All experiments involving live viruses were performed in negative-pressure isolators with HEPA filters in a biosafety level 3 (BSL3) facilities in accordance with the institutional biosafety manual.

Highly pathogenic H5N1 strain A/Chicken/Jiangsu/k0402/2010(CK10) that shows high virulence in mouse was isolated from a dead chicken (Hu et al., 2013b). A549 (human type II alveolar epithelial), 293T (human embryonic kidney) and MDCK (Madin-Darby canine kidney) cells were maintained in Dulbecco's modified Eagle's medium (DMEM; Life Technologies) supplemented with 10% fetal calf serum (FCS; Life Technologies) and 50 KU/L antibiotics and were cultured at 37°C under 5% CO₂.

The PA gene was amplified using high fidelity DNA polymerase (Invitrogen) based on CK10 cDNA and cloned into pCDNA3.1 vector to generate pCDNA-PA. The NCL and eEF1A1 genes were amplified from A549 cells. The sequences of NCL and eEF1A1 have aligned with sequences from GenBank (accession No: NM005381.2, KJ891086). The NCL and eEF1A1genes were cloned into the pires-hrGFP-1a vector (with a 3 × Flag tag in the C-terminus) to generate pires-hrGFP-1a-NCL and pires-hrGFP-1a-eEF1A1, respectively. Meanwhile, these two genes were also cloned into the pCDNA3.1-Myc-C vector (with a 3 × Myc tag in the C-terminus) to generate pCDNA-Myc-NCL and pCDNA-Myc-eEF1A1, respectively. All the positive clones were further validated by sequencing.

Protein samples were separated by electrophoresis on 10% (w/v) SDS-PAGE and transferred to polyvinylidene fluoride (PVDF) membranes (Bio-Rad). The membranes were then blotted with corresponding antibodies. Subsequently, the membranes were washed four times with TBST (0.05% tween-20 in Tris-buffered Saline) and incubated with horseradish peroxidase (HRP) conjugated goat-anti-mouse (Sigma-Aldrich) or goat-anti-rabbit IgG (Sigma-Aldrich). The enhanced chemiluminescence (ECL) system (Thermo) was utilized to detect the blotted proteins.

A549 cells were seeded in 6-well plates and cultured overnight. Then the cells were infected with CK10 virus at a multiplicity of infection (MOI) of 1. CK10-infected A549 cells were harvested at 12, 24, 36, 48, and 60 h post-infection (p.i.). The protein samples were subjected to western blot assay using anti-PA antibody and anti-β-actin antibody subsequently. The protein bands were analyzed by Image J.

Statistical analysis was performed using the SPSS statistics software. The independent samples T-test was used for data analysis. p < 0.05 was considered as significant.

For IP, CK10-infected A549 cells were lysed in IP buffer and incubated at 4°C on a shaker for 30 min, followed by centrifugation at 12,000 rpm for 20 min. A total of 600 μl of the supernatant was incubated with anti-PA ployclonal antibody (GeneTex) or the irrespective IgG (Beyotime) at 4°C overnight. Protein A+G Agarose beads (Beyotime) were then added, and the mixture was incubated with gentle rocking at 4°C for 5 h. The beads were washed five times with cold IP buffer and eluted with Glycline-HCL (PH = 3.0) for further LC-MS/MS analysis.

For co-IP, the 293T cells were co-transfected with pCDNA-PA or pCDNA3.1 alone and pires-hrGFP-1a-NCL, pires-hrGFP-1a-eEF1A1, or pires-hrGFP-1a, respectively. The transfected cells were then lysed in IP buffer at 48 h post-transfection. The cells lysates were precipitated with appropriate antibodies in conjunction with beads as described above. The beads were then washed five times with cold IP buffer and boiled with 5 × SDS loading buffer (Cwbio) for 5 min. The immunoprecipitated proteins were then detected by western blot.

The LC-MS/MS was performed on a mass spectrometer coupled to Easy nLC (Thermo Fisher Scientific). A volume of 6 μl of each fraction was injected for nanoLC-MS/MS analysis. The peptide mixture was loaded onto a C18-reversed phase column (Thermo Scientific Easy Column, 2 cm long, 100 μm inner diameter,5 μm resin) in buffer A (0.1% Formic acid) and separated with a linear gradient of buffer B (80% acetonitrile and 0.1% Formic acid) at a flow rate of 300 nL/min controlled by IntelliFlow technology over 140 min. MS data was acquired using a data-dependent on Top10 method dynamically choosing the most abundant precursor ions from the survey scan (300–1800 m/z) for HCD fragmentation. Determination of the target value is based on predictive Automatic Gain Control (pAGC). Dynamic exclusion duration was 60 s. Survey scans were acquired at a resolution of 70,000 at m/z 200 and resolution for HCD spectra was set to 17,500 at m/z 200. Normalized collision energy was 30 eV and the underfill ratio, which specifies the minimum percentage of the target value likely to be reached at maximum fill time, was defined as 0.1%. The instrument was run with peptide recognition mode enabled. MS/MS spectra were searched using MASCOT engine (Matrix Science, version 2.2) against Universal Protein (UniProt) database.

The functional annotation and classification of all the identified proteins were performed using Gene Ontology (GO) analysis tool in the Database for Annotation Visualization and Integrated Discovery (DAVID) (version 6.7). KEGG pathway database was used for pathway analysis. The Protein-Protein interaction network was constructed using the Cytoscape software. The Diseases and Functions analysis was conducted using Ingenuity Pathway Analysis (IPA) software.

293T cells were seeded on coverslips in 24-well plates and cultured overnight, then co-transfected with pCDNA-PA or the empty vector (pCDNA3.1) alone and pCDNA-Myc-NCL, pCDNA-Myc-eEF1A1, or the empty vector (pCDNA-Myc-C), respectively. At 24 h post-transfection, the cells were fixed with 4% paraformaldehyde for 30 min at room temperature, and permeabilized with 0.1% Triton-X-100 for 15 min. After blocking in 5% bovine serum albumin (BSA) for 30 min, the cells were incubated for 1 h with the primary antibodies at 37°C. After washing in PBS for three times, the cells were stained with Alexa Fluor 555-conjugated goat anti-rabbit (Beyotime) and FITC-conjugated goat anti-mouse secondary antibodies (Beyotime) for 1 h at 37°C. After washing in PBS for three times, cell nuclei were stained with DAPI (Beyotime). The cells were then observed by using the Leica TCS SP8 STED 3X confocal microscope.

To determine the peak of PA expression, A549 cells that infected with CK10 virus at a MOI of 1 were harvested at 12, 24, 36, 48, and 60 h p.i. and the samples were then subjected to western blot assay by using anti-PA antibody as first antibody. At the same time, the expression patttern of β-actin protein was served as control. Three independent experiments were performed. The western blot protein bands were analyzed by Image J subsequently. And the results showed that the expression level of PA increased gradually in the process of CK10 infection, and reached a peak at 36 h p.i. The expression level of PA at 36 p.i. was significantly higher than those of other time points (Figures 1A,B). Therefore, we then collected the samples at 36 h p.i. for subsequent interactome analysis.

To efficiently precipitate the PA protein from the virus-infected cells and identify the host proteins that interact with PA, A549 cells were infected with CK10 at a MOI of 1. The cells were then harvested at 36 h p.i. and were immunopricicipitated with ployclonal antibody against PA or the irrespective antibody as control. Three independent experiments were carried out for both experimental group and control group. Subsequently, all samples were subjected to LC-MS/MS analysis. In addition, all proteins present in the negative controls were excluded and only proteins appeared at least twice in the triplicate analyses were reserved to generate interaction data. As a result, 278 proteins specifically precipitated with PA antibody were detected by LC-MS/MS analysis when compared to those precipitated with the irrespective IgG. A detailed summary of these proteins were given in Table 1 and Table S1.

To obtain an overall functional profile of the interactome associated with the PA protein, the 278 identified proteins were assigned for further bioinformatics analysis. As shown in Figure 2, three major types of annotations, biological process, cellular components, and molecular functions, were generated in the GO consortium website. In addition, we found that subclasses associated with gene expression, viral infectious cycle and viral transcription were highly enriched in biological process category (Figure 2A). As for the most enriched subclasses in cellular component, the major types, such as cytosol, integral to membrane, and ribonucleoprotein complex, were enriched (Figure 2B). Moreover, molecular function analysis showed that protein binding, ATP binding, structural constituent of ribosome and RNA binding were highly enriched (Figure 2C). A more detailed summary of GO annotation of the identified proteins was provided in Table S2.

In addition, KEGG pathway analysis revealed an enrichment of 165 pathways related with 130 proteins of the infection network (Table S3). Six types of pathways were covered, including Human Diseases, Genetic Information Processing, Organismal Systems, Metabolism, Cellular Processes, and Environmental Information Processing (Figure 3A). To be noted, pathways involved in translation, infectious disease (contains infecious disease: Viral, infecious disease: Bacterial and infecious disease: Parasitic), cancers: overview and signal transduction were highly enriched in the PA-host interactome (Figure 3B, Table S3).

After detatied functional analysis, we then focused on two targeted novel cellular partners (NCL and eEF1A1) that might interact with PA, cellular proteins. As shown in Figure 4, the interactome profile indicated the important protein-protein interactions of the identified proteins associated with NCL and eEF1A1. The analysis revealed that NCL and eEF1A1 are cross-linked with several proteins with interesting interactions, including HNRNPA2B1-PTBP1-NCL-RPL5 and AHR-HSP90AA1-EEF1A1-RPL5. By GO annotation, we identified that these proteins have important functions in RNA transport, RNA splicing, apoptotic process, and translation which are closely asociated with IAV infection.

Meanwhile, the top Diseases and Functions analysis of the identified 278 cellular proteins was also performed using IPA software (Figures 5A,B, Figure S1). Surprisingly, we found that both NCL and eEF1A1 were related with Organismal Injury and Abnormalities. In addition, NCL was also involved in Cell Death and Survival. It has been demonstrated that immune injury and cell death contribute to the pathogenicity of IAV (Iwai et al., 2013; Duan and Thomas, 2016; Sridhar, 2016). Therefore, these resluts indicated that NCL and eEF1A1 might contribute to the virulence of IAV.

In order to systematically analyze the screened PA-interacting host proteins, we then made a detailed comparison of the interactome data associated with PA from different studies. Using affinity purification-mass spectrometry (AP-MS), Wang et al. identified 134 H5N1 IAV PA interacting proteins in chicken cells (Wang et al., 2016). Watanabe et al. also characterized 304 proteins that interact with the PA protein of the H1N1 influenza virus in human cells using LC/MS (Watanabe et al., 2014). In the present study, 278 human proteins were found to interact with H5N1 PA protein using LC-MS/MS in A549 cells. As a result, our dataset shared 50 and 10 host proteins with data from H1N1-Human study and H5N1-Chicken study respectively (Figure 6A; Table S4). It is worth noting that NCL and eEF1A1 are identifed as PA-interacting factors both in our study and H1N1 study (Figure 6B). These resluts further indicated that these two proteins may play an important role in influenza virus infection.

Considering the NCL and eEF1A1 were shared in this study and the H1N1 study, we then verifed their interactions with the PA protein. 293T cells were co-transfected with pCDNA-PA or the empty vector (pCDNA3.1) alone with the Flag-tagged NCL expression plasmid (pires-hrGFP-1a-NCL), Flag-tagged eEF1A1 expression plasmid (pires-hrGFP-1a-eEF1A1) or the empty vector (pires-hrGFP-1a) respectively. After 48 h post-transfection, the co-IP assay was performed using anti-PA or anti-FLAG antibody. The immune-complexes were detected in western blot assay using anti-FLAG or anti-PA antibody subsequently. The results showed that both NCL and eEF1A1 were detected only in the presence of PA, but not in the presence of empty vector (Figures 7A,B). In addition, the PA protein was detected only in the presence of NCL and eEF1A1, but not in the presence of empty vector (Figures 7C,D).

To demonstrate the co-localization of NCL and PA or eEF1A1 and PA, three plasmid-pCDNA-Myc-NCL, pCDNA-Myc-eEF1A1, and pCDNA-PA were constructed. Cells grown on coverslips were co-transfected with pCDNA-PA or the empty vector (pCDNA3.1) alone with pCDNA-Myc-NCL, pCDNA-Myc-eEF1A1, or the empty vector (pCDNA-Myc-C), respectively. The cells were fixed at 24 h post-transfection. Then the cells were subjected to confocal microscopy analysis by using primary antibodies of Myc-tag, PA and secondary antibodies with two different fluorophores. Co-localization of two proteins in cells was visualized by confocal microscope. As shown in Figure 7E, NCL and PA were co-localized both in the nucleus and cytoplasm, while eEF1A1 and PA were only detected in the cytoplasm of the cell. Thus, these results clearly confirmed that the PA protein interacts well with the cellular proteins NCL or eEF1A1.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.