Different Subtypes of Influenza Viruses Target Different Human Proteins and Pathways Leading to Different Pathogenic Phenotypes

Different subtypes of influenza A viruses (IAVs) cause different pathogenic phenotypes after infecting human bodies. Analysis of the interactions between viral proteins and the host proteins may provide insights into the pathogenic mechanisms of the virus. In this paper, we found that the same proteins (nucleoprotein and neuraminidase) of H1N1 and H5N1 have different impacts on the NF-κB activation. By further examining the virus–host protein–protein interactions, we found that both NP and NA proteins of the H1N1 and H5N1 viruses target different host proteins. These results indicate that different subtypes of influenza viruses target different human proteins and pathways leading to different pathogenic phenotypes.


Introduction
Influenza A virus (IAV) belongs to the Orthomyxoviridae family. Its genome consists of eight segmented, negative-strand RNAs [1][2][3]. IAVs are further typed into many different subtypes based on the antigenicity of hemagglutinin (HA) and the neuraminidase (NA) [4], the proteins on the surface of the virus. Currently, 18 HA(H1-H18) and 11 NA (N1-N11) subtypes have been identified in IAVs [5]. Different subtypes of IAVs may have evolved various mechanisms to co-opt host processes and suppress host defenses, inducing different infectious phenotypic outcomes. For example, the highly transmissible seasonal H1N1 virus usually causes mild illness, while the highly pathogenic avian H5N1 virus o en leads to severe cases when infecting humans. e pathogenic mechanism under this phenomenon needs more research efforts to reveal. e innate immune system of human plays an important role in the hosts' resistance to viruses. e NF-κB signaling pathway [6][7][8][9] is a key component of the innate immune system during its antiviral process. In mammals, there are two major NF-κB signaling pathways: the canonical NF-κB pathway and the noncanonical NF-κB pathway [10,11]. Under the stimulation of pathogenic microbial infections or their induced inflammatory factors, NF-κB is normally activated by classical signaling pathways [12]. Diverse bacterial and viral pathogens target NF-κB signaling pathway to evade host immune defenses. Some viruses suppress NF-κB activation to dampen the host immune responses to maintain latency [13]. For example, human bocavirus (HBoV) proteins NS1 and NS1-70 [14], the mumps virus (MuV) small hydrophobic protein (SH) [15], and molluscum contagiosum virus (MCV) protein MC005 [16] inhibit TNF-α-mediated activation of NF-κB by reducing the phosphorylation of IKKβ, IκBα, and p65 as well as the translocation of p65 into the nucleus. Some pathogenic microorganisms activate NF-κB for viral gene expression, replication and spread. For example, the K15 protein of KSHV and the early protein Nef of most primate lentiviruses enhances NF-κB activation to initiate proviral transcription [17,18]. IAVs can also interfere with antiviral responses by regulating the NF-κB signaling pathway. Studies have shown that influenza virus proteins interact with NF-κB to promote viral replication. For example, the IAV NS1 protein specifically inhibits IKK-mediated NF-κB activation and production of the NF-κB induced antiviral genes by physically interacting with IKK through the C-terminal effector domain [19].
In this paper, we used the luciferase assay and the co-immunoprecipitation experiment to study how the NP and NA viral proteins of both HIN1 and H5N1 interact with NF-κB signaling pathway. We found that the H1N1 NP has little impact on the TNF-α-induced NF-κB transcriptional activation, while the H5N1 NP inhibits the TNF-α-induced NF-κB transcriptional activation by interacting with IKKα. Furthermore, the H5N1 NP protein promotes the expression of IκBα proteins, meanwhile, suppresses the phosphorylation of IκBα, and restrains the translocation of p65 into the nucleus of infected 293 T cells. We also found that the H5N1 NA activates the IL-1β-induced NF-κB transcriptional activation through TAB2, while H1N1 NA inhibits the IL-1β-induced NF-κB transcriptional activation by coimmunoprecipitating with IKKβ. All these results imply divergent pathogenic mechanisms of different subtypes of IAVs.

Immunoprecipitation, and
Immunoblotting. Immunoprecipitation was performed using IP buffer (1% Nonidet P-40, 50 mM Tris-HCl [pH 7.5], 150 mM NaCl, and Complete TM protease inhibitor cocktail-EDTA (Roche)). Whole cell extracts were prepared a er transfection and incubated with indicated antibodies together with Protein A/G beads (Roche) overnight. Beads were then washed 4 times with IP buffer, and immunoprecipitates were eluted with SDS loading buffer (TransGen Biotech) and resolved in SDSPAGE gels. e proteins were transferred to PVDF membrane (Bio-Rad) and further incubated with the indicated antibodies. e antigen-antibody complexes were visualized by the Immobilon TM chemiluminescent detection kit (Millipore).

Immunofluorescence
Analysis. HeLa cells cultured on coverslips were fixed in 4% paraformaldehyde for 10 min and washed with PBS for three times. A er that, cells were fixed with 0.1% Triton X-100 on ice for 5 min, washed in PBS, and blocked in 1% BSA for 20 min. e coverslips were incubated with 1% BSA containing primary antibodies for 1 h, followed by PBS wash for three times. e cells were further stained with Alexa Fluor® conjugated secondary antibodies (Abcam). Images were acquired on an Olympus FV1000 fluorescence microscope.

Statistical Analysis.
Analyses were done with the statistical so ware SAS/STAT. Data analysis over time was undertaken by repeated-measures analysis with SAS/STAT. Differences were considered statistically significant if the value was <0.05.

H5N1 NP and H1N1 NP Have Different Impacts on the TNF-α-Induced NF-κB Activation.
e nucleoprotein of negative-strand RNA viruses forms a major component of the ribonucleoprotein complex that is responsible for viral transcription and replication, and NP proteins of IAV support the progress of the polymerase during the elongation phase [20]. e dual-luciferase activity assay was performed to determine the impacts of the NP proteins from different subtypes of IAVs on the TNF-α-induced NF-κB activation. e results showed that the H5N1 NP clone significantly inhibited TNF-α stimulated NF-κB promoter activity, while the H1N1 NP had little impact on it (Figures 1(a) and 1(b)).
It is known that activated NF-κB signaling pathway can phosphorylate IκBα and cause subsequent degradation [21]. To further verify the impact of the NP proteins on the NF-κB signaling pathway, we examined the phosphorylation and degradation of IκBα in the NF-κB signaling pathway by immunoblotting. When treated with TNF-α, the decreasing amount of IκBα in H5N1 NP-expressing cells was less than that in empty vector-transfected cells (Figure 1(c)). Furthermore, the level of phosphor-IκBα in the H5N1 NP-expressing cells was diminished ( Figure 1(d)). e decreasing amount of IκBα in the H1N1 NP-expressing cells was similar to that in the empty vector-transfected cells, which were both treated with TNF-α ( Figure 1(e)). e level of phosphor-IκBα in the H1N1 NP-expressing cells was similar to the empty vector-transfected cells (Figure 1(f)). ese results demonstrated that H5N1 NP protein suppressed TNF-α-mediated IκBα phosphorylation and degradation while H1N1 NP protein had little impact on them.  1: H5N1 NP and H1N1 NP have different impacts on TNF-α-induced NF-κB activation. 293 T cells in 24-well plates were cotransfected with 125 ng pNF-κB-luc, 25 ng pRL-TK and indicated amount of H5N1 NP (a) and H1N1 NP (b) expression plasmid, or empty vector for 24 h. Cells were then mock-treated or treated with TNF-α (10 ng/ml) for 6 h. Reporter activity was determined by dual-luciferase reporter assays. e resultant ratios were normalized to the fold-change value by that of TNF-α-untreated cells cotransfected with empty vector, pNF-κB-luc and pRL-TK. Data shown represent three independent experiments, with each determination performed in duplicate (mean ± SD of fold change). Asterisks indicate significant differences between groups ( * * 푝 < 0.05, Student's t-test). 293 T cells transfected with empty vector, H5N1 NP or H1N1 NP-expressing plasmid were stimulated with TNF-α (20 ng/ml) for indicated durations. Equal amounts of cell lysates were analyzed by immunoblotting with the anti-IκBα antibody or the anti-phospho-IκBα antibody (c-f). HeLa cells were transfected with H5N1 NP, H1N1 NP expression plasmid, or empty vector for 30 h. e cells were then mock-treated or treated with TNF-α (10 ng/ml) for 30 min. HeLa cells were subjected to immunofluorescence staining for detection of p65 subcellular localization by using rabbit anti-p65 and FITC-conjugated secondary Ab (green). H5N1 NP and H1N1 NP expression levels were detected using a mouse anti-Flag tag and Texas Red-conjugated secondary Ab (red). Nuclei were stained by Hoechst 33258 (blue) (g, h).
BioMed Research International 4 the cytoplasm a er treatment with TNF-α in the H5N1-NP-GFP-transfected cells (Figure 1(g)), while p65 transported from cytoplasm to nucleus in the H1N1-NP-GFP-transfected e nuclear translocation of NF-κB is a prerequisite for the promotion of downstream genes [22]. Immunofluorescence also showed that almost all the p65 proteins were retained in  each determination performed in duplicate (mean ± SD of fold-change). Asterisks indicate significant differences between groups ( * * 푝 < 0.05, Student's t-test). 293 T cells were transfected with Flag-H5N1 NP or either HA-TAK1 orSS HA-TAB1 or HA-TAB2 or HA-IKKα or HA-IKKβ HA-vector expression plasmids for 30 h. All cells were then treated with TNF-α (10 ng/ml) for 30 min. e cells were lysed and subjected to immunoprecipitation (IP) using the mouse anti-HA tag. IP products and 5% input samples were analyzed by immunoblotting (f), 293 T cells transfected with empty vector, H5N1 NP -expressing plasmid were stimulated with TNF-α (20 ng/ml) for indicated durations. Equal amounts of cell lysates were analyzed by immunoblotting with the anti-phospho-IKKα antibody (g).
e activation of IKKα by phosphorylation is required to the phosphorylation of IκB [23]. We therefore tested whether H5N1 NP played a role in inhibiting the phosphorylation of the IKKα. e results showed that H5N1 NP protein may have blocked the TNF-α-mediated IKKα phosphorylation ( Figure  2(g)).

H5N1 NA and H1N1 NA Have Different Impacts on the IL-1β-Induced NF-κB Activation. Neuraminidase (NA) is an
integral membrane glycoprotein and a second major surface antigen of the IAV. During the infection process of IAV, NA cleaves terminal sialic acid from glycoproteins or glycolipids to free virus particles from host cell receptors and facilitate virus spread [6,24].
To study the different impacts of the NA proteins of H5N1 and H1N1 on the NF-κB signaling pathways, reporter plasmid pNF-κB-luc and internal control plasmid pRL-TK, together with pH5N1 NA and pH1N1 NA or empty vector, were cotransfected into 293 T cells. At 24-h post-transfection, cells were mock-treated or treated with human IL-1β for 6 h. ese results showed that the H5N1 NA significantly promoted cells (Figure 1(h)). ese findings indicated that H5N1 NP and H1N1 NP had different impacts on the TNF-α-induced NF-κB activation.

H5N1 NP Inhibits the NF-κB Signaling Pathway by Targeting IKKα.
e results above indicated that H5N1 NP had an inhibitory effect on the NF-κB signaling pathway. Next, we intended to investigate which host protein the NP protein targeted on to get this effect. Firstly, we examined the effects of H5N1 NP protein on the luciferase activity mediated by overexpression of TNF-α signaling transducers along the NF-κB signaling pathway. H5N1 NP protein inhibited TRAF2, TAK1 + TAB1, TAB2, IKKα-induced NF-κB activation in a dose-dependent manner (Figures 2(a)-2(d)) and had little effect on the IKKβ-mediated luciferase activity (Figure 2(e)).
ese results showed that the potential targets of H5N1 NP on the NF-κB signaling pathway may include the IKKα and the complex of TAK. Results of co-immunoprecipitation showed that H5N1 NP binds to IKKα, indicating that there were protein-protein interactions between H5N1 NP and the TAK1 and IKKα in cells (Figure 2(f)).  Cells were then mock-treated or treated with IL-1β (10 ng/ml) for 6 h. Reporter activity was determined by dual-luciferase reporter assays. e resultant ratios were normalized to the fold-change value by that of IL-1β-untreated cells cotransfected with empty vector, pNF-κB-luc and pRL-TK. Data shown represent three independent experiments, with each determination performed in duplicate (mean ± SD of fold change).
Asterisks indicate significant differences between groups ( * * 푝 < 0.05, Student's t-test). 293 T cells were transfected with Flag-H5N1 NA (c) or Flag-H1N1 NA (d) or either HA-TAK1 or HA-TAB1 or HA-TAB2 or HA-IKKα or HA-KKβ or HA-vector expression plasmids for 30 h. All cells were then treated with IL-1β (10 ng/ml) for 30 min. e cells were lysed and subjected to immunoprecipitation (IP) using the mouse anti-HA tag. IP products and 5% input samples were analyzed by immunoblotting.
However, in our experiment with H1N1 NP as the experimental object, we found that H1N1 NP hardly affected the NF-κB signaling pathway (A/Beijing/501/2009 (H1N1), Figure  1(b)). We did not observe any significant change in the degradation and phosphorylation of IκBα protein. Nuclear translocation of p65 was also not affected by H1N1 NP. We speculate that the different effects between the two viral NP proteins and the NF-κB signaling pathway lead to differssent pathogenicity to some extent, but a more detailed mechanism needs further investigation.
NA is one of the major antigenic targets of the humoral immune response to IAV and the target of the antiviral drugs oseltamivir and zanamivir [30]. Our studies on the neuraminidase protein showed that the NA protein of H5N1 virus promotes the transcriptional activation of IL-1β-induced NF-κB, while the NA protein of H1N1 influenza virus inhibits the corresponding activation.
Considering the cells we used in the experiments, it is true that 293 T and HeLa cells are not the main target organ of influenza virus and do not support its productive infections; however, here in this study, we tried to evaluate the possible host target of the NP and NA proteins in the NF-κB signaling pathway, a process that happened a er virus infection. Nevertheless, A549 is a more suitable cell model for studying influenza virus, which we will use in our future study. e comparison of the above two groups of results indicates that the interaction between influenza virus proteins and host signaling pathway proteins has subtype-specific characteristics, which can respond to the specific pathogenic mechanism of influenza virus of different subtypes. e viral proteins employed to target at these conserved pathways are different for different subtypes of IAVs, suggesting more subtype-specific pathogenic mechanisms need to be revealed in the future research work.

Conclusion
Our study shows that different proteins of varied influenza viruses will produce distinct interaction effects with host NF-κB signaling pathway. e interaction between influenza viral proteins and host proteins is subtype-specific. is specificity may due to the specific pathogenic mechanisms of different subtypes of influenza viruses. Our discovery may suggest a new way of thinking about flu treatment and provide potential avenues to study pathogenesis of avian influenza. IL-1β-stimulated NF-κB promoter activity (Figure 3(a)), while the H1N1 NA had an inhibitory impact on it (Figure 3(b)).

Conflicts of Interest
We inferred that the potential targets of the H5N1 NA on the NF-κB signaling pathway was the complex of TAK ( Figure  S1), and the potential targets of the H1N1 NA were TAB2 and IKKβ ( Figure S2). Similarly, by implementing co-immunoprecipitation essays, we found that H5N1 NA interacted with TAB2 of the NF-κB signaling pathway (Figure 3(c)), while H1N1 NA interacted with IKKβ in the NF-κB signaling pathway (Figure 3(d)). Based on these results, we conclude that H5N1 NA and H1N1 NA had different impacts on the IL-1βinduced NF-κB activation.

Discussion
NF-κB signaling pathway is widely found in almost all animal cells and plays a key role in the innate immune defense of host against pathogenic microorganisms invaded [25]. Once being invaded by viruses, the host manipulates the NF-κB signaling pathway to clear them. To survive, microbial pathogens have evolved strategies to modulate the NF-κB signaling pathway, which is an important pathogenic mechanism to evade innate immunity and promote replication.
In the process of influenza virus infecting the host, viral proteins or other pathogenic factors will interact with host proteins and signaling pathways. Previous research found that the IAV NS1 protein specifically inhibits IKK-mediated NF-κB activation and production of the NF-κB induced antiviral genes [19]. One report showed that PB1-F2 binds to IKKβ and impairs DNA-binding of NF-κB [26]. In contrast, another study reports on an NF-κB intensifying activity of PB1-F2 [27]. In our study, we focused on the regulation of NF-κB signaling pathway by NP and NA proteins.
Firstly, we found the NP protein of H5N1 virus inhibited the transcriptional activation of the TNF-α-induced NF-κB signaling pathway (A/goose/Jilin/hb/2003 (H5N1), Figure  1(a)) [28]. In the absence of stimuli, NF-κB dimers are retained in the cytosol through association with an inhibitor of κB activity, termed IκB [29]. Stimulus-induced degradation of IκB molecules leads to nuclear accumulation of p65, which is the basis of NF-κB activation. In our study, H5N1 NP protein abolished IκkBα phosphorylation and degradation of IκBα, which makes NF-κB dimers immobilized in the cytoplasm. e results of immunofluorescence experiments in Figure 1 verified that nuclear factors were still in the cytoplasm. Our study reveals a potential mechanism by which H5N1 virus evades human innate immune responses.
Although the H5N1 NP inhibited TNF-mediated NF-κB activation and the H5N1 NA activated NF-κB downstream of IL-1β in this paper, it has been reported that the overall NF-κB activation status upon infection by the H5N1 subtype is activated. We speculated that the inhibitory effects of NP protein on TNF-α-mediated NF-κB activation can inhibit the antiviral immune response of the hosts in the early stage of invasion and promote the massive replication of the virus, while the NA protein promotes the IL-1β-mediated NF-κB activation, which may promote apoptosis of host cells. Both of the two facts of the two proteins may benefit the spread of the virus.