Pathological Characteristic And Complement-mediated Pathogenesis In Mice With Highly Pathogenic Avian Inlfuenza H5N1Virus Infection

Objective: A high fatality rate of the highly pathogenic avian influenza (HAPI) H5N1virus and the frequent emerging influenza viruses have caused the public health concern.However, the pathological and physiological injury after avian influenza virus infectionhave not been fully clarified, while the pathological characteristic and host immuneresponse after different influenza viruses infection need to be studied. Furthermore,although the disregulated host innate immune response has been considered the cause ofacute lung injury/acute respiratory distress syndrome (ALI/ARDS) after avian influenzavirus infection, it is not well understood the role of abberant complement activation onthe lung injury with H5N1virus infection. In our study, influenza mouse models wereestablished with infection of HAPI H5N1virus, and the pathological characteristic andhost immune response were analyzed comparatively among H5N1virus, H7N9virus orH1N1virus respectively. Furthermore, the mechanism of abberant complementactivation mediated acute lung injury was investigated in detail using the HAPI H5N1virus-infected model. Collectively, the study indicated that excessive complementactivation plays an important role in mediating H5N1-induced ALI and that inhibitionof complement may be an effective clinical intervention and adjunctive treatment forH5N1-induced ALI.Methods:(1) Establishment of mouse model infected with HAPI H5N1virus. Mice werechallenged intranasally with1050%lethal doses of H5N1virus with a volume of20l in an approved biosafety level3facility. All mice were killed at1d,3d and5drespectively, the control mice were killed at the indicate times. The survival rate andclinical symptoms of mice were evaluated, while the virus prederendum, virus titer anddistribution in tissue with H5N1virus infection were preformed by MDCK cells andimmunohistochemistry. Furthermore, pathological lesions of lung were analyzed byhematoxylin-eosin (H&E) staining, the infiltration of neutrophils were measured byimmunohistochemistry. To further identify the imflammatory systematically,inflammatory cytokines sucha as interleukin6(IL-6) and tumor necrosis factor α (TNFα) and lung tissue myeloperoxidase (MPO) were measured by ELISA assay, and theinterferon α/β R and interferon γ were analyzed by relative real time PCR.(2) Comparative analysis of pathological characteristic of mice infected withH5N1virus, H7N9virus and H1N1virus. Mice were challenged intranasally with1050%lethal doses of H7N9, H5N1or H1N1virus respectively in an approved biosafetylevel3facility. The survival rate, clinical symptoms, virus distribution, pathologicallesions, neutrophils infiltration and proliferating cell nuclear antigen (PCNA) wereanalized comparativelyafter virus infection.(3) Complement activation in mice after H5N1virus infection. The deposition ofcomplement C3and C5b-9and MBL-c composition in lung tissue were detected byimmunohistochemistry assay using the ALI mouse model induced by H5N1virusinfection, while the receptors of complement activation product such as C3aR andC5aR were also presented by the same assay. In addition, the transcription level ofC3aR, C5aR and MBL-associated serine protein2(MASP2) were also detected byreal time PCR assay.(4) Protective role of complement activation inhibition in ALI/ARDS mouse modelinfection with H5N1virus. To further identify the protective role of complementactivation inhibition after H5N1virus infection, cobra venom factor (CVF), C3a receptor antagonist (C3aRa) and anti-C5a antibody (Anti-C5a Ab) were used toexhaust the complement, inhibit the combination of C3a and C3aR, and block C5aactivity, respectively. Mice were challenged intranasally with1050%lethal doses ofH5N1virus and divided into four groups,①CVF group: mice were intravenouslyadministered CVF (2.5ug/mouse) at24and48hours before H5N1virus infectionrespectively.②C3aRa group, mice were treated intravenously with C3aRa (1mg/kg)immediately after H5N1virus infection.③Anti-C5a Ab group: mice were treatedintravenously with Anti-C5a Ab (400ug/mouse) immediately after H5N1viruschallenge.④PBS group: mice were treated intravenously with the same volume ofPBS after H5N1virus challenge. The analysis and comparison of survival rate, clinicalsymptoms, virus titer, pathological lesions, neutrophils infiltration and host immuneresponses between treatment group and infected only group were presented usingseveral assays according the above methods.Result:(1) To systematically evaluate the pathological and physiological characteristic afterH5N1virus challenge, we first established a mouse model of H5N1virus inducedALI/ARDS which was similar to clinical patients infected by H5N1virus. Comparedwith control mice received saline only, mice at3days after H5N1virus challengestarted to present with early symptoms and become more severe at5days whichpresented with ruffled fur, hunched back, and reduced activity, and absolutely diedwithin10days. The virus replication reach to peak level in mouse lungs at5dayspost-challenge and the virus distribution was migrate from trachea in early stage topulmonary in later period. There was accompanied by pulmonary histopathologicaldamages similar to clinical signs. Only some parenchymal expansion and low levels ofneutrophil congestion were seen at1day after H5N1virus infection, and At Day5, lung damage was more severe with high numbers of alveolar epithelial cells showingdegeneration and collapse, infiltration of inflammatory cells accompanied by largequantities of exudates, and severe edema, especially around vessels, and minor lesionwere found in other organs such as brain, spleen and liver. Furthermore, low numbersof infiltrating neutrophils were observed close to vessels at1day after H5N1virusinfection, and at5days, a significantly increased neutrophil infiltration was occurred,especially in parenchymal lung tissue. MPO activity was consistent with neutrophilsinfiltration. Serum concentrations of the inflammatory cytokines TNF-α and IL-6werealso significantly higher at5days after H5N1virus infection compared with the controlmice. Finally, relative mRNA expression of IFN-γ was increasing significantly at1and5days after infection, but the expression of IFN-α/βR mRNA decreased after H5N1infection. These data demonstrate that the pathogen and inflammatory of H5N1infected mouse model was not only similar to clinical features of patients infected byH5N1virus, but also represent a potency to explore pathogenic mechanism andevaluate relative vaccines and drugs.(2) Furthermore, the pathological and physiological indicator were analyzedcomparatively in H5N1, H7N9or H1N1virus infected mice model. No death case wasfound in H7N9infected mice, but all mice died after H5N1and H1N1virus challengeand the survival time infected with H5N1virus was shorter than mice after H1N1virusinfection. The distributions of the virus antigens were mainly on the bronchialepithelial cells, some stromal cells and alveolar epithelial cells. Lung injury of micewhich presented vasculitial and interstitial pneumonia infected with H7N9virus wasslighter than that of mice infected with H5N1or H1N1virus induced ALI/ARDS,while the damage of spleen in H5N1and H1N1virus inoculated mice was more severethan that in H7N9virus inoculated mice on5days after virus inoculation. There wasmore neutrophils infiltration and lower PCNA expression in the lung of H5N1or H1N1virus infected mice than that of H7N9virus infected mice. These results provideuseful information for the further study on the virulence of the avian influenza and theexploration of pathogenic mechanism after influenza virus infection.(3) To determine whether complement activation plays a key role in the occurrence ofALI induced by H5N1virus infection, we detected complement activation productsand the expression of anaphylatoxin receptors in the lungs of mice challenged withH5N1virus. Less deposition of C3was observed at the location of bronchial epithelialcells1day after challenge, but later C3deposition increased with time, especially onepithelial cells of bronchioles and on terminal bronchioles, pneumocytes, andinfiltrating lymphocytes. Similar to C3, deposition of C5b-9or MBL-C and expressionof C3aR or C5aR was seen in lung tissues at Day1after challenge. In addition,deposition of C5b-9, MBL-C and expression of C3aR and C5aR in lung tissuesincreased gradually in the following days until Day5after H5N1infection, especiallythe presence of pneumocytes and infiltrating lymphocytes. To confirm theimmunohistochemical observations, we analyzed the gene expression of C3aR, C5aR,and MASP2in lung tissues by relative real-time quantitative PCR after H5N1viruschallenge. The expression of C3aR and C5aR mRNA was increased at Day1afterchallenge, and expression levels of C5aR mRNA increased gradually in the followingdays. The expression of MASP2also increased significantly at Day1after H5N1infection and increased in the following days. These data indicate that complement ismassively activated after H5N1infection and that, based on the deposition of MBL-Cand increased expression of MASP2, complement activation is involved in the lectinpathway.(4) Several lines of evidence demonstrated that aberrant activation of complementcontributes to the pathogenesis of many inflammatory and immunological diseases. Toinvestigate whether complement activation is involved in mediating lung inflammation and injury after H5N1virus infection, we depleted complement from mice with CVFbefore H5N1virus challenge. At Day1after H5N1virus challenge, lung damage andinflammatory cell infiltration were attenuated in the H5N1+CVF group compared withthe H5N1+PBS group. Attenuation of injury and inflammatory cell infiltration was alsoseen on Day5after infection in the CVF-treated group, and the expression level ofIFN-gmRNA was significantly lower in the CVF-treated group compared with thePBS-treated group on Days1and5after infection. However, the dynamic changes ofIFN-a/bR mRNA were similar between the two groups. A lower virus titer on Day5and a higher survival rate were also observed in CVF-treated mice compared withPBS-treated mice. Similar to CVF treatment, C3aR antagonist and anti-C5a Abtreatment also reduced lung injury and neutrophil infiltration especially on Day5afterH5N1virus infection, while treatment of C3aRa and anti-C5a Ab increased survivalrate simultaneously. Also, inhibition of complement activation with C3aR antagonistdecreases virus load in lung tissues and serum TNF-α level after H5N1virus infection.These data further indicate an important role for complement in lung inflammation andinjury after H5N1infection, suggesting that complement inhibition may be atherapeutic option.Conclusion: In summary, our study has established acute lung injury mice modelinduced with H5N1virus infection which had similar clinical presentations in patientsafter H5N1virus infection. The different pathological and physiological characteristicswere evaluated among mice infected with H5N1, H7N9or H1N1virus. H5N1virusinfection results in complement activation via the activation pathways, including theMBL pathway, which mediated the inflammatory response and ALI. Furthermore,inhibition of complement activation could significantly decrease inflammatoryresponses and attenuate ALI. These data not only provide useful information for thefurther study on the virulence of the avian influenza and the exploration of pathogenic mechanism after influenza virus infection, but also indicate that inhibition ofcomplement or complement activation products represents an alternative andadjunctive therapeutic option for treating ALI induced by H5N1virus infection andthat the combined treatment of antivirus and anti inflammation would berecommended.