|Home | About | Journals | Submit | Contact Us | Français|
The 1918 pandemic influenza virus has demonstrated significant pathogenicity in animal models and is the progenitor of ‘classical’ swine and modern seasonal human H1N1 lineages. Here we characterize the pathogenicity of an early ‘classical’ swine H1N1 influenza A virus isolated in 1931 compared to the pathogenicity of the 1918 pandemic virus and a seasonal H1N1 virus in mice and ferrets. A/Swine/Iowa/31 (Sw31) and the 1918 influenza viruses were uniformly lethal in mice at low doses and produced severe lung pathology. In ferrets, Sw31 and 1918 influenza viruses caused severe clinical disease and lung pathology with necrotizing bronchiolitis and alveolitis. The modern H1N1 virus caused little disease in either animal model. These findings revealed that in these models the virulence factors of the 1918 influenza virus are likely preserved in the Sw31 virus, and suggest that early swine viruses may be a good surrogate model to study 1918 virulence and pathogenesis.
Influenza A viruses have unpredictably caused pandemics throughout human history (Taubenberger and Morens, 2009) and today cause approximately 36,000 deaths in the US annually in inter-pandemic form (Thompson et al., 2003). In addition to humans, influenza A viruses infect multiple avian and mammalian species, including horses, swine, and poultry (Webster et al., 1992). Swine have been frequently infected with different strains of influenza globally (Alexander and Brown, 2000; Brown, 2000; Brown et al., 1998; Thacker and Janke, 2008; Van Reeth, 2007; Vincent et al., 2008). Interestingly, the first recognized swine influenza infections were identified clinically in 1918, during the largest known influenza pandemic in humans (Dimoch, 1918-1919; Koen, 1919). Since 1918 swine influenza viruses have continued to cause disease in populations of pigs (Vincent et al., 2008) worldwide, and much like human influenza (Rambaut et al., 2008), swine influenza viruses have been shown to be evolving dynamically by both antigenic drift and frequent reassortment (Brown et al., 1997; Dunham et al., 2009).
The human 1918 pandemic influenza virus has been reconstructed and characterized (Taubenberger, Hultin, and Morens, 2007; Taubenberger and Morens, 2006; Taubenberger et al., 2005; Tumpey et al., 2005). Its pathogenicity has been studied in mice (Kash et al., 2006; Pappas et al., 2008; Tumpey et al., 2005), macaques (Kobasa et al., 2007), guinea pigs (Van Hoeven et al., 2009), swine (Weingartl et al., 2009), and ferrets (Tumpey et al., 2007). In comparison to contemporary human H1N1 viruses, this virus has been shown to be extremely virulent in mice and macaques causing severe disease and high mortality. It also causes significant disease in ferrets and pigs, although it has been reported to be less lethal in these models. The virulence factors that increased the lethality of the 1918 virus in humans during the pandemic are the subject of continued investigation (Morens, Taubenberger, and Fauci, 2008). Characterizing these virulence factors is essential to anticipating how current and future pandemic influenza viruses arise and cause significant human disease, as well as to provide a basis for development of novel therapeutics.
While clinically identified in pigs in 1918 (Dimoch, 1918-1919; Koen, 1919), swine influenza viruses were first isolated from pigs in 1930 and until 1998 had been primarily of the ‘classical’ H1N1 lineage . Similar disease to what was observed in humans was seen in pigs during the 1918 pandemic, possibly due to concurrent infection of pigs and humans. Sequence analysis has shown that the classical swine influenza viruses first isolated in 1930s were direct descendants of the 1918 pandemic virus (Taubenberger and Morens, 2006; Taubenberger, Reid, and Fanning, 2000). Thus, swine H1N1 influenza viruses isolated in the 1930’s are the closest naturally occurring relatives to the 1918 pandemic virus. Patrick Laidlaw, one of the co-authors of the 1933 paper describing the isolation of the first human influenza A virus (Smith, Andrewes, and Laidlaw, 1933), wrote in 1935, “the virus of swine influenza is really the virus of the great pandemic of 1918, adapted to the pig and persisting in that species ever since” (Laidlaw, 1935).
Swine influenza viruses have caused sporadic human infections over the past decades (Gray et al., 2007; Myers, Olsen, and Gray, 2007; Vincent et al., 2009). A classical swine H1N1 virus caused the well-known outbreak at Ft. Dix in New Jersey in 1976 (Gaydos et al., 2006). Recently in 2009, a large-scale human outbreak of a novel swine influenza lineage has developed into the first pandemic of the 21st century (Garten et al., 2009; Smith et al., 2009; WHO, 2009). Thus, a better understanding of the role that influenza A virus lineages and evolution in swine play in the formation of strains capable of infecting humans remains vitally important for pandemic preparedness (Dunham et al., 2009; Morens, Taubenberger, and Fauci, 2009).
To determine if H1N1 viruses isolated in the 1930s still retain some of the virulence properties of the 1918 influenza virus, we compared the pathogenicity of the reconstructed 1918 influenza virus with a classical swine H1N1 virus isolated in 1931, A/Iowa/swine/31 (Sw31), in ferrets and mice.
A/NY/312/2001/H1N1 [NY312] was obtained from the New York State Department of Health Wadsworth Center—Griffin Laboratory in Slingerlands, NY and passaged three times (Qi et al., 2009). A/Iowa/Swine/1931 [Sw31] was obtained from Jack Bennick in the Laboratory of Viral Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health. The 1918 virus containing the A/South Carolina/1/1918 HA plus 7 segments from A/Brevig Mission/1/1918 was rescued using standard technique at the Centers for Disease Control and Prevention (CDC) Laboratory in Atlanta, GA (Tumpey et al., 2005). Viruses were passaged in MDCK cells (ATCC CCL-34) in Dulbecco’s Modified Eagle’s Medium (DMEM, Quality Biological Inc., Gaithersburg, MD) in the presence of 1ug/ml TPCK-Trypsin (Sigma-Aldrich, St. Louis, MO). Viral titers were determined by plaque assay in MDCK cells, and expressed as plaque forming units per milliliter (PFU/ml), as previously described (Qi et al., 2009).
All animal experiments at the NIH were performed following NIH Institutional Animal Care and Use Committee approved protocols and guidelines in an ABSL2 or ABSL3 facility. All animal experiments at the CDC were performed following CDC Institutional Animal Care and Use Committee approved protocols and guidelines in an enhanced ABSL3 facility.
Groups of 8-10 week old female BALB/c mice (The Jackson Laboratory, Bar Harbor, ME) were lightly anesthetized in a chamber with isoflurane supplemented with O2 (1.5 L/min) and were intranasally inoculated with between 102 and 105 PFU of influenza A virus in a total volume of 25 to 50 μL under enhanced ABSL3 conditions. Body weights were measured daily and mice were humanely euthanized if they lost more than 20% of their starting body weight. Lungs were collected for pathology at 4 and 6 days post-infection (dpi). For each virus and time point, lungs from 3 animals were collected for pathology. Lungs collected for pathology were inflated with 10% neutral buffered formalin at time of isolation to prevent atelectasis.
Inoculation of ferrets with NY312 and Sw31 was performed at the NIH and 1918 virus inoculation took place at the CDC. Prior to inoculation, serum samples were analyzed for the presence of neutralizing antibodies against A/New Caledonia/20/00/H1N1, A/NY/55/04/H3N2, and B/Jiangsu/10/03. Animals with no evidence of neutralizing antibodies to the above influenza A virus reference strains were used in this study. Sixteen six-month-old male ferrets were allowed to acclimate for five days, and then were weighed and examined. Ferrets were anesthetized using Ketamine (100mg/ml, Vedco, St. Joseph, MO), Atropine (0.4mg/ml, Baxter Healthcare, Deerfield, IL), and Xylazine (100mg/ml, Akom, Decatur, IL) and then inoculated intranasally with 106 PFU of virus in 1ml of DMEM (500ul per nostril). Four ferrets received the Sw31 virus, four received NY312, and four received DMEM at the NIH. Four ferrets received the 1918 virus at the equivalent dose at the CDC.
Body temperature, weight, and physical exam were assessed daily. A clinical score was assigned from 0-5 by the investigator based on five criteria. One point was assigned for each of the following: nasal crusting or discharge, decreased activity, decreased eating habits or diarrhea, active sneezing, and respiratory distress. The score was calculated as a sum of the assigned points and recorded.
On 4 and 14 dpi, two ferrets from each group were humanely euthanized and necropsy performed. Whole lungs were harvested, half the lung from each animal was frozen for virus titration, and the rest was fixed in 10% neutral-buffered formalin for histopathological analysis. Nasal turbinate tissue was harvested and stored for virus titration.
Frozen ferret tissues were thawed, weighed, and homogenized in sterile cold Leibovitz’s medium (Gibco L-15, Invitrogen Corporation Carlsbad, CA) containing 1X penicillin, streptomycin, and amphotericin B (Gibco Anti-Anti, Invitrogen Corporation Carlsbad, CA). Solid debris was pelleted by centrifugation and viral titer determined using standard plaque assay technique on MDCK cells. Lung titers are expressed as plaque forming units per gram of tissue (PFU/g); nasal turbinate titers are expressed as total plaque forming units (PFU).
Tissue samples, including the fixed, inflated lungs, mediastinal lymph nodes, and spleen were dehydrated, embedded in paraffin, and 5 μm sections on positively charged slides were stained with hematoxylin and eosin for histopathologic examination. Influenza virus antigen distribution was evaluated by immunohistochemistry using manufacturer’s protocol (Invitrogen Corp., Carlsbad, CA). Antigen retrieval was performed in 10mM Sodium Citrate, 0.05% Tween 20, buffer, pH 6.0, using the 2100 Retriever model pressure cooker following the manufacturer’s instructions (Pickcell Laboratories, Amsterdam, Netherlands). The primary antibody was an anti-influenza A goat polyclonal (Abcam Inc., Cambridge, MA) and was detected using a biotinylated rabbit anti-goat IgG (Invitrogen Corp., Carlsbad, CA); the chromogen was Aminoethylcarbazole (AEC). A single pathologist reviewed the histopathology and immunohistochemistry in a blinded fashion.
Differences in tissue viral titers and clinical parameters were tested for significance using the Student’s t test. 95% confidence intervals were generated to evaluate differences in tissue viral titers. Graphpad Prism software was used for all statistical analysis.
Sequences were downloaded from the NCBI Influenza Virus Resource (http://www.ncbi.nlm.nih.gov/genomes/FLU/FLU.html). Alignments were generated using the Megalign program (Lasergene 7.2, DNAStar, Madison, WI).
The ten major open reading frames of the influenza virus (PB2, PB1, PA, HA, NP, NA, M1, M2, NS1, NS2 (NEP)) consist in aggregate of 4468 codons. The Sw31 virus genome has a 96.3% amino acid identity with the 1918 virus genome.
To determine the murine pathogenicity of Sw31, mice were inoculated with 105 PFU of Sw31 or a recent human H1N1 virus (A/New York/312/2001; NY312, as previously described (Qi et al., 2009)). Mice inoculated with Sw31 virus became acutely ill within 2 to 3 days as assessed by rapid weight loss, and 100% of animals reached end-point criteria on average by day 5. In contrast, mice inoculated with NY312 became mildly ill and showed minimal weight loss with recovery by day 7 (Figure 1). Sw31 showed 100% mortality at the lowest dose tested showing the Sw31 LD50 < 102 PFU; while NY312 was non-lethal at the highest dose tested (LD50 > 5×105 PFU).
Mice inoculated with the Sw31 virus developed severe disease, similar to disease previously described in 1918 inoculated mice (Kash et al., 2006; Tumpey et al., 2005). Histopathologic changes at 4 dpi with the Sw31 virus similarly showed moderate-to-marked necrotizing bronchiolitis and alveolitis, with a neutrophil predominant inflammatory infiltrate, and areas of acute alveolar edema and hemorrhage (Figure 2) that were very similar to findings reported for 1918 (Kash et al., 2006; Tumpey et al., 2005).
Signs of clinical illness were observed after 24 hours in all ferrets inoculated with virus. Approximately 5% weight loss was observed by the end of 14 days in all ferrets inoculated with the NY312 virus. The 1918 and Sw31 inoculated ferrets demonstrated significantly more weight loss than NY312 by day 3 (p=0.0042 and 0.0059 respectively), and reached nearly 15% weight loss by days 6-7 (Figure 3A). Mean weight loss was similar between the two groups throughout the infection and no significant difference was seen between weights at the of peak weight loss on days 4, 5, 6, or 7 (p=0.4639, 0.4795, 0.8818, 0.2818 respectively). Both groups of ferrets showed similar trends of recovery and weight gain after day 7. The control ferrets showed consistent weight gain (Figure 3A).
Ferrets inoculated with Sw31 reached peak body temperature elevation (+2.125°C) by day 2, significantly different from NY312 (+0.750°C) and control (+0.675°C, p=0.0056, 0.0005 respectively). The 1918 virus infection caused ferrets to reach a peak temperature elevation (+2.250°C) on day 4 similar to that of Sw31 on day 2 (p=0.3890). NY312 caused no significant increase of body temperature from baseline at any time during the infection. The body temperature change in both Sw31 and the 1918 virus inoculated ferrets after day 5 were similar to that of the control animals and remained so for the duration of the experiment (Figure 3B).
All virus inoculated ferrets developed symptoms. Ferrets inoculated with the 1918 and Sw31 virus reached similar mean peak clinical scores of 5.00 and 4.75 respectively (p=0.5415). These were significantly higher than the peak clinical score reached by ferrets inoculated with NY312 (1.75; p=0.0014, 0.0106). Sw31 and NY312 inoculated ferrets exhibited higher clinical scores earlier, with Sw31 inoculated ferrets reaching peak clinical score 1 day earlier than 1918 virus inoculated ferrets. One of the two ferrets in the 1918 group that was followed beyond day 4 was found dead on day 8. The other ferret in the 1918 group as well as all ferrets in the other groups survived until the end of the experiment on day 14. After day 4 the NY312 inoculated ferrets began to have a reduction in clinical score, and after day 6, both the surviving 1918 and Sw31 inoculated ferrets showed a similar reduction in clinical score (Figure 4).
All three viruses were recovered from the nasal turbinates on day 4. The 1918 and Sw31 viruses grew to a 2.5 to 4 log higher titer in the nasal turbinates on day 4 than did NY312 (Figure 5A). Although the 1918 virus grew to 1.5 log higher titer from the nasal turbinates than Sw31 on day 4, statistical significance was not achieved (p=0.0971). No virus was detected on day 14 in the nasal turbinates of any animal (data not shown).
The 1918 and Sw31 viruses grew to similar titers from the lung tissue on day 4 (1.8×106 PFU/g, 2.0×105 PFU/g; p=0.1135). This was 3.5 to 5 logs higher than NY312 (88 PFU/g) on day 4 (Figure 5B). No virus was recovered from the lungs of any animal on day 14 (data not shown).
Histopathological examination of lung tissue on day 4 revealed a remarkably similar pattern of pathology in the 1918 and Sw31 inoculated ferrets (Figure 6). There was a moderate-to-marked necrotizing bronchiolitis and alveolitis, consisting of a mixed inflammatory infiltrate composed predominantly of neutrophils. Focal areas of acute alveolar edema and hemorrhage were also noted. Widespread viral antigen was detected in bronchiolar epithelial cells by immunohistochemistry in both the lungs of 1918 and Sw31 inoculated ferrets (Figure 6).
The lungs of NY312 inoculated ferrets in contrast showed minimal histopathologic changes, consisting of mild focal bronchiolitis without alveolitis. A single focus of acute alveolar edema was observed. There was also focal acute inflammation of the submucosal glands of the trachea and mainstem bronchi (data not shown). The control ferrets displayed no histopathologic abnormalities.
On day 14, the 1918 and Sw31 inoculated ferret lung sections showed a very similar pathology, consisting of focal interstitial fibrosis and chronic inflammation with remodeling. A single focus of bronchiolitis obliterans with organizing pneumonia was seen in one Sw31 animal. The day 14 lung tissues of both the NY312 and control ferrets showed no pathologic changes.
The virulence of the reconstructed 1918 virus has been extensively demonstrated in animal models, and efforts to understand the molecular basis of this virulence are ongoing (Pappas et al., 2008; Qi et al., 2009). In this study we demonstrate that this virulence is not unique to the 1918 virus, but that a descendant influenza A/H1N1 virus isolated from swine in 1931 demonstrates similar virulence and pathogenicity in mice and ferrets.
Previously published studies have shown that the 1918 virus causes severe pathology in mice and has an LD50 of 103.25 - 103.5 PFU (Kash et al., 2006; Pappas et al., 2008; Tumpey et al., 2005). The histopathologic changes in the respiratory tree of mice infected with the 1918 virus have been previously described (Kash et al., 2006; Tumpey et al., 2005) and consist of moderate-to-marked necrotizing bronchiolitis and alveolitis, with a neutrophil predominant inflammatory infiltrate. Acute alveolar edema and hemorrhage was also noted. We have demonstrated that the LD50 of Sw31 in mice is even lower than what has been previously described for the 1918 virus, and that lung histopathology during Sw31 and 1918 infections are identical.
In ferrets, the similarities between the 1918 virus and Sw31 were also apparent as we observed nearly identical disease course and histopathologic phenotypes. No significant difference was observed in clinical disease, viral growth in nasal turbinates and lungs on day 4, or lung pathology. Although 1 ferret died on day 8 following infection with 1918, ferrets inoculated with Sw31 were given aggressive supportive care consisting of special food and hydration by the animal facility veterinarian while those inoculated with 1918 were not. This may have reduced morbidity and mortality in the SW31 group.
In stark contrast to what was seen in the SW31 and 1918 groups in ferrets, mild clinical disease, low viral tissue titers, and little clinical signs of disease were seen in ferrets inoculated with the NY312 virus, a contemporary human H1N1 virus.
The evolution of swine influenza viruses has progressed at a slower rate than in humans (Taubenberger, Reid, and Fanning, 2000; Taubenberger et al., 2001), and the similarity of pathogenicity in mice and ferrets reported here suggests that virulence factors of the 1918 virus may have been retained in early classical swine viruses such as Sw31. In contrast, the NY312 H1N1 virus, having been subjected to decades longer, and comparatively more rapid, evolution in humans, appears to have lost the pathogenicity of its predecessors. Because of techniques at the time of its isolation, Sw31 is likely to have been passaged through mice and eggs; therefore it is possible that laboratory adaptation could account for some of its virulence in mice. Despite this possibility, there is an overall phenotypic similarity between the reconstructed 1918 virus and the Sw31 virus across animal models, and this should be considered as we continue to try to determine the virulence factors that may have contributed to the high morbidity and mortality of the 1918 pandemic.
Influenza A virus receptor-binding specificity is mediated by the viral hemagglutinin (HA), which binds to receptors containing terminal sialic acids on cell surface glycans. The reconstructed 1918 virus used in this study (Tumpey et al., 2005) contained the A/South Carolina/1/1918 hemagglutinin (HA) gene (Reid et al., 1999) that has been demonstrated to have a receptor-binding specificity for α2-6 linked sialic acids (Stevens et al., 2006). Classical swine H1N1 virus HA proteins, including Sw31, have been shown to have a blended receptor-binding specificity for both α2-6 and α2-3 linked sialic acids (Gambaryan et al., 1997). Previous work has shown that the 1918 virus had two co-circulating HA variants that differed in receptor-binding specificity (Reid et al., 2003; Stevens et al., 2006), the South Carolina HA variant (with Asp190, Asp225) bound exclusively to α2-6 receptors, while the New York variant, which differed only by one residue (Gly225), had mixed α2-6/α2-3 specificity. The two 1918 variants had similar pathogenicity in ferret and mouse models (Tumpey et al., 2007; Qi et al., 2009). Since the Sw31 virus has an HA specificity like that of the 1918 New York HA, it is unlikely that differences in receptor specificity are playing a key role in pathogenicity in the ferret model.
In the present study, Sw31 caused significant disease in both mice and ferrets, and exhibited severe mortality in mice and high levels of viral replication in the lungs of ferrets associated with significant temperature elevation, weight loss, and clinical disease. The phenotypic features of this virus are strikingly similar to those of the 1918 virus, demonstrating that the 1918 virus is not uniquely virulent in these models. Thus Sw31 likely contains important virulence factors that may have been maintained in other H1N1 influenza viruses isolated after 1918. Thorough and careful investigation of the conserved virulence factors present in this and other viruses from this time period is important in furthering our understanding of influenza virus virulence, host adaptation, and evolution.
The Intramural Research Program of the NIH and the NIAID supported this work.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.