Summary

Genetic and environmental effects on incidence of Porcine Circovirus Associated Disease (PCVAD) and immune responses to Porcine Circovirus 2 (PCV2), and their relationships with body weights were studied in 3,440 pigs of the Nebraska litter size selection lines.

Pigs were weighed at birth, weaning, 65 and 180 days of age and scored for symptoms of PCVAD every 10 days from 70 to 180 days of age. Necropsies were performed to confirm accuracy of scoring. PCV2 viraemia, and antibodies to PCV2, Porcine Reproductive and Respiratory Syndrome Virus, Mycoplasma hyopneumoniae and Actinobacillus pleuropneumoniae were measured in serum from blood samples drawn at various ages from live pigs and in tissues of pigs necropsied. PCV2b genotype was confirmed to be the pathogen causing PCVAD; other pathogens studied were not involved.

Pigs with no symptoms of PCVAD had significantly greater weights (0.22, 1.12, 6.6 and 46.0 lb, at birth, weaning, 60 days and 180 days, respectively) than affected pigs. Heritability of PCVAD score was 16 per cent ± 4 per cent. The location in which pigs were raised accounted for 22 per cent of the variation in PCVAD score. Nearly all pigs were non-viraemic until 90 days of age but many had antibody titres at weaning and at 60 days of age. These maternal antibodies appeared to protect pigs from PCVAD until approximately 90 days of age. Heritability of viraemia level at 90 days of age was greater than at 125 days of age (38 ± 11 per cent vs 11 ± 8 per cent). Genetic variation existed for antibody titres at 90 (h² = 55 ± 21 per cent) and 125 days of age (h²= 10 ± 8 per cent).

Incidence of PCVAD was correlated genetically with body weights, PCV2 viraemia level, PCV2 antibody titres, and body weights. Expected response to direct selection for reduced PCVAD score was very low (-0.89 per cent in one generation), whereas expected response to index selection for 65-day weight and PCV2 viraemia and antibody titres at 90 days of age was -8.0 per cent, 998 per cent greater than direct selection. Genomic selection for decreased incidence of PCVAD is feasible.

Introduction

Porcine Circovirus Associated Disease (PCVAD), caused by Porcine Circovirus 2 (PCV2), causes high economic losses to pork producers. Symptoms of PCVAD in the University of Nebraska–Lincoln swine research herd were first observed in 2002. Not all pigs on the farm were infected and the incidence rate varied depending on the genetic makeup of the pigs, their location at the farm, and season of the year. The incidence rate in crossbred pigs was very low, but a significant number of pigs of the UNL lines selected for increased litter size were affected. Some pigs seemed to be highly sensitive to the disease whereas others in the same pen remained healthy and showed no symptoms.

Usually, only one or two pigs in a pen were affected, but it was common for a high percentage of pigs within some litters to be affected, even when raised in different locations.

These observations pointed to underlying genetic variation in the immune response of pigs to the PCV2 virus. A study of PCVAD conducted at another institution supports this hypothesis as the incidence rate was greater in some breed crosses than others. However, sample size in that experiment was too small to determine the degree of genetic variation (heritability) in incidence rate of PCVAD and in immune responses to PCV2 virus. If sufficient genetic variation exists, then greater resistance to PCV2 and reduced incidence of PCVAD through selection are possible. When practiced in nucleus breeding populations, greater resistance achieved through selection can be transmitted through the breeding pyramid to commercial producers, possibly reducing the need for vaccination.

The authors therefore conducted a study in which pigs were systematically scored for PCVAD, weighed and bled to create a database for genetic analyses. Because secondary pathogens are often thought to be involved in expression of PCVAD, pigs were also characterised for Porcine Reproductive and Respiratory Syndrome Virus (PRRVS), Mycoplasma hyopneumoniae and Actinobacillus pleuropneumoniae to determine whether these pathogens were involved along with PCV2 in expression of PCVAD.

Pigs were from generations 24 to 26 of UNL selection lines, Lines 2 and 45, and control lines, Lines 16 and 61, that were derived from a Large White-Landrace composite population formed in 1979. Lines 2 and 45 have been selected for increased litter size (generations 1 to 20) and increased litter size, increased growth, and decreased backfat, generations 21 to 27. Lines 16 and 61 were randomly selected. Each line is maintained with 40 to 45 litters per generation. Line 2 and 16 litters were born in July and August (denoted contemporary group 1, CG1) and Line 45 and 61 litters were born in January and February (CG2).

Pig management, PCVAD scoring and serum sampling

Within two days of farrowing, fostering of pigs among sows both within and between lines was practised. Fostering could not be accomplished uniformly in all litters. Total number and number of live pigs per litter, pig birth (n=3,440) and weaning weights (n=3,438), and number after foster and weaned per foster dam (NW) were recorded.

Pigs were weaned at approximately 17 days and grouped by age in nursery pens of 30 pigs per pen. At 60 to 65 days of age, four boars from each of the 15 largest litters and four to five females from each of the largest 20 litters in Lines 2 and 45 were identified as candidates for selection as breeders. Two sons per sire and one to two gilts per litter were randomly selected as candidates for breeding in Lines 16 and 61. Final selections occurred at 180 days of age.

All breeder candidates were placed in one of six rooms, eight pens per room and 10 pigs per pen, in a confined, mechanically ventilated building denoted as Location 1. Remaining pigs were placed in one of three locations. Location 2 was a confined building, 25 pens of 10 pigs per pen, with natural ventilation regulated by thermostatically controlled curtains over windows. Location 3 was a confined building, 23 pens of 10 pigs per pen, with natural ventilation controlled manually by adjusting doors over windows. Location 4 was five outdoor lots containing a small hoop structure with straw bedding. There were 50 to 60 pigs per lot. Pigs in CG1 were in Locations 1, 2, and 3; pigs in CG2 were in Locations 1, 2 and 4.

Pigs with symptoms of PCVAD were first observed in 2002 (generation 21). Observations in subsequent years led the researchers to a protocol that produced data for genetic analyses. During Generations 24, 25 and 26 pigs were systematically weighed, scored for PCVAD, and blood samples were drawn. Weight at 65 days was recorded for 2,646 pigs (some pigs were not weighed) when they were placed in finishing pens and weight at 180 days (n=3,115) were recorded. Beginning seven days after pigs were placed in finishing pens, they were scored for symptoms of PCVAD once every 7 to 10 days, approximately 10 scores per pig, until 180-day weight was recorded. Pigs with no symptoms received a score of 0, pig with minor symptoms a score of 1, and pigs with definitive symptoms a score of 2. Scores were based on degrees of muscle wasting, growth retardation, rough hair coat, diarrhoea and respiratory distress. A score of 1 was used only to identify pigs for more careful future observation. Only pigs receiving one or more score of 2 were considered positive for PCVAD.

Blood was collected at 60, 90 and 125 days of age from all pigs from Generation 25, CG2, and Generation 26, CGs 1 and 2. Blood was collected from pigs at weaning in Generation 25, CG2 and from sows in Generation 26, CG1 when their pigs were weaned. PCV2 viraemia, a measure of the pigs ability to replicate virus, was measured in serum of all pigs with PCVAD score of 2, in serum from a randomly selected pen mate with scores of only 0, in samples of a full sib from another pen, and in samples from two pigs drawn randomly from each birth litter in which no pigs were positive for PCVAD. Serum was sent to Iowa State University Veterinary Diagnostic Laboratory, Ames, Iowa, where Porcine Circovirus II C-ELISA PCR and PCV2 Quantitation (qPCR) were performed to obtain PCV2 viraemia and antibody levels.

ELISA was used to test sub-samples of serum for Mycoplasma hyopneumoniae (MH), Actinobacillus pleuropneumoniae (APP), and Porcine Reproductive and Respiratory Virus (PRRSV). Samples collected at 90 (n=261) and 125 days (n=228) from pigs from generation 25, CG2, were tested for MH; samples collected at 125 days of age from generation 25, CG2 (n=228) and Generation 26, CG1 and CG2 (n=511) were tested for APP. The UNL swine herd is free of PRRSV. To confirm this status, serum from 52 pigs from generations 24 to 26 was tested for PRRSV.

Necropsies were performed on samples of pigs with a PCVAD score of 2 from generation 24, CG1 (n=10) and CG2 (n=11), and generation 25, CG2 (n=17) and in 11 randomly selected pigs with PCVAD score of 0 from generation 25, CG2. Pigs were from all locations and only one pig from any one litter was selected for necropsy. Immunohistochemistry and RT-PCR for PCV2 were performed in lung, cervical lymph node, mesenteric lymph node, tonsil, kidney and ileum of these pigs. Nasal swabs for RT-PCR testing were collected from five pigs; two of these pigs had no lesions suggestive of PCVAD. Necropsies and RT-PCR were done at the Veterinary Diagnostic Center of the University of Nebraska Department of Veterinary and Biomedical Sciences. These pigs were also tested for PRRSV antibodies by ELISA. Serum from three pigs that were necropsied and that were diagnosed with PCVAD were submitted to the Veterinary Diagnostic Laboratory at Iowa State University, Ames, Iowa where Porcine Circovirus II C-ELISA PCR-PCV2 Quantitation in which the virus was sequenced to determine the specific PCV2 transcript in this herd.

Statistical procedures

Data were analysed with procedures appropriate for genetic analyses. Two traits had binomial distributions as the outcome was either yes or no, coded as 0 or 1. These were PCVAD score (1 = positive, 0 = negative) and whether a pig was viraemic (0 = no viral replication, 1 = viral replication, viraemia level greater than 0). For those pigs that were viraemic (coded score of 1) the observed viraemia (genomic copies per ml) were expressed as log¹⁰ to normalise the distribution. PCV2 antibody titres and body weights were normally distributed.

Table 1 contains a description of traits and numbers of records. Table 2 contains the joint distributions of PCVAD scores, PCV2 viraemia, and PCV2 antibody titres. Genetic analyses used a pedigree file containing 12,032 pigs, all those with phenotypes in the present study and all parent animals tracing back to the base generation. Traits analysed were PCVAD Score, birth weight, weaning weight, 65-day weight, 180-day weight, viraemia scores at 90 days and 125 of age (the 0 or 1 code), log₁₀ of viraemia level at 90 days and 125 days of age in pigs with positive viraemia scores, and PCV2 antibody titres at 60, 90, and 125 days of age. Antibody titres .5 and greater are considered to be positive, evidence that the pig had been exposed to virus, titres of .2 to .49 are in the suspect range, and those below .2 are considered to be negative.

Results

Observations and Fixed Effects

Overall, 14.4 per cent of pigs had at least one positive PCVAD score, but the incidence varied greatly across generations (Figure 1). Mortality rate of pigs with positive score was 35.4 per cent. Genetic lines did not differ significantly in incidence of PCVAD.

Figure 1. Percentage of pigs scored positive for PCVAD

Nearly all serum samples collected at 60 days of age (94.6 per cent) were negative for PCV2 viraemia; therefore viraemia at 60 days of age was not analysed. All but two serum samples collected at 90 days of age from pigs in generation 26, CG2 had negative viraemia (Table 2); therefore, data from that group were deleted from analyses of 90-day viraemia level. All but two serum samples collected from weaned pigs were negative for PCV2 viraemia but 37 of 77 of their dams were positive. Average antibody titres for sows and progeny were 1.07 ± 0.15 and 0.90 ± 0.16, respectively.

Antibody titres for PCV2 at 60, 90, and 125 days of age for each contemporary group are shown in Figure 2. Titres at 60 and 90 days were similar in generation 25, CG2, and generation 26, CG1, greatest in Generation 26.

Figure 2. Mean PCV2 antibody titres

Frequency of viraemic pigs varied greatly across generations, contemporary groups and ages (Figure 3). In generation 26, CG2, 53.9 per cent of the pigs were non-viraemic at 90 days of age but had antibody titres greater than 0.5 (Table 2). Only 5.9 per cent of the pigs in other groups were non-viraemic and had high antibody titres. Non-viraemic pigs at 125 days of age with antibody titres less than 0.2 occurred in 60.8 per cent of pigs from generation 26 CG2, but in only 10.2 per cent of the pigs in the other groups.

Figure 3. Percentage of viraemic pigs at 90 and 125 days

Incidence of PCVAD was greater in males than females and males had greater (P<0.05) PCV2 antibody titres at 60 (0.03) and 90 days (0.05). The probability of being viraemic at 90 days of age was less for females than males (-0.30, P<0.05).

Pigs with 0 PCVAD score weighed more (P<0.0001) at birth, weaning, 65 days and 180 days (0.22, 1.12, 6.6 and 46.0 lb, respectively) than pigs with score of 1. They also had lower PCV2 viraemia levels at 90 (0.26±0.15) and 125 days (0.85±0.12) and higher antibody titres (0.04±0.01 and 0.05±0.02, respectively).

PCV2 sequence

All three pigs whose PCV2 mRNA was sequenced were positive for PCV2b genotype. The sequence for one pig was 100 per cent identical to a PCV2 isolate previously characterised and described in the National Center for Biotechnology Information database. The sequences for the other two pigs were not identical to any sequence in the database. One of them had a single base change at position 116 (G to A); the other one had two base changes, one at position 116 (G to A) and one at position 465 (C to G). These findings confirm that PCV2b was the causative virus for PCVAD in this population but that mutations had occurred causing slightly different nucleotide sequences from those previously characterised.

Necropsy findings

Tissue samples from 36 of the 38 pigs that were positive for PCVAD were also positive for PCV2. Tissues of the other two were negative, but their nasal swabs were positive. Tissues from all 11 pigs scored as negative for PCVAD were negative for PCV2. Pigs with PCVAD had severe wasting, weight loss, rough hair coat, enlarged mesenteric lymph nodes, lesions indicative of pneumonia, and chronic colitis. Lymphocyte depletion within mesenteric lymph nodes and giant cells in lymphoid follicles, splenic follicles, and Peyer’s Patches, and thymic atrophy due to lymphocyte depletion were observed. Nineteen pigs had symptoms of Mycoplasma hyopneumoniae; 18 had symptoms of Streptococcus suis, and four had symptoms of Lawsonia intracellularis.

ELISA screening

All serum samples tested for PRRSV and for Mycoplasma hyopneumoniae were negative. Thirty two of the 228 serum samples collected at 125 days of age from generation 25, CG2 pigs were positive for APP. Twenty one of 513 serum samples collected at 90 days of age from pigs of generation 26 were positive for APP, 16 had titres in the suspect range, and the rest were negative. Eight pigs that were positive for APP and five with titres in the suspect range had positive PCVAD scores. However, 142 pigs with positive PCVAD scores were negative for APP. Thus, PRRSV and Mycoplasma hyopneumoniae can be ruled out as secondary pathogens, and APP was not likely a secondary pathogen, involved in expression of PCVAD in this population.

Genetic and environmental parameters

Percentages of total variation due to genetic and environmental effects for PCVAD score and body weights are in Table 3 and those for PCV2 viraemia and antibody titres are in Table 4. Direct heritability, a measure of the relative importance of genes of the pig, of PCVAD score was 16 ± 4 per cent. Heritabilities of body weights ranged from 16 per cent for weaning weight to 27 per cent for birthweight. Genes of the dam, maternal heritability, were important for birth and 180-day weights. The location in which pigs were raised accounted for the most variation (22 per cent) in PCVAD score whereas several environmental sources of variation contributed to variation in body weights. For pigs with positive viraemia score (levels greater than zero), heritability estimate of viraemia level at 90 days of age was greater than at 125 days of age (38 ± 11 per cent versus 11 ± 8 per cent). Genetic variation existed for antibody titres at 125 days of age (h² = 10 ± 8 per cent), but not at 90 days of age when all data were included in analyses. However, when data for generation 26, CG2 were deleted, genetic variation in PCV2 antibody titre at 90 days of age was high (h²=55 ± 21 per cent). A large percentage of the variation in viraemia score at 90 and 125 days of age was due to either the location or the room in which pigs were raised. These sources of variation were relatively small for viraemia level and antibody titres.

Genetic and residual correlations among traits were calculated but are not presented here. The important ones were that PCVAD score was quite highly correlated genetically with day 90 viraemia (r_g = 0.75) and antibody level (r_g = -0.67) and moderately correlated with 65-day weight (r_g = -0.53). Neither birth weight nor weaning weight were significantly correlated genetically with viraemia or antibody levels. However, weaning weight was highly correlated genetically with viraemia at 125 days of age (r_g=-0.73) and 180-day weight was negatively correlated with viraemia at both 90 and 125 days of age and positively correlated with antibody titres at 90 days of age. viraemia level at 90 and 125 days of age were positively correlated genetically (r_g=0.59) and viraemia and antibody titres at 90 days of age were negatively correlated (r_g=-0.51). Antibody titres at 90 and 125 days of age were positively correlated (r_g=0.59).

Although several environmental correlations among traits were significant, none was especially strong. Most notable were that environmental effects on score for PCVAD and for body weights through 65 days of age were correlated (-0.47e=-0.38), and viraemia at 90 days of age was negatively correlated with body weights (-0.26 < r_e < -0.16).

Discussion

The authors conclude from this study and others in the literature that genetic variation in incidence of PCVAD and measures of immune responses to PCV2 exists. The heritability of PCVAD score was 16 per cent and heritabilities of PCV2 viraemia and antibody levels at 90 days of age were 38 per cent and 55 per cent, respectively. Genetic variation in viraemia and antibody levels at 125 days of age were less and were not significant, perhaps because death of pigs with PCVAD between 90 and 125 days resulted in fewer records and decreased the genetic variance in remaining pigs.

Viraemia and antibody levels at 90 days of age and 65-day weight were quite highly correlated genetically with PCVAD score. Thus selection for these traits may be an effective way to decrease incidence of PCVAD. Expected selection responses were calculated for different selection strategies. These were 1) direct selection – selection of breeders only from pigs that never displayed symptoms of PCVAD, 2) single-trait selection for weight, viraemia, or antibody levels, and 3) index selection for correlated traits or PCVAD score plus correlated traits.

Even though PCVAD score is heritable, direct selection was relatively ineffective, resulting in a reduction of incidence of PCVAD of only -0.8 per cent in the first generation. This result occurred because 85 per cent of the pigs, those with score of 0, were candidates for selection, resulting in a very low selection rate. The other traits are continuously distributed so pigs vary across the entire range of the distribution. Assuming 10 per cent of the males and 30 per cent of the females are selected for high PCV2 antibody titres at 90 days resulted in the greatest correlated expected response to single-trait selection (generation 1 response = -6.5 per cent). Greatest expected response was for a three-trait index including 65-day weight and PCV2 viraemia and antibody titres at 90 days of age (-8.0 per cent). A four-trait index of these traits and PCVAD score produced the same expected response per generation. Expected selection responses were 414 to 998 per cent greater when correlated traits were used than from direct selection.

Necropsy results confirmed that PCV2 was the main causative agent of PCVAD in this population. Genotyping PCV2 revealed that PCV2b was likely the causative agent. However, at least three allelic forms, and possibly more, existed. Thus, mutations have occurred in the PCV2b genome but the consequence is not known.

Porcine Reproductive and Respiratory Virus, Mycoplasma hyopneumoniae, and Actinobacillus pleuropneumoniae were not secondary pathogens involved in expression of PCVAD. In other work, co-infection of pigs with PCV2 and Gram-negative bacteria induced viral replication of PCV2. Diagnostics for these pathogens were not performed, but infection of pigs with pathogens such as Escherichia coli and Haemophilus influenzae, Gram-negative bacteria known to be present in this population, may have increased the risk of PCVAD.

A strong relationship between PCV2 viraemia level at 90 days of age and incidence of PCVAD across contemporary groups existed. Incidence of positive PCVAD Scores was 3.9 per cent in generation 26, CG2, and only one per cent of these pigs were viraemic at 90 days of age. The incidence of PCVAD in generation 25, CG2, and Generation 26, CG1, was greater than 20 per cent; 75 per cent and 93 per cent of the pigs in these respective groups were viraemic.

Pigs with PCVAD differed in birth weight, weaning weight, 65-day weight, and 180-day weight, even though some of these weights were recorded well before pigs expressed symptoms of PCVAD. The genetic correlations of PCVAD score with birth and weaning weights were positive, but not significant. However, environmental correlations were negative and significant, indicating that the phenotypic relationship was largely due to environmental effects that reduced early body weights and increased risk of PCVAD.

Mean PCV2 antibody titres in pigs at weaning was high, most likely due to maternal PCV2 antibodies because all these pigs were non-viraemic. Antibody titres decreased from weaning to 60 days of age, but 94.6 per cent of pigs were still non-viraemic at 60 days of age. Antibody titres continued to decrease from 60 to 90 days of age, suggesting that maternal antibodies were deteriorating . Effects associated with the birth litter were a major source of variation in antibody titres at 60 and 90 days of age, providing further evidence for significant variation associated with maternal antibodies. Other work has shown that high titres of maternal PCV2 antibodies are generally protective against PCV2, whereas low titres are not; however, maternal antibodies do not fully prevent PCV2 infections. Thus, pigs in our study were likely at least partially protected from PCV2 by maternal antibodies to at least 60 days of age.

Implications

Immune responses to PCV2 are heritable. Thus, genetic selection could be a useful tool to reduce incidence of PCVAD. Even though progress is permanent, several generations of selection will be required to greatly reduce the incidence and this selection must be practised in nucleus herds and then transmitted through the breeding pyramid to commercial populations. Consequently, it would take considerable time for such a strategy to significantly reduce the incidence of PCVAD in commercial herds.

Furthermore, response to selection for PCVAD scores, viraemia, and antibody levels will occur only if all pigs in nucleus populations are exposed to PCV2 so that variation reflects genetic variation in the traits. Marker-assisted or genomic selection may be more effective as once marker panels with known relationships with response variables are available, selection can be practised in any population without exposure of pigs to PCV2.

Thus genomic selection for resistance to PCV2 and decreased incidence of PCVAD may be the most effective long-term selection strategy.

Further Reading

-

You can view other papers in the 2009 Nebraska Swine Report by clicking here.

Further Reading

-

Find out more information on Post-Weaning Multisystemic Wasting Syndrome (PMWS) by clicking here.