Development and characterization of a human monoclonal antibody probably detecting the leukocyte differentiation antigen CD39 U
N. Ikewaki, S. Takabe, K. Tsuji. Development and characterization of a
Nobunao Ikewaki*, Sachiko Takabe
human monoclonal antibody probably detecting the leukocyte differentiation antigen CD39. Tissue Antigens 1992: 39: 174-181.
and Kimiyoshi Tsuji
Abstract: A human monoclonal IgM antibody, referred to as TU223, has been produced. The reactivity of TU223 was tested in various cells and cell lines by complement-dependent microcytotoxicity test and fluorescenceactivated cell sorter analysis. The antigen defined by TU223 was expressed on Epstein-Barr virus-transformed B-cell lines and on some Burkitt’s lymphoma cell lines, but was not expressed on normal T cells, B cells or erythrocytes. In addition, expression of the antigen defined by TU223 was also induced on B cells activated by Epstein-Barr virus or pokeweed mitogen, and on T cells activated by phytohemagglutinin, concanavalin A, pokeweed mitogen or recombinant interleukin-2. However, no expression of the antigen detected by TU223 was induced at all on recombinant interleukin-4-treated B cells or macrophage-like cell line U937. When the ability of TU223 and various mouse monoclonal antibodies to bind to human differentiation antigens was compared, interestingly, the reactivity of TU223 was found to be very similar to that of mouse monoclonal antibody CD23 (H107), which reacts with Fcs receptor 11. Two-color analysis revealed that the antigen defined by TU223 is expressed on the cell surface of certain lymphoid cells expressing CD23 antigen. However, it can be concluded that the antigen defined by TU223 is clearly distinct from Fcs receptor 11, based on assay of cross-blocking between HI07 and TU223. The surface antigen on B85 cells recognized by TU223 had the molecular size of 80-82 kiloDaltons as determined by immunoblotting analysis. In biological function, TU223 inhibited the generation of plaqueforming cells when added to a pokeweed mitogen-derived IgG-secreting system on days 0 to 4. These findings suggest that TU223 recognizes a novel and functional antigen expressed on certain activated lymphoid cells.
ttroduction Over the past several years, many mouse monoclonal antibodies (mAb) which react with human differentiation antigens on a variety of cells have been established and characterized. These mAb have played an important role in analyzing cell activation a n d / or differentiation mechanisms (1). On the other hand, human mAb t o a variety of * Present address: Department of Microbiology, Kitasato University School of Nursing, Kitasato, Sagamihara, Kanagawa 228, Japan. 174
Department of Transplantation, Tokai University School of Medicine, Bohseidai, Isehara, Kanagawa, Japan
Key words: a novel antigen - human monoclonal antibody - lymphoid cells Received 31 October, accepted for publication 15 November 1991
antigens have also been established using a number of techniques, and these human mAb have been
very useful clinically (2-8). We recently succeeded in establishing an EpsteinB a n virus (EBV)-transformed B-cell line, derived from B cells obtained from the peripheral blood mononuclear cells (PBMCs) of a pregnant woman, which steadily secretes a human IgM mAb. In this report, we describe the properties of this human mAb, referred to as TU223.Based on its reactivity, it was found that TU223 recognizes a novel and functional antigen expressed on certain activated lymphoid cells.
A new human monoclonal antibody Material and Methods Antibodies
Mouse mAb CD19 (B4) (9), CD20 (Bl) (lo), CD21 (B2) (ll), B5 (12) were purchased from Coulter Immunology Laboratories (Hialeah, FL). CD25 (IOT-14) (13) was purchased from Immunotech Laboratories (Marseilles). CD5 (NU-T PAN) (14), CDlO (NU-Nl) (15), CD23 (H107) (16) were purchased from Nichirei Laboratories (Higashimurayama, Tokyo). CD24 (OKB2) (17), CD38 (OKTlO) (18) were obtained from Ortho Diagnostic System Laboratories (Raritan, NJ). HU20 (,19) was kindly provided by M. Aizawa, M.D. (Department of Pathology, Hokkaido University School of Medicine). YU3-20A (20) or Sa3 (21) were produced in our own laboratory. The specificities of the mAb used in this study are summarized in Table 1. Fluorescein isothiocyanate (F1TC)-conjugated goatanti-mouse IgF(ab’)2, FITC-conjugated goat-antihuman IgMF(ab’)2, non-conjugated goat-anti-human IgG, goat-anti-human IgM, goat-anti-human IgA, horseradish peroxidase (HRP0)-conjugated goat-anti-human IgF(ab’)2 and phycoerythrin (PE)-conjugated goat-anti-mouse IgF(ab’)2 were purchased from Cappel Laboratories (Malvern, PA). Preparation of lymphoid cells
Peripheral blood mononuclear cells (PBMCs) obtained from 37 healthy donors were isolated by Ficoll-Paque (Pharmacia, Uppsala) density sedimentation as previously described (22). T and B cells were separated into erythrocyte (E) rosettepositive (E+)and E rosette-negative (E-) populations using sheep erythrocytes (23). Human cell lines
All human cell lines used in the present investigation, except EBV-transformed B-cell lines and Burkitt’s lymphoma BT-l, were supplied by the Japanese Cancer Research Resource Bank (Yoga, Tokyo). EBV B-cell lines B85 and HID were established by J. Minowada, M.D. of New York State University and in our own laboratory, respectively. Other EBV B-cell lines were exchanged at the 10th International Histocompatibility Workshop Meeting. Burkitt’s lymphoma cell line BT-1 was kindly provided by H. Yakura, M.D. of the Department of Pathology, Asahikawa Medical College.
Flow cytometric analysis The binding activities of the mAb were analyzed by direct or indirect immunofluorescence using a
fluorescence-activated cell sorter (FACStar, Becton Dickinson, Mountain View, CA) as previously described (24). In some experiments, the cellular distribution of antigens detected by mAb was determined by two-color flow cytometry. Complement-dependent microcytotoxicity test
Complement-dependent microcytotoxicity test was performed using a standard microcytotoxicity test procedure employing Terasaki typing trays and rabbit complement as previously described (25). Establishment of a human mAbproducing 8-cell line
PBMCs obtained from a pregnant woman, SU, were isolated by Ficoll-Paque density gradient centrifugation. B cells were separated on the basis of rosette formation. Enriched B cells were suspended in complete RPMI 1640 medium (GIBCO, Grand Island, NY) supplemented with 10% fetal calf serum (FCS), 100 units/ml penicillin, 100 pg/ml streptomycin and 2 mM L-glutamine. The suspension was distributed in 24-well plates (Costar, Cambridge, MA) with approximately lo5 cells in 900 pl. In order to transform the B cells, the filtered EBV-containing culture supernatant from marmoset cell line B95-8 was added to each well, and incubation was performed for 60 min at 37°C. EBV-transformed B cells began to appear, usually within 3 weeks. The amount of human Ig in the culture supernatant was determined by enzymelinked immunosorbent assay (ELISA) (24). The transformed cells were cloned by limiting dilution, and selected cells were then recloned twice, resulting in the establishment of a cell line which has been steadily producing large amounts of human IgM, even in a new culture medium, GIT (Wako Pure Chemistories, Tokyo), deprived of FCS. This Table 1. Specificity of the mouse mAb used in this study CD no.
mAb
specificity
lg class
CD5 CDlO CO19 CD20 C021
NU-T PAN NU-N1 84 B1 B2 H107 OKB2 IOT-14 OKTlO 85 Sa3 HU20 YU3-20A
Pan-T cells CALW Early B cells Pan-B cells BceU subset FCZ receptor II Pan B cells 11-2 receptor (achain) T progenitorfl blasts Activated B cells HLA-DR.DQ0P HLA-OR HLA-A,B,C
lgG2a IgG1 lgGl lgG2a IgM lgG2b lgG1 lgG2a lgG1 IgM 1963 lgG2b IgM
CD23
CD24 CD25 CD38
-
*
CALLA; common acute lymphoblastk leukemia antigen.
175
Ikewaki et a]. Table 2. Binding of TU223 to cultured cell lines Degree of binding of mAb”
Cell lines
TU223
H107
0 0 0
0 0
+
++ + + +
T-cell leukemia GEM MOLT4 MOLT3 EBV h e l l lines 885 WAR BM16 RML . AKlBA Kl12 MANN HID Burkitt’s lymphoma BT-1 Raji Daudi Ramos B-cell ALL
CCRF-SB Erythroid cells K562 Myeloid cells u937 THP1
NU-N1
84
85
B1
82
OKB2
Om10
HU20
Sa3
N F
0 0 0
0
NT
+
NT
0 0
0 0
0
0
0 0
++
NT NT
NT 0
NT
0 0
++
0 0
+
NT
NT
++
0
+
0
+
f
+ +
++ ++
0
0
0 0 0
++
+
++ ++ ++ ++
+
++ ++ ++
0 0 0 0 0
NT NT NT
++ ++ ++
++ ++ ++ ++
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
+ +-
++ ++
+
++ ++
++ ++
0 0
0 0
0
++
++
0 0 0
+
NT NT NT NT
+
+
+ + + + 4.
0
++
0
0
+
+ +
f
+
+
NT
+
0 0
NT
r+
+
+
NT
++
++ +
0
0
0
0 0
0 0 0
+
+
0 0
0 0
0 0 0
0
0 0 0 0
+ + +
+
0 0
+
+ 0
+
0
0
++ ++ ++ 0
++
++ ++ ++
++
++ ++ ++
YU3-20A IOT-14 NU-TPAN
0 NT
+ +
+ + +
++ ++ ++
++ ++
+
0
+
+
++
0
0
0
++
++
++
0
NT
0
0
NT
NT
0
NT
NT
NT
NT
NT
0
0
NT
NT
0
0 0
NT NT
NT NT
0
0 0
NT NT
NT NT
NT NT
NT NT
0
0
++ ++
NT 0
0 0
0
+
+
++
0
0
Binding activities of mAb were determined by complement-dependent microcytotoxicity test or a flow cytometry. Binding of mAb was scored as follows: (0) indicates < lo%, ( A )indicates 11 to 20%. (+) indicates 21 to 70%. (++) indicates > 71%. ** NT; not tested. *
mAb was assigned the designation TU223*TU223 has been maintained for more than 4 yr without loss of Ig-secreting activity, The isotype of TU223 is IgM. The binding titer of TU223 to B85 cell line was 1 5 4 in the unconcentrated culture supernatant (data not shown).
Assay of cross-blocking between CD23 (H107) and TU223
Culture with activators
Reverse hemolytic plaque assay for IgG-secreting cells
PBMCs in 1 ml of complete RPMI 1640 medium were put into 24-well plates and cultured with 1 pg/ml phytohemagglutinin (PHA) (DIFCO), 10 pg/ml concanavalin A (ConA) (Sigma), pokeweed mitogen (PWM) (GIBCO) in a 1:lOO dilution and 1 U/ml recombinant interleukin-:! (rIL-2) (Shionogi Pharmacy, Osaka) at 37°C. After activation, T or B cells were isolated on the basis of rosette formation. On the other hand, B cells were cultured in 900 p1 complete RPMI 1640 medium in 24-well plates with 100 pl EBV or 100 U/ml recombinant interleukin-4 (rIL-4) (Genzyme). U937 cells were also treated with 100 U/ml rIL-4. Expression of the antigen defined by TU223 and CD23 antigen was detected by complement-dependent microcytotoxicity test or two-color flow cytometry.
PBMCs (1 x lo’) were isolated by Ficoll-Paque density gradient centrifugation and cultured in triplicate in v i m for 7 days with PWM (1 pg/ml) in the presence (10 pg/ml) or absence of TU223, CD23 (H107) or Sa3. On d 7 the cells were harvested and washed in RPMI 1640 medium. IgG-secreting cells were determined by the reverse plaqueforming cell (PFC) assay with some modifications as previously described (26).
176
In order to determine whether TU223 reacts with an epitope of CD23, a cross-blocking assay was performed by FACS analysis as previously described (24).
Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting
EBV-B cell line B85 or Burkitt’s‘lymphoma cell line Ramos were lysed in lysis buffer (50 mM TrisHCI pH 7.4, 150 mM NaCl, 5 mM EDTA, 0.5%
A new human monoclonal antibody
NP40, 1 mM PMSF). After centrifugation, the solubilized proteins were separated on 10% SDS gel and transferred electrophoretically to nitrocellulose sheets, which were then incubated with TU223, CD23 (H107) or B5. After incubation, immunoperoxidase staining was performed using an avidinbiotin complex (ABC) kit (Vector Laboratories Inc., Burlingame, CA), followed by visualization by the diaminobenzidine chromogen reaction. Control immunoblotting was also performed using non-reactive normal human IgM, mouse IgG or IgM.
Table 3. Binding activity of TU223 to normal lymphoid cells and erythrocytes
96 of positive cells' mAb
NU-T PAN 81 TU223
T cells ( ~ 3 7 ) 95 2 0.1
B cells (17-37)
Degree of agglutination" Erythrocytes (A,
B. 0, AB)
3 64 0.7
Analyzed by complementdependent rnicrocytotoxicity test. The average percentage of positive cells in 37 healthy donors is shown. * Analyzed using an agglutination test (-); negative.
Results Binding activity of TU223 to a panel of cultured cell lines
The reactivity of TU223 was assessed using a panel of 19 established -cell lines and complement-dependent microcytotoxicity test or an immunofluorescence assay. Table 2 shows that all EBV B-cell lines used in this experiment and certain Burkitt's lymphoma cell lines, Raji and BT-1, were found to bind strongly. The reactivity of TU223 with these cell lines was found to be similar to that of CD23
(H107), but not CD5 (NU-T PAN), CDlO (NUNl), CD19 (B4), CD20 (Bl), CD21 (B2), B5, CD24 (OKB2), CD25 (IOT-14), CD38 (OKTlO), HU20, YU3-20A or Sa3. The cellular distributions of CD23 and the antigen defined by TU223 were also tested by two-color analysis and expression of the antigen defined by TU223 was detected on the cell surface of certain lymphoid cells bearing CD23 antigen (Fig. 1). Reactivity of TU223 with normal lymphoid cells and erythrocytes
AKIBA
The binding activity of TU223 to T or B cells obtained from 37 healthy donors was determined by complement-dependent microcytotoxicity test, and binding to erythrocytes (human) was tested by means of the agglutination test. All lymphoid cells and erythrocytes yielded negative results with regard to response to TU223 (Table 3). Binding activity of TU223 to activated cells
8-85
Reactivity of TU223 to activated cells was tested by means of the complement-dependent microcytotoxicity test. Table 4 shows that percentages of cytotoxic activity of TU223 against T cells activated by PHA, ConA, PWM or rIL-2 were approximately 58%, 48%, 70% and 35%, respectively, after 5 d. On the other hand, percentage of cytotoxic activity of TU223 against B cells activated by PWM was approximately 65%. Binding activity of TU223 to rlL-4-treated B or U937 cells
GREEN FLUORESCENCE Tu223 Figure 1. Twocolor analysis of TU223 binding to certain cultured B-cell lines. The cells were stained with CD23 (H107)and TU223 in the first stage. PE-conjugated goat-anti-mouse IgG or FITC-conjugated goat-anti-human IgM, respectively, were used in the second stage. Relative fluorescence intensity was expressed by logarithmic amplification.
Reactivity of TU223 with rIL-&treated B or U937 cells was tested by two-color flow cytometric analysis. Fig. 2 shows that CD23 expression was strongly induced on both B and U937 cells after treatment with rIL-4; however, no expression of the antigen defined by TU223 was induced on these cells at all. 177
Ikewaki et al. Table 4. Cytotoxic activity of TU223 against T or B cells actiiated in viffo Cells obtained from culture with
% of positive cells’
T cells
Medium only PHA ConA PWM rll-2
(n-5) (n-5)
5 58
(n=5) (n-5)
4a
(n-5)
B cells
means of FACS analysis. As shown in Fig. 4, CD23 (H107) did not block the reactivity of TU223 with AKIBA cells at all. Thus, these findings indicate that TU223 recognizes an epitope distinct from that of CD23 antigen. .
a NT NT 65
70 35
NT
Analyzed by complementdependent microcytotoxicity test, average percen5 healthy donors are shown. * NT; not tested. tages of positive cells in
Kinetics of expression of the antigen defined by TU223 on EBV blasts and on PWM-activated B cells
We examined the kinsics of expression of the antigen defined by TU223 on EBV blasts and on PWM-activated B cells, Expression of the antigen defined by TU223 was detectable on B cells apProximately 10 to 12 d after activation by EBy On the other hand, it was found that expression of this antigen on PWM-activated B cells clearly occurred about 2 to 3 d after activation (Fig. 3).
Inhibition of
We investigated the possibility that the antigen defined by TU223 plays a regulatory role in the differentiation of resting B cells to IgG-secreting cells. As shown in Table 5, the presence of TU223 markedly reduced the PWM-derived induction of PFC when added to cultures on d 0 to 4. However, CD23 (H107) did not inhibit PFC inductions. On the other hand, it was shown that addition of Sa3 to this system on d 0 to 2 induced significant inhibition of the development of PFC. Biochemical characterization of the cell surface antigen recognized by TU223
Solubilized proteins from B85 or Ramos cells were electrophoresed on 10% polyacrylamide gel, trans-
x
Evidence that the antigen defined by TU223 is distinct from CD23 antigen
PFC induction
A
01
wsilin clllr
/-
In order to test whether TU223 reacts with CD23 antigen, a cross-blocking assay was performed by
dm
after xtiration
B
medium only
rlL-4(100U/ml)
.
“ S 0
1
2
1
4
5
dap allcr a c l w a l m
GREEN FLUORESCENCE W223 Figure 2. Expression of the antigen defined by TU223 on B and U937 cells after treatment with rIL-4. Cells treated with rIL-4 (100 Ulml) were stained with HI07 and TU223 in the first stage. PE-conjugated goat-anti-mouse IgG or FITC-conjugated goat-anti-human IgM, respectively, were used in the second stage. Fluorescence intensity was tested by FACStar.
178
Figure 3. Kinetics of cellular expression of the antigen defined by TU223. The kinetics of induction of expression of the antigen defined by TU223 on B cells activated by EBV (panel A) and B cells activated by PWM (panel B) is shown. The percentage of positive cells was determined by complement-dependent microcytotoxicity test. The average percentage of positive cells in 2 healthy donors is shown.
A new human monoclonal antibody
Discussion
A
i
B
In the present study, we succeeded in establishing an EBV-transformed B-cefl line that has been producing a human n A b referred to as TU223 for more than 4 yr. When the reactivity of TU223 was tested on various cells and cell lines, it was found that the antigen defined by TU223 was expressed on certain Burkitt's lymphoma cell lines, on all EBV B-cell lines used in this experiment, on T cells activated by PHA, ConA, PWM or rIL-2, and on EBV or PWM blast cells, but was not expressed on resting T or B cells. Thus, it appears that TU223 reacts with a particular epitope on certain activated lymphoid cells. Since the antigen defined by TU223 is expressed on rIL-2-activated T cells, this antigen may be the IL-2 receptor. The antigen defined by TU223, however, differs from the IL-2 receptor (a chain) defined by CD25 (IOT-14) in its pattern of cellular expression. Antigens expressed on activated lymphoid cells have been defined and char-
FLUORESCENCE INTENSITY F&rc 4. Cross-blocking assay between CD23 (H107) and TU223. Cell assays were carried out by incubating with CD23 (H107) (panel A) or CD20 (BI) (panel B) in the first stage. In the second stage, TU223 was incubated, and FITC-conjugated goat-mouse IgG or human IgM were used in the third stage, respectively. Fluorescence intensity was measured by FACStar. The control indicates background staining without mAb.
A
B
KO
- 116
q-
97
i
i
ferred to nitrocellulose sheets, and then immunoblotted with TU223, CD23 (H107) or B5. As illustrated in Fig. 5, TU223 precipitated a polypeptide chain having an apparent molecular weight (m.w.) of 80-82 kiloDaltons (kD).On the other hand, CD23 (H107) or B5 precipitated a polypeptide chain with a m.w. of 42-45 kD and 70-75 kD, respectively.
-
45
*
Table 5. Inhibition of PFC by TU223 as measured by the reverse hemolytic plaque assay IgGPFC/l Oh mAb added time (h) Medium TU223 NHulgM" H107
0 24 48 72 96 PWM alone no PWM
-
548
1758
605 428 475 412
1258 1100
ion
1168
Sa3
1203 90 1080 128 1000 568 1050 1028 1160 1005
1220 23
IgGsecreting cells were measured by the reverse plaque assay. * NHulgM; normal human IgM.
1
2
3
4
5
6
7
Figure 5. Western blotting analysis of the antigen defined by TU223. Solubilized proteins from B85 (panel A) or Ramos (panel B) cells were subjected to 10% SDS-PAGE, and electrotransferred to nitrocellulose sheets. Strips were immunoblotted with antibodies as follows: lane 1, phosphate-buffered saline (PBS)containing normal human IgM (1 pg/ml); lane 2, TU223; lane 3, PBS containing normal mouse IgG (1 pg/ml); lane 4, CD23 (H107);lane 5, PBS containing normal mouse IgM ( 1 pglml); lane 6, B5. Molecular weight markers stained with amido black are as follows, 8-galactosidase, 116000; phosphorylase b, 97 000; bovine serum albumin, 66 000; egg albumin, 45 000 seen in lane 7.
179
Ikewaki et al.
acterized using several mAb over the last several years. Especially, new B cell-associated antigens were defined, e.g., CD19 (9), CD20 (lo), CD21 (1 I), CD23 (16), CD39 (27), B5 (12) and BB-I/ LB- 1 (28). These antigens play a key role in human B-cell proliferation and/or differentiation. In this study, the sero-reactivity of TU223 was found to be similar to that of CD23 defined by H107, which reacts with Fc &‘receptor11. In addition, it was found by two-color analysis that the antigen defined by TU223 is expressed on certain CD23 antigen-bearing lymphoid cells, suggesting that TU223 might recognize CD23 antigen. Cross;blocking assay, immunoblotting analysis and the pattern of cellular expression on 1-11-4-treated B or U937 cells, however, showed that the antigen defined by TU223 is not identical to CD23 antigen. Thus, these findings suggest that the regulation of the antigen defined by TU223 and CD23 has something in common with cellular expression on certain lymphoid cells. On the other hand, mAb AC2, which reacts with CD39 antigen associated with the EBV-transformed state, has been reported by Rowe et al. (27). AC2 binds with all EBV-transformed B-cell lines and some EBV genome-positive Burkitt’s lymphoma cell lines but not with EBV genomenegative cell lines. Furthermore, it was found that AC2 precipitates a polypeptide chain having an apparent m.w. of 80-82 kD, and binds to some cell populations activated with PHA or PWM. This study shows that TU223 is very similar to CD39 (AC2) in molecular weight and cellular distribution. Our data indicate that TU223 is very closely associated with certain EBV-transformed cell stages, suggesting that the antigen defined by TU223 is also equal to CD39 as a B cell-restricted activation antigen. However, as we do not have mAb AC2 in our laboratory, detailed differences between TU223 and AC2 are not clear at this time, and are currently being investigated. As the results of this study indicated that the cellular expression of the antigen defined by TTJ223 is similar to that of CD23 antigen, we examined the differences between TU223 and H107 mAb in biological function. CD23 is now understood to be a B cell-specific differentiation marker whose expression is restricted to a certain stage in the maturation of B cells (29-32). Furthermore, it was found that induction of EBV nuclear antigen 2 (EBNA 2) expression on B cells is dependent on direct or indirect transactivation of CD23 antigen (33). Regarding biological function, we showed that addition of TU223 to PWM-stimulated cultures for PFC assay on d 0 to 4 induced significant suppression of the development of PFC. In contrast to these observations, we found that CD23 (H107) 180
did not significantly suppress or enhance the PFC response under the same conditions, while addition of Sa3 to this system on d 0 to 2 induced significant suppression of PFC production. These findings indicate that TU223 recognizes an antigen expressed on B or T cells activated by PWM in the late phase of B-cell differentiation, and that this antigen plays an important role in differentiating human B cells into IgG-secreting cells, and also showed that the biological function of the antigen defined by TU223 differs from that of CD23 or HLA-class I1 antigens in the PFC suppression pattern. Our studies show that TU223 reacts with a novel and functional antigen on the cell surface of some activated lymphoid cells; however, detailed properties of this antigen are not clear at the present time. Finally, why B cells obtained from PBMCs of a pregnant woman produce this intriguing mAb is not fully understood at the moment. Studies to solve these interesting problems are currently in progress. As the sero-reactivity or molecular size of TU223 was found to be very similar to that of CD39 (AC2), further studies will be necessary to analyze in more detail the sequential immune precipitation between TU223 and CD39 mAb.
References 1. Nadler LM, Ritz J, Griffin JD, Todd I11 RF, Reinherz
EL, Schlossman SF. Diagnosis and treatment of human leukemias and lymphomas utilizing monoclonal antibodies. Prog Haematol 1981: 12: 187-225. 2. Croce CM, Linnenbach A, Hall W.Production of human hybridomas secreting antibodies to meales virus. Nature 1980: 228 488-9. 3. Eisenbarth GS, Linnenbach A, Jakson R. Human hybridomas secreating anti-islet autoantibodies. Nature 1982: 300: 264-7. 4. Houghton AN, Brooks H, Cote RJ, Taormina MC, Oettgen HF, Old LJ. Generation of human monoclonal antibodies reactive with cellular antigens. Proc Nut1 Acad Sci USA 1983: 80: 2026-30. 5. Olsson L, Andreasen RB, Ost A, Christensen B, Biberfeld P. Antibody producing human-human hybridomas. I1 Derivation and characterization for an antibody specific for human leukemia cells. J Exp Med 1984 159 537-50. 6. Effros RB, Hulette CM, Ettenger R, Dillard LC, Zeller E. A human-human hybridoma secreting anti-HLA class I1 antibody. J Immunol 1986: 137: 1599-603. 7. Alpert SD, Turek PJ, Foutlg SKH, Engleman EG. Human monoclonal anti-T cell antibody from a patient with juvenile rheumatoid arthritis. J Immunol 1987: 138: 104-8. 8. Hansen T, Bratlie A, Johansen B, Hannestad K.Two cytotoxic human-human hybridoma antibodies to HLA: TrAHlO (anti A3.1) and TrAG2 (anti B7, Bw42). Tissue Antigens 1989: 33: 520-30. 9. Nadler LN, Anderson KC, Marti G, et al. B4, a human B lymphocyte-associated antigen expressed on normal, mitogen-activated, and malignant B lymphocytes. J Immunol 1983: 131: 2 6 5 0 . 10. Stashenko P, Nadler LM, Hardy R, Schlossman SF. Characterization of a human B lymphocyte-specific antigen. J Immunol 1980 125: 1678-85.
A new human monoclonal antibody 11. Nadler LM, Stashenko P, Hardy R, Agthoven A, Terhorst C. Schlossman SF. Characterization of a human B cellspecific antigen (B2) distinct BI. J Zmmunol 1981: 126 1941-7. 12. Freedman AS, Boyd AW. Anderson KC, Fisher DC, Schlossman SF, Nadler LM. B5, a new B-cell-restricted activation antigen. J Immunol 1985: 134: 2228-35. 13. Hemler ME, Malissen B, Rebai N, et al. A 55,000 Mr surface antigen on activated human T lymphocytes defined by a monoclonal antibody. Human Immunol 1983: 8:
24. 25.
26.
153-65. 14. Hagiwara S, Mori E, Matsuo Y, Kondo H, Mon S. A
CD5 monoclonal antibody that cross-reacte with platelet membrane antigen (NU-T PAN). Acta Haematol Jpn 1987: 5 0 626-32 (in Japanese). 15. Matsuo Y, Hagiwara S, Okubo K, Matsumoto K, Sagawa K, Yokoyama MM. Monoclonal antibodies (NU-N1 and NU-N2) reactive with common acute lymphoblastic leukemia antigen. Acta Haemafol Jpn 1985: 48: 775-83 (in Japanese). 16. Noro N, Yoshioka A, Adachi M, Yasuda K, Masuda T, Yodoi J. Monoclonal antibody (H107) inhibiting IgE binding to FCER( +) human lymphocytes. J Zmmunol1986: 137: 1258-63. , 17. Mittler RS, Talk MA, Carpenter K, Rao PEO,Goldstein G. Generation and characterization of monoclonal antibodies reactive with human B lymphocytes. J Immunol 1983: 131: 175461. 18. Reinherz EL, Kung PC, Goldstein G, Levy RH, Schlossman SF. Discrete stage of human intrathymic differentiation: analysis of normal thymocytes and leukemic lymphoblasts of T-cell lineage. Proc Nafl Acad Sci USA 1980: 77: 1588-92. 19. Ikeda H, Kasahara M, Ogasawara K, et al. Evidence for polymorphism of MB3 antigens among three HLA-D clusters associated with HLA-DR4. Inmunogenetics 1984: 19: 38 1-90. 20. Nose Y, Araki N, Yamashita H, et al. Analyses of HLA antigens especially HLA-A and B interlocus antigen on platelets with monoclonal antibodies and ELISA assay. Acra Haemutol Jpn 1987: 33: 745-9 (in Japanese). 21. Ikewaki N, Nakatsuji T, Inoko H, et al. Monoclonal antibody against HLA-DP antigen on transfected cells. Proc 3rd AOHW 1986: 826-9. 22. Bayum A. Separation of white blood cells. Nature 1964: 204: 793-4. 23. Jondal M, Holm G, Wigzell H. Surface markers on human T and B lymphocytes. I. A large population of lymphocytes
27.
28.
29.
30.
31. 32.
33.
forming nonimmune rosettes with sheep red blood cells. J Exp Med 1972: 136: 207-15. Ikewaki N. Production of human B cell hybridomas with interleukin 2 receptors selected by a new selective medium, HAT-neo. Kitmato Arch Exp Med 1988: 61: 55-65. Mittal KK, Mickey MR, Singal DP, Terasaki PI. Serotyping for homotransplantation. XVIII. Refinement of microdroplet lymphocyte cytotoxicity test. Transplantation 1968: 6 913-27. Gronowin E, Coutino A, Melchers F. A plaque assay for all cells secreting Ig a given type of class. Eur J Immunol 1976: 6 588-90. Rowe M, Hildreth JEK, Rickinson AB, Epstein MA. Monoclonal antibodies to Epstein-Barr virus-induced, transformation-associated cell surface antigens: binding pattern and effect upon virus-specific T-cell cytotoxicity. Int J Cancer 1982: 29: 373-81. Yokochi T, Holly RD, Clark EA. B lymphoblast antigen (BB-1) expressed on Epstein-Barr virus-activated B cell blast, B lymphoblastoid cell lines, and Burkitt's lym-' phomas. J Immunol 1982: 128: 823-7. Kikutani H, Suemura M, Owaki H, et al. FCEreceptor, a specific differentiation marker transiently expressed on mature B cells prior to isotype switching. J Exp Med 1986: 164: 1455-69. Gordon J, Rowe M, Walker L, Guy G. Ligation of the CD23, $5 (Blast-2, EBVCS) antigen triggers the cell-cycle progression of activated B lymphocytes. Eur J Immunol 1986: 16: 1075-80. Yukawa K, Kikutani H, Owaki H, et al. A B cell-specific differentiation antigen, CD23, is a receptor for IgE (FcER) on lymphocytes. J Immunol 1987: 138: 2576-80. Crow MK, Kushner B, Jover J, Friedman SM, Mechanic SE, Stohl W. Human peripheral blood T hekper cell-induced B cell activation results in B cell surface expression of the CD23 (BLAST-,) antigen. Cell Itntnunol 1989: 121: 99-1 12. Wang F, Gregory CD, Rowe M, et al. Epstein-Barr virus nuclear antigen 2 specifically induces expression of the Bcell activation antigen CD23. Proc Nut1 Aciid Sci USA 1987: 84: 3452-6.
Address: Kimiyoshi Tsrq'i, kf.D. Professor, Department of Transplantation Tokai University School of Medicine Bohseidai Isehara Kanagawa 259- 1 1 Japan
181
N. Ikewaki, S. Takabe, K. Tsuji. Development and characterization of a
Nobunao Ikewaki*, Sachiko Takabe
human monoclonal antibody probably detecting the leukocyte differentiation antigen CD39. Tissue Antigens 1992: 39: 174-181.
and Kimiyoshi Tsuji
Abstract: A human monoclonal IgM antibody, referred to as TU223, has been produced. The reactivity of TU223 was tested in various cells and cell lines by complement-dependent microcytotoxicity test and fluorescenceactivated cell sorter analysis. The antigen defined by TU223 was expressed on Epstein-Barr virus-transformed B-cell lines and on some Burkitt’s lymphoma cell lines, but was not expressed on normal T cells, B cells or erythrocytes. In addition, expression of the antigen defined by TU223 was also induced on B cells activated by Epstein-Barr virus or pokeweed mitogen, and on T cells activated by phytohemagglutinin, concanavalin A, pokeweed mitogen or recombinant interleukin-2. However, no expression of the antigen detected by TU223 was induced at all on recombinant interleukin-4-treated B cells or macrophage-like cell line U937. When the ability of TU223 and various mouse monoclonal antibodies to bind to human differentiation antigens was compared, interestingly, the reactivity of TU223 was found to be very similar to that of mouse monoclonal antibody CD23 (H107), which reacts with Fcs receptor 11. Two-color analysis revealed that the antigen defined by TU223 is expressed on the cell surface of certain lymphoid cells expressing CD23 antigen. However, it can be concluded that the antigen defined by TU223 is clearly distinct from Fcs receptor 11, based on assay of cross-blocking between HI07 and TU223. The surface antigen on B85 cells recognized by TU223 had the molecular size of 80-82 kiloDaltons as determined by immunoblotting analysis. In biological function, TU223 inhibited the generation of plaqueforming cells when added to a pokeweed mitogen-derived IgG-secreting system on days 0 to 4. These findings suggest that TU223 recognizes a novel and functional antigen expressed on certain activated lymphoid cells.
ttroduction Over the past several years, many mouse monoclonal antibodies (mAb) which react with human differentiation antigens on a variety of cells have been established and characterized. These mAb have played an important role in analyzing cell activation a n d / or differentiation mechanisms (1). On the other hand, human mAb t o a variety of * Present address: Department of Microbiology, Kitasato University School of Nursing, Kitasato, Sagamihara, Kanagawa 228, Japan. 174
Department of Transplantation, Tokai University School of Medicine, Bohseidai, Isehara, Kanagawa, Japan
Key words: a novel antigen - human monoclonal antibody - lymphoid cells Received 31 October, accepted for publication 15 November 1991
antigens have also been established using a number of techniques, and these human mAb have been
very useful clinically (2-8). We recently succeeded in establishing an EpsteinB a n virus (EBV)-transformed B-cell line, derived from B cells obtained from the peripheral blood mononuclear cells (PBMCs) of a pregnant woman, which steadily secretes a human IgM mAb. In this report, we describe the properties of this human mAb, referred to as TU223.Based on its reactivity, it was found that TU223 recognizes a novel and functional antigen expressed on certain activated lymphoid cells.
A new human monoclonal antibody Material and Methods Antibodies
Mouse mAb CD19 (B4) (9), CD20 (Bl) (lo), CD21 (B2) (ll), B5 (12) were purchased from Coulter Immunology Laboratories (Hialeah, FL). CD25 (IOT-14) (13) was purchased from Immunotech Laboratories (Marseilles). CD5 (NU-T PAN) (14), CDlO (NU-Nl) (15), CD23 (H107) (16) were purchased from Nichirei Laboratories (Higashimurayama, Tokyo). CD24 (OKB2) (17), CD38 (OKTlO) (18) were obtained from Ortho Diagnostic System Laboratories (Raritan, NJ). HU20 (,19) was kindly provided by M. Aizawa, M.D. (Department of Pathology, Hokkaido University School of Medicine). YU3-20A (20) or Sa3 (21) were produced in our own laboratory. The specificities of the mAb used in this study are summarized in Table 1. Fluorescein isothiocyanate (F1TC)-conjugated goatanti-mouse IgF(ab’)2, FITC-conjugated goat-antihuman IgMF(ab’)2, non-conjugated goat-anti-human IgG, goat-anti-human IgM, goat-anti-human IgA, horseradish peroxidase (HRP0)-conjugated goat-anti-human IgF(ab’)2 and phycoerythrin (PE)-conjugated goat-anti-mouse IgF(ab’)2 were purchased from Cappel Laboratories (Malvern, PA). Preparation of lymphoid cells
Peripheral blood mononuclear cells (PBMCs) obtained from 37 healthy donors were isolated by Ficoll-Paque (Pharmacia, Uppsala) density sedimentation as previously described (22). T and B cells were separated into erythrocyte (E) rosettepositive (E+)and E rosette-negative (E-) populations using sheep erythrocytes (23). Human cell lines
All human cell lines used in the present investigation, except EBV-transformed B-cell lines and Burkitt’s lymphoma BT-l, were supplied by the Japanese Cancer Research Resource Bank (Yoga, Tokyo). EBV B-cell lines B85 and HID were established by J. Minowada, M.D. of New York State University and in our own laboratory, respectively. Other EBV B-cell lines were exchanged at the 10th International Histocompatibility Workshop Meeting. Burkitt’s lymphoma cell line BT-1 was kindly provided by H. Yakura, M.D. of the Department of Pathology, Asahikawa Medical College.
Flow cytometric analysis The binding activities of the mAb were analyzed by direct or indirect immunofluorescence using a
fluorescence-activated cell sorter (FACStar, Becton Dickinson, Mountain View, CA) as previously described (24). In some experiments, the cellular distribution of antigens detected by mAb was determined by two-color flow cytometry. Complement-dependent microcytotoxicity test
Complement-dependent microcytotoxicity test was performed using a standard microcytotoxicity test procedure employing Terasaki typing trays and rabbit complement as previously described (25). Establishment of a human mAbproducing 8-cell line
PBMCs obtained from a pregnant woman, SU, were isolated by Ficoll-Paque density gradient centrifugation. B cells were separated on the basis of rosette formation. Enriched B cells were suspended in complete RPMI 1640 medium (GIBCO, Grand Island, NY) supplemented with 10% fetal calf serum (FCS), 100 units/ml penicillin, 100 pg/ml streptomycin and 2 mM L-glutamine. The suspension was distributed in 24-well plates (Costar, Cambridge, MA) with approximately lo5 cells in 900 pl. In order to transform the B cells, the filtered EBV-containing culture supernatant from marmoset cell line B95-8 was added to each well, and incubation was performed for 60 min at 37°C. EBV-transformed B cells began to appear, usually within 3 weeks. The amount of human Ig in the culture supernatant was determined by enzymelinked immunosorbent assay (ELISA) (24). The transformed cells were cloned by limiting dilution, and selected cells were then recloned twice, resulting in the establishment of a cell line which has been steadily producing large amounts of human IgM, even in a new culture medium, GIT (Wako Pure Chemistories, Tokyo), deprived of FCS. This Table 1. Specificity of the mouse mAb used in this study CD no.
mAb
specificity
lg class
CD5 CDlO CO19 CD20 C021
NU-T PAN NU-N1 84 B1 B2 H107 OKB2 IOT-14 OKTlO 85 Sa3 HU20 YU3-20A
Pan-T cells CALW Early B cells Pan-B cells BceU subset FCZ receptor II Pan B cells 11-2 receptor (achain) T progenitorfl blasts Activated B cells HLA-DR.DQ0P HLA-OR HLA-A,B,C
lgG2a IgG1 lgGl lgG2a IgM lgG2b lgG1 lgG2a lgG1 IgM 1963 lgG2b IgM
CD23
CD24 CD25 CD38
-
*
CALLA; common acute lymphoblastk leukemia antigen.
175
Ikewaki et a]. Table 2. Binding of TU223 to cultured cell lines Degree of binding of mAb”
Cell lines
TU223
H107
0 0 0
0 0
+
++ + + +
T-cell leukemia GEM MOLT4 MOLT3 EBV h e l l lines 885 WAR BM16 RML . AKlBA Kl12 MANN HID Burkitt’s lymphoma BT-1 Raji Daudi Ramos B-cell ALL
CCRF-SB Erythroid cells K562 Myeloid cells u937 THP1
NU-N1
84
85
B1
82
OKB2
Om10
HU20
Sa3
N F
0 0 0
0
NT
+
NT
0 0
0 0
0
0
0 0
++
NT NT
NT 0
NT
0 0
++
0 0
+
NT
NT
++
0
+
0
+
f
+ +
++ ++
0
0
0 0 0
++
+
++ ++ ++ ++
+
++ ++ ++
0 0 0 0 0
NT NT NT
++ ++ ++
++ ++ ++ ++
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
+ +-
++ ++
+
++ ++
++ ++
0 0
0 0
0
++
++
0 0 0
+
NT NT NT NT
+
+
+ + + + 4.
0
++
0
0
+
+ +
f
+
+
NT
+
0 0
NT
r+
+
+
NT
++
++ +
0
0
0
0 0
0 0 0
+
+
0 0
0 0
0 0 0
0
0 0 0 0
+ + +
+
0 0
+
+ 0
+
0
0
++ ++ ++ 0
++
++ ++ ++
++
++ ++ ++
YU3-20A IOT-14 NU-TPAN
0 NT
+ +
+ + +
++ ++ ++
++ ++
+
0
+
+
++
0
0
0
++
++
++
0
NT
0
0
NT
NT
0
NT
NT
NT
NT
NT
0
0
NT
NT
0
0 0
NT NT
NT NT
0
0 0
NT NT
NT NT
NT NT
NT NT
0
0
++ ++
NT 0
0 0
0
+
+
++
0
0
Binding activities of mAb were determined by complement-dependent microcytotoxicity test or a flow cytometry. Binding of mAb was scored as follows: (0) indicates < lo%, ( A )indicates 11 to 20%. (+) indicates 21 to 70%. (++) indicates > 71%. ** NT; not tested. *
mAb was assigned the designation TU223*TU223 has been maintained for more than 4 yr without loss of Ig-secreting activity, The isotype of TU223 is IgM. The binding titer of TU223 to B85 cell line was 1 5 4 in the unconcentrated culture supernatant (data not shown).
Assay of cross-blocking between CD23 (H107) and TU223
Culture with activators
Reverse hemolytic plaque assay for IgG-secreting cells
PBMCs in 1 ml of complete RPMI 1640 medium were put into 24-well plates and cultured with 1 pg/ml phytohemagglutinin (PHA) (DIFCO), 10 pg/ml concanavalin A (ConA) (Sigma), pokeweed mitogen (PWM) (GIBCO) in a 1:lOO dilution and 1 U/ml recombinant interleukin-:! (rIL-2) (Shionogi Pharmacy, Osaka) at 37°C. After activation, T or B cells were isolated on the basis of rosette formation. On the other hand, B cells were cultured in 900 p1 complete RPMI 1640 medium in 24-well plates with 100 pl EBV or 100 U/ml recombinant interleukin-4 (rIL-4) (Genzyme). U937 cells were also treated with 100 U/ml rIL-4. Expression of the antigen defined by TU223 and CD23 antigen was detected by complement-dependent microcytotoxicity test or two-color flow cytometry.
PBMCs (1 x lo’) were isolated by Ficoll-Paque density gradient centrifugation and cultured in triplicate in v i m for 7 days with PWM (1 pg/ml) in the presence (10 pg/ml) or absence of TU223, CD23 (H107) or Sa3. On d 7 the cells were harvested and washed in RPMI 1640 medium. IgG-secreting cells were determined by the reverse plaqueforming cell (PFC) assay with some modifications as previously described (26).
176
In order to determine whether TU223 reacts with an epitope of CD23, a cross-blocking assay was performed by FACS analysis as previously described (24).
Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting
EBV-B cell line B85 or Burkitt’s‘lymphoma cell line Ramos were lysed in lysis buffer (50 mM TrisHCI pH 7.4, 150 mM NaCl, 5 mM EDTA, 0.5%
A new human monoclonal antibody
NP40, 1 mM PMSF). After centrifugation, the solubilized proteins were separated on 10% SDS gel and transferred electrophoretically to nitrocellulose sheets, which were then incubated with TU223, CD23 (H107) or B5. After incubation, immunoperoxidase staining was performed using an avidinbiotin complex (ABC) kit (Vector Laboratories Inc., Burlingame, CA), followed by visualization by the diaminobenzidine chromogen reaction. Control immunoblotting was also performed using non-reactive normal human IgM, mouse IgG or IgM.
Table 3. Binding activity of TU223 to normal lymphoid cells and erythrocytes
96 of positive cells' mAb
NU-T PAN 81 TU223
T cells ( ~ 3 7 ) 95 2 0.1
B cells (17-37)
Degree of agglutination" Erythrocytes (A,
B. 0, AB)
3 64 0.7
Analyzed by complementdependent rnicrocytotoxicity test. The average percentage of positive cells in 37 healthy donors is shown. * Analyzed using an agglutination test (-); negative.
Results Binding activity of TU223 to a panel of cultured cell lines
The reactivity of TU223 was assessed using a panel of 19 established -cell lines and complement-dependent microcytotoxicity test or an immunofluorescence assay. Table 2 shows that all EBV B-cell lines used in this experiment and certain Burkitt's lymphoma cell lines, Raji and BT-1, were found to bind strongly. The reactivity of TU223 with these cell lines was found to be similar to that of CD23
(H107), but not CD5 (NU-T PAN), CDlO (NUNl), CD19 (B4), CD20 (Bl), CD21 (B2), B5, CD24 (OKB2), CD25 (IOT-14), CD38 (OKTlO), HU20, YU3-20A or Sa3. The cellular distributions of CD23 and the antigen defined by TU223 were also tested by two-color analysis and expression of the antigen defined by TU223 was detected on the cell surface of certain lymphoid cells bearing CD23 antigen (Fig. 1). Reactivity of TU223 with normal lymphoid cells and erythrocytes
AKIBA
The binding activity of TU223 to T or B cells obtained from 37 healthy donors was determined by complement-dependent microcytotoxicity test, and binding to erythrocytes (human) was tested by means of the agglutination test. All lymphoid cells and erythrocytes yielded negative results with regard to response to TU223 (Table 3). Binding activity of TU223 to activated cells
8-85
Reactivity of TU223 to activated cells was tested by means of the complement-dependent microcytotoxicity test. Table 4 shows that percentages of cytotoxic activity of TU223 against T cells activated by PHA, ConA, PWM or rIL-2 were approximately 58%, 48%, 70% and 35%, respectively, after 5 d. On the other hand, percentage of cytotoxic activity of TU223 against B cells activated by PWM was approximately 65%. Binding activity of TU223 to rlL-4-treated B or U937 cells
GREEN FLUORESCENCE Tu223 Figure 1. Twocolor analysis of TU223 binding to certain cultured B-cell lines. The cells were stained with CD23 (H107)and TU223 in the first stage. PE-conjugated goat-anti-mouse IgG or FITC-conjugated goat-anti-human IgM, respectively, were used in the second stage. Relative fluorescence intensity was expressed by logarithmic amplification.
Reactivity of TU223 with rIL-&treated B or U937 cells was tested by two-color flow cytometric analysis. Fig. 2 shows that CD23 expression was strongly induced on both B and U937 cells after treatment with rIL-4; however, no expression of the antigen defined by TU223 was induced on these cells at all. 177
Ikewaki et al. Table 4. Cytotoxic activity of TU223 against T or B cells actiiated in viffo Cells obtained from culture with
% of positive cells’
T cells
Medium only PHA ConA PWM rll-2
(n-5) (n-5)
5 58
(n=5) (n-5)
4a
(n-5)
B cells
means of FACS analysis. As shown in Fig. 4, CD23 (H107) did not block the reactivity of TU223 with AKIBA cells at all. Thus, these findings indicate that TU223 recognizes an epitope distinct from that of CD23 antigen. .
a NT NT 65
70 35
NT
Analyzed by complementdependent microcytotoxicity test, average percen5 healthy donors are shown. * NT; not tested. tages of positive cells in
Kinetics of expression of the antigen defined by TU223 on EBV blasts and on PWM-activated B cells
We examined the kinsics of expression of the antigen defined by TU223 on EBV blasts and on PWM-activated B cells, Expression of the antigen defined by TU223 was detectable on B cells apProximately 10 to 12 d after activation by EBy On the other hand, it was found that expression of this antigen on PWM-activated B cells clearly occurred about 2 to 3 d after activation (Fig. 3).
Inhibition of
We investigated the possibility that the antigen defined by TU223 plays a regulatory role in the differentiation of resting B cells to IgG-secreting cells. As shown in Table 5, the presence of TU223 markedly reduced the PWM-derived induction of PFC when added to cultures on d 0 to 4. However, CD23 (H107) did not inhibit PFC inductions. On the other hand, it was shown that addition of Sa3 to this system on d 0 to 2 induced significant inhibition of the development of PFC. Biochemical characterization of the cell surface antigen recognized by TU223
Solubilized proteins from B85 or Ramos cells were electrophoresed on 10% polyacrylamide gel, trans-
x
Evidence that the antigen defined by TU223 is distinct from CD23 antigen
PFC induction
A
01
wsilin clllr
/-
In order to test whether TU223 reacts with CD23 antigen, a cross-blocking assay was performed by
dm
after xtiration
B
medium only
rlL-4(100U/ml)
.
“ S 0
1
2
1
4
5
dap allcr a c l w a l m
GREEN FLUORESCENCE W223 Figure 2. Expression of the antigen defined by TU223 on B and U937 cells after treatment with rIL-4. Cells treated with rIL-4 (100 Ulml) were stained with HI07 and TU223 in the first stage. PE-conjugated goat-anti-mouse IgG or FITC-conjugated goat-anti-human IgM, respectively, were used in the second stage. Fluorescence intensity was tested by FACStar.
178
Figure 3. Kinetics of cellular expression of the antigen defined by TU223. The kinetics of induction of expression of the antigen defined by TU223 on B cells activated by EBV (panel A) and B cells activated by PWM (panel B) is shown. The percentage of positive cells was determined by complement-dependent microcytotoxicity test. The average percentage of positive cells in 2 healthy donors is shown.
A new human monoclonal antibody
Discussion
A
i
B
In the present study, we succeeded in establishing an EBV-transformed B-cefl line that has been producing a human n A b referred to as TU223 for more than 4 yr. When the reactivity of TU223 was tested on various cells and cell lines, it was found that the antigen defined by TU223 was expressed on certain Burkitt's lymphoma cell lines, on all EBV B-cell lines used in this experiment, on T cells activated by PHA, ConA, PWM or rIL-2, and on EBV or PWM blast cells, but was not expressed on resting T or B cells. Thus, it appears that TU223 reacts with a particular epitope on certain activated lymphoid cells. Since the antigen defined by TU223 is expressed on rIL-2-activated T cells, this antigen may be the IL-2 receptor. The antigen defined by TU223, however, differs from the IL-2 receptor (a chain) defined by CD25 (IOT-14) in its pattern of cellular expression. Antigens expressed on activated lymphoid cells have been defined and char-
FLUORESCENCE INTENSITY F&rc 4. Cross-blocking assay between CD23 (H107) and TU223. Cell assays were carried out by incubating with CD23 (H107) (panel A) or CD20 (BI) (panel B) in the first stage. In the second stage, TU223 was incubated, and FITC-conjugated goat-mouse IgG or human IgM were used in the third stage, respectively. Fluorescence intensity was measured by FACStar. The control indicates background staining without mAb.
A
B
KO
- 116
q-
97
i
i
ferred to nitrocellulose sheets, and then immunoblotted with TU223, CD23 (H107) or B5. As illustrated in Fig. 5, TU223 precipitated a polypeptide chain having an apparent molecular weight (m.w.) of 80-82 kiloDaltons (kD).On the other hand, CD23 (H107) or B5 precipitated a polypeptide chain with a m.w. of 42-45 kD and 70-75 kD, respectively.
-
45
*
Table 5. Inhibition of PFC by TU223 as measured by the reverse hemolytic plaque assay IgGPFC/l Oh mAb added time (h) Medium TU223 NHulgM" H107
0 24 48 72 96 PWM alone no PWM
-
548
1758
605 428 475 412
1258 1100
ion
1168
Sa3
1203 90 1080 128 1000 568 1050 1028 1160 1005
1220 23
IgGsecreting cells were measured by the reverse plaque assay. * NHulgM; normal human IgM.
1
2
3
4
5
6
7
Figure 5. Western blotting analysis of the antigen defined by TU223. Solubilized proteins from B85 (panel A) or Ramos (panel B) cells were subjected to 10% SDS-PAGE, and electrotransferred to nitrocellulose sheets. Strips were immunoblotted with antibodies as follows: lane 1, phosphate-buffered saline (PBS)containing normal human IgM (1 pg/ml); lane 2, TU223; lane 3, PBS containing normal mouse IgG (1 pg/ml); lane 4, CD23 (H107);lane 5, PBS containing normal mouse IgM ( 1 pglml); lane 6, B5. Molecular weight markers stained with amido black are as follows, 8-galactosidase, 116000; phosphorylase b, 97 000; bovine serum albumin, 66 000; egg albumin, 45 000 seen in lane 7.
179
Ikewaki et al.
acterized using several mAb over the last several years. Especially, new B cell-associated antigens were defined, e.g., CD19 (9), CD20 (lo), CD21 (1 I), CD23 (16), CD39 (27), B5 (12) and BB-I/ LB- 1 (28). These antigens play a key role in human B-cell proliferation and/or differentiation. In this study, the sero-reactivity of TU223 was found to be similar to that of CD23 defined by H107, which reacts with Fc &‘receptor11. In addition, it was found by two-color analysis that the antigen defined by TU223 is expressed on certain CD23 antigen-bearing lymphoid cells, suggesting that TU223 might recognize CD23 antigen. Cross;blocking assay, immunoblotting analysis and the pattern of cellular expression on 1-11-4-treated B or U937 cells, however, showed that the antigen defined by TU223 is not identical to CD23 antigen. Thus, these findings suggest that the regulation of the antigen defined by TU223 and CD23 has something in common with cellular expression on certain lymphoid cells. On the other hand, mAb AC2, which reacts with CD39 antigen associated with the EBV-transformed state, has been reported by Rowe et al. (27). AC2 binds with all EBV-transformed B-cell lines and some EBV genome-positive Burkitt’s lymphoma cell lines but not with EBV genomenegative cell lines. Furthermore, it was found that AC2 precipitates a polypeptide chain having an apparent m.w. of 80-82 kD, and binds to some cell populations activated with PHA or PWM. This study shows that TU223 is very similar to CD39 (AC2) in molecular weight and cellular distribution. Our data indicate that TU223 is very closely associated with certain EBV-transformed cell stages, suggesting that the antigen defined by TU223 is also equal to CD39 as a B cell-restricted activation antigen. However, as we do not have mAb AC2 in our laboratory, detailed differences between TU223 and AC2 are not clear at this time, and are currently being investigated. As the results of this study indicated that the cellular expression of the antigen defined by TTJ223 is similar to that of CD23 antigen, we examined the differences between TU223 and H107 mAb in biological function. CD23 is now understood to be a B cell-specific differentiation marker whose expression is restricted to a certain stage in the maturation of B cells (29-32). Furthermore, it was found that induction of EBV nuclear antigen 2 (EBNA 2) expression on B cells is dependent on direct or indirect transactivation of CD23 antigen (33). Regarding biological function, we showed that addition of TU223 to PWM-stimulated cultures for PFC assay on d 0 to 4 induced significant suppression of the development of PFC. In contrast to these observations, we found that CD23 (H107) 180
did not significantly suppress or enhance the PFC response under the same conditions, while addition of Sa3 to this system on d 0 to 2 induced significant suppression of PFC production. These findings indicate that TU223 recognizes an antigen expressed on B or T cells activated by PWM in the late phase of B-cell differentiation, and that this antigen plays an important role in differentiating human B cells into IgG-secreting cells, and also showed that the biological function of the antigen defined by TU223 differs from that of CD23 or HLA-class I1 antigens in the PFC suppression pattern. Our studies show that TU223 reacts with a novel and functional antigen on the cell surface of some activated lymphoid cells; however, detailed properties of this antigen are not clear at the present time. Finally, why B cells obtained from PBMCs of a pregnant woman produce this intriguing mAb is not fully understood at the moment. Studies to solve these interesting problems are currently in progress. As the sero-reactivity or molecular size of TU223 was found to be very similar to that of CD39 (AC2), further studies will be necessary to analyze in more detail the sequential immune precipitation between TU223 and CD39 mAb.
References 1. Nadler LM, Ritz J, Griffin JD, Todd I11 RF, Reinherz
EL, Schlossman SF. Diagnosis and treatment of human leukemias and lymphomas utilizing monoclonal antibodies. Prog Haematol 1981: 12: 187-225. 2. Croce CM, Linnenbach A, Hall W.Production of human hybridomas secreting antibodies to meales virus. Nature 1980: 228 488-9. 3. Eisenbarth GS, Linnenbach A, Jakson R. Human hybridomas secreating anti-islet autoantibodies. Nature 1982: 300: 264-7. 4. Houghton AN, Brooks H, Cote RJ, Taormina MC, Oettgen HF, Old LJ. Generation of human monoclonal antibodies reactive with cellular antigens. Proc Nut1 Acad Sci USA 1983: 80: 2026-30. 5. Olsson L, Andreasen RB, Ost A, Christensen B, Biberfeld P. Antibody producing human-human hybridomas. I1 Derivation and characterization for an antibody specific for human leukemia cells. J Exp Med 1984 159 537-50. 6. Effros RB, Hulette CM, Ettenger R, Dillard LC, Zeller E. A human-human hybridoma secreting anti-HLA class I1 antibody. J Immunol 1986: 137: 1599-603. 7. Alpert SD, Turek PJ, Foutlg SKH, Engleman EG. Human monoclonal anti-T cell antibody from a patient with juvenile rheumatoid arthritis. J Immunol 1987: 138: 104-8. 8. Hansen T, Bratlie A, Johansen B, Hannestad K.Two cytotoxic human-human hybridoma antibodies to HLA: TrAHlO (anti A3.1) and TrAG2 (anti B7, Bw42). Tissue Antigens 1989: 33: 520-30. 9. Nadler LN, Anderson KC, Marti G, et al. B4, a human B lymphocyte-associated antigen expressed on normal, mitogen-activated, and malignant B lymphocytes. J Immunol 1983: 131: 2 6 5 0 . 10. Stashenko P, Nadler LM, Hardy R, Schlossman SF. Characterization of a human B lymphocyte-specific antigen. J Immunol 1980 125: 1678-85.
A new human monoclonal antibody 11. Nadler LM, Stashenko P, Hardy R, Agthoven A, Terhorst C. Schlossman SF. Characterization of a human B cellspecific antigen (B2) distinct BI. J Zmmunol 1981: 126 1941-7. 12. Freedman AS, Boyd AW. Anderson KC, Fisher DC, Schlossman SF, Nadler LM. B5, a new B-cell-restricted activation antigen. J Immunol 1985: 134: 2228-35. 13. Hemler ME, Malissen B, Rebai N, et al. A 55,000 Mr surface antigen on activated human T lymphocytes defined by a monoclonal antibody. Human Immunol 1983: 8:
24. 25.
26.
153-65. 14. Hagiwara S, Mori E, Matsuo Y, Kondo H, Mon S. A
CD5 monoclonal antibody that cross-reacte with platelet membrane antigen (NU-T PAN). Acta Haematol Jpn 1987: 5 0 626-32 (in Japanese). 15. Matsuo Y, Hagiwara S, Okubo K, Matsumoto K, Sagawa K, Yokoyama MM. Monoclonal antibodies (NU-N1 and NU-N2) reactive with common acute lymphoblastic leukemia antigen. Acta Haemafol Jpn 1985: 48: 775-83 (in Japanese). 16. Noro N, Yoshioka A, Adachi M, Yasuda K, Masuda T, Yodoi J. Monoclonal antibody (H107) inhibiting IgE binding to FCER( +) human lymphocytes. J Zmmunol1986: 137: 1258-63. , 17. Mittler RS, Talk MA, Carpenter K, Rao PEO,Goldstein G. Generation and characterization of monoclonal antibodies reactive with human B lymphocytes. J Immunol 1983: 131: 175461. 18. Reinherz EL, Kung PC, Goldstein G, Levy RH, Schlossman SF. Discrete stage of human intrathymic differentiation: analysis of normal thymocytes and leukemic lymphoblasts of T-cell lineage. Proc Nafl Acad Sci USA 1980: 77: 1588-92. 19. Ikeda H, Kasahara M, Ogasawara K, et al. Evidence for polymorphism of MB3 antigens among three HLA-D clusters associated with HLA-DR4. Inmunogenetics 1984: 19: 38 1-90. 20. Nose Y, Araki N, Yamashita H, et al. Analyses of HLA antigens especially HLA-A and B interlocus antigen on platelets with monoclonal antibodies and ELISA assay. Acra Haemutol Jpn 1987: 33: 745-9 (in Japanese). 21. Ikewaki N, Nakatsuji T, Inoko H, et al. Monoclonal antibody against HLA-DP antigen on transfected cells. Proc 3rd AOHW 1986: 826-9. 22. Bayum A. Separation of white blood cells. Nature 1964: 204: 793-4. 23. Jondal M, Holm G, Wigzell H. Surface markers on human T and B lymphocytes. I. A large population of lymphocytes
27.
28.
29.
30.
31. 32.
33.
forming nonimmune rosettes with sheep red blood cells. J Exp Med 1972: 136: 207-15. Ikewaki N. Production of human B cell hybridomas with interleukin 2 receptors selected by a new selective medium, HAT-neo. Kitmato Arch Exp Med 1988: 61: 55-65. Mittal KK, Mickey MR, Singal DP, Terasaki PI. Serotyping for homotransplantation. XVIII. Refinement of microdroplet lymphocyte cytotoxicity test. Transplantation 1968: 6 913-27. Gronowin E, Coutino A, Melchers F. A plaque assay for all cells secreting Ig a given type of class. Eur J Immunol 1976: 6 588-90. Rowe M, Hildreth JEK, Rickinson AB, Epstein MA. Monoclonal antibodies to Epstein-Barr virus-induced, transformation-associated cell surface antigens: binding pattern and effect upon virus-specific T-cell cytotoxicity. Int J Cancer 1982: 29: 373-81. Yokochi T, Holly RD, Clark EA. B lymphoblast antigen (BB-1) expressed on Epstein-Barr virus-activated B cell blast, B lymphoblastoid cell lines, and Burkitt's lym-' phomas. J Immunol 1982: 128: 823-7. Kikutani H, Suemura M, Owaki H, et al. FCEreceptor, a specific differentiation marker transiently expressed on mature B cells prior to isotype switching. J Exp Med 1986: 164: 1455-69. Gordon J, Rowe M, Walker L, Guy G. Ligation of the CD23, $5 (Blast-2, EBVCS) antigen triggers the cell-cycle progression of activated B lymphocytes. Eur J Immunol 1986: 16: 1075-80. Yukawa K, Kikutani H, Owaki H, et al. A B cell-specific differentiation antigen, CD23, is a receptor for IgE (FcER) on lymphocytes. J Immunol 1987: 138: 2576-80. Crow MK, Kushner B, Jover J, Friedman SM, Mechanic SE, Stohl W. Human peripheral blood T hekper cell-induced B cell activation results in B cell surface expression of the CD23 (BLAST-,) antigen. Cell Itntnunol 1989: 121: 99-1 12. Wang F, Gregory CD, Rowe M, et al. Epstein-Barr virus nuclear antigen 2 specifically induces expression of the Bcell activation antigen CD23. Proc Nut1 Aciid Sci USA 1987: 84: 3452-6.
Address: Kimiyoshi Tsrq'i, kf.D. Professor, Department of Transplantation Tokai University School of Medicine Bohseidai Isehara Kanagawa 259- 1 1 Japan
181
