Journal of Virological Methods, 10 (1985) 215-224

Elsevier

JVM 00375

215

ENZYME-ANTIBODY CONJUGATION BY A HETEROBIFUNCTIONAL

REAGENT AND ITS APPLICATION IN ENZYME-LINKED

IMMUNOSORBENT ASSAY (ELISA) FOR THE DETECTION OF

HEPATITIS B SURFACE ANTIGEN

DEEPAK A. GADKARI’. HOWARD A. FIELDS? and JAMES E. MAYNARD2

‘National .Instirute qf Virology, Post Office Box 1 I, Pune 41IOO1, India: *Division of Viral Diseases,* Center

for Ir$ectkms Diseases. Centers for Disease Control, Public Health Service. U.S. Department qf Health and

Human Services. 1600 Clifton Road, N.E.. Atlanta, GA 85014. U.S.A.

(Accepted 29 October 1984)

The heterobifunctional reagent 3-(2.pyridyldithio)propionate (SPDP) was used to prepare defined

conjugates composed of horseradish peroxidase (HRP) and goat anti-HBs IgG. The modification of HRP

and IgG ,wtth SPDP was dependent on both the SPDP: protein molar ratio and the pH of the buffer.

Conjugates were separated by a single affinity chromatographic step using concanavalin A-Sepharose 48

equilibrated with 0.1 M Tris-HCI buffer, pH 7.4, containing 0.5 M KCI and I mM EDTA. The conjugate

was eluted with 10 or 100 mM a-methyl-o-mannoside and appropriate pools were made reflecting various

HRP/IgCi molar ratios. Each pool was examined for performance in an enzyme-linked immunosorbent

assay for HBsAg. Conjugate composed of an HRP/lgG molar ratioof 2.5-4 yielded the greatest sensitivity.

SPDP conjugation conjugate purification ELISA HBsAp

INTRODUCTION

Since its introduction in 1971 by Engvall and Pearlman (1971) enzyme-linked

immunosorbent assay (ELISA) has been widely used for the measurement of bac-

terial, viral, parasitic and fungal antigens, as well as immunoglobulins, hormones,

drugs, serum proteins and tumor antigens (Yolken, 1982). Advances in substrate

selection, solid-phase adsorbents and instrumentation have resulted in ELISA achiev-

ing similar sensitivities as radioimmunoassay without the concomitant disadvan-

tages associated with the use of radioactive reagents. In 1976, Wolters et al. introduced

ELISA for the detection of hepatitis B surface antigen. Since this initial report, there

have been many publications on this subject (Caldwell and Barrett, 1977; Halbert and

Anken, 1977: Wei et al., 1977).

* World Health Organization Collaborating Centre for Reference and Research on Viral Hepatitis

0166-0934/85/$03.30 0 1985 Elsevier Science Publishers B.V. (Biomedical Division)

216

Various methods for conjugating IgG to HRP using either glutaraldehyde (Avra-

meus, 1969; Avrameus and Ternynck, 1971) or meta-periodate (Wilson and Nakane,

1978) have been utilized in developing ELISA. These methods have several drawbacks

including the production of large molecular weight conjugates, homopolymers, and

low enzyme yield. A heterobifunctional reagent, N-succinimidyl 3-(2-pyridyldithio)

propionate (SPDP), synthesized by Carlsson et al. (1978) has been successfully used

by Nilsson et al. (1981) to prepare defined conjugates of IgG and HRP without the

drawbacks associated with the other conjugation methods. This paper extends Nils-

son’s conjugation results and presents a novel one-step procedure for separating

conjugates into defined HRP : IgG molar ratios.

MATERIALS AND METHODS

Reagents

SPDP was purchased from Pharmacia Fine Chemicals (Piscataway, NJ). Stock

solutions were prepared in absolute ethanol and the concentrations were determined

by the method recommended by the manufacturer. SPDP solutions were aliquoted

and stored at -70°C until used. All other reagents were purchased or prepared as

previously described (Fields et al., 1981, 1983).

SPDP thiolation

HRP and IgG were thiolated at different SPDP: protein molar ratios. The degree of

substitution (DOS) (number of thiol groups per mole of protein) was determined by

the method of Grassetti and Murray (1967). Briefly, after reacting HRP or IgG with

SPDP at different molar ratios for 30 min, the modified protein was desalted on a I .6

X 30 cm column of Sephadex G-25 (Pharmacia Fine Chemicals) equilibrated with 0.1

M sodium acetate buffer, pH 4.6, containing 0.1 M NaCl (buffer A) to remove

N-hydroxysuccinimide released during the reaction. One molar dithiothreitol (DTT)

was added to the modified protein preparation to a final DTT concentration of 50

mM. After incubation for 30 min at room temperature, the reaction mixture was

desalted on a second Sephadex G-25 column equilibrated with 0.1 M Tris-HCl buffer,

pH 7.4, containing 0.5 M KC1 and 1 mM EDTA (buffer B). The eluted protein was

concentrated to a volume of 4 ml by ultrafiltration by means of a 25 mm diameter

YMlO ultrafilter and an 8 MC stirred cell (Amicon, Lexington, MA). Equal volumes

of thiolated protein and 1.5 mM 2,2’-dipyridyl disulfide (2-PDS) were added to a

cuvette and the concentration of liberated 2-thiopyridone (2-TP) was determined

spectrophotometrically at a wavelength of 343 nm after subtracting absorbance due to

thiolated protein alone and 2-PDS alone. One mole of liberated 2-TP is equivalent to 1

thiol group.

Conjugate preparation

SPDP (15-30 mM) was added dropwise to 15-20 mg of goat (gt) anti-HBs IgG

217

prepared in 0.1 M phosphate buffer, pH 7.2, containing 0.1 M NaCl to yield a molar

ratio (SPDP : IgG) of 54 : 1. The modified protein was desalted on G-25 equilibrated

with buffer B, concentrated by ultrafiltration and stored at 4°C. HRP was modified

with SPDP at a molar ratio of 50 : 1. The modified protein was desalted on Sephadex

G-25 equilibrated with buffer A, treated with 50 mM DTT, desalted on Sephadex

G-25 equilibrated with buffer B, concentrated to 4 ml by ultrafiltration, and used

immediately for the conjugation reaction. The modified IgG containing 2-pyridyl

disulphide residues and the thiolated HRP were mixed together and stirred gently for

16 h at 2.3”C. A 1.5-5 molar excess of thiolated HRP was used to favor the thiol-disul-

phide bond exchange reaction.

Separation of conjugate

A 0.9 X 16 cm column of concanavalin A-Sepharose 4B (Con A-Seph) (Sigma

Chemical Co., St. Louis, MO) equilibrated with 10 bed volumes of buffer B was used

to separate the conjugate. The conjugation mixture was chromatographed at a flow

rate of 5 ml/h. Unadsorbed protein was eluted with buffer B until the absorbance at

280 nm (A,,,) returned to baseline and then the conjugate was eluted with 10 and 100

mM a-methyl-o-mannoside (AMDM) prepared in buffer B. Fractions containing the

conjugate were selectively pooled, lyophilized, and stored at -20°C. The conjugate has

remained stable for more than 2 yr.

HRP: IgG molar ratio

The molar ratio of HRP : IgG was determined spectrophotometrically by measuring

the AZ80 and A,,,,. The extinction coefficients (Ey:z ) andmolecular WeightsforHRPand

IgG used were 1.95 and 4.0 X lo5 and 1.35 and 1.6 X 105, respectively. A standard

curve is presented in Fig. 1.

Enzyme-linked immunosorbent assay (ELBA)

ELISA was performed according to the method described by Fields et al. (1983).

Briefly, 0.125 ml of guinea pig anti-HBs IgG appropriately diluted in 0.05 M carbo-

nate buffer, pH 9.6, was added to flat bottom microtiter strips (Immulon I or II,

Dynatech, Alexandria, VA) for 16 h at 23°C. The wells were then washed 5 times with

phosphate-buffered saline (PBS), pH 7.2, containing 0.5% (w/v) bovine serum albu-

min (BSA), and 0.1 ml of a standard dilution of HBsAg-positive chimpanzee plasma

was added. After incubation at 45°C for 3 h the wells were washed 7 times and 0.1 ml of

various dilutions of conjugate were added to each well. After 2 h of incubation at 37°C

the wells were washed 10 times and 0.2 ml of substrate (H,O,) and chromophore

(o-phenylenediamine) prepared in phosphate-citrate buffer, pH 5.0, were added in the

dark. After 40 min at 23°C the reaction was stopped by the addition of 0.1 ml of 4 N

H,SO,. Absorbance was measured by reading the A,,, in each well using an automatic

Dynatech reader (MR-580). Results were scored by comparing positive to negative

(P/N) quotients. It has been previously demonstrated that the higher the P/N ratio,

the greater the sensitivity (Fields et al. 1983).

218

20-

5

I 5-

8 \lO- 8

ob O 5-

2 4 6 8 IO 12 14 16 I8 HRP/lgG molar ratio

0

Fig. I. Spectrophotometric determlnatlon 01. HRP/lpG molar ratios. The calculation was performed as

follows:

1) HRP protein concentration (me/ml) = A,,,,/E’\j’jl

2) IgG protein concentration (mg/ml) = (A 2R,,-A,,,,/RZ value for HRP)/E”,i’,,

3) HRPAgG molar ratio = (HRP protein concentration/molecular weight)/(IgG protein concentra-

tion/molecular *eight)

Pro rein assay

Protein concentration was determined by the Bio-Rad protein assay (Richmond,

CA) using protein standard I for IgG and a prepared HRP standard for HRP.

RESULTS

Degree of substitution

Fig. 2 demonstrates the degree of substitution (number of thiol groups/mole of

protein) as a function of the SPDP : IgG molar ratio. At a molar ratio of 2 : 1, a degree

of substitution (DOS) of 1.8 was obtained which increased linearly up to a DOS of 3.5

at a molar ratio of 18 : 1. The DOS could not, however, be increased further even at a

molar ratio of 54 : 1.

In an attempt to increase the DOS, we examined the SPDP thiolation reaction as a

function of pH (Fig. 3). The goat anti-HBs IgG was dialyzed against the appropriate

buffer before being reacted with SPDP at a molar ratio of 54 : 1, The buffers used were

as follows: (1) 0.1 M phosphate buffer, pH 7.2, (2) 0.1 M phosphate buffer, pH 8.0, and

(3) 0.1 M carbonate buffer, pH 9.0. All buffers contained 0.1 M NaCl. As the pH was

increased from 7.2 to 9.0 the DOS decreased from 7.5 to 1.5.

HRP was considerably more refractory to thiolation than IgG (Table 1). At a molar

ratio of 15 : I, a DOS of 0.8 1 was obtained. By increasing the molar ratio by a factor of

5, only a 2-fold increase of DOS was observed. For the preparation of conjugates a

SPDP: HRP molar ratio of 50: 1 was used which resulted in a DOS of close to 1.

219

5 IO 15 20 25 30 35 40 45 50 !

SPDP/ IgG molar ratlo 7.0 80

PH

9.0

Fig. 2. Degree of substitution of goat anti-HBs IgG as a function ofSPDP: I@ molar ratios obtained at pH

8.0. All values are averages of at least 3 separate experiments.

Fig. 3. Degree ofsubstitution ofgoat anti-HBs IgG asa function ofpH at a constant SPDP: IgG molar ratio

of 54: 1.

Separation of conjugate

SPDP-modified IgG was mixed with the modified and reduced HRP as described in

Materials and Methods. After the incubation period the conjugate was fractionated by

lectin affinity chromatography using Con A-Seph 4B (Fig. 4). The unbound fractions

(fractions 7-49) contained little anti-HBs-HRP conjugate activity based upon low

reactivity in the ELISA for HBsAg. This peak was subsequently rechromatographed

on a 2.6 X 87 cm gel filtration column of Bio-Gel A-O.5 m agarose equilibrated with

PBS and previously calibrated with HRP and iodinated IgG. This experiment de-

monstrated (data not shown) that the unbound fraction from the Con A-Seph 4B

chromatography was composed of free IgG and free HRP as indicated by their

appearances corresponding to their expected elution volumes. Conjugate was eluted

with 10 mM AMDM from the Con A-Seph 4B column when the absorbance returned

to baseline. Fractions 65-75 contained considerable conjugate activity based upon its

TABLE I

Degree of substitution in HRP by SPDP

SPDP/HRP Degree of

molar ratio Substitution

15: 1 0.81

25: I 0.80

50: 1 1.18

75: 1 1.65

All values are average of 2 or more experiments.

220

0 4 a I2 16 20 24 28 32 36 40 44 40 52 56 60 64 68 72 76 80 84 00 92

FRACTION NUMBER

Fig. 4. Separation of SPDP-prepared conjugates on concanavalin A-Sepharose 4B. The column (0.9 X 16

cm) was equilibrated with 0.1 M Tr~s-HCI buffer. pH 7.4, containing 0.5 M KCI and I mM EDTA (buffer

B). The flow rate was 5 ml/h and 2 ml size fractions were collected.

performance in an ELISA. When the AMDM concentration was increased to 100

mM, a second absorbance peak was observed between fractions 85 and 92. This peak

also displayed conjugate activity by ELISA. No more conjugate could be eluted

despite the presence of a brownish colored material characteristic of HRP adsorbed to

the top of the column.

ELISA

The assay was performed using conjugates obtained by pooling various fractions

eluted from the Con A-Seph 4B column. The HRP/IgG ratio was determined

according to the standard curve (Fig. 1). Table 2 lists the pooled fractions comprising

the various conjugate preparations and their corresponding HRP/IgG molar ratio.

Conjugates A-E were compared with a conjugate previously prepared by the sodium

meta-periodate method (Fields et al., 1983).

The performance of each conjugate was assessed by comparing a calculated P/N

TABLE 2

HRP/IgG molar ratios of selected pooled fractions from the concanavalin A-Sepharose 4B fractionation of

SPDP-prepared conjugates’

Conjugate Fraction Nos. ‘42,” &J HRP/lgG

pooled (Fig. 3) (molar ratio)

A 65-67 0.290 0.458 8.64

B 68-74 0.085 0.086 4.10

c 75-84 0.032 0.023 2.57

D 85-87 0.085 0.119 6.89

E 88-89 0.077 0.095 5.56

221

TABLE 3

Performance of HRP-IgC conjugates in ELISA for the detection of HBsAg”

Conjugate HRP/IgG Reciprocal of conjugate dilution

20 40 80 160 320 640

A 8.64 9 13 25h 16

B 4. I 56 72 66 F 25

C 2.51 29 43 49 43 30 46

D 6.89 25 27 30 32 19 12 - E 5.56 18 12 26 F ND 20 52 r _

a The results are expressed as P/N values.

b Underlined values indicate end-point dilutions determined visually

’ Not determined.

value obtained at various dilutions normalized to the same protein concentration. The

underscored values represent the optimal dilutions determined by visual inspection

(Table 3). Conjugate B yielded the highest P/N ratio of 72 at a conjugate dilution of

1 : 80. This conjugate was composed of an average of 4.1 enzyme molecules per IgG

molecule. The same conjugate performed almost as well at a dilution of 1 : 160.

Conjugate C also demonstrated a high P/N ratio at a 1 : 80 dilution. This conjugate

was composed of an average of 2.6 enzyme molecules per IgG molecule. As the

number of enzyme molecules increased per IgG molecule, the conjugate performance

decreased as indicated by a lower P/N value. In all cases high concentrations of

conjugate resulted in a lower P/N value due to a higher Adq3 value of the denominator

(N value).

DISCUSSION

The preparation of IgG-HRP conjugates by a thiol-disulphide bond exchange

reaction after modifying the proteins with SPDP has several advantages as compared

with other methods: (1) absence of IgG-IgG homopolymers, (2) better control over

the molecular composition (HRP/IgG ratio), and (3) production of a conjugate

possessing higher activity by ELISA. In addition, we present a single one-step affinity

chromatographic method of purifying the conjugate which eliminates the requirement

of a second gel-filtration step previously reported by Nilsson et al. (1981).

SPDP modification of protein involves the introduction of protected thiol groups

(Zpyridyl disulphide structures) covalently bound to E-amino groups of lysine. The

reaction is efficient and the DOS can be controlled by varying either the molar ratio of

SPDP : protein or the pH. The low DOS obtained at pH 9.0 at a molar ratio of 54 : 1 for

IgG was probably due to base catalyzed hydrolysis of SPDP which actually resulted in

222

a lowered molar ratio of intact SPDP to IgG. The 2-pyridyl disulphide (2-PDS)

protecting groups can be removed by a single reduction procedure which spares native

protein disulphide bridges.

The DOS of proteins is dependent upon the number of accessible E-amino groups

of lysine. IgG possesses between 50 and 70 groups while HRP possesses only 4

(Carlsson et al., 1978). Even though IgG contains many lysine residues only an

average of 7.5 2-PDS groups could be introduced indicating that most ofthese groups

are not accessible for SPDP modification or that optimal conditions were not used. It

was, therefore, not surprising that HRP, possessing only 4 lysine residues, required a

molar ratio of 75 : 1 (SPDP: HRP) to achieve a modest DOS of 1.65.

The lectin concanavalin A has been shown to bind HRP (Bio-Rad, Richmond, CA,

1982). This observation has led to the use of immobilized Con A for the purification of

IgG-HRPconjugates (Lanneret al., 1978; Arends, 1979; Nilsson etal., 1981). Unreact-

ed IgG does not bind to Con A-Seph 4B whereas conjugates and unreacted HRP bind

to the lectin necessitating a subsequent gel-filtration step. The separation method

described in this report uses a Con A-Seph 4B column equilibrated with a buffer

containing Tris, KC1 and EDTA (buffer B). The reaction mixture also contained the

same buffer. Under these conditions unreacted IgG and HRP were not significantly

adsorbed to the column, although HRP was slightly retarded. Most of the conjugate,

however, was bound to the column and was eluted with AMDM. Conjugate eluted

with 10 mM AMDM was qualitatively similar and differed only quantitatively to the

conjugate eluted with 100 mM AMDM. The separation procedure was performed on

three separate trials and similar results were obtained demonstrating consistent

reproducibility.

Why unreacted HRP did not bind to the affinity column is not understood. It is

known that Ca2’ and Mn?’ are required for Con A activity. In the presence of EDTA

(buffer B) these ions are chelated which might interfere with the binding of unreacted

HRP. However, in the presence of EDTA the majority of the conjugate remained

bound to the affinity column. This might have occurred due to conformation changes

in HRP molecules as a result of the conjugation reaction. Although the precise

mechanism for the preferential binding of IgG-HRP conjugates to Con A-Seph is not

understood, the method described was highly reproducible and eliminated the require-

ment of a second gel-filtration step. Furthermore, this method yielded conjugates

with different molar ratios of HRP/IgG which could be separated by pooling selected

fractions. HRP/IgG molar ratios in these pools varied from 2.6 to 8.6. Despite an

average DOS of 7.5 for IgG, not all of the modified groups participated in a disulphide

bond formation with the reduced HRP. Some of the modified groups probably were

located within the IgG molecule and were not accessible for the reaction. For

large-scale preparations, however, conjugates can be eluted with 100 mM AMDM in a

single pool. This pool had an average HRP/IgG ratio of 5.2 and could be used at a

dilution of 1 : 160 in an ELISA (data not shown).

The ELISA data indicated that conjugates composed of 2.5-4 HRP molecules per

223

IgG molecule yielded the greatest sensitivity. This observation is based upon a

comparison of P/N ratios obtained with various conjugates differing in their

HRP/IgG molar ratios. It has been previously demonstrated (Fields et al., 1983) that

P/N ratios are directly related to sensitivity. The decreased P/N values obtained with

conjugates composed of >4 HRP molecules per IgG molecule may be due to a loss of

IgG affinity or specificity for HBsAg due to a stereoconformational change of the

antibody binding site. These high molecular weight conjugates yielded a higher

background (N value) which decreased the overall P/N quotient.

This paper not only supports and extends the results of Nilsson et al. (198 l), but

also points out some important differences. First, the conjugates reported in the

present study possessed an average HRP : IgG molar ratio of 4.1 as compared to 2.0 by

Nilsson. This difference is probably due to the slightly higher DOS ( 1.18) achieved for

HRP which may have r,esulted in some degree of intermolecular bridging due to more

than one reaction site on some of the HRP molecules. Indeed, some conjugates

possessed molar ratios of greater than 8.0 (Table 2). Secondly, this paper examines the

parameters of SPDP modification of IgG as a function of SPDP : protein molar ratios

and pH, Finally, this paper describes a one-step method to separate conjugates from

free reactants. The mechanism responsible for this separation, however, remains

enigmatic.

REFERENCES

Arends, J.. 1979. J. Immunol. Methods 25. 171.

Avrameas. S.. 1969. Immunochemistry 6. 43.

Avrameas. S. and Ternynck, T., 1971, Immunochemistry 8, 1175.

Belanger. L., Hamel. D., Dufour, D. and Pouliot. M., 1976. Clin. Chem. 22, IX.

Bio-Rad Catalogw. January 1982, Richmond. CA, 4X.

Caldwell. C.W. and Barrett. J.T.. 1977. Chn. Chim. .4cta 81, 305.

Carlsaon, .I.. Drevin, H. and Axen. R., 1978. Biochem. J.. 173. 723.

Enpvall. E. and Pearlman, P.. 1971. Immunochemistry 8. 871.

Fields, H.A.. Davis. C.L.. Drecsman. G.R.. Bradley, D.W. andMaynard,J.E.. 19XI.J. Immunol. Methods

47. 145.

Fields, H.A.. Daws, C.L.. Bradley. D.W. and Maynard, .J.E.. 1983, Bull. Wrld Hlth Org. 61. 135.

Glshaw. J.B.. Hu. M.W.. Singh, P., Miller. J.G. and Schneider. R.S.. 1977, Clin. Chem. 23, 1144.

Grassetti. D.R. and Murray, J.F., 1967. Arch. Biochem. Biophys. 119, 41.

Halbert. S.P. and Anken, M., 1977, J. Infect. Dis. 136, 5318.

Lanner. M.. BerquiSt. R. and Carlsson. J.. 1978. in: Affinity Chromatography (Prrganon Press) p. 237.

Masseyeff. R.. 1978, Stand. J. Immunol. 8 (Suppl. 7). 83.

Nilsson. P.. Berqulst. N.R. and Grundy, M.S.. 1981. J. Immunol. Methods 41. 81.

Pal, S.B.. 1978. Walter de Gryuter. Berlin.

Pepple. J.. Moyom. E.R. and Yolken, R.N.. 1980, J. Pediat. 97. 233.

Voller, A.. Bartlett. A. and Bidwell. D.E.. 1978, Trap. Med. Hyg. 70. 98.

Warren. R.C.. Bartlett. A.. Bidwell. D.E.. Richardson. M.D.. Voller, A.and White. L.O., 1977,Br.Mtd..J.

I. 11x3.

Wel. R. ct al.. 1977. Clan. Chem. 23. 813.

224

Wilson, B.M. and Nakane, P.K., 1978, in: Immunofluorescence and Related Staining Techniques (Elsevier/

North-Holland Biomedical Press) p. 215.

Walters, G., L. Kuijpers, Kalaki, J. and Schuurs, A., 1976, J. Clin. Pathol. 29. 873.

Yolken, R.H., 1982, Rev. Infect. Dis. 4, 35.