Journal of Virulogical Methods, 31 (1991) 77-92

Elsevier 17

VI~MET 01106

Sensitive detection of anti-human T-cell leukemia virus type I IgG in human serum by a novel enzyme immunoassay (immune

complex transfer enzyme immunoassay) using recombinant gag-env hybrid protein as antigen

H. Kohno’, T. Kohno’, I. S&coda3 and E. Ishikawa’ ‘Fuji Research Laboratories, Kyowa Medex Co. Ltd., Shizuoka, Japan

2Department of Biochemistry, Medical College of Miyazaki, Kiyotake, Miyazaki, Japan and jMiyazaki Blood Center, The Japanese Red Cross. Miyazaki, Japan

(Accepted 17 September 1990)

Summary

Anti-hums T-cell leukemia virus type I IgG (anti”~V-I IgG) in human serum was detected with high sensitivity by a novel enzyme immunoassay (immune com- plex transfer enzyme immunoassay) using recombinant gag{ 14-139)-env( 197-295) hybrid protein. Anti-HTLV-I IgG in test serum was reacted- simultaneously with dinitrophenyl bovine serum albumin-recombinant gag-env hybrid protein conjugate and recombinant gag-env hybrid protein-horseradish peroxidase conjugate. The complex formed of the three components was trapped onto polystyrene balls coated with vanity-paled anti-dinitrophenyl group IgG. After washing the polystyrene balls to eliminate nonspecific IgG in the test serum and excess of the peroxidase conjugate, the complex was eluted from the polystyrene balls with dinitrophenyl-l- lysine and transferred to polystyrene balls coated with affinity-purified anti-human IgG y-chain IgG. Peroxidase activity bound to the polystyrene balls was assayed by fluorometry. The nonspecific binding of peroxidase. activity was remarkably re- duced by transfer of the complex and the detection limit of anti-FITLV-I IgG in serum was lowered 300 to 3000-fold compared with that by Western blotting and the conventional enzyme immunoassay, in which a recombinant gag-env hybrid protein-coated polystyrene ball was incubated with the test serum and, after wash-

Cormspondence to: E. Ishikawa, Department of Biochemistry, Medical College of Miyazaki. Kiyotake, Miyazaki 889- 16, Japan.

Oar-8510~1/$03.5~ 1991 Elsevier Science Publishers B.V. (Biomedical Division)

78

ing, with anti-human IgG y-chain Fab’-peroxidase conjugate. The immune complex transfer enzyme immunoassay may overcome some difficulties with currently used methods.

Antibody; Adult T-cell leukemia; Peroxidase; Western blotting; Gelatin particle agglutination; ELISA

Introduction

Human T-cell leukemia virus type I (HTLV-I) is etiologically associated with adult T-cell leukemia (Hinuma et al., 1981) and adult T-cell cancers (Gallo et al., 1983). The virus is transmitted by blood transfusion (Okochi et al., 1984) and breast-feeding (Ando et al., 1989). In order to prevent transmission of the virus, plasma or serum of blood donors and pregnant women has been screened for anti-HTLV-I antibodies by gelatin particle agglutination (Ikeda et al., 1984) or enzyme immunoassay (Taguchi et al., 1983). However, there are some difficulties with these methods. Firstly, HTLV-I used as antigen has to be produced by cell culture using TCL-Kan cell line (Kannagi et al., 1983) or MT-2 cell line (Miyoshi et al., 1981), and has to be purified by density gradient centrifugation (Ikeda et al., 1984) or by ultracentrifugation (Taguchi et al., 1983). This is hazardous and inefficient. Secondly, uncertain test results on pregnant women (Maeda et al., 1989) raise the serious problem of whether neonates may be infected by breast feeding (Ando et al., 1989). Thirdly, the methods may not be sufficiently sensitive as a basis for complete prevention of virus transmission.

This paper describes a novel enzyme immunoassay (immune complex transfer enzyme immunoassay) with high sensitivity, using recombinant gag (14-139)-env(197-295) hybrid protein (Kuga et al., 1988) as antigen for de- tection of anti-HTLV-I IgG in serum. This technique overcomes in part the difficulties described above.

Materials and Methods

Buffers

The buffers used regularly were 20 mmol/l sodium phosphate buffer, pH 7.0, containing 2 mmol/l EDTA (buffer A), 0.1 mol/l sodium phosphate buffer, pH 6.0, containing 5 mmol/l EDTA (buffer B), 10 mmol/l sodium phosphate buffer, pH 7.0, containing 0.1 mol/l NaCl and 1.0 g/l bovine serum albumin (fraction V, Armour Pharmaceutical Co., Kankakee, Illinois, U.S.A.) (buffer C) and 10 mmol/l sodium phosphate buffer, pH 7.5, containing 0.14 mol/l NaCl (buffer D).

79

Antibodies

Rabbit anti-dinitrophenyl bovine serum albumin serum was obtained from ICN ImmunoBiologicals (Lisle, Illinois, U.S.A.). Rabbit anti-human IgG y-chain IgG was obtained from Medical and Biological Laboratories Co., (Nagoya, Japan). IgG was prepared from serum using Na2S04 and diethylaminoethyl cellulose (Ishikawa et al., 1983). F(ab’)2 was prepared by digestion of IgG with pepsin, and Fab’ was prepared by reduction of F(ab’)z (Ishikawa et al., 1983). The amount of IgG and of its fragments was calculated from the absorbance at 280 nm (Ishikawa et al., 1983).

Dinitrophenyl bovine serum albumin

Thiol groups were introduced into bovine serum albumin molecules using S- acetylmercaptosuccinic anhydride and were reacted with maleimide groups intro- duced into dinitrophenyl-L-lysine molecules (Kohno et al., 1988). The amount of bovine serum albumin was calculated from the absorbance at 280 nm by taking the extinction coefficient and molecular weight to be 0.63 g-‘vlcm- (Web- ster, 1970) and 66 200 (Peters, 1975), respectively. The number of dinitrophenyl groups was calculated from the absorbance at 360 nm by taking the molar extinc- tion coefficient to be 17400 mol-‘lcm-’ (Eisen et al., 1954). The amount of dinitrophenyl bovine serum albumin was calculated from the absorbance at 280 nm and 360 nm (Webster, 1970; Peters, 1975; Eisen et al., 1954). The average number of dinitrophenyl groups introduced per bovine serum albumin molecule was 7.4.

Protein-sepharose 4B

Dinitrophenyl bovine serum albumin (10 mg) and human IgG (10 mg) were coupled to CNBr-activated Sepharose 4B (1.0 g, Pharmacia Fine Chemicals AB, Uppsala, Sweden) according to the instructions of Pharmacia.

AfJinity-purification of antibodies

Anti-dinitrophenyl bovine serum albumin IgG and anti-human IgG r-chain IgG were affinity-purified by elution at pH 2.5 from columns of dinitrophenyl bovine serum albumin-Sepharose 4B and human IgG-Sepharose 4B, respectively (Kohno et al., 1988).

Recombinant gag+nv hybrid protein (preparations I and II)

Recombinant gag( 14-l 39)-env( 197-295) hybrid protein, containing 5 cysteine residues with low solubility, was produced in transformed Escherichia coli and was purified by centrifugal precipitation, solubilization with urea and cation exchange chromatography (Kuga et al., 1988). This preparation (preparation I)

80

contained trace amounts of three components of Escherichia cofi. The three components were removed by gel filtration after reduction. Prepa-

ration I (1.5 mg) in 1.8 ml of buffer A containing 7 mol/l urea and 120 mmol/l NaCl was incubated with 0.2 ml of 0.1 mol/l 2-mercaptoethylamine at 50°C for 30 min and was subjected to gel filtration on a column (1.5 x 45 cm) of Toyopearl TSK-HW 55s (Tosoh Corporation, Tokyo, Japan) with a liquid chromatography apparatus (FPLC, Pharmacia Fine Chemicals AB). The buffer for elution was prepared by dissolving 6 mol/l guanidineeHC1 in 20 mmol/l sodium phosphate buffer, pH 7.5, containing 2 mmol/l EDTA and adjusting pH to 7.5. Elution was carried out at a flow rate of 0.5 ml/min under a pressure of 0.2 MPa. Fractions containing recombinant gag-em hybrid protein were pooled, concentrated with an ultrafiltration apparatus (Amicon 8010, Amicon Co., Danvers, Massachusetts, U.S.A.) using a DIAFLOW membrane (PM-10 membrane, Amicon Co.) and dialyzed against buffer A containing 6 mol/l urea. Homogeneity of this prepara- tion (preparation II) was confirmed by polyacrylamide gel electrophoresis in the presence of sodium dodecylsulfate (Laemmli, 1970).

The concentration of recombinant gag-env hybrid protein was determined by a commercial protein assay kit (Bio-Rad Protein Assay Kit, Bio-Rad Laboratories, California, U.S.A.) using bovine serum albumin as standard.

Dinitrophenyl bovine serum albumin-recombinant gag*nv hybrid protein

conjugate

Recombinant gag-env hybrid protein containing 5 cysteine residues per mol- ecule (1 mg, preparation II purified to homogeneity) in 0.75 ml of 20 mmol/l sodium phosphate buffer, pH 6.0, containing 6 mol/l urea and 2 mmol/l EDTA was incubated with 7.5 id of 13 mol/l 2-mercaptoethanol and 7.5 ,~l of 100 g/l sodium dodecylsulfate at 50°C for 30 min and was subjected to gel filtration on a column (1 x 30 cm) of Sephadex G-25 (Pharmacia Fine Chemicals AB) using buffer B containing 1.0 g/l sodium dodecylsulfate. The average number of thiol groups generated per recombinant gag-env hybrid protein molecule was 4.8 (Ishikawa et al., 1983).

Maleimide groups were introduced into dinitrophenyl bovine serum albumin using N-succinimidyl-6-maleimidohexanoate (Dojindo Laboratories, Kumamoto, Japan) (Hashida et al., 1984). The average number of maleimide groups intro- duced per dinitrophenyl bovine serum albumin molecule was 13 (Ishikawa et al., 1983).

Reduced recombinant gag-env hybrid protein (0.8 mg) in 0.32 ml of buffer B containing 1.0 g/l sodium dodecylsulfate was incubated with maleimide- dinitrophenyl bovine serum albumin (2.2 mg) in 0.2 ml of buffer B at 30°C for 60 min. The reaction mixture was subjected to gel filtration on a column (1.5 x 100 cm) of Ultrogel AcA 34 (IBF Biotechnics, Villeneuve-la-Garenne, France) using 0.1 mol/l sodium phosphate buffer, pH 7.0, containing 1.0 g/l sodium dodecylsulfate. The average number of recombinant gag-env hybrid protein molecules conjugated per dinitrophenyl bovine serum albumin molecule was

81

1.7, which was calculated from the concentration of dinitrophenyl bovine serum albumin and the total protein concentration determined by the Bio-Rad protein assay kit described above. The amount of dinitrophenyl bovine serum albumin- recombinant gag-env hybrid protein conjugate was calculated as described for dinitrophenyl bovine serum albumin.

Recombinant gag-env hybrid protein-horseradish peroxidase conjugate

Recombinant gag-env hybrid protein (preparation I) was reduced with 2- mercaptoethanol as described above. Maleimide groups were introduced into horseradish peroxidase (grade I, Boehringer Mannheim GmbH, Mannheim, F.R.G.) using N-succinimidyl-6-maleimidohexanoate (Hashida et al., 1984). The amount of peroxidase was determined from the absorbance at 403 nm (Ishikawa et al., 1983). The average number of maleimide groups introduced per peroxidase molecule was 1.2 (Ishikawa et al., 1983).

Reduced recombinant gag-env hybrid protein (0.9 mg) in 0.45 ml of buffer B containing 1.0 g/l sodium dodecylsulfate was incubated with maleimide- peroxidase (1.4 mg) in 0.13 ml of buffer B at 30°C for 60 min. The reaction mixture was subjected to gel filtration on a column (1.5 x 45 cm) of Ultro- gel AcA 34 using 0.1 mol/l sodium phosphate buffer, pH 7.0, containing 1.0 g/l sodium dodecylsulfate. The average number of recombinant gag-env hybrid protein molecules conjugated per peroxidase molecule was 1.0, which was cal- culated from the concentration of peroxidase and the total protein concentration determined by the Bio-Rad protein assay kit described, above. The amount of recombinant gag-env hybrid protein-peroxidase conjugate was calculated from peroxidase activity (Ishikawa et al., 1983).

Protein-coated polystyrene balls

Polystyrene balls (3.2 mm in diameter, Immunochemical Co., Okayama, Japan) were coated by physical adsorption (Kohno et al., 1988) with affinity-purified anti-dinitrophenyl bovine serum albumin IgG (0.1 g/l), affinity-purified anti- human IgG y-chain IgG (0.1 g/I) in 0.1 mol/l sodium phosphate buffer, pH 7.5, containing 1 g/l NaN3, and recombinant gag-env hybrid protein (0.1 s/i, Preparation I) in 0.1 mol/l sodium phosphate buffer, pH 7.5, containing 7 mol/l urea and 5 mmol/l EDTA.

Anti-human IgG y-chain Fab’-peroxidase conjugate

Rabbit anti-human IgG y-chain Fab was conjugated to horseradish peroxidase using N-succinimidyl-6-maleimidohexanoate (Hashida et al., 1984). The amount of the conjugate was calculated from peroxidase activity (Ishikawa et al., 1983).

82

Immune complex transfer enzyme immunoassay for anti-HTLV-I IgG

The test serum (20 ~1) was incubated simultaneously with both dinitrophenyl bovine serum albumin-recombinant gag-env hybrid protein conjugate (100 fmol) and recombinant gag-env hybrid protein-peroxidase conjugate (100 fmol) in 0.13 ml of 10 mmol/l sodium phosphate buffer, pH 7.0, containing 0.46 mol/l NaCl and 1.0 g/l bovine serum albumin at 20°C for 3 h. Subsequently, the reaction mixture was incubated with two affinity-purified anti-dinitrophenyl bovine serum albumin IgG-coated polystyrene balls at 20°C for 3 h and at 4OC overnight. The polystyrene balls were washed twice with 2 ml of 10 mmol/l sodium phosphate buffer, pH 7.0, containing 0.1 mol/l NaCl and incubated with 0.15 ml of 1 mmol/l dinitrophenyl-L-lysine in buffer C at 20°C for 1 h. After removal of the polystyrene balls, the eluate was incubated with two affinity-purified anti-human IgG y-chain IgG-coated polystyrene balls at 20°C for 3 h. The polystyrene balls were washed as described above, and bound peroxidase activity was assayed at 30°C for 150 min by fluorometry using 3-(4-hydroxyphenyl)propionic acid as hydrogen donor (Imagawa et al., 1983). The fluorescence intensity was measured relative to 0.2 mg/l quinine in 50 mmol/l HzS04.

Conventional enzyme immunoassay for anti-HTLV-I IgG

Test serum (20 ~1) was mixed with 130 ~1 of 10 mmol/l sodium phosphate buffer, pH 7.0, containing 0.46 mol/l NaCl, 1.0 g/l bovine serum albumin and 1.0 g/l NaN3 and was incubated with a recombinant gag-env hybrid protein-coated polystyrene ball at 37’C for 3 h. The polystyrene ball was washed twice with 2 ml of 10 mmol/l sodium acetate buffer, pH 5.0, containing 0.5 g/l Tween-20 (Nacalai Tesque, Inc., Kyoto, Japan) and was incubated with 50 ng of anti-human IgG y-chain Fab’-peroxidase conjugate in 0.15 ml of buffer C at 37°C for 3 h. The polystyrene ball was washed as described above, and bound peroxidase activity was assayed at 30°C for 3 min as described above. The fluorescence intensity was measured relative to 1.0 mg/l quinine in 50 mmol/l H2SO4.

Expression of the detection limit of anti-HTLV-I IgG in serum

The detection limit of anti-HTLV-I IgG in serum by enzyme immunoassays was expressed as the maximal dilution of serum containing anti-HTLV-I IgG with pooled normal serum which gave a bound peroxidase activity significantly in excess of that in the presence of normal serum (background). A signifi- cant difference from the background was confirmed by the t-test (n=5, P < 0.001).

Western blotting

Antigens (preparation I of recombinant gag-env hybrid protein and the three components from Escherichia coli) were subjected to polyacrylamide gel elec-

83

trophoresis in the presence of sodium dodecylsulfate (Laemmli, 1970) and were transferred electrophoretically to a nitrocellulose sheet (GVHP, Millipore Co., Bedford, Massachusetts, U.S.A.) at 200 mA for 3 h using 25 mmol/l Tris, 192 mmol/l glycine and 20% (v/v) methanol, pH 8.3 as electrode buffer (Bumette, 1981). The three components from Escherichiu coli were prepared in the same way as Preparation I of recombinant gag-env hybrid protein. Strips of nitrocel- lulose were incubated at room temperature with buffer D containing 50 g/l skim milk (Morinaga Milk Industries Co. Ltd., Tokyo, Japan) for 3 h and subsequently with test serum (30 ~1) in 3 ml of the same buffer for 3 h. After washing with buffer D containing 0.5 g/l Tween-20, the nitrocellulose strips were incubated with 1.0 pg of anti-human IgG y-chain Fab’peroxidase conjugate in 3 ml of buffer D containing 1.0 g/l bovine serum albumin (Oriental Yeast Co., Ltd., Tokyo, Japan) at room temperature for 3 h, washed as described above and dyed with 1.2 g/l 4-chloro-1-naphthol in 10 mmol/l Tris-HCl buffer, pH 7.4, containing 17% (v/v) methanol and 50 mg/l H202.

Western blotting using HTLV-I produced by TCL-Kan cell line was generously performed by Fujirebio Inc., Tokyo (Japan).

Gelatin particle agglutination for anti-HTLV-I antibodies

Measurement of anti-HTLV-I antibodies by gelatin particle agglutination was carried out using a commercial kit with HTLV-I produced by TCL-Kan cell line as antigen (Serodia-Atla, Fujirebio Inc., Tokyo, Japan) (Ikeda et al., 1984). The test serum was diluted &fold or more with the diluent included in the kit. The diluted serum (25 ,ul) was mixed with a particle suspension (25 ~1) in ‘U’ shaped wells of microplates and allowed to stand at room temperature for 3 h.

Results and Discussion

An immune complex transfer enzyme immunoassay for anti-HTLV-I IgG was developed using recombinant gag(14_139)-env( 197-295) hybrid protein (Kuga et al., 1988). Anti-HTLV-I IgG in serum was reacted simultaneously with dinitrophenyl bovine serum albumin-recombinant gag-env hybrid protein conjugate and recombinant gag-em hybrid protein-peroxidase conjugate. The complex formed of anti-HTLV-I IgG and the two conjugates was trapped onto two anti-dinitrophenyl group IgG-coated polystyrene balls. After washing the polystyrene balls to eliminate nonspecific IgG and excess of the peroxidase conjugate, the complex was eluted from the polystyrene balls with excess of dinitrophenyl-L-lysine and was transferred to two anti-human IgG y-chain IgG- coated polystyrene balls. Peroxidase activity bound to the polystyrene balls was assayed by fluorometry. Nonspecific binding of peroxidase activity was remarkably reduced, by transfer of the complex, and the sensitivity was improved to a great extent.

84

TABLE 1 No significant correlation between the fluorescence intensity obtained by the immune complex transfer enzyme immunoassay and the level of antibody IgG against three components (I, II and III) from Escherichia coli which were present in preparation I of recombinant gag-env hybrid protein

Serum No. Level of antibody IgG against Escherichia coli components shown by Western blotting

I II III

Fluorescence intensity for bound peroxidase activity

1 + ++ - 2.0 2 + ++ _ 2.3 3 + ++ - 4.0 4 ++ + - 3.9 5 + + ++ 2.4 6 - + + 3.1 7 - ++ + 2.3

The molecular weights of the three components I, II and III from Escherichiu coli were 33 000, 43 000 and 50 000, respectively.

Specificity of the immune complex transfer enzyme immunoassay

Recombinant gag-env hybrid protein preparation purified to homogeneity (Preparation II) was conjugated to dinitrophenyl bovine serum albumin, although recombinant gag-env hybrid protein preparation containing trace amounts of three components from Escherichia co/i (Preparation I) was conjugated to horse- radish peroxidase. Therefore, antibody IgG against components of Escherichia co/i might not have been detected by the immune complex transfer enzyme im- munoassay.

In addition, seven normal human sera, which were shown by Western blot- ting to contain IgG antibody against the three components of Escherichia coli at different levels, were subjected to the immune complex transfer enzyme im- munoassay. No significant correlation was observed between bound peroxidase activity and antibody IgG level shown by Western blotting. This indicated that antibody IgG against the components of Escherichia coli was not detected sig- nificantly by immune complex transfer enzyme immunoassay (Table 1).

Detection limit of anti-HTLV-I IgG in human serum

The detection limit of anti-HTLV-I IgG in human serum by the immune com- plex transfer enzyme immunoassay was compared with that by other methods, including the conventional enzyme immunoassay using recombinant gag-env hy- brid protein, Western blotting using recombinant gag-env hybrid protein, West- em blotting using HTLV-I produced by TCL-Kan cell line and gelatin particle agglutination using HTLV-I produced by TCL-Kan cell line. Four human sera containing anti-HTLV-I IgG from HTLV-I carriers were serially diluted with normal serum and were tested. For gelatin particle agglutination, the four sera serially diluted with normal serum were further diluted 8-fold with the diluent

85

Western blotting (gag-env) l + l + : + + + - + + + - : * *

1

t

Western blotting + + l + (HTLV-I) l + + +

1 - t l

- - T + Gelatin particle + + + +

- + + +

a, 105 104 103 102 10

Dilution of Serum Containing Anti-HTLV-I IgG

with Normal Serum (-fold)

Fig. 1. Dilution curves of sera containing anti-HTLV-I IgG by the immune complex transfer enzyme immunoassay (open symbols) and the conventional enzyme immunoassay (closed symbols). Four sera containing anti-HTLV-I IgG from HTLV-I carriers were diluted serially with normal serum and subjected to both immunoassays. Results by Western blotting using HTLV-I produced by TCL-Kan cell line, by Western blotting using recombinant gag-env hybrid protein and by a commercial kit based on gelatin particle agglutination using HTLV-I produced by TCL-Kan cell line were also shown. For gelatin particle agglutination, sera containing anti-HTLV-I antibodies were diluted serially with normal serum

and further diluted 8-fold with the diluent included in the kit.

included in the kit. The detection limit by the immune complex transfer enzyme immunoassay was 300 to 3000-fold lower than that by other methods (Fig. 1).

Assay variation of the immune complex transfer enzyme immunoassay

The assay variation in the immune complex transfer enzyme immunoassay was examined using serum samples which showed the fluorescence intensity for bound peroxidase activity at three different levels (9.1, 58 and 304 for within- assay and 40, 147 and 641 for between-assay). The coefficients of variation for within-assay and between-assay were 4.9-9.0% (n=15) and 8.7-9.5% (n=lO), respectively.

TA

BL

E 2

C

ompa

riso

n of

te

st

resu

lts

for

271

seru

m

sam

ples

by

va

riou

s m

etho

ds

Gro

up

No.

of

ser

um

sam

ples

w

ith

the

num

ber

of s

ampl

es

in p

aren

thes

es

Gro

up

1

(103

) G

roup

2

Gro

up

3

(168

) (3

3)

Gro

up

4 G

roup

5”

(135

) (2

1)

Gro

up

6b

(114

)

Gel

atin

pa

rtic

le

Wes

tern

W

este

rn

Fluo

resc

ence

in

tens

ity

for

boun

d ag

glut

inat

ion

blot

ting

1 w

ith

blot

ting

II

pero

xida

se

activ

ity

by:

with

H

TL

V-I

H

TL

V-I

w

ith

gag-

env

Imm

une

com

plex

C

onve

ntio

nal

tran

sfer

en

zym

e en

zym

e im

mun

oass

ay

imm

unoa

ssay

(-)

NT

N

T

0.5-

8.4

4.0-

65

(+)

(-)

(-)

0.4-

7.5

5.91

21

Gro

up

7 (+

) (-

) (-

) 15

2, 2

89

77,

59

(2)

Gro

up

8 (+

) (+

) (-

) 55

2,

1,28

0,

3,32

0 87

, 56

, 61

(3)

Gro

up

9 (+

) (+

) (+

) 37

@4,

300

18-1

56

(16)

(+

) N

T

NT

5J

3-57

,300

37

-237

aTw

enty

-one

sa

mpl

es,

for

whi

ch

the

max

imal

di

lutio

n w

ith

the

dilu

ent

incl

uded

in

the

kit

to

caus

e ge

latin

pa

rtic

le

aggl

utin

atio

n w

as

8 to

64-

fold

. w

ere

test

ed

by W

este

rn

blot

ting.

hO

ne h

undr

ed

four

teen

sa

mpl

es,

for

whi

ch

the

max

imal

di

lutio

n of

sam

ples

w

ith t

he d

iluen

t in

clud

ed

in t

he k

it t

o ca

use

gela

tin

part

icle

ag

glut

inat

ion

was

12

8 to

16

,384

-fol

d,

wer

e no

t te

sted

by

W

este

rn

blot

ting.

N

T,

not

test

ed.

87

0.1-1 "' "".'_ c-s-‘-d 1 ‘a,,*..1 * ~~~.LLIJ Negative 10 100 1000 lOODo IOODDO

Maximal Dilution of Serum with Uuffer to Cause Gelatin Particle Agglutination (-fold)

Fig. 2. Comparison of test results by the immune complex transfer enzyme immunoassay, gelatin particle agglutination and Western blotting. Open and closed squares indicate negative and positive samples, respectively, by Western blotting using HTLV-I produced by TCL-Kan cell line. Open triangles indicate

samples which were not tested by Western bloning.

Detection of anti-HTLV-I IgG in serum samples

Two hundred and seventy-one serum samples were tested by the immune complex transfer enzyme immunoassay and other methods (Table 2 and Fig. 2). The 271 samples were divided into two groups by gelatin particle agglutination using HTLV-I produced by TCL-Kan cell line as antigen: 103 negative samples (Group 1) and 168 positive samples (Group 2). When Group 1 was subjected to the immune complex transfer enzyme immunoassay, fluorescence intensities for bound peroxidase activity ranged from 0.5 to 8.4. Group 2 was divided into two groups by the immune complex transfer enzyme immunoassay. One group of 33 samples (Group 3) showed fluorescence intensities between 0.4 and 7.5, and was shown to be negative by Western blotting using HTLV-I produced by TCL-Kan cell line (Western blotting I) and by Western blotting using recombinant gag-env hybrid protein (Western blotting II). It was very likely that test results of Group 3 by gelatin particle agglutination were false-positive. This is consistent with a previous report that the gelatin particle agglutination kit used in this study often gives false-positive results (Maeda et al., 1989). The remaining 135 samples (Group 4) that were positi.ve by gelatin particle agglutination showed

88

fluorescence intensities between 152 and 57 300. The lowest value of fluorescence intensities obtained with Group 4 was 18-fold higher than the highest value of fluorescence intensities obtained with Groups 1 and 2 which were negative by the immune complex transfer enzyme immunoassay and gelatin particle agglutination or Western blotting. Group 4 was divided into two groups: Group 5 (21 samples), for which the maximal dilution of samples in the diluent included in the kit to cause gelatin particle agglutination was 8 to 64-fold, and Group 6, for which the maximal dilution was 128 to 16 384-fold. Group 5 was divided into three groups (Groups 7-9) by Western blottings I and II. Two samples (Group 7) in Group 5 were negative by both Western blottings, three samples (Group 8) were positive by Western blotting I but negative by Western blotting II, and 16 samples (Group 9) were positive by both Western blottings. Fluorescence intensities obtained with the two samples (Group 7) that were negative by both Western blottings, were 152 and 289. These were the lowest values among those for the 135 samples (Group 4) that were positive by the immune complex transfer enzyme immunoassay. These two samples (Group 7) might have contained anti-HTLV-I IgG at low concentrations which were detectable by the immune complex transfer enzyme immunoassay but not by Western blottings I and II.

Fluorescence Intensity for Bound Peroxidase Activity by the Conventional Enzyme Imnunoassay

Fig. 3. Comparison of the immune complex transfer enzyme immunoassay and the conventional enzyme immunoassay. Open and closed squares indicate negative and positive samples, respectively, by Western blotting using HTLV-I produced by TCL-Kan cell line. Open triangles indicate samples which were not

tested by Western blotting.

89

A A

160 -

Maximal Dilution of Serum with Buffer to Cause Gelatln Particle Agglutination (-fold)

Fig. 4. Comparison of test results by the conventional enzyme immunoassay, gelatin particle agglutina- tion and Western blotting. Open and closed squares indicate negative and positive samples, respectively. by Western blotting using HTLV-I produced by TCL-Kan cell line. Open triangles indicate samples

which were not tested by Western blotting.

The 271 samples were tested by the conventional enzyme immunoassay using recombinant gag-env hybrid protein, and the results were compared with those by other methods including the immune complex transfer enzyme immunoassay, Western blottings I and II and gelatin particle agglutination (Table 2, and Figs. 3 and 4). Positive and negative samples were not clearly discriminated, and the test results by the conventional enzyme immunoassay were not consistent with those by any of other methods. Moreover, some samples that were positive by Western blottings I and II, were apparently negative or not clearly positive by the conventional enzyme immunoassay.

In summary, the immune complex transfer enzyme immunoassay detected anti- HTLV-I IgG in serum at levels below those detectable by the conventional enzyme immunoassay and Western blotting and discriminated negative and positive samples more clearly than the conventional enzyme immunoassay. However, anti-HTLV-I IgG must be concentrated by affinity chromatography from serum samples that are negative by Western blotting but positive by the immune complex transfer enzyme immunoassay and be investigated by Western blotting as described above. Also a greater number of samples must be examined

90

to see whether the immune complex transfer enzyme immunoassay detects anti-HTLV-I IgG in serum at levels below those detectable by gelatin particle agglutination.

In preliminary experiments, the detection limit of anti-HTLV-I IgG in serum by the immune complex transfer enzyme immunoassay was examined with shorter incubations. The test serum was incubated simultaneously with the two conjugates and anti-dinitrophenyl,group IgG-coated polystyrene balls for 2 h. The polystyrene balls were washed and incubated simultaneously with dinitrophenyl- L-lysine and anti-human IgG y-chain IgG-coated polystyrene balls for 3 h. The detection limit of anti-HTLV-I IgG in serum was still 100 to 300-fold lower than that by the conventional enzyme immunoassay.

Acknowledgements

We are grateful to Dr Seiga Itoh, Tokyo Research Laboratories, Kyowa Hakko Co., Ltd., Tokyo for production and partial purification of recombinant gag-env hybrid protein, to Mr Hideharu Mori, Fuji Research Laboratories, Kyowa Medex Co., Ltd., Shizuoka for test by Western blotting using recombinant gag-env hybrid protein and to Fujirebio Inc., Tokyo for test by Western blotting using HTLV-I produced by TCL-Kan cell line.

References

Ando, Y., Saito, K., Nakano, S., Kamimoto, K., Furuki, K., Tanigawa, T., Hashimoto, H., Moriyama, I., Ichijo, M. and Toyama, T. (1989) Bottle-feeding can prevent transmission of HTLV-1 from mothers to their babies. J. Infect. 19, 25-29.

Bumette, W.N. (1981) ‘Western blotting’: electrophoretic transfer of proteins from sodium dodecyl sulfate-polyacrylamide gels to unmodified nitrocellulose and. radiographic detection with antibody and radioiodinated protein A. Anal. Biochem. 112, 195-203.

Eisen, H.N., Carsten, M.E. and Belman, S. (1954) Studies of hypersensitivity to low molecular weight substances. III. The 2,4-dinitrophenyl group as a determinant in the precipitin reaction. .I. Immunol. 73, 296-308.

Gallo, R.C., Kalyanaraman, V.S., Sarngadharan, M.G., Sliski, A., Yonderheid, E.C., Maeda, M., Nakao, Y., Yamada, K., Ito, Y., Gutensohn, N., Murphy, S., Bunn, Jr. P.A., Catovsky, D., Greaves, M.F., Blayney, D.W., Blattner, W., Jarrett, W.F.H., zur Hausen, H., Seligmann, M., Brouet, J.C., Haynes, B.F., Jegasothy, B.V., Jaffe, E., Cossman, J., Broder, S., Fisher, RI., Golde, D.W. and Robert- Guroff, M. (1983) Association of the human type C retrovirus with a subset of adult T-cell cancers. Cancer Res. 43, 3892-3899.

Hashida, S., Imagawa, M., Inoue, S., Ruan, K-h. and Ishikawa, E. (1984) More useful maleimide compounds for the conjugation of Fab’ to horseradish peroxidase through thiol groups in the hinge. J. Appl. B&hem. 6, 56-63.

Hinuma, Y., Nagata, K., Hanaoka, M., Nakai, M., Matsumoto, T., Kinoshita, K., Shirakawa, S. and Miyoshi, I. (1981) Adult T-cell leukemia: antigen in an ATL cell line and detection of antibodies to the antigen in human sera. Proc. Natl. Acad. Sci. U.S.A. 78, 647-80.

Ikeda, M., Fujino, R., Matsui, T., Yoshida, T., Komoda, H. and Imai, J. (1984) A new agglutination test for serum antibodies to adult T-cell leukemia virus. Gann 75, 845-848.

Imagawa, M., Hashida, S., Ishikawa, E., Mori, H., Nakai, C., Ichioka, Y. and Nakajima, K. (1983)

91

A highly sensitive sandwich enzyme immunoassay for insulin in human serum developed using capybara anti-insulin Fab’-horseradish peroxidase conjugate. Anal. Lett. 16, 1509-1523.

Ishikawa, E., Imagawa, M., Hashida, S., Yoshitake, S., Hamaguchi, Y. and Ueno, T. (1983) Enzyme- labeling of antibodies and their fragments for enzyme immunoassay and immunohistochemical staining. J. Immunoassay 4, 209-327.

Kannagi, M., Sugamura, K., Sato, H., Okochi, K., Uchino, H. and Hinuma, Y. (1983) Establishment of human cytotoxic T-cell lines specific for human adult T-cell leukemia virus-bearing cells. J. Immunol. 130, 2942-2946.

Kohno, T., Mitsukawa, T., Matsukura, S. and Ishikawa, E. (1988) Novel enzyme immunoassay (immune complex transfer enzyme immunoassay) for anti-thyroglobulin IgG in human serum. J. Clin. Lab. Anal. 2, 209-214.

Kuga, T., Yamasaki, M., Sekine, S., Fukui, M., Yokoo, Y., Itoh, S., Yoshida, M., Hattori, T. and Takatuki, K. (1988) A gag-env hybrid protein of human T-cell leukemia virus type I and its application to serum diagnosis. Jpn. J. Cancer Res. 79, 1168-1173.

Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685.

Maeda, Y., Imai, J., Kiyokawa, H., Kanamura, M. and Hino, S. (1989) Performance certification of gelatin particle agglutination assay for anti-HTLV-1 antibody: inconclusive positive results. Jpn. J. Cancer Res. 80, 915-919.

Miyoshi, I., Kubonishi, I., Yoshimoto, S., Akagi, T., Ohtsuki, Y., Shiraishi, Y., Nagata, K. and Hinuma, Y. (1981) Type C virus particles in a cord T-cell line derived by co-cultivating normal human cord leukocytes and human leukaemic T cells. Nature 294, 770-77 1.

Okochi, K., Sato, H. and Hinuma, Y. (1984) A retrospective study on transmission of adult T cell leukemia virus by blood transfusion: Seroconversion in recipients. VOX Sang. 46, 245-253.

Peters, Jr. T. (1975) Serum albumin. In: F.W. Putnam (Ed.), The Plasma Proteins, Structure, Function, and Genetic Control, pp. 133-181. Academic Press, New York.

Taguchi, H., Sawada, T., Fujishita, M., Morimoto, T., Niiya, K. and Miyoshi, I. (1983) Enzyme-linked immunosorbent assay of antibodies to adult T-cell leukemia-associated antigens. Gann 74, 185-187.

Webster, G.C. (1970) Comparison of direct spectrophotometric methods for the measurement of protein concentration. B&him. Biophys. Acta 207, 371-373.