key: cord-341667-ayl71jpc
authors: Van Reeth, Kristien; Nauwynck, Hans; Pensaert, Maurice
title: Bronchoalveolar Interferon-α, Tumor Necrosis Factor-α, Interleukin-1, and Inflammation during Acute Influenza in Pigs: A Possible Model for Humans?
date: 1998-04-17
journal: J Infect Dis
DOI: 10.1086/517398
sha: 
doc_id: 341667
cord_uid: ayl71jpc

Biologically active interferon-α, tumor necrosis factor-α (TNF-α), and interleukin-1 (IL-1) were detected in bronchoalveolar lavage (BAL) fluids of 3-week-old cesarian-derived colostrum-deprived pigs inoculated with H1N1 influenza virus. Cytokine titers and lung virus titers were significantly higher 18–24 h after inoculation than at 48–72 h after inoculation in all 4 litters of pigs examined. All three cytokines were positively correlated with a 3- to 4-fold increase in BAL cell numbers (P < .036) and with a drastic neutrophil infiltration (24%–77% of BAL cells vs. 0–1.5% in controls) (P < .001). In addition, cytokine production coincided with the onset of general and respiratory symptoms of influenza and with the development of a necrotizing bronchopneumonia. This study is the first demonstration of TNF-α and IL-1 in BAL fluids of a natural influenza virus host. It documents that pigs may be a highly valuable experimental model in human influenza virus pneumonia.

Biologically active interferon-a, tumor necrosis factor-a (TNF-a), and interleukin-1 (IL-1) were detected in bronchoalveolar lavage (BAL) fluids of 3-week-old cesarian-derived colostrum-deprived pigs inoculated with H1N1 influenza virus. Cytokine titers and lung virus titers were significantly higher 18 -24 h after inoculation than at 48 -72 h after inoculation in all 4 litters of pigs examined. All three cytokines were positively correlated with a 3-to 4-fold increase in BAL cell numbers (P õ .036) and with a drastic neutrophil infiltration (24% -77% of BAL cells vs. 0 -1.5% in controls) (P õ .001). In addition, cytokine production coincided with the onset of general and respiratory symptoms of influenza and with the development of a necrotizing bronchopneumonia. This study is the first demonstration of TNF-a and IL-1 in BAL fluids of a natural influenza virus host. It documents that pigs may be a highly valuable experimental model in human influenza virus pneumonia.

Swine influenza (SI) is a highly important respiratory disease onset of fever, anorexia, tachypnea, dyspnea, and coughing. of pigs. The causative viruses are type A influenza viruses of Considerable economic losses result from growth retardation H1N1 and H3N2 subtype, which are antigenically related to or weight loss. Experimental viral infections of pigs have docuhuman influenza viruses. Typical outbreaks involve an abrupt mented massive viral replication in lung epithelial cells, accompanied by polymorphonuclear leukocyte infiltration and epithelial degeneration [1] . However, the exact mechanisms by which SI virus produces lung pathology and disease have not been studied.

phragmatic lung were collected for standard histopathology, influorexia, and weight loss, particularly if produced at higher levels enza virus titrations, and fluorescent antibody stainings [7] . The [5] . Evidence for a role of TNF-a and IL-1 in influenza virus right lung was lavaged with 60 mL of PBS. Bronchoalveolar lavage pathogenesis and disease is growing [4, 6] .

(BAL) cells were separated by centrifugation at 400 g and counted,

The pathogenesis of human influenza has been studied aland cytospin preparations were stained with DiffQuik (Baxter, most exclusively in volunteers and in small laboratory animals. Here we studied the production of IFN-a, TNF-a, and IL-1 fluid samples in 96-well microtiter plates. IFN-a activity was dein the lungs of pigs and its relation with disease or inflammation termined in a cytopathic effect reduction assay by use of MDBK cells and vesicular stomatitis virus [8] . was defined as the reciprocal of the dilution producing 50% inhibition of cytopathic effect. IFN-a specificity was demonstrated by neutralization of samples with rabbit antiserum against recombi-

nant porcine IFN-a (gift from C. La Bonnardière, Jouy en Josas, France). TNF-a was assayed as cytotoxic activity in PK(15) sub-Experimental design, virologic examinations, and lung inflamclone 15 cells (gift from G. Bertoni, Bern, Switzerland) in the matory parameters. Four litters (18 total) of 3-week-old cesarianpresence of actinomycin D [9] . The plates were stained with crystal derived colostrum-deprived (CDCD) pigs were used. They were violet and read spectrophotometrically. The number of units of inoculated intratracheally with 10 7.0 EID 50 of influenza virus TNF-a per milliliter was defined as the reciprocal of the dilution (H1N1 A/Sw/Belgium/1/83), third passage in embryonated eggs.

producing 50% cytotoxicity. TNF-a specificity was established Control pigs were left either uninoculated or inoculated with PBS by neutralization with rabbit anti-human TNF-a (Innogenetics, or with sterile allantoic fluid ( in all groups. Maximum IL-1 titers were 535, 245, 520, and 150 U/mL in the 4 respective groups. Only in group 1 pigs was IL-1 also found at 72 h after inoculation.

Levels of the three cytokines were significantly higher 18 -24 h after inoculation than at 48 -72 h after inoculation (P õ Clinical responses, influenza virus titers, BAL cell numbers, percentage of neutrophils, and cytokine titers of individual pigs .016 for all three cytokines). Significant correlations were noted between production of all three cytokines and BAL cell num-are summarized in table 1.

Clinical symptoms. Control pigs remained healthy. In in-bers (P õ .036 for each) and percentage of neutrophils (P õ .001 for each). Cytokine production coincided with the onset fluenza virus -inoculated pigs, lethargy, shivering, anorexia, tachypnea, and labored abdominal respiration developed be-of clinical disease and development of bronchopneumonia. tween 18 and 24 h after inoculation. Recovery started between 48 and 72 h after inoculation. Pigs in group 1 were most Discussion severely affected.

Influenza virus replication. Control pigs tested negative for At the start of this study, it was unknown whether the lungs of 3-week-old CDCD pigs are fully capable of TNF-a and virus. Virus titers and immunofluorescence scores were similar in the 4 influenza virus -inoculated groups (P ú .577). Titers IL-1 production. Thus far, TNF-a and IL-1 have only been demonstrated in the lungs of pigs 6 -8 weeks old and older, in apical and diaphragmatic lung lobes were significantly higher 18 -24 h after inoculation than at 48 -72 h after inoculation (P either conventional [11] or gnotobiotic [12] . In our study, the use of CDCD pigs was required, since we regularly detect õ .016 and .008, respectively). Fluorescence was evident in all sections examined. Eighteen and 24 h after inoculation, TNF-a and/or IL-1 in BAL fluids from conventional pigs in the absence of experimental viral infections. The age of 3 weeks bronchi/bronchioli and alveoli had, respectively, 90% and 30% of their epithelial cells fluorescing. By 48 -72 h, more of the was selected for practical reasons. Because sanitary status and age may influence cytokine production, we performed a prelim-alveolar tissue became involved.

Lung inflammatory changes. Control pigs did not have inary experiment in 3-week-old CDCD pigs. Two pigs were inoculated intratracheally with 17 mg of Escherichia coli O111: macroscopic or microscopic lung pathology. BAL cell numbers were between 35 and 60 1 10 6 . Fewer than 1.5% of cells were B4 lipopolysaccharide, a known potent TNF-a and IL-1 inducer. BAL fluids collected 6 and 12 h after inoculation re-neutrophils; ú95% had macrophage morphology.

After influenza virus inoculation, gross lung lesions appeared vealed TNF titers of 184 and 128 U/mL and IL-1 titers of 564 and 596 U/mL in the respective bioassays. Consequently, 3-between 48 and 72 h after inoculation. At that time, Ç85%, 18%, 18%, and 8% of lung tissue was affected in groups 1, 2,

week-old CDCD pigs were found suitable for further influenza virus -cytokine studies. 3, and 4, respectively. On histopathology, bronchi/bronchioli and, to a lesser degree, alveoli showed epithelial necrosis and To our knowledge, this is the first demonstration of influenza virus -induced TNF-a and IL-1 in BAL fluids of a natural virus massive neutrophil infiltration at 18 -24 h after inoculation. Forty-eight and 72 h after inoculation, bronchioli and alveoli host. Interestingly, influenza virus was an equally effective inducer of TNF-a and IL-1 as was E. coli endotoxin. TNF-a were filled with exudate containing necrotic debris and macrophages and only few neutrophils. Histologic changes were most and IL-1 have remarkably overlapping and synergistic effects, several of which are consistent with clinicopathologic manifes-dramatic in group 1. BAL cell numbers were between 112 and 160 1 10 6 at 18 -24 h after inoculation and consisted of a tations of influenza [5] . Our findings in pigs are in agreement with previous reports in mice [3, 4] and further substantiate the maximum of 56% -77% neutrophils in groups 1, 2, and 3. In / 9d43$$ap48 03-04-98 19:03:33 jinfa UC: J Infect

Group 1 pigs, which had consistently higher values of TNF than the other groups, showed most severe disease and lesions

Effect of intratracheal challenge of fattening pigs previously immunised with an inactivated influenza H1N1 vaccine

Group 4 pigs, with barely detectable TNF levels, also had lower

Interferon production by leukocytes Interferon-a was detected at extremely high titers in some infiltrating the lungs of mice during primary influenza virus infection. pigs. As for TNF-a and IL-1, IFN-a production was tightly

Production of interleukin 1 and tumour necrosis factor activities in and peak IFN production coincided with the appearance of bronchoalveolar washings following infection of mice by influenza viillness. These findings support the idea that interferons contribrus

Peper RL, Van Campen H. Tumor necrosis factor as a mediator of inflam-IFN-a is intrinsically pyrogenic [13] and can mediate neutromation in influenza A viral pneumonia

Tumour necrosis factor and interleukin 1: cytokines with vitro that IFN-a may enhance the neutrophil respiratory burst multiple overlapping biological activities

Kluger a in pig BAL fluids, IFN could considerably add to the pyro-MJ. Thermal and behavioral effects of lipopolysaccharide and influenza genic and inflammatory effects of TNF-a and IL-1. in interleukin-1b -deficient mice

Porcine respiratory coronavirus -mediated interference against influenza virus replication in the respiratory tract the other groups. Most striking was the virtual lack of TNF of feeder pigs

High interferon titer in newborn pig intestine and previous groups cannot be attributed to differences in viral during experimentally induced viral enteritis

Improved bioassay for the belonged to a genetically different line. It would be worthwhile detection of porcine tumor necrosis factor using a homologous cell line: PK(15)

Mur-Although influenza viruses in humans primarily infect the taugh MP. Inflammatory cytokine expression in swine experimentally upper respiratory tract, influenza pneumonia yearly causes high infected with Actinobacillus pleuropneumoniae

Increased levels of tumor necrosis factor elderly. Experimental research thus remains of high priority, and interleukin 1 in bronchoalveolar lavage fluids from pigs infected and commonly used animal models have some limitations

ferrets, for example, upper respiratory tract infection predomi

Tumor necrosis factor (cachectin) is an endogenous pyrogen and induces production of innounced [15]. Mice and guinea pigs, on the other hand, are terleukin 1

Interferon-a enhances neutrophil respito them. Also, mice show a fall in body temperature instead of ratory burst responses to stimulation with influenza A virus and FMLP. fever and, with more virulent strains, the infection is invariably

Lessons for human influenza from pathogenicity studies with ferrets

hypothesis that TNF-a and IL-1 contribute to clinicopathologic effects of influenza. First of all, both cytokines were positively We thank R. Ducatelle for help with histology and D. J. Shaw correlated (r ú .879) with neutrophil recruitment to the lungs, for statistics. and peak cytokine production coincided with the onset of clinical disease. Furthermore, there was a clear association between individual TNF levels on the one hand and the extent of neutrophil infiltration and severity of lung pathology on the other.