The Identification Distinct Antiviral Factors Regulated Influenza Pandemic H1N1 Infection

Influenza pandemic with H1N1 (H1N1pdms) causes severe lung damage and “cytokine storm,” leading to higher mortality and global health emergencies in humans and animals. Explaining host antiviral molecular mechanisms in response to H1N1pdms is important for the development of novel therapies. In this study, we organised and analysed multimicroarray data for mouse lungs infected with different H1N1pdm and nonpandemic H1N1 strains. We found that H1N1pdms infection resulted in a large proportion of differentially expressed genes (DEGs) in the infected lungs compared with normal lungs, and the number of DEGs increased markedly with the time of infection. In addition, we found that different H1N1pdm strains induced similarly innate immune responses and the identified DEGs during H1N1pdms infection were functionally concentrated in defence response to virus, cytokine-mediated signalling pathway, regulation of innate immune response, and response to interferon. Moreover, comparing with nonpandemic H1N1, we identified ten distinct DEGs (AREG, CXCL13, GATM, GPR171, IFI35, IFI47, IFIT3, ORM1, RETNLA, and UBD), which were enriched in immune response and cell surface receptor signalling pathway as well as interacted with immune response-related dysregulated genes during H1N1pdms. Our discoveries will provide comprehensive insights into host responding to pandemic with influenza H1N1 and find broad-spectrum effective treatment.


Introduction
With the evolution, reassortment, and transmission of infuenza virus, infuenza pandemics caused severe pneumonia and higher mortality, leading to public health emergence and economic losses [1].Infuenza H1N1 strains have resulted in two pandemics in history spreading worldwide and killing many individuals [2,3].H1N1pdm1918 and H1N1pdm2009 originated from a series of reassortments among avian, swine, and human infuenza viruses and then transmitted to humans, leading to acute lung injury [4,5].Infuenza virus reassortments usually afect the efcacy of vaccination, which is the most efcient approach to prevent and control infuenza circulation [6], and only four anti-infuenza drugs (oseltamivir, zanamivir, peramivir, and baloxavir) were used, although the usage of these drugs may lead to the emergence of resistant infuenza strains [7][8][9][10][11].Terefore, novel broad-spectrum treatments are needed to be explored and developed.
A large number of host factors and cellular processes, for example, host dependency factors, host restriction factors, apoptosis, and autophagy are involved in the replication cycle of infuenza virus [12][13][14][15][16].It was all known that faced with infection, host will quickly respond to virus clearance and tissue function maintenance for the host survives by releasing antiviral signalling [17] .Host antiviral responses activated by infuenza virus infection, in turn, can prevent viral infection by inhibiting the fusion of viral and host membranes, inducing viral protein degradation and strengthening the innate immune response and antiviral signalling of MAVS [18][19][20].Moreover, host apoptosis directly targets infuenza virus-infected cells [21] and autophagy induces damaged oranges containing viral particles into the lysosome for viral elimination [22].But the virus can employ host compounds and processes to promote its replication and induce lung injury.Sialic acid receptor on the host cell surface is a key for the initiation of infuenza virus infection that depends on cellular endocytosis [23,24].Host factors CMAS and ST3GAL4 knockout inhibited the synthesis of sialic acid receptors [25].Glucosylceramidase (GBA) regulates infuenza virus entry and cellular endocytosis [26].Additionally, infuenza virus infection induces the release of proinfammatory cytokines that promote infuenza virus-related lung injury [27].Tese data indicate that the interaction between the host and virus is complex.Tus, the releasing mechanisms of virus-host interaction are particularly important, and analyses of the lung transcriptomic pattern in response to viral infections are a useful paradigm.
In this study, we performed an integrative analysis of transcriptomic expression profle data of mouse lung to assess host response patterns to diferent H1N1pdm strains and identify diferentially expressed genes (DEGs).Ten, gene ontology and pathway enrichment analyses were performed to clarify the function of these assessed DEGs, and protein-protein interaction (PPI) network analysis was also conducted and revealed key genes.We identifed distinct dysregulated genes during H1N1pdms infection.Tese results contribute to understanding the host response mechanisms to H1N1pdm virus.

Analysis of Gene Ontology and Pathway Enrichment.
To explain the role of DEGs in response to H1N1 virus infection, gene ontology (GO) enrichment analysis for the biological process was performed using the cluster profler package [41] in R software, and a p value <0.05 was considered statistically signifcant.Reactome pathway and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were combined to evaluate the functional pathways of the H1N1-associated genes by KEGG and Reactome databases with p value <0.05 [42,43].

Protein-Protein Interaction (PPI) Analysis and Network
Construction.Te common DEGs were selected for proteinprotein interaction analysis by using the STRING database (https://string-db.org/).Ten, the cluster analysis of the PPI network was performed with the MCL infation parameter in STRING.Te signifcant cluster networks were exported to Cytoscape for further visualization processing [44].

Transcriptomic Profles of the H1N1pdm2009-Infected
Mouse Lung.Te mRNA expression profles of the mouse lung infected with a pandemic infuenza virus (H1N1pdm09) were performed in here.Tree independent datasets (GSE43764, GSE40091, and GSE63786) with A/California/04/ 2009 H1N1 infection (a subtype of H1N1pdm09) were integrated to identify diferentially expressed genes (DEGs) using the limma package.Te result showed that a subtype of H1N1pdm09, A/California/04/2009 H1N1, produced numerous DEGs at three-day postinfection (3 dpi) and a timedependent increase of DEGs was detected after 5/6 dpi (Figures 1(a) and 1(d), Supplementary data1).Moreover, to look for the functional characteristics of these DEGs, gene ontology (GO) classifcation and pathway enrichment analysis were performed.Consistently, we found that H1N1pdm09-induced DEGs in both 3 dpi and 5 dpi were mainly associated with the response to virus and innate immune process, where a part of DEGs showed a role in organelle fssion and nuclear division process in 5/6 dpi (Figures 1(b) and 1(e)).KEGG pathway analysis indicated that these DEGs were overrepresented in pathways associated with viral protein interaction with cytokine and cytokine receptors, NOD-like receptor signalling, and infuenza A (Figures 1(c) and 1(f)).Furthermore, collecting all DEGs in 3 dpi and 5/6 dpi, we identifed 231 common DEGs, including 19 downregulated and 209 upregulated genes in response to H1N1pdm09 infection (Supplementary data1).Te downregulation was observed for genes (1500035N22Rik, 1700012B09Rik, 2410066E13Rik, Aass, Abcg5, Asgr1, Cd207, Ces1f, Cyp2a4, Cyp4f15, F2, Fabp1, Fmo3, Hepacam2, Hmgcs2, PiPOx, Pon1, Scgb1c1, and Uox) involved in the Llysine catabolic process to acetyl-CoA, acetyl-CoA metabolic, carboxylic acid catabolic, and small molecule catabolic process, while the biological process terms of the upregulated genes were enriched in response to virus, negative regulation of the viral process, and regulation of innate immune response.With MCL infation parameters based on the STRING database, protein-protein interaction (PPI) networks were generated (p value <1.0e − 16).173 of the upregulated DEGs were divided into fve signifcant clusters (Figure 2).STAT1, Cxcl10, and IRF7 nodes have the highest degree of connectivity (degree ≥ 90) in the PPI networks.

2
International Journal of Microbiology In addition, to identify whether other subtypes of H1N1pdm09 strain show a similar transcriptomic profle, three independent microarray datasets of the mouse lung responding to two diferent strains of H1N1pdm09 (A/ Jena/5258/09 and A/California/7/2009) infection were jointly analysed (Supplementary data2).A similar conclusion was reached that the number of overlapped DEGs (42 in 3 dpi and 109 in 5 dpi) in all two datasets was correlated with time points after infection (Figures 3(a) and 3(b), Supplementary data2).We found that all overlapped DEGs were upregulated and displayed in a volcano plot (Figure 3(c)).Furthermore, functional enrichment and PPI network analysis of these DEGs indicated that genes in Cluster1 were involved in the cytokine-mediated signalling pathway, neutrophil chemotaxis, and infammatory response, when genes in Cluster2 mainly played a role in the type I interferon signalling pathway, defence response, and innate immune response in both 3 dpi and 5 dpi (Figures 4(a) and 4(b)).Integrated with gene expression profles in diferent strains of H1N1pdm2009, we fnally identifed 38 and 87 overlapped DEGs in all analysed datasets infected with H1N1pdm09 strains at 3 dpi and 5 dpi, respectively (Supplementary data3).International Journal of Microbiology 5

Diferential Gene Expression Profle between the H1N1pdm1918-Infected Lung and Healthy Controls.
Another pandemic infuenza virus 1918 H1N1 strain (H1N1pdm1918) caused the deadly infuenza pandemic and severe lung injury.To explore the regulatory mechanism of lung in-host defence against H1N1pdm1918 virus infection, we performed a comparative gene expression profling by using the publicly available array expression profling datasets (GSE38112 and GSE70445).A direct comparison analysis of up or down trends in expression showed that there were 263 and 650 overlapped DEGs in 3 dpi and 5 dpi, respectively (Figures 5(a     International Journal of Microbiology process, those DEGs in 5 dpi were enriched in adaptive immunity-related antigen processing and presentation and phagosome process (Figures 5(c) and 5(f )).To reveal whether there are common dysregulated genes during H1N1pdm1918 virus infection, we continuously analysed time-course gene expression profling in 3 dpi and 5 dpi.192 overlapped DEGs were identifed with Venn diagrams (Figure 6(a) and Supplementary data4) and signifcantly enriched in host defence, immune system, and interferon signalling.Additionally, the diferential protein-protein interaction network was constructed and showed two signifcant clusters: one cluster (80 genes) enriched interferon signalling and antiviral mechanism by IFN-stimulated genes, another cluster (24 genes) enriched chemokine receptor binding chemokines and regulating of IFNG signalling (Figures 6(b) and 6(c)).By integrating all DEGs in H1N1pdms, we acquired 32 and 74 dysregulated genes at 3 dpi and 5 dpi, respectively (Supplementary data4).

Identifcation of Distinct DEGs in H1N1pdm Strains
Compared with Nonpandemic H1N1.To make clear H1N1pdms-induced distinct host responses, gene expression profles of lungs infected with nonpandemic H1N1 (nH1N1pdm) were analysed and then compared with the DEGs induced by H1N1pdms.Similarly, a direct comparison of up or down trends showed that infection of nH1N1pdm triggered a strong and persistent innate immune response due to the production of many innate immune-related upregulated genes (Figures 7(a

Discussion
H1N1 infuenza pandemic (H1N1pdms) causes severe public health emergency, resulting in severe pneumonia and high mortality rates.Te diferent strains of H1N1pdms can show distinct infection patterns and interaction with the host, suggesting that the study of H1N1pdms-host interaction is essential.Due to higher genetic mutations and reassortment, the diferent strains of H1N1pdms can utilize diferent mechanisms to induce host injury.Tus, explaining the host response to H1N1pdms infection and identifying critical genes and signalling pathways will provide novel treatment strategies in infuenza pandemic.In this study, we performed multiple gene expression profles and used bioinformatical approaches to investigate host responses to diferent H1N1pdms and identify distinct diferentially expressed genes in H1N1pdms infection compared with nonpandemic H1N1.H1N1pdms elicit acute hyperinfammatory response, causing lung damage and respiratory failure as well as death [45], and host resistance and tolerance to H1N1pdms-induced lung injury refer to host genes' expression level [46].However, the mechanisms of hyperinfammatory activation during H1N1pdms infection and the interaction of host-H1N1pdms are unclear.Our transcriptomic profling and biological processes analysis explored that H1N1pdms-induced dysregulated genes were mainly involved in defence to infection, chemokine receptors binding chemokines, and regulating of IFNG signalling.Importantly, most of ten identifed distinct DEGs (AREG, CXCL13, GATM, GPR171, IFI35, IFI47, IFIT3, ORM1, RETNLA, and UBD) in H1N1pdms infection in our study are involved in host response to viruses, and the discovery of these molecular biomarkers may provide new insights into diagnosis and treatment against H1N1pdms infection.IFI35, IFI47, and IFIT3 are associated with the immune and defence process.IFI35 can increase H5N1 infuenza disease and has been identifed as a promising biomarker and therapeutic target for syndromes induced by SARS-CoV-2 or infuenza virus [47,48].Amphiregulin (AREG) is an epidermal growth factor that plays an important role in regulating virus-infected lung repair [49].AREG expression has been reported in epithelial cell layers and various immune cells, including dendritic cells, neutrophiles, and CD4 + T cells [50,51], and is constitutively upregulated in response to infammation or infection [52].AREG can promote alveolar remodelling and integrity during infuenza virus infection.Innate lymphoid cells (ILCs) that are critical in immune response and tissue homeostasis can produce AREG, which in turn restores lung function and airway remodel [53].Previous studies have shown that infuenza viruses bind to sialic acid receptors and then lead to the activation of EGFR, promoting virus entry [54], suggesting that AREG-EGFR signalling could function in host immune response to infuenza virus and tissue tolerance.In addition, C-X-C motif chemokine ligand 13 (CXCL13) is also involved in receptor-mediated signalling pathways, except for its proinfammatory function [55].International Journal of Microbiology HIV-1-infected and COVID-19 patients have higher levels of plasma and serum CXCL13 concentration, and CXCL13 has been identifed as a biological signature of COVID-19 patients and HIV-1 patients [56][57][58].Moreover, high levels of CXCL13 expression have been proved to be associated with pulmonary fbrosis that is the prominent feature of infection with 2009 pandemic infuenza A (H1N1) virus [59,60], suggesting that CXCL13 may play an important role in pulmonary diseases caused by infuenza virus infection and still need to be further investigated.Resistin-like alpha (RETNLA), a cysteine-rich secreted family of Fizz/Resistin-like molecules and a M2 macrophage marker that modulates lung fbrosis and infammation, has been revealed to act as a marker of activated macrophages and involved in the immune response-induced pulmonary vascular remodelling [61][62][63].It is all known that the mRNA levels of RETNLA can refect M2 macrophage polarization and infuenza virus infection-induced cell apoptosis [64].Our result shows the upregulated RETNLA expression in H1N1pdms, indicating that H1N1pdms infection may increase M2 macrophage apoptosis.In addition, overexpression of RETNLA can decrease allergic lung infammation by reducing infltration of immune cells and T2 cytokine production, suggesting that the host may increase RETNLA expression to trigger M2 macrophage polarization and promote lung repair during H1N1pdms infection [65].A previous study has shown that glycine amidinotransferase (GATM) was upregulated in M2polarized macrophages.GATM deletion inhibited the expression of RETNAL and blocked M2 polarization [66].Based on these, we speculate that GATM may regulate RETNLA to afect M2 macrophage polarization during H1N1pdms infection.
Te interaction between DEGs and transcription factors (TFs) was explored to know about how the DEGs regulate infuenza virus at the transcriptional level.Our analysis of the TFs-DEGs network found that BATF2 was the most signifcant TF as the regulator of DEGs.We found that BATF2 was upregulated during H1N1pdms infection.In previous analysis, BATF2 is an important regulator of the innate immune system and has high expression in human lung structural cells infected with infuenza [67], indicating that BATF2 could play a critical role in host antiviral immune, but further studies are needed.

Conclusion
In our study, based on integration microarray datasets of the mouse lung infected with diferent H1N1pdms, host cells perform the similar immune response to diferent H1N1pdms.We further identifed ten distinct DEGs (AREG, CXCL13, GATM, GPR171, IFI35, IFI47, IFIT3, ORM1, RETNLA, and UBD) diferentially expressed genes during H1N1pdms infection compared with nonpandemic H1N1.Tese distinct dysregulated genes may have important regulation efects, and our future work will focus on revealing the function of these distinct dysregulated genes during infuenza virus infection for the development of novel treatment strategies.

Figure 1 :Figure 2 :
Figure 1: Identifcation and function enrichment analysis of diferentially expressed genes (DEGs) of the lung during H1N1pdm2009 infection.(a, d) Gene expression profle analyses of the infected lung with A/California/04/2009 strains indicating the common and distinct gene sets in 3 day-postinfection and 5/6 day-postinfection via Venn diagram; (b, e) Gene ontology analysis showing the biological process of DEGs; (c, f ) pathway enrichment analysis of DEGs.

Figure 3 :Figure 4 :
Figure 3: Gene expression profle analyses of the infected lung with other H1N1pdm2009 strains.(a, b) Gene expression profle analyses of the infected lung with three other H1N1pdm2009 strains indicating the common and distinct gene sets in 3 dpi and 5 dpi via Venn diagram; (c) volcano plot representation of DEGs in three diference microarray datasets (GSE63786, GSE67241, and GSE70882).Red and blue colours indicate the genes increased or decreased expression, respectively.Te overlapped DEGs are separately displayed.
) and 5(d), Supplementary data4).Importantly, H1N1pdm1918-induced DEGs represented the similar function characteristic with those DEGs in H1N1pdm09 (Figures5(b) and 5(e)).Compared with DEGs in 3 dpi that were mostly involved in the innate immune

Figure 5 :
Figure 5: Global transcriptomic profles change of the lung during H1N1pdm1918 infection.(a, d) Heatmap of diferentially expressed patterns of genes in the H1N1pdm1918-infected lung from two microarray datasets (GSE38112 and GSE70445); (b, e) gene ontology analysis showing the biological process of DEGs; (c, f ) KEGG pathways of the DEGs.Signifcant top 20 enriched by diferentially expressed genes were shown.

Figure 6 :
Figure 6: Te overlapped DEGs in H1N1pdm1918.(a) Venn diagram of the overlapped genes between diferent data.A total of 192 genes were common to the H1N1pdm1918-infected lung at 3 dpi and 5 dpi; (b, c) the common DEGs were divided into two clusters and visualized using Cytoscape.
) and 7(d), Supplementary data5).Te biological process enrichment analysis of 296 DGEs in 3 dpi and 342 DEGs in 5 dpi showed signifcant enrichment of response to virus, defence response to virus, and response to interferon-gamma (Figures7(b) and 7(e)).KEGG analysis showed a signifcant enrichment of upregulated genes involved in the NODlike receptor signalling pathway, Toll-like receptor signalling pathway, TNF signalling pathway, and cytokine-cytokine receptor interaction (Figures7(c) and 7(f )).Besides, we found that DEGs in H1N1pdms infection were completely present in nonpandemic H1N1 infection at 3 dpi, while only ten distinct DEGs (AREG, CXCL13, GATM, GPR171, IFI35, IFI47, IFIT3, ORM1, RETNLA, and UBD) in H1N1pdm strains infection were identifed in comparison with nonpandemic H1N1 at 5 dpi (Figures8(a) and 8(b)), suggesting that H1N1pdm induces host distinct response in the later stage of infection.Moreover, we found that these distinct DEGs were continuously upregulated in H1N1pdms infection, while there was no change or rapid up-and -downregulation in nH1N1pdms infection.Furthermore, the biological process of these distinct DEGs was involved in immune response and cell surface receptor signalling pathway, and these distinct DEGs can interact with immune response-related dysregulated genes (Figure8(c)).

Figure 7 :
Figure 7: Identifcation and function enrichment analysis of diferentially expressed genes (DEGs) of the lung during A/PR/8/34 infection.(a, d) Gene expression profle analyses of the infected lung with A/PR/8/34 strains indicating the common and distinct gene sets in 3 dpi and 5 dpi via Venn diagram; (b, e) gene ontology analysis showed the biological process of DEGs; (c, f ) cell signalling pathway analyses showing pathway enrichment of the DEGs.Signifcant top 20 enriched by diferentially expressed genes were shown.

Figure 8 :
Figure 8: Identifcation of the distinct DEGs in H1N1pdms strains compared with nH1N1pdms.(a, b) Venn diagram showing the distinct DEGs in H1N1pdms at 3 dpi and 5 dpi, respectively.(c) Te distinct DEGs in H1N1pdms interacted with H1N1pdms-induced dysregulation genes related with immune response.Yellow circle represents the identifed distinct DEGs; red line represents the interaction with distinct DEGs.