Structural analysis of the CD69 early activation antigen by two monoclonal antibodies directed to different epitopes

Mokwdar Immunology, Vol. 28, No. l/2, pp. 159-168, 1991 Printedin Great Britain.

0161-5890/91 $3.00 + 0.00

Pergamon Press plc

STRUCTURAL ANALYSIS OF THE CD69 EARLY ACTIVATION ANTIGEN BY TWO MONOCLONAL ANTIBODIES DIRECTED TO DIFFERENT EPITOPES FRANCA GEROSA,* TULLIO POZZAN,~

MARINATOMMASI,*MARIA SCARDONI,* ROBERTO S. ACCOLLA,* LIBONATI,~ GIUSEPPE TRIDENTE* and GIUSEPPE CARRA$$

MASIMO

*Istituto di Scienze Immunologiche, Universita di Verona, Italy; TIstituto di Patologia Universita di Ferrara, Italy; fIstituto di Chimica Biologica, Universita di Verona,

(First received 20 April 1990; accepted in revised form

Generale, Italy

18 July 1990)

Abstract-The biochemical structure of CD69 early activation antigen has been characterized by means of two newly isolated mAb, namely Cl.18 and E16.5. Upon analysis by SDS-PAGE, Cl.lS-reactive molecules immunoprecipitated from ‘*‘I-surfacelabeled PMA activated PBL consisted of a 32 + 32 kD dimer, a 32 + 26 kD dimer, a 26 + 26 kD dimer and a 21 + 21 kD dimer. E16.5reactive molecules consisted of a 26 + 26 kD dimer and a 21 + 21 kD dimer. Cross absorption experiments showed that E 16.5 mAb reacts with an epitope of the CD69 molecule distinct from the one recognized by C 1.18mAb and present only on a subpopulation of the CD69 molecular pool. The patterns of migration of C1.18and El 6.5reactive molecules in two-dimensional gel-electrophoresis, under reducing conditions before and after treatment with Endoglycosidase F enzyme suggest that the two mAb recognize the same glycoprotein structure, but in two distinct glycosylation forms, both expressed on the cell surface membrane. Finally, ~32, p26 and p21 of CD69 complex obtained from three distinct normal donors did not show appreciable structural polymorphism, by two-dimensional peptide mapping, not only among single subunits within the same individual, but also among homologous subunits in distinct individuals. Further, it was found that CD69 complex is expressed at the cell surface of resting PBL, although at a very reduced level in comparison to PMA activated cells. Cl. 18 and E16.5 mAb induced comparable cell proliferation and IL-2 production in PBL in the presence of PMA. Cl. 18 mAb increased intracellular free calcium concn in PMA activated PBL after cross-linking with goat anti mouse Ig, while the effect induced by E16.5 mAb after cross-linking was consistently lower. Finally, it was found that Sepharose-linked Cl.18 mAb, in the presence of rIL-2 or PMA, did not induce TNF release from 6 NK cell clones.

INTRODUCTION

Following interaction with different agents such as antigens, mitogens and mAb to specific surface structures alone or in the presence of tumor promoter phorbol esters, T lymphocytes acquire the expression of newly synthesized surface molecules defined activation antigens. Some of these molecules function as growth factor receptors [IL-2 R (CD25), insulin and transferrin receptors] (Leonard et al. 1982; Trowbridge and Omary, 1981; Helderman and Strom, 1978), while others (such as 4F2, Act-l) (Haynes et al. 1981; Lazarovits et af. 1984) have still an elusive function. Recently, an activation antigen designated EA-1 (for early activation antigen) (Hara et al. 1986) has been described. This structure is synthesized and expressed at the cell surface l-2 hr following stimulation with the tumor promoter PMA. The functional §Author to whom correspondence should be addressed at: Istituto di Chimica Biologica, Universiti di Verona, Strada Le Grazie, 37100 Verona, Italy. Abbreviations used in this paper: ZD-PM, two-dimensional peptide mapping; [Ca2+li free cytosolic calcium concn; LPGO, lactoperoxidase-glucoseoxidase; SN, supernatant; ZD-PAGE, two-dimensional gel electrophoresis.

importance of EA-1 has been deduced from the fact that different mAb, such as anti-AIM, MLR3, P8 (Cebrian et al. 1988; Cosulich et al. 1987; Nakamura et al. 1989), trigger, in the presence of PMA, T lymphocyte proliferation, IL-2 production and CD25 expression. Structurally, EA-1 has been described as a series of 32 + 32 kD, 32 + 28 kD and 28 + 28 kD disulfide linked glycoproteins (Bjorndahl et al. 1988; Lanier et al. 1988). At the 4th International Workshop on Human Leukocyte Differentiation Antigens mAb directed to this structure have been assigned to the new determinant cluster (CD) 69 (Knapp et al. 1989). In the present report we analyzed the biochemical structure of CD69 antigen by means of two newly isolated mAb, namely Cl. 18 and E16.5. We found that these two antibodies recognize distinct epitopes on differently glycosylated forms of CD69 antigen. CD69 appears as a complex structure composed of a series of differently glycosylated, disulfide linked dimers i.e. 32 + 32 kD, 32 + 26 kD, 26 + 26 kD, as described, and by a 21 + 21 kD disulfide linked dimer. Furthermore, these subunits did not show appreciable structural polymorphism, by two-dimensional peptide mapping (2D-PM), not only among single subunits within the same individual, but also 159

160

F.

GEROSA et al.

among homologous subunits in distinct individuals. We found that CD69 is expressed also at the cell surface membrane of resting PBL. Finally, we investigated whether CD69 could be involved in functional triggering of NK cells. We found that NK cell clones could not be activated through this antigen to produce TNF, suggesting that CD69 antigen expressed on NK cells (Lanier et al. 1988) might not be associated to a functional activity. MATERIALS

AND METHODS

Isolation and culture of lymphocytes Lymphocytes were isolated from PBL by centrifugation on Ficoll-Hypaque density gradient. NK cell clones (CD3-, CDl6+) were obtained from PBL treated with a mixture of mAb directed to CD3, CD4 and CD8 (CBT-3M, kindly provided by F. Malavasi, University of Torino, OKT4 and OKT8, Ortho, Raritan, NJ) and complement and centrifugation on a gradient of Ficoll-Hypaque to remove dead cells. CD33, CDl6+ PBL were cultured under limiting dilution in RPM1 1640 containing 10% FCS in the presence of irradiated spleen feeder cells with 1% (v/v) PHA and rIL-2 (kindly provided by Glaxo, Geneva, Switzerland). Cultures were supplemented weekly with irradiated feeder cells and rIL-2. Microcultures obtained from cells plated at 2 or less than 2 cells/well were operationally considered as clones, and selected for high cytolytic activity against the K562 human erythromyeloid cell line and the Hu126 human glioma cell line (Gerosa et al. 1988). Production qf Cl.18

and E16.5 mAb

Balb/c mice were immunized with a mixture of 3NK clones (CD3-, CDl6+) by two weekly i.v. injection of 10’ cells. After 15 days the mice received a booster injection of 1.5 x 10’ cells. Three days later, splenocytes from one mouse were fused with the murine P3Ul myeloma cell line. Hybridoma supernatants (SN) were screened by the capacity to induce [3H]TdR incorporation of PBL in the presence of submitogenic doses of PMA (Sigma Chemical Co., St Louis, MO). Three positive hybridomas were cloned under limiting dilution conditions; two hybrid clones, Cl.18 and E16.5, are described in the present report. Cl.18 and E16.5 mAb induced, after 3 days [3H]TdR incorporation in PBL in the presence of 0.6 ng/ml PMA (respectively, 92 x lo3 and 82 x lo3 cpm; background 6 x 103cpm). Cl.18 and El6.5mAb were identified as IgGl by double immunodiffusion with anti mouse Ig subclass-specific antibodies (ICN Biochemicals, Irvine, CA). Proliferative assay PBL (2 x 10’ cells/well) were incubated with Cl .18, E16.5 (10 pg/ml of purified antibody) and 0KT3 (l/4 dilution of culture SN), in the presence of PMA (Zng/ml). Cultures were incubated for 72 hr and 0.5 I.rCi [3H]TdR were added 6 hr before harvesting.

SN were collected after 48 hr of incubation to detect IL-2 activity, and, after replacing with fresh medium, PBL were cultured for additional 24 hr to evaluate proliferative activity. Assay for IL-2 activity The murine CTLL T cell line was used as indicator system. These cells (4 x 10’) were cultured for 24 hr in the presence of serial dilutions of SN from cultured PBL. [3H]TdR was added 6 hr before harvesting. The arbitrary value of 1 U/ml was assigned to the IL-2 concn inducing 50% maximal response of the standard rIL-2. Stimulation of NK cell clones for TNF production NK cell clones were incubated for 18 hr with PMA (1 ng/ml) or rIL-2 (250 U/ml) and with the following mAb linked to CNBr-Sepharose 4B (Pharmacia, 6 mg antibody/l ml Sepharose, 5 ~1 Uppsala; heads/2x lo5 cells): Cl.18, Bl6.15 (directed to CD16) (Gerosa et al. 1989), B9.12.1 (directed to a monomorphic determinant of class I HLA antigens). After culture SN were collected and TNF was detected with a TNF Elisa test kit (T Cell Sciences, Cambridge, MA). [Ca 2+J measurement [Ca2+], was measured with the fluorescent probe fura- essentially as described previously (Treves et al. 1987). PMA treated PBL loaded with fura- were washed and resuspended in a medium containing: 125 mM NaCl, 5 mM KCI, 1 mM MgSO,, 1 mM CaCl,, 5.5 mM glucose, 1 mM Na,HPO, and 20 mM Hepes (pH 7.4 at 37°C). External cell surface radiolabeling PBL cultured for 24 hr in the presence of PMA (2 ng/ml) or freshly isolated PBL were surface radiolabeled with 12’1by the lactoperoxidase-glucose oxidase (LPGO) catalyzed iodination method originally described by Schenkein et al. (1972) with minor modifications (Carra and Accolla, 1987). Briefly 3 x 10’ cells were extensively washed in cold RPM1 medium followed by several washes in PBS (0.01 M sodium phosphate; 0.14 M NaCl, pH 7.4). Cells were then resuspended in 1 ml of PBS plus 11 mM glucose and the following reagents were added: 20~1 Kl (5 x 10m5M), 1 mCi of Na’? carrier free (lOOmCi/ml, Amersham Corp., UK) and 50~1 of lactoperoxidase (1 mg/ml solution in PBS, Sigma Chemical Co., St Louis, MO). Reaction was started by addition of 30 p 1 of glucose oxidase at 19.55 U/ml (Sigma Chemical Co.). The glucose oxidase step was repeated six times at 10min intervals. All reaction steps were carried out at 4°C to avoid cytoplasmic internalization of the labeling enzymes. The reaction was stopped by adding 12 ml of cold RPM1 supplemented with 5 mM L-cysteine hydrochloride (Sigma Chemical Co.). Cells were then washed twice with cold RPMI, twice with cold PBS before lysis.

Structurat analysis of the CD69 antigen

161

Viability of the cells was higher than 99% in all experiments.

three times with cold acetone. The pellets were dried under vacuum, and finally resuspended in IEFsample buffer and analyzed by 2D-PAGE.

Cell lysates were prepared by incubating the cells for 60 min at 4°C in 1 ml of T&buffered saline, pH 8, containing 1% NP40 detergent, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride (PMSF), 1 mg/ml BSA, according to the method of Owen et al. (1981). Insoluble material was removed by centrifugation at 100,OOOg for 30 min. Cell extracts were incubated for 2 hr at 4°C with killed Stuphyiococcus aureus Cowan 1 stain bacteria (SaCl) (Pansorbin, Calbiochem, La Jolla, CA) under rotation to remove material that could interfere with subsequent specific immuuoprecipitates. This step was repeated three times. Detergent-solubilized cell extracts were reacted with specific mAb, and the resulting immunoprecipitate recovered by incubation with SaCl. Washes were done several times with a buffer containing 125 mM Tris-HCl, 0.5 M NaCl, 0.5% NP40, 10 mM EDTA, pH 8.2, at 4°C. After elution from immunoabsorbent by adding 80 ~1 of SDS-sample buffer (Laemli, 1970) to the packed immunoabsorbents and boiling this for 5 min, the specific immunoprecipitate was analyzed by one-dimensional electrophoresis (SDS-PAGE) on 11% polyacrylamide gels by using the discontinuous buffer system according to Laemmli (1970). Immunoprecipitates eluted by adding IEF-sample buffer were analyzed by two-dimensional gel electrophoresis according to O’Farrell et al. (1977) (NEPHGE/ SDS-PAGE) (2D-PAGE). It must be stressed that to avoid protein carbamylation (Bobb et al., 1971), elution with IEF-sample buffer was performed at room temp (Kessler, 1981) for 30 sec. Employing these elution conditions, we obtained 80-90% elution efficiency (Carra, personal observation). For 2D-PM, molecules were first separated by SDS-PAGE in reducing conditions. Then subunits of various Cl.18 and El65 reactive molecules, labeled by the LPGO method, were separately cut from the dried gels, eluted, reduced, alkylated, and then pepsin digested. The digest was then analyzed by 2D-PM as described (Accolla, 1984). For Endo F digestion (see below), elution from immunoabsorbents was carried out by adding 70 ~1 of extraction buffer (100 mM Tri-HCl pH 7.5, I % SDS, 1% Z-ME) and then boiling for 5 min.

RESULTS

Endo F deglycosyiation

Antigen eluted from immunoabsorbent by extraction buffer, was diluted 10 times with 0.112 M sodium phosphate, pH 6.1, 55.56 mM EDTA, 1.12% NP-40, 1.12% 2-ME. Endo F (Boehringer Mannheim) was added to a final concn of 15 U/ml, and samples were incubated for 18 hr at 37°C in the presence of 1 mg/ml leupeptin, 1 mg/ml pepstatin and 1 mg/ml antipain (Sigma). After digestion, the samples were precipitated with 20% TCA (final concn) for 2 hr at 4°C. Antigens were pelleted on microfuge and washed

Cl.18 and E16.5mAb CD 69 antigen

recognize distinct epitopes of

The molecular nature of the structures recognized by Cl.18 and E16.5mAb was investigated on PMA activated PBL surface labeled with 1251by the lactoperoxidase catalyzed method. Cell extracts were immunoprecipitated with the two mAb and immunoprecipitates analyzed by SDS-PAGE. Molecules recognized by Cl .18 mAb consisted of two diffuse and equally heavily labeled bands of 54 kD and 50 kD, respectively (Fig. 1, A), while molecules recognized by E16.5 mAb consisted of one diffuse band of 50 kD (Fig, 1, B), under non reduced conditions. After reduction, the 54 kD and 50 kD bands of C 1.1&reactive molecules were resolved as two bands of 32 kD and 26 kD (Fig. 1, C). The 26 kD band was three times intensely labeled with respect to the 32 kD band, as determined cutting the respective bands on the dried gel and evaluating the amount of radioactive material. The 50 kD band of E16.5-reactive molecules was resolved as one band of 26 kD (Fig. 1, D). Moreover, after a longer exposure of the gel, an additional poorly labeled band of 43 kD was observed both in Cl.l8- andin E16.5-reactive molecules in non reducing conditions (data not shown), which after reduction migrated as single band of 21 kD (Fig. 1, C’, D’). Molecular weight and migration pattern of the antigen recognized by Cl. 18 mAb, in conjunction with the capacity of the antibody to trigger proliferative activity of PBL in the presence of PMA (see below) suggested that Cl.18 mAb could recognize the CD69 early activation antigens as defined by the IV Leukocyte Typing Workshop (Knapp et al. 1989). Indeed, when PMA activated PBL surface labeled with rz51 were lysed and cell extracts exhaustively depleted of CD69 molecules employing the MLR3 mAb (kindly provided by Cosulich and Risso, University of Genova), no residual activity of Cl. 18 mAb was observed. Moreover, after depletion of C1.18-reactive material, no residual activity of MLR3 mAb was retained (data not shown). These data clearly identified Cl. 18-reactive molecules as the CD69 activation antigen. E16.5 mAb recognizes a distinct epitope of CD69 antigen expressed only on a subunit of the complex. In order to investigate this point, crossabsorption experiments were performed. PMA activated PBL, were vectorially labeled with lz51by LPGO, lysed and cell extracts analyzed for reactivity with Cl.18 and E16.5 mAb before and after exhaustive depletion of the E16.5- and C1.18-reactive materials, respectively. Figure 2 shows, that after complete depletion of Cl. 1S-reactive molecules, no residual activity of

E. GEROSA et al.

162

32 26 -21

A

B

Fig. 1. Autoradiographs of 1 I % SDS-PAGE of unreduced (lanes A, B) and reduced (lanes C, D and C’, D’) ‘Z51-lactoperoxidase labeled Cl .18- (lanes A, C and C’) and E16.5-reacting molecules (lanes B, D and D’) from PMA activated PBL. C’ and D’ show p21 band from Cl.18 and E16.5 specific immunoprecipitates, respectively, after longer exposure of the gel. The electrophoresis was carried out by using the discontinuous Tris buffer system. M, x lo-’ of the major detectable bands and of mol. wt standards

(arrows) are indicated. El65 mAb was observed. In contrast, after exhaustive depletion of E16.5-reacting molecules, the reactivity of Cl. 18 mAb, although drastically reduced, was not completely abrogated. These residual molecules appeared as two relatively diffuse bands of 54 and 52 kD respectively, which under reducing conditions migrated as two bands of 32 kD and 26 kD.

The 32 kD band was three times intensely labeled with respect to the 26 kD band, suggesting that ~54 band represents a 32 + 32 kD disulfide linked dimer and ~52 band represents a 32 + 26 kD disulfide linked dimer. From these results we conclude that the CD69 molecules recognized by the Cl.18 mAb are composed of several cell surface structures in which there

93,

32

30 b

26

AB

C

D

E

F

G

H

Fig. 2. Autoradiographs of 11% SDS-PAGE of i2Wactoperoxidase labeled CD69 antigens from PMA activated PBL. The electrophoresis were carried out as described in Fig. 1. Lanes show CD69 molecules eluted from the following mAb immunoabsorbents: (A, E) C1.18; (B, F) E16.5; (C, G) C1.18, after exhaustive depletion of ElbS-reactive molecules; (D, H) E16.5, after exhaustive depletion of Cl.18 reactive molecules. Lanes A, B, C and D show the non reducing condition pattern of migration and lanes E, F, G and H the reducing condition, respectively. To maximize visual comparison, autoradiographs in lanes A, B, E and F were obtained by running a portion, approximately equivalent in counts, of the total immunoprecipitated material. Conversely, autoradiographs in lanes C, D, G and H represent the totality of immunoprecipitated counts obtained from corresponding samples. M, x lo-’ of the major detectable bands and of mol. standards (arrows) are indicated.

Structural

analysis

are at least a 32 + 32 kD, a 32 + 26 kD, a 26 + 26 kD and finally a 21 + 21 kD disulfide linked dimers. On the other hand molecules recognized by E16.5 mAb are composed only by 26 + 26 kD and 2 1 + 21 kD disulfide linked dimers. Finally, the cross-absorption experiments indicated that E16.5 epitope is expressed on a subpopulation of the 26 kD structure of CD69 antigen, i.e. in the 26 + 26 kD dimer, but not in the 32 + 26 kD dimer. 2D-PAGE and Endoglycosidase F C1.18- and E 16.5reactive molecules

treatment

of

To further investigate the biochemical nature of the Cl. 18- and E16.5-reactive molecules, 2D-PAGE analysis was performed on immunoprecipitates before and after treatment with Endoglycosidase F enzyme, which specifically cleaves both high mannose and complex N-linked oligosaccharides. Under reducing conditions E16.5-reactive molecules (Fig. 3, B) were composed of a set of 10 basic spots which migrated with an apparent molecular weight of 26 and 28 kD, respectively. The 21 kD structure, although poorly labeled, was detected in all experiments as a single basic spot (insert). C1.18-reactive molecules (Fig. 3, A) were composed of a set of spots that, both in isoelectric point and in apparent molecular weight, were super-imposable to the E16.5 specific spots. As in the case of E16.5, the ~21 NEPHGE

of the CD69 antigen

163

structure was always detected after longer exposure of the gel (insert). In addition, a set of 3-4 more acidic spots, migrating at an apparent molecular weight of 32 kD were observed. However, after Endo F deglycosylation, 2D-PAGE patterns of Cl. 18 and E16.5 specific immunoprecipitates were virtually superimposable: both Cl.18 and E16.5 molecules were composed of a set of 4-5 spots migrating with an apparent molecular weight of 22 kD and with a more basic isoelectric point with respect to their glycosylated components (Fig. 3, C and D). These results strongly suggest that Cl.18 and E16.5 mAb recognize the same molecular complex, but in two distinct glycosylation forms, both expressed at the cell surface membrane. 2D-peptide mapping

Structural characteristics of subunits recognized by Cl.18 and E16.5mAb were further investigated by 2D-PM analysis of PMA activated PBL from three normal donors. Figure 4 shows the fingerprints of p32 (panels A, B), p26 (panels C, D) and p21 (panels E, F) of Cl. 18-reactive molecules obtained from two normal donors, respectively. Comparison of corresponding ~32, p26 and p21 fingerprints demonstrated a striking degree of homology between donors and also among subunits of the same CD69 complex. Results obtained from a third donor were virtually

-

labeled CD69 Fig. 3. Two-dimensional gel electrophoresis in reducing conditions of ‘Z51-lactoperoxidase antigens from PMA activated PBL. First dimension was NEPHGE, with acidic end to the left and basic end to the right; second dimension was 11% SDS-PAGE from the top to the bottom. Autoradiographs A and B show the CD69 molecules recognized by Cl.18 and E16.5 mAb, respectively. Autoradiographs C and D show the same corresponding reactive molecules after Endo F deglycosylation treatment. Arrows in panels A and B point to the p21 spot that can be seen after longer exposure of the gels as depicted on the lower right inserts. M, x 1O-3 of the major detectable spots are indicated.

F. GEROSAet al.

164

CHROMATOGRAPHY

-

Fig. 4. Autoradiographs of peptide maps of ‘251-lactoperoxidase-labeled p32 (panels A, B), ~26 (panels C, D), and p21 (panels E, F) of CD69 complex obtained from PMA activated PBL derived from two normal donors, respectively. (data not shown). These data indicate the lack of polymorphism in the CD69 complex. 2D-PM fingerprints of ~26 and p21 subunits of CD69 molecular complex recognized by the E16.5 mAb were virtually superimposable to those of corresponding ~26 and ~21 subunits of C1.18” reactive molecules (data not shown). This finding further supports the conclusion that E16.5 and Cl. i 8 mAbs recognize the same protein structure but in different post-translational modifications. superimposable

CD49 surface expression in ”freshly _ isolated PBL

Previous reports have suggested the absence of expression of CD69 antigen on resting lymphocytes (CebriBn et al. 1988). Indeed no CD69+ cells could be detected by indirect immu~o~uorescence with Cl.18 mAb. However, when freshly isolated PBL

were surface labeled with i25Iby the LPGO catalyzed method and cell lysates were immunoprecipitated with C 1.18 mAb, canonical CD49 antigen was found on these cells (Fig. 5), at very reduced amounts (200 times less than on PMA activated PBL). That the cells analyzed are indeed resting PBL could be demonstrated by the fact that the CD25 activation antigen could not be found after exhaustive immunoprecipitation with MAR108 mAb (CebriBn et al. 1988), under the same experimental conditions (Fig. S), nor by indirect immunofluores~n~. Functional activity of Cl.18

crnd E16.5 mAb

Cl.18 and El 6.5 mAb induced a comparable proliferative activity and IL-2 production in PBL in the presence of PMA (Table 1). Moreover, the proliferative response increased linearly as a function of

Structural

analysis

of the CD69 antigen

165

parameter of activation TNF production was evaluated in SN after 18 hr of culture. Results obtained with 5 representative NK cell clones are reported in Table 2. Sepharose-linked Cl. 18 mAb did not induce TNF release in the presence of either rIL-2 or PMA, although both rIL-2 and PMA treatments induced CD69 antigen expression in all the clones tested (see legend to Table 2). On the contrary, the same clones produced high amounts of TNF in presence of Sepharose-linked B 16.15 (anti-CD1 6) mAb. It can therefore be concluded that under these culture conditions CD69 antigen is not involved in functional triggering of NK cells leading to production of TNF. DISCUSSION

A

c

6

Fig. 5. Cell surface expression of CD69 antigen on resting PBL. Autoradiographs of 11% SDS-PAGE of reduced rz51-LPGO labeled C1.18-reactive molecules on PMA activated (lane A) and resting PBL (lane B). Lane C indicate the control immunoprecipitation with MAR108 (anti-CD25) mAb. To maximize visual comparison, the autoradiograph in lane A has been obtained by running a portion of the total immunoprecipitated material. Conversely, autoradiographs in lane B and C represent the totality of immunoprecipitated counts obtained from the corresponding samples. M, x 10e3 of the major detectable band and of mol. wt standards (arrows) are indicated.

increasing mAb concn (from 20 ng/ml up to 50 pg/ml), irrespective of the mAb used (data not shown). One of the earliest response of lymphocytes to surface receptor stimulation is a rise in [Ca*+],. Figure 6 shows that Cl. 18 mAb increased [Ca’+], after cross-linking with goat anti-mouse Ig antibody. The increase induced by E 16.5 mAb, after cross-linking, was consistently lower than [Ca*+], rise induced by Cl.18 mAb. It has been reported that CD 16 cross-linking at the cell membrane induces activation of NK cells (Anegbn et a/. 1988). To evaluate whether CD69 might be an activatory molecule for NK cells, 10 NK cell clones were stimulated with Sepharose-linked Cl.18 mAb in the presence of rIL-2 or PMA. As Table I. Cl.18 and E16.5 mAb induced comparable proliferative activity and IL-2 production in PBL in the presence of PMA _ [‘H]TdR Incorporation (cpm x lo-“) IL-2 Production (U/ml)

PMA (2 “g/ml) Cl.18 E16.5

5.0

42.6

40.0

0

20

16

PBL (2 x IO5 cells/well) were cultured in the presence of Cl.18 or E16.5 mAb (lO~g/ml) or 0KT3 mAb (l/4 of culture SN). IL-2 production and proliferative were assayed on the same cultures after 2 and 3 d of respectively. Results are expressed as means from triplic& cultures. results were obtained in three different experiments.

OKT3 52.1 50 purified dilution activity culture, Similar

The detailed structural analysis of the CD69 antigen expressed at the surface of PMA activated PBL was performed by means of the two distinct Cl. 18 and E16.5 mAb. Purification of CD69 molecules after vectorial labeling of externally disposed plasma membrane proteins was selected as a method to specifically analyze mature CD69 molecules at the structural level by SDS-PAGE and by 2D-PAGE analysis and thus minimize problems due to comparison of molecular entities at various steps of biosynthesis. From our studies CD69 appears as a complex of at least four disulfide linked dimers of 32 + 32 kD, 32 + 26 kD, 26 + 26 kD and 2 1 + 21 kD, coexpressed on the cell membrane. However, while Cl.18 mAb immunoprecipitated the entire family of CD69 molecular forms, E16.5 mAb immunoprecipitated only a portion of them. Analysis by 2D-PAGE indicated that CD69 complex is a heavily glycosylated structure both as quantities and as different types. In this structure Cl. 18 mAb recognized the entire costellation of molecules included the more glycosylated forms, composed by addition of acidic sugar moieties, while E 16.5 mAb recognized less extensively glycosylated structures. Furthermore after Endo F deglycosylation both Cl. 18- and E 16.5-reactive molecules showed similar patterns of migration. 2D-PAGE analysis after Endo F treatment revealed that C1.18- and E16.5-reactive molecules still retained a well characterized heterogeneity as isoelectric point, suggesting the existence of other sugar moieties Endo F insensitive (O-linked sugars) or, more unlikely, protein structure microheterogeneity. Carbamylation processes might constitute a source of heterogeneity in isoelectric point. However, under our experimental conditions (i.e. complete elution at room temp and for very short time, by urea), it is reasonable to exclude the production of cyanate ions, from degradation of urea, which reacts with N-terminal residues to produce carbamoyl adducts to the proteins. Structural analysis by 2DPM of CD69 molecular complex at the level of single subunit resulted in a striking homology not only among ~32, p26 and p21

F. GEROSAet al.

230 -

100 -

--

t

t

Cl.13 (7w/ml)

GaM (3OWrnl1

(6)

E.a+], (nM) 2mln H

06.5

(7avnll)

GIM (%g/ml1

Fig. 6. Comparison of the effects of Cl.18 and E16.5 mAb on [Car+], in PMA activated PBL. (A) cells, incubated for 24 hr with PMA (2 ng/ml), were loaded with fura-2. Where indicated by arrows Cl.18 mAb, affinity purified goat anti-mouse IgG (GaM, Cappel, West Chester, CA) and PHA (1% v/v) were added. On the left the [CaZ+],calibration of fura- signal is shown. (B) Conditions as in (A), but the scale of the time was compressed two fold. Where indicated by arrows E16.5 mAb and GaM were added. Because of the slow kinetic of E16.5 induced [Ca*+], rise the cells were also treated with 200 PM sulphinpyrazone (Di Virgilio et al. 1988) to avoid leakage of fura-2. In control experiments it was demonstrated that sulphinpyrazone did not affect the rise of [Ca*+], by either mAb. Typical traces repeated in at last three occasions on two donors. within the same individual, but also among homologous subunits in distinct individuals. High similarity between peptide digests of subunits like ~32, p26 and ~21, with strong differences in mol. wt, strongly suggests that the protein backbone of the various subunits is highly similar, if not identical. It further reinforces the notion (see results with Endo F) that differences in mol. wt are the result of glycosylation processes and that this glycosylation does not interfere with the capacity of the various subunits to be labelled by LPGO iodination. From these results we conclude that CD69 is a nonpolymorphic molecular complex resulting of a core structure of 21-22 kD disulfide linked homodimer which has the potential to be expressed on the cell surface under distinct glycosylated forms. Thus

21 + 21 kD dimers, 26 + 26 kD dimers as well as more mature 32 + 26 kD and 32 + 32 kD dimers may be expressed. It derives that antibodies of different specificities can be found which discriminate the various molecular species. A likely hypothesis to explain our results could therefore be that E16.5 mAb recognizes an epitope on the less mature form of CD69 complex (the 26 + 26 kD dimer) which is lost during the terminal maturation of one unit of the dimer toward the 32 kD higher glycosylated subunit. On the contrary, the Cl.18 mAb would recognize an epitope on the 26 kD subunit (either one) which is not modified by the glycosylation process. Previous results showed that CD69 antigen was not expressed on resting PBL (Cebrian et al. 1988). By indirect immunofluorescence we could not detect

Structural analysis of the CD69 antigen Table 2. Effect of sepharose linked Cl.18 mAb on TNF production by NK cell clones Inducers [IL-2b rIL-2 + sepharoseB9.12.1 riL-2 + sepharose-B 16.15 rIL-2 + sepharose-Cl.18 PMA” PMA + sepharose-B9.12.1 PMA t sepharose-Bl6.15 PMA + sepharose-Cl. 18

TNF (pgiml) NK.57’ NK60 NK51 NK48 0 0 1SO0 0 70 ndd 1000

100

0 0 1400 0 140 nd 1000 190

0 0 1800 0 250 200 3500 350

4.5 20 2000 45 300 300 2400 320

NK43 0 10 2000 10 60 100 1200 170

WKS7, NK60, NKSl, NK48 and NK43 are CD3--, CD16+ cytolytic cell clones derived from three normal donors. Cells were cultured in duplicate and SN collected after 18 hr. CD69 antigen expression was evaluated on the cells after culture with rIL-2 or with PMA. In all cases cells were more than 90% CD69+ with mean fluorescence intensity ranging from 250 to 580. Mean background fluorescence was always less than 15. %IL-2: 250 U/ml. ‘PMA: 1 ng/ml. “nd: not determined.

appreciable amounts of CD69 antigen on the cell surface of freshly isolated PBL. However, after radiolabeling and immunopr~ipitation a low but detectable expression of CD69 antigen was found. At the moment it is not possible to exclude that an exiguous number of activated cells might contribute to such expression. However, since CD25 antigen, whose presence directly correlates with a state of cell activation, could not be found by immunoprecipitation under the same experimental conditions, we believe that CD69 expression detected in our study reflects the presence of low amounts of this antigen on cell surface of resting cells. Comparison of functional activity of Cl.18 and E16.5mAb showed that triggering of PBL in the presence of PMA via a subpopulation of CD69 molecules (E16.Sreacting molecules) results in cell proliferation and IL-2 production comparable to those obtained by triggering via the entire CD69 molecular pool (C 1,18-reactive molecules). Early [Caz+li has been reported after cross-linking of CD69 antigen (Nakamura &f al. 1989; Testi et af. 1989). Optimal [Ca”], rise was obtained only by Cl. 18 mAb mediated triggering. Differences in [Ca2+Ji rise induced by the two mAb might be explained by a differential cross-linking of CD69 complex at the cell surface. The functional role of CD69 has been described for T and B lymphocytes {Cebrian et ai. 1988; Risso ei al. 1989); however no functional studies have been reported on NK cells. Our results directly demonstrate the lack of functional involvement of CD69 complex at least for the capacity of NK cells to produce TNF after triggering of relevant celi surface antigens. Other activities of NK cells are now under study to exclude that CD69 antigen is a molecule involved in other activation pathways of NK cells. In this regard, preliminary experiments on cytolytic activity of NK effector cells against the P815 cell line in the presence of Cl.18 and E16.5 mAb, further strengthen the hypothesis that CD69

167

antigen expressed on NK cells is not associated to a functional activity. ~c~~Q~Ze~ge~e~z~-We thank Centro Trasfusionaie of Policlinico Borg0 Roma for normal blood samples. This work was supported in part by CNR grants “Oncology” and “Biotechnology”, by grants from AIRC, by Regione Veneto and Minister0 Pubblica Istruzione.

REFERENCES

Accoila R. S. (1984) Analysis of the structural heterogeneity and polymorphism of human Ia antigens. J. exp. Med. 159, 3788393. Anegdn I., Cuturi M. C., Trinchieri G. and Perussia B. (1988) Interaction of Fc Receptor (CD16) ligands induces transcription of interleukin 2 receptor (CD25) and lymphokine genes and expression of their products in human natural killer cells. J. exp. Med. 167, 452472. Bjorndahl J. M., Nakamura S., Yung L. K. L. and Fu S. M. (1988) The 28 kD/32 kDa activation antigen EA 1. Further characterization and signal requirements for its expression. J. Zmmun. 141, 40944100. Bobb D. and Hofstee B. H. J. (1971) Gel isoelectric focusing for following the successive carbamylation of amino groups in ch~otrypsinogen A. Analyz ~~ochem. 40, 209-217. Carra G. and Accolla R. S. (1987) Structural analysis of human Ia antigens reveals the existence of a fourth molecular subset distinct from DP. DQ and DR. molecules. J. exp. Med. 165, 47-63. Cebribn M., Yagiie E., Rinccm M., Lopez-Botet M., De LandLuri M. 0. and Sanchez-Madrid F. (1988) Triggering of T cell proliferation through AIM, an activation inducer molecule expressed on activated human lymphocytes. J. exp. Med. 168, 1621-1637. Cosulich M. E., Rubartelli A., Risso A., Cozzolino F. and Bargellesi A. (1987) Functional characterization of an antigen involved in an early step of T-cell activation. Proc. natn. Acad. Sci. U.S.A. 84, 4205-4209. Di Virgilio F., Fasolata C. and Steinberg T. H. (1988) Inhibitors of membrane transport system for organic anions block fura- excretion from PC12 and N2A cells. Biochem. J. 256, 959-963. Gerosa F., Tommasi M., Spiazzi A. L., Azzolina L. S., Carra G.. Maffei A.. Accolla R. S. and Tridente G. (1988) Heterogeneity of lymphokine-activated killer (LAK) populations at the clonal level: both NK and CD3+, CD4-, CDS- clones efficiently mediate tumor cell killing. &Yin. Immun. imm~opat~oI. 49, 91-100. Gerosa F., Scardoni M., Tommasi M., Medaina N., Libonati M., Tridente G. and Carra G. (1989) Role of CD16 in human NK clones analyzed by a new monoclonal antibody. J. Immunol. Res. 1, 133-138. Hara T., Yung L. K. L., Bjorndahl J. M. and Fu S. M. (1986) Human T cell activation. III. Rapid induction of a phospho~lated 28 kD/32 kD disulfide-linked early activation antigen (EA I) by 12-o-tetread~anoyl phorbol-13acetate, mitogens and antigens. J. exp. Med. 164, 1988-2005. Haynes B. F., Hemler M. E., Mann D. L., Eisenbarth G. S., Shelhamer J., Mostowski H. S., Thomas C. A., Strominger J. L. and Fauci A. S. (1981) Characterization of a monoclonal antibody (4F2) that binds to human mono~ytes and to a subset of activated lymphocytes. J. Immun. 126, 1409-1414. Helderman J. H. and Strom T. B. (1978) Specific insulin binding sites on T and B lymphocytes as a marker of cell activation. Nature 274, 6263. Kessler S. W. (1981) Use of Protein A-bearing Staphylococci for the immunoprecipitation and isolation of antigens from cells. Meth. Enzymof. 73, 442-459.

168

F. GEROSA et al.

Knapp W., Rieber P., D&ken B., Schmidt R. E., Stein H. and van der Borne A. E. G. K. (1989) Towards a better definition of human leucocvte surface molecules. Immun. Today 10, 253-258. . Laemmh U. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage - _ T4. Nature 227. 680-685: Lanier L. L., Buch D. W., Rhodes L., Ding A., Evans E., Barney C. and Phillips J. H. (1988) Interleukin 2 activation of natural killer cells rapidly induces the expression and phosphorylation of the Leu-23 activation antigen. J. exp. Med. 167, 1572-1585. Lazarovits A. I., Moscicki R. A., Kurnick J. T., Camerini D., Bhan A. K., Baird L. G., Erikson M. and Colvin R. B. (1984) Lymphocyte activation antigens. I. A monoclonal antibody, anti-Act I, defines a new late lymphocyte activation antigen. J. Immun. 133, 185771862. Leonard W. J., Depper J. M., Uchijama T., Smith K. A. Waldmann T. A. and Green W. C. (1982). A monoclonal antibody appears to recognize the receptor for human T cell growth factor; partial characterization of the receptor. Nature 300, 267-269. Nakamura S., Sung S. J., Bjorndahl J. M. and Fu S. M. (1989) Human T cell activation. IV. T cell activation and

proliferation via the early activation antigen EA 1. J. exp. Med. 169, 677689. O’Farrell P. Z.. Goodman H. H. and O’Farrell P. H. (1977) High resolution two dimensional electrophoresis of basic as well as acidic proteins. Cell 12, 1133-l 142. Owen M. J.. Kissonerghis A. M.. Lodish H. F. and Crumpton M. J. (1981) Biosynthesis and maturation of HLA-DR antigens in vivo. J. biol. Chem. 256, 8987-8993. Risso A., Cosulich M. E., Rubartelli A., Mazza M. R. and Bargellesi A. (1989) MLR3 molecule is an activation antigen shared by human B, T lymphocytes and T cell precursors. Eur. J. Immun. 19, 323-328. Schenkein I., Levy M. and Uhr J. M. (1972) The use of glucose-oxidase-as a generator of H,O, in the enzymatic radioiodination of components of cell surface. Cell. Immun. 5, 490493. Testi R., Phillips J. H. and Lanier L. L. (1989) T cell activation via Leu-23 (CD69). J. Immun. 143, 1123-l 128. Treves S., Di Virgilio F., Cerundolo V., Zanovello P., Collavo D. and Pozzan T. (1987) Calcium and inositolphosphates in the activation of T cell-mediated cytotoxicity. J. exp. Med. 166, 3242. Trowbridge I. S. and Omary M. B. (1981) Humancell surface glycoprotein related to cell proliferation is the receptor for transferrin. Proc. nafn. Acad. Sci. U.S.A. 78, 3039-3043.