|Home | About | Journals | Submit | Contact Us | Français|
Glycoprotein 5 (GP5) of porcine reproductive and respiratory syndrome virus (PRRSV) has been studied extensively as a target for vaccine development. This study evaluated the serodiagnostic application of PRRSV GP5 by enzyme-linked immunosorbent assay (ELISA). Two immunodominant peptides (VR #1 and VR #2) and two neutralizing ectodomain-containing peptides (Ecto #1 and Ecto #2), as well as recombinant GP5 (rGP5) as a control, were prepared. Serum from unvaccinated pigs was screened for the antibodies that bind to these peptide and protein antigens. The results were compared with those from a commercially available diagnostic ELISA kit (HerdChek), which uses the nucleocapsid (N) protein as an antigen. Only VR #1+#2 showed a result statistically similar to that of N protein. Ecto #1 and Ecto #2 had a lower sensitivity than VR #1+#2 and rGP5. The peptides and rGP5 showed significant associations with the N protein (P < 0.05 or 0.01), which suggests that GP5 may also be a candidate serodiagnostic antigen. Since antibodies against GP5 persist much longer than those against the N protein, GP5 itself and some of its fragments are thought to be good targets for serodiagnosis. In addition, the presence of antibodies against the PRRSV structural antigens showed significant antigen-dependent differences.
La glycoprotéine 5 (GP5) du virus du syndrome reproducteur et respiratoire porcin (PRRSV) a été étudiée de manière exhaustive à titre de cible pour le développement d’un vaccin. La présente étude a évalué la GP5 du PRRSV pour une application séro-diagnostique avec une épreuve immuno-enzymatique (ELISA). Deux peptides immunodominants (VR#1 et VR#2) et deux peptides contenant des ectodomaines neutralisants (Ecto#1 et Ecto#2), de même que de la GP5 recombinante (rGP5) comme témoin ont été préparés. Des échantillons de sérum provenant de porcs non-vaccinés ont été testés pour la présence d’anticorps se liant à ces peptides et ces protéines antigéniques. Les résultats ont été comparés à ceux obtenus avec une trousse ELISA commerciale (HerdChek) qui utilise la protéine de la nucléocapside (N) comme antigène. Des résultats statistiquement similaires à ceux obtenus avec la protéine N n’ont été observés qu’avec VR#1+#2. Ecto#1 et Ecto#2 avaient une sensibilité plus faible que VR#1+#2 et rGP5. Les peptides et rGP5 ont montré des associations significatives avec la protéine N (P < 0,05 ou 0,01), ce qui suggère que GP5 pourrait également être un candidat comme antigène pour le sérodiagnostic. Étant donné que les anticorps contre GP5 persistent beaucoup plus longtemps que ceux dirigés contre la protéine N, il est logique de croire que GP5 elle-même ou ses fragments seraient de bonnes cibles pour le sérodiagnostic. De plus, la présence d’anticorps contre les antigènes structuraux du PRRSV ont montré des différences antigènes-dépendants significatives.
(Traduit par Docteur Serge Messier)
Porcine reproductive and respiratory syndrome virus (PRRSV), an enveloped RNA virus (1), has emerged as a pathogen (2,3) and is considered to be one of the most economically important pathogens of pigs (4–6). Strains of PRRSV have been grouped into North American (VR2332, VR) and European (Lelystad, LV) types on the basis of antigenic and genetic differences between the isolates from the 2 continents (7). The diagnostic situation was simple in the late 1980s and early 1990s: any PRRSV isolated in Europe was certain to be the LV type, and only the VR type was isolated in North America. However, the genotypes soon became mixed. In Europe, VR was introduced in 1996 through the use of a live attenuated PRRSV vaccine (8,9), and LV was introduced around 1999 in North America (5).
The virus core contains the 15-kb genome packaged in a capsid containing 14-kDa nucleocapsid protein (N) subunits. In the envelope, up to 6 structural proteins have been identified. The most abundant are the 17-kDa nonglycosylated membrane protein (M) and the 25-kDa glycosylated GP5 protein, which occur as disulfide-linked heterodimers (10,11). Among the envelope proteins, GP5 appears to be one of the key viral structures. It is believed that attachment and entry to the target cells is mediated by GP5 or GP5-M heterodimers (12). In addition, the neutralization epitope of PRRSV is located in the middle of the GP5 ectodomain (13–17).
Pigs mount a rapid antibody response to PRRSV infection, but the antibodies are directed mainly to the N and M proteins and are not neutralizing (13,18,19). The main neutralization epitope of some North American PRRSVs is in the middle of the GP5 ectodomain (amino acids 37 to 45) (14,16). Neutralizing antibodies are generated more slowly, and their titers remain low (13,19,20), but this may differ in individual pigs. Moreover, neutralizing antibodies against GP5 persist in infected pigs much longer than antibodies against the N protein (17).
Current diagnostic enzyme-linked immunosorbent assays (ELISAs) use extracts from PRRSV-infected cell cultures or recombinant N protein (HerdChek; IDEXX Laboratories, Westbrook, Maine, USA) as an antigen (15,21–24), and the ELISAs based on the open reading frame 7 antigen can easily differentiate between the antibodies induced by the VR and LV types (23,24).
For a more precise diagnosis, peptide ELISAs are used for virus typing in cases of suspected infection with human immunodeficiency virus (HIV), hepatitis C virus, foot-and-mouth disease virus (25–27), and PRRSV (15,17,24,28–30). Peptide ELISAs have cost and throughput advantages over typing methods based on nucleotide sequence (28). In this study, the serodiagnostic application of PRRSV GP5 was evaluated by ELISAs using the GP5 peptides and recombinant GP5 (rGP5).
Two synthetic immunodominant peptides (VR #1 and VR #2) and two neutralizing ectodomain-containing (15,17,24) peptides (Ecto #1 and Ecto #2) were synthesized by Anygen (Gwangju, Korea); the amino acid sequences are presented in Table I, and Figure 1 shows the location of each peptide within GP5. To confirm the antigenicity of the peptides, polyclonal antibodies against the peptide antigens were produced from 6-wk-old New Zealand White rabbits, 3 rabbits per antigen, by Orient Bio (Daejeon, Korea). Each peptide was conjugated with keyhole limpet hemocyanin, as described elsewhere (31,32), and the polyclonal antibody inductivity was interpreted as the immunogenicity of each peptide. Seven days after the final booster inoculation (a total of 3 injections were given, at 2-wk intervals), blood samples were collected from the ear vein and used for the ELISA.
Briefly, 96-well plates (MaxiSorp; Nunc, Roskilde, Denmark) were coated with each of the unconjugated peptides (1 μg/well/90 μL). The plates were then blocked with phosphate-buffered saline (PBS) containing 1% skim milk and washed with PBST (PBS + 0.05% Tween-20, pH 7.4). The serum for IgG analysis was prepared at 1:100 dilutions (dilution factor optimized previously). The antibody samples in 90 μL of PBST were added to the plates, which were then incubated for 1 h at 37°C, emptied, and washed 3 times with PBST. The bound IgG was detected by means of horseradish peroxidase (HRP)-conjugated goat IgG against rabbit antigen (Pierce, Rockford, Illinois, USA) diluted 1:500. The secondary antibody was added and the plates were incubated for 1 h at 37°C, emptied, and washed 3 times with PBST. The presence of any bound secondary antibody was determined with the use of a substrate solution consisting of 10 mL of 100 mM citric acid buffer (pH 4.0), 250 μL of ABTS stock solution (100 mg of ABTS in 4.5 mL of distilled water), and 50 μL of H2O2. The plates were developed in the dark at room temperature for 15 min. The absorbance at 405 nm was read with an ELISA reader (Multiskan EX; Thermo LabSystems, Beverly, Massachusetts, USA).
The significance of the variation between the different groups was determined by 1-way analysis of variance, and the difference between the groups was determined by a Dunnett multiple-comparisons test with the use of GraphPad Instat Software, version 3.05 (GraphPad Software, La Jolla, California, USA). The polyclonal antibodies produced were used as positive controls in the ELISA. To check the cross-reactivity between peptide antigens, each of the peptides was cross-reacted with the antibodies developed against the other peptides. The results were expressed as the mean optical density (OD) ± standard deviation.
In preparing the rGP5, Escherichia coli strains JM109 and BL21(DE3)pLysS and the pRSET vector (Invitrogen, Carlsbad, California, USA) were used for protein expression. The PRRSV strain VR2332 (Ingelvac, Boehringer Ingelheim Vetmedica, St. Joseph, Missouri, USA) was used to prepare total RNA, with Trizol reagent (Invitrogen), and the cDNA was synthesized with the use of a SuperScript III One-Step RT-PCR System (Invitrogen). The E. coli manipulations were carried out according to the manufacturer’s instructions. The standard DNA, RNA, and protein manipulations were carried out as described elsewhere (33,34).
The polymerase chain reaction (PCR) primers were designed with the 5’ primer containing a BamHI (underlined) restriction site (GP5-F: 5′-AATT GGATCC ATGAGCAACGACAGCAGCTCCCA-3′) and the 3′ primer containing a HindIII (underlined) restriction site (GP5-R: 5′-GGCC AAGCTT CTAAGGACGACCCCATTGTT-3′). The PCR products were purified by means of a PCR purification kit (Qiagen, Hilden, Germany), digested with 2 restriction enzymes (BamHI and HindIII [B+H]), and then cloned into B+H-digested pRSET to generate pRSET-GP5. The constructs were transformed into JM109 cells, and the plasmids were prepared as described. The inserted sequence of the construct was confirmed by DNA sequencing. The construct was then transformed into E. coli BL21(DE3)pLysS host cells. The recombinant protein was expressed by adding isopropyl β-d-1-thiogalactopyranoside (final concentration 1 mM) and purified by means of a Probond purification system (Invitrogen) according to the manufacturer’s instructions. The protein concentration was then measured by means of a protein assay kit (Bio-Rad Protein Assay; Bio-Rad, Hercules, California, USA). The anti-rGP polyclonal antibodies were also produced in the rabbits, as described above, and used as the positive control in the ELISA.
Six types of antigens (VR #1, VR #2, VR #1+#2, rGP5, Ecto #1, and Ecto #2) were used for indirect ELISA to detect the peptides and rGP5-specific antibodies in pig serum. At 20 pig farms in Jeonnam province 258 serum samples were obtained from nonvaccinated pigs of both sexes and various breeds that were more than 2 mo old and had no clinical signs (about 13 pigs per farm, randomly chosen). The samples were diluted 1:100 in PBS. Five serum samples from pigs vaccinated against VR2332 were used as positive controls. Serum samples from 10 colostrum-deprived newborn piglets of sows that were free of specific pathogens, including PRRSV, were used as negative controls. As the secondary antibody, HRP-conjugated goat IgG against pig antigen (Serotec, Oxford, England), diluted 1:500, was used. An absorbance above the mean OD of the negative controls plus 3 standard deviations was considered positive. All the data from each ELISA for the 258 field samples were compared to calculate the overall sensitivity and specificity for each antigen. The OD values from each experiment were analyzed with Pearson’s correlation coefficient (SPSS, version 14.0; SPSS, Chicago, Illinois, USA) to determine the strength of the association (R-value) between ELISA reactions.
The peptide antigens and rGP5 showed no cross-reactivity with the antibodies developed in rabbits (Table II).
With the use of HerdChek indirect ELISA to detect PRRSV GP5 antigen-specific antibodies in pig serum, 209 of the 258 field samples (81%) were found to be seropositive. These results were used as a control for the sensitivity and specificity calculations. Only rGP5 and Ecto #2 showed relatively high sensitivities, 99.52% and 75.12%, but the latter was lower than that of VR#1+2, 86.60% (a nonsignificant difference; P = 1.0); the other differences were very or extremely significant. This result indicates that VR #1+#2 can be substituted for the N protein in ELISAs. In the Pearson’s correlation coefficient analysis the N protein showed some correlation with each of the GP5 antigens except VR #2. Interestingly, VR #1+#2 showed the highest correlation with rGP5 (R-value 0.6) (Table III). It is hard to explain why the correlation was greater when the 2 peptide antigens were coated together than when they were tested separately. Perhaps neither antigen alone binds sufficiently with GP5 antibodies. It appears that VR #1+#2 is a better candidate as an antigen for serodiagnosis than the ectodomain (neutralization epitope)-containing antigens.
Ecto #1, which is known as a neutralization epitope, showed a relatively low sensitivity (35.41%), and only 83 samples were seropositive when it was used in the ELISA. Ecto #2 showed higher sensitivity and specificity than Ecto #1. In a previous study (17), antibodies against VR-P3 (a longer peptide containing Ecto #1) were not blocked by the addition of competitive smaller peptides. This suggests that the pigs in the current study also produced antibodies against the epitopes located downstream of the neutralizing domain (Ecto #1). Moreover, Ecto #2 may be a better candidate as an antigen for serodiagnosis than Ecto #1.
Previously, some serum samples were found to have low levels of antibodies against Ecto #1 even though the serum was negative according to HerdChek and lacked neutralizing activity (17). In addition, there was no correlation between the anti-N protein and neutralizing antibody formation (24). In this study, some samples were seropositive (noted as false-positive in Table IV) even though they were seronegative according to HerdChek. Since 208 samples were true-positive (208/209; 99.5%) and 45 were false-positive according to ELISA using rGP5, 45 pigs have antibodies against GP5 but not against the N protein, which suggests that antibodies against GP5 and N protein may present differently.
It has been reported that GP5 peptide-binding antibodies appeared in the serum within 30 d after farrowing, reaching the highest level at 100 to 200 d (17). Moreover, half of the maximum titer was maintained by approximately 400 d. In contrast, N protein-binding antibodies appeared within 7 d, reaching the highest level at 100 d and decreasing below detectable levels at approximately 200 d. Considering the previous and present results, it is likely that antibodies against PRRSV antigens GP5 and N are induced or maintained, or both, with different patterns. Therefore, to perform more precise serodiagnosis, researchers should consider the time of PRRSV infection, which can affect the presence of the proper amount of antibodies.
In conclusion, GP5 and some of its fragments were found to be good targets for serodiagnosis. In addition, the presence of antibodies against PRRSV structural antigens appears to have substantial time-dependent and antigen-dependent differences.
The authors acknowledge a graduate fellowship provided by the Korean Ministry of Education and Human Resources Development through the Brain Korea 21 project. They also thank Ms. Myeonghwa Kim (College of Veterinary Medicine, Chonnam National University, Gwangju, Republic of Korea, for her excellent technical assistance.