In Thymocyte Maturation Developing

8B4/20, A Private CD43 Epitope on Developing Human Thymocytes, Is Involved in Thymocyte Maturation

This information is current as Marina Fabbi, Jens Geginat, Micaela Tiso, Dunia Ramarli, of September 26, 2021. David Parent, Antonio Bargellesi and Eileen Remold-O'Donnell J Immunol 1999; 163:5964-5970; ; http://www.jimmunol.org/content/163/11/5964 Downloaded from

References This article cites 51 articles, 30 of which you can access for free at: http://www.jimmunol.org/content/163/11/5964.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 26, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1999 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. 8B4/20, A Private CD43 Epitope on Developing Human Thymocytes, Is Involved in Thymocyte Maturation1

Marina Fabbi,2* Jens Geginat,† Micaela Tiso,‡ Dunia Ramarli,§ David Parent,¶ Antonio Bargellesi,‡ and Eileen Remold-O’Donnell¶

The 8B4/20 Ag is a 120-kDa molecule whose expression on human thymocytes varies according to the differentiation stage: high density on immature CD3؊/low thymocytes, reduced density on CD3medium and double-positive thymocytes, and absent on CD3high and single-positive thymocytes and on circulating T cells. In this paper we present immunological and biochemical evidence demonstrating that 8B4/20 Ag is a variant of CD43. We show that 8B4/20-expressing molecules, which are a subset of the CD43 molecules on thymocytes, are heterogeneous in charge, suggesting varying sialylation levels. The 8B4/20 epitope was mapped to the peripherally exposed N-terminal region of CD43, and the 8B4/20 antigenic determinant was characterized by requirement for the sialic acid exocyclic polyhydroxyl side chain, a feature shared with ligands of CD22. Altogether, 8B4/20-CD43 expression Downloaded from pattern and biochemical characteristics suggest its participation in carbohydrate-based interactions in the thymus. We therefore used speciﬁc Ab to mimic putative 8B4/20 interactions with natural ligand and examined the effect on isolated thymocytes. Treatment with 8B4/20 had no effect on in vitro apoptosis of isolated thymocytes. In contrast, 8B4/20 ligation enhanced the conversion of isolated thymocytes to differentiated phenotypes. Increased numbers were found in 8B4/20-treated cultures of CD3high and single-positive thymocytes and decreased numbers of CD3؊/low and double-positive thymocytes, strongly suggesting that engagement of 8B4/20 delivers a positive signal that favors completion of the thymocyte maturation program. The ability of http://www.jimmunol.org/ 8B4/20 mAb to drive thymocyte maturation in vitro suggests that CD43 molecules bearing the 8B4/20 epitope participate in early events of thymic selection. The Journal of Immunology, 1999, 163: 5964–5970.

n previous studies of thymic development, we produced a O-glycosylated (ϳ80 O-glycans) (2, 5, 6) and sialylated extracel- murine mAb 8B4/20 that detects an Ag expressed on human lular region (7–9), suggesting an unfolded mucin structure as I thymocytes and absent on most leukocytes except NK cells shown by transmission electron microscopy (10). The large ex- and a minor fraction of monocytes (1). The 8B4/20 Ag was char- panded structure of CD43 (45 nm in length) (10) together with its acterized as a 120-kDa molecule whose expression level on thy- high copy number (ϳ150,000 molecules on T lymphoid cells) (5) by guest on September 26, 2021 mocytes varies inversely with the degree of differentiation of the and the negative charge of multiple sialic acid residues provide a cell. 8B4/20 is expressed after pro-T cells enter the thymus with repulsive barrier to the cell surface with potential to interfere with the highest levels on early differentiation stages including cells that receptor interactions on opposing cells. Indeed, the role of CD43 have not yet acquired TCR (triple-negative thymocytes). Surface as an anti-adhesion molecule was demonstrated by transfection density decreases with increasing maturity of the thymocytes, and and gene-targeting studies, which showed that CD43 prevents cell: the Ag is lost before the cells leave the thymus. 8B4/20 Ag is cell and cell:matrix interactions involving diverse receptor-ligand high absent on the CD3 thymocytes, which have undergone thymic pairs (11–13). CD43 functions, in addition, as a positive adhesion selection, and is absent on circulating T lymphocytes (1). receptor in select systems (including thymocytes) via restricted In this study, we examined the relationship of 8B4/20 to CD43 epitopes (e.g., Ref. 14) (see Discussion). Cross-linking of CD43 ϳ (sialophorin/leukosialin). CD43 is also an 120-kDa Ag on thy- with specific mAbs delivers activation signals that vary depending mocytes (2, 3), but unlike 8B4/20, CD43 is broadly expressed on on the cell type including inducing proliferation and activation of most leukocytes (4). It is a transmembrane protein with a highly T cells (e.g., Refs. 15–19). In humans, a single copy gene encodes the single CD43 polypeptide, and all molecular variation results from posttransla- * Istituto Nazionale per la Ricerca sul Cancro, Genova, Italy; †FB Biologie Chemie Pharmatie, FU Berlin, Berlin, Germany; ‡Dipartimento Medicinia Sperimentale, Se- tional modifications (8, 9). Early studies showed variation of gly- zione Biochimica, Universita`di Genova, Genova, Italy; §Azienda Ospedaliera e Uni- cosylation resulting in different m.w. isoforms in different cells (3, versita`di Verona, Policlinico Borgo Roma, Verona, Italy; and ¶The Center for Blood ϳ Research, Harvard Medical School, Boston, MA 02115 6, 20). A well-characterized mAb, T305 (21), reacts with 135- kDa CD43 on activated T cells and neutrophils and not with ϳ115- Received for publication March 22, 1999. Accepted for publication September 21, 1999. kDa CD43 on resting T lymphocytes (6, 22, 23). The costs of publication of this article were defrayed in part by the payment of page In this study, we present immunological and biochemical evi- charges. This article must therefore be hereby marked advertisement in accordance dence that 8B4/20 Ag is a CD43 variant and that 8B4/20-express- with 18 U.S.C. Section 1734 solely to indicate this fact. ing CD43 molecules constitute a subset of the CD43 molecules in 1 This work was supported by grants from Consiglio Nazionale delle Ricerche (stra- thymocytes. We also generate mapping data and chemical charac- tegic project: Cell Cycle and Apoptosis), and Ministero dell’Universitaé della Ricerca Scientifica e Tecnologica (project: Regulatory Signals in Lymphocyte Life and Death) teristics for the epitope recognized by 8B4/20 mAb. In addition, (to A.B.), and National Institutes of Health Grants AI29880 and AI39574 (to E.R.-O.). the presence of the 8B4/20 mAb in overnight cultures is shown to J.G. was recipient of a Deutscher Akademischer Austauschdienst fellowship. enhance the conversion of isolated thymocytes to differentiated 2 Address correspondence and reprint requests to Dr. Marina Fabbi, Centro Biotec- nologie Avanzate, Largo R. Benzi 10, 16132 Genova, Italy. E-mail address: phenotypes, suggesting that CD43 molecules expressing the [email protected] 8B4/20 epitope play a role in thymocyte maturation.

Materials and Methods pelleting in cold PBS. The cells were incubated in freshly prepared 20 mM ϳ Cells sodium borohydride in PBS for 15 min at 22°C, washed twice with cold PBS, and prepared for flow cytometry. Human T lymphoblastoid tumor cell lines CEM (24) and MOLT-4 (25) were cultured in RPMI 1640 with 10% heat-inactivated FCS and antibi- Flow cytometry otics. CEM cells from the American Type Culture Collection (Manassas, Cells (106) were incubated with mAb (0.5% ascites, 30% culture superna- VA) were used in Fig. 3 and Table I. CD43-deficient homologous recom- tant) in 200 ␮l of binding medium (phenol red-free HBSS with 4% FCS) binant CEM cells (12) and corresponding wild-type CEM subline A3.01 at 4°C for 60 min, washed in cold-binding medium, incubated with 1 ␮g cells, provided by Dr. B. Ardman (New England Medical Center, Boston, Ј ␮ of FITC-labeled F(ab )2 of goat anti-mouse IgG or IgM in 100 l of bind- MA), were studied in Figs. 2 and 4. HeLa cells stably transfected with ing medium at 4°C for 30 min, washed in cold PBS, and fixed in 1% CD43 and CD43-negative wild-type HeLa cells were as described (11). formaldehyde in PBS. When a direct staining protocol was used, PE- and Thymocytes, isolated by teasing thymic tissue obtained from children FITC-labeled mAb to CD3, CD4, CD8, and isotype-matched controls were undergoing corrective cardiac surgery, were purified by density centrifu- diluted in binding medium according to the manufacturer’s instructions and gation on Ficoll-Hypaque (Pharmacia Biotech, Uppsala, Sweden). incubated at 4°C for 30 min. Stained cells were washed and fixed and analyzed on a FACScan or FACStar (Becton Dickinson). Nonviable cells Antibodies were excluded by gating on light scatter profiles. A total of 5000 events Murine anti-8B4/20 mAb is an IgM (1). IgG1 isotype anti-CD43 mAbs were acquired for each sample. L10 (26) and T305 (21), provided by Dr. R. I. Fox (Scripps Clinic, La Jolla, Isoelectrofocusing CA), react, respectively, with a sialic acid-independent epitope within the distal (N-terminal) region and a sialic acid-dependent epitope proximal to Reduced samples were isoelectrofocused as described (30) at 300 V for the lipid bilayer (23, 27). Isotype-matched control mAb were from Sigma 17hand500Vfor2hincylindrical gels containing pH 3.5–10, pH 5–7,

(St. Louis, MO). Peroxidase-labeled streptavidin and unlabeled, peroxi- pH 6–8, and pH 2.5–4 ampholytes (Pharmacia Biotech) at a ratio of 10: Downloaded from dase-, PE-, and FITC-labeled goat anti-mouse-IgG and anti-mouse IgM 1:1:1.5. The focused samples were separated in a second dimension on 8% Abs were from Southern Biotechnology Associates (Birmingham, AL). polyacrylamide-SDS gels under reducing conditions, and the CD43 region PE- and FITC-labeled Abs to CD3, CD4, and CD8 Ags as well as isotype- was analyzed by immunoblot. matched controls were from Becton Dickinson (San Jose, CA). Treatment of thymocytes and evaluation of apoptosis Immunoblotting Thymocytes (5 ϫ 105/ml) were incubated with 5 ␮g/ml 8B4/20 mAb, IgM Cell extracts were fractionated by SDS-electrophoresis (28) on 8% poly- control mAb, or 10 ␮M dexamethasone (Calbiochem) for 15 h at 37°C in http://www.jimmunol.org/ acrylamide gels under reducing conditions. The separated polypeptides RPMI 1640 with 10% heat-inactivated FCS and antibiotics, after which the were transferred electrophoretically to nitrocellulose (Hybond-C Extra, cells were harvested, washed in PBS, counted, and processed for surface Amersham, Little Chalfont, U.K.). Protein Ags were detected by incubat- Ag expression (flow cytometry) and evaluation of apoptosis. ing the membrane with mAb (ascites at 1:1000 or purified Ab at 5 ␮g/ml) Apoptosis was evaluated by binding of FITC-annexin V (Bender Sys- in 10 mM Tris-HCl buffer (pH 8.0), 150 mM NaCl, and 0.01% Tween 20 tem, Boehringer Ingelheim, Heidelberg, Germany) according to the man- with 2% w/v low fat milk solids), followed by peroxidase-conjugated goat ufacturer’s instructions and by TUNEL labeling. For the latter (31), thy- Abs recognizing murine IgG or IgM. When proteins were labeled with mocytes (107/ml) were incubated in 1% paraformaldehyde in PBS for 15 biotin (described below), the nitrocellulose membranes were incubated min at 4°C with gentle shaking, washed twice in PBS, and permeabilized with peroxidase-labeled streptavidin. Peroxidase was revealed by enhanced by 70% ethyl alcohol precooled at Ϫ20°C. The cells were held at Ϫ20°C chemiluminescence (ECL, Amersham) detected by autoradiography (Hy- for 1 day, rehydrated in PBS, and resuspended at 2 ϫ 106 to 50 ␮lin45 perfilm, Amersham). ␮l of dUTP-FITC containing reaction buffer and 5 ␮l of TdT solution by guest on September 26, 2021 (reagents from Boehringer Mannheim). After1hat37°C, the cells were Surface labeling and immunoprecipitation washed three times with PBS and incubated for 30 min at ϳ22°C in the dark with 0.5 ml of DNA staining solution (5 ␮g/ml propidium iodide and CEM cells (107) in 1 ml of PBS with 1 mM CaCl , 1 mM MgCl , and 10 2 2 100 ␮g/ml RNase in PBS) and analyzed by a FACStar flow cytometer. mM glucose were combined as described (29) with sulfosuccinimidyl 6-(biotinamido)hexanoate (Pierce, Rockford, IL) (50 ␮l of 25 mg/ml in Results DMSO). After 5 min at ϳ22°C, an additional 50 ␮l out, s.c. of biotin solution was added for 10 min. The cells were washed in PBS and lysed Immunological relatedness of 8B4/20 Ag and CD43 with 1% Nonidet P-40 in PBS with 1 mM EDTA, 1 mM EGTA, 1 mM The putative relationship of 8B4/20 Ag to CD43 was investigated NaF, 1 mM PMSF, and 5 ␮g/ml leupeptin. The lysates were clarified by centrifugation at 500 ϫ g for 5 min and precleared sequentially with Pan- by using L10 mAb, which reacts with various human CD43 iso- sorbin (Calbiochem, La Jolla, CA) and isotype-matched control mAb forms (3, 26). Thymocytes were double stained with 8B4/20 and bound to protein A-Sepharose (Pharmacia Biotech) by means of goat anti- L10 mAbs and analyzed by flow cytometry. The 8B4/20ϩ popu- mouse IgG or IgM Abs. The precleared lysates were incubated overnight lation, although extremely variable in size among individuals at 4°C with 8B4/20 or L10 mAb absorbed onto protein A-Sepharose by (mean 52 Ϯ 20%, range 19–77% of positive cells; nϭ6), was means of goat anti-mouse IgM or IgG, respectively. Immunoprecipitates were washed with lysis buffer alternating with lysis buffer with 0.5 M NaCl contained completely within the L10 (CD43) positive population and were extracted for electrophoresis at 100°C for 5 min with SDS- and (Fig. 1A). Isolated thymocytes were then lysed and immunopre- mercaptoethanol-containing buffer (28). cipitated with 8B4/20 and L10 mAbs and the immunoprecipitates Enzyme treatments examined by immunoblotting. L10 was found to stain a 120-kDa band in L10 precipitates and 8B4/20 precipitates (Fig. 1B, left). Crude extracts were incubated for 16 h at 37°C with 1 IU/ml of sialidase The 8B4/20 mAb also stained a 120-kDa thymocyte band in L10 (Vibrio cholerae) (Boehringer Mannheim, Indianapolis, IN). CEM cells at 107/ml in Ca2ϩ/Mg2ϩ-free HBSS were incubated with 0.02 IU/ml of siali- precipitates and in 8B4/20 precipitates (Fig. 1B, right). Similar dase (neuraminidase) (V. cholerae, 20 IU/mg protein) (Calbiochem) for 30 results were obtained when cells of the T lymphoblastoid line min at ϳ22°C. MOLT-4 or CEM cells at 1.5 ϫ 107/ml in Ca2ϩ/Mg2ϩ-free CEM were examined (data not shown). The most straightforward HBSS were incubated with 25 ␮g/ml neutrophil elastase (Elastin Products, explanation for these findings is that 8B4/20 Ag and CD43 are Owensville, MO) at 37°C for 10 min. Cell viability was monitored by precipitated because they are immunologically related molecules; trypan blue exclusion. Immunoprecipitates from CEM cells were incubated for 16 h at 37°C with 1 IU/ml of sialidase (V. cholerae) (Boehringer Mann- however, an alternative explanation, namely that 8B4/20 Ag and heim) in 0.05 M sodium acetate (pH 5.5), 0.15 M NaCl, and 9 mM CaCl2 CD43 are unrelated coprecipitating molecules, could not be with 25 mg/ml human serum albumin or with 10 IU/ml of endoglycosi- excluded. dase-F (Calbiochem) in 0.1 M sodium phosphate buffer (pH 6.1), 50 mM Nonhematopoietic cells transfected with CD43 cDNA and lym- EDTA, and 1% Nonidet P-40. For periodate oxidation and borohydride reduction, CEM cells were washed in cold PBS, resuspended at 107/ml in phoid cells depleted of CD43 by homologous recombination were freshly prepared 2 mM sodium periodate in PBS, and incubated on ice in also used to compare 8B4/20 and CD43. Flow cytometry showed the dark for 15 min. The reaction was stopped by two cycles of dilution and that HeLa cells are negative for L10 and 8B4/20 Ag as expected 5966 A CD43 VARIANT PROMOTES MATURATION OF ISOLATED THYMOCYTES IN VITRO

FIGURE 1. Comparison of 8B4/20 Ag and CD43 by ﬂow cytometry, immunoprecipitation, and immunoblotting. A, Flow cytometric proﬁle of thymocytes double stained with L10 and 8B4/20 mAb. Numbers indicate the percentage of cells in each quadrant. B, Human thymocyte lysates were precipitated with 8B4/20 or L10 mAb, and the precipitates were analyzed by immunoblot. The left four lanes show L10 blots, and the right four lanes show 8B4/20 blots of precipitates of IgG1 isotype control, L10, IgM isotype control, and 8B4/20. Molecular weight marker positions are indicated on the left. Downloaded from and that HeLa cells transfected with CD43 were positive for L10 but failed to stain with 8B4/20 mAb (data not shown). Although lack of reactivity of 8B4/20 mAb with transfected HeLa cells could indicate that 8B4/20 Ag is not a CD43 molecule, an alter-

native explanation is that HeLa cells lack posttranslational en- http://www.jimmunol.org/ zymes required to generate the 8B4/20 epitope. On examination of ϩ CEM cells (CD43 ) and CEM cells depleted of CD43, L10 and FIGURE 2. 8B4/20 mAb does not bind to CD43-depleted CEM cells. A, 8B4/20 mAbs both stained the wild-type cells and both failed to The dark flow cytometric profiles show CD43-expressing CEM cells (top stain the cells from which the CD43 gene had been deleted (Fig. panels) and CD43Ϫ/Ϫ homologous recombinant CEM cells (bottom) 2), consistent with 8B4/20 being a CD43 molecule. stained with L10 (left) or 8B4/20 mAb (right). Isotype control-stained cells are indicated by white profiles. B, CD43-expressing CEM cells and Chemical characteristics of 8B4/20 Ag CD43Ϫ/Ϫ CEM cells lysates were subjected to immunoblotting with L10 (left) and 8B4/20 mAb (right). The 8B4/20 blot was overexposed to show CD43, although heavily O-glycosylated, has only one N-linked Ϫ Ϫ the lack of 120-kDa signal in the CD43 / CEM cell lysates. glycan and, when treated with the N-glycan cleaving enzyme en- by guest on September 26, 2021 doglycosidase-F (32), its SDS-PAGE mobility is not altered (33). When immunoprecipitated biotin-labeled 8B4/20 molecules from alter charge properties of the molecules. To characterize charge CEM cells were treated with endoglycosidase-F, SDS-PAGE mobil- properties, 8B4/20 and L10 precipitates from thymocyte lysates ity remained unaltered (Fig. 3), suggesting that 8B4/20 is not heavily were fractionated by isoelectrofocusing and the Ag detected by N-glycosylated, consistent with its being a CD43 molecule. immunoblotting with L10. The isoelectrofocusing profiles, shown A more diagnostic biochemical feature of CD43 is the large decrease of SDS-PAGE mobility (increase of apparent m.w.) that results from enzymatic desialylation (26). Whereas mock-treated immunoprecipitated 8B4/20 Ag migrated as expected at 120 kDa, sialidase-treated 8B4/20 immunoprecipitates migrated at an apparent molecular mass of 150 kDa, showing the identical mobility of L10 Ag (Fig. 3). The large decrement in its SDS-PAGE mobility following desialylation strongly suggests that 8B4/20 Ag is a CD43-like molecule. In addition, no residual biotinylated protein remained at 120 kDa in the sialidase-treated 8B4/20 precipitates (Fig. 3), indicating that the CD43-like molecule is the sole surface molecule in 8B4/20 immunoprecipitates. Cumulatively these findings indicate that 8B4/20-reactive molecules are present in L10 precipitates and L10-reactive molecules in 8B4/20 precipitates (Fig. 1B) because the molecules are immunologically related. Altogether, the com- bination of immunological relatedness and shared biochemical fea- tures demonstrate that 8B4/20 Ag is a CD43 molecule. FIGURE 3. 8B4/20 Ag characterized by enzyme treatments. L10 and 8B4/20 immunoprecipitates from biotin-labeled CEM cells were mock- 8B4/20 mAb detects a subset of thymocyte CD43 molecules treated or treated with endoglycosidase-F or sialidase. The precipitates were separated by SDS-PAGE and transferred to nitrocellulose, and biotin- The broad distribution of CD43 and restricted distribution of labeled Ag was detected with streptavidin reagent. A positive control for 8B4/20 Ag suggest that the latter is an epitope expressed on a endoglycosidase-F activity was provided by the altered mobility of endog- subset of CD43 molecules. The extent of sialylation of thymocyte enous IgG or IgM heavy chains (lower panel). Note that the mobility of surface molecules increases dramatically during trafficking and both L10 and 8B4/20 Ags is unaffected by endoglycosidase-F and substan- maturation in the thymus (34, 35), and this variation is expected to tially decreased by sialidase. The Journal of Immunology 5967

Table I. Effect of cell treatments on CD43 epitopes

Epitope Expression (linear ﬂuorescence values)a

Treatment L10 8B4/20

Mock 914 649 Sialidaseb 950 41

Mockc 823 552 Periodated 749 55

a Fluorescence values are expressed as mean linear ﬂuorescence units. b Cells were treated with 0.02 IU sialidase/ml in PBS. c Cells were treated with PBS followed by incubation with 20 mM sodium borohydride. d Cells were treated with 2 mM sodium periodate at 4°C followed by sodium borohydride. One experiment representative of three.

dase-sensitivity of the 8B4/20 epitope was veriﬁed also by ﬂow cy-

tometry, which showed loss of the 8B4/20 epitope in sialidase-treated Downloaded from CEM cells and, as a control, retention of L10 staining (Table I). FIGURE 4. 8B4/20 mAb precipitates a subset of CD43 molecules of CEM cells were also subjected to mild periodate treatment. thymocytes. The 8B4/20 and L10 precipitates from thymocyte lysates of Whereas sialidase, which removes entire sialic acid moieties, alters three individuals (#1 to #3) were fractionated by isoelectrofocusing and structural and charge properties of molecules, mild periodate treat- detected by SDS-PAGE and L10 immunoblotting. pH gradient is indicated. ment converts the nine-carbon sialic acid moiety to a seven-carbon derivative that retains acidic properties (5-acetaido-3,5-dideoxy-L- http://www.jimmunol.org/ arabino-2-heptulosonic acid) (36, 37). After periodate treatment, in Fig. 4 for thymi from three individuals (#1 to #3), contain mul- CEM cells bound L10 mAb as anticipated but no longer bound tiple CD43 charge variants that may span a wide range of isoelec- 8B4/20 mAb (Table I). This finding indicates that the terminal trofocusing points, consistent with variable sialylation of CD43. polyhydroxylated tail of sialic acid participates in the 8B4/20 The isoelectrofocusing profiles, although reproducible for individ- epitope (see Discussion). ual samples (data not shown), vary substantially among individuals (Fig. 4). Paired comparisons of 8B4/20 and L10 precipitates indi- Mapping of the 8B4/20 epitope cate that the 8B4/20-expressing thymocyte molecules do not cor- To localize the 8B4/20 epitope, we treated 8B4/20-expressing respond to defined charge species but rather constitute a subset of

MOLT-4 T lymphoblastoid cells with neutrophil elastase (25 ␮g/ by guest on September 26, 2021 the differently charged CD43 molecules. ml) sufﬁcient to cleave and release the N-terminal ϳ40- to 52-kDa Chemical characterization of the 8B4/20 epitope fragment from the linear extracellular region of CD43 (27). The released CD43 fragment contains the L10 epitope, and the residual To examine whether sialic acid is critical for expression of the cell-associated fragment contains the T305 epitope (27). Elastase epitope, thymocyte lysates were examined after treatment with treatment, which caused loss of Ͼ90% of L10-binding sites (Fig. sialidase. On immunoblotting, L10 stained desialylated CD43, but 6, left) and did not effect T305 binding (right), led to the loss of no bands were detected with 8B4/20 mAb, indicating that the 8B4/20-binding sites (90 Ϯ 4% decrease; n ϭ 3) (center), strongly 8B4/20 epitope has a requirement for sialic acid (Fig. 5). Likewise, suggesting that the 8B4/20 epitope is located within the N-terminal 8B4/20 immunoprecipitates from CEM cells became undetectable ϳ40- to 52-kDa region of CD43. by 8B4/20 mAb after treatment with sialidase (data not shown). Siali- 8B4/20 enhances thymocyte maturation The expression of 8B4/20-CD43 on early but not late thymocytes, together with previous ﬁndings showing that early thymocytes en- riched for 8B4/20 expression are more sensitive to both “spontaneous” and induced apoptosis than the 8B4/20Ϫ mature ones (38),

FIGURE 6. 8B4/20 epitope is released from T-lymphoid cells by neutrophil elastase. Shown are flow cytometric profiles of MOLT-4 cells that were mock-treated (gray profiles) or treated with elastase (white profiles FIGURE 5. 8B4/20 epitope is sialidase-sensitive. Mock-treated and with dark outlines). The cells were stained with L10 (left), 8B4/20 (center), sialidase-treated thymocyte lysates were subjected to immunoblotting with or T305 mAbs (right). Isotype control staining is indicated by dashed pro- L10 and 8B4/20 mAbs. Note that L10 stained desialylated CD43, but no files. Note that the 8B4/20 epitope, like the L10 epitope and unlike the T305 bands were detected with 8B4/20 mAb. The 8B4/20 blot was overexposed epitope, is lost from elastase-treated cells. Similar results were obtained for to show the lack of signal in the sialidase-treated sample. 8B4/20 and L10 staining of elastase-treated CEM cells (data not shown). 5968 A CD43 VARIANT PROMOTES MATURATION OF ISOLATED THYMOCYTES IN VITRO

Table II. 8B4/20 mAb does not induce apoptosis in isolated human in the fraction of CD3medium cells (region M3) in the 8B4/20- thymocytes treated cultures. This pattern of change was observed in thymi from six of six individuals; the mean 8B4/20-induced decrease of % Apoptotic Cells after 15-h Culture with CD3Ϫ cells was Ϫ11 Ϯ 6% (range 4–19%), and the mean increase high ϩ Ϯ Thymus Control IgM 8B4/20 mAb Dexamethasone of the CD3 population was 10 3% (range 3–14%). Double staining showed that thymocytes cultured with 8B4/20 12a 319mAb signiﬁcantly decreased the number of CD4ϩCD8ϩ double- a 29 552positive cells and increased the number of CD4 and CD8 single- 322b 37 56 47b 931positive cells relative to cultures with control IgM (Fig. 7B). The 8B4/20-induced increase was greater for CD4ϩ cells than CD8ϩ. a Thymus 1 and 2 were analyzed by TUNEL labeling. b Thymus 3 and 4 were analyzed by annexin V-FITC binding. This pattern of change was observed in six of six thymi; the mean decrease of double-positive cells was Ϫ9 Ϯ 4%, and the mean increase of single-positive cells was ϩ11 Ϯ 4% (ϩ8 Ϯ 4% for suggests that these molecules participate in T cell maturational CD4ϩ and ϩ3 Ϯ 1% for CD8ϩ). 8B4/20 expression relatively to events. To study this putative role, we examined the effects of CD3 after culture with the Ab was evaluated by further addition of 8B4/20 ligation on in vitro apoptosis of isolated thymocytes and on 8B4/20 mAb followed by PE-goat anti-IgM and CD3-FITC. As the phenotype of the surviving cells. Isolated thymocytes were shown in Fig. 7C, the surface density of 8B4/20 Ag decreased over cultured with 8B4/20 mAb and then examined for apoptosis by culture (Ϫ45 Ϯ 6% 8B4/20 positive cells, n ϭ 5), and the resulting

TUNEL assay or FITC-annexin V binding. After 15 h with 8B4/20 phenotype was consistent with progression of maturation because ma- Downloaded from mAb, the extent of thymocyte apoptosis was not significantly dif- ture CD3high thymocytes are 8B4/20Ϫ. These findings strongly sug- ferent from the spontaneous apoptosis of thymocytes cultured with gest that 8B4/20-CD43 molecules function in thymocyte maturation. control IgM, in contrast to thymocytes cultured with dexamethasone, which had significantly higher numbers of apoptotic cells (Table II). Total surviving cells were also quantified. The fraction Discussion

of recovered surviving cells varied in the four experiments, rang- Immunological and biochemical findings identified 8B4/20 as an http://www.jimmunol.org/ ing from 65% of seeded cells in thymus 3–85% in thymus 1 (mean epitope on human thymic CD43 molecules. The immunological 78 Ϯ 9%). Within individual experiments, however, the fraction of comparisons were made with L10, the mAb used in the original recovered thymocytes in the 8B4/20 culture differed by Ͻ5% from definition of CD43 (39). L10 was found to react with molecules the control IgM culture (data not shown). In two additional exper- precipitated by 8B4/20, and 8B4/20 reacts with L10-precipitated iments, the 8B4/20 mAb was added to the dexamethasone culture, molecules (Fig. 1B). L10 and 8B4/20 mAbs both stained wild-type and FITC-annexin V binding was evaluated after 15-h culture. Again, CD43ϩ CEM cells, and both failed to stain CEM cells depleted of no dramatic effect was induced by the Ab on dexamethasone-induced CD43 by homologous recombination (Fig. 2). Biochemical char- apoptosis (data not shown). Together, these findings indicate that acteristics shared by 8B4/20 Ag and L10 Ag (CD43) include an 8B4/20 mAb does not influence in vitro thymocyte apoptosis. atypical large decrease of SDS-PAGE mobility (increase of appar- by guest on September 26, 2021 When examined for surface marker expression, viable thymo- ent m.w.) on removal of sialic acid (Fig. 3). 8B4/20 Ag, like CD43, cytes harvested after 15 h with 8B4/20 mAb included a substan- is sensitive to low levels of neutrophil elastase (Fig. 6) and insen- tially smaller number of cells that were CD3Ϫ (Fig. 7A, M1 region) sitive to endoglycosidase-F (Fig. 3). On the other hand, 8B4/20- and a greater number of cells that were CD3high (region M4) com- expressing molecules and L10-expressing molecules are not iden- pared with cultures with control IgM. An increase was noted also tical because the former show a differentiation-dependent,

FIGURE 7. 8B4/20 mAb triggers maturation of human thymocytes in vitro. Flow cytometric profiles of thymocytes after 15-h culture with control IgM (left panels) or 8B4/20 mAb (right panels). A, Har- vested thymocytes were stained with CD3-FITC; the table indicates the percentage of cells in each marker region. B, Harvested thymocytes were double stained with CD4-FITC and CD8-PE; the table indicates the percentage of cells in each quadrant. One thymus representative of six is shown. C, Harvested thymocytes were double stained with CD3-FITC and 8B4/20 mAb, followed by PE-labeled goat anti IgM serum. The table indicates the percentage of cells in each quadrant. One thymus representative of three and different from the one in A and B is shown. The Journal of Immunology 5969 primarily thymic distribution, and the latter are broadly expressed delivery of intracellular signals via binding to an unidentified on leukocytes. counterreceptor. Isoelectrofocusing showed that 8B4/20 molecules constitute a Altogether, its expression pattern and biochemical characteris- subset of the thymocyte molecules reactive with L10, and also that tics suggest the participation of 8B4/20-CD43 in carbohydrate 8B4/20-CD43 molecules are heterogeneous in charge, suggesting based interactions of developing thymocytes. Discrete stages of variable content of sialic acid (Fig. 4). Previous studies showed thymic development are distinguishable by sequential phenotypic that developmentally regulated CD43 epitopes result from post- changes: from the immature triple-negative (CD4ϪCD8ϪCD3/ translational events primarily glycosylation (40). For thymocyte TCRϪ) to double-positive (CD4ϩCD8ϩCD3/TCRϪ/low/medium)to CD43, two enzymatic activities are known to be important: 1) core the immunocompetent mature single-positive (CD4ϩCD8ϪCD3/ 2 ␤ 1–6-N-acetylglucosaminyl transferase (C2GnT), expressed in TCRhigh or CD4ϪCD8ϩCD3/TCRhigh) thymocytes. T cell reper- early but not late thymocytes (14), which generates branched O- toire selection occurs primarily at the double-positive stage and is glycans (6), and 2) sialyltransferase, which determines the extent driven by both positive and negative selection signals (50, 51). of sialylation (3, 41). Carbohydrate-mediated interactions have a major role, at least C2GnT is required to generate the T305 epitope on thymic for negative (apoptotic) selection mechanisms. Galectin-1, a lectin CD43 (42). Because T305 and 8B4/20 are similarly distributed (on product of thymic epithelium, interacts with galactose-bearing re- immature thymocytes and not on CD3high cells) (14, 21), we ceptors on immature thymocytes. Interaction with galectin-1 in- mapped 8B4/20 with respect to T305. The mapping data localized duces apoptosis of immature thymocytes in vitro and is thought to 8B4/20 to the N-terminal region of CD43, thereby distinguishing deliver a second apoptotic signal to developing thymocytes, which this epitope from T305, which is located proximal to the phospho- synergizes with CD3-TCR engagement (52). Prominent among the Downloaded from lipid bilayer (Fig. 6). It is not known whether 8B4/20 requires core thymocyte surface molecules that bind galectin-1 is the T305 vari- 2 glycosylation; however, its expression on strains of CEM cells ant of CD43 (14). expressing primarily low m.w. CD43 and negative for T305 to- To characterize the function of the 8B4/20 variant of CD43, we gether with its absence on neutrophils, which are strongly positive assumed that its interaction with natural ligand could be mimicked for T305 (23), suggest that 8B4/20 does not require core 2 glyco- by specific Ab, and we therefore investigated whether mAb addi-

sylation. Thus, 8B4/20 and T305 represent two distinct glycosy- tion to cultures of isolated thymocytes would alter the rate of in http://www.jimmunol.org/ lation-dependent CD43 structures expressed in a developmentally vitro apoptosis or the phenotype of the surviving cells. The pres- regulated fashion on thymocytes. ence of the 8B4/20 mAb in overnight cultures did not induce apo- The other enzymatic activity known to be relevant for CD43 ptosis of isolated thymocytes above the spontaneous levels occur- epitopes is sialyltransferase. Several sialyltransferases are ex- ring in control cultures (Table II). pressed in human thymus, some of which show differentiation- In contrast, 8B4/20 mAb increased the conversion of thymo- dependent expression (43, 44). In particular, increased expression cytes to differentiated phenotypes. The fraction of CD3Ϫ cells was of Gal-␤-1,3GalNAc ␣2,3-sialyltransferase in mature thymocytes decreased in 8B4/20 cultures, the fraction with high CD3 expres- correlates with increased sialylation of O-glycans (43). Sialylation sion increased substantially and the CD3med fraction increased was found to be essential to the 8B4/20 epitope, as recognition by slightly (Fig. 7). These changes in CD3 expression in 8B4/20- by guest on September 26, 2021 its mAb was abrogated by sialidase treatment (Fig. 5 and Table I) treated thymocytes were accompanied by a decrease of double- and also by mild oxidation conditions that specifically truncate the positive cells and an increase of single-positive cells. In the ab- exocyclic polyhydroxyl side chain of sialic acid without destroying sence of increased cell death, the finding of higher numbers of the acidic moiety (36, 45) (Table I). Despite the strict epitope CD3high cells and single-positive cells strongly suggests that requirement for a sialic acid moiety, the level of 8B4/20 expression 8B4/20 ligation delivers a positive signal that favors completion of is high on immature thymocytes, which are minimally sialylated, the thymocyte maturation program. These findings together with and low on mature thymocytes, which are heavily sialylated. The the higher expression of 8B4/20 on CD3Ϫ/low cells suggest that decrease of 8B4/20 expression as thymocytes mature could reflect 8B4/20-CD43 participates in the early events of thymic selection, decrease of an enzyme required for synthesis of the epitope. Al- where it might deliver information on the microenvironmental ternatively, decrease of 8B4/20 could result from epitope “mask- context permitting CD3/TCR up-regulation. Assuming these ef- ing” due to extensive sialylation. The latter possibility is suggested fects as physiological, then a dual role for CD43 in T cells emerg- by analogy with peanut agglutinin (PNA) binding epitopes, which es: as carrier of epitopes interacting with ligand(s) relevant for also decrease as thymocytes mature (34, 35). Binding sites for maturation signals in thymocytes and as regulator of cell surface PNA, consisting of Gal-GalNac epitopes (46), are masked to vary- interactions during activation in circulating T cells (53). A role for ing extent on CD43 molecules and can be exposed by enzymatic 8B4/20-CD43 surface molecules in influencing the balance be- desialylation (26, 41). tween negative and positive selection is consistent with the current A situation possibly related to 8B4/20 expression exists in ger- model of intrathymic development, in which a series of overlap- minal center B cell interactions in which developmentally regu- ping signals provided by growth factors, cytokines, and adhesion lated expression of the sialic acid-requiring epitope CDw75 on molecules integrate the effects of T cell Ag receptors engagement, surface molecules such as IgM regulates binding of the cells to the guiding thymocyte survival, expansion, and differentiation. germinal center specific lectin CD22 (45, 47, 48). Interestingly, Taken together, these findings demonstrate that select CD43 both the CDw75 and 8B4/20 epitopes require the exocyclic poly- thymocyte molecules carry a differentiation epitope, 8B4/20, that hydroxyl side chain of sialic acid. In the B cell situation, interac- is structurally and functionally distinct from the previously de- tions of the CDw75 ectodomain, which provide the cell with in- scribed T305 epitope. Biochemical characterization indicates that formation on its microenvironmental context, are physiologically 8B4/20-CD43 could serve as ligand for carbohydrate-based inter- relevant to the regulation of other signaling pairs, fine-tuning the actions to deliver cues to the developing thymocyte on its micro- strength and quality of the Ag receptor signal (reviewed in Ref. environmental context. The ability of 8B4/20 mAb to drive thy- 49). The developmentally regulated sialylation-dependent 8B4/20 mocyte maturation in vitro strongly suggests that 8B4/20-CD43 epitope might similarly provide developing T cells with the surface molecules participate in early events of thymic selection, condu- structure required to regulate intrathymic localization and/or the cive to thymocyte differentiation. 5970 A CD43 VARIANT PROMOTES MATURATION OF ISOLATED THYMOCYTES IN VITRO

Acknowledgments 23. Remold-O’Donnell, E., and D. Parent. 1994. Two proteolytic pathways for down- regulation of the barrier molecule CD43 of human neutrophils. J. Immunol. We thank B. Ardman (Department of Hematology-Oncology, New En- 152:3595. gland Medical Center, Boston MA) for performing immunofluorescence on 24. Foley, G. E., H. Lazarus, S. Farber, B. G. Uzman, B. A. Boone, and R. E. McCarthy. 1965. Continuous culture of human lymphoblasts from periph- CD43 stably transfected HeLa cells and for providing CD43-depleted CEM eral blood of a child with acute leukemia. Cancer 18:522. cell lines; R. Fox for providing T305 Ab; M. E. Cosulich for helpful dis- 25. Minowada, J., T. Onuma, and G. E. Moore. 1972. Rosette-forming human lym- cussion; and Istituto “G. Gaslini” Cardiac Surgery, for providing thymus phoid cell lines. I. Establishment and evidence for origin of thymus-derived lym- specimens. phocytes. J. Natl. Cancer Inst. 49:891. 26. Remold-O’Donnell, E., D. Kenney, R. Parkman, L. Cairns, B. Savage, and F. S. Rosen. 1984. Characterization of a human lymphocyte surface sialoglyco- References protein that is defective in Wiskott-Aldrich syndrome. J. Exp. Med. 159:1705. 27. Remold-O’Donnell, E., and D. Parent. 1995. Specific sensitivity of CD43 to 1. Fabbi, M., M. Tiso, R. M. R. Gangemi, A. Favre, P. Demartini, and neutrophil elastase. Blood 86:2395. A. Bargellesi-Severi. 1994. A novel 120-kDa antigen shared by immature thy- 28. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the mocytes and long-term-activated T cells. Eur. J. Immunol. 24:1. head of bacteriophage T4. Nature 227:680. 2. Brown, W. R. A., A. N. Barclay, C. A. Sunderland, and A. F. Williams. 1981. 29. De Rossi, G., D. Zarcone, F. Mauro, G. Cerruti, C. Tenca, A. Puccetti, Identification of a glycophorin-like molecule at the cell surface of rat thymocytes. F. Mandelli, and C. E. Grossi. 1993. Adhesion molecule expression on B-cell Nature 289:456. chronic lymphocytic leukemia cells: malignant cell phenotypes define distinct 3. Remold-O’Donnell, E., C. Zimmerman, D. Kenney, and F. S. Rosen. 1987. Ex- disease subsets. Blood 81:2679. pression on blood cells of sialophorin, the surface glycoprotein that is defective 30. O’Farrell, P. H. 1975. High resolution two-dimensional electrophoresis of pro- in Wiskott-Aldrich syndrome. Blood 70:104. teins. J. Biol. Chem. 250:4007. 4. Horejsi, V., and H. Stockinger. 1997. CD43 workshop panel report. In Leukocyte 31. Gorczyca, W., J. Gong, B. Ardelt, F. Traganos, and Z. Darzynkiewicz. 1993. The Typing VI: White Cells Differentiation Antigens. T. Kishimoto, H. Kikutani, cell cycle related differences in susceptibility of HL60 cells to apoptosis induced A. E. G. Kr. von dem Borne, S. M. Goyert, D. Y. Mason, M. Miyasaka, by various antitumor agents. Cancer Res. 53:3186. L. Moretta, K. Okumura, S. Shaw, T. Springer, et al. eds. Garland, New York, 32. Tarentino, A. L., C. M. Gomez, and T. H. Plummer, Jr. 1985. Deglycosylation of Downloaded from p. 494. asparagin-linked glycans by peptide:N-glycosidase F. Biochemistry 24:4665. 5. Remold-O’Donnell, E., A. E. Davis III, D. Kenney, K. Ramakrishnan-Bhaskar, 33. Wilson, A. P., and C. C. Rider. 1992. Evidence that leukosialin, CD43, is in- and F. S. Rosen. 1986. Purification and chemical composition of gpL115, the tensely sulfated in the murine T lymphoma line RDM-4. J. Immunol. 148:1777. human lymphocyte surface sialoglycoprotein that is defective in Wiskott-Aldrich 34. Reisner, Y., M. Linker-Israeli, and N. Sharon. 1976. Separation of mouse thy- syndrome. J. Biol. Chem. 261:7526. mocytes into two subpopulations by the use of peanut agglutinin. Cell. Immunol. 6. Carlsson, S. R., and M. Fukuda. 1986. Isolation and characterization of leuko- 25:129. sialin, a major sialoglycoprotein on human leukocytes. J. Biol. Chem. 261:12779. 35. Sharon, N. 1983. Lectin receptors as lymphocyte surface markers. Adv. Immunol.

7. Killeen, N., A. N. Barclay, A. C. Willis, and A. F. Williams. 1987. The sequence 34:213. http://www.jimmunol.org/ of rat leukosialin (W3/13 antigen) reveals a molecule with O-linked glycosylation 36. Van Lenten, L., and G. Ashwell. 1971. Studies on the chemical and enzymatic of one third of its extracellular amino acids. EMBO J. 6:4029. modiﬁcation of glycoproteins: a general method for the tritiation of sialic acid- 8. Shelley, C. S., E. Remold-O’Donnell, A. E. Davis III, G. A. P. Bruns, containing glycoproteins. J. Biol. Chem. 246:1889. F. S. Rosen, M. C. Carroll, and A. S. Whitehead. 1989. Molecular characteriza- 37. Schauer, R. 1982. Chemistry, metabolism and biological functions of sialic acids. tion of sialophorin (CD43), the lymphocyte surface sialoglycoprotein defective in Adv. Carbohydr. Chem. Biochem. 40:131. Wiskott-Aldrich syndrome. Proc. Natl. Acad. Sci. USA 86:2819. 38. Tiso, M., R. Gangemi, A. Bargellesi Severi, S. Pizzolitto, M. Fabbi and A. Risso. 9. Pallant, A., A. Eskenazi, M.-G. Mattei, R. E. K. Fournier, S. R. Carlsson, 1995. Spontaneous apoptosis in human thymocytes. Am. J. Pathol. 147:434. M. Fukuda, and J. G. Frelinger. 1989. Characterization of cDNAs encoding hu- 39. Borche, L., F. Lozano, R. Vilella, and J. Vives. 1987. CD43 monoclonal anti- man leukosialin and localization of the leukosialin gene to chromosome 16. Proc. bodies recognize the large sialoglycoprotein of human leukocytes. Eur. J. Immu- Natl. Acad. Sci. USA 86:1328. nol. 17:1523. 10. Cyster, J. G., D. M. Shotton, and A. F. Williams. 1991. The dimensions of the T 40. Fukuda, M. 1991. Leukosialin, a major O-glycan-containing sialoglycoprotein

lymphocyte glycoprotein leukosialin and identification of linear protein epitopes defining leukocyte differentiation and malignancy. Glycobiology 1:347. by guest on September 26, 2021 that can be modified by glycosylation. EMBO J. 190:893. 41. Brown, W. R. A., and A. F. Williams. 1982. Lymphocyte cell surface glycop- 11. Ardman, B., M. A. Sikorski, and D. E. Staunton. 1992. CD43 interferes with roteins which bind to soybean and peanut lectins. Immunology 46:713. T-lymphocyte adhesion. Proc. Natl. Acad. Sci. USA 89:5001. 42. Bierhuizen, M. F. A., K. Maemura, and M. Fukuda. 1994. Expression of a dif- 12. Manjunath, N., R. S. Johnson, D. E. Staunton, R. Pasqualini, and B. Ardman. ferentiation antigen and poly-N-acetyllactosaminyl O-glycans directed by a 1993. Targeted disruption of CD43 gene enhances T lymphocyte adhesion. J. Im- cloned core 2–1,6-N-acetylglucosaminyltransferase. J. Biol. Chem. 269:4473. munol. 151:1528. 43. Gillespie, W. G., J. C. Paulson, S. Kelm, M. Pang, and L. G. Baum. 1993. 13. Manjunath, N., M. Correa, M. Ardman, and B. Ardman. 1995. Negative regula- Regulation of ␣2,3-sialyltransferase expression correlates with conversion of tion of T-cell adhesion and activation by CD43. Nature 377:535. peanut agglutinin (PNA)ϩ to PNAϪ phenotype in developing thymocytes. J. Biol. 14. Baum, L. G., M. Pang, N. L. Perillo, T. Wu, A. Delegeane, C. H. Uittenbogaart, Chem. 268:3801. M. Fukuda, and J. J. Seilhamer. 1995. Human thymic epithelial cells, express an 44. Baum, L. G., K. Derbin, N. L. Perillo, T. Wu, M. Pang, and C. H. Uittenbogaart. endogenous lectin, galectin-1, which binds to core 2 O-glycans on thymocytes 1996. Characterization of terminal sialic acid linkages on human thymocytes: and T lymphoblastoid cells. J. Exp. Med. 181:877. correlation between lectin-binding phenotype and sialyltransferase expression. 15. Mentzer, S. J., E. Remold-O’Donnell, M. A. V. Crimmins, B. E. Bierer, J. Biol. Chem. 271:10793. F. S. Rosen, and S. J. Burakoff. 1987. Sialophorin, a surface sialoglycoprotein 45. Sgroi, D., A. Varki, S. Braesch-Andersen, and I. Stamenkovic. 1993. CD22, a B defective in the Wiskott-Aldrich syndrome, is involved in human T lymphocyte cell-specific immunoglobulin superfamily member, is a sialic acid-binding lectin. proliferation. J. Exp. Med. 165:1383. J. Biol. Chem. 268:7011. 16. Axelsson, B., Y. Youseffi-Etemad, S. Hammarstrom, and P. Perlmann. 1988. 46. Swamy, M. J., D. Gupta, S. K. Mahanta, and A. Surolia. 1991. Further charac- Induction of aggregation and enhancement of proliferation and IL-2 secretion in terization of the saccharide specificity of peanut (Arachis hypogea) agglutinin. human T cells by antibodies to CD43. J. Immunol. 141:2912. Carbohydr. Res. 213:59. 17. Alvarado, M., C. Klassen, J. Cerny, V. Horejsi, and R. E. Schmidt. 1995. 47. Stamenkovic, I., D. Sgroi, A. Aruffo, M. S. Sy, and T. Anderson. 1991. The B MEM-59 monoclonal antibody detects a CD43 epitope involved in lymphocyte lymphocyte adhesion molecule CD22 interacts with leukocyte common antigen activation. Eur. J. Immunol. 25:1051. CD45RO on T cells and ␣2,6 sialyltransferase, CD75, on B cells. Cell 66:1133. 18. Sperling, A. I., J. M. Green, R. L. Mosley, P. L. Smith, R. J. DiPaola, J. R. Klein, 48. Leprince, C., K. E. Draves, R. L. Geahlen, J. A. Ledbetter, and E. A. Clark. 1993. J. A. Bluestone, and C. B. Thompson. 1995. CD43 is a murine T cell co-stim- CD22 associates with the human surface IgM-B cell antigen receptor complex. ulatory receptor that functions independently of CD28. J. Exp. Med. 182:139. Proc. Natl. Acad. Sci. USA 90:3236. 19. Fanales-Belasio, E., G. Zambruno, A. Cavani, and G. Girolomoni. 1997. Anti- 49. Cyster, J. C., and C. C. Goodnow. 1997. Tuning antigen receptor signaling by bodies against sialophorin (CD43) enhance the capacity of dendritic cells to clus- CD22: integrating cues from antigens and the microenvironment. Immunity ter and activate T lymphocytes. J. Immunol. 159:2203. 6:509. 20. Carlsson, S. R., H. Sasaki, and M. Fukuda. 1986. Structural variations of O- 50. von Boehmer, H. 1994. Positive selection of lymphocytes. Cell 76:219. linked oligosaccharides present in leukosialin isolated from erythroid, myeloid, 51. Page, D. M., L. P. Kane, T. M. Onami, and S. M. Hedrick. 1996. Cellular and and T-lymphoid cell lines. J. Biol. Chem. 261:12787. biochemical requirements for thymocytes negative selection. Semin. Immunol. 21. Fox, R. I., M. Hueniken, S. Fong, S. Behar, I. Royston, S. K. Singhal, and 8:69. L. Thompson. 1983. A novel cell surface antigen (T305) found in increased 52. Perillo, N. L., C. H. Uittenbogaart, J. T. Nguyen, and L. G. Baum. 1997. Galec- frequency on acute leukemia cells and in autoimmune disease states. J. Immunol. tin-1, an endogenous lectin produced by thymic epithelial cells, induces apoptosis 131:762. of human thymocytes. J. Exp. Med. 185:1851. 22. Piller, F., F. LeDeist, K. I. Weinberg, R. Parkman, and M. Fukuda. 1991. Altered 53. Sperling, A. I., J. R. Sedy, N. Manjunath, A. Kupfer, B. Ardman, and O-glycan synthesis in lymphocytes from patients with Wiskott-Aldrich syn- J. K. Burkhardt. 1998. TCR signaling induces selective exclusion of CD43 from drome. J. Exp. Med. 173:1501. the T cell-antigen-presenting cell contact site. J. Immunol. 161:6459.