Vol. 30, No. 2

JOURNAL OF CLINICAL MICROBIOLOGY, Feb. 1992, p. 351-358

0095-1137/92/020351-08$02.OO/O Copyright © 1992, American Society for Microbiology

Monoclonal Antibodies and Chemiluminescence Immunoassay for Detection of the Surface Protein of Human T-Cell Lymphotropic Virus LAWRENCE D. PAPSIDERO,1* RAYMOND P. DITTMER,1 LOUIS VAICKUS,2 AND BERNARD J. POIESZ3 Cellular Products, Inc.,' and Roswell Park Cancer Institute,2 Buffalo, New York 14202, and SUNY Health Science Center, Syracuse, New York 132103 Received 16 July 1991/Accepted 1 November 1991

Monoclonal antibodies (MAbs) raised against human T-cell lymphotropic virus type I (HTLV-I) recognized five distinct antigenic domains of viral env gene-encoded proteins. By using recombinant env proteins and synthetic peptides as mapping antigens, it was determined that the most immunogenic region represented a central portion of the retroviral surface protein (domain 2; amino acids 165 to 191). However, only a single MAb was able to react strongly with native viral proteins. This antibody (clone 6C2) was directed to an epitope within domain 4 (amino acids 210 to 306) of the retroviral env gene and reacted with envelope proteins in both HTLV-I and HTLV-ll, as determined by immunoprecipitation, solid-phase binding, and immunoblotting. No reactivity against envelope components of other human retroviruses, including human immunodeficiency virus types 1 and 2, was present. Flow cytometry data demonstrated that MAb 6C2 reacted with cell lines chronically infected with HTLV-I or HTLV-II and also with surface antigens expressed on fresh adult T-cell leukemia cells, following up-regulation with interleukin-2. By a chemiluminescence immunoassay procedure, picogram amounts of viral surface protein could be detected in the unconcentrated supernatants of HTLV-infected cell lines and in diagnostic cultures. Levels of env and gag proteins released by cells into culture supernatants were not directly related to percent expression of cell surface viral-coat proteins. Further, the molar ratio of p19 to gp46 in conditioned media varied from strain to strain, possibly reflecting differences in viral assembly or packaging mechanisms. MAb 6C2 will be of value in characterizing the biochemical and immunological behavior of retroviral env gene proteins and in studying the interaction of HTLV-I and HTLV-H with their receptors.

Human T-cell lymphotropic virus type I (HTLV-I) is associated with various forms of lymphoproliferation and leukemogenesis, including polyclonal lymphocytosis of T cells (44), adult T-cell leukemia-lymphoma (ATL) (12, 42, 54), and B-cell chronic lymphocytic leukemia (33). Seroepidemiologic and molecular biology studies have also associated this viral infection with encephalomyeloneuropathies (2, 17, 20, 27) and polymyositis (53). HTLV-I infections have been reported as the leading cause of hematopoietic neoplasms in Japan and in the Caribbean, where the rate of infection among asymptomatic individuals can be as high as 37% (3, 14, 18). HTLV-I infections and associated leukemias may be more common in the United States than previously recognized (31, 35, 46). In addition, infections with a related retrovirus, HTLV-II, have most recently been demonstrated at high prevalence within certain risk groups (10, 19, 26, 47). Envelope glycoproteins of animal and human retroviruses, encoded by the retroviral env gene, contain antigenic regions which play an important role in viral neutralization and also in determining the host range, strain specificity, and relative infectivity of the virus (1). Evidence also suggests that the HTLV env glycoprotein may be a critical antigen for targeting by the host immune system. For example, the HTLV-I env glycoprotein is the major viral antigen recognized by antibodies in the sera of infected individuals (5, 8, 29, 37, 48), and these antibodies are able to neutralize the infectivity of pseudotype virions which carry HTLV-I env proteins (6). In other retroviral infections, particularly human immunodeficiency virus (HIV) infections, the most important neutraliz*

ing antibodies react with envelope proteins (for a review, see reference 4). In order to further characterize the env gene products of HTLV-I, we have generated a panel of murine monoclonal antibodies (MAbs) to the surface protein of the virus. In addition to the capacity of these antibodies to identify various antigenic domains, one antibody (clone 6C2) was a sensitive probe for the detection of HTLV-I and HTLV-II env proteins on live cells and has also been utilized to construct a chemiluminescence immunoassay (CLIA) for the quantitative measurement of HTLV surface proteins. MATERIALS AND METHODS Abbreviations. Abbreviations used in this paper are as follows: BSA, bovine serum albumin; HIV type 1, HIV-1; IL-2, interleukin-2; mlg, nonimmune murine immunoglobulin; PBMC, peripheral blood mononuclear cells; PBS, 50 mM phosphate-0.9% sodium chloride (pH 7.2); PBS-M, PBS containing 5% (wt/vol) nonfat bovine milk solids; PBS-T, PBS containing 0.05% Tween 20; RIPA, radioimmunoprecipitation assay; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Cell culture. Continuous cell lines were grown in RPMI 1640 medium containing 10% fetal bovine serum and antibiotics (complete medium). Cell-free culture supernatants were obtained at log phase of growth (0.8 x 106 to 1.2 x 106 cells per ml) by sedimentation of the cells at 5,000 x g; the supernatants were then filtered through 0.45-,um-pore-size filters. Viral cultures were done as described previously (38). Recombinant proteins. Escherichia coli-expressed recombinant proteins from the HTLV-I env gene were obtained

Corresponding author. 351

PAPSIDERO ET AL.

352

HTLVLI

L

Clew"

pi2'-3131

gee

-F

wa

V6-?

RIEl1

-A

J. CLIN. MICROBIOL.

2111

6l I REssRE-3, ICPPll

17l_ 7]>

C .6-164 (i2L..§

,s H1"1

I

F

a

"

p6; J

2

3

Andwnbssu&

4

FIG. 1. Locations of recombinant protei and synthetic peptides used in the present study along with tns the antigenic domains which correspond to regions of the HTLV-I env gene. Numbers in parentheses refer to linear sequences of am ino acids according to the numbering system of Seiki et al. (49).

from Repligen Corp. The precise locatic ns of these proteins within the viral env gene are shown in Fi was a recombinant p24 core protein deirived from the gag gene of the HTLV-I genome (p249ag; Triton Bfomctengag Inc.). Synthetic peptides. Peptides from the HTLV-I env gene were synthesized (34) at Multiple Peptidle Systems, Inc. The precise location of each peptide within tihe HTLV-I env gene is shown in Fig. 1. The synthetic peptidle CP-11 sequence is VDAPGYDPIWFLNTEPSQLPPTAPP] L; the synthetic peptide CP-12 sequence is LLPHSNLDHI]LEPSIPWKSK. Virus isolation. Sucrose-banded virionIs were isolated from cell-free supernatants of the cell lines i IUT-102 (42), Mo T (21), LAV-2 (7), and HIV-1/HUT-78 (4(0). MAb production. Female BALB/c miice were primed and boosted after 3 months with 50 ,ug of HTLV-I protein emulsified in complete Freund's adjuv,ant. Thereafter, the animals were boosted with recombinanIt HTLV-I envelope proteins (RE-1, RE-3, and RE-5 [10 p.Eg each]). The fusion protocol utilized P3x63.Ag8.653 myelorna cells in the presence of 40% polyethylene glycol (16). C ontrol MAbs used in the present study included clones lC1l (anti-HTLV-I gp46)

(37), W6/32 (anti-human leukocyte antigen, monomorphic determinant) (41), and F5 (anti-prostate antigen) (39). Immunoblotting. Viral proteins were subjected to immunoblotting as described previously (38), except that chemiluminescent substrate (ECL Western Blot Kit; Amersham) was used. Solid-phase enzyme immunoassays. Polystyrene microwells were coated with target antigens in 50 mM carbonatebicarbonate buffer (pH 9.6) for 18 h at 4°C and allowed to react with experimental antibodies as described previously (9). Competitive inhibition with synthetic peptides. Synthetic peptides were dissolved in PBS-T. Solid-phase target antigen represented microwells which were coated with recombinant HTLV-I env proteins. For competition analysis, various concentrations of

synthetic peptides

were

allowed to react

within antigen-coated microwells in the presence of MAbs developed as for 60 min at 37°C. The immune reactions were ... described above, and percent specific inhibition was calculated as described previously (40). RIPA. Cells were labeled in culture media containing [35S]cysteine and [35S]methionine and subjected to RIPA as described elsewhere (9, 10). CLIA for detection of HTLV env protein. Samples were applied to nitrocellulose paper (0.2-p.m pore size; Schleicher &

Schuell)

Schleicher

with &

a

dot

Schuell).

manifold apparatus (Minifold the entire

Thereafter,

II;

sheet was

blocked for 1 h at 37°C in PBS-M. After the membrane was rinsed in PBS, it was incubated with MAb 6C2 (10 jig/ml in PBS-M) for 1 h at 37°C and then overnight at 4°C, with agitation. The membrane was washed (three times, 5 min each time) in RIPA buffer and then allowed to react with biotin-conjugated caprine antibodies to murine IgG (Jackson Laboratory) (1:500 dilution in PBS-M) for 1 h at 37°C. Following another wash step, streptavidin-peroxidase was applied (1:1,000 dilution in PBS-M) for 30 min at 37°C. The

washed membrane

was

then incubated for 1 min in chemi-

luminescent substrate (ECL reagents; Amersham) and used to expose XAR-2 film. The developed film was subjected to visible-light spectrometry (405-nm wavelength) in order to quantitate antigen levels against a dose-response curve generated with serial dilutions of recombinant HTLV-I env protein. The concentrations of viral surface antigen in samples were determined by interpolation of the standard curve by linear regression.

TABLE 1. Immunoreactivity of MAbs to recombinant proteins and synthetic peptides Clone

4B1 3A2

5B5 3A5 5A2 4B4 5B1 SA1

lClld 6C2 6A4

Isotype

IgG2b IgGl IgGl IgGl IgGl IgG2b IgG2b IgGl IgG IgG IgG

Reactivity with: RE-la

RE-3a

RE-5a

CP_11b

CP-12b

++ +++ +++ + +++ +++ ++ +++

-

-

NEG POS POS POS POS POS POS POS NEG NEG NEG

NEG NEG NEG NEG NEG NEG NEG NEG POS NEG NEG

+++ +++ + +++ +++

+++ +++

-

-

-+++ -+++ -+++

Domain'

1 2 2 2 2 2 2 2 3 4 5

a Immunoreactivity to RE-1, RE-3, and RE-5 recombinant proteins was determined by solid-phase enzyme immunoassay. Signal-to-noise designations are: -, c2; +, >2 to 4; + +, >4 to 8; + + +, >8. b Immunoreactivity to CP-11 and CP-12 synthetic peptides was determined by competitive inhibition of MAb binding to solid-phase target antigen. NEG, no inhibition; POS, >95% inhibition of antibody binding. ' Domains correspond to antigen domains within the HTLV-I env gene open reading frame diagrammed in Fig. 1. d Reference MAb to HTLV-I gp46 (37).

IMMUNODETECTION OF HTLV SURFACE PROTEIN

VOL. 30, 1992

353

1.25

1.00I

~0

0.75 1

E

z

0.50

E C

0

0.00

U,

ii

-

/E

I-.rS r-r L IsSFI 100

50

0)

/4

0

[

0.25

0

0

-

150

mAb 6C2 (ng/ml)

U

~0 0 D

0 (A

Log Fluorescence FIG. 2. Immunoreactivity of MAb 6C2 with the surfaces of live cells as determined by flow cytometry. HUT-102 cells were treated with MAb 6C2, MAb W6/32 (anti-human leukocyte antigen framework epitope) (41), or control mlg (solid histogram). MAb 6C2 staining was dose dependent (results for 1-,ug, 0.1-,ug, and 100-ng doses not shown) and was detectable at concentrations as low as 10 ng.

Immunofluorescence phenotyping. Surface-marker phenotyping was performed as previously described (52). Immunofluorescence was expressed as percent positive cells after subtraction of background staining with mlg. HTLV-I p19 antigen quantitation. Measurements of HTLV-I p19 (viral matrix protein) were performed by antigen capture immunoassay, as described previously (38), with commercial reagents (Retro-Tek HTLV-I Antigen ELISA kit; Cellular Products, Inc.). This immunoassay is HTLV-I specific and does not cross-react with HTLV-II (38).

2.0 1

..

.

/0

1.51 / A

HTV-I

1.0 0.5

1-2 ""_.

0.0 I 0

500

*\1 1000

1500

mAb 6C2 (ng/ml) FIG. 3. Immunoreactivity of MAb 6C2 with recombinant and native viral antigens with which microwells were coated. (A) Recombinant viral proteins. RE-1, RE-3, and RE-5 are recombinant proteins corresponding to regions of the HTLV-I env gene, as illustrated in Fig. 1. p249aB is a recombinant protein derived from the gag gene of HTLV-I. All wells were coated with 20 ng of protein. (B) Natural viral proteins. Sucrose-banded preparations of HTLV-I, HTLV-II, HIV-1, and HIV-2 were used. All wells were coated with 100 ng of protein. Abscissa, concentration of MAb 6C2 (nanograms per milliliter); ordinate, optical density.

RESULTS

Antigenic domains of HTLV-I envelope proteins. Hybridantibodies were screened for immunoreactivity against recombinant HTLV-I env proteins RE-1, RE-3, and RE-5 (Fig. 1). For comparison, a reference MAb was incorporated into the experimental scheme. This antibody, lCll, has been shown to identify domain 3 (Fig. 1) of the HTLV-I gp46 molecule (37). Of the 10 experimental hybridomas studied, 1 (clone 4B1) reacted solely with protein RE-1, 1 (clone 6C2) reacted solely with RE-3, and 1 (clone 64A) reacted solely with protein RE-5 (Table 1). In contrast, 7 of the 11 hybridomas reacted with both RE-1 and RE-3 proteins. This indicated a high degree of immunogenicity within the overlap region of these recombinant proteins (i.e., amino acids 165 to 209). To further resolve the RE-1-RE-3 overlap region identified above, competition analyses with synthetic peptides were performed. The synthesized peptides completely spanned the overlap region and represented antigenic domains 2 and 3 (Fig. 1). Peptide CP-12 at a dose of 10 ,ug completely inhibited the immunoreactivity of reference MAb lC1l, corroborating previous data with synthetic peptide 4A (37). In contrast, the antigen-binding activity of none of the seven experimental MAbs which interacted with the RE-1RE-3 overlap region was perturbed in the presence of oma

peptide CP-12. However, the immunological activity of each of these antibodies was neutralized in the presence of 10 ,ug of peptide CP-11, which corresponds to the upstream region of the RE-1-RE-3 overlap domain (i.e., amino acids 165 to 191) (Fig. 1; Table 1). Each MAb was also examined for its capacity to react with naturally occurring viral protein(s). Within the entire group of antibodies, only one strongly reacted with native viral proteins. This MAb (clone 6C2) bound to the surface of infected cells, and results indicated levels of surface antigen approximately 10-fold lower than those of human leukocyte antigen sites (Fig. 2). Antigenic specificity of MAb 6C2. In direct-binding immunoassays, MAb 6C2 strongly reacted with recombinant HTLV-I env protein RE-3, derived from the C-terminal portion of gp46 (Fig. 3A). No significant binding to env proteins RE-1 or RE-5 was detectable, nor was the antibody reactive with recombinant HTLV-I p249aB protein. In addition, the antibody reacted with native viral proteins in sucrose-banded preparations of HTLV-I and HTLV-II (Fig. 3B). HIV-1 and HIV-2, purified under identical conditions, were nonreactive. Structural analysis of the protein(s) recognized by MAb 6C2 was performed by RIPA and immunoblotting tech-

354

~ ~.

PAPSIDERO ET AL.

Ml4-2

J. CLIN. MICROBIOL.

HUT--102

-.

5LBE- 1

4-

MoT

+i

kDa~~~3a6~

68

45 kWo 3OkWo

:.......

FIG. 4. Structural analysis of native HTLV proteins which are recognized by MAb 6C2 as determined by RIPA. Cells were biosynthetically labeled with [35S]cysteine and [35S]methionine, precipitated with MAb 6C2 or control Ig, and then analyzed by SDS-PAGE. Radiolabeled molecular-mass markers are shown in lane 1, and their sizes are indicated in kilodaltons. They included BSA (68 kDa), ovalbumin (45 kDa), and carbonic anhydrase (30 kDa). Lanes marked "+" represent immunoprecipitations performed with MAb 6C2; lanes marked "-" represent immunoprecipitations performed with a control subclass-matched murine MAb (clone F5, anti-prostate antigen) (39). Asterisks indicate proteins which were specifically precipitated with MAb 6C2.

niques. As seen in Fig. 4, the antibody precipitated proteins between 46 and 60 kDa from HTLV-I- and HTLV-II-infected cells. The presence of mature viral surface protein, gp46, was best resolved in the HUT-102 cell line (Fig. 4). Specifically immunoprecipitated proteins were not detected from uninfected T-cell lines (CEM and HUT-78) or in cells infected with HIV-1 or HIV-2. Reduced and denatured viral proteins which reacted with MAb 6C2 were detected by immunoblotting. Immune reactions with viral surface proteins from both HTLV-I and HTLV-II were observed. As shown in Fig. 5, reactive proteins exhibited molecular masses between 46 and 48 kDa. No reactivity was present with HIV-1 or HIV-2 specimens. Reactivity of MAb 6C2 with live cells. The fate of MAb 6C2

1

2

3

4

92.5-

m

-

45

-

z 0

Log Fluorescence FIG. 6. Fate of MAb 6C2-gp46 surface complex. HUT-102 cells (106) were treated with 10 FLg of MAb 6C2 for 30 min at 4°C, washed twice, and placed in culture at 37°C for 20 h (open histograms). As a negative control, HUT-102 cells were treated identically, except for the omission of MAb 6C2 (solid histogram). After 20 h in culture, aliquots of MAb 6C2-treated and control cells were stained with mlg alone (top panel) or with 10 pg of MAb 6C2 (bottom panel) followed by the secondary fluoresceinated antibody (caprnne F(ab')2 antimurine Ig) (both panels). The top panel demonstrates residual MAb 6C2 on the cell surface of 6C2-treated cells (open histogram). The bottom panel demonstrates that most surface protein-binding sites on MAb 6C2-treated cells are available and that additional MAb 6C2 can bind (open histogram) with a fluorescence pattern similar to that of untreated cells (solid histogram).

.;; .'.t ,. s,

,KigwmL.

:7ppwi ...:.:,:

*0

":

30-

21.514,4

FIG. 5. Structural analysis of reduced and denatured viral proteins which are recognized by MAb 6C2 as determined with immunoblotting. The migration of molecular-mass standards is indicated on the left in kilodaltons. Lane 1, HTLV-I; lane 2, HTLV-II; lane 3, HIV-1; lane 4, HIV-2.

on the cell surface (6C2-gp46 complex) was examined by staining HUT-102 cells with saturating doses of MAb 6C2, washing away excess MAb, and incubating the cells at 37°C for 20 h. After the incubation, the cells were probed with fluoresceinated caprine F(ab')2 anti-murine Ig to detect surface-bound MAb 6C2. Additional MAb 6C2 was added to a separate aliquot of pretreated cells to determine whether free MAb 6C2 binding sites were available (Fig. 6). Significant amounts of MAb 6C2 remained bound to the cells after the 20-h incubation. However, additional MAb 6C2 could bind to pretreated cells with an intensity similar to that of binding to HUT-102 cells not pretreated with MAb 6C2. This suggested that during incubation, most MAb-gp46 complexes were shed from the cell surface or were internalized

IMMUNODETECTION OF HTLV SURFACE PROTEIN

VOL. 30, 1992

355

TABLE 2. Coordinate expression of HTLV antigens in culture supernatants and on the surfaces of live cells

Description HUT 102 MT-2 SLB-1 Mo T H9/HTLV-IIIB HUT-78/HIV-1 HUT-78/HIV-2 HUT-78 MOLT-4 1937 CEM K562 Raji Normal T Normal Td ATL Tf ATL Tf CLL9 Neutrophilsd

Concn (ng/ml) of antigen:

(reference)

gp46b

HTLV-I infected (42) HTLV-I infected (54) HTLV-I infected (24) II infected (21) HIV-1 infected (43) HIV-1 infected (40) HIV-2 infected (7) Uninfected T cell (13) Uninfected T cell (51) Uninfected monocytoid (15) Uninfected T cell (11) Uninfected erythroleukemia (32) Uninfected B cell (45) Noncultured Post-culture in IL-2e Noncultured Post-culture in IL-2e Noncultured Noncultured

0.2 9.6 1.9 200