Swine influenza: variations on an old theme

Prof. Kristien Van Reeth
Laboratory of Virology, Faculty of Veterinary Medicine,
Ghent University, Belgium

Swine influenza (SI) is an acute viral infection of the respiratory tract and an important cause of respiratory disease in pigs. Influenza A viruses of H1N1 and H3N2 subtypes have been circulating in the European swine population for about 20 years. A swine influenza virus of H1N2 subtype, however, has emerged only recently and this has complicated the control and diagnosis of SI.

One virus, multiple subtypes
Swine influenza viruses (SIVs) are influenza A viruses and they belong to the family Orthomyxoviridae. These viruses also occur in wild birds, poultry, horses and humans, but interspecies transmission is a rare event. All influenza A viruses have the same basic structure, but there are considerable differences between influenza viruses of different species and between different “subtypes” that may co-circulate in one species. The surface of an influenza virus particle is covered by two types of glycoproteins: haemagglutinin (HA) and neuraminidase (NA). HA and NA are important because they elicit an antibody response to the influenza virus. Fifteen different HAs (H1 to H15) and 9 different NAs (N1 to N9) have been recognized by virologists and these form the basis for classification of influenza viruses into “subtypes”.
Pigs are notable among mammalian influenza virus hosts in that they harbour three different subtypes: H1N1, H1N2 and H3N2. New influenza virus subtypes can emerge in two different ways and this is referred to as antigenic “shift”. One possible mechanism is interspecies transmission. The H1N1 subtype circulating in the European swine population, for example, is originally an avian virus that has spread from wild waterfowl into the pig population in 1979. A second possibility is genetic “reassortment”. This can occur when two different influenza viruses simultaneously infect the same host. The genome of an influenza virus is “segmented” – it comprises eight single-stranded negative-sense RNA’s - and the 8 gene segments of one virus can then mix and match with that of the other virus to form a new combination virus. Both the H3N2 and the H1N2 SIVs circulating in Europe are reassortants. The H3N2 virus is human in origin and it’s HA and NA resemble those of the human virus that caused the “Hong Kong flu” pandemic in 1968. This virus has later reassorted with the “avian-like” swine H1N1 virus, from which it derived the internal proteins. The H1N2 virus was first isolated in Great Britain in 1994 and has a very complex history of origin. The virus appears to be a combination of the HA of a human H1N1 virus from the early 1980’s and the NA of the swine H3N2 virus. The internal proteins, on the other hand, are “avian-like” as in the swine H1N1 and H3N2 viruses. The European H1N2 virus, therefore, is remarkably distinct from the H1N2 virus that has been previously reported in Japan and from H1N2 in the US.
Smaller antigenic changes in the HA or NA of an already existing influenza virus subtype are called antigenic “drift”. Generally spoken, the immunity that develops after infection with a given influenza virus subtype protects only partially against a “’drift variant” and not at all against a new subtype. Antigenic drift is less common with swine than with human influenza viruses. Still, genetic reassortment of influenza viruses appears to occur more frequently in pigs than in any other influenza virus host. This can be explained by the fact that pigs are susceptible to infection with a variety of influenza A virus strains, both of avian and human origin. In addition, pigs have regular close contacts with humans and birds.

Three SIV subtypes circulate concurrently in Europe
SIV spreads readily by the air and by contact with respiratory secretions from infected pigs. It is therefore impossible to prevent SIV infections by sanitary measures alone, and the virus is enzootic in most densely swine populated regions of Europe. Table 1 shows the seroprevalence of H1N1, H3N2 and H1N2 SIVs in unvaccinated sows in Flanders, Belgium, between 2001 and 2003. In this region, only one out of 100 farms sampled was completely negative for any SIV subtype. Most individual sows had antibodies to a combination of two (48%) or to all three (31%) subtypes, indicating infection with multiple subtypes during their lifetime. A similar situation has been reported in Germany, Italy and Spain, where the three subtypes are also widespread.
Most pigs are thus born from immune sows and will have maternally-derived antibodies during their first 7 to 8 weeks of life. If pigs become exposed to the influenza virus in the presence of maternal antibodies, they may be (partially) protected against the infection and they are less likely to develop disease. Most SIV infections, therefore, occur after 10 weeks of age. By the end of the fattening period, however, pigs rarely test negative for SIV antibodies. In recent serologic examinations at the slaughterhouse in Flanders, only 9% of the 168 pigs examined were SIV negative and 44% had antibodies to two different SIV subtypes.


The disease
SIV is one of the rare primary respiratory pathogens of swine. This means that the virus can induce disease and lung lesions on its own. Typical swine “flu” outbreaks are characterized by a rapid onset of high fever, dullness, loss of appetite, labored abdominal breathing and coughing. Weight loss can be considerable, but mortality is low and recovery occurs within 7-10 days. In studies of acute respiratory disease outbreaks on pig farms in the Netherlands in 1996-1998, SIV was diagnosed in nearly 50% of the cases. All three virus subtypes have been associated with disease and there are no indications for differences in virulence between subtypes or strains. However, the infection is much more frequent than the disease. Subclinical infections are very common and many pigs become infected with the virus without ever showing clinical signs. Another possibility is that the infection with SIV does not induce the typical flu symptoms, but that it interacts with other respiratory pathogens to induce multifactorial disease. Mycoplasma hyopneumoniae, porcine reproductive and respiratory syndrome virus (PRRSV), P. multocida, H. parasuis, S. suis and B. bronchiseptica have been found in association with SIV and SIV is considered a contributor to the so-called porcine respiratory disease complex (PRDC). Also, secondary infections with bacteria may cause more severe and prolonged disease and even mortality in typical clinical cases of SI.

Understanding disease mechanisms
From the pathogenetic viewpoint, SIV is the typical example of an acute respiratory tract infection. The virus replicates in epithelial cells of the entire respiratory tract, from the nose to the deeper lungs, but almost never enters other tissues. Immunofluorescence and immunohistochemical studies have shown that bronchiolar and alveolar epithelial cells are the major target cells (Fig. 1). Epithelial cell necrosis and an influx of neutrophils into the lung accompany the typical respiratory disease. These inflammatory cells cause obstruction of the airways and substantial lung damage by release of their enzymes. Both the infection and disease are very transient. Virus excretion in nasal swabs and virus replication in the lungs last for 6-7 days at most.
Why does not every infection with SIV result in the typical clinical symptoms? Based on our experiments, we believe that disease severity relates to the amount of virus produced in the lungs and the resulting release of inflammatory mediators by the host. Under experimental conditions, the infection can be easily reproduced, but the typical disease and lung pathology only result when pigs are inoculated with high virus doses directly into the trachea. In such intratracheal infection studies, we found massive virus titers in the lungs and high levels of several cytokines, or “signal molecules”, in lung lavage fluids. The cytokines included interferon-alpha (IFN-), tumor necrosis factor-alpha (TNF-), interleukin-1 (IL-1) and IL-6. These cytokines are known to cause lung inflammation, functional lung disturbances, fever, malaise and loss of appetite, and they can strongly enhance each other’s effects. These symptoms were seen within 24 hours after intratracheal inoculation of pigs with SIV and they were associated with the peak of virus replication and peak cytokine levels. In contrast, inoculation by the less invasive intranasal or aerosol inoculation routes resulted in lower virus titers in the lungs. These infections remained clinically mild or subclinical and failed to induce the massive production of cytokines in the lungs (Fig. 2). It is noteworthy that intratracheal infection studies with other respiratory viruses, such as the porcine respiratory corona virus (PRCV) or PRRS virus, also failed to induce high levels of multiple cytokines as well as obvious respiratory disease.
Together, these data support that a massive SIV replication in the lungs is required to induce high cytokine levels and the associated disease. Any factors (partial active or passive immunity, sanitary measures…) likely to reduce the extent of virus replication are thus likely to prevent disease.


Fig. 1. Lungs of a pig 24 hours after experimental infection with SIV. (A) Immunofluorescence staining of bronchiolar epithelium, (B) Neutrophil infiltration in the bronchiolar lumen, (C) Neutrophils are the predominant cells in lung lavage fluids.

Effectiveness of the current SI vaccines

Commercial, inactivated SIV vaccines for intramuscular administration have been in use in Europe since the early 1980s. They are based on whole inactivated influenza virus, or its immunogenic proteins, HA and NA, and an oil adjuvant. Most vaccines used in Belgium and in the Netherlands contain the human New Jersey/76 (H1N1) and Port Chalmers/73 (H3N2) strains, but none of the vaccines contains an H1N2 component. The first vaccination, usually around the age of 10 weeks, should consist of two injections 3 to 4 weeks apart. In sows, bi-annual revaccinations are recommended. These inactivated vaccines rely largely on circulating antibodies to the viral HA for protection. There appears to be some diffusion of these antibodies from the circulation into the lungs, where they can neutralize the virus in case of an infection. Consequently, there is a tight correlation between HI antibody titers induced by the vaccine and protection.
Of all vaccines against respiratory viruses of pigs, SI vaccines are among the most effective. Under experimental conditions, a double vaccination of SIV seronegative pigs can provide complete protection against a severe intratracheal challenge with a high dose of H1N1 or H3N2 SIV. In our experiments, pigs that were not completely protected against the infection show a highly significant reduction of the challenge virus titer in their lungs when compared to unvaccinated control pigs. Cytokine titers in the lungs of these vaccinated pigs were also 10 to 100 times lower than in the control pigs, and all vaccinated pigs were completely protected against disease. This led us to conclude that even a minimal reduction of the influenza virus titer in the lungs of vaccinated pigs strongly reduces cytokine levels and that this reduction is sufficient to prevent the typical disease.
Another important point is that the commercial vaccines, which contain older H1N1 and H3N2 strains, do still adequately protect against the H1N1 and H3N2 strains that are currently circulating in swine. There has been much debate as to whether antigenic drift of H1N1 and H3N2 SIVs would require replacement of the vaccine strains by more recent ones, as occurs with human and equine flu vaccines. Our and other experimental studies have convincingly shown that such a replacement is not needed. One weakness of the current vaccines, however, is that they did not protect against infection or disease by the H1N2 virus in our experiments. This can be explained by the dramatic antigenic difference between the H1 of H1N1 and H1N2 viruses. As such, the antibodies induced by the H1N1 vaccine strain cannot neutralize the H1N2 virus and therefore not prevent the infection. Thus, some consideration should be given to the inclusion of an H1N2 strain in SIV vaccines for use in pigs in Europe.
In the field, the efficacy of SI vaccination will depend on numerous factors. Maternal antibodies, for example, may interfere with effective vaccination of piglets. This is one reason why SI vaccines are mainly used in sows, but in most European countries less than one quarter of the sow population is vaccinated. Though there is little published efficacy data on SI vaccination, serologic investigations have shown significantly higher H1N1 and H3N2 HI antibody titers, up to 640-2560, in vaccinated than in unvaccinated sows. Sow vaccination clearly results in higher maternal antibody levels in the young piglets, which may persist until the age of 12-14 weeks. It will control disease in suckling pigs and seems to protect pigs throughout the nursery phase.


Fig. 2.
SIV titers in the pig lung, clinical outcome and cytokine levels (log10 values) in lung lavage fluids in 3 different experimental situations: (1) intratracheal inoculation of SIV seronegative pigs, (2) intranasal inoculation of SIV seronegative pigs, (3) intratracheal inoculation of previously vaccinated pigs. The results suggest that any measures that reduce the infection pressure and/or the viral load in the lungs will reduce cytokine production and the resulting symptoms.

Some cross-protection between subtypes after infection with SIV
Apart from protection following vaccination, we have also studied the protective immune response after a true infection with SIV. We have focused on the question as to whether pigs with infection-derived immunity to H1N1 or H3N2 are still fully susceptible to the antigenically different H1N2 virus. To examine this, several groups of pigs were inoculated intranasally with two (H1N1-H1N2, H3N2-H1N2) or three subtypes (H1N1-H3N2-H1N2) one month apart. H1N2 challenge virus titers in nasal swabs were determined to evaluate protection against challenge. Serum HI antibody titers to each of the three subtypes were also determined at several time points. These data were used to gain insights in the interpretation of serological profiles from the field.
Prior infection with H1N1 or H3N2 did not protect against infection with the H1N2 virus, while H1N2 infection-immune pigs were fully protected. Still, H1N2 virus excretion was up to two days shorter in H1N1 or H3N2 immune pigs than in fully naïve pigs. This points towards a very limited and partial cross-protection between two different SIV subtypes. A surprisingly solid cross-protection to the H1N2 virus was found in pigs with infection-immunity to both H1N1 and H3N2. The H1N2 virus was not or barely detectable in nasal swabs as well as in lung tissue of these pigs. Following intratracheal H1N2 challenge, the pigs did not show any clinical signs. Despite the protection against the H1N2 subtype, these pigs only had antibodies to H1N1 and H3N2 at the time of the H1N2 challenge. There was thus no serological cross-reaction between the subtypes in the HI test.
Thus, though cross-protection between subtypes does not occur after vaccination, it may be induced by exposure to infectious SIV. This suggests that viral proteins other than the HA, which is the major immunogenic protein in the vaccines, may have some role. A mucosal or cell-mediated immune response, which is only stimulated by infection, is probably involved.
The exact significance of these experimental findings for the field situation is not yet clear. It is possible that this “heterosubtypic” protection contributes to the mild clinical course of many SIV infections in the field.

The diagnosis of SIV is relatively straightforward, but the timing of sample selection is a critical issue. Because SIV replicates in the respiratory tract for only 5 to 7 days after infection, samples should be collected from febrile, acutely affected pigs. Both lung tissue or nasal swabs can be used to isolate the virus. In the diagnostic laboratory, embryonated chicken eggs or cell cultures are used for virus isolation. Virus isolation remains among the most sensitive detection methods, but it takes at least several days and up to 1-2 weeks to determine the SIV subtype. The fluorescent antibody test on frozen sections of lung is a more rapid (2-3 hours) alternative. Because only small portions (4 mm sections) of lung are examined in this test, it is less sensitive than virus isolation. A commercial membrane immunoassay (Directigen TM, Becton Dickinson) to demonstrate influenza virus in nasal swabs is also available. This test was designed to detect influenza viruses in throat and nasal swabs of humans, but is works equally well for any influenza A virus from other species, including swine, poultry and horses. Speed is the greatest advantage of this test because it takes only 20 minutes, but it is not subtype-specific.
The haemagglutination inhibition (HI) test is the most widely used serologic test for SIV. HI antibody titres in serum peak by 2-3 weeks after an infection and they begin to decline by 3-6 months. Paired sera, collected at the time of the presumed SIV outbreak (acute serum) and approximately 3 weeks later (convalescent serum), are needed for the serologic diagnosis of SIV. If there has been a recent infection with a given SIV subtype, the convalescent sera will show rising antibody titers to that subtype. The HI test is highly subtype-specific and separate tests must be performed with H1N1, H3N2 and H1N2 strains as antigens. Furthermore, the test is most sensitive if these strains resemble the current field strains.
Our pig infection experiments with multiple SIV subtypes have confirmed the lack of serologic cross-reaction between subtypes. HI antibodies to a given SIV subtype generally point towards a previous infection with that subtype. On the other hand, the interpretation of HI antibody profiles may become much more complex if pigs have been subsequently exposed to different SIV subtypes. To illustrate, pigs that become infected with H1N1 followed by H1N2 or vice versa, may show a serologic booster to the first infecting subtype after the second infection. The other way round, seroconversion to a given SIV subtype can be less pronounced if pigs are partially protected against that subtype.

SIV: a zoonotic risk?
As mentioned higher, influenza viruses are species-specific and this is also true for swine influenza viruses. There are only a few documented cases of transmission of SIVs to humans, often children. In these rare cases, there was no further transmission of SIV between humans. For a still unknown reason, SIVs do not manage to spread efficiently in the human population. The single exception here was the so-called “New Jersey” incident in the US in 1976, during which some 500 humans became infected with a swine H1N1 influenza virus. Still, most of these infections were subclinical and there was no real swine flu pandemic. Recent serologic investigations have shown higher levels of seropositivity to H1N1 SIV in people who had close contact with pigs than in urban residents. However, these data must be interpreted with caution, because it is difficult to distinguish between antibodies to human and swine influenza viruses by the classical serologic methods.
One difference between swine and other mammalian influenza virus hosts is that swine are clearly more susceptible to infection with avian influenza viruses. In the past, it was assumed that spread of avian influenza viruses into the human population can only occur after passage or reassortment of these viruses in the pig, but this hypothesis had to be rejected recently. Indeed, some of the so-called “highly pathogenic” avian influenza viruses may occasionally transmit directly to humans. Recent examples are the avian H5N1 virus in Hong Kong in 1997 and in Vietnam and Thailandin 2004-05, as well as the H7N7 virus in The Netherlands in 2003. Interestingly, the H7N7 virus was also found in a limited number of swine at the time of the avian flu outbreak, and the H5N1 virus had been occasionally detected in pigs in China in 2001 and 2003. However, transmission of such viruses between pigs appears to be minimal and both viruses disappeared from the swine population. Most important, both swine and humans appear to have been infected directly by poultry, by close contact with infected chickens. It remains uncertain, therefore, whether swine can play a role in the transmission of avian influenza viruses to humans and this question definitely merits more research.

The establishment of the H1N2 virus in the European swine population has added an extra level of complexity to the control and serologic diagnosis of SI in Europe. To afford maximum protection, SI vaccines may need to contain all three SIV subtypes. On the other hand, some cross-subtype protection seems to occur following natural SIV infections. The presence of a third SIV subtype is thus not necessarily detrimental. Finally, continuous surveillance of the swine population for influenza viruses remains essential, because the epidemiological situation can change rapidly.

Selected references
BROWN I.H., HARRIS P.A., McCAULEY J.M., ALEXANDER D.J.: Multiple genetic reassortment of avian and human influenza A viruses in European pigs resulting in the emergence of an H1N2 virus of novel genotype. J. Gen. Virol., 1998, 79, 2947-2955.
BROWN I.H.: The epidemiology and evolution of influenza viruses in pigs. Vet. Microbiol., 2000, 74, 29-46.
CAMPITELLI L., DONATELLI I., FONI E., et al. : Continued evolution of H1N1 and H3N2 influenza viruses in pigs in Italy. Virology, 1997, 232, 310-318.
CLAAS E.C.J.: Pandemic influenza is a zoonosis, as it requires introduction of avian-like gene segments in the human population. Vet. Microbiol., 2000, 74, 133-139.
DE JONG J.C., VAN NIEUWSTADT A.P., KIMMAN T.G., et al.: Antigenic drift in swine influenza H3 haemagglutinins with implications for vaccination policy. Vaccine, 1999, 17, 1321-1328.
DE JONG J.C., HEINEN P.P., LOEFFEN W.L., et al.: Antigenic and molecular heterogeneity in recent swine influenza A (H1N1) virus isolates with possible implications for vaccination policy. Vaccine, 2001, 19, 4452-4464.
HAESEBROUCK F., PENSAERT M.B.: Effect of intratracheal challenge of fattening pigs previously immunised with an inactivated influenza H1N1 vaccine. Vet. Microbiol., 1986, 11, 239-249.
HEINEN P.P., VAN NIEUWSTADT A.P., POL J.M., et al.: Systemic and mucosal isotype-specific antibody responses in pigs to experimental influenza virus infection. Viral. Immunol., 2000, 13, 237-247.
HEINEN P.P., DE BOER-LUIJTZE E.A., BIANCHI A.T.J.: Respiratory and systemic humoral and cellular immune responses of pigs to a heterosubtypic influenza A virus infection. J. Gen. Virol., 2001, 82, 2697-2707.
HEINEN P.P., VAN NIEUWSTADT A.P., DE BOER-LUIJTZE E.A., BIANCHI A.T.J.: Analysis of the quality of protection induced by a porcine influenza A vaccine to challenge with an H3N2 virus. Vet. Immunol. Immunopathol., 2001, 82, 39-56.
LARSEN D.L., KARASIN A., ZUCKERMANN F., OLSEN C.W.: Systemic and mucosal immune responses to H1N1 influenza virus infection in pigs. Vet. Microbiol., 2000, 74, 117-131.
LOEFFEN W.L.A., KAMP E.M., STOCKHOFE-ZURWIEDEN N., et al.: Survey of infectious disease agents involved in acute respiratory disease in finishing pigs. Vet. Rec., 1999, 145, 175-180.
MAROZIN S., GREGORY V., CAMERON K., et al.: Antigenic and genetic diversity among swine influenza A H1N1 and H1N2 viruses in Europe. J. Gen. Virol., 2002, 83, 735-745.
MURTAUGH M.P., BAARSCH M.J., ZHOU Y., SCAMURRA R., LIN G.: Inflammatory cytokines in animal health and disease. Vet. Immunol. Immunopathol., 1996, 54, 45-55.
OLSEN C.W., BRAMMER L., EASTERDAY B.C., et al.: Serologic evidence of H1 swine influenza virus infection in swine farm residents and employees. Emerg. Infect. Dis., 2002, 8, 814-819.
SCHRADER C., SUSS J.: Genetic characterization of a porcine H1N2 influenza virus strain isolated in Germany. Intervirology, 2003, 46, 66-70.
SWENSON S.L., VINCENT L.L., LUTE B.M., et al.: A comparison of diagnostic assays for the detection of type A swine influenza virus from nasal swabs and lungs. J. Vet. Diagn. Invest., 2001, 13, 36-42.
VAN REETH K., NAUWYNCK H., PENSAERT M.: Bronchoalveolar interferon, tumor necrosis factor, interleukin-1, and inflammation during acute influenza in pigs: a possible model for humans? J. Infect. Dis., 1998, 177, 1076-1079.
VAN REETH K., BROWN I.H., PENSAERT M.: Isolations of H1N2 influenza A virus from pigs in Belgium. Vet. Rec., 2000, 146, 588-589.
VAN REETH K., LABARQUE G., DE CLERCQ S., PENSAERT M.: Efficacy of vaccination of pigs with different H1N1 swine influenza viruses using a recent challenge strain and different parameters of protection. Vaccine, 2001, 19, 4479-4486.
VAN REETH K., VAN GUCHT S., PENSAERT M.: Correlations between lung proinflammatory cytokine levels, virus replication and disease after swine influenza virus challenge of vaccination-immune pigs. Viral Immunol., 2002, 15, 583-594.
VAN REETH K., GREGORY V., HAY A., PENSAERT M.: Protection against a European H1N2 swine influenza virus in pigs previously infected with H1N1 and/or H3N2 subtypes. Vaccine, 2003, 21, 1375-1381.
VAN REETH K., VAN GUCHT S., PENSAERT M.: Investigations of the efficacy of European H1N1- and H3N2-based swine influenza vaccines against the novel H1N2 subtype. Vet. Rec., 2003, 153, 9-13.