulates the signal transduction pathway Ca 2 + “calcineurin“ NFAT The reporter construct pNFAT-TA-Luc driven by multiple copies of NFAT binding sites was included to test the activity of ML I, ML IA, ML IB and of CD2 mAbs for induction of transcription of the reporter gene Luc. The results are summarized in Fig. 9. Stimulation of transiently transfected E6.1 cells with 5.0 mg/ml ML IB induced a 19.3-fold increase of luciferase

activity when compared with control incubations (C). Inhibition of the calcium/calmodulin-regulated serine/ threonine phosphatase calcineurin by pretreatment of cells with the immunosuppressive drugs CsA or FK506 [23] completely blocked the ML IB induced induction of the reporter. Costimulation with two CD2 mAbs (each at 1.25 mg/ml) recognizing different epitopes increased the luciferase activity 10.9-fold. When compared with control incubations (C), ML I as well as ML IA inhibited the transcription of the reporter caused by the cell toxicity of the A-chain. Around 10-fold decreased luciferase activities relative to the control (C) were measured in cell lysates from E6.1 cells transiently transfected with the pTAL-Luc construct (negative control). ML IB failed to induce this construct (not shown).

4. Discussion The results of the present study show that ML I recognizes appropriately glycosylated cell surface glycoproteins of the human lymphoblastoid T-cell line Jurkat. Among them, the T-cell activation antigen CD2 has been identified to be a target for ML I. T-cell activation via the CD2 pathway induces the same events as those stimulated by T-cell receptor or CD3 ligands such as increase in [Ca2 + ]i [24], generation of phosphoinositol pathway-related second messengers [25], opening of voltage-insensitive calcium channels [26], and tyrosine phosphorylation of the CD3 z chain

Fig. 9. NFAT reporter gene assay. Jurkat E6.1 cells were transiently transfected with the pNFAT-TA-Luc cis-reporter vector. The cells were cultured at 5 ×106 cells/4 ml supplemented cell culture medium for 5 h at 37 °C in the presence of ML I, ML IA, ML IB, and CD2 mAbs as indicated in the figure. For CsA and FK506 treatment, transfected cells were pretreated with the drugs for 30 min at 37 °C. Luciferase activity is expressed relative to control (C).

H. Walzel et al. / Immunology Letters 78 (2001) 57–66

[27]. In Jurkat E6.1 cells the lectin induces the sustained increase in [Ca2 + ]i consisting of both an initial transient release from intracellular stores and an altered influx across the plasma membrane. Studies with isolated A- and B-chains provided evidence that the lectininduced rise in [Ca2 + ]i is B-chain mediated. Among the competitors tested the saccharidic structures available on bovine fetuin and asialofetuin [19,20] efficiently inhibited the ML I and ML IB stimulated rise in [Ca2 + ]i at the micromolar range. To get evidence whether the maturation of Asn-linked glycans is involved in the formation of recognition structures for ML IB, 1-dMJ an inhibitor of the Golgi-located mannosidase I [21], was included. The processing of the initial N-linked core glycosylation product Glc3Man9GlcNAc2-protein starts in the endoplasmic reticulum by the action of a1,2-glucosidase, a1,3-glucosidase, and a1,2-mannosidase. Glycan processing is continued in the cis-region of the Golgi by mannosidase I that removes up to four a1,2-linked mannoses [28,29] and in the medial Golgi by GlcNAc transferase and mannosidase II that hydrolyzes a1,3- and a1,6-linked mannoses from the a1,6branch and results in the generation of GlcNAcMan3GlcNAc2-protein [30]. This intermediate is important in the formation of complex oligosaccharides since GlcNAc- and galactose residues as well as sialic acid and fucose are added by a variety of glycosyltransferases in the trans Golgi network [31]. When the cellular processing of the mannose-rich N-linked glycans is blocked by 1-dMJ ML IB-FITC binding to Jurkat E6.1 cells and the increase in [Ca2 + ]i stimulated by ML IB were strongly reduced. In contrast to ML IB, no inhibitory effects on the increase in [Ca2 + ]i were recorded after stimulation of 1-dMJ pretreated cells with a CD3 mAb (clone UCHT1) that reacts with a peptide epitope on CD3 o [32]. From these experiments we conclude that in Jurkat E6.1 cells the processing of Asn-linked glycans is an important event in the formation of recognition structures involved in ML IB induced signaling. Benzyl-a-GalNAc that inhibits the formation of fully glycosylated mucin in a mucin-producing colon cancer cell line caused by inhibition of the UDP-Gal:GalNAc-b1,3-galactosyltransferase [22] has only slight decreasing effects on ML IB-FITC binding to Jurkat E6.1 cells. Pretreatment of the cells with benzyl-GalNAc at 4 mM was without effects on the increase in [Ca2 + ]i stimulated by ML IB. To determine differences between cell calcium mobilization stimulated by ML I and that elicited by CD2 and CD3 mAbs Jurkat E6.1 cells were treated with the protein kinase inhibitors staurosporine and herbimycin A as well as with cholera toxin under conditions that abrogated the mAb-stimulated Ca2 + -response [18]. Pretreatment of the cells with staurosporine also abrogated the cellular calcium response induced by ML I whereas herbimycin A pretreatment decreased the lectin-medi-

65

ated increase in [Ca2 + ]i to about 20% when compared with non-treated cells. T-cell activation via CD2 requires the presence of CD3-z [33]. Inhibition by cholera toxin of GTP-binding to CD3-z uncouples the receptor from further signaling steps [34]. Pretreatment of E6.1 cells with cholera toxin reduced the ML I stimulated Ca2 + -signal to about 36% when compared with nontreated cells. Further evidence for the involvement of CD3 in ML I stimulated elevation in [Ca2 + ]i was provided by applying a CD3-deficient Jurkat T-cell clone. In CD2+ CD3− Jurkat 31-13 T-cells ML I was inefficient to induce an elevation in [Ca2 + ]i. Although the CD2 cytoplasmic domain has no intrinsic protein tyrosine kinase activity and no tyrosine residues, which might serve as docking sites for SH2 domains upon phosphorylation, stimulation via CD2 induces the tyrosine phosphorylation of several intracellular proteins. Coupling to SH3 domains of non-receptor protein tyrosine kinases of the Src family is achieved by prolinerich sequences of the CD2 tail [35]. By recognition of galactosidic structures on T-cell activation markers galectin-1 [18,36] and galectin-3 [37] trigger the increase in cytoplasmic free Ca2 + . Calcium is one of the most versatile second messengers involved in cell growth, differentiation, and apoptosis [38]. Therefore, the question becomes interesting how the cell decides on the molecular level between the proliferative or the apoptotic pathway. There is evidence that upon T-cell stimulation the InsP3-receptor 1 (InsP3R1) becomes specifically activated by tyrosine phosphorylation resulting in InsP3-mediated calcium release from endoplasmic reticulum, which finally leads to IL-2 production and proliferation [39]. T-cells lacking InsP3R1 due to the presence of antisense RNA fail to increase [Ca2 + ]i or to synthesize IL-2 after T-cell receptor stimulation [40]. InsP3R1 deficient T-cells were resistant to apoptosis by dexamethasone, T-cell receptor stimulation, ionizing radiation, and Fas [41]. However, in lymphocytes undergoing apoptosis in response to dexamethasone, the synthesis of InsP3R3 was specifically induced and blocking of InsP3R3 expression with antisense oligonucleotides inhibited the induction of apoptosis [42]. These data indicate that InsP3-mediated Ca2 + -release is involved whether stimulation leads to proliferation or apoptosis. In contrast to galectin-3 that stimulates the uptake of extracellular calcium in Jurkat T-lymphocytes [37], galectin-1 elevates [Ca2 + ]i by depletion of intracellular stores and by uptake of Ca2 + from the extracellular space [18]. May be that receptor crosslinking by galectin-1 or ML IB having two carbohydrate recognition domains [7,43] is required for the activation of both phases involved in the increase in [Ca2 + ]i. It has been shown that galectin-1 stimulates apoptosis in thymocytes and activated T-cells [44,45]. Therefore, it is conceivable that ML IB has the potency to modulate the immune response by depletion of acti-

66

H. Walzel et al. / Immunology Letters 78 (2001) 57–66

vated T-cells. On the other hand, the biphasic rise in [Ca2 + ]i stimulates a sustained activation of the calcium/ calmodulin-dependent calcineurin phosphatase [23,46]. After dephosphorylation of cytoplasmic NFATp by calcineurin, NFAT translocates to the nucleus, interacts with the transcription factor AP-1 [47] and induces the expression of several cytokine genes [48,49]. We provided evidence by NFAT reporter gene assays that ML IB stimulates the Ca2 + -dependent signal transduction pathway leading to induction of transcription and activation of the reporter gene by endogenous NFAT. Inhibition of calcineurin with cyclosporin A or FK506 abrogated the ML IB stimulated expression of the reporter.

References [1] L. Barbieri, M.G. Batelli, F. Stirpe, Biochim. Biophys. Acta 1154 (1993) 237 – 282. [2] H.-J. Gabius, H. Walzel, S.S. Joshi, J. Kruip, S. Kojima, V. Gerke, H. Kratzin, S. Gabius, Anticancer Res. 12 (1992) 669 – 676. [3] Y. Endo, K. Tsurugi, H. Franz, FEBS Lett. 231 (1988) 378 – 380. [4] A. Wiedlocha, K. Sandvig, H. Walzel, C. Radzikowsky, S. Olsnes, Cancer Res. 51 (1991) 916 –920. [5] H. Walzel, L. Jonas, J. Brock, A. Wiedlocha, C. Radzikowsky, in: E. van Driessche, H. Franz, S. Beeckmans, U. Pfu¨ ller, A. Kallikorm, T.C. Bøg-Hansen (Eds.), Lectins: Biology, Biochemistry, Clinical Biochemistry, vol. 8, Textop, Hellerup, Denmark, 1993, pp. 55 – 63. [6] R.T. Lee, H.-J. Gabius, Y.C. Lee, J. Biol. Chem. 267 (1992) 23722 – 23727. [7] A.M. Wu, L.-K. Chin, H. Franz, U. Pfu¨ ller, A. Herp, Biochim. Biophys. Acta 1117 (1992) 232 –234. [8] R.J. Joule, D.M. Neville, J. Biol. Chem. 257 (1982) 1598 – 1601. [9] J. Beuth, H.L. Ko, H.-J. Gabius, H. Burrichter, K. Oette, G. Pulverer, Clin. Invest. 70 (1992) 658 –661. [10] T. Hajto, K. Hostanska, H.-J. Gabius, Cancer Res. 49 (1989) 4803 – 4808. [11] T. Hajto, K. Hostanska, K. Frei, C. Rohrdorf, H.-J. Gabius, Cancer Res. 50 (1990) 3322 –3326. [12] J. Beuth, H.L. Ko, L. Tunggal, G. Buss, J. Jeljaszewicz, G. Pulverer, Arzneim.-Forsch. Drug Res. 44 (1994) 1255 –1258. [13] J. Beuth, Anti-Cancer Drugs 8 (1997) S53 – S55. [14] J.G. Altin, E.B. Pagler, Anal. Biochem. 224 (1995) 382 – 389. [15] U.K. Laemmli, Nature (London) 227 (1970) 680 – 685. [16] G. Grynkiewicz, M. Poenie, R.Y. Tsien, J. Biol. Chem. 260 (1985) 3440 – 3450. [17] M.H. Soler, S. Stoewa, W. Voelter, Biochem. Biophys. Res. Commun. 246 (1998) 596 –601. [18] H. Walzel, M. Blach, J. Hirabayashi, K.-I. Kasai, J. Brock, Glycobiology 10 (2000) 131 –140.

.

[19] E.D. Green, G. Adelt, J.U. Baenziger, S. Wilson, H. Van Halbeek, J. Biol. Chem. 263 (1988) 18253 – 18268. [20] B. Nilsson, N.E. Norden, S. Svensson, J. Biol. Chem. 254 (1979) 4545 – 4553. [21] H.Y. Naim, G. Joberty, M. Alfalah, R. Jacob, J. Biol. Chem. 274 (1999) 17961 – 17967. [22] S.-F. Kuan, J.C. Byrd, C. Basbaum, Y.S. Kim, J. Biol. Chem. 264 (1989) 19271 – 19277. [23] S. Wesselborg, D.A. Fruman, J.K. Sagoo, B.E. Bierer, J. Burakoff, J. Biol. Chem. 271 (1996) 1274 – 1277. [24] A. Alcover, M.J. Weiss, J.F. Daley, E.L. Reinherz, Proc. Natl. Acad. Sci. U.S.A. 83 (1986) 2614 – 2618. [25] G. Pantaleo, D. Olive, A. Poggi, W.J. Kozumbo, L. Moretta, A. Moretta, Eur. J. Immunol. 17 (1987) 55 – 60. [26] P. Gardner, A. Alcover, M. Kuno, P. Moingeon, C.M. Weyand, J. Goronzy, E.L. Reinherz, J. Biol. Chem. 264 (1989) 1068 – 1076. [27] E. Monostori, D. Desai, M.H. Brown, D.A. Cantrell, M.J. Crumpton, J. Immunol. 144 (1990) 1010 – 1014. [28] E. Bause, E. Bieberich, A. Rolfs, C. Vo¨ lker, B. Schmidt, Eur. J. Biochem. 217 (1993) 535 – 540. [29] A. Lal, J.S. Schutzbach, W.T. Forsee, P.J. Neame, K.W. Moremen, J. Biol. Chem. 269 (1994) 9872 – 9881. [30] K.W. Moremen, O. Touster, J. Biol. Chem. 260 (1985) 6654 – 6662. [31] P. Stanley, Glycobiology 2 (1992) 99 – 107. [32] I. Hannet, F. Erkeller-Yuksel, P. Lydyard, V. Deneys, M. DeBruyere, Immunol. Today 13 (1992) 215 – 218. [33] P. Moingeon, J.L. Lucich, D.J. McConkey, F. Letourneur, B. Malissen, J. Kochan, H.-C. Chang, H.R. Rodewald, E.L. Reinherz, Proc. Natl. Acad. Sci. U.S.A. 89 (1992) 1492 – 1496. [34] G.F. Bo¨ l, B.M. Haack, K. Resch, NATO ASI Ser. H76 (1993) 343. [35] T. Pawson, G.D. Gish, Cell 71 (1992) 359 – 362. [36] H. Walzel, J. Hirabayashi, K.-I. Kasai, J. Brock, P. Neels, Glycoconjugate J. 13 (1996) 99 – 105. [37] S. Dong, R.C. Hughes, FEBS Lett. 395 (1996) 165 – 169. [38] P. Nicotera, S. Orrenius, Cell Calcium 23 (1998) 173 – 180. [39] T. Jayaraman, K. Ondrias, E. Ondriasova, A.R. Marks, Nature 272 (1996) 1492 – 1494. [40] T. Jayaraman, E. Ondriasova, K. Ondrias, D.J. Harnick, A.R. Marks, Proc. Natl. Acad. Sci. U.S.A. 92 (1995) 6007 – 6011. [41] T. Jayaraman, A.R. Marks, Mol. Cell. Biol. 17 (1997) 3005 – 3012. [42] A.A. Khan, M.J. Soloski, A.H. Sharp, G. Schilling, S.-H. Sabatini, C. Li, A. Ross, S.H. Snyder, Science 273 (1996) 503 –507. [43] J. Hirabayashi, Trends Glycosci. Glycotechnol. 5 (1993) 251 – 270. [44] N.L. Perillo, K.E. Pace, J.J. Seilhamer, L.G. Baum, Nature 378 (1995) 736 – 739. [45] H. Walzel, U. Schulz, P. Neels, J. Brock, Immunol. Lett. 67 (1999) 193 – 202. [46] D.A. Fruman, C.B. Klee, B.E. Bierer, S.J. Burakoff, Proc. Natl. Acad. Sci. U.S.A. 89 (1992) 3686 – 3690. [47] J. Jain, Z. Miner, A. Rao, J. Immunol. 151 (1993) 843 –848. [48] E.S. Masuda, R. Imamura, Y. Amasaki, K. Arai, N. Arai, Cell. Signal. 10 (1998) 599 – 611. [49] A. Rao, C. Luo, P.G. Hogan, Annu. Rev. Immunol. 15 (1997) 707 – 747.