Construction and insect larval expression of recombinant vesicular stomatitis nucleocapsid protein and its use in competitive ELISA

Journal of Virological Methods ELSEVIER

Journal of Virological Methods 54 (1995) 145-157

Construction and insect larval expression of recombinant vesicular stomatitis nucleocapsid protein and its use in competitive ELISA J.B. Katz

*,

Natlonal Veterinary Sercices Laboratories,

A.L. Shafer, K.A. Eernisse US Department

of Agriculture,

P.O. 844, Ames, IA 50010, USA

Accepted 3 April 1995

Abstract The gene encoding the nucleocapsid (N) protein of Indiana 1 serotype vesicular stomatitis virus (VSV-IN,) was transferred into the genome of Autographa cnlifornica nuclear polyhedrosis virus (baculovirus) as a full-length non-fusion construct under the control of the polyhedrin gene promoter. Recombinant N protein was obtained from Trichoplusiu ni insect larvae inoculated 72-96 h previously with the recombinant baculovirus. Polyclonal antibody (PAB) against VSV-IN, was produced in mice using VSV-IN, whole virus antigen concentrated from virus-infected cell culture fluids. The N protein and the PAB were used without further purification in a competitive enzyme-linked immunosorbent assay (C-ELISA) for detection of bovine, porcine, and equine origin serum antibodies against VSV-IN,. A limited number of field origin, experimental, and reference VSV antisera were evaluated using the C-ELISA and with a standard serum neutralization (SN) procedure. Sensitivity of the C-ELISA was comparable to the serotypically homologous SN procedure. Subject to further validation, similar C-ELISA tests for the other VSV serotypes, used in conjunction with the test described here, may offer the best combination of rapidity, sensitivity, simplicity, economy, and laboratory biosafety of any of the methods yet developed for VSV serodiagnosis. Keywords: Vesicular munosorbent assay

* Corresponding

stomatitis

virus;

Nucleocapsid

protem;

Baculovirus;

author: Fax: + 1 (515) 239 8348

0166-0934/95/$09.50 0 1995 Elsevier Science B.V. All rights reserved SSDI 0166-0934(95)00036-4

Competitive

enzyme-linked

im-

146

J.B. Katz et al. /Journal

of Virologtcal Methods 54 (1995) 145-157

1. Introduction Vesicular stomatitis virus (VSV) is the prototypic member of the genus vesiculovirus, family Rhabdoviridae, and is responsible for enzootic and occasionally epizootic disease in cattle, swine, and horses (Gibbs, 1993). Human beings are also readily susceptible to VSV infection with the production of influenza-like illness (Gibbs, 1993). In cattle, swine, and horses, VSV infections cause severe oral and pedal erosive and ulcerative lesions, resulting in considerable economic damage. These lesions are easily mistaken for those produced during etiologically unrelated viral diseases, including foot-and-mouth disease, swine vesicular disease, and other exanthemous viral livestock diseases (Gibbs, 1993). Vesicular stomatitis virus occurs only in the western hemisphere, and its rapid differential serodiagnosis has been a long-sought goal of animal health authorities in that part of the world (Ferris and Donaldson, 1988; Alonso et al., 1991; DeAnda et al., 1992; Afshar et al., 1993). Serodiagnosis is complicated by the limited antigenic crossreactivity between the two serotypes of VSV, designated VSV Indiana (VSV-IN) and VSV New Jersey (VSV-NJ) (Gelata and Holbrook, 1966; Kang and Prevec, 1970; Cartwright and Brown, 1972; Charan et al., 1987). Serum neutralization tests are used to further differentiate VSV-IN isolates into 3 subtypes: VSV-IN, (classic subtype) and the much less frequently encountered South American subtypes VSV-IN, (Coca1 and Argentina subtype) and VSV-IN, (Brazil subtype) (Gelata and Holbrook, 1966; Federer et al., 1967; Emerson, 1976). Complement-fixation (CF) tests, serum neutralization (SN) tests, and enzyme-linked immunosorbent assay (ELlSA) are used for VSV serodiagnosis (Ferris and Donaldson, 1988; Alonso et al., 1991; DeAnda et al., 1992). The ELISA is more sensitive than the CF test and as sensitive as the SN test. Additionally, the ELISA is more rapid than the CF and SN tests and is less dependent upon the quality of serum specimens (Ferris and Donaldson, 1988; Alonso et al., 1991; DeAnda et al., 1992). Two independent recent technical advances were recently reported in attempts to further enhance the VSV ELISA serodiagnostic approach. A competitive ELISA was developed (Afshar et al., 1993), and this obviated the need for the multiple, speciesspecific, anti-immunoglobulin enzyme conjugates required for the indirect ELISA. Separately, recombinant DNA techniques were used to generate a serologically useful VSV-NJ nucleocapsid (N) protein antigen (Ahmad et al., 1993). Use of this protein could help reduce the exposure of laboratory workers to a zoonotic disease hazard in the course of preparing and using infectious whole virus antigen. The present study describes a novel combination of both the recombinant DNA and the C-ELISA strategies for the development of a recombinant antigen-based C-ELISA for VSV-IN, serodiagnosis.

2. Materials

and methods

2.1. VSV antisera, polyclonal

anti-VW-IN1

mouse ascites, and SN procedure

Serotype specific VSV-IN, and VSV-NJ antisera were produced in gnotobiotic or conventionally reared ponies using previously described methods (Charan

calves et al.,

J.B. Katz et al./Journal

of Virological Methods 54 (1995) 145-157

147

1987; Alonso et al., 1991). Specific hyperimmune antisera against VSV-IN, and VSV-IN, (SN endpoint titers = 750) were provided by the US Department of Agriculture, Foreign Animal Disease Diagnostic Laboratory, Orient Point, New York. Field origin convalescent sera from cattle previously infected with VSV-NJ or VSV-IN were obtained from The School of Veterinary Medicine, Tropical Disease Program, Heredia, Costa Rica. A panel of sera was also obtained using young swine intradermally inoculated with 10’ ‘-lo6 ’ median cell culture infective doses (CCID,,) of VSV-NJ or VSV-IN,. Sera were collected from those swine daily for 11 days postinoculation (DPI). The SN endpoint titers of all sera were measured twice using a previously described microplate cell culture serum neutralization assay (Afshar et al., 1993). The cell cultures used in these tests were observed for cytopathic effect (CPE) 48 h after inoculation with the indicator virus-serum mixtures. Serum antibody titers were then calculated as reciprocals of the greatest dilutions of sera preventing CPE. Mouse ascites fluid containing anti-VSV-IN, antibodies was produced in 6-week-old mice by intraperitoneal (i.p.) injection of VSV-IN, mixed with adjuvant as previously detailed (Overkamp et al., 1988; Afshar et al., 1993). Subsequent i.p. injections of viral antigen combined with lo3 180/TG sarcoma cells resulted in VSV-IN, hyperimmune ascites fluids that were pooled, stored at -20°C and titrated for antibody using the SN procedure. 2.2. Construction

of recombinant

baculovirus

A molecularly cloned VSV-IN, N protein encoding gene (Gallione et al., 1981; Bannerjee et al., 1984) was obtained in a pBR322 plasmid vector (Sprague et al., 1983). The complete gene and attached non-coding flanking sequences were excised by Xhol digestion followed by gel purification. This fragment was ligated (Sambrook et al., 1989), both unmodified and following limited exonucleolytic digestion, into the previously restricted and dephosphorylated Smal site of the polyhedrin promoter-based baculovirus transfer plasmid pVL1393 (Fig. 1). Exonucleolytic treatment (Henikoff, 1984) (Fig. 1) was designed to remove most of the 5’-non-coding sequence preceding the N protein initiation codon. This was intended to possibly maximize the expression of the N gene (Luckow and Summers, 1988; Webb and Summers, 1990) by minimizing the length of extraneous nucleic acid intervening between the polyhedrin promoter and the N gene initiation site. Plasmids resulting from infection of both the unmodified (~5-10) and the trimmed N gene (~4-7) segment were sequenced @anger et al., 1977) to confirm the correct sequences and the correct junctional orientations of the inserted genes (Figs. 1 and 2). Linearized and Bsul 36 restricted (Kitts et al., 1990) Autogrupha californica nuclear polyhedrosis virus (baculovirus) DNA (Baculogold, Pharrnigen, Inc., San Diego, CA) was cotransfected (0.5 pug) in separate mixtures containing 5.0 pg of either ~4-7 or ~5-10 and 30 pg of lipid microsomal transfection reagent (DOTAP, Boehringer Mannheim, Inc., Indianapolis, IN). Spodopteru frugiperdu (SF-9) 25 cm2 cell monolayers were used for transfection (Kitts et al., 1990; Bremer et al., 1994; O’Reilly et al., 1994). Recombinant baculoviruses were plaque purified (Kitts et al., 1990; Bremer et al., 1994; O’Reilly et al., 1994), and N gene insertion was confirmed using a polymerase chain reaction (PCR) screening procedure followed by nucleic acid sequencing of the

J.B. Katz et al. /Journal

148

of Virologxal Methods 54 (1995) 145-157

Bsu I 36 Digested

pVL 1393

pJS 223

Lig. Fig. 1. Strategy used in the construction of baculoviruses expressing the VSV-IND, N protein. The N protein gene was excised from pJS223 by XhoI digestion. That fragment was blunt-ended both with and wtthout a preceding exposure to exonuclease III (Exe III) and Sl nuclease (Sl NW.). These fragments were ligated into pVL1393, previously cleaved at the SmaI cloning site. The resulting ~4-7 and ~5-10 plasmids were contransfected with BspI 36 digested baculovirus DNA. Plasmid rescue resulted in two viruses expressing the N protein, BV-N4-7 and BV-NS-10. BAP, bacterial alkaline phosphatase; T1 Lig., Td DNA ligase: Klenow, Klenow (large) fragment of Escherichra coli DNA polymerase.

insertional junctions using the same pair of oligonucleotide primers. These primers flanked the PVL1393 cloning site. The S-primer was a 24-mer (S-TITACTGTITTCGTAACAGTTITG-3’1, and the 3’-primer was a 21-mer 6’CGGATTTCCTTGAAGAGAGTG-3’) separated from the S-partner by 182 nucleotides in the absence of a gene insert. The insertion of the two approximately 1.3 kb N gene-containing fragments from ~4-7 and ~5-10 thus generated two recombinant baculoviruses, BV-N4-7 and BV-NS10 (Figs. 1 and 2).

*

* 4-l 5 - 10

*

CCTATAAATATTCCG T I, q ” ” I #I. I, ,I II” ” II ” Y ” II 0 I II0, ,I I, ” ” ,I II II 0, I ” ,I U I * ,I . I I, l I, ” ,I ” I, Y I, I, I, ,I ” q I, II ” II IIII I II II I, I, II II Y . * II II IIY II” AGTTTTGTAAT AAAAAAACCTATAAATATTCCG T a

AGTTTTGTAAT-

TTCATACCGTCCCAC II II II ,I I ,I YI . H #III q I II II Y* (4II * w 11,I I ” )I II . Y II &I TTCATACCGTCCCAC

SMA I l

4-l

,,,~,,~,"~~-~~---_--__--__-__________~_ 0," " I,* . I II q " q II IIIIII*

5-10

CGCGGATCTCGAGGTCAGGAGAAACTTTAACAGTAATC

l

Fig. 2. Nucleotide sequences of ~4-7 and ~5-10 at the 5’-junction between the clomng site and the ATG initiation codon of the N gene. The dashed line indicates sequence removed by exonuclease. The N gene of ~4-7 is 28 nucleotides closer than ~5-10 to the natural translation initiation site (+ 11 of non-recombinant baculovirus. The natural polyhedrin gene ATG site at + 1 has been mutated to ATI in pVL1393 to allow translation to begin at the first subsequent ATG.

J.B. Katz et al. /Journal

2.3. Production

of Vwologrcal Methods 54 (1995) 145-157

and analysis of recombinant

149

VSV-IN, N protein

Baculoviruses BV-N4-7 and BV-NS-10 were used to infect SF-9 cells seeded into 96-well plates or 25 cm2 flasks (multiplicity of infection = 0.1). After 24-, 48-, 72-, and 96-h incubations at 27°C the plate monolayers were fixed with an 80/20% cold acetone-water mixture. Total cell pellets (Bremer et al., 1994) were harvested from the flasks and frozen for Western blot analysis. An immunoperoxidase staining monolayer assay (IPMA) (O’Reilly et al., 1994) was carried out using anti-VSV-IN, gnotobiotic calf antiserum and horseradish peroxidase conjugated protein G (Zymed, Inc., San Francisco, CA). Techniques reported for Western blotting (Laemmli, 1970; Towbin et al., 1979) were also employed using the same antiserum to detect and confirm the molecular mass of the expression product. Trichoplusia ni insect eggs were obtained 24-48 h after oviposition. The eggs were hatched and larvae allowed to develop at 27°C on artificial media until they weighed 120-150 mg (8-9 days post-hatching) (Medin et al., 1990). At that time, 5 ~1 of sterile saline solution containing lo4 3 CCID,, of either BV-N4-7 or BV-NS-10 was injected subcuticularly into each larvae (Medin et al., 1990; O’Reilly et al., 1994). When larvae became pale, swollen, and lethargic (72-96 h), they were frozen, pooled, and homogenized (10% w/v) for 20 s (Polytron homogenizer, Brinkmann, Inc., Westbury, NY) in a solution of 25 mM Tris buffer, pH 7.4, 5 mM dithiothreitol, 2 mM EDTA, 0.001% Triton X-100, and 1 mM Pefabloc, a serine protease inhibitor (Boehringer Mannheim, Inc., Indianapolis, IN). Total protein concentrations of the homogenates were determined (Biorad, Inc., Richmond, CA), and the homogenates were then diluted to 1.1 mg/ml total protein in pH 7.2 phosphate-buffered saline (PBS) containing 0.01% Triton X-100. This material was then used directly as the C-ELISA antigen or frozen ( - 70°C) for later use. 2.4. C-ELISA for VSV-INI antibody detection The C-ELISA antigen (50 pi/well) was coated onto 96-well plates (Polysorp plates, Nunc, Inc., Naperville, IL) for at least 16 h at 4°C. Antigen was then decanted and wells blocked at 4°C for 2 h with PBS, pH 7.2, containing 5% (w/v) non-fat dry milk. Wells were then rinsed 3 times with PBS containing 0.05% Tween 20. Unknown and control sera were then placed (50 PI/ well), each in duplicate wells, for 30 min at 37°C. Sera were previously diluted 4-fold in PBS containing 0.01% Triton X-100 and 1% non-fat dry milk (serum diluent solution). Without removing test or control sera, an aliquot (50 ~1) of the mouse anti-VSV-IN, PAB was added to each well. The PAB had been freshly diluted 2400-fold in serum diluent solution. After another 30 min incubation at 37°C wells were washed 3 times with PBS containing Tween 20. A goat-origin anti-mouse immunoglobulin peroxidase antibody conjugate (Zymed, Inc., San Francisco, CA) diluted 1: 500 (v/v) in serum diluent was added (50 pl/ well) for 30 min at room temperature. Wells were rinsed again and then filled with 50 pi/well of peroxidase substrate (two-component ABTS-hydrogen peroxide substrate system, Kirkegaard and Perry, Inc., Gaithersburg, MD). Color development at 405 nm (OD,,) was monitored, and the OD,,, of all wells was recorded when negative serum control wells displayed an OD,,, between 0.4 and 0.6. The percentage inhibitions in OD,,, of test serum wells

J.B. Katz et al./Journal

150

of Virologlcal Methods 54 (1995) 145-157

relative to duplicate negative serum control wells were calculated inhibition = [l - (OD,, test well/OD,,, control wells)] X 100).

(i.e.,

percentage

3. Results 3.1. Construction

and analysis of recombinant

baculooiruses

The pVL1393 plasmid was designed as a non-fusion protein expression transfer vector (Luckow and Summers, 1988; O’Reilly, 1994). This plasmid has a SmaI cloning site 29 nucleotides downstream from the usual polyhedrin gene initiation codon. In pVL1393, this codon is mutated (Fig. 2) to allow translation to begin at the first ATG codon inserted after the SmaI site. Two recombinant baculoviruses, differing in the length of the S-untranslated sequence preceding this initiation codon, but both containing the complete VSV-IND, N gene, were derived in this study. The recombinant virus BV-NS-10 possessed the entire XhoI fragment containing the VSV-IND, N gene. This virus had 34 untranslated nucleotides between the cloning site and the N protein initiation codon. Nucleotide sequencing (Fig. 2) confirmed that 31 of these 34 untranslated nucleotides had been removed by exonucleolytic digestion of the same XhoI fragment during construction of the recombinant virus BV-N-47. Both viruses, however, directed the abundant expression of VSV-IN, immunoreactive protein upon infection of SF-9 cell cultures, as visualized by the IPMA procedure (Fig. 3). Western blotting (Fig. 4) of SF-9 cells or insect larvae infected with either virus revealed an intensely immunoreactive band of approximately 47 kDa, consistent with the reported calculated molecular mass of the N protein of 47,355 Da (Gallione et al., 1981; Bannerjee et al., 1984). Quantitative protein expression studies were not carried out, but IPMA and Western blot time course studies did not reveal overt differences between the viruses in recombinant protein accumulation after 48 h of infection in cell culture or 72 h of infection in larvae (data not shown). Two pools of 20 randomly selected larvae each were infected with 104j CCID,, of BV-N4-7 or BV-NS-10 and harvested 72 h later. Homogenates of each pool (1 .l mg/ ml total protein) were evaluated by C-ELISA using

A

B

C

Fig. 3. Immunoperoxidase monolayer assay (IPhL4) of recombinant VSV-IND, N protein expressed in SF-9 cells. Gnotobiotic calf serum primary antibody against VSV-IND, was followed by a protein G-horseradish peroxidase conjugate and aminoethylcarbazole substrate. A: X 110. B: same as A except BV-NS-10 was used, X 220. C: uninfected SF-9 cell monolayer, X 220.

J.B. Katz et al./Journal

of Virological Methods 54 (1995) 145-157

151

Fig. 4. Western blot of N protein expressed by larvae infected with BV-N4-7. Lane 1, prestained ovalbumin reference marker, 43 kDa; lane 2, additional prestained protein mass markers of indicated size; lane 3, uninfected T. ni larval homogenate; lane 4, homogenate of larvae inoculated 96 h previously with lo4 3 CUD,, of BV-N4-7. Blot was developed using gnotobiotic calf VSV-IN, antiserum (1:500), followed by protein G-HRP conjugate (1: 500), and a precipitable tetramethylbenzidine @MB) substrate (Kirkegaard and Perry, Gaithersburg, MD). Immunoreactive band in lane 4 is consistent with calculated 47 kDa mass of VSV-IN, N protein.

constant dilutions of several reference VSV-IN, antisera competing against serially increasing dilutions of the PAB (Fig. 5). The antibody competition in wells coated with BV-NS-10 homogenates was slightly, but obviously, more stringent than with those coated with BV-N4-7 homogenates (Fig. 5, rows H3-E14). These differences were noted with subsequent batches of C-ELISA antigen and may have resulted from slightly less N protein expression during larval infections with the BV-NS-10 virus. 3.2. C-ELBA

results

Titrations of recombinant VSV-IN, N protein antigen, PAB, and reference VSV-IN, antisera from cattle, swine, and horses (Fig. 5) were used to adjust antibody competition to achieve results that were diagnostically useful. Strong positive (SP, SN = 512), weak positive (WP, SN titer = 8), and VSV seronegative control sera were used to calibrate the percentages of OD,, inhibition and to evaluate well-to-well reproducibility. Negative, WI’, and SP sera were each placed separately into 6 scattered wells on each of two plates. The means and standard errors 6 f S.E.M.) in percentages OD,,, inhibition for the SP and WP control sera were 94.6 f 1.0% (n = 12) and 66.4 f 4.0% (n = 12), respectively. In that test, the negative serum control OD,, values were 0.41 f 0.01 (? f S.E.M., n = 12). Using results with those controls as cut-off points, sera were defined as VSV-IN, seronegative if OD,,, percentage inhibition was < 50%, as

J B. Katz et al. /Journal

152

ofVirologmzl Methods 54 (1995) 145-157

Hl H2 H3 H4 E9

VSV “N” Clone 5-10

El0 El4

Hl H2 H3 H4

VSV “N” Clone 4-7

E9 El0 El4

Fig. 5. VW-IN, C-ELISA using unpurified homogenates (1.1 mg/ml) of larvae inoculated with BV-NS-10 or BV-N4-7. For each plate, the bottom row of wells received only serially diluted mouse anti-VSV-IN, ascitic fluid as indicated (1:200-l : 6,400). All other rows received the same dilutions of ascitic fluid plus one of several VW-IN, reference antisera in constant 1: 5 dilutions. Rows Hl and H2, VSV-IN, hyperimmune cattle sera; rows H3 and H4, hyperimmune swine sera; rows E9, ElO, and E14, VSV-IN, hyperimmune pony sera. Competition by murine immunoglobulins was detected using mouse-specific antibody conjugate as described in text. Greater optical densities in wells of the 4-7 plate (e.g. rows H3 and E9) relative to corresponding wells of the 5-10 plate suggest less competition (i.e., more antigen) in wells coated with BV-N4-7 larval homogenate. This difference was modest, but consistently noted with equivalently diluted BV-N4-7 and BV-NS-10 larval homogenates.

indeterminate if the percentage inhibition was 50-66%, and as positive if the percentage inhibition was > 66%. Endpoint estimates of C-ELISA seropositivity were thus defined as the greatest dilutions of sera producing > 66% inhibition in OD,,,. Comparison of SN and C-ELISA endpoint titers using reference hyperimmune VSV (Table 1) antisera revealed that they were virtually identical using the VSV-IN, antisera, but that the C-ELISA titers of the VSV-NJ hyperimmune reference antisera substantially exceeded the SN-IN, titers of those sera. In contrast, the SN-IN, and C-ELISA titers of sera from Costa Rican cattle naturally infected with VSV-NJ were comparable. The SN-NJ, SN-IN,, and C-ELISA titers of the VSV-IN, and VSV-IN, reference hyperimmune sera were all < 16, markedly less than the reported endpoint SN titers when using SN indicator viruses of the homologous subtypes (neutralization titer of VSV-IN, serum = 750, and of VSV-IN, serum = 750). Evaluation of sera obtained sequentially from swine inoculated with either VSV-IN, or VSV-NJ revealed a pattern similar to that seen with the sera from Costa Rican cattle:

J.B. Katz et al. /Journal Table 1 Evaluation

of Virological Methods 54 (1995) 145-157

of bovine origin VSV antisera using the C-ELBA

Serum identification Field origrn sera b 49 52 56 57 54 58 Hyperimmune reference sera ’ 8201 8401 8601 8901 8602 8501

SN-NJ

and SN procedures SN-IN,

153

a C-ELBA

32 32 32 <8

128 > 512 128 32

>512 > 512 > 512 128

128 > 512

32 8

32 <8

<8 <8 <8 <8

> 8192 512 > 8192 > 8192

> 8192 512 > 8192 2048

> 8192 > 8192

<8 <8

2048 512

a Titers are expressed as the highest serum dilution yielding a positive test result by SN or C-ELBA. SN-IND, and SN-NJ were serum neutralization tests using 100 CCID,, OF Indiana, and New Jersey VSV indicator viruses, respectively, while C-ELISA was the competitive enzyme immunoassay utilizing insect larval origin recombinant Indiana, serotype viral N protein. b Sera were from field cases of natural VSV infection in Costa Rican cattle. Sera 49, 52, 56, and 57 were from cattle diagnosed infected with VSV-IN,; sera 54 and 58 were from cattle diagnosed infected with VSV-NJ. ’ Sera 8201, 8401, 8601, and 8901 were from gnotobiotic calves inoculated initially with inactivated VSV-IN,, then repeatedly with virulent VSV-IN,. Sera 8602 and 8501 were from similarly reared calves exposed to inactivated VSV-NJ followed by repeated inoculations with virulent VSV-NJ.

C-ELISA and SN-IN, titers were usually comparable to each other, but generally differed from the SN-NJ titers of the same sera. Antibody responses were detected by 6 days postinoculation in most of the swine using the serotypically appropriate procedure; the C-ELISA and the SN-IN, performed similarly in this regard.

4. Discussion The molecularly cloned gene (Sprague et al., 1983) encoding VSV-IN, nucleocapsid (N) protein was expressed as a full-length non-fusion protein in recombinant baculovirus-infected insect cells and larvae. Two virus constructs were made with the N gene, which was inserted either 33 or 61 nucleotides downstream from what would have been the initiation codon of the polyhedrin gene (O’Reilly et al., 1994). Accumulated experience with the baculovirns expression system suggests that removal or reduction in length of inserted non-translated S-sequences preceding the recombinant gene initiation codon may help maximize protein production (Luckow and Summers, 1988; Webb and Summers, 1990; O’Reilly et al., 1994). Consistent with this concept, modest differences in concentrations of recombinant N protein were observed by C-ELISA using homogenates of BV-N4-7and BV-NS-lo-infected larvae. Protein expression is a func-

154

J.B. Katz et al. /Journal

of Virological Methods 54 (1995) 145-157

tion of many complex transcriptional and translational control factors, and the differences between BV-N4-7 and BV-NS-10 in the lengths of their inserted 5’-untranslated sequences may not have substantially altered the level of N protein expression. Trichoplusia ni larvae were an easily reared, inexpensive, and easily processed source of antigen. The larval homogenates functioned reproducibly without further purification of antigen. Mouse polyclonal ascitic fluids were easily prepared and used without extensive purification (Overkamp et al., 1988). The potential advantages of polyclonal antibodies as indicator reagents in C-ELISA serodiagnostic methods have been previously discussed (Afshar et al., 1993) and include, most importantly, their ability to compete with test serum antibodies over a range of immunoglobulin classes, affinities, and avidities for multiple epitopes on the target antigen. The antigen used for ascites production was concentrated from Vero cell cultures infected with VSV-IN, and not the insect origin N protein homogenate. Although ascites fluid antibodies to Vero (mammalian) cell antigens were likely produced, they were not crossreactive with the insect cell antigens found in the N protein-containing larval homogenates used to coat C-ELISA plates. The strategy of using antigenically heterologous systems to produce the recombinant antigen and the PAB obviated the need to purify either reagent and is an important practical advantage of this approach. Both the C-ELISA and the SN-IN, procedures detected antibody in VSV-IN, infected swine as early as 5-8 days postinoculation. Using sera from convalescent field-origin animals and hyperimmune serum donor animals, the C-ELISA and the SN-IN, performed with comparable endpoint sensitivity with respect to VSV-IN, serodiagnosis. However, the limited numbers of animals and their widely variable individual serologic responses precluded rigorous statistical comparisons between the SN and C-ELISA test methods. Sera from naturally exposed or experimental animals infected with either live VSV-IN, or VSV-NJ were typically less serotype-specific than those of the animals used to produce highly specific reference neutralizing VSV-IN, and VSV-NJ antisera (Gelata and Holbrook, 1966; Charan et al., 1987; Alonso et al., 1991). This difference in the degree of serotypic specificity results from the non-natural administration routes, virus dosages, and virus antigen adjuvantation and inactivation methods used to produce the hyperimmune typing sera. Serotype-specific serum neutralizing antibodies to VSV-IN, and VSV-NJ are directed primarily against the surface glycoprotein (G protein) and not against the internal N protein of those viruses (Kelley et al., 1972; Emerson, 1976; Yilma et al., 1985). The N proteins of these viruses, however, were the antigens against which serologic responses were detected by the C-ELISA. The N proteins of VSV-IN, and VSV-NJ are relatively more conserved than their surface glycoproteins (Emerson, 1976; Gallione et al., 1981; Bannerjee et al., 1984; Yilma et al., 1985). The nucleotide sequences of the prototypic VSV-NJ and VSV-IN, N genes are 68% homologous; their putative identical amino acid sequence homology is 69%, and nearly 80% if conservative amino acid substitutions are considered (Gallione et al., 1981; Bannerjee et al., 1984). Thus, it was not surprising that C-ELISA titers tended to be less serotype-specific than titers measured by the SN test when sera from VSV-NJ infected animals were evaluated (e.g. 8602 and 8501, Table 1; 733, Table 2). The C-ELISA reported here, being N protein based, does not replace the SN procedure for serotyping purposes (Federer et al., 1967; Kang and Prevec, 1970; Cartwright and

J.B. Katz et al. /Journal Table 2 Temporal C-ELBA

serum antibody responses and SNT procedures

of Virological Methods 54 (1995) 145-157

of swine infected

with vesicular

stomatitis

viruses

1.55

as detected

Animal a

Test procedure

o-4

5

6

726

SNT-IND, SNT-NJ C-ELBA

<8 <8 <8

<8 8 8

32 32 128

128 128 128

>512 128 > 512

735

SNT-IND, SNT-NJ C-ELISA

<8 <8 <8

8 <8 <8

64 <8 32

2 512 16 32

> 512 32 > 512

741

SNTLIND, SNT-NJ C-ELISA

<8 <8 <8

8 <8 <8

16 8 8

> 512 128 >, 512

> 512 32 2 512

745

SNT-IND, SNT-NJ

<8 <8

<8 <8

C-ELISA

<8

<8

<8 <8 <8

128 <8 32

128 <8 128

SNT-IND,

<8

SNT-NJ C-ELISA

<8 <8

<8 <8 <8

<8 16 <8

<8 64 <8

<8 256 <8

SNT-IND,

<8 <8 <8

<8 <8 <8

<8 128 <8

<8 256 32

<8 > 512 128

729

733

SNT-NJ C-ELISA

Days following

a Animals 726, 735, 741, and 745 were inoculated were inoculated with VSV (New Jersey serotype).

by the

virus inoculation 8

with VSV (Indiana,

serotype)

10-11

and animals

729 and 733

156

J.B. Katz et al. /Journal

of Vtrologrcal Methods 54 (1995) 145-157

VSV-IN, and VSV-NJ N proteins would minimize exposure to VSV zoonotic hazards in the laboratory. Several other indirect ELISA and C-ELISA methods have been reported for VSV serodiagnosis (Ferris and Donaldson, 1988; Alonso et al., 1991; DeAnda et al., 1992; Afshar et al., 1993). Unlike the present approach, all of those tests have required the use of live virus in the course of preparing natural viral antigens as test components. The indirect ELISA methods also required multiple species-specific anti-immunoglobulin conjugate reagents, a need obviated by the C-ELISA. VSV ELISA methods reported previously have been comparable or superior to their SN homologs in sensitivity. In this study, the C-ELBA was comparable in sensitivity to the VSV-IN, SN test. A larger, diagnostic survey examined statistically will be needed to validate the recombinant protein VSV-IN, C-ELISA as a replacement for the current equivalent SN-IN, serodiagnostic method. However, the combination of recombinant antigen production and C-ELISA methodology may offer the most promising approach to date for rapid, safe, and convenient serodiagnosis of this important disease.

Acknowledgements We thank Drs. J. House and M. Schubert for VSV-IND, and VSV-IND, antisera and plasmid pJS223, respectively. The cooperation of Dr. L. Rodriguez in providing serum samples from Costa Rica is gratefully acknowledged.

References Afshar, A., Shakarchi, N.H. and Dulac, G.C. (1993) Development of a competitive enzyme-linked immunosorbent assay for detection of bovine, ovine, porcine, and equine antibodies to vesicular stomatitis virus. J. Clin. Microbial. 31, 1860-1865. Ahmad, S., Bassiri, M., Banergee, A.K. and Yilma, T. (19931 Immunological characterization of the VSV nucleocapsid (Nl protein expressed by recombinant baculovirus in Spodoptera exigua larva: use in differential diagnosis between vaccinated and infected animals. J. Virol. 192, 207-216. Alonso, A., Martins, M.A., Games, M., Allende, R. and Sondahl, M.S. (1991) Development of and evaluation of an enzyme-linked immunosorbent assay for detection, typing, and subtyping of vesicular stomatitis virus. J. Vet. Diagn. Invest. 3, 287-292. Bannerjee, A.K., Rhodes, D.P. and Gill, D.S. (1984) Complete nucleotide sequence of the mRNA coding for the N protein of vesicular stomatitis virus (New Jersey serotype). Virology 137. 432-438. Bremer, C.W., du Pleiss, D.H. and van Dijk, A.A. (19941 Baculovirus expression of non-structural protein NS2 and core protein VP7 of African horsesickness virus serotype 3 and their use as antigens in an indirect ELISA. J. Virol. Methods 48, 245-256. Cartwright, B. and Brown, F. (1972) Serological relationships between different strains of vesicular stomatitis virus. J. Gen. Virol. 16, 391-398. Charan, S., Hengartner, H. and Zinkernagel, R.M. (19871 Antibodies against two serotypes of vesicular stomatitis virus measured by enzyme-linked immunosorbent assay: immunodominance of serotype-specific determinants and induction of asymmetrically cross-reactive antibodies. J. Virol. 61, 2509- 2514. DeAnda, J.H., Salman, M.D., Webb, P.A., Keefe, T.J., Arevalo, A.A. and Mason, J. (19921 Evaluation of an enzyme-linked immunosorbent assay for detection of antibodies to vesicular stomatitis virus in cattle in an enzootic region of Mexico. Am. J. Vet. Res. 53, 440-443. Emerson, S.V. (1976) Vesicular stomatitis virus: structure and function of virion components. In: W. Arber (Ed.), Current Topics in Microbiology and Immunology, Vol. 73, Springer-Verlag, Berlin, pp. l-34.

J.B. Katz et al./Journal

of Virological Methods 54 (1995) 145-157

157

Federer, K.E., Burrows, R. and Brooksby, J.B. (19671 Vesicular stomatitis virus - the relationship between some strains of the Indiana serotype. Res. Vet. Sci. 8, 103-117. Ferris, N.P. and Donaldson, AI. (1988) An enzyme-linked immunosorbent assay for the detection of vesicular stomatitis virus antigen. Vet. Microbial. 18, 243-258. Gallione, C.J., Greene, J.R., Iverson, L.E. and Rose, J.K. (19811 Nucleotide sequence of the mRNA’s encoding the vesicular stomatitis virus N and NS proteins. J. Virol. 39, 529-535. Gelata, J.N. and Holbrook, A.A. (19661 Vesicular stomatitis: patterns of complement-fixing and serum-neutralizing antibodies in serum of convalescent cattle and horses. Am. J. Vet. Res. 22, 713-719. Gibbs, E.P.J. (1993) Rhabdoviridae. In: F.J. Fenner, E.P.J. Gibbs, F.A. Murphy, R. Rott, M.J. Studdert and D.O. White (Eds.), Veterinary Virology, 2nd edn. Academic Press, San Diego, CA, pp. 489-510. Henikoff, S. (1984) Unidirectional digestion with exonuclease III creates targeted breakpoints for DNA sequencing. Gene 28, 351-359. Kang, C.Y. and Prevec, L. (1970) Proteins of vesicular stomatitis virus: II. Immunological comparisons of viral antigens. J. Virol. 6, 20-27. Kelley, J.M., Emerson, S.V. and Wagner, R.R. (1972) The glycoprotein of vesicular stomatitis virus is the antigen that gives rise to and reacts with neutralizing antibody. J. Virol. 10, 1231-1235. Kitts, P.A., Ayres, M.D. and Possee, R.D. (1990) Linearization of baculovirus DNA enhances the recovery of recombinant virus expression vectors. Nucleic Acids Res. 18, 5667-5672. Laemmli, V.K. (1970) Cleavage of structural proteins during assembly of the head of bacteriophage T4. Nature 227, 680-685. Luckow, V.A. and Summers, M.D. (1988) Trends in the development of baculovirus expression vectors. Biotechnology 6, 47-55. Medin, J.A, Hunt, L., Gathy, K., Evans, R.K. and Coleman, M.S. (1990) Efficient, low cost protein factories: expression of human adenosine deaminase in baculovirus-infected larvae. Proc. Natl. Acad. Sci. USA 87, 2760-2764. O’Reilly, D.R., Miller, L.K. and Luckow, V.A. (1994) Baculovirus Expression Vectors: A Laboratory Manual. Oxford University Press, Oxford. Overkamp, D.S., Mohammed-Ali, S., Cartledge, C. and Lando, J. (1988) Production of polyclonal antibodies in ascitic fluid of mice: technique and applications. J. Immunoassay 9, 51-68. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sanger, F., Nicklen, F. and Coulson, A.R. (19771 DNA sequencing with chain terminating inhibitors. Proc. Natl. Acad. Sci. USA 74, 5463-5467. Sprague, J., Condra, J.H., Amheiter, H. and Lazzarini, R.A. (19831 Expression of recombinant DNA gene coding for the vesicular stomatitis virus nucleocapsid protein. J. Virol. 45, 773-781. Towbin, H.T., Stahelin, T. and Gordon, J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA 76, 4350-4354. Webb, N.R. and Summers, M.D. (1990) Expression of proteins using recombinant baculoviruses. Technique 2, 173-178. Yilma, T., Breeze, R.G., Ristow, S., Gorham, J. and Leib, S.R. (1985) Immune responses of cattle and mice to the G glycoprotein of vesicular stomatitis virus. Adv. Exp. Med. Biol. 150, 101-115.