Int. J . Cancer: 46, 675-681 (1990) 0 1990 Wiley-Liss, Inc.

Publication of the International Union Against Cancer Publication de I’Union lnternationale Contre le Cancer

GENERATION AND CHARACTERIZATION OF MONOCLONAL ANTIBODIES AGAINST MULTIPLE EPITOPES ON THE C-TERMINAL HALF OF ENVELOPE gp46 OF HUMAN T-CELL LEUKEMIA VIRUS TYPE-I (HTLV-I) Yuetsu TANAKA’.~, Masazumi YASUMOTO’, Hiroshi NYUNOYA~, Tsutomu OGURA’, Masayoshi KIKUCH13, Kunitada SHIMOTOHNO~, Hiroshi SHIRAKI~, Naotaka KURODA~, Hisatoshi SHIDA~ and Hideki TOZAWA’ ‘Department of Immunology, School of Hygienic Sciences, Kitasato University, Kitasato 1-15-1, Sagamihara, Kanugawa 228; 2Virology and Biochemistry Divisions, National Cancer Center Research Institute, Tsukiji, Tokyo; 3Tsukuba Research Laboratories, Tokuyama Soda Co. Ltd., Tsukuba, Ibaraki; 4Fuku0kaRed Cross Blood Center, Chikushinio, Fukuoka; and 51nstitutefor Virus Research, Kyoto University, Sakyoku, Kyoto, Japan. In order to study the antigenicity of envelope 46 kDa glycoprotein (gp46) of human T-cell leukemia virus type-I (HTLV-I), we have generated monoclonal anti-gp46 antibodies (MAbs), REY-7, REY-I I, REY-16, REY-30, M E T 4 and MET3 from rats and mice. lmmunoblot and immunofluorexence assays showed that these MAbs recognize gp46 and its related antigens, and specifically stained HTLV-I-bearing cells. All MAbs reacted with a recombinant gp46 antigen, N147, expressing the 147 amino acids in the C-terminal half of gp46. By using various synthetic peptides corresponding to the gp46 sequence, epitopes recognized by REY-7 and MET-3, REY-I I and REY-16, and REY-30 were mapped to regions corresponding to the amino acids 175-1 99,253-282 and 288-3 12, respectively. MET-2 did not react with any of the peptides used. These results indicate that the present MAbs are directed against at least 4 distinct epitopes expressed on the Cterminal half of gp46. The binding of these MAbs to gp46 was specifically inhibited by sera from HTLV-I-infected individuals, but none of these MAbs inhibited the cell fusion activity of HTLV-I.

Human T-cell leukemia virus type-I (HTLV-I) is a C-type retrovirus etiologically associated with a human malignant Tcell disorder, adult T-cell leukemia (ATL) (Poiesz et al., 1980; Hinuma et al., 1981; Yoshida et al., 1982; Blattner et al., 1982) and some neurologic disorders (Osame et al., 1986; Jacobson et al., 1988). This virus has 2 external glycoproteins cleaved from a glycosylated precursor protein encoded by HTLV-I env gene, one of which is a major 46-kDa glycoprotein (gp46) linked to the other transmembrane glycoprotein, gp21 (Lee et al., 1984; Schneider et al., 1984; Hattori et al., 1984). Both envelope proteins are also expressed at the surface of HTLV-I-bearing cells. These HTLV-I glycoproteins are considered to be involved in both viral adsorption to and penetration into cells through specific virus receptors expressed on the surface of target cells. In fact, antisera raised against recombinant HTLV-I envelope polypeptides (Kiyokawa et al., 1984; Nakamura et al., 1987) or vaccinia virus containing HTLV-I env gene (Shida et al., 1987) were shown to neutralize HTLV-I infectivity. Monoclonal antibodies are useful in characterizing biological and immunological properties of viral antigens. For characterization of gp46, 2 MAbs had been generated from human (Matsushita et al., 1986) and mouse (Palker et al., 1989) origins. Both are reactive with one of the 2 major immunogenic regions of gp46 for humans (Palker et al., 1989; Ralston et al., 1989). For more information on gp46 antigenicity, new antigp46 MAbs directed to the other epitopes are required. In this report, we describe the generation and characterization of 6 anti-gp46 MAbs generated from mice and rats immunized with various immunogens.

respectively. The animals were maintained at the Institute of Laboratory Animal Science, School of Hygienic Sciences, Kitasato University. Cells HTLV-I-bearing T-cell lines used were MT-2 (Miyoshi et al., 1981), HuT102 (Poietz et al., 1980), F-Taj and F-Aki (Tanaka et al., 1985) and the ILT-MOR-I1 cell line, which was derived from a healthy HTLV-I carrier and expresses a cellsurface phenotype of CD4-CD8+ (data not shown). HTLVI-negative cell lines were Molt-4 (Minowada et al., 1972), CCRF-CEM (Foley et al., 1965), HPB-ALL and HPB-MLT (Morikawa et al., 1978), TALL-1 (Miyoshi et al., 1977), Daudi (Klein et al., 1968), Raji (Pulvertaft, 1965), K-562 (Lozzio and Lozzio, 1973), and HL-60 (Collins et al., 1978). Myeloma cell lines used were Sp2/0-Ag14 (Shulman et al., 1978) and YB2/0 (Galfre and Milstein, 1981) of mouse and rat origin, respectively. All these cell lines, except ILT-MOR-11, were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum (FCS), 100 unitdm1 penicillin and 100 pg/ml streptomycin (referred to as RPMI-1). The ILT-MOR-I1 cell line was cultured in RPMI-1 medium supplemented with 40 units/ml of IL-2 (Shionogi, Osaka, Japan). HeLa cells were obtained from the Japanese Cancer Research Resources Bank (JCRB, Tokyo, Japan) and cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% FCS and 20 pg/ml kanamycin (referred to as DMEM). An HTLV-11-bearing cell line, Mo-B (Chen et al., 1983) was obtained from the American Type Culture Collection (Rockville, MD), and maintained in Iscove’s modified Dulbecco’s medium supplemented with 20% FCS. Recombinant vaccinia virus Recombinant vaccinia viruses (RVV) containing HTLV-I entire envelope gene (WR-proenvl), gag gene (WR-gag) and pX gene (WR-27”) have been reported previously (Shida et al., 1987; Nam et al., 1988). For the expression of HTLV-I antigen by RVV, HeLa cells were infected with RVV for 1 hr at a multiplicity of infection (m.0.i.) of 10 and incubated for 16-18 hr in DMEM at 37°C. Immunogens The following 3 kinds of immunogen were used. (a)RW-infected cells. YB2/0 cells (5 X lo6 cells) infected with WR-proenvl at an m.0.i. of 4 and incubated for 18 hr, were extensively washed, resuspended in 0.5 ml phosphatebuffered saline (PBS) and disrupted by sonication.

MATERIAL AND METHODS

Animals Six-week-old female BALB/c mice, WKA/H rats and LEW rats were purchased from SLC Japan and Charles River Japan,

6To whom correspondence and reprint requests should be addressed. Received: April 20, 1990 and in revised form June 12, 1990.

676

T A N A K A ET A L .

(b) N147. Production of a recombinant envelope protein, N147, by a baculovirus expression system will be reported in detail elsewhere. Briefly, a part of the HTLV-I genomic DNA containing two-thirds of the HTLV-I env gene (AccI-PstI) was inserted in the polyhedrin gene of a transfer vector, pBF124 (Maeda et al., 1989). The resultant transfer plasmid pBF124N147 and wild-type Bombyx mori nuclear polyhedrosis virus (BmNPV) DNA were co-transfected into BmN cells to generate a recombinant virus BmNPV-N147. Infection of this virus led to production of a fused protein with 44 amino acids of the N-terminal part of polyhedrin followed by the C-terminal half of gp46 that was estimated to consist of 147 amino acids from Asp-166 to Arg-312 (data not shown). The recombinant protein was purified by preparative SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and dissolved in 0.1 M TridO.1 M tricineio. 1% SDS buffer (PH 8.25). (c) Partially purified gp46 (pp-gp46). Anti-gp46 MAb, MET-2, generated in the present study, was produced in ascites form, purified by Sephadex G-200 gel filtration and coupled to CNBr-activated Sepharose 4B (Pharmacia, Uppsala, Sweden) (2 mg IgG/ml gel) to form the bed of an immunoaffinity column. gp46 was collected from HTLV-I virus-free culture supernatants of either MT-2 or TCL-Kan cells (obtained from Fujirebio, Tokyo) by passing the material over the column at 4°C. After washing the column with a washing buffer (20 m~ Tris-HC1, PH 7.6,0.5 M NaC1, 1 m~ EDTA, 0.5% NP-40, 1% desoxycholate) and then with PBS, the bound materials were eluted with 0.2 M glycine-HC1, PH 2.6, and neutralized immediately with Tris-base. The eluates were concentrated by the Amicon ultra-filtration system with PMlO filter (Amicon, Danvers, MA). SDS-PAGE showed that this pp-gp46 had about 10% purity for gp46. Immunization (a) LEW rats were injected i.v. with the disrupted 5 x lo6 YB2/0 cells infected with WR-proenvl . Two weeks later, the rats were injected i.v. with 100 pg N147. (b) BALB/c mice were injected in each foot-pad with 40 pg N147 emulsified with Freund’s complete adjuvant (FCA). Two and 4 weeks later, each mouse was injected S.C. with N147 (40 pg) emulsified with Freund’s incomplete adjuvant (FIA). Two more weeks after final immunization, each mouse was injected i.v. with 40 pg N147. (c) WKNH rats and BALB/c mice were injected i.p. with pp-gp46 (50 p g and 25 p,g for rat and mouse, respectively) emulsified with FCA. Two and 4 weeks later, each rat or mouse was injected S.C. with pp-gp46 (50 pg and 25 pg for rat and mouse, respectively) emulsified with FIA. Two weeks later, each animal was injected i.v. with 10 kg PP-gP46. Preparation of hybridomas Immune spleen cells were isolated from mice and rats 3 days after the final booster immunization. Spleen cells from rats and mice and YB210 and Sp-2/0-Ag14 myeloma cells, respectively, were fused at a spleen-to-myeloma cell ratio of 10:1 and 4: 1 in the presence of 50% polyethyleneglycol 2000 (Tanaka el al., 1984). After HAT selection, hybridomas producing antibodies reactive with HTLV-I-bearing, but not HTLVI-negative cells were screened by an indirect immunofluorescence assay and cloned by limiting dilution. Immunojluorescence assay Live cells and cells smeared onto glass slides and fixed with methanol for 5 min at -20°C were stained with MAbs by an indirect immunofluorescence (IF) method using FITC-labelled sheep IgG anti-rat IgG or FITC-labelled goat IgG anti-mouse IgG (Cappel, Cochranville, PA) as the secondary reagent (Tanaka et al., 1986). Fluorescent cells were examined under

fluorescence microscope, or membrane fluorescence was analyzed by FACScan (Becton Dickinson, Mountain View, CA) after fixation of cells with 0.5% paraformaldehyde in PBS. For positive and negative control MAbs, TA21 anti-HTLV-I gp21 mouse MAb (Tanaka et al., 1985), a mouse MAb produced by P3IX63-Ag8 cells (X63) and a rat MAb anti-mouse IL-2R, 7D4 (Malek et al., 1983), were used. 7D4-producing cell line was obtained from the JCRB. lmmunoblot Cell lysates were obtained by lysis of 2 X lo7 cells with 1 ml of a low-salt extraction buffer (10 m~ Tris-HC1, PH 8.0, containing 0.14 M NaCl, 3 m~ MgCl,, 1 m~ dithiothreitol, 2 m~ phenylmethylsulfonyl fluoride, and 0.5% NP40) on ice for 20 min, followed by centrifugation at 12,000 g for 10 min at 4°C. The cell lysates were mixed with an equal volume of 2-fold concentrated sample buffer (Laemmli, 1970). For enrichment of glycoproteins, 1-ml cell lysates or 10 ml cell-free culture supernatants containing 0.5% NP40 were mixed with 0.1 ml Con A-Sepharose 4B beads (Pharmacia) at 4°C overnight. The beads were then washed in PBS, resuspended in 0.1 ml of the sample buffer. These samples were boiled for 3 min, subjected to SDS-PAGE and transferred to blotting sheets. Bindings of MAbs were visualized by a modification of the method of Manson et al. (1982), the “PAP method” (Lee et al., 1989). ELISA pp-gp46 antigen at 10 pg/ml in bicarbonate buffer, PH 9.6, was plated into the wells of U-bottomed 96-well plates (Corning 25802) (25 pl/well) and incubated at 4°C overnight. The wells were saturated with 0.5% gelatine in PBS (200 pl/well) at 37°C for 1 hr before reaction with MAbs. For competitive binding assays, the gp46-coated wells were pre-incubated with human sera (50 yllwell) for 1 hr at room temperature, and then with added MAbs (25 pl/well) diluted to give half maximum antibody binding values for 30 min. Culture supernatants of MAbs were used. After washing with PBS containing 0.5% Tween 20 (T-PBS), wells were incubated with horse-radish peroxidase (POD)-labelled rabbit IgG anti-rat or mouse IgG, which were non-reactive with human IgG, for 30 min at room temperature. Enzyme activities bound to wells were measured in 0.1% H,O, and 0.04% ortho-phenylenediamine in 0.1 M sodium citrate buffer, PH 5 . 5 , with a plate reader at 492 nm. Epitope mapping Materials and methods for synthesis of peptides and ELISA for epitope mapping have been reported by Kuroda et al. (1990). Briefly, peptides were synthesized manually with a stepwise solid-phase procedure (Mitchel et al., 1978), using tert-Butyloxycarbonylaminoacyl-4-oxymethyl-Pamresin as solid support. Peptide solution dissolved in 5% acetic acid was applied to a Sephadex G-25 column equilibrated with 5% acetic acid, and eluted with the same solution to remove byproducts of low molecular weight. The purity of individual peptides was examined by reverse-phase high-performance liquid chromatography using an octadecyl (C,,)-silicated column. The majority of peptides were isolated as a single peak and used for ELISA without further purifications. Wells of 96-well plates coated with peptides (500 ng/well) were incubated with 1:lO diluted culture supernatants of MAbs for 1 hr at 37°C. After washing, the wells were further incubated with PODlabelled goat IgG specific for either mouse or rat IgG. Bound enzyme activity was measured as described above. HTLV-I-mediatedcell fusion inhibition assay ILT-MOR-I1 cells were suspended in RPMI-1 at 2 X lo6 cells/ml, and aliquots (50 pl/well) were incubated with 100 pl of either culture supernatants or purified MAbs, human sera or

677

MAb AGAINST HTLV-1 ENVELOPE gp46 TABLE I - MAbs GENERATED IN THIS STUDY MAbs

REY-7

REY-11 REY- 16 MET-2 MET-3 REY-30

Originating from

Immunized with

Animalistrain

Rat/LEW Rat/LEW Rat/LEW Mouse/BALB/c

Ig class

First

Final

WR-proenvl’ WR-proenvl WR-proenvl

N1472 N147 N147 N147 PP-gP46 PP-gP46

N147

Mouse/BALB/c ~ p - g p 4 6 ~ Rat/WKA/H pp-gp46

IgG2, IgG2, IgG2, IgG, IgG, IgG2,

‘Recombinant vaccinia virus with HTLV-I env gene.-’Purified recombinant HTLV-I gp46 antigen expressing the C-terminal half 147 amino acids of g ~ 4 6 . - ~ g p 4 6 antigen purified from cell- and virus-free culture supernatants of HTLV-I producer cells with MET-2IgG affinity column.

medium alone in wells of U-bottomed 96-well plates (Coming 25850) at 37°C for 15 min, and then 50p1 of Molt4 cell suspensions (4 x lo6 celldml) were added to each well. After incubation at 37°C for 16 hr in a 5% CO, incubator, giant multinuclear cells (Kikukawa-Itamura et al., 1987) in the cultures were counted under a microscope with a hemocytometer.

new MAbs, only REY-30 reacted with Mo-B cells bearing HTLV-11. Membrane IF assays showed that these MAbs, except MET2, specifically stained the surface of HTLV-I-bearing cells. Figure 1 shows flow cytometry analysis of cell-surface staining of HTLV-I bearing MT-2 and HuT102 cells with REY-7, MET-2 and anti-HTLV-I gp21 TA-21MAb. REY-7 stained the cell surface of MT-2 and HuT102 cells. Similar staining was observed with FEY-11, FEY-16, REY-30 and MET-3 (not shown). In contrast, MET-2 did not stain the surface of these cells. These results indicate that the new MAbs react specifically with HTLV-I envelope antigen-positive cells, and that epitopes recognized by these MAbs, except MET-2, are also expressed at the surface of HTLV-I-bearing cells. Immunoblot analysis Immunoblot assays were performed to examine antigens recognized by these MAbs. Since REY-7 had been generated first, we examined the reactivity of REY-7 in detail. Figure 2 shows that REY-7 reacted preferentially with gp68 (a fusion protein between gp46 and pX antigen found only in MT-2 and related cells) in MT-2 cell lysates, gp68, gp62 and gp46 in the glycoprotein fractions from the cell lysates and gp46 in the cell-free culture supernatant glycoproteins. Reactivity with purified HTLV-I virion was not examined because sucrose-gradient purified HTLV-I virion contains little gp46. In the cell lysate glycoprotein fractions of MT-2 cells, REY-7 detected several additional proteins with molecular weights of 56, 49, 41, 38, 37 and 34 kDa. Some of these may be degraded or intermediate products of either the gp46, gp62 or gp68. In HuT102 cells, REY-7 detected 58-kDa protein and gp46 mainly in cell lysates and in cell-free supernatant glycoproteins, respectively, and additional 49- and 34-kDa proteins in cell iysate glycoproteins. In lysates of HeLa cells infected with WR-proenvl , REY-7 detected gp62 predominantly and gp46 and 70 kDa protein faintly. These antigens were not detected in WR-gag-infected HeLa cells. REY-7 did not react with Raji cells. These results, together with those obtained by IF assays, indicated that REY7 specifically reacted with HTLV-I gp46 and its related antigens.

RESULTS

To generate MAbs directed against various epitopes of gp46, we immunized both mice and rats with a recombinant vaccinia virus expressing HTLV-I env gene (WR-proenvl), a recombinant gp46 antigen, N147, expressing 147 amino acids located in the C-terminal half of the gp46, or affinity-purified gp46, and fused their spleen cells with mouse or rat myeloma cells. In this study, we have succeeded in the generation of 6 MAbs, REY-7, REY-11, REY-16, REY-30, MET-2 and MET-3. Table I shows the origins and Ig clasdsubclass of these MAbs. Cell specijicity Reactivity of the MAbs with various cells was examined (Table 11). IF assays with methanol-fixed cells showed that all the MAbs stained HTLV-I-bearing cells and HeLa cells infected with a RVV encoding the entire HTLV-I env gene, but not HTLV-I-negative cells or HeLa cells infected with control RVV encoding HTLV-I gag or pX gene. Among these

TABLE 11- CELL SPECIRCITY OF MAbs DETERMINED BY INDIRECT IMMUNOFLUORESCENCE ASSAY’ Reactivitv of MAb Cell

Rat IgG MAb REY-7

(HTLV-I-bearing cells) MT-2 HuT102 F-Aki F-Taj ILT-MOR-I1 (HTLV-I-negative cells) Molt4 HPB-ALL CCRF-CEM TALL-1 Raji Daudi K-562 PHA-PBL (n = 3)4 HeLa cells infected withS WR-proenvl

+

+3

++ ++

++

++ ++

-

-

Mouse IgG MAb

REY-16

REY-30

MET$

MET-3

++ ++ ++ ++

++

++ ++ ++ ++ ++

+ + + + +

++ ++ ++ ++ ++

-

-

-

-

-

-

-

-

-

-

-

-

-

+-+

+-+

+-+

+-+

+-

+-+

-

-

-

-

-

++

-

-

-

-

(HTLV-11-bearing cells) Mo-B

++ ++ ++

++ -

WR-gag WR-27”

REY-I1

-

-

-

-

-

-

-

‘Live and fixed cells were e~amined.-~MET-2 did not stain cell surface of live ~ells.-~Fluorescence intensity was expressed as + , strongly positive; positive ; and - , negati~e.-~Normal peripheral blood lymphocytes stimulated with PHA for 3 day~.-~HeLa cells were infected with RVVs and incubated for 16 hr.

+

+

678

TANAKA ET AL.

MT-2

CELL

HUT102 CELL

FLUORESCENCE INTENSITY FIGURE1 - Immunofluorescence of HTLV-I-bearing cells stained with REY-7, MET-2 and TA-21 anti-HTLV-I gp21. Live MT-2 and HuT102 cells were stained with either REY-7, MET-2 or TA-21 (antiHTLV-I gp21 mouse MAb). After fixation with paraformaldehyde, the cells were analyzed by FACScan. Broken lines show background staining of the cells with negative control MAbs, rat anti-mouse IL-2R MAb, 7D4, and mouse MAb x63.

The reactivities of the other MAbs generated subsequently were examined using MT-2 and HeLa cells infected with WRproenvl or WR-gag (Fig. 2). Like REY-7, all MAbs reacted with gp46 antigen in cell-free culture supernatants and gp62 expressed in WR-proenvl-infected HeLa cells. The reactivity of REY-30 was different from that of the others, as it did not react with gp68 antigen, which was confirmed by an immunoprecipitation assay using 35S-cysteine-labelledMT-2 cells (not shown), and it detected gp46 antigen more efficiently than the others, even in glycoprotein-non-enrichedcell lysates. HTLVI-non-specific reactivity was seen in REY-11, REY-16 and MET-2 which reacted with 32-kDa protein in HeLa cells infected or uninfected (not shown) with RVV. Then we examined the reactivity of the new MAbs with a recombinant gp46 antigen, N147, expressing the 147 amino acids in the C-terminal half of gp46 (Fig. 3). REY-7, REY-11, REY-16 and MET-2, which were generated from animals immunized with N147, and MET-3 reacted strongly with N147. REY-30 did not detect N147 under the present conditions (Fig. 3); however, when a larger dose of N147 (100 ngllane) was blotted, REY-30 detected a very weak band of N147 (not shown). Epitope mapping To locate the epitopes recognized by these MAbs, their reactivities with a panel of synthetic peptides corresponding to various regions of gp46 and gp21 were examined by ELISA (Table 111). REY-7 and MET-3, REY-11 and REY-16, and REY-30 bound specifically to peptide 175-199, peptide 253282, and peptide 288-317, respectively. MET-2 did not bind to any peptides used. The specificity of the binding of these MAbs to immobilized peptides was further demonstrated by

FIGURE2 - Immunoblot analysis of antigen recognized by MAbs. Samples tested were whole lysates of MT-2, HuT102 or Raji cells (lane l), glycoprotein fractions collected with Con-A from the cell lysates (lane 2), cell-free culture supernatants from MT-2 and HuT102 cells (lane 3) and whole-cell lysates of HeLa cells infected with either WR-proenvl or WR-gag. These samples were separated by SDSPAGE and transferred to blotting sheets, and then reacted with MAbs or PBS alone (control) as listed below. Bindings of MAbs were visualized by the PAP method.

inhibition of binding of MAbs by 10-fold excess concentrations of soluble peptides (not shown). These results, and the fact that MET-2 reacted with the N147, indicated that all of the present and anti-gp46 MAbs were directed against the C-terminal half of gp46 molecule, and that there exist at least 4 distinct epitopes on the C-terminal half of gp46. Competitive binding assay To determine the relation between epitopes recognized by these MAbs and those recognized by antibodies produced in HTLV-I-infected humans, competitive binding assays by

679

MAb AGAINST HTLV-1 ENVELOPE gp46 Binding of REY-7

Binding of MET-3 10

08 06 04

02

n.. o t

1000

100

10

10

1

1000

100

Binding of REY-16

Binding of REY-11

08

06

04

02

FIGURE3 - Immunoblot analysis of reactivities of MAbs with N147. N147 (50 ng/lane) separated by SDS-PAGE and transferred to blotting sheets were reacted with either anti-gp46 MAbs or control MAbs (rat anti-mouse IL-2R MAb, 7D4, and mouse MAb x63), or stained with amido black. Bindings of MAbs were visualized by the PAP method. The molecular weight of N147 is 18 kDa, and several bands with higher molecular weights detected by anti-gp46 MAbs may be either aggregated N147 or debris of aggregated N147.

ELISA were performed (Fig. 4). Wells coated with culturesupematant-derived gp46 antigen purified by MET-2 affinitycolumn chromatography were pre-incubated with sera from either HTLV-I-infected or uninfected humans. The binding of all the MAbs was inhibited specifically by pretreatment of gp46-coated wells with sera from HTLV-I-infected humans, but not with those from normal humans. These results indicate that epitopes recognized by these MAbs are identical to, or locate adjacent to, those recognized by antibodies of HTLVI-infected humans. Cell fusion inhibition assay Antibodies from HTLV-I-infected humans and also animal antibodies raised against the gp46 neutralized HTLV-I infectivity, as shown by inhibition of cell fusion of HTLVI-negative cells co-cultivated with HTLV-I-bearing cells (Hoshino et al., 1983; Nagy et al., 1983; Weiss et al., 1985; Kikukawa-Itamura et al., 1987). As shown in Table IV, the fusion of Molt4 cells by HTLV-I-producer ILT-MOR-I1 cells was completely inhibited by 1 5 0 dilution of an ATL patient serum. However, none of the MAbs inhibited the cell fusion under the present conditions. Similar results were obtained when purified IgG of each MAb was tested at up to 100 Fg/ml (not shown). DISCUSSION

In the present study, we succeeded in the generation of

00 10

1

1000

100

Bindina of REY-30

:I

06

.

00

,

10

REY-7 MET-3 MY-I1 REY-16 REY-30 MET-2 TA-21

1000

sera from HTLV-I- infected (0,. ) and uninfected (0, 0)humans were pre-incubated in microtiter wells coated with pp-gp46. Then the binding of anti-gp46 MAbs to the wells was determined by ELISA. various anti-HTLV-I gp46 MAbs from mice and rats. On the basis of their reactivity with synthetic peptides corresponding to gp46 sequence, the present MAbs were classified into 4 groups; (1) REY-7 and MET-3, (2) REY-11 and REY-16, (3) REY-30 and (4) MET-2. So far, 2 anti-gp46 MAbs, l l C l (Palker et al., 1989) and 0.5a(Matsushita et al., 1986), that had similar or identical reactivities with REY-7 and MET-3, have been generated from a mouse and a human, respectively. Peptide mapping showed that l l C l and 0.5a epitopes located in regions within the amino acids 190-209 (Palker et al., 1989) and 186-195 (Ralston et al., 1989), respectively. This region around the amino acid 190 was determined to be a highly immunogenic region for humans (Palker et al., 1989; Kuroda et al., 1990). Generation of REY-7 from rats indicates that the highly immunogenic region for humans is also immunogenic for rats.

gp21 peptide

4 9

175 -199

253 -282

O.0OO2

0.000

0.000

0.004 0.001

0.372 2.548 0.000

0.000 0.000 0.025 0.376

0.000 0.000

0.000

0.000

0.001

0.000

0.003

0.000

0.003

0.000 0.000

0.000

0.000

,

RECIPROCAL OF SERUM DILUTION

gp46 peptide

~

,

FIGURE4 - Inhibition of the binding of anti-gp46 MAbs to gp46 by sera from HTLV-I-infected, but not uninfected, humans. Dilutions of

89 -115

20

, , , ,100 ,,

1

TABLE I11 - MAPPING OF EPITOPES RECOGNIZED BY ANTI-HTLV-I gp46 MAbs WITH SYNTHETIC PEPTIDES BY ELISA MAbs

1000

100

Bindina of MET-2

oil I

w ,y ,A 10

1

0.000

0.000

0.002

288 -317

350 -386

400 426

458 487

0.005

0.001 0.000

0.000 0.004

O.OO0

0.000 0.000

O.OO0

0.006 0.005 0.002 0.845 0.002 0.004

0.001 0.OO0 0.001

0.003

0.002

0.004 0.003

0.000

~

0.1223

0.000 0.000 0.000 0.001

~~

‘MAbs (1:lO diluted culture supernatants)were reacted with various synthetic peptides (500 nglwell) at 37” C for 2 hr. Bound MAbs were detected with POD-labelled anti-rat or mouse IgG -’Values of OD,,2.-3Non-specific binding, because the soluble peptide, 458-487, could not inhibit the binding of MET-3 (not shown).

680

TANAKA ET AL.

TABLE IV - HTLV-I FUSION INHIBITION ASSAY' Antibody

tite?

Final dilution

ATL patient serum control MAb ( x 6 3 ) REY-7 REY-11 REY-16 REY-30 MET-2 MET-3

320

1.50

IF

0 640 320 160 160 40 160

1:2 1.2 1:2 1:2 1.2

1.2 1.2

Number of giant cells per well (xIo-~)~