Introduction

Porcine reproductive and respiratory syndrome (PRRS) has caused severe economic damage to the swine industry worldwide since its emergence in the late 1980s [1]. This syndrome is characterized by reproductive disorders in sows, including abortion, reduced numbers of weaned pigs due to an increase in the number of stillborn pigs, mummified fetuses, and weakness, and respiratory disorders in pigs from nursery to finishing.

Because of the economic losses caused by PRRSV outbreaks, various types of PRRSV vaccines have been developed and implemented on pig farms with varying degrees of success. Several vaccination trials have shown the replication of live immunogen in pigs to be a crucial requirement for generating robust protective immunity against PRRSV infection [2, 3]. Therefore, a modified-live vaccine (MLV) rather than a killed or subunit vaccine has been deemed to be the most efficacious type of vaccine against PRRSV infection to date and has been employed regularly in both experimental and field-scale trials since its first introduction in 1994.

Currently, various types of MLV vaccines, including genotypes I and II, are commercially available. In regions where a single infection with either genotype I or II has been reported, MLV targeting the circulating genotype should be used. However, in co-infected herds, which genotype of PRRSV MLV should be used to successfully control the disease is less obvious. Criteria for vaccine selection could be the induction of immune responses and the shedding pattern of the vaccine virus. The induction of immune responses following MLV administration may vary according to the virus isolate used to produce the vaccine [2, 4]. Occasionally, vaccination with genotype II MLV can yield undesired outcomes, such as delayed immune responses, low potency of humoral or cell-mediated immune activation, the induction of regulatory IL-10 and/or T-cells (Tregs), the suppression of pro-inflammatory cytokine production, and a reduced level of type I (α/β) and type II (γ) interferon. These undesired outcomes could potentially lead to reduced protective efficiency of a vaccine or, in the worst case, to increased susceptibility to infection by other pathogens [5]. In addition, the safety of MLV of all genotypes remains doubtful because the persistence of MLV, the development of viremia, transmission of a vaccine to non-vaccinated pigs, and clinical signs in vaccinated pigs have been documented following vaccination [6]. Therefore, this study aimed to evaluate the induction of humoral immune responses and viral vaccine shedding following vaccination with either genotype I or II MLV. In the present study, the induction of antibody responses and shedding patterns of six different MLVs were compared.

Materials and methods

Ethics statement

All animal procedures were conducted in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Research Council of Thailand according to protocols approved by the Chulalongkorn University IACUC.

Experimental design

A cohort of 90 seven- to eight-week-old castrated male PRRSV-free pigs were randomly assigned based on a stratification by weight into the following seven treatment groups: NEG, EU1, EU2, US1, US2, US3 and US4. Groups of pigs were housed in separate rooms with separate air spaces (Table 1). The NEG group included 30 pigs that were left unvaccinated. Pigs in the EU1 and EU2 groups were vaccinated intramuscularly with AMERVAC® PRRS and Porcillis® PRRS, respectively, which are both PRRSV genotype I MLVs. The US1, US2, US3 and US4 groups were vaccinated intramuscularly with Fostera™ PRRS, Ingelvac® PRRS MLV, Ingelvac® PRRS ATP and PrimePac™ PRRS+ , respectively, which are all PRRSV genotype II MLVs. The dosage and route of administration were in accordance with the respective manufacturer’s directions.

Table 1 Experimental design
Full size table

Following vaccination, all groups were monitored for changes in physical condition and were scored for clinical respiratory disease. Blood samples were collected at 0, 3, 5, 7, 14, 21, 28, 35, 42 and 84 days post-vaccination (DPV). Sera were separated from blood samples and assayed for the presence of antibody using ELISA and a serum neutralization (SN) assay against both homologous and heterologous isolates. The viral load in serum was measured using real-time quantitative PCR (qPCR). Tonsil scraping samples were collected at 60, 70, 84 and 90 DPV and assayed for the presence of viral RNA by RT-PCR. Individual pigs were restrained using a snare, and samples were then collected by scraping the palatine tonsil with an elongated spoon. Scraping were mixed with 1 ml of DMEM supplemented with 50 μg of gentamicin per ml and filtered through a 0.22-μm nitrocellulose membrane. Filtrates were then stored at -80 °C for later use.

Vaccines and viruses

Homologous and heterologous viruses were used to perform a serum neutralizing (SN) assay. Homologous virus refers to a vaccine isolate. To retrieve homologous virus, each vaccine (except Fostera™ PRRS) was re-constituted in DMEM media. Then, the virus was propagated in MARC-145 cells using a previously described method [7]. Virus was harvested by a cycle of freezing and thawing. Supernatant containing the virus was stored at -80 °C before subsequent use. Because of the inability of the Fostera™ PRRS vaccine virus to be generated using MARC-145 cells, the homologous virus used to generate to Fostera™ PRRS was a virus that was isolated from pigs that were previously vaccinated with Fostera™ PRRS. Heterologous viruses refer to the SB_EU02 and ST_US02 isolates, which are Thai PRRSV genotype I and II field isolates. SB_EU02 and ST_US02 were isolated from farms experiencing PRRS outbreaks.

Clinical evaluation

Rectal temperature was recorded daily for two consecutive weeks by the same personnel at the same time. The severity of clinical respiratory disease was evaluated daily for two consecutive weeks following vaccination, and on a weekly basis for 2 more weeks using a scoring system for each pig following stress induction as described previously [8]. In brief, a score of 0 is normal. Pigs with scores of 1 and 2 display mild dyspnea and/or tachypnea when stressed and when at rest, respectively. Scores of 3 and 4 indicate moderate dyspnea and/or tachypnea when stressed and when at rest, respectively. Pigs with a score of 5 displayed severe dyspnea and/or tachypnea when stressed or at rest.

Quantification of PRRSV RNA in serum and RT-PCR in tonsil scraping samples

Total RNA was extracted from serum and tonsil scraping samples using NucleoSpin® RNA Virus (Macherey-Nagel, Germany) according to the instructions provided by the manufacturer. The RNA quality was measured using a NanoDrop spectrophotometer (Colibri spectrometer, Titertek Berthold, Germany). cDNA was synthesized from viral RNA immediately after the extraction process. The reaction contained 1X M-MuLV reverse transcriptase reaction buffer, 0.5 mM each dNTP, 2.5 µM random hexamers, 13.2 U of RNase inhibitor (RiboLock™, Fermentas, Vilnius, Lithuania), 6.6 U of M-MuLV reverse transcriptase (New England Biolabs, Ipswich, UK), 0.5 µg of viral RNA, and RNase-free water up to 25 µl. The reaction was carried out at 42 °C for 60 min, followed by the inactivation of reverse transcriptase at 95 °C for 10 min. All cDNA samples were kept at -20 °C until used for quantitative PCR (qPCR).

qPCR assays for determination of genomic copy number of PRRSV RNA in serum samples were conducted using an ABI PRISM 7500® Real-Time PCR platform (Applied Biosystems, USA). Primers specific for the ORF5 EU and US strains were used for qPCR [9]. Each qPCR reaction contained 0.1 µg of cDNA, 0.2 µM ORF5EU or ORF5US primers (as appropriate for each sample), 1x EvaGreen real-time-PCR master mix E4® (GeneOn, Germany), and deionized water to yield a 20-µl final volume. pGEM®-T Easy Vector (Promega, WI, USA) containing an insert of 704 bp from ORF5EU or 780 bp from ORF5US were used to construct standard curves in these qPCR assays. Thermocycling conditions for qPCR started with an initial denaturation step at 94 °C for 4 min, followed by 35 cycles of denaturation at 94 °C for 45 s, annealing at 55 °C for 45 s, and extension with fluorescence acquisition at 72 °C for 45 s. A standard curve was made for each pair of primers.

To detect the presence of virus in tonsil scrapings, RT-PCR and PCR amplification were performed using GoTaq® Green Master Mix (Promega, WI, USA). All reactions were performed as described above for qPCR. PCR amplicons were visualized by agarose gel electrophoresis.

Antibody detection

Sera were assayed for the presence of PRRSV-specific antibody by ELISA and serum neutralization (SN) assay. PRRSV Idexx ELISA (HeardCheck PRRS X3, Idexx Laboratories Inc., Westbrook, Maines, USA) was used in accordance with the manufacturer’s instructions. The presence or absence of antibody was determined by calculating the sample-to-positive control (S/P) ratio of the test. Serum samples were considered positive for PRRSV antibody if the S/P ratio was greater than 0.4.

SN assays were performed using the homologous isolate along with two different heterologous PRRSV isolates, SB_EU01 and ST_US02 as described previously [10]. The SN antibody titers were reported as the highest serum dilution that resulted in a 90% reduction in the number of fluorescent focus units per well. Geometric mean titers were calculated.

Sentinel pigs

Viral shedding patterns were monitored by placing two groups of two age-matched sentinel pigs in contact with the principal pigs at 60 and 84 DPV for 7 days. Subsequently, sentinel pigs were removed and housed in a separate unit for an additional 2 weeks to monitor them for seroconversion, using ELISA.

Statistical analysis

Data from repeated measurements were analyzed using multivariate analysis of variance (ANOVA). Continuous variables were analyzed for each day by ANOVA to determine whether there were significant differences between treatment groups. If the p-value in the ANOVA table was ≤ 0.05, differences between treatment groups were evaluated by pairwise comparisons using least significant differences at the p ≤ 0.05 rejection level.

Results

Rectal temperature and clinical observations

The rectal temperatures of the pigs in the control group and all vaccinated groups were within normal physiological ranges throughout the experimental period. None of the pigs in any of the groups displayed any clinical respiratory disease throughout the study except for those in the EU2 and US3 group. There were 5/10, 3/10 and 6/10 of pigs in the EU2, US2 and US3 groups, respectively, that showed respiratory signs at 20 DPV. At 21 DPV, all pigs in the EU2 group were injected once intramuscularly with turathromycin. At 28 DPV, all pigs in the EU2, US1 and US2 groups were in-feed medicated with amoxycillin at 300 ppm for 7 consecutive days. There was one pig in the US3 group that died at 35 DPV, and the necropsy revealed paleness of skin and gastric ulceration. PCR results from organ samples, including lung, bronchial and mesenteric lymph nodes, were positive for PRRSV. At 35 DPV, two pigs from the EU2 and US3 groups were euthanized because of the severity of clinical disease.

Quantification of PRRSV RNA in serum

PRRSV RNA was not detectable in the NEG and EU1 groups throughout the study (Fig. 1). In the US2 and US3 groups, the viral RNA copy numbers were highest at 3 DPV and then slowly declined until they were below the limit of detection at 21 and 28 DPV, respectively. In the EU2 group, the viral RNA copy number peaked at 3 DPV and decreased until it was below the limit of detection at 7 DPV. In the US1 group, the viral RNA copy number was detectable after 3 DPV and peaked at 7 DPV. Then, the viral RNA copy number level gradually declined until it could not be detected at 28 DPV.

Fig. 1
figure 1

Mean values of the genomic copy number of PRRSV RNA in serum of the NEG (), EU1 (), EU2 (), US1(), US2 (), US3 () and US4 () groups. Variation is expressed as the standard deviation. Different letters in superscript indicate a statistically significant difference (p < 0.05) between groups

Full size image

The viral RNA copy number of the EU1 group was significantly lower compared with the other vaccinated groups at 3 DPV. At 7 DPV, the viral RNA copy number in the US1 group was not different from those of the other genotype II MLV groups but was significantly higher than those of the genotype I MLV groups.

RT-PCR in tonsil scrapings

Viral RNA was not detected in any pigs in the EU1 group. In contrast, viral RNA in the EU2 groups was detected in 3 of 9, 2 of 9 and 1 of 9 pigs at 60, 70 and 90 DPV, respectively (Table 2). For genotype II MLVs, viral RNA was detected in 1 of 10 and 2 of 10 pigs in the US1 and US4 group, respectively, at 60 DPV. In the US2 group, viral RNA was detected on all sampling days and was still detectable in 2 of 10, 3 of 10, 1 of 10 and 1 of 10 of pigs at 60, 70, 84 and 90 DPV, respectively. In the US3 group, viral RNA remained detectable in 1 of 8, 2 of 8 and 1 of 8 of pigs at 60, 70 and 84 DPV, respectively.

Table 2 Detection of viral RNA in tonsil scrapings from vaccinated pigs and sentinel pigs
Full size table

Antibody responses as measured by ELISA

Pigs in the NEG group remained serologically negative throughout the experiment (Fig. 2). Our findings revealed a similar pattern in all vaccinated groups. Antibody responses were first detected at 7 DPV at the earliest in some pigs of the vaccinated groups, but the average antibody level was below the cutoff level (S/P ratio at 0.4). At 14 DPV, the average antibody responses of the EU2, US1, US2 and US3 groups were significantly higher than those of the EU1 and US4 groups, in which the average antibody levels were below the cutoff level. At 21 DPV, the average antibody responses of all vaccinated groups were above the cutoff level and remained constant until the end of the experiment. There were no differences between any of the vaccinated groups from 21 to 84 DPV.

Fig. 2
figure 2

Mean values of PRRSV-specific antibodies as measured by ELISA of the NEG (), EU1 (), EU2 (), US1 (), US2 (), US3 () and US4 () groups. Variation is expressed as the standard deviation. Different letters in superscript indicate statistical significant difference (p < 0.05) between groups. A dashed line indicates the cutoff level (S/P ratio of 0.4)

Full size image

Antibody responses as measured by serum neutralization (SN) assay

Pigs in the NEG group remained serologically negative throughout the experiment. In all vaccinated groups, the SN assay against homologous virus in all vaccinated groups (Fig. 3A) showed a similar pattern of SN titers that could be detected as early as 28 DPV. Titers reached a peak level at 35 or 42 DPV and then declined by 84 DPV. At 28 DPV, the EU2 group had significantly higher SN titers against the homologous virus compared with the other vaccinated groups. In contrast, the US4 group had significantly lower SN titers compared with the other vaccinated groups. At 35 DPV, the EU2 and US2 groups had significantly higher SN titers compared with the other vaccinated groups. At 42 DPV, the EU2, US1, US2 and US4 groups had significantly higher SN titers compared with the EU1 and US3 groups.

Fig. 3
figure 3

A Antibody responses as measured by serum neutralizing (SN) assay using homologous virus as a recall antigen in the NEG (), EU1 (), EU2 (), US1 (), US2 (), US3 () and US4 () groups. Variation is expressed as the standard deviation. Different letters in superscript indicate a statistical significant difference (p < 0.05) between groups. B. Antibody responses as measured by serum neutralization (SN) assay using heterologous genotype I virus (SB_EU02) as a recall antigen of the NEG (), EU1 (), EU2 (), US1 (), US2 (), US3 () and US4 () groups, respectively. Variation is expressed as standard deviation. Different letters in superscript indicate a statistically significant difference (p < 0.05) between groups. C. Antibody responses as measured by serum neutralizing (SN) assay using heterologous genotype II virus (ST_US01) as a recall antigen in the NEG (), EU1 (), EU2 (), US1 (), US2 (), US3 () and US4 () groups. Variation is expressed as the standard deviation. Different letters in superscript indicate a statistically significant difference (p < 0.05) between groups

Full size image

For the heterologous genotype I virus (SB_EU02), the kinetics of the SN response differed in a manner that was dependent on the MLV (Fig. 3B). Compared with homologous virus, the SN titers were relatively low. The SN titers remained at a similar level from 28 to 42 DPV and then declined by 84 DPV. However, the SN titers in the US1 group were significantly higher than in the other vaccinated groups from 28 to 84 DPV and were significantly higher than those for the homologous virus.

In the case of heterologous genotype II virus (ST_US01), the kinetics of the SN response were similar to those evoked using homologous genotype I virus (SB_EU02) (Fig. 3C). SN titers were lower than with homologous genotype I virus in all groups, except for the EU2 group. The SN titers of the EU2 group were significantly higher on 35 and 42 DPV compared with the other vaccinated groups.

Sentinel pigs

Sentinel pigs introduced to the EU2, US2 and US3 groups at 60 DPV seroconverted, but those in the EU1, US1 and US4 groups did not. Sentinel pigs that were placed in contact with pigs of all groups on 84 DPV did not seroconvert over a 7- or 14-day period of observation.

Discussion

We compared the efficacy of six different PRRSV MLVs in the induction of antibody responses in PRRSV-free pigs. All six MLVs rapidly induced antibody responses as measured by ELISA. The antibody responses were detected as early as 7-14 DPV in an MLV-genotype-dependent manner. The antibody levels in all vaccinated groups were similar at 21 DPV. It was notable that there was a difference in early antibody detection between two genotype I MLVs. The EU1 group had a significantly lower S/P ratio than the EU2 group at 14 DPV, and the S/P ratio was below 0.4, the cutoff level. Surprisingly, the EU2 group produced an antibody response at a similar level when compared with the genotype II MLVs. In summary, there was no difference in the antibody responses as measured by ELISA for any of the MLV genotypes. The results of the present study suggest that the MLV genotype is not the key factor in the induction of immunity, but the specific virus isolate used for the vaccines might play an important role. Moreover, it is notable that the antibody detection in the present study was performed using Idexx ELISA, which can simultaneously detect specific antibodies against PRRSV genotype I and II infections. However, these findings may not be applicable when using other diagnostic kits.

Following MLV vaccination, antibody responses measured by the SN assay were delayed regardless of the MLV genotype and isolate in the vaccines. The responses were detected as early as 28 DPV. In addition, the response was isolate-specific. Homologous responses generated using a homologous virus induced a higher, although delayed response compared with the heterologous responses generated using either heterologous genotype I or II viruses. Heterologous responses were lower and shorter in duration. Our finding of differences in responses was not surprising. PRRSV isolates differed in their susceptibility to neutralization [11], and the mechanisms associated with this susceptibility remain poorly characterized, although the influence of N-linked glycosylation in decoy epitope regions could be one key factor [9, 12, 13]. A previous study demonstrated that a heterologous response could be higher or lower [4], depending on the isolates that were used in the assay.

In addition to their ability to induce an immune response, the shedding patterns of the vaccine viruses were investigated using three different measurements, including the duration of viremia, the detection of viral RNA in tonsils, and infection of sentinel pigs. After vaccination, we detected a difference in the shedding patterns between MLVs. The two genotype I MLVs had a shorter viremic phase compared with genotype II MLVs. However, the magnitude of viral titers was not different. In addition, there was a difference in the shedding pattern of genotype I MLV, although one genotype I MLV had a shedding pattern that resembled that of a genotype II MLV. Viremia could not be detected in one genotype I MLV. This finding could indicate the absence of viremia or that the quantity of virus in the serum was lower than the limit of detection of the real-time PCR assay. Within genotype II MLVs, all three MLVs caused viremia as early as 3 DPV, and it then declined thereafter. In contrast, the titers of one genotype II MLV continued to increase until 7 DPV and then declined. These findings suggested that the viremic phase of MLV was associated with the virus isolate used in the vaccine, not the virus genotype.

To further evaluate the viral shedding pattern, sentinel pigs were used. Sentinel pigs were housed along with principal pigs of the EU2, US2 and US3 groups in the same pen beginning on day 60, and they were found to undergo seroconversion. However, sentinel pigs introduced at 84 DPV remained uninfected, as indicated by their failure to seroconvert. The shedding patterns of vaccine viruses over this long period of time have not yet been investigated. Compared to wild-type PRRSV, vaccine viruses should be shed to sentinel pigs over a shorter time. Previous studies conducted by several investigators to characterize several wild-type field isolates of PRRSV and the duration of PRRSV shedding to sentinel pigs have suggested that virus shedding to sentinel pigs occurred on average 60 to 70 days after exposure [14]. Although vaccine viruses were not transmitted to sentinel pigs at 84 DPV, the detection of viral-RNA-positive samples in the tonsil scraping samples might represent a risk factor for the shedding of vaccine viruses.

The PCR results from tonsil scrapings at 84 DPV indicated that vaccinated pigs still harbored viral RNA. However, whether the RNA-positive samples represented infectious viruses was not determined. The detection of viral RNA does not necessarily indicate the isolation of infectious virus. Any viral genomic material needs to be tested further to determine whether the pigs may still be infectious and contagious. Using a swine bioassay, it was demonstrated that homogenates from tonsils collected from pigs infected with the PRRSV strain VR-2332 at 105 days post-exposure remained infectious [15]. Viral RNA was detected in the tonsils, suggesting that viruses remained present in both groups of pigs but were not transmitted to contact sentinel pigs. Determining whether virus shedding can be reinitiated will require further study.

In conclusion, based on the induction of immune responses, all MLV genotypes yield a similar immune-response pattern. Measurement of the antibody response by ELISA is quick, but the response measured using an SN assay is delayed and isolate-specific. However, the shedding pattern of a vaccine virus is influenced by the isolate that is used to manufacture the vaccine. The criteria for MLV selection should be based on the shorter duration of vaccine virus shedding and the broader response against heterologous virus.