Viral Disease in Swine: An Immunologist’s Perspective with Focus on PRRS

Introduction

The controversy that disease was transmitted by invisible life forms stretches back centuries dating to Aristotle (300BC) who believed that all life originated from soil and Virgil (40BC) who believed that bees came from honey and maggots originated from warm meat. It was Spallanzani who showed that no maggots arose from boiled meat while others showed that simple cotton filters could prevent spontaneous generation of life from warm meat. The idea that invisible substances caused disease was proposed by Varro already in the second century BC, but a millennium passed before Fracatorius proposed that syphilis was one of these invisible “germs”. Finally, Leeuwenhoek in the 17th century invented the microscope and provided the first evidence that these potential “germs” were actually visible. In the 150 years that followed, many showed that some of these microscopic forms were responsible for disease. However, it would be Iwanowski in 1892, who showed that substances not removed by bacterial filters could still cause disease and especially noteworthy was that these small “viruses” could not be grown in free culture like bacteria but required a cell source; alas, viruses must be cell parasites. Consistent with the chronology of discovery, it is not surprising that the first contagious diseases characterized were caused by bacteria, the organisms held back by the cotton filters. While the last portion of the 19th century and the beginning of the 20th century was dominated by the identification of bacteria-borne diseases, the last half of the 20th and the 21th century witnessed the regular discovery of diseases caused by viruses. As the rate of their discovery increased, it became known as the era of emerging viral disease.

In humans, HIV, SARS, MERS, Ebola, Marburg, Lyssa, Dengue, Hendra, Nipah and new coronaviruses emerged. Neurotropic paramyxoviruses like Japanese encephalitis viruses (JEV) and West Nile (WNV) had devastating effects. Since its appearance in 1999, WNV has caused >300 deaths in the US, and more than 50,000 have been ill from JEV, of which half were in China. Add to this the ~ 500 million cases of Dengue reported worldwide and the rise in enterovirus infections (Wilson 2013). Some human viral diseases are zoonotic and infect the species around which this International Congress revolves (Tab. 1). These include influenza A, Nipah and SARS. Nipah has led to hundreds of deaths and the slaughter of a million pigs (Sezzad et al., 2013). In a study by VanderWaal and Deen (2018) based on articles cited in PubMed, seven of the top 10 diseases of swine are of viral origin, although enteric bacteria (Salmonella and E. coli) still topped the list (Fig. 1). We did a PubMed survey for 2019 and found a major shift in apparent importance. PRRSV previously in sixth place, vaulted to first place for all swine disease, with PCV-2 close behind and Nipah moved from 37th to sixth. Figure 1 is a reminder that the once invisible and non-filtrable life form we call a virus, has become the dominant pathogen in 70% of the major diseases of swine. Figure 1 also shows that PRRSV and PCV-2 replaced E. coli and Salmonella at the top of the list. This shift illustrates that veterinary microbiology is now more concerned with viral disease than enterobacteria. Table 1 summarizes the major swine viruses, emphasizing their economic impact, whether they have zoonotic potential and whether vaccines are available (Lager and Buckley, 2019).

Table 1: Viral Pathogens of Swine

Another virus for which no vaccine exists, and which is receiving more attention is Nipah, which like influenza A, is especially zoonotic and uses fruit bats as a major vector (Fig. 1 and Table 1). Influenza A especially affects humans and swine, and some regard swine as the “mixing bowel” for the influenza A genome (Urbaniak and Markowska-Daniel, 2014). Treating influenza A in humans, requires that the CDC decides each year on a killed vaccine that their prediction models suggest will provided protection against the variants expected to be present in a particular year. Currently, much effort is underway to produce a truly universal polyvalent vaccine that might result in some unemployment at the CDC but would also result in a world safer from influenza A.

In the last half century, inefficient family farm swine production has been replaced by industrial scale farming which presents challenges in the control of especially emerging viral diseases. While improved security and use of antibiotics can often control enteric bacterial pathogens, there is no universal pharmaceutical control of viral infection, which depends on the host immune system, both active and passive. In the absence of effective vaccines, diseases like ASF can only be controlled by euthanization of infected herds and destruction of their carcasses. ASF was for a long period confined to regions of Africa, with periodic and controlled outbreaks in Spain and Portugal in 1957-1960. Sadly, ASF spread to the Caucasus and surrounding region in 2007 resulting in the destruction of >140,000 animals in Romania and >800,000 in that general region of Europe (Sanchez-Codon et al., 2018). ASF emerged in 2018 in China and has disseminated one-third of the swine population of that country (Lu et al., 2019). Without an effective vaccine, ASF remains a disease out of control. FMD is one of the longest studied animal viruses and its control history follows a similar path as now being used to control ASF. While vaccines are available, they tend to be serotype specific and lack cross protection for the seven known FMD variants (Mahapatra and Pandi, 2018). While there are a number of PRRS vaccines, full efficacy has not been obtained and like FMD, are often serotype specific, presumably because of the enormous strain diversity of this RNA virus. Currently > 30,000 variants are known but only a small number have been tested in vitro or in vivo for their virulence (Faaberg et al., 2012). The danger posed by the genetic variability of PRRSV is exacerbated by the ability of these viruses to modulate the host’s immune system. This is the major focus of this article.

Figure 1: Frequency of publications in PubMed on diseases in swine from 1996-2016 (blue) and the frequency of publication cited in PubMed in 2019 (red). In order to use a single Y-axis, values for 2019 were inflated 20-fold. Numbers in parenthesis for each pathogen indicates their rank in the 1996-2016 study

Frequency of publications in PubMed on diseases in swine from 1996-2016 (blue) and the frequency of publication cited in PubMed in 2019 (red). In order to use a single Y-axis, values for 2019 were inflated 20-fold. Numbers in parenthesis for each pathogen indicates their rank in the 1996-2016 study

The Critical Window of Immunological Development

Mammals are born with a functional cardiovascular, digestive, and skeletal-muscle systems but their immune system must develop during the period we call the Critical Window of Immunological Development (Butler et al., 2006; 2014). Mammals like swine are born with only an innate immune system, and receptors capable of distinguishing the biomolecules made by prokaryotes, i.e. bacteria and viruses, from those common among vertebrates and multicellular eukaryotes. Best known among the innate receptors that recognize prokaryotic biochemistry, are the Toll-like receptors (TLRs; Beutler and Rietschel 2003; Giradin et al., 2003). Toll, being a German word for beautiful was used to describe them when first encountered in fruit flies (Hoffmann, 1995). Table 2 summarizes the major TLRs and their specificities. In contrast to innate immunity, newborn mammals inherit only the rudimentary building blocks for developing their adaptive immune system that must be somatically generated and cannot be passed on to their offspring (Fig. 2). Adaptive immunity can be considered as the evolutionary response of higher vertebrates to the high rate of genomic change capable in bacteria and viruses (Table 3). We call this “The response of the Jedi Knight Lymphocytes to the Microbial Empire”.

Table 2: Specificity of Some Common Toll-like Receptors

Table 3: The Microbial Empire versus the Jedi Lymphocytes

Figure 2. Ontogeny of lymphocytes development from progenitor cells to virus-specific B and T cells. Pathways include those genomes encoded (green brackets), events stimulated by encounter with receptors of the innate immune system (orange brackets) and events that are somatically generated and cannot be passed to the offspring (blue brackets). [Reference Table 3]

Ontogeny of lymphocytes development from progenitor cells to virus-specific B and T cells. Pathways include those genomes encoded (green brackets), events stimulated by encounter with receptors of the innate immune system (orange brackets) and events that are somatically generated and cannot be passed to the offspring (blue brackets). [Reference Table 3]

The development of adaptive immunity in mammals, in particular in swine, requires colonization of the gastrointestinal tract (GIT) by bacteria or viruses or exposure to ligands recognized by the TLRs (Butler et al., 2002, 2005; Fig.3). This exposure triggers the somatic processes that lead to development of specific and effective protective VN antibodies (Fig.2). Development of T- and B cell repertoires in swine starts in utero (Fig. 2). B cell development begins in the fetal yolk sac, moves to the fetal liver and continues for life in the bone marrow and through the process of somatic gene recombination, an array of B cells with their B cell receptors (BCRs), generates a pre-immune repertoire (Sinkora and Butler 2016: Fig. 2). Meanwhile, the same occurs in the fetal thymus which generates a corresponding preimmune T cell repertoire (Fig.2). Development proceeds so that only B cells and especially T cell populations that do not recognize self-antigens, are allowed to survive (Fig.4). It is these surviving B- and T cells populations that are charged with the responsibility of recognizing foreign antigens in newborns. Viruses able to gain access to the fetus, such as PRRSV, parvovirus and Nipah virus, would theoretically be recognized as “self”, and thymocytes recognizing them would be deleted, thus after birth, the neonates would be “blind” to any infectious agent that had thrived in the fetus. Fortunately, the healthy fetus is normally sterile, so tolerance to pathogens does not normally develop.

Figure 3. A. Bacterial colonization is necessary to allow isolator piglets to develop adaptive immune responses to model bacterial and viral antigens. B. TLR ligands alone enable germ-free isolator piglets to make IgM and IgG antibody responses. MDP=muramyl dipeptide, a ligand for TLR 2 and CpG, a ligand for TLR 9. (Reference Table 2).

A. Bacterial colonization is necessary to allow isolator piglets to develop adaptive immune responses to model bacterial and viral antigens. B. TLR ligands alone enable germfree isolator piglets to make IgM and IgG antibody responses. MDP=muramyl dipeptide, a ligand for TLR 2 and CpG, a ligand for TLR 9. (Reference Table 2).

Figure 4. Illustration showing that developing thymocytes that are infected with PRRSV, are eliminated thus reducing the size of the peripheral T cell repertoire. Also illustrated is the apoptosis of developing thymocytes that recognize self-antigens. In infected fetuses, the antigens of pathogens would be considered as self, thus removing potential pathogen-specific T cells that emerge and populate the periphery.

Illustration showing that developing thymocytes that are infected with PRRSV, are eliminated thus reducing the size of the peripheral T cell repertoire. Also illustrated is the apoptosis of developing thymocytes that recognize selfantigens. In infected fetuses, the antigens of pathogens would be considered as self, thus removing potential pathogenspecific T cells that emerge and populate the periphery.

Developing offspring also benefit from passive adaptive immunity. In humans, this begins during fetal life by in utero transfer of maternal antibodies. In swine and other Artiodactyls and Perrisodactyls, this transfer of passive antibodies occurs only after birth via colostrum and milk (Fig.5).

All of the events described are necessary for the adaptive immune system to develop and these occur within a concentrated neonatal “window of life” that stretches from conception to weaning in piglets and to puberty in humans. Within this window, the highest frequency of mortality and morbidity occurs for offspring of all multicellular life forms including invertebrates, oak trees and flowering plants. This is also the period when infectious disease claims the offspring of vertebrates as their victims. We have called this the Critical Window of Development (Butler et al., 2006; 2014).

Figure 5. Transmission of passive immunity from mother to young among common mammals. The size of the symbols for the different immunoglobulins, e.g. IgA, IgG etc., indicates their relative levels in colostrum. From Butler 1971, Butler et al., 2017.

Transmission of passive immunity from mother to young among common mammals. The size of the symbols for the different immunoglobulins, e.g. IgA, IgG etc., indicates their relative levels in colostrum. From Butler 1971, Butler et al., 2017.

PRRS: A Porcine Pandemic

Porcine reproductive and respiratory syndrome is a pandemic caused by an Arterivirus (PRRSV) that affects all life stages, but especially animals in the Critical Window of Development. Some estimate that PRRSV is responsible for a yearly $600,000 lost to the US pork industry (Holtkamp et al., 2013). The syndrome was recognized in the 1980s and the virus identified in 1991 (Collins et al., 1992; Wensvoort et al., 1991). Since then >3,000 articles have been published on this disease it has become the major studied disease of swine (Fig.1).

PRRSV is one of several known viruses to cross the placenta that includes parvovirus and Nipah. Infected fetuses suffer from vasculitis, especially of the umbilicus which could be the cause of their death. Fetal death is not uncommon in swine, but it is not a cause of abortion. However, infection of the sow can cause abortion “storms” even if the litter is not affected, possibly through some neuro-endocrine mechanism. Since PRRSV is highly contagious, an infected sow can readily infect her healthy newborn litter causing a respiratory disease that increases the risk of secondary infection (Done and Paton 1995; Van Reeth et al., 1996, Renukaradhya et al., 2010, reviewed in Butler et al., 2014). Altogether these pathologies explain the origin of the name PRRS.

The strategy for many viruses is to modulate or dysregulate the host’s immune system, just enough to delay the anti-viral response and expulsion of the parasite, but not by killing the host; “Good parasites do no kill their hosts”. The first response of the infected host is the responsibility of the innate immune system, through recognition by TLRs, which stimulates the production of type I interferons (IFNs) which induces resistance to viral infection by inhibiting replication of the viral genome and by up-regulating MHC I expression on infected cells. This enhances the ability to display viral antigens that can be recognized by T and B cell receptors, which triggers the engagement of the adaptive immune system (Fig. 2 and 3). Suppressing MHC I expression on infected cells is a common tactic used by viruses to prevent the immune surveillance system from recognizing a virus-infected cell. Suppression of innate immunity by PRRSV is well-documented which delays a strong and effective VN-antibody response for 4-6 weeks (Lopez et al., 2007; Yoo et al., 2010; Benfield et al., 1992; reviewed in Butler et al., 2014). Specifically in the case of PRRSV, type I IFN expression is suppressed or altered (Buddaert et al., 1998; Calzada et al., 2011; Patel et al., 2010; Wang et al., 2013) but can be reversed using an adenovirus vectors expressing IFN-α (Brockmeier et al., 2009).

One of the first recorded pathological effects of PRRSV infection was thymic atrophy (Rowlands et al., 2003). This effect is virulence dependent; high path strains cause severe atrophy (Li et al., 2014; Fig. 6). In this situation, PRRSV infects the CD14+ APCs as well as CD3+ thymocytes, which causes their apoptosis (Figures 4 and 7). APCs infected in utero could present PRRSV epitopes to developing CD3+ thymocytes (Fig. 7).

Figure 6. Thymic atrophy in piglets infected with high path PRRSV (HuN4).

Studies using Landrace-Yorkshire isolator piglets and fetal piglets infected with the VR 2332 lab strain, provided evidence for a third form of immunopathology. We showed that PRRSV infection resulted in lymphoid adenopathy, hypergammaglobulinemia (Fig. 8A) and rapid differentiation of B cells into Ig-secreting plasma cells (Fig. 8B; Lemke et al., 2004; Sun et al., 2012; Sinkora et al., 2014). This rapid conversion of B cells to plasma cells correlates with the observation that <1% of the elevated IgG levels (Fig. 8A) were virus-specific (Lemke et al., 2004). We later showed that the elevated Ig came from plasma cells expressing an undiversified B cell repertoire (Butler et al., 2007; 2008). This effect can be quantified by measuring the repertoire diversification index (RDI; Fig.9).

Figure 7. TOP: CD3+ thymocytes infected with PRRSV undergo apoptosis. BOTTOM: CD14+ APCs in piglets infected with HuN4 PRRSV, are infected with PRRSV.

Discussion

We believe that all three of the immune dysregulatory effects described act together to compromise the anti-PRRSV immune response including delaying the production of effective VN-antibodies. First, by interfering with type I IFN production, innate viral immunity is suppressed which delays the recognition by the host that it is infected which delays activation of the adaptive immune system (Fig. 2). Second, by causing apoptosis of PRRSV-infected CD3+ thymocytes (Fig. 4) randomly reduces the size of the pre-immune repertoire so there would be fewer T cells that could recognize PRRSV, other pathogens and foreign antigens. Evidence in support of this contention is shown in Figure 10 in which infection with high-path PRRSV (HuN4) significantly reduces the proportion of double-positive T cells (CD4+ CD8+) compared to controls. This can also be seen after infection with the low-path CH-1a (Fig. 10B). This results in an alteration of the T cell repertoire, especially Vβ families IV-VI as shown by their restricted spectratype (Fig. 10C). The atrophy shown in Figure 6 could be explained by the loss of infected thymocytes by apoptosis. While this is one possible effect of infection of the thymus, there is yet another. We illustrate in Figure 4 that developing CD3+ thymocytes that encounter antigens during fetal development, are “educated” those eliminating all thymocytes that are presented a self-antigen by an APC during fetal life. Since PRRSV presented by APCs to developing thymocytes would be recognized as self, thereby causing their elimination, maturing T cells leaving the thymus and forming the piglets´ T cell repertoire, would be unable to recognize PRRSV as a pathogen.

The third dysregulatory impact of PRRSV that causes hypergammaglobulinemia and accelerates B cell differentiation to plasma cells (Fig. 8) might allow insufficient time for the somatic generation and selection of high affinity B cells and production of anti-PRRSV antibodies (Fig. 2). This could explain why <1% of the elevated IgG levels was specific for PRRSV (Lemke et al., 2009) and that the plasma cell population responsible is still expressing an undifferentiated B cell repertoire (Butler et al., 2007. 2008; Fig. 9). Furthermore, it is known that high serum levels of IgG are involved in a feedback loop that down-regulates B cell activity (Cerottini et al., 1969; Heyman 2003). We observed this to be the case when maternal IgG provided in colostrum, suppressed Ig synthesis by the suckling piglet (Klobasa et al., 1980). Interestingly, labile factors in colostrum also down-regulate B cell activity (Klobasa et al., 1990; Butler et al., 2019) which may explain why investigators using conventional pigs did not observe the severe hypergammaglobulinemia we observed using isolator piglets (Lemke et al., 2004; Fig.8A).

Figure 8. A. Hypergammaglobulinemia in isolator piglets infected with the VR 2332 PRRSV, SIV and PCV-2. Blue arrows with red bars indicate the means Ig levels in age-matched conventionally-reared piglets. B. The expansion of lymphocytes and B cell populations in PRRSV-infected isolator piglets compared to control animals and isolator piglets infected with swine influenza A virus (SIV). CD2- CD21+ is the phenotype for naïve B cells and CD2+ CD21- is the phenotype of plasma cells in swine. From Butler et al., 2019 and Sinkora et al., 2014.

Figure 9. Antibody repertoire development after 35 days in fetal and isolator piglets infected with PRRSV and control piglets. Note that the RDI is a log scale so that the mean RDI for PRRSV-infected fetuses is > 2-fold lower than controls. Similarly, in 35-day old isolator piglets, the RDI is nearly 10-fold lower than in colonized piglets. This is nearly 100-fold lower than colonized piglets infected with swine influenza (S-FLU). SP= spleen and represents the systemic immune system while IPP= ileal Peyer patches, and reflects the mucosal GIT response. In all cases, repertoire development in suppressed in both fetal and isolator piglets infected with VR 2332. From Sun et al., 2012.

Figure 10. The impact of PRRSV infection on the proportion of double positive (CD4+ CD8+) thymocytes in piglets infected with high-path PRRSV (HuN4), a less virulent form (CH-1a) versus in age-matched control piglets. A. Review of the pathways of thymocyte development. Double positive cells are in gray. B. The proportion of double-positive T cells in control piglets and those infected with high-path PRRSV (HuN4) or low-path (CH-1a) PRRSV. C. Spectratypic analysis of T cells from control and PRRSV-infected piglets. Brackets indicate that certain T cell clones in the Vβ IVVI families are absent from piglets infected with PRRSV. P= cells from peripheral blood; L= cells from bronchial-alveolar lavage. Roman numerals indicated different V- families of porcine T cells receptors as described by Butler et al., 2005.

IPVS - Viral Disease in Swine: An Immunologist’s Perspective with Focus on PRRS - Image 1

So, what are the major take-home messages? First PRRSV is a highly variable RNA virus capable of changing its genome and offering new challenges to the immune system. PRRSV resembles influenza A and FMD but with the added effect of immune dysregulation. A polyvalent influenza A vaccine may soon be available and perhaps such a vaccine might be developed to combat PRRS. In a recent comparative study, live PRRSVs were more effective than killed version (Toman et al., 2019). However, constructing a live polyvalent vaccine moves the issue to uncharted waters. Preventing “abortion storms” might be handled by a separate and complementary approach using hormone therapy to prevent PRRSV-infected sows from aborting their perfectly heathy litters. If placental transfer of PRRSV is responsible for infection of the thymus, which causes thymic atrophy and restricts the T cell repertoire, perhaps understanding how PRRSV is able to cross the placenta and devising a pharmaceutical antidote, might also lead to parallel therapy to control this pandemic viral disease of swine.

Published in the proceedings of the International Pig Veterinary Society Congress – IPVS2020. For information on the event, past and future editions, check out https://ipvs2022.com/en.