An Indirect Double-Antibody Sandwich Enzyme-Linked

CLINICAL AND DIAGNOSTIC LABORATORY IMMUNOLOGY, Mar. 1996, p. 167–174 Vol. 3, No. 2

1071-412X/96/$04.0010

Immunosorbent Assay (ELISA) using Baculovirus-Expressed Antigen

for the Detection of Antibodies to Glycoprotein E of Pseudorabies

Virus and Comparison of the Method with Blocking ELISAs

TJEERD G. KIMMAN,

* OLAV DE LEEUW,

GRAZYNA KOCHAN,

BOGUSLAW SZEWCZYK,

EUGENE VAN ROOIJ,

LIESBETH JACOBS,

JOHANNES A. KRAMPS,

AND BEN PEETERS

Department of Porcine and Exotic Viral Diseases

and Department of Bovine Virology,

Institute for

Animal Science and Health ID-DLO, 8200 AJ Lelystad, The Netherlands, and

Department of Biochemistry, University of Gdansk, Poland

Received 27 September 1995/Returned for modiﬁcation 21 November 1995/Accepted 20 December 1995

Antibodies in porcine sera against glycoprotein E (gE) of pseudorabies virus (PRV) are usually measured in

blocking enzyme-linked immunosorbent assays (ELISAs) with one or two murine monoclonal antibodies

(MAbs) directed against gE. Our aim was to develop a conﬁrmation assay which is based on another principle

and which is able to detect antibodies directed against most potential binding sites on gE with high speciﬁcity.

Therefore, we developed an indirect double-antibody sandwich assay (IDAS) using recombinant gE expressed

by baculovirus (BacgE960). A fragment of the gE gene consisting of nucleotide positions 160 to 11020 of gE,

coding for the major antigenic sites of gE but not the transmembrane region, was cloned behind the signal

sequence of PRV gG and the p10 promoter in a baculovirus vector. Immunoblot analysis showed that the

expressed protein reacted with MAbs directed against ﬁve of the six antigenic sites on gE. Although the

conformation of some antigenic sites, notably antigenic sites E and C, was not identical to their natural

conformation, the expressed protein bound gE-speciﬁc antibodies in porcine sera in Western blots (immuno-

blots) and ELISAs. For the IDAS, a coating MAb directed against the nonimmunodominant antigenic site A

on gE was chosen. A major obstacle in binding ELISAs, such as the IDAS, appeared to be the high nonspeciﬁc

binding activity observed in porcine sera. As a result, sera could be tested only in relatively high dilutions in

the BacgE960 IDAS, in contrast to the testing of sera in blocking ELISAs. The sensitivity and speciﬁcity of the

newly developed BacgE960 IDAS were evaluated and compared with those of ﬁve commercially available

blocking ELISAs by using several sets of sera of known PRV disease history. The BacgE960 IDAS assay had

a high diagnostic speciﬁcity and a moderate sensitivity. The ﬁve blocking ELISAs differed remarkably in

sensitivity and speciﬁcity, thereby illustrating the need for standardization and conﬁrmation. We conclude that

the BacgE960 IDAS is a useful and speciﬁc additional (conﬁrmatory) test for the detection of antibodies to gE.

Vaccination of pigs against pseudorabies virus (PRV) (syn-

onyms, Aujeszky’s disease virus and suid herpesvirus type 1)

with marker vaccines that lack the nonessential glycoprotein E

(gE) (previously called gI) enables the detection of wild-type

(wt) PRV-infected pigs in vaccinated populations by using

serologic assays that detect antibodies to gE (37). This princi-

ple is currently used in control and eradication campaigns

worldwide. A critical factor in these campaigns is the reliabil-

ity, in terms of sensitivity, speciﬁcity, reproducibility, and ro-

bustness, of the serotests for gE-speciﬁc antibodies in porcine

sera. Current serotests for gE-antibodies are blocking enzyme-

linked immunosorbent assays (ELISAs) in which one or two

monoclonal antibodies (MAbs) directed against gE are used.

Reported difﬁculties in the detection of antibodies to gE in-

clude false-negative results, false-positive results, nonspeciﬁc

reactions, high rates of doubtful and weakly positive test re-

sults, high interassay variability between batches of commercial

test kits, low and variable gE-speciﬁc antibody responses in

pigs with maternal antibodies, and the so-called single reactors

(2, 31, 34, 36). A single reactor is a single gE-seropositive pig

in a herd and may indeed reﬂect the only seropositive pig in a

herd, i.e., a recently infected animal which has not or not yet

resulted in transmission of the infection, or a false-positive test

result (2). In addition, the success of vaccination-eradication

campaigns results in a very low seroprevalence, which may be

accompanied by a high rate of false-positive test results. To

conﬁrm with high predictive value such positive results, it is

desirable to have a speciﬁc conﬁrmation assay.

With murine MAbs, six antigenic domains, A to F, on gE

have so far been identiﬁed (13). Sera of PRV-infected pigs

predominantly recognize the conformational antigenic domain

E and to a somewhat lesser degree recognize the antigenic

domains C and D. Only a minority of infected pigs develop

antibodies directed to the antigenic domains A, B, and F (12).

The presently available gE serotests are blocking ELISAs

which measure the inhibition of binding of murine MAbs di-

rected against domain E or C or both by porcine antibodies

directed to these antigenic domains. The purpose of the

present study was to develop an additional, conﬁrmation assay

for the detection of antibodies to gE which is able (i) to detect

antibodies in an assay which makes use of a principle other

than the available blocking ELISAs used for screening, (ii) to

detect gE-speciﬁc antibodies with high speciﬁcity without los-

ing much sensitivity, and (iii) to detect antibodies directed

against as many antigenic sites on gE as possible. This would

also enable the serologic diagnosis of infections with potential

* Corresponding author. Present Address: RIVM, Research Labo-

ratory for Infectious Diseases, P.O. Box 1, 3720 BA Bilthoven, The

Netherlands. Phone: 31 30 2742330. FAX: 31 30 274444 9. Electronic

mail address: [email protected].

167

aberrant PRV strains that have mutations in the antigenic

domain E or C of gE (16). For that purpose, we expressed part

of the gE gene, containing all antigenic sites identiﬁed so far,

in the baculovirus system and developed an indirect double-

antibody sandwich assay (IDAS) using the baculovirus-ex-

pressed gE as antigen (BacgE960 IDAS).

MATERIALS AND METHODS

MAbs. MAb 23.3.1a is directed against porcine immunoglobulin G (IgG) (38).

Eleven MAbs directed against PRV gE have been described previously (13).

MAbs 1, 3, and 5 are directed against the discontinuous but almost linear

antigenic domain A (amino acids [aa] 78 to 239); MAbs 4, 8, and 11 are directed

against the continuous antigenic domain B (aa 52 to 67); MAbs 6 and 9 are

directed against the discontinuous antigenic domain C (aa 78 to 239); MAb 7 is

directed against the continuous antigenic domain D (aa 68 to 82); MAb 2 is

directed against the discontinuous antigenic domain E (aa 78 to 239); and MAb

10 is directed against the discontinuous antigenic domain F (amino acid range

unknown). Antigenic domain E and, to a lesser extent, antigenic domains C and

D are immunodominant in pigs. Only few pigs respond to antigenic domains A,

B, and F (12). The MAbs were puriﬁed from mouse ascites ﬂuid by 50%

ammonium sulfate precipitation, which was followed by dialysis overnight against

phosphate-buffered saline (PBS; pH 7.4).

Cells and viruses. The Autographa californica nuclear polyhedrosis virus

(AcNPV) was used as vector for the expression of gE under the control of the

p10 promoter. The methods have been described previously (9, 41). Brieﬂy,

AcNPV was grown in monolayers of the Spodoptera frugiperda cell line Sf21

(ATCC CRL 1711) (39) at 288C in TC100 medium (GIBCO-BRL) (7) containing

10% heat-inactivated fetal bovine serum (Sebak 31054) and antibiotics (100 IU

of penicillin per ml, 110 mg of streptomycin per ml, and 2.5 mg of fungizone per

ml). For cotransfection, Sf21 cells were grown in Grace’s insect tissue culture

medium (GIBCO-BRL) supplemented with 10% heat-inactivated fetal bovine

serum and antibiotics. For production of ELISA antigen, Sf21 cells were grown

in SF900 serum-free insect culture medium (GIBCO-BRL 10900-066) supple-

mented with antibiotics. Virus titers were determined by end point dilution as

described elsewhere (32).

Construction and characterization of recombinant baculovirus expressing gE.

For the construction of recombinant baculovirus, the DNA of PRV strain NIA-3

(22) was digested with BamHI. The BamHI 7 fragment was subsequently cloned

in vector pBR322 (Boehringer Mannheim), generating pBR322-Bam7. Plasmid

DNA was recovered by standard methods (29). A fragment of 1,914 bp contain-

ing 1,020 bp of the gE gene was excised from vector pBR322-Bam7 by using

ApaLI. The ends of the fragment were ﬁlled in with Klenow, and the fragment

was subsequently cloned into the HincII site of pUC19 (Biolabs) (pUC19-

ApaLIfr.). A fragment of 1,264 bp containing the ﬁrst 370 bp of the gE gene was

removed from pUC19-ApaIfr by using BamHI and NcoI. This fragment was

replaced by a 310-bp fragment consisting of nucleotide positions 160 to 1370

of gE, which was obtained in a PCR with Taq polymerase (Perkin-Elmer). A

BamHI restriction site was introduced at the 160 position during PCR ampliﬁ-

cation. The resulting plasmid (pUC19-gE) contains the gE fragment without the

gE signal sequence and without the C-terminal transmembrane region and was

used to realize an in-frame connection with the signal sequence of the PRV

envelope protein G (gG) in plasmid pARK6gGss. The pARK6gGss vector was

obtained by cloning the gG signal sequence from plasmid pAcAs3gX into

pARK6 (9). The gE-ApaLI fragment was excised from pUC19-gE by using

BamHI and HindIII and was cloned into pARK6gGss digested with BamHI and

XbaI. The sticky ends of HindIII and XbaI were ﬁlled in with Klenow before

digestion with BamHI. The resulting vector, pARKgE960, was used for cotrans-

fection with AcNPV wt DNA.

For cotransfection, conﬂuent monolayers of Sf21 cells (2 3 10

) grown in T25

ﬂasks were cotransfected in a 49-cm

petri dish (Falcon) with 2.5 mg of plasmid

pARKgE960 DNA and 0.5 mg of baculovirus AcNPV wt DNA by the calcium

phosphate coprecipitation method (32). Recombinant viruses expressing b-ga-

lactosidase were puriﬁed by six rounds of plaque puriﬁcation under agarose with

2.5 mg of Bluo-Gal (GIBCO-BRL) per ml in a TC100 agar overlay as the

indicator. Expression of gE was examined by Western blotting (immunoblotting)

(see below). One recombinant expressing gE was used to prepare high-titer virus

stocks (.10

50% tissue culture infective doses [TCID

] per ml) for further

experiments and was designated BacgE960.

The optimal time and localization of expression of gE in cells and the super-

natant of BacgE960-infected Sf21 cells were determined by infecting Sf21 cells at

a multiplicity of infection of 10 TCID

per cell. Cells and cellular fractions were

harvested each day until day 6 postinfection and were analyzed by Western

blotting and ELISA for the expression of gE. Fractionation of cells was per-

formed according to the method described by Summers and Smith (32). Brieﬂy,

the cells were harvested, centrifuged, and washed with PBS. The cells were

suspended in lysis buffer (0.03 M Tris-HCl [pH 7.5], 0.01 M Mg acetate, 1%

Nonidet P-40) and were incubated on ice for 10 min with occasional stirring.

Then, the nuclei and membrane fractions were spinned down for 5 min at 2,000

3 g, and the supernatant containing the cytoplasmic fraction was transferred to

a new tube. The pellet was washed and resuspended in PBS.

Western blotting. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis

(SDS-PAGE) and immunoblotting were performed as described by Sambrook et

al. (29). Brieﬂy, proteins were precipitated with 10% TCA for 15 min. The pellets

were washed twice in cold acetone and resuspended in sample buffer (10 mm

Tris-HCl [pH 8.0], 1 mM EDTA, 10% glycerol, 2% SDS, 5% 2-mercaptoethanol,

0.25% bromphenol blue). Resuspended proteins were denatured by being boiled

for 5 min in sample buffer and were then subjected to electrophoresis in an

SDS-12% PAGE gel using Laemmli’s (19) discontinuous system in a Mini-

Protean II Electrophoresis Cell (Bio-Rad). Separated proteins were electro-

phoretically transferred to a 45-mm-pore-size Protran cellulose nitrate mem-

brane (Schleicher and Schuell) with a Mini Trans-Blot Electrophoretic Transfer

Cell (Bio-Rad). The membranes were washed twice in PBS containing 0.5 M

NaCl, 5% horse serum (Sera-Tech 9904419), and 0.05% (vol/vol) Tween 80

(PBS-NHT buffer). The membranes were subsequently incubated for1hatroom

temperature with an optimal dilution of MAbs directed against gE. The mem-

branes were then washed three times for 5 min in PBS-NHT buffer and were

incubated with rabbit anti-mouse IgG-horseradish peroxidase conjugate (DAKO

A/S; article no. P260) for1hatroom temperature. The membranes were washed

three times for 5 min in PBS-gT buffer and once in PBS and were ﬁnally

incubated in substrate solution (0.5 mg of 39-39-diaminobenzidine, 0.001%

Tunicamycin treatment. The cells were infected with BacgE960 at a multiplic-

ity of infection of 10 and were incubated in serum-free medium containing 4 mg

of tunicamycin (Sigma) per ml (14). Seventy-two hours postinfection, the cells

and medium were harvested and analyzed by SDS-PAGE and immunoblotting.

Production of ELISA antigen. For production of ELISA antigen, approxi-

mately 2 3 10

Sf21 cells were infected with BacgE960 or, as a control, wt

baculovirus (wt AcNPV) in a T175 ﬂask at a multiplicity of infection of 2 to 5

TCID

per cell for 1.5 h at room temperature. The virus was removed, and the

cells were supplied with 20 ml of SF900 serum-free insect culture medium and

were grown at 288C until all cells contained visible polyhedrins, which was usually

at day 5 postinfection. The medium was collected, clariﬁed by centrifugation for

10 min at 600 3 g, and subsequently ﬁltered through a 0.2-mm-pore-size ﬁlter.

The antigen was stored at 2208C before use in the IDAS.

IDAS. The test was designed to detect the binding of porcine IgG antibodies

to recombinant gE (BacgE960) bound to the wells of a microtiter plate by a

catching MAb directed to a linear nonimmunodominant epitope on gE (MAb 5).

All reagents were added in 100-ml quantities. After each incubation step, the

plates were washed six times with a solution of 0.05% Tween 80 in tap water. For

coating, ELISA plates (Costar E.I.A. 3590) were incubated overnight at 378C

with MAb 5 dissolved in PBS. The plates were subsequently incubated for1hat

378C with an optimal dilution of the baculovirus-expressed gE antigen (clariﬁed

and ﬁltered culture supernatant of BacgE960-infected Sf21 cells). A control plate

was similarly incubated with supernatant of wt AcNPV-infected Sf21 cells. Serum

test samples, diluted 1:20 in test buffer consisting of PBS containing 20% horse

serum (Sera-Tech 9904419), 0.5 M NaCl, and 2% Tween 80, were then pipetted

into the wells of the antigen-containing and control plate and incubated for 30

min at 378C. To determine ELISA titers, serial twofold dilutions of the test

samples were made in the test buffer. An optimal dilution of the conjugate was

prepared in PBS containing 4% fetal bovine serum, and the dilution was incu-

bated for 60 min at 378C. Conjugate was prepared by coupling the anti-porcine

IgG MAb 23.3.1a to horseradish peroxidase (Boehringer Mannheim; article no.

814407) according to the method described by Wilson and Nakane (44). The

conjugate was dialyzed against PBS, after which glycerol was added to a ﬁnal

concentration of 50%. The conjugate was stored at 2208C and, immediately

before use, was diluted in PBS containing 4% fetal bovine serum. Finally, 100 ml

of the substrate-chromogen mixture, consisting of 3,39,5,5-tetramethylbenzidine

(Sigma; 1 mg/ml) and H

(0.005%) in 0.1 M Na-acetate buffer (pH 6.0), was

added. After incubation for 15 min at room temperature, the reaction was

stopped by the addition of 100 mlof0.5MH

. Color development was

measured at 450 nm with a microplate reader (EAR 400; SLT-Labinstruments,

Vienna, Austria).

The following controls were included in each assay. As mentioned, every test

was done in a plate with gE antigen, yielding an optical density at 450 nm for gE

[OD

450

(gE)], and in a plate with control antigen, yielding an OD

450

(control).

Four wells in row 12 of each plate contained twofold dilutions of a positive

reference serum sample, two wells were ﬁlled with a negative serum sample, and

two wells were ﬁlled with buffer during the serum incubation step. A sample (or

sample dilution) was scored as positive when the OD

450

(gE) minus OD

450

(control) was .0.4. The test was considered valid when (i) the positive control

serum sample scored positive in all four dilutions, (ii) the negative control serum

sample scored negative in both wells, and (iii) the control wells tested without

serum showed OD

450

of #0.04.

Validation of the BacgE960 IDAS and comparison with blocking ELISAs. The

newly developed BacgE960 IDAS was validated with 395 well-deﬁned sera and

was compared with ﬁve commercially available blocking ELISAs for the detec-

tion of antibodies to PRV gE (Svanova, Uppsala, Sweden; Rhone Merieux,

Lyon, France; Eurodiagnostica, Apeldoorn, The Netherlands; Idexx, Portland,

Maine, European Veterinary Laboratories, Woerden, The Netherlands). The

ﬁve commercially available blocking ELISAs were designated (in random order)

168 KIMMAN ET AL. CLIN.DIAGN.LAB.IMMUNOL.

from A to E. The samples were analyzed in the blocking ELISAs strictly accord-

ing to the instructions of the manufacturers. Tests B and E are based on the

principle of the complex trapping blocking ELISA (8), which makes use of a

preincubation step of test serum and antigen in a separate plate; transfer of the

serum-antigen mixture to the test plate, which is coated with one anti-gE MAb;

and a conjugate step with a second anti-gE MAb. Hence, these tests detect

porcine antibodies against one or two antigenic sites on gE. Tests A, C, and D are

based on the principle of a direct blocking ELISA, which makes use of an

antigen-coated plate, a serum incubation step, and a conjugate step with an

anti-gE MAb. Hence, these tests detect porcine antibodies against one antigenic

site on gE. In ELISAs B, C, and E, undiluted serum is used, whereas in ELISAs

A and D, serum samples must be diluted 1:2.

The following serum collections were available for this purpose (i). Ten weakly

positive and 17 negative serum samples were obtained from speciﬁc-pathogen-

free (SPF) pigs from our institute and several other European laboratories. The

pigs had been vaccinated and infected according to various protocols to induce

low to very low levels of antibodies to gE or to give high levels of antibodies to

PRV without antibodies to gE. Some of these serum samples have been selected

by the Subcommittee on Aujeszky’s disease of the European Union Scientiﬁc

Veterinary Committee. (ii) Sixteen positive and 56 negative swine serum samples

were obtained from pigs with a known history of PRV. Positive serum samples

were obtained from pigs that were infected with wt PRV or that were vaccinated

with gE-positive vaccines. Negative samples were obtained from PRV-free herds,

either vaccinated with gE-negative vaccines or not vaccinated, in the United

Kingdom and the Netherlands. (iii) An additional set of 249 negative samples

was obtained from PRV-free herds from the United Kingdom. This set was used

only to examine the speciﬁcity of the BacgE960 IDAS. (iv) To determine the

detection limit of each test, serial twofold dilutions (1:2 to 1:2,048 for the

blocking ELISAs and 1:40 to 1:40,960 for the BacgE960 IDAS) of 14 positive

serum samples were made in negative SPF serum. (v) In addition, a collection of

six well characterized serum samples were used to further analyze the detection

limit of each method. These six samples were used in an interlaboratory ex-

change program according to ISO-5725 (10) to check the sensitivity, speciﬁcity,

reproducibility, and repeatability of gE antibody tests in different laboratories in

the Netherlands. (vi) The ability of each method to detect antibodies early after

infection was evaluated by testing serum samples obtained from a pig that was

vaccinated intranasally with the PRV Bartha strain and challenge inoculated

intranasally with 10

PFU of the virulent NIA-3 strain (22) at 3 weeks postvac-

cination. Blood samples were collected at days 8, 11, 15, 18, 21, and 76 after

challenge inoculation. (vii) To further determine the speciﬁcity of the tests, 21

serum samples of different origins were analyzed in each test: eight nonimmune

calf serum samples, four bovine herpesvirus type 1 (BHV1)-immune calf serum

samples, one equine arteritis virus-immune horse serum sample, one porcine

parvovirus-immune pig serum sample, one porcine inﬂuenza virus H3N2-im-

mune pig serum sample, and one porcine inﬂuenza virus H1N1-immune pig

serum sample, and ﬁve Streptococcus suis (type 2)-immune pig serum samples

were available for this purpose. With the exception of the 249 negative serum

samples from the United Kingdom, all test samples were analyzed by each of the

six methods. On the basis of the results obtained with the negative and positive

serum samples, speciﬁcity and sensitivity were determined, respectively. The

speciﬁcity of an ELISA was expressed as the percentage of the negative serum

samples obtained from animals never infected with PRV which gave unambig-

uously negative results. According to a worst-case scenario, we considered doubt-

ful results positive. The sensitivity of an ELISA was expressed as the percentage

of the positive serum samples obtained from animals (experimentally) infected

with PRV which gave unambiguously positive results. Doubtful results were

considered negative.

RESULTS

Expression of PRV gE by baculovirus. PRV gE was cloned

behind the N-terminal signal sequence of PRV gG to allow efﬁ-

cient intracellular transport. The C-terminal transmembrane

anchor was removed from the construct to allow efﬁcient se-

cretion into the supernatant. The construction of the pARKgE

960 transfer vector is schematically shown in Fig. 1. One re-

combinant baculovirus, which was designated BacgE960, was

used to prepare high-titer virus stocks, to characterize the ex-

pressed gE product, and to develop the IDAS for the detection

of porcine antibodies to gE.

The localization of the recombinant BacgE960 product was

checked by Western blot analysis of the supernatant of BacgE

960-infected Sf21 cells, as well as of nuclear, membrane, and

cytoplasmic fractions. BacgE960 was demonstrated in all frac-

tions. The cytoplasmic fraction contained a mixture of heter-

ogeneously glycosylated proteins with molecular masses of 42

to 46 kDa (data not shown), which is in accordance with the

molecular mass predicted from the DNA sequence (about 44

to 46 kDa). The nuclear and membrane fractions contained a

gE protein with a molecular mass of approximately 42 kDa,

suggesting that this form of gE is poorly glycosylated. In the

supernatant, gE was mainly expressed as a doublet with mo-

lecular masses of 44 and 46 kDa. Tunicamycin treatment of

BacgE960-infected Sf21 cells and then Western blot analysis

resulted in one band with an apparent molecular mass of 38

kDa, therewith revealing heterogeneous N-linked glycosylation

of the expressed gE polypeptides of 42 to 46 kDa (data not

shown). The heterogeneity of the glycoprotein bands from 42

to 46 kDa is consistent with the presence of four potential

glycosylation sites as determined by computer sequence anal-

ysis. Tunicamycin treatment reduced the amount of secreted

gE, while the 38-kDa form accumulated intracellularly. These

results conﬁrm those of Jarvis and Summers (14), who showed

that glycosylation is a prerequisite for efﬁcient secretion of a

protein in baculovirus-infected insect cells.

FIG. 1. Construction of transfer vector pARKgE960. Fragments of PRV gE

were cloned into the multicopy vector pUC19 and then subcloned into the

transfer vector pARK6. Arrows indicate the directions of transcription. Ac,

AcNPV DNA; Ap, ampicillin resistance gene; Tc, tetracycline resistance gene;

p10p, p10 promoter; gGss, PRV gG signal sequence; LacZ, Escherichia coli lacZ

gene; STOP STOP STOP, stop codons in three reading frames.

VOL. 3, 1996 IDAS FOR PRV gE ANTIBODY DETECTION 169

When cells were infected at a multiplicity of infection of 10

TCID

per cell, the expression reached a maximum level at

day 3 postinfection before the cells began to lyse.

Antigenicity of BacgE960. To determine whether the bacu-

lovirus-expressed antigen contained all identiﬁed antigenic do-

mains of gE, Western blot analyses were performed with mu-

rine MAbs representative for each of the antigenic domains A

to F (Fig. 2). With the exception of MAb 10, directed against

the discontinuous antigenic domain F, the MAbs recognized

the 44- and 46-kDa protein doublet and sometimes a faint and

diffuse 42-kDa band as well. However, MAb 2 and MAbs 6 and

9, respectively, directed against the discontinuous antigenic

domains E and C (12), reacted with faint or intermediate

bands. These positive results were somewhat surprising, be-

cause we have previously shown that the conformation of the

discontinuous antigenic domains E and C (but not of the linear

antigenic domains A and D) is partly dependent on the non-

covalently linked complex between gE and gI (11). However,

BacgE960 reacted poorly (in comparison with wt NIA-3 anti-

gen) in an ELISA with MAbs 2 and 9 as catching and detecting

antibodies, respectively. We concluded that the BacgE960 an-

tigen contained most of the major identiﬁed antigenic domains

on gE. However, the conformation of the antigenic domains E

and C is most likely not completely identical to its natural

conformation.

Development of the BacgE960 IDAS. The ELISA with the

BacgE960 as antigen was set up by using 30 negative and 20

positive serum samples with known history as reference mate-

rial. Incubation temperatures and periods, type and pH of

buffers, washing conditions, coating material, serum dilutions,

blocking solutions, concentrations of reagents, and cutoff val-

ues were chosen in a series of checkerboard titrations. Because

the purpose of the work was to develop an assay which can be

used to conﬁrm the results of screening ELISAs in the latter

stages of eradication-control programs (in which the preva-

lence of the disease may be very low), the assay was set up to

have a high speciﬁcity with an acceptable level of sensitivity. It

further appeared necessary to diminish the nonspeciﬁc binding

of the porcine serum samples to the ELISA plate as much as

possible. Sera from conventional pigs from the ﬁeld gave

higher nonspeciﬁc binding than sera from SPF pigs.

During the development of the assay, it appeared that the

sensitivity could be improved by coating a MAb directed

against PRV to catch the antigen to the wells of the microtiter

plate instead of coating the antigen directly. Because the use of

a coating MAb could block the binding to the antigen of

antibodies in test samples (and thus reduce the sensitivity of

the test), we selected MAb 5, which is directed against anti-

genic domain A on gE and to which only a small proportion of

PRV-infected pigs develop antibodies (12). The ﬁnal setup of

the assay was described in Materials and Methods and is shown

schematically in Fig. 3.

To test the reproducibility of the BacgE960 IDAS, two

strongly positive and two weakly positive serum samples were

tested 7 times on different days over a 3-month period. The

mean [OD

450

(gE) 2 OD

450

(control)] 6 standard deviation

values obtained with these serum samples were 1.82 6 0.10,

1.56 6 0.09, 1.03 6 0.16, and 0.62 6 0.09, yielding low and

acceptable coefﬁcients of variation of 5.5, 5.8, 15.6, and 14.6%,

respectively.

Attempts were further made to develop a Western blot assay

with BacgE960 as the antigen for the detection of antibodies

directed against gE in porcine sera. These attempts were not

pursued because of the lower degree of sensitivity of Western

blot analysis in comparison with the IDAS.

Validation of the BacgE960 IDAS and comparison with

blocking ELISAs. The reliability of the newly developed BacgE

960 IDAS was determined and simultaneously compared with

those of ﬁve commercially available blocking ELISAs (A to E)

by using several sets of well-deﬁned sera. All samples were

tested according to the protocol of the BacgE960 IDAS or the

instructions of the manufacturer. However, tests A and E did

not fulﬁll the criteria of validity as given by the manufacturer,

i.e., these tests failed to reach the prescribed results with con-

trol sera.

First, the tests were evaluated by testing 10 weakly positive

and 17 negative serum samples from SPF pigs which had been

vaccinated and infected in different laboratories. The pigs were

treated according to protocols to give low to very low levels of

antibodies to gE or to give high levels of antibodies to PRV

without antibodies to gE. As shown in Table 1 (ﬁrst row), the

BacgE960 IDAS failed to detect these low levels of gE anti-

bodies in 5 of 10 samples but showed a high speciﬁcity. Tests A

to E showed various results. Second, the tests were evaluated

by using 16 positive and 56 negative serum samples from pigs

with a well-known PRV history (Table 1, second row). The

combined data obtained with all positive (10 plus 16) and

negative (17 plus 56) serum samples are given in Table 1, third

row. These data show that the BacgE960 IDAS combines a

high level of speciﬁcity (99%) with a slightly lower level of

FIG. 2. Antigenicity of BacgE960 examined by gel electrophoresis and West-

ern blot analyses with BacgE960 antigen (Ag) (1) and control antigen (2)

(supernatant of wt AcNPV-infected cells). Lane samples were incubated with

MAb 5 (representative of antigenic domain A), MAb 8 (representative of anti-

genic domain B), MAb 9 (representative of antigenic domain C), MAb 7 (rep-

resentative of antigenic domain D), MAb 2 (representative of antigenic domain

E), and MAb 10 (representative of antigenic domain F). Molecular mass cali-

brations are indicated to the left of the blot.

FIG. 3. Schematic representation of the BacgE960 IDAS. MAbagE(A),

coating MAb directed against antigenic domain A of gE; BacgE960, baculovirus-

expressed gE antigen (BacgE960) containing antigenic sites A to E; SwIgagE,

test serum containing antibodies directed against gE; MAbaSwIgGPO; peroxi-

dase (PO)-labelled MAb directed against swine IgG.

170 KIMMAN ET AL. CLIN.DIAGN.LAB.IMMUNOL.

sensitivity (81%). Remarkably, the commercially available

ELISAs A to E differed strongly in sensitivity and speciﬁcity.

Tests B and D combined high levels of sensitivity and speci-

ﬁcity, with test D being slightly more superior in speciﬁcity. In

addition to not fulﬁlling the criteria of validity given by the

manufacturer, tests A and E showed a relatively low level of

sensitivity and a moderate level of speciﬁcity. Test C revealed

the highest sensitivity of the tests. However, in contrast to the

high level of sensitivity was its relatively low speciﬁcity of 85%.

In a number of cases, negative samples were scored as doubtful

by test C. Third, the speciﬁcity of the BacgE960 IDAS was

further examined by testing 249 serum samples from PRV-free

herds from the United Kingdom. Only three of these samples

reacted positively, giving a speciﬁcity of 99%. Fourth, to com-

pare the detection level of each test, 14 selected positive serum

samples were serially diluted twofold in negative SPF swine

serum and analyzed. The highest dilution of each serum sam-

ple giving a positive result was calculated (Table 2). These data

indicate that on average the BacgE960 IDAS gave the highest

dilution giving a positive sample result. Fifth, to further deter-

mine the detection limit of each method, six well-characterized

serum samples were analyzed in each test. These six samples

are used in an interlaboratory exchange program (10) to check

the sensitivity, speciﬁcity, reproducibility, and repeatability of

gE antibody tests in different laboratories. On the basis of the

results obtained over a 2-year period in the participating lab-

oratories, the sera had arbitrarily been designated as sera con-

taining no, very low, low, medium, high, or very high levels of

gE antibodies. As shown in Table 3, the BacgE960 IDAS

detects gE antibodies in the samples containing medium and

higher levels of gE antibodies, but not in the samples contain-

ing no, very low, or low levels of gE antibodies. This apparent

contradiction between a high detection limit as determined by

diluting positive sera and a somewhat lower diagnostic sen-

sitivity when the test is used at a single dilution is due to the

high sample dilution (i.e., 1:20) which had to be used in the

BacgE960 IDAS to circumvent the nonspeciﬁc binding of por-

cine sera. Sixth, the ability of each method to detect gE anti-

bodies shortly after infection was evaluated by analyses of

serially collected serum samples obtained from a pig which was

vaccinated and subsequently infected. With the exception of

tests A (day 76) and C (day 11), all tests detected infection at

day 15 postinfection. Seventh, to further determine the speci-

ﬁcities of the tests, 21 serum samples of different origins were

analyzed in each test. These serum samples gave negative

scores in all but one test; test E gave doubtful scores for two

BHV1-seronegative calf serum samples (data not shown).

DISCUSSION

We report the synthesis of a truncated gE antigen from a

recombinant virus expression system for application in a diag-

nostic test. Several previous attempts in our laboratory to de-

velop recombinant gE for use in an immunodiagnostic assay

were unsuccessful, either because of inappropriate processing

and conformation of recombinant gE in bacterial cells, a low

level of expression of recombinant gE in eukaryotic 3T3 cells,

or unsuccessful expression of full-length gE in bacterial and

TABLE 1. Results obtained by BacgE960 IDAS and ELISAs A to E with known positive and negative serum samples

Assay

Result (%) for

Sera from experimentally inoculated

SPF animals (n 5 27)

Sera from pigs with a known

PRV history (n 5 72)

Totals (n 5 99)

Speciﬁcity

Sensitivity

Speciﬁcity

Sensitivity

Speciﬁcity

Sensitivity

BacgE960 IDAS 100 (17/17) 50 (5/10) 98 (55/56) 100 (16/16) 99 (72/73) 81 (21/26)

Test A 100 (17/17) 80 (8/10) 95 (53/56) 94 (15/16) 96 (70/73) 88 (23/26)

Test B 100 (17/17) 90 (9/10) 96 (54/56) 100 (16/16) 97 (71/73) 96 (25/26)

Test C 88 (15/17) 100 (10/10) 84 (47/56) 100 (16/16) 85 (62/73) 100 (26/26)

Test D 100 (17/17) 90 (9/10) 100 (56/56) 100 (16/16) 100 (73/73) 96 (25/26)

Test E 100 (17/17) 70 (7/10) 91 (51/56) 100 (16/16) 93 (68/73) 88 (23/26)

Numbers in parentheses are the numbers of negative test results/numbers of tested negative sera.

Numbers in parentheses are the numbers of positive test results/numbers of tested positive sera.

TABLE 2. Sensitivities of the BacgE960 IDAS and ELISAs A to E

as deﬁned by the highest dilution giving a positive test result

Serum

sample

no.

Highest dilution giving a positive result for:

BacgE960IDAS Test A Test B Test C Test D Test E

1 #20 16 8 16 8 2

2 320 64

16 64 16 4

3 #20 32 16 64 16 8

4 #20 16

64 64

5 128 256 128 128 64 32

6 2,560 1,024 1,024 1,024 1,024 128

7 640 32 16 32 32 4

840244824482

9#20 16 2 32 #1

10 640 64 32 128 32 16

11 1,280 64 64 256 64 16

12 2,560 $2,048 $2,048 $2,048 $2,048 512

13 #20 2 #14#1#1

14 10,240 512

128 512 128 64

Mean

200 63 32 79 40 10

Serial dilutions gave inconclusive results.

Geometric means. To allow calculation, negative values were set at the

detection level and values greater than or equal to and less than or equal to were

read as equal to the respective values.

TABLE 3. Sensitivities and speciﬁcities of the BacgE960IDAS and

ELISA A to E as deﬁned by six test serum samples

Serum sample

(antibody level)

Result with:

BacgE960IDAS Test A Test B Test C Test D Test E

1 (negative) 222222

2 (very low) 262621

3 (low) 211112

4 (medium) 111111

5 (high) 111111

6 (very high) 111111

Level of gE-speciﬁc antibodies.

VOL. 3, 1996 IDAS FOR PRV gE ANTIBODY DETECTION 171

insect cells (results not shown). The last ﬁnding may be due to

the toxicity of mature full-length gE, as suggested by the in-

stability of recombinant baculovirus-expressing full-length gE

(results not shown). A possible explanation for toxicity may be

that accumulation of full-length gE, including the transmem-

brane region, in the membranes of the endoplasmic reticulum

may inhibit protein syntheses, resulting in low levels of pro-

duction of the expressed protein (9). Therefore, we decided to

express a truncated part of gE containing the major antigenic

part of gE in the baculovirus expression system (13). Because

most aspects of the intracellular modiﬁcation of proteins in

insect cells are similar to those in eukaryotic cells (1, 18, 21,

25), conformation-dependent antigenic sites of baculovirus-

expressed proteins may largely have the native structure, there-

with facilitating their use in immunodiagnostic assays, as pre-

viously shown (4, 27, 33, 35, 40, 42). A correct folding of the

expressed gE is important, because we previously showed that

the major immunodominant antigenic sites on gE are confor-

mation dependent (12, 13). In addition, the level of expression

of heterologous proteins in the baculovirus-insect cell system is

high (18, 21, 25). The PRV gG signal sequence was introduced

to allow efﬁcient intracellular transport. Because the expressed

BacgE960 contained no transmembrane region, the expressed

protein was excreted, although partly, in the supernatant of

BacgE960-infected insect cells, allowing an easy harvest of

recombinant proteins. Because it has been previously shown

that secretory and membrane proteins are excreted properly

only if they are correctly folded and oligomerized in the endo-

plasmic reticulum (9, 20, 26), excretion of the BacgE960 prod-

uct in the supernatant of infected cells is a further indication of

a correct processing and folding of the recombinant antigen.

Western blotting (Fig. 2) indicated that the expressed

BacgE960 antigen contained most of the identiﬁed antigenic

sites, including the immunodominant conformational antigenic

domains C and E.

Although BacgE960 was clearly antigenic in the Western

blot analysis, two lines of evidence suggested that the antige-

nicity of the expressed BacgE960, especially of the immuno-

dominant conformational antigenic domains, is not identical to

that of gE in wt PRV virions or virus-infected cells. First,

BacgE960 reacted poorly (in comparison with complete viral

antigen) in a blocking ELISA with MAbs to antigenic sites E

and C as catching and detecting antibodies. Second, we have

previously shown that the binding of MAbs directed to anti-

genic domains E and C was enhanced by complexing gE to gI

(previously called gp63) (11). In PRV virions and in PRV-

infected cells, gE forms a noncovalently linked complex with gI

(43, 46). Not only in PRV, but also in herpes simplex virus type

1 (15), BHV1 (28), varicella-zoster virus (45), and Marek’s

disease virus (3), gE and gI homologous proteins are nonco-

valently bound. In PRV, the gE-gI complex (previously called

gI-gp63 complex) is a functional and apparently also antigenic

entity (6, 11, 43, 46). These ﬁndings indicate that application of

a gE-gI complex as antigen for the detection of gE-speciﬁc

antibodies may have advantages. This is possible in blocking

ELISAs but impossible in binding assays such as the IDAS;

note that vaccinated pigs may develop gI-speciﬁc antibodies.

Nonetheless, despite the intrinsic shortcomings of the ex-

pressed BacgE960, it appeared sufﬁciently antigenic for use in

ELISAs and Western blot assays. The developed BacgE960

IDAS detects gE-speciﬁc porcine IgG. In an antibody capture

assay, the antigen would also be able to detect gE-speciﬁc IgM

(17).

An advantage of an IDAS with recombinant or puriﬁed (23)

gE is its potential ability to detect antibodies directed against

several antigenic determinants on the protein, in contrast to

that of the current commercially available blocking ELISAs,

which detect antibodies directed against only one or two

epitopes. Therefore, the BacgE960 IDAS may have the ability

to detect infection with PRV strains that have mutations in

antigenically important regions of gE. The occurrence of such

strains may impair PRV eradication campaigns in which the

blocking ELISAs are used as a diagnostic test to discriminate

infected from vaccinated pigs. One such strain has been de-

scribed by Katz and Pederson (16). In addition, a partially

gE-deleted PRV isolate from the ﬁeld with a low level of

expression of a truncated form of gE has been described else-

where (24). The importance of gE serotests for PRV eradica-

tion campaigns illustrates the necessity of both the character-

ization of the gE phenotype of new isolates and of serological

surveillance by the BacgE960 IDAS.

Because the prevalence of gE-seropositive pigs in vaccinated

herds may be very low (5, 31), a sensitive test and large sample

size are needed to detect positive herds. However, because

high sensitivity usually combines with low speciﬁcity (see, for

example, test C), this will likely result in a large number of

false-positive test results. We developed the BacgE960 IDAS

to have a high speciﬁcity so that it can be used to conﬁrm such

positive test results in herds with a low prevalence, thereby

enhancing the predictive value of positive test results.

For the evaluation of the diagnostic quality of the BacgE960

IDAS and the commercial tests A to E, we used several sets of

well-characterized sera from pigs with known PRV histories

regarding both natural and experimental infections. The ad-

vantage of this approach is that sensitivity and speciﬁcity can

be calculated unambiguously, which is difﬁcult or impossible

when sera from the target population itself are used, because

the infection history is usually unknown. However, the choice

of sera may bias the results. In addition, some sera from ex-

perimental infections may not be representative for the ﬁeld

situation (30). For example, the ﬁrst set of sera (Table 1, ﬁrst

row) contained experimental sera with low to very low concen-

trations of gE antibodies. Because the BacgE960 IDAS was

developed to have a high speciﬁcity, it conversely had a rela-

tively lower sensitivity, which was especially clear with this

collection of sera. The second set of sera (Table 1, second row)

may be more representative for the ﬁeld situation. With this set

of sera, the BacgE960 IDAS gave a higher-sensitivity score in

addition to its high speciﬁcity. Other sets of sera (including

sera from PRV-free herds from the United Kingdom) were

further included to test the speciﬁcity of the BacgE960 IDAS.

Because PRV gE shows sequence homology with gEs of other

alphaherpesviruses, we also examined the possibility that the

BacgE960 IDAS detected anti-BHV1 gE-speciﬁc antibodies in

calf sera cross-reacting with PRV gE. However, the four

BHV1-immune calf serum samples all reacted negatively in the

BacgE960 IDAS. We concluded from the validation of tests

that several commercially available ELISAs differed remark-

ably in quality (test B and especially test D had high levels of

sensitivity and speciﬁcity), and that the BacgE960 IDAS had a

high level of speciﬁcity and an acceptable level of sensitivity.

The different degrees of reliability of the commercial tests and

their high interassay variability (34) point to the need for rigid

standardization as well as for use of a conﬁrmatory test. When

used to conﬁrm positive results of screening assays, the

BacgE960 IDAS will increase the predictive value of ultimately

positive test results. The BacgE960 IDAS may further be use-

ful as an additional conﬁrmation assay because some commer-

cial assays (notably tests A, C, and E) detect only antibodies

against a single epitope. Nonetheless, when the BacgE960

IDAS is used as a conﬁrmation assay, it should be kept in mind

that the test may fail to detect a sample with a low level of gE

172 KIMMAN ET AL. CLIN.DIAGN.LAB.IMMUNOL.

antibodies. This will not have serious implications for eradica-

tion of PRV, provided that the swine population is thoroughly

vaccinated (31).

In conclusion, we have obtained a high level of expression of

a truncated gE in the baculovirus expression system. The ex-

pressed BacgE960 was sufﬁciently antigenic to develop an

IDAS for the detection of antibodies directed against gE in

swine sera. The developed BacgE960 IDAS had a high level of

speciﬁcity and an acceptable level of sensitivity, both of which

make the test a useful complementary or conﬁrmatory test for

the diagnosis of wt PRV infections. The test may be further

improved by circumventing the nonspeciﬁc binding of porcine

sera. For that purpose, we are attempting to develop a recom-

binant gE antigen with a speciﬁc tag.

ACKNOWLEDGMENTS

We thank M. Banks (Central Veterinary Institute, Weybridge,

United Kingdom), H. van der Heijden (Regional Animal Health Ser-

vice Laboratory in the Southern Netherlands), and U. Vecht (ID-

DLO) for providing sera with known disease histories. M. Hulst and J.

Quak were helpful in introducing the baculovirus expression system

and providing technical assistance.

This work was supported by KBN grant 411619101 and a grant from

the Animal Health Service.

REFERENCES

1. Altmann, F., G. Kornfeld, T. Dalik, E. Staudacher, and J. Glossl. 1993.

Processing of asparagine-linked oligosaccharides in insect cells. N-Acetylglu-

cosaminyltransferase I and II activities in cultured lepidopteran cells. Gly-

cobiology 3:619–625.

2. Annelli, J. F., R. B. Morrison, S. M. Goyal, M. E. Bergeland, W. J. Mackey,

and D. G. Thawley. 1991. Pig herds having a single reactor to serum antibody

tests to Aujeszky’s disease. Vet. Rec. 128:49–53.

3. Brunovskis, P., X. Chen, and L. F. Velicer. 1992. Analysis of Marek’s disease

glycoproteins D, I, and E, p. 118–121. Proceedings of the 19th World’s

Poultry Congress. Amsterdam, the Netherlands.

4. Chuma, T., H. Le Blois, J. M. Sanchez-Vizcaino, M. Diaz-Laviada, and P.

Roy. 1992. Expression of the major core antigen VP7 of African horsesick-

ness virus by a recombinant baculovirus and its use as a group-speciﬁc

diagnostic reagent. J. Gen. Virol. 73:925–931.

5. Duffy, S. J., R. B. Morrison, and D. G. Thawley. 1990. Spread of pseudora-

bies virus within quarantined swine breeding herds. JAMA 199:61–65.

6. Enquist, L. W., J. Dubin, M. E. Whealy, and J. P. Card. 1994. Complemen-

tation analysis of pseudorabies virus gE and gI mutants in retinal ganglion

neurotropism. J. Virol. 68:5275–5279.

7. Grace, T. 1962. Establishment of four strains of cells from insect tissues

grown in vitro. Nature (London) 195:788–789.

8. Houwers, D. J., and J. Schaake. 1987. An improved ELISA for the detection

of antibodies to maedi-visna and caprine arthritis-encephalitis virus, employ-

ing monoclonal antibodies in a novel one-step blocking concept. J. Immunol.

Methods 98:151–154.

9. Hulst, M. M., D. F. Westra, G. Wensvoort, and R. J. M. Moormann. 1993.

Glycoprotein E1 of hog cholera virus expressed in insect cell protects swine

from hog cholera. J. Virol. 67:5435–5442.

10. International Organization for Standardization. 1986. International stan-

dard ISO 5725. Precision of test methods—determination of repeatability

and reproducibility for a standard test method by inter-laboratory tests, 2nd

ed. Publication no. UDC 519.248:620.1. International Organization for Stan-

dardization, Lausanne, Switzerland.

11. Jacobs, L., O. de Leeuw, F. Hanssens, and T. G. Kimman. 1995. Complex

between glycoproteins gE and gI of pseudorabies virus is important for

optimal presentation of conformation-dependent immunodominant epi-

topes on gE, p. 132–136. In M. Schwyzer and M. Ackermann (ed.), Immu-

nobiology of viral infections. Proceedings of the 3rd Congress of the Euro-

pean Society of Viral Virology.

12. Jacobs, L., and T. G. Kimman. 1994. Epitope-speciﬁc antibody responses

against glycoprotein E of pseudorabies virus. Clin. Diagn. Lab. Immunol. 1:

500–505.

13. Jacobs, L., R. H. Meloen, A. L. J. Gielkens, and J. T. van Oirschot. 1990.

Epitope analysis of glycoprotein I of pseudorabies virus. J. Gen. Virol. 71:

881–887.

14. Jarvis, D. L., and M. D. Summers. 1989. Glycosylation and secretion of

human tissue plasminogen activator in recombinant baculovirus-infected

insect cells. Mol. Cell. Biol. 9:214–223.

15. Johnson, D. C., and V. Feenstra. 1987. Identiﬁcation of a novel herpes

simplex virus type 1-induced glycoprotein which complexes with gE and

binds immunoglobulin. J. Virol. 61:2208–2216.

16. Katz, J. B., and J. C. Pederson. 1992. Analysis of glycoprotein I (gI) negative

and aberrant pseudorabies viral diagnostic isolates. Am. J. Vet. Res. 53:

2259–2263.

17. Kimman, T. G. 1993. Characterization of the pseudorabies virus-speciﬁc

immunoglobulin M response and evaluation of its diagnostic use in pigs with

preexisting immunity to the virus. J. Clin. Microbiol. 31:2309–2314.

18. King, L., and R. Possee (ed.). 1992. The baculovirus expression system: a

laboratory guide. Chapman and Hall, London.

19. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of

the head of bacteriophage T4. Nature (London) 227:680–685.

20. Lodisch, H. F. 1988. Transport of secretory and membrane glycoproteins

from the rough endoplasmic reticulum to the Golgi. J. Biol. Chem. 263:

2107–2110.

21. Luckow, V., and M. Summers. 1988. Trends in development of baculovirus

expression vectors. Bio/Technology 6:47–55.

22. McFerran, J. B., and C. Dow. 1975. Studies on immunization of pigs with the

Bartha strain of Aujeszky’s disease virus. Res. Vet. Sci. 19:17–22.

23. Mellencamp, M. W., N. E. Pfeiffer, B. T. Suiter, J. R. Harness, and W. H.

Beckenhauer. 1989. Identiﬁcation of pseudorabies virus-exposed swine with

a gI glycoprotein enzyme-linked immunosorbent assay. J. Clin. Microbiol.

27:2208–2213.

24. Mettenleiter, T. C., C. Schreurs, H. J. Thiel, and H.-J. Rziha. 1987. Vari-

ability of pseudorabies virus glycoprotein I expression. Virology 158:141–

146.

25. Miller, L. 1988. Baculovirus as gene expression vectors. Annu. Rev. Micro-

biol. 42:177–199.

26. Pfeffer, S., and J. Rothman. 1987. Biosynthetic protein transport and sorting

by the endoplasmic reticulum and Golgi. Annu. Rev. Biochem. 56:829–

852.

27. Reid-Sanden, F. L., J. W. Sumner, J. S. Smith, M. Fekadu, J. H. Shaddock,

and W. J. Bellini. 1990. Rabies diagnostic reagents prepared from a rabies N

gene recombinant expressed in baculovirus. J. Clin. Microbiol. 28:858–863.

28. Rijsewijk, F. A. M. Unpublished data.

29. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a

laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Har-

bor, N.Y.

30. Stegeman, A., M. C. M. de Jong, H. M. J. F. van der Heijden, A. R. W. Elbers,

and T. G. Kimman. Evaluation of tests for detection of antibodies to pseu-

dorabies virus glycoprotein E in the target population. Unpublished data.

31. Stegeman, A., T. G. Kimman, J. T. van Oirschot, M. M. J. M. Tielen, and

W. A. Hunneman. 1994. Spread of pseudorabies virus within pig herds in an

intensively vaccinated region. Vet. Rec. 134:327–330.

32. Summers, M. D., and E. Smith. 1987. A manual of methods for baculovirus

vectors and insect cell culture procedures. Bulletin no. 1555. Texas Agricul-

tural Experimental Station, College Station, Tex.

33. Urukawa, T., and P. Roy. 1988. Bluetongue virus tubules made in insect cells

by recombinant baculoviruses: expression of the NS1 gene of bluetongue

virus serotype 10. J. Virol. 62:3919–3927.

34. Van der Heijden, H. M. J. F. Unpublished data.

35. Vanderheijden, N., L. De Moerlooze, D. Vandenbergh, G. Chappuis, A.

Renard, and C. Lecomte. 1993. Expression of the bovine viral diarrhoea virus

Osloss p80 protein: its use as ELISA antigen for cattle serum antibody

detection. J. Gen. Virol. 74:1427–1431.

36. Van Oirschot, J. T., F. Daus, T. G. Kimman, and D. van Zaane. 1991.

Antibody response to glycoprotein I in maternally immune pigs exposed to a

mildly virulent strain of pseudorabies virus. Am. J. Vet. Res. 52:1788–1793.

37. Van Oirschot, J. T., A. L. J. Gielkens, R. J. M. Moormann, and A. J. M.

Berns. 1990. Marker vaccines, virus protein-speciﬁc antibody assays and the

control of Aujeszky’s disease. Vet. Microbiol. 23:85–101.

38. Van Zaane, D., and M. M. Hulst. 1987. Monoclonal antibodies against

porcine immunoglobulin isotypes. Vet. Immunol. Immunopathol. 16:23–36.

39. Vaughn, J., R. Goodwin, G. Tompkins, and P. McCawley. 1977. The estab-

lishment of two cell lines from the insect Lepidoptera noctuidae. In Vitro 13:

213–217.

40. Verschoor, E. J., A. L. W. van Vliet, H. F. Egberink, W. Hesselink, M. C.

Horzinek, and A. de Ronde. 1993. Expression of feline immunodeﬁciency

virus gag and env precursor proteins in Spodoptera frugiperda cells and their

use in immunodiagnosis. J. Clin. Microbiol. 31:2350–2355.

41. Vlak, J. M., A. Schouten, M. Usmany, G. J. Belsham, E. C. Klinge-Roode,

A. J. Maule, J. W. M. Van Lent, and D. Zuidema. 1990. Expression of

cauliﬂower mosaic virus gene I using a baculovirus vector based upon the p10

gene and a novel selection method. Virology 179:312–320.

42. Wan, C.-H., M. I. Riley, R. R. Hook, C. L. Franklin, C. L. Besch-Williford,

and L. K. Riley. 1995. Expression of Sendai virus nucleocapsid protein in a

baculovirus expression system and application to diagnostic assays for Sendai

virus infection. J. Clin. Microbiol. 33:2007–2011.

43. Whealy, M. E., J. P. Card, A. K. Robbins, J. R. Dubin, H.-J. Rziha, and L. W.

Enquist. 1993. Speciﬁc pseudorabies virus infection of the rat visual system

requires both gI and gp63 glycoproteins. J. Virol. 67:3786–3797.

VOL. 3, 1996 IDAS FOR PRV gE ANTIBODY DETECTION 173

44. Wilson, M., and P. K. Nakane. 1978. Recent developments in the periodate

method of conjugating horseradish peroxidase (HRPO) to antibodies, p.

215–224. In W. Knap, K. Holubar, and G. Wick (ed.), Immunoﬂuorescence

and related staining techniques. Elsevier/North Holland Biomedical Press,

Amsterdam.

45. Yao, Z., W. Jackson, B. Forghani, and C. Grose. 1993. Varicella-zoster virus

glycoprotein gpI/gpIV receptor: expression, complex formation, and antige-

nicity within the vaccinia virus T7 RNA polymerase transfection system. J.

Virol. 67:305–314.

46. Zuckermann, F. A., T. C. Mettenleiter, C. Schreurs, N. Sugg, and T. Ben-

Porat. 1988. Complex between glycoproteins gI and gp63 of pseudorabies

virus: its effect on virus replication. J. Virol. 62:4622–4626.

174 KIMMAN ET AL. CLIN.DIAGN.LAB.IMMUNOL.