Which Mediates Cellular Cytotoxocity and Neutralization

JOURNAL OF VIROLOGY, Feb. 1990, p. 936-940 Vol. 64, No. 2 0022-538X/90/020936-05$02.00/0 Copyright C) 1990, American Society for Microbiology A Mon...
2 downloads 0 Views 642KB Size
JOURNAL OF VIROLOGY, Feb. 1990, p. 936-940

Vol. 64, No. 2

0022-538X/90/020936-05$02.00/0 Copyright C) 1990, American Society for Microbiology

A Monoclonal Antibody to Human Immunodeficiency Virus Type 1 Which Mediates Cellular Cytotoxocity and Neutralization PER A.

BROLIDEN,l.2.3* KRISTINA LJUNGGREN,2'3 JORMA HINKULA,1 ERLING NORRBY,2

LENNART AKERBLOM,4 AND BRITTA WAHREN1 Department of Virology, National Bacteriological Laboratory,l* Departments of Virology2 and Immunology,3 Karolinska Institute, Stockholm, and Department of Veterinary Microbiology, Biomedical Center, Uppsala,4 Sweden Received 5 June 1989/Accepted 13 October 1989

Monoclonal antibodies (MAbs) were raised against human immunodeficiency virus type 1 gpl20. One MAb, P4/D10, was found to mediate highly efficient antibody-dependent cellular cytotoxicity and virus neutralization. The reactivity was located to a major neutralizing region (amino acids 304 to 323) on gpl20. Five other MAbs with a similar epitopic reactivity did not show any antibody-dependent cellulan cytotoxicity activity but had a virus-neutralizing capacity.

Understanding the biological function of specific antibodies against human immunodeficiency virus (HIV) is important for the development of a vaccine. One of the potentially important mechanisms of protection against viral spread is antibody-dependent cellular cytotoxicity (ADCC) (12). In this reaction, HIV-specific antibodies bind to HIV antigens on the surface of infected cells, which are then killed by Fc receptor-positive effector cells. It has been shown that ADCC to HIV-infected cells is mediated by human immunoglobulin Gl (IgGl) (13) and also that HIV envelope glycoprotein constitutes a target for this reaction (12, 15). HIV type 1 (HIV-1)-specific neutralizing antibodies that are able to inhibit viral infection in vitro have been reported

seronegative controls including nonimmunized mice were included in each test. Virus neutralization was performed as follows. Virus supernatant (reverse transcriptase titer, 40.000) was preincubated with serial dilutions (six steps starting with 1:20) of the MAbs for 60 min at 37°C. The serum-virus mixture was added to 5 x 104 peripheral blood mononuclear cells or HUT-78 cells for 60 min at 37°C. After being washed, the cells were cultured in 96-well plates for 8 days. Supernatants were then analyzed by HIV antigen capture enzyme-linked immunosorbent assay (ELISA) (V. A. Sundqvist, J. Albert, E. Ohlsson, J. Hinkula, E. M. Fenyo, and B. Wahren, J. Med. Virol., in press). Neutralization was defined as >80% reduction of p24 viral antigen production. Solid-phase-synthesized (8) 15-amino-acid (aa) peptides with an overlapping sequence of 10 aa representing the complete region of the envelope (Env) protein based on the human T-cell lymphotropic virus type "'B sequence (23) were used as antigens in ELISA. They were a kind gift from J. Rosen (Johnson & Johnson Biotechnology, La Jolla, Calif.). The recombinant proteins pE3 (gpl20), pBl (gpl20), and penv9 (gp4l) were generous gifts from J. Ghrayeb (Centocor, Malvern, Pa.), Scott Putney (Repligen, Corp., Cambridge, Mass.), and S. Petteway (Du Pont Co., Rockville, Md.), respectively. Amino acid numbering of the Los Alamos database (17) was used. Peptide ELISA has been described previously (3, 27a). To further evaluate the specificity of the reactivity, we used the peptides as soluble antigens to inhibit the MAb reactivity in the ELISA (J. Hinkula, J. Rosen, V.-A. Sundqvist, T. Stigbrand, and B. Wahren, Mol. Immunol., in press). MAbs were preincubated with each peptide for 120 min at 37°C and then transferred to pBl-coated microplates, where ELISA was performed. One of the MAbs, P4/D10, was able to mediate high HIV-specific ADCC (Fig. 1). This ADCC activity could also be shown with human HIV antibody-positive sera but not with the other MAbs F58/H3 and T1.1 or with human HIV-seronegative sera. Nor was the reactivity seen when P4/D10 was tested against uninfected target cells. Of the five MAbs with similar epitopic reactivity and the ability to mediate virus neutralization, two are presented in Table 1. In addition, an MAb with a different epitopic reactivity to the N-terminal end of gpl20 is presented as a control. The neutralizing capacities of P4/D10 and F58/H3

(24, 28), although the protective role of these antibodies against HIV infection in humans is controversial. Several authors have reported that HIV-1-neutralizing antibodies can be produced against various regions of gpl20 (7, 11, 16, 19, 21), gp4l (9), and p17 (20). One major site inducing neutralizing antibodies has been described as a hypervariable loop of gpl20 (4, 25). Even though several neutralizing monoclonal antibodies (MAbs) (16, 26) have been produced, no MAb mediating ADCC was previously described. In this report, we present an MAb which mediates both ADCC and neutralization. The region of gpl20 to which it is directed was identified by peptide mapping. The gpl20 used as an immunogen was prepared from culture fluid of HIV-1 (human T-cell lymphotropic virus type III)-infected H9 cells and was a kind gift from Larry Arthur, Frederick Cancer Research Center, Frederick, Md. NMRI mice (National Veterinary Institute, Uppsala, Sweden) were immunized with gpl20 five times, given at monthly intervals. Fusion of spleen cells was performed with Sp20x Agl4 mouse myeloma cells. MAbs were characterized by isotypes, Western blot (immunoblot) (27), and immunofluorescence. The ADCC was determined as described previously (12, 14). Cells of the monocytoid line U-937 clone 2, chronically infected with HIV-1 strain human T-cell lymphotropic virus type IIIB used as targets, and peripheral blood mononuclear cells from normal healthy donors as effector cells were incubated with serum or MAb dilutions in a 3-h 51Cr release assay. The spontaneous release never exceeded 10%. HIV antibody-positive sera with known ADCC titers as well as *

Corresponding author. 936

VOL. 64, 1990

NOTES

937

60 -

50 w V-4 0

u,

ej

40~

30~ 10

*

-

20 7

I

30

90

270

810

2430

7290

I

Ly

Reciprocal serum dilutions FIG. 1. HIV-specific ADCC titers for the MAbs P4/D1O and F58/H3 compared with positive and negative human serum controls. Symbols: [D, P4/D1O; *, F58/H3; *, positive human sera; *, negative human sera; *, Ly.

were high (Table 1). Both neutralized human T-cell lymphotropic virus type IIIB infection of peripheral blood lymphocytes and the CD4+ cell line HUT-78. MAb F58/H3, which was unable to mediate ADCC, had a high neutralizing titer, while the ADCC-positive P4/D1O MAb had a less efficient but still significant neutralizing capacity. Neither of the MAbs showed any toxic effects in the assays. In ELISA, P4/D1O and F58/H3 showed a strong reactivity with pBl. Fine mapping of the epitopes of these two MAbs was performed with synthetic peptides representing the whole pBl sequence (Fig. 2a and b). A specific reactivity of both P4/D1O and F58/H3 was found only to peptides C53 and C54, which represent aa 304 to 323. In addition, specific peptide blocking was performed (Fig. 2c and d). A complete inhibition of reactivity of both MAbs was seen with peptide C53, while C54 only partly inhibited the MAb binding to pBl. Thus, an MAb that mediates HIV-specific ADCC and neutralization was identified, and the specificity appeared to be directed toward an epitope located on two overlapping peptides, of which the common sequence is IQRGPGRAF (Table 1). GPGRA is a conserved sequence within a variable putative loop region which has been found to induce neutralizing antibodies in animals (25). No clear correlation has previously been demonstrated between neutralization and ADCC, but it is reasonable to suggest that there are several ADCC and neutralizing epitopes. However, only one ADCC-mediating epitope at the very C-terminal part of gpl20 has been suggested previously (2). A correlation was

found between high ADCC titers and reactivity to aa 304 to 313 in a few human sera (unpublished data). The importance of the isotype of murine MAbs mediating ADCC with human effector cells varies in different studies. IgG2a and IgG3 were the predominant isotypes for ADCC (1, 6, 10), but efficient ADCC mediated by IgGl and IgG2b has also been reported (5, 18, 22). The two MAbs described here reacted with the same peptides but differed in ADCC and neutralizing activity. The region which induces neutralizing antibodies (aa 296 to 331) might include a large proportion of the protein, while the epitope inducing ADCCmediating antibodies appears to be a limited sequence within this neutralization-inducing region. -Neutralizing antibodies are by themselves capable of inhibiting infection by binding directly to the virus, whereas the ADCC-mediating antibodies act by binding to infected cells and depend on a second binding of an effector cell. The ADCC activity might therefore be more sensitive to small steric changes of the binding site. Other possible explanations for the difference in functional activities could be individual variations in the Fc receptors of the human effector cells or in the Fc portions of the MAbs. Further studies are necessary to characterize such differences. Epitope mapping with sequentially overlapping pentadecapeptides allows determination of epitopes of 5 to 15 aa but small variations in the reactive site not detected by the linear peptides could apparently elicit a variation of functional activities of the corresponding antibodies. The ability to

938

J. VIROL.

NOTES QQOOOO00000000000 OQV VV V V V V V V V V VV V

Hi

ON

C

-

o

0-4 o

-0

+

G)0. 0 0

0. I.

.1

00

v

0

.

C0

0O

.

v O o

.

u0o

v v v C° Cs VVV0°

OO°QON 000

VV OOO

v v

Q to

+

o

0+

0

+

-

2

r-

0

O

O

00

t4l)

QOO 8 O o -0 v

)

C>

OOQQQ o 0 o '-4 '-4 6 C00

o 8 o ON o 0 00

.

vO

tn

,,

v v v g8

O00

0o

4

00t

v v

6 co 00 \o ON ON 0

oN oN 0

e'- 0

C(

o 0 0.

6O

0

13)U,

HH

U

PA P4 pq Eo E21

C,

Hp 04

a' CFR

13) 13)

0

p

HSl

H

H G

H

0'

P4 pe.4 E-4

E- p qp E-4 00M

HO

Ag PP4 E-4 EqOz H

A

E-4 E-

V2

HH

0'@

E 131 0 ._

0

0

cU .0 o.

0

0

ea

130 (13~~~~~~~~

13)

,D .0 N

'O

U.) u

0

co

_

e

ro ^ o