Abstract

   As a highly pathogenic influenza virus, H5N1 avian influenza virus
   (AIV) poses a great threat to poultry production and public health.
   H5N1 AIV has a small genome and, therefore, relies heavily on its host
   cellular machinery to replicate. To develop a comprehensive
   understanding of how H5N1 AIV rewires host cellular machinery during
   the course of infection, it is crucial to identify which host proteins
   and complexes come into physical contact with the viral proteins. Here,
   we utilized affinity purification mass spectrometry (AP-MS) to
   systematically determine the physical interactions of 11 H5N1 AIV
   proteins with host proteins in chicken DF1 cells. We identified with
   high confidence 1,043 H5N1 AIV–chicken interactions involving 621
   individual chicken proteins and uncovered a number of host proteins and
   complexes that were targeted by the viral proteins. Specifically, we
   revealed that chicken Staufen double-stranded RNA-binding protein 2
   interacts with AIV non-structural protein 1 (NS1) and promotes the
   replication of the virus by enhancing the nuclear export of NS1 mRNA.
   This dataset facilitates a more comprehensive and detailed
   understanding of how the host machinery is manipulated during the
   course of H5N1 AIV infection.

   Keywords: H5N1 AIV, AP-MS, Stau2, NS1, chicken

Introduction

   Influenza A virus (IAV) is a segmented, single-stranded, negative-sense
   RNA virus that has adapted to infect multiple species. This virus
   causes annual epidemics and recurring pandemics, which have huge
   impacts on public health. IAV particles have two viral surface
   glycoproteins (hemagglutinin, HA; neuraminidase, NA) and one matrix-2
   protein (M2). Inside the virion, all eight viral RNA (vRNA) segments
   bind three RNA polymerases (polymerase acid protein, PA; polymerase
   basic protein 1, PB1; and 2, PB2) and are encapsidated by the
   nucleoprotein (NP) to form the viral ribonucleoprotein (vRNP) complexes
   ([33]1–[34]5). The vRNPs are surrounded by a layer of the matrix
   protein, M1, which lines the envelope ([35]6). Upon infection by
   influenza, the host cells detect the viral RNA through pathogen
   sensors, and the major gene products of the influenza virus mediate the
   viral life cycle and modulate cellular processes ([36]7).

   Viruses rely on host cellular functions to replicate, and thus, they
   hijack the host cell machinery and rewire it for their own needs.
   Several proteomic studies have used affinity purification mass
   spectrometry approaches to identify a series of cellular factors that
   interact with IAV proteins ([37]8–[38]10). However, knowledge of common
   and strain-specific interactions remains incomplete, and how these
   interactions control host defense and viral infection remains to be
   fully elucidated. A comprehensive understanding of host–virus
   interactions would greatly improve our understanding of the viral life
   cycle and host resistance mechanisms. Here, we applied the AP-MS
   technology to uncover a wide array of host proteins, complexes, and
   pathways that are hijacked by H5N1 avian influenza virus (AIV) during
   the course of infection. We constructed an H5N1 AIV–chicken protein
   interaction map, and in addition to replicating previously identified
   host factors, we uncovered several novel interactions. Among the novel
   factors, chicken Staufen double-stranded RNA-binding protein 2 (STAU2)
   was found to be a crucial component when the viral mRNA is transported
   during the replication stage of the viral life cycle. Importantly,
   STAU2 interacts with the influenza NS1 protein and promotes the
   replication of H5N1 AIV by promoting the transport of NS1 mRNA from the
   nucleus to the cytoplasm.

Materials and Methods

Cells and Virus

   Chicken embryonic fibroblast (DF1) cells, Madin–Darby canine kidney
   (MDCK) cells, and human embryonic kidney cells (293T) were cultured in
   Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal
   bovine serum (Gibco), 100 μg/ml streptomycin, and 100 U/ml penicillin
   at 37°C under a humidified atmosphere of 5% CO[2]. The highly
   pathogenic H5N1 strain A/mallard/Huadong/S/2005 (SY) ([39]11) was
   propagated in 10-day-old specific pathogen-free embryonic chicken eggs.
   The H5N1 influenza virus strain 178 (GenBank Accession No.
   [40]AY737296-737300) was isolated from a chicken in Guangdong, China,
   by the MOA Key Laboratory for Animal Vaccine Development, P.R. China.
   The experiments that involved live viruses were performed in a
   biosafety cabinet with HEPA filters in a biosafety level 3 laboratory
   in Yangzhou University or South China Agricultural University.

Plasmid Construction

   The H5N1 AIV genes were amplified by high-fidelity DNA polymerase
   (TransGen), and cDNA derived from the H5N1 virus
   (A/Chicken/ShanXi/2/2006) was used as the template. To construct the
   FLAG-tagged C-terminus fusion proteins, a 3 × FLAG tag was inserted
   into the C-terminus of the pcDNA 3.1 vector, and the PA, PB1, PB2, NP,
   HA, NA, M1, and NS1 genes were cloned upstream of the tag using the
   Seamless Assembly Cloning Kit (CloneSmarter). For the GFP-tagged
   proteins, a GFP tag was inserted into the C-terminus of the pcDNA 3.1
   vector, and the NS2, M2, and PB1-F2 genes were cloned upstream of the
   GFP tag using the same method. All the expression vectors were
   validated by sequencing.

Antibodies

   The following antibodies were used in this study: anti-FLAG [Abmart, #
   M20008L, WB (1:2,000)], anti-Myc [Abmart, # M20002L, WB (1:2,000)], and
   anti-Lamin B1 [Abcam, # ab16048, WB (1:2,000)].

Protein Co-immunoprecipitation and Western Blotting

   The transiently transfected cells were washed twice with
   phosphate-buffered saline (PBS) and were then lysed in
   radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris–HCl, pH 7.4,
   150 mM NaCl, 0.25% deoxycholic acid, 1% NP-40, 1 mM EDTA, and 0.5% SDS
   supplemented with protease inhibitor, Roche). The whole-cell lysate was
   firstly precleared with a protein A/G slurry (Millipore) and was then
   incubated with 40 μl of the anti-Flag affinity gel (Sigma-Aldrich) at
   4°C for 2 h or with 10 μl of the GFP antibody at 4°C overnight, and
   then, it was incubated with 40 μl of the protein A/G slurry (Millipore)
   at 4°C for 2 h. The immunoprecipitated samples were washed four times
   with RIPA buffer and twice with 54K buffer (50 mM Tris–HCl, pH 7.4, 150
   mM NaCl, and 0.25% Triton X-100 supplemented with protease inhibitor).
   The FLAG tag-associated proteins were eluted with 250 ng/μl of the Flag
   peptide (Sigma) by rocking the samples on a tilted rotator at 4°C for 2
   h. The GFP tag-associated proteins were eluted with ammonium hydroxide
   at 4°C for 2 h, and the supernatant was collected with a vacuum
   centrifugal concentrator.

   For Western blotting, SDS electrophoresis was performed, and the
   separated proteins were transferred to polyvinylidene fluoride (PVDF)
   membranes. The membranes were blocked and then incubated with the
   corresponding antibodies. The proteins were visualized using the
   Immobilon Western Chemiluminescent HRP Substrate (Millipore).

Mass Spectrometry

   Strict experimental controls were used for the MS analysis. The
   immunoprecipitated samples from the empty FLAG-transfected cells (empty
   FLAG control)/empty GFP-transfected cells (empty GFP control) and the
   protein complexes that were pulled down by the normal IgG (IgG control)
   were also subjected to mass spectrometry for identification. All the
   proteins identified in these two sets of controls were excluded from
   consideration as H5N1 AIV interacting proteins. The authentic
   FLAG-precipitated proteins and the GFP-precipitated proteins associated
   with H5N1 AIV were examined in triplicate. Proteins that were enriched
   by co-immunoprecipitation were separated by SDS-PAGE, and the entire
   lane was cut and sent for tryptic digestion.

   Proteins pulled down by the anti-FLAG beads or the anti-GFP antibody
   were digested with trypsin for 20 h at room temperature. The peptides
   were extracted twice with 50% aqueous acetonitrile containing 0.1%
   formic acid, dried in a SpeedVac, and then desalted using Sep-Pak C18
   cartridges. Tandem mass tag (TMT) reagents (Thermo Fisher) were used to
   label the purified peptides according to the manufacturer's
   instructions. Briefly, the TMT labeling reagents, in anhydrous
   acetonitrile, were carefully added to the desalted peptides and
   incubated for 1 h at room temperature, after which the reactions were
   stopped by hydroxylamine, and the TMT-labeled peptides separated by
   reverse phase (RP) chromatography. The first dimension RP separation by
   micro-liquid chromatography (LC) was performed on an Ultimate 3000
   System (Thermo Fisher) using an Xbridge C18 RP column (5 μm, 150 Å, 250
   mm × 4.6 mm i.d., Waters). The samples were reconstituted in 5–10 ml of
   mobile phase A (2% acetonitrile, 0.1% formic acid, pH adjusted to 10.0
   with NH[4]OH). The peptides were monitored at 214 nm, and 1-min
   fractions were collected, dried, and reconstituted in 20 μl of 0.1%
   (v/v) formic acid in water for the nano-LC-MS/MS analyses, in which the
   fractions were further separated on a C18 column (75 μm inner diameter,
   150 mm length) with a flow rate of 250 nl/min. A gradient was formed,
   and the peptides were eluted with increasing concentrations of solvent
   B (98% acetonitrile, 0.1% formic acid, pH to 10.0, as above). This
   second-dimension separation was performed on an Orbitrap Q Exactive
   mass spectrometer that was operated in data-dependent acquisition mode
   using Xcalibur 3.0 software. The scan range was from m/z 300 to 1,800,
   with a resolution of 70,000 at m/z 400. A full scan followed by 20
   data-dependent MS/MS scans was acquired with collision-induced
   dissociation having a normalized collision energy of 35%. The MS/MS
   spectra obtained from each LC-MS/MS run were searched against a protein
   database using the Proteome Discoverer searching algorithm. The
   precursor ion mass tolerance was set at 20 ppm, and the fragment ion
   mass tolerance was 20 mmu. One missed cleavage by trypsin was allowed.
   Oxidation (Met) was chosen as the variable modification, and
   carbamidomethyl (Cys) and TMT6plex were chosen as the fixed
   modifications.

Significance Analysis of Interactome

   The MS data were analyzed using the Significance Analysis of
   Interactome (SAINT) tool, and only the results with a SAINT score >0.9
   were considered for the subsequent analysis. To define the
   high-probability interaction sets, we selected the interactions at a
   probability threshold of 0.9, which was approximately equivalent to an
   estimated false discovery rate (FDR) of 2% ([41]12).

Gene Ontology Term Enrichment and Pathway Enrichment Analysis

   Gene Ontology (GO) enrichment and pathway enrichment were conducted
   using the Database for Annotation, Visualization and Integrated
   Discovery (DAVID) ([42]13).

Domain Enrichment Analysis

   The protein domain enrichment analysis was conducted for the H5N1
   AIV-associated host proteins using FunRich
   ([43]http://funrich.org/index.html).

Immunofluorescence Staining and Confocal Microscopy Analysis

   Chicken DF1 cells transiently transfected with FLAG-NS1 and Myc-STAU2
   plasmids were cultured for 24 h and were then fixed in 4%
   paraformaldehyde for 30 min at room temperature and were permeabilized
   with 0.1% Triton X-100 (Sigma, # T8787) for 15 min. FLAG-NS1 cells were
   incubated with an anti-FLAG antibody (Abmart, # M20008L) and Myc-STAU2
   cells with an anti-Myc antibody (Abmart, # M20002L), both at a 1:1,000
   dilution for 2 h. After three washes in PBS, the cells were
   subsequently incubated with a FITC-conjugated secondary antibody
   (Abcam, # ab6785) for FLAG and with a Cy3.5-conjugated secondary
   antibody (Abcam, # ab6954) for Myc, both at a 1:1,000 dilution for 1 h.
   The nuclei were stained with 6-diamidino-2-phenylindole (DAPI) (Sigma,
   # D9542, 1:500). Finally, the cells were visualized with a confocal
   microscope (Nikon A1 R MP).

RNA Interference

   All the small interfering RNAs (siRNAs) used in this study were
   designed and synthesized by Guangzhou Ruibo (Guangzhou, China). Chicken
   DF1 cells, at 90% confluence in six-well plates, were transfected with
   100 nM of an effective siRNA specific for the chicken STAU2 gene (Gene
   ID: [44]420184: siSTAU2, sense 5′-CCTACAAGCTCTCCAGAAT-3′; siNC, sense
   5′-GUGAACGAACUCCUUAAUUTT-3′). Human 293T cells, at 65% confluence in
   six-well plates, were transfected with 100 nM of an effective siRNA
   specific for the human STAU2 gene (Gene ID: [45]27067: siSTAU2, sense
   5′-CCAAGGGAUGAACCCUAUUTT-3′). All the siRNAs were transfected into the
   cells using Lipofectamine 3000 (Life Technologies).

Cell Culture and Viral Transfections

   The highly pathogenic H5N1 strain A/mallard/Huadong/S/2005 (YS),
   showing a high virulence in mice, was isolated from a mallard. Chicken
   DF1 embryonic fibroblast cells, 293T cells, and MDCK cells were
   cultured in DMEM supplemented with 10% fetal bovine serum (Gibco), 100
   μg/ml streptomycin, and 100 units/ml penicillin at 37°C under a
   humidified atmosphere of 5% CO[2]. The transfection with siRNA or
   plasmid was performed with Lipofectamine 3000 (Life Technologies). The
   cells were harvested for protein extraction 48 h after the
   transfection.

Virus Titration

   The 50% tissue culture infectious dose assay (TCID[50]) was used to
   evaluate progeny virus production. The siRNA-STAU2- or
   siRNA-NC-transfected DF1 cells grown in six-well plates were infected
   with H5N1 AIV, A/mallard/Huadong/S/2005 (SY), at a multiplicity of
   infection (MOI) of 0.1. After a 1-h incubation, the supernatants were
   discarded, and the cells were washed twice with PBS. Subsequently, 2 ml
   of DMEM without fetal bovine serum was added. The supernatants were
   collected at 12 and 18 h postinfection (h.p.i.), and the virus titers
   were determined in the MDCK cells. The virus titer was calculated as
   the TCID[50] per 0.1 ml using the Reed and Muench method. The data are
   shown as the means ± standard deviations (SD) from three independent
   experiments. An independent-sample t-test was used to analyze the
   TCID[50] results. For all the tests, P ≤ 0.05 was considered
   significant.

Extraction of Nuclear and Cytoplasmic RNA

   The DF1 cells were transfected with an effective siRNA specific for
   STAU2 for 12 h and then were transfected with the NS1-FLAG plasmid for
   24 h. The DF1 cells and 293T cells were infected with the H5N1 avian
   influenza virus for 10 h. The fresh cells were collected and were then
   separated into the nuclear and cytoplasmic fractions to extract the
   nuclear RNA and cytoplasmic RNA, respectively, using the PARIS™ Kit
   (Life Technologies, USA).

Real-Time PCR

   Total cellular RNA was extracted by TRIzol (Invitrogen) according to
   the manufacturer's instructions. Next, cDNA was generated from 1 μg of
   RNA using the PrimeScript RT reagent kit with gDNA Eraser (TaKaRa). The
   target mRNAs were quantified by real-time PCR using SYBR Green Master
   Mix. The data were normalized to the expression of the housekeeping
   genes GAPDH and β-actin. NS1 expression in the nucleus was normalized
   to the expression of U6, which is only expressed in the nucleus. The
   sequences of the PCR primers used to amplify target genes are listed as
   follows: STAU2-chicken: sense 5′-AGTGCCTAAAATTTTCTATGT-3′, antisense
   5′-GCTTCCCCATTCTGAGGTAAT-3′; STAU2-human: sense
   5′-AAGCACTGCAGAATGAA-3′, antisense 5′-GCAATTTCAAACACTAA-3′; NS1: sense
   5′-TGCGGGAAAGCAGATAGT-3′, antisense 5′-TGGGCATGAGCATGAACC-3′;
   GAPDH-chicken: sense 5′-CCCCCATGTTTGTGATGGGT-3′, antisense
   5′-TGATGGCATGGACAGTGGTC-3′; GAPDH-human: sense
   5′-AACCATGAGAAGTATGAC-3′, antisense 5′-GATGGCATGGACTGTGGT-3′;
   β-actin-chicken: sense 5′-GAGAAATTGTGCGTGACATCA-3′, antisense
   5′-CCTGAACCTCTCATTGCCA-3′; β-actin-human: sense
   5′-CTCCATCCTGGCCTCGCTGT-3′, antisense 5′-GCTGTCACCTTCACCGTTCC-3′;
   U6-chicken: sense 5′-CTCGCTTCGGCAGCACATAT-3′, antisense
   5′-TGGAACGCTTCACGAATTTG-3′; and U6-human: sense 5′-TTCGGCAGCACATATA-3′,
   antisense 5′-ATATGGAACGCTTCAC-3′.

Results

Identification of Host Proteins Associated With 11 Viral H5N1 AIV Proteins

   We aimed to systematically and quantitatively identify host proteins
   associated with H5N1 AIV proteins using AP-MS. To this end, eight
   FLAG-tagged viral proteins (i.e., PA, PB1, PB2, NP, HA, NA, M1, and
   NS1) and three GFP-tagged viral proteins (i.e., PB1-F2, M2, and NS2) of
   an avian influenza virus (A/Chicken/ShanXi/2/2006, H5N1 subtype) were
   individually expressed in chicken DF1 cells. Affinity purification was
   carried out using FLAG tag or GFP tag antibodies to immunoprecipitate
   the host proteins associated with the tagged viral proteins. The
   Western blot results showed that all the viral genes were expressed in
   the chicken cells ([46]Supplementary Figure 1). The immunoprecipitated
   samples from the plasmid-transfected cells were separated by SDS-PAGE
   for MS analysis ([47]Figure 1A). In order to generate AP-MS data with a
   high degree of confidence, all the samples (including those from the
   empty FLAG/GFP-transfected cells and the IgG control samples) were
   subjected to the SAINT analysis ([48]12), and only the results with a
   SAINT score >0.9 were considered for the subsequent analysis. The MS
   analysis of the coprecipitated proteins identified 621 host proteins in
   total, and among these, 103, 41, 23, 131, 81, 213, 151, 38, 17, 186,
   and 62 coprecipitated with the viral PA ([49]14), PB1, PB1-F2, PB2, NP,
   HA, NA, M1, M2, NS1, and NS2 proteins, respectively ([50]Figure 1B,
   [51]Supplementary Table 1).

Figure 1.

   [52]Figure 1
   [53]Open in a new tab

   Affinity purification of H5N1 avian influenza virus (AIV) proteins. (A)
   Flowchart of the proteomic affinity purification mass spectrometry
   (AP-MS) method used to define the H5N1 AIV–host interactome. (B) The
   number of host proteins identified by AP-MS that interact with
   individual H5N1 avian influenza virus proteins.

H5N1 AIV–Chicken Interactome

   We next plotted a network representation of the 621 H5N1 AIV–chicken
   protein interactions identified in this study ([54]Figure 2A). This
   network contained nodes corresponding to 11 H5N1 AIV proteins (purple)
   and the 621 host proteins derived from the chicken DF1 cells (pink and
   green). The outermost pink nodes and the innermost green nodes
   represent the host proteins that interact with only one viral protein
   or with multiple viral proteins, respectively. The network structure
   suggests that a large proportion of the interacting host proteins are
   not only factors for this particular influenza virus but may have more
   general biological functions and interact with multiple viral proteins.

Figure 2.

   [55]Figure 2
   [56]Open in a new tab

   Global landscape of the H5N1 AIV–chicken protein complexes. In total,
   1,043 H5N1 AIV–chicken interactions are represented, linking 11 H5N1
   AIV proteins and 621 chicken factors. (A) Types of host interacting
   proteins. Chicken proteins that interact with H5N1 AIV proteins (purple
   nodes in the middle ring) are denoted with two colors: the pink nodes
   in the outer ring represent the host proteins that interact with only
   one specific virus protein, while the green nodes in the inner ring
   represent the host proteins that interact with a variety of viral
   proteins. (B) Network representation of the H5N1 AIV–human PPIs. The
   host–pathogen interaction map for H5N1 AIV contained 1,475 interactions
   between 11 influenza virus proteins (red nodes) and 621 cellular
   proteins (blue and pink nodes). Some of the host proteins interact with
   the virus in the form of protein complexes (pink nodes).

   Next, we analyzed all of the H5N1 AIV–host protein interactions using
   Search Tool for the Retrieval of Interacting Genes/Proteins (STRING)
   ([57]15) and plotted a network representation ([58]Figure 2B,
   [59]Supplementary Table 2) incorporating both the MS and STRING
   analyses ([60]Supplementary Table 3) using Cytoscape. In this network,
   the red nodes represent the viral proteins, the remaining are the host
   interaction proteins, and the pink nodes form the subnetworks in the
   protein–protein interaction (PPI) map. The green lines represent the
   interactions among the virus and host proteins, while the gray lines
   represent interactions among the host proteins. In addition to
   interactions with the viral proteins, it was evident that many of the
   host interacting proteins also interacted with the other host proteins,
   and some even associated with the viral proteins in the form of protein
   complexes ([61]Figure 2B).

GO, Pathway, and Domain Enrichment Analysis of the Host Interacting Partners
of H5N1 AIV

   We performed a GO analysis and pathway enrichment on the host
   interactors of the H5N1 AIV proteins using the DAVID analysis tool. A
   number of crucial GO terms and pathways were identified as enriched in
   the AIV–chicken interacting proteins ([62]Tables 1, [63]2), indicating
   that these signaling pathways and associated genes play important roles
   in the life cycle of the influenza virus.

Table 1.

   Gene Ontology terms enriched among host proteins interacting with H5N1
   AIV.
   GO term Count Involved genes/total genes (%) P-value
   Molecular function
   Poly(A) RNA binding 116 22.4 1.30E-37
   ATP binding 106 20.5 2.50E-15
   RNA binding 37 7.1 1.10E-11
   Nucleotide binding 35 6.8 4.40E-10
   ATP-dependent RNA helicase activity 13 2.5 4.80E-07
   Actin filament binding 13 2.5 1.10E-06
   Structural constituent of ribosome 21 4.1 2.30E-05
   mRNA binding 14 2.7 2.70E-05
   Translation initiation factor activity 10 1.9 5.20E-05
   Double-stranded RNA binding 9 1.7 3.30E-04
   Cellular component
   Extracellular exosome 149 28.8 2.90E-22
   Membrane 86 16.6 1.40E-19
   Myelin sheath 29 5.6 2.30E-13
   Nucleoplasm 89 17.2 7.40E-11
   Focal adhesion 39 7.5 1.10E-10
   Catalytic step 2 spliceosome 16 3.1 5.60E-08
   Mitochondrial inner membrane 23 4.4 4.10E-07
   U1 snRNP 7 1.4 3.90E-06
   Nucleolus 40 7.7 7.50E-06
   U5 snRNP 7 1.4 1.30E-05
   U2 snRNP 7 1.4 1.30E-05
   Biological process
   RNA secondary structure unwinding 10 1.9 3.90E-06
   Translation 20 3.9 1.50E-05
   Protein folding 15 2.9 2.80E-05
   IRES-dependent viral translational initiation 5 1 3.00E-05
   mRNA splicing, via spliceosome 11 2.1 2.20E-04
   Negative regulation of translation 8 1.5 3.60E-04
   Osteoblast differentiation 11 2.1 4.00E-04
   Viral translational termination-reinitiation 4 0.8 4.80E-04
   mRNA processing 10 1.9 7.40E-04
   Regulation of translational initiation 6 1.2 8.50E-04
   Protein import into nucleus 8 1.5 1.00E-03
   [64]Open in a new tab

Table 2.

   Pathway enrichments of host proteins interacting with H5N1 AIV.
   Term Count Involved genes/total genes (%) P-value
   Spliceosome 28 5.4 8.40E-11
   Proteasome 11 2.1 4.40E-05
   RNA transport 22 4.2 4.50E-05
   Citrate cycle (TCA cycle) 9 1.7 2.10E-04
   Ribosome 19 3.7 3.00E-04
   DNA replication 9 1.7 3.60E-04
   Carbon metabolism 16 3.1 7.40E-04
   Protein processing in endoplasmic reticulum 20 3.9 1.50E-03
   Pyruvate metabolism 8 1.5 4.20E-03
   Salmonella infection 11 2.1 7.00E-03
   Biosynthesis of antibiotics 21 4.1 8.50E-03
   Aminoacyl-tRNA biosynthesis 8 1.5 1.40E-02
   Ribosome biogenesis in eukaryotes 10 1.9 2.00E-02
   Mismatch repair 5 1 3.20E-02
   Pentose phosphate pathway 5 1 4.40E-02
   [65]Open in a new tab

   Protein domains provide insight into the preferential structure of an
   interaction partner. To survey the binding preferences of the H5N1 AIV