Production of lipopolysaccharide-induced tumour necrosis factor during influenza virus infection in mice coincides with viral replication and respiratory oxidative burst

Increased morbidity and mortality occur regularly during influenza epidemics. The exact mechanisms involved are not well defined but bacterial superinfection of influenza virus infected patients is considered to play an important role. In the present study, the effect of influenza virus infection on in vivo production of turnout necrosis factor (TNF) in response to bacterial stimuli was investigated. Release of TNF in mice infected by an aerosol of influenza virus was significant after administration of bacterial lipopolysaccharide (LPS) at 72 h, whereas administration of homologous influenza virus produced only modest amounts of TNF at 96 h. Significant production of TNF was observed 48 h after intravenous administration of infectious influenza in response to LPS but not with the homologous virus. TNF induced after influenza virus infection could be blocked by a specific murine anti-TNF monoclonal antibody. Higher TNF production following aerosol influenza infection correlated with peak titres of influenza virus in the lungs of infected mice and with enhanced generation of luminoldependent chemiluminscence.


INCREASED morbidity and mortality occur regularly
Production of lipopolysaccharideduring influenza epidemics. The exact mechanisms involved are not well defined but bacterial induced tumour necrosis factor superinfection of influenza virus infected patients is considered to play.an important role. In the during influenza virus infection in present study, the effect of influenza virus infecmice coincides with viral tion on in vivo production of turnout necrosis factor (TNF) in response to bacterial stimuli was replication and respiratory investigated. Release of TNF in mice infected by oxidative burst an aerosol of influenza virus was significant after administration of bacterial lipopolysaccharide (LPS) at 72h, whereas administration of homologous influenza virus produced only modest K.N. Masihi Elderly persons and those with underlying health problems have more severe illness and Tumour necrosis factor ot (TNFa) is a pleioincreased complications from influenza. Additropic cytokine produced mainly by macrotional deaths above the normal winter increase phages.
The potent immunomodulatory occur regularly during influenza epidemics and properties of TNF include proliferation of B and are used as one of the non-virologic indicators T 2 lymphocytes, as well as augmentation of cytofor influenza surveillance. 12 Excess mortality of 3 4 56 toxic T, NK cell, and neutrophil' activities. In at least 10000 has been documented in each addition, TNF has been shown to snergize with of the 19 epidemics in the United States from other cytokines such as interferon-,. These TNF-1957 to 1986.12 The exact mechanisms involved associated host defence mechanisms can play a in increased morbidity and mortality during pivotal role in restricting the spread of microbial influenza infection are not as yet well defined. pathogens. Bacterial superinfection of influenza virus infec-There is, however, increasing evidence that ted patients is considered to play an important excessive and deleterious production of TNF can role in exacerbating virus-associated pathophybe triggered during severe episodes of certain siology during influenza epidemics. 3  bated for 18h at 37C in the presence of 5% Mice: Fourto 5-week-old female NMRI mice CO2. Following incubation, the wells were were obtained from the animal facility maintained aspirated and stained with 0.05% crystal violet in by the Federal Health Office.
20% ethanol for 15 min. After thoroughly washing the cells with distilled water, the plates Pretreatment with influenza virus: Influenza A/ were air dried. The stained cells were solubilized PR/8/34 (HIN1) was grown in the allantoic cavity by adding 250 1 of ethylene glycol monomethyl of l 1-day embryonated eggs using standard pro-ether to each well. The plates were shaken in cedures. Infectious virus was administered to the dark for l h and the optical density was mice using a 1-2 Im diameter particle aerosol of read at 600 nm in a Microplate Reader MR700 mouse-adapted influenza virus in a Middlebrook spectrophotometer (Dynatech). The percentage Airborne Infection Apparatus (Tri-R Instruments, cytotoxicity was determined. Units of TNF were Rocklle Center, USA). A 6 ml suspension of calculated by plotting the regression lines of the influenza virus was nebulized at 20C using a log of the reciprocal dilution of the test sample venturi nebulizer. The LD50 for mice was 10 -4.5 VS. the percentage cytotoxicity. One unit equals by the intranasal and 10 -v by the aerosol route, the highest reciprocal of the supernatant dilution Some groups of animals received 10 1 000 haewhich resulted in 50% lysis. Germany) containing 125 units. Test serum saline and quickly frozen at-70C. After thawing, alone, anti-TNF alone and normal mouse serum lungs were homogenized in a tissue grinder in 1 plus anti-TNF antibody were also included in the ml of cold MEM medium containing 10% fetal panel. Sera were incubated for 1 h at 37C with calf serum and antibiotics for each lung. Ten-fold gentle shaking every 15 min. The plates were dilutions of each supernatant obtained after cenplaced for 30 min at 4C and the TNF assay was trifugation of the lung homogenate were inocuperformed as described above. lated in 25 !1 amounts into confluent monolayers of MDCK cells in 24-well microtitre plates. Quadruplicate cultures were set up for each dilution.
Chemiluminescence assay: Aliquots of 500 !1 of The virus titre was determined 2 days later by spleen cell suspensions in round-bottomed vials haemadsorption with chicken red blood cells, were mixed with 10 !1 of luminol at a con- The results are expressed as logm of 50% tissue centration of 1 mg/ml in phosphate buffered culture infective doses (TClD) calculated accord-saline containing 0.4 triethylamine and incubated ing to the Reed and Muench method, at 37C for 10 min. Chemiluminescence (CL) measurements were performed at 37 C in a Treatment with lipopolysaccharide (LPS): Mice specially developed Biolumat model 9505 pretreated with influenza virus were administered (Berthold, Wildbad, Germany) which permits 25 l.tg of phenol-chloroforom-petroleum ether simultaneous reading of six samples. After the extracted LPS from Salmonella minnesota measurement of background for 3 min, CL was (Sigma Chemical Company) by the intravenous generated by addition of 10 I1 of particulate route at various time intervals. Animals were bled stimuli zymosan (Sigma) suspended in buffered from the retro-orbital plexus 1.5-2h later and saline at a concentration of 50 mg/ml. CL was sera were stored at-20C, continuously monitored on a programmed microcomputer.
TNF bioassay: TNF activity in serum samples was determined by a standard fibroblast cytotoxicity Statistical analysis: Results are expressed as assay. Briefly, 100 I1 of 2.5 x 10 4 cells per ml of mean 4-S. D. Statistical comparison with approa sensitive L929 cell line were added to 96-well priate control groups was performed using twomicrotitre plates (NUnc) and incubated overnight tailed Student's t-test. 7 The specificity of TNF induced by influenza the lungs was determined by titration of lung virus was tested using an anti-TNF monoclonal homogenates on confluent monolayers of MDCK antibody. TNF induced 72 h after influenza virus cells. The results presented in Fig. 3 show the infection could be blocked by a specific murine presence of detectable virus at all time periods anti-TNF monoclonal antibody, as is shown in tested. Infectivity titres from the lungs of mice of African swine fever virus production. 25 The kinetics of TNF production by the influ-Discussion enza virus exhibited a marked dependence on the time elapsed after influenza infection and Influenza virus infection can predispose the administration of LPS. Significant TNF production host to bacterial superinfection. Human and was evident 72h after aerosol virus infection. experimental animals infected with influenza TNF can profoundly affect the functional caphave been shown to have increased incidence of ability of phagocytic cells, and has been shown bacterial infections due to diverse bacteria to enhance the response of neutrophils to parti- 26 including Streptococcus species, Escherichia coli culate stimuli, 5 alter surface receptors and and Haemophilus influenzae. 1 The results of prime for enhanced production of oxygen radithe present study demonstrate that in vivo influ-cals. 5' Administration of TNF in mice can result enza virus infection by the natural aerosol route in an increase of the percentage of splenic maccan prime the host for subsequent release of rophages. 27 In the present study, a high level of TNF in response to bacterial stimuli and also splenic cell CL activity was observed at 72h in extend previous in vitro observations showing influenza virus-infected animals that had received LPS-induced potentiation of the production of LPS. It is noteworthy that the time period of TNF from influenza virus-infected macrophages. 17 enhanced CL coincides with the peak TNF pro-enza virus which were also detected 72 h after infection.
Influenza virus or viral components have been shown to stimulate the production of interferon-728 and IL-1.16 '29 In numerous studies, TNF has been shown to synergize with other cytokines including interferon-7. The intricate synergistic interaction between the putative cytokines released during the period of peak influenza virus production and resultant sequelae remain to be elucidated. The results of the present study suggests that the TNF response of the influenza virus infected host on the subsequent homologous influenza virus challenge is of a low magnitude. This could be a reflection of the protection afforded by primary influenza infection. In contrast, high levels of TNF were induced in influenza virus infected animals administered bacterial LPS. An understanding of the kinetics required for the influenza virus to sensitize a host for induction of TNF in response to bacterial stimuli may help define the critical period when complications could occur in institutionalized elderly patients and other persons at risk during influenza epidemics.