scientificscientificreport report Structure homology and interaction redundancy for discovering virus–host protein interactions Benoıˆt de Chassey 1,Laure`ne Meyniel-Schicklin 2,3, Anne Aublin-Gex 2,3, Vincent Navratil 2,3,w, Thibaut Chantier 2,3, Patrice Andre´1,2,3 &VincentLotteau2,3+ 1Hospices Civils de Lyon, Hoˆpital de la Croix-Rousse, Laboratoire de virologie, Lyon , 2Inserm, U1111, Lyon , and 3Universite´ de Lyon, Lyon, France

Virus–host interactomes are instrumental to understand global purification coupled to mass spectrometry, led to the publication perturbations of cellular functions induced by infection and of the first virus–host interactomes [3–4]. Although incorporation discover new therapies. The construction of such interactomes is, of data from different methods or variation of the same method [5] however, technically challenging and time consuming. Here we has improved the quality of the data sets, diversification of describe an original method for the prediction of high-confidence methods is still clearly needed to generate high-quality interactions between viral and human proteins through comprehensive virus–host interactomes. In addition, regarding a combination of structure and high-quality interactome data. the size of host genome and the huge diversity of viruses, millions Validation was performed for the NS1 protein of the influenza of binary interactions remain to be tested. Therefore, accurate and virus, which led to the identification of new host factors that rapid methods for the identification of cellular interactors control viral replication. controlling viral replication is a major issue and a crucial step Keywords: interactome; prediction; protein interaction; towards the selection of original therapeutic targets and drug structure; virus development. This could benefit from predictive methods EMBO reports advance online publication 6 September 2013; doi:10.1038/embor.2013.130 preselecting subsets of putative interactions. Computational methods are essentially dedicated to predict intraspecies interac- tions [6] but applications to pathogen–host protein inter- actions are now emerging [7]. These methods are generally INTRODUCTION evaluated by the overlap between predicted and published Viruses are obligate parasites that rely on cellular functions to interaction data sets. Here, we present and experimentally replicate. Identifying the cellular functions that a virus must use, validate an original method for the prediction of virus–host counteract or interfere with to assure its replication is therefore a interactions combining protein structure homology and major issue for both basic knowledge and drug discovery. interaction redundancy. Biological processes being largely dependent on protein–protein interactions, cellular functions involved in viral replication can be identified by mapping virus–host protein–protein interactions and RESULTS AND DISCUSSION by constructing virus–host interactomes. Until 2007, viral protein– Principle of the method host protein interactions have essentially been identified in Several methods have proposed to use structural information to low-throughput experiments. Several databases are integrating predict protein–protein interactions [8–10]. The method described these data spread in the literature [1–2]. The construction of here relies on the assumption that when two proteins are physical viral ORFeomes and the development of high-throughput structurally homologous, they are more likely to have interactors technologies, such as yeast two-hybrid (Y2H) and tandem affinity in common (Fig 1A). Using the Structural Classification of Proteins from SCOP database [11] and the high-quality data sets of 1Hospices Civils de Lyon, Hoˆpital de la Croix-Rousse, Laboratoire de virologie, 103 grande rue de la Croix-Rousse, Lyon, F-69004, protein–protein interactions from the VirHostNet database [2], we 2Inserm, U1111, 21 Avenue Tony Garnier, Lyon, F-69007, showed that this assumption is a general feature of human 3Universite´ de Lyon, Lyon, France and viral proteins (supplementary information online). Briefly, wPresent address: Laboratoire de Biome´trie et Biologie E´volutive, CNRS, Universite´ Claude Bernard Lyon 1, 43 boulevard du 11 Novembre 1918, 69622 Villeurbanne for a viral protein having a solved structure, structurally Cedex, France. homologous human and viral proteins are first selected. Inter- +Corresponding author. Tel: þ 33 437262971; Fax: þ 33 437282341; actors of these homologous proteins are then identified. These E-mail: [email protected] proteins are considered as putative interactors and ranked Received 14 February 2013; revised 29 July 2013; accepted 31 July 2013; according to a score that favors proteins independently identified published online 6 September 2013 from multiple structural homologues.

&2013 EUROPEAN MOLECULAR BIOLOGY ORGANIZATION EMBO reports 1 Prediction of virus–host protein interactions scientificreport B. de Chassey et al

A Protein of interest Viral and human Human Best predicted New picture of (full-length 3D structural interactors interactors the interactome or domain) homologs

Scoring

Protein-protein interaction db (VirHostNet)

B 16,187 HH ppi VirHostNet high–quality dataset 72,339 HH ppi

Chain 22,052 PDB 14,422 H protein chains (3,163 Splitting 344 homologous 1,377 predicted structures H proteins) H proteins H interactors

1/all Influenza NS1 full- pairwise 2< 5 % Experimental length structure structure z-score 69 predicted Scoring benchmarking PDB (3F5TA) alignments <20 best H interactors (Dali)

Score 4,933 PDB 3,130 V protein Chain 42 homologous 86 predicted >3 chains (440 V structures Splitting V proteins H interactors proteins)

26 predicted VirHostNet 2,539 VH ppi high- H interactors 4,458 VH ppi quality data set

Fig 1 | Description of the method. (A) Principle of the method. (B) Detailed steps of the method applied to influenza A virus NS1 protein. The score 2 2 favors independent predictions provided by interaction redundancies. Score (putative interactor) ¼ [(nH /kH) þ (nV /kV)] a. With nH and nV number of distinct human interactors and viral interactors, respectively; kH and kV degree of the protein in the human protein–protein interaction data set and in the virus–human protein–protein interaction data set, respectively; a ¼ 3 for a prediction from both human and virus–human data sets, a ¼ 2 from only virus–human data set, 1 else. H, human; V, virus.

Prediction of influenza A virus NS1 protein interactions of protein–protein interactions from the VirHostNet database [2], As a proof of concept, we applied this method to identify new a total of 1,384 human proteins were predicted to interact with partners of the influenza A NS1 protein whose full-length structure NS1- 1,377 interacting with human proteins that are homologous has been solved by x-ray crystallography [12], (Fig 1B). A total of to NS1 and 86 interacting with viral proteins that are homologous 22,052 human protein structures and 4,933 viral protein structures to NS1. These 1,384 proteins were then ranked according to respectively corresponding to 3,163 human proteins and 440 viral the number and origin of the NS1 homologues weighted by proteins were extracted from the Protein Data Bank (PDB). The their degree. The 5% top-ranked candidates, corresponding to standalone version 3.1 of the Dali programme was used to perform 69 proteins, were selected for further validation (Fig 1B, one-against-all pairwise alignments between NS1 and these large supplementary Table S1 online). In silico mapping of the predicted data sets of viral and human protein structures [13]. NS1 structural interactions can also be performed when the structure of domains homologues with a Dali z-score between 2 and 20 were retained or fragments are available. For NS1, an RNA-binding domain to include weakly homologous proteins and exclude fully and an effector domain have been structurally defined (PDB homologous NS1 proteins [14]. All homologues that are full- identifiers: 2Z0A and 3D6R, supplementary Table S2 online). Fifty length or part of NS1 proteins of influenza A viruses were also eight of the 69 predicted interactors for the full-length NS1 protein excluded. Hence, known NS1–human interactors were not favored. were predicted to interact with the RNA-binding domain while Influenza NS1 protein was found structurally homologous to 344 none of them were predicted to interact with the effector domain human proteins and 42 viral proteins. Using high-quality data sets (supplementary Table S1 online).

2 EMBO reports &2013 EUROPEAN MOLECULAR BIOLOGY ORGANIZATION Prediction of virus–host protein interactions B. de Chassey et al scientificreport

C 22.0 AB20.0 Final predicted Interaction siRNA 18.0 cellular interactor validation assay 16.0 Official GST pull Neuraminidase 14.0 Ensembl ID Y2H PFU name down activity 12.0 ENSG00000002330 BAD × × 10.0 ENSG00000153094 BCL2L11 × × ×× 8.0 ENSG00000107223 EDF1 × × 6.0 ENSG00000110324 IL10RA × × ×× ENSG00000169047 IRS1 × × 4.0

ENSG00000143393 PIK4B × × Replication fold increase 2.0

ENSG00000132155 RAF1 × × ××GST GST–NS1 GST–NS1–RBD GST–NS1–effector 0.0 ENSG00000160691 SHC1 × × ENSG00000136560 TANK × × IB: myc BID CTRL TAF1 ENSG00000175564 UCP3 × × ENSG00000168646 AXIN2 × ×× D BCL2L11 ENSG00000087088 BAX × ×× Pull 1.2 ENSG00000015475 BID × ×× down 1 ENSG00000179151 EDC3 × ×× IB: GST ENSG00000030110 BAK1 ENSG00000157764 BRAF 0.8 ENSG00000101224 CDC25B 0.6 ENSG00000189229 CDK11B ENSG00000107566 CHUK 0.4 ENSG00000143466 IKBKE IB: myc Input ENSG00000185950 IRS2 0.2 ENSG00000155506 LAFP1 ×× ENSG00000116191 RALGPS2 Replication fold increase 0 ENSG00000147133 TAF1 ×× RA ENSG00000103197 TSC2 BAX CTRL BRAF RAF1 AXIN2 EDC3 ENSG00000138592 USP8 IL10 LARP1

Fig 2 | Functional characterization of the 26 best-predicted interactors. (A) The physical and genetic interactions have been tested (white) or not (grey). Essential host factors are in green and restriction host factors are in red. (B) BID–NS1 interaction mapping. Interaction of myc–BID with GST–NS1 and GST–NS1 RBD but not with GST alone or GST–NS1 effector domain. (C,D) Impact of silencing indicated genes on viral replication. Values are the number of PFUs normalized to control siRNA. The mean values ±- s.d. from three independent experiments are shown. (C) Restriction host factors. (D) Essential host factors. GST, glutathione S-transferase; PFU, plaque-forming unit; RBD, RNA-binding domain; siRNA, short interfering RNA.

Functional analysis of NS1 interactors score 4 ¼ 3. The validation rate was 83% for predicted interactor Sixty-four of the 69 predicted interactors are known to be having a score o3 (supplementary Table S1 online). Therefore, expressed in the normal lung or trachea in accordance with using a score 4 ¼ 3 as a threshold to select the best interactors, the tropism of the virus (supplementary Table S1 online). Analysis the prediction method allowed the identification of very high- of enriched gene ontology terms and KEGG pathways corres- confidence interactors. Accordingly, from the 69 interactors ponding to the 69 predicted interactors of NS1 confirms and initially predicted, a very high-confidence set of 26 NS1 provides more molecular supports to the interplay of NS1 with interactors was retained (Figs 1B and 2A). Performance assessment essential cellular functions. A significant proportion of the of the prediction method shows that this score threshold leads predicted interactors being involved in apoptosis (Benjamini to an optimal success regarding specificity and precision P-value ¼ 6.31 10 6), this new data set might help clarifying (supplementary information online). Twenty-five out of these 26 the complex and controversial role of NS1 and apoptosis in viral interactors were new, illustrating the potential of this method to replication [15–18]. In the enriched RIG-I signalling pathway predict novel interactions (supplementary Table S1 online). From (Benjamini P-value ¼ 5.4 10 3), NS1 is known [19,20] to the 26 interactors, 24 were predicted to interact with the RNA- interfere with the signalling of viral RNA recognition by the binding domain but not the effector domain of NS1 (supplementary pathogen recognition receptors by interacting with RIG-I and Table S1 online). This was confirmed by GST pull down for the TRIM25 [21,22]. Here, the interaction of NS1 with 4 IKK proteins NS1–BID interaction further validating the method (Fig 2B). further supports its ability to interfere with this signalling at a more As the overlap of predicted interactors for full-length NS1 and downstream step. Predicted interaction of NS1 with nine proteins the RBD is not 100%, both domains and full-length proteins of the mitogen-activated protein kinase pathway (Benjamini are essential to explore the full potential of this method. P-value ¼ 1.9 10 3) is also demonstrative of the crucial role of this pathway whose late activation is required for efficient Impact of silencing NS1 interactors on viral replication replication of the virus [23]. The 26 high-confidence NS1 predicted interactors were further tested for their ability to control influenza virus replication. Validation of NS1 interactors Systematic gene silencing was performed with three independent Experimental validation of the predicted interactions was small interfering RNA per gene and virus production was first performed by Y2H and GST pull down for 32 candidates. evaluated 24 h post infection by measuring neuraminidase activity Eighty-one percent of the interactions were confirmed by Y2H in the culture supernatant. When viral production was inhibited or and 72% by co-affinity precipitation (supplementary Fig S1 enhanced by at least twofold with at least two short interfering online). Interestingly, 100% of the interactions were validated RNAs (siRNAs) in at least two out of three experiments, viral by at least one method when the predicted interactor had a production was further quantified with a plaque-forming unit

&2013 EUROPEAN MOLECULAR BIOLOGY ORGANIZATION EMBO reports 3 Prediction of virus–host protein interactions scientificreport B. de Chassey et al

CHUK BBC3

BNIP1 CDK11B IL10RA BAX BAD CDC25B

TANK BAK1 BID IKBKE FKBP8 M1 BRAF TAF6 NP EDF1 CBL LIFR NS1 PB1 AXIN2 IRS1 NAP1L1 EDC3 UCP3 GRB10 RAF1 BCL2L11 SHC1 LARP1 RALGPS2 TSC2 USP8 PI4KB HA

IRS2 TAF1 MECR BRD8 Viral protein NR0B2

Human protein PA MED1 NRBF2 Human protein/host factor NCOA2 Influenza-human PNRC2 protein–protein interaction

Fig 3 | Predicted influenza–human protein–protein interaction network. Network composed of 108 interactions involving 6 influenza viral proteins and 41 human proteins. Twelve of 41 human proteins have been identified to be host factors for influenza virus replication (from literature and this work). The CHUK-NS1 interaction was already reported in the literature.

assay. Using this criteria, 10 new cellular interactors of NS1 were interact with several viral proteins. Multi-targeting of specific found to directly modulate the replication of influenza A virus— host factors by different proteins of the same virus has been three were functioning as restriction host factors and seven as observed previously, highlighting the importance of such essential host factors (Figs 2A,C,D). None of the tested siRNAs proteins for the virus [24]. Interestingly, viral proteins sharing affected cell survival (supplementary Fig S2 online). Therefore, common interactors are those displaying a certain degree of the method generated a data set that is strongly enriched structural homology (supplementary Table S5 online). Finally, an in modulators of replication—43.5 versus 0.05–1.2% for up-to-date influenza–human interactome combining literature siRNA pangenomic screens. and predicted data are presented in Fig 4.

Proteome-wide prediction of virus–host interactions METHODS Having validated the method, predictions were extended to other Retrieval and annotation of viral and human structures. All proteins of the same virus to construct a predicted influenza– sequences from viral and human protein structures stored at the human interactome. Structural information was available for PDB were retrieved in a Fasta format and blasted to be annotated 13 structures corresponding to nine viral proteins. Following the with a GenBank identification for viral proteins and Ensembl process described above for NS1, 108 interactions were protein identification for human proteins as it is the case in the predicted connecting 41 cellular proteins to six viral proteins VirHostNet database. A total of 4,933 PDB sequences (viral (Fig 3, supplementary Tables S3 and S4 online). The low overlap GenBank identification) and 22,052 PDB sequences (human of our set of interactions with literature reflects the incomplete- Ensembl protein identification) have been retrieved and assigned ness of previously reported influenza virus–host protein to 440 distinct viral proteins and 3,163 distinct human proteins. interaction data (supplementary information online). In addition, For the full-length influenza A virus NS1 protein, the chain A of structural and interactomic data availability restricts the list of the 3F5T PDB structure has been used. predictable interactors, as illustrated for NS1 in supplementary High-quality protein–protein interaction data from VirHostNet. information online. Some human proteins are predicted to Protein–protein interaction data were retrieved from the VirHostNet

4 EMBO reports &2013 EUROPEAN MOLECULAR BIOLOGY ORGANIZATION Prediction of virus–host protein interactions B. de Chassey et al scientificreport

NBPF22P PPP2R5C LY6D NUP54 KIAA0586 C1QA DNM2 GNB2L1 VPS28 TRIM27 NR0B2 PIK3R3HMBOX1 ATM ITM2B HSPA8 IL6R SNAPC4 ZMAT3 ACOT9 UROS NCOA2 PPIAP22 CDC42 NXF1 SP100 GLYAT SSBP2 BCAP29 IGHM FUS SYNCRIP RPS3 HNRNPK TTC12 LYPLA1 OLA1 DAZAP2 IGF2BP2 YIPF6 PCBD1 HNRNPF PNRC2 CD74 DHX15 PCBP2 IGF2BP1DHX30 MCM2 GSN IPO9 HIST1H4B ILF2 MED1 ZBTB1 MCM5 NRF1 RPL8 C16orf45 SSX2IP IKBKB MLH1 SLC25A4 PIK3CB LRRFIP1 RNF5 TMEM86B ELP4 AXIN2 NXTJ EIF2AK2 ABLIM1 SDCBP2 RP52 MCM7 CLNS1A TRIM25 NISCH C7 IL10RA DDX58 HNRNPA3 C14orf166 DDX5 RPL11 PRKRA ACTB STAU1 ATXN2L YWHAZ PTPMT1 RP59 BBC3 UMO1 RBMX SLPI UXS1 BNIP1 YWHAG TUBB2O PIK3R1 LIFR NRBF2 DDX39B ZMAT4 TAF1 KIAA1143 TANK UCP3 HNRNPL PIK3CA HNRNPH3 PA SIAH1 TUBB LZTS2 POLR2A SECISBP2 MCM3 IPO5 M1 HNRPDL HNRNPR NQO2 USP8 TSC2 IKBKE HNRNPU RPS10 PA2G4 LARP1 BRD8 CDK11B EEF1A2 RAE1 HNRNPAB PIK3R2 MAGED1 CHUK PTBP1YWHAB GLYR1 YWHAQ MCM4 MECR TRIP6 KPNA3 BAD IR51 MATR3 C10orf35 COMMD1 CSAG2 CEP152 PI4KB CDC25B GABPB1 PB1 BAK1 SLC25A5 CHMP6 SEPT1 HNRNPA1IRS4 ATP6V1G1 PLAC8 BCL2L11 PABPN1 NS1 YWHAE BID UBE2I ILF3 NUP214 EEF1D CHD6 IKZF3 FKBP8RALGPS2 BAX STX5 PIGQ EDC3 MTAP NEFH HNRNP DABPC1 FTH1 RPS5 HNRNPT PCBP1 DST EEF1A1 KHDRBS1 IRS2 PABPC4 SLC16A9 TAF6 GRB10 NP ZNF346 RAF1 MRI1 NAP1L1 MOV1O TARBP2 RPS24 HLA-B PSMA7 NCAPH2 BRAF GAS8 LNX2 RPLP1 CBL CDC42EP4 SHC1 HNRNPA2B1 PCID2 HA MEOX2 DDB1 HTATSF1 KHDRBS3 C22orf2SLC25ARALY WDR6 MPI DVL2 KPNA4 IGF2BP3 KIAA1967 KPNA5 HNRNPA0 FXR2 MAVS NPM1 BANP ATL1 NBPF8 GLUL DVL3 AKT2 CSDA HRNR COL4A3BP ZBTB25 ISG15 NOMO3 MAGEA11 XPO1 LGALS3BP CRK HNRNPUL1 BHLHE40 YBX1 KPNA1 CRKL ELAVL1 USHBP1 NCL CEP70 GMCL1 LINC00525 NA MT2A KPNA6 EDF1 MZ HSP90AA1 HNRNPH1 EXOSC8 TFCP2 KPNA2 DHX9 CTSB CBS RBPMSTCF1Z TUBA1A CPSF4 UPF1 ERH MAPK9 SETBP1 TRAF1 GOLGA2 CREB3 MIPOL1MAGEA2B DOCK8 CRX OCDC33

BECN1 HSPD1 ZXDC POLR3F TXNIP STIP1 CCDC172 DDX54 MYC TACC1 AKAP1 DNAJB5 Viral protein OCT2 CAPN2 MVP C9orf135 DCAF6 PB2 Human protein ANXA1 POMP DNAJB13 Literature FANCG DUSP14 Human protein/ C1orf94 TRAF2 UBC GCOM1 ZNF251 FLOT1 HSP90AB1NPC2 host factor NSMCE4A ASXL1 NS2 ANKRD18CP ALDH1L1 Influenza-human CHAF1A PPIA COL7A1 FAM156B CES1 protein–protein interaction POLR2B EIF3L NCOA7 TSNAXIP1 FAM96B C1QBP CMTM5 Human protein MCM8 NUP98 Prediction BLZF1 KIF6 CDK9 SMARCB1 NF74 HOOK1 Human protein/ DNAJB1 PNMA1 host factor PCNA FKBP5 CALCOCO1 PLA2G4A PSMD8 RABGEF1 TRIM29 AIMP2 Influenza-human NMI DCTN1 protein–protein interaction NCC27HSPA1A DYNLL2 COPS5 BCR Both Human protein KCNRG CRYAB BPIFB1 SMURF2 PCCA Influenza-human MNAT1 protein–protein interaction ZOCHC17

Fig 4 | Influenza–human interactome combining predicted and literature data. Network composed of 515 interactions involving 10 influenza viral proteins and 364 human proteins. One-hundred and six interactions were obtained from predictions, 407 from literature and 2 from both. Forty-seven of 364 human proteins have been identified to be host factors directly controlling viral replication (literature and this work). database [2]. A high-quality data set of human protein–protein human homologues are favored. The prediction from viral interactions were defined by selecting interactions identified homologues only is also better considered than the prediction by two different methods or in two independent papers. from human homologues only. Finally, as a correlation was Indeed, the human interactome is an integration of data from observed between the score and the degree of a putative interactor several databases, and might be polluted by low-confidence (that is, the number of partners in the protein interaction data set), interactions [25]. A high-quality data set of virus–human protein– the score was weighted by the degree. Altogether, the score 2 2 protein interactions were constructed with exclusively manually formula is as follows: score (putative interactor ¼ [(nH /KH) þ (nV / curated interactions. To experimentally validate the predicted kV)] a, with: (i) nH and nV number of distinct human interactors interactions, an additional level of stringency was added by and viral interactors, respectively. (ii) kh and kV degree of the focusing on interactions described by classical laboratory tech- protein in the human protein–protein interaction data set and in nologies. Altogether, the predictions are on the basis of 16,187 the virus–human protein–protein interaction data set, respectively. human protein–protein interactions involving 6,378 distinct human (iii) a ¼ 3 for a prediction from both human and virus–human data proteins and 2,539 virus–host protein–protein interactions involving sets, a ¼ 2 from only virus–human data set, 1 else. (iv) H: human, 434 viral proteins and 1,395 human proteins (supplementary V: virus. Tables S6 and S7 online). Cells and virus. A549 human lung epithelial cell line and Madin- Score definition. All putative interactors were individually Darby canine kidney (MDCK) cell line were grown in DMEM scored according to several criteria. First, an interactor is more media supplemented with 100 U ml 1 penicillin/streptomycin confident when predicted by several independent NS1 homo- and 10% fetal calf serum at 37 1C and 5% CO2. The epidemic logues. Interactors independently predicted by both viral and A/H1N1/New Caledonia/2006 strain was propagated in MDCK

&2013 EUROPEAN MOLECULAR BIOLOGY ORGANIZATION EMBO reports 5 Prediction of virus–host protein interactions scientificreport B. de Chassey et al

cells in DMEM supplemented with 1 mgml 1 modified and magnesium and 50 mlof20mM 4-MUNANA. After 1 h trypsin TPCK in absence of fetal calf serum. Virus stocks were incubation at 37 1C, 100 ml of glycine 0.1 M 25% ethanol titrated by standard plaque assay on MDCK cells using an agar pH 10.7 was added. Measures were done with TECAN overlay medium. infinite M1000 instrument at 365 nm excitation and 450 nm GST pull down. Cellular proteins (corresponding to the 32 cDNAs emission wavelengths. available in orfeotech ORFeome v3.1 [26], description of the Plaque-forming unit assay. To quantify infectious virus particles hORFeome v3.1 in supplementary information online) were in infected cell culture supernatants, 300,000 MDCK cells amplified from the MGC cDNA plasmid collection using KOD were seeded in 6-well plates. Three days later, cells were washed polymerase (Toyobo), cloned by recombinational cloning twice with DMEM supplemented with 100 U ml 1 penicillin/ into pDONR207 (Life technologies) and transferred to pDEST- streptomycin. Dilutions of infected cell culture supernatants were myc or pCIneo3xFlag (kind gift of Y. Jacob). H5N1 NS1 was dispensed on the cells. After 1 h 30 min incubation at 37 1C and 5 transferred into pDEST27 (GST fusion in N-term). A total of 4.10 5% CO2 cells were washed again and covered with overlay HEK293T cells were then co-transfected (6 ml JetPei, Polyplus) medium containing MEM 1 , 1% agarose (Lonza) and modified with 1.5 mg of each pair of plasmid. Two days after transfection, trypsin TPCK 1 mgml 1. Plates were incubated upside down at cells were harvested and lysed (0.5% NP-40, 20 mM Tris–HCl (pH 37 1C and 5% CO2 up to 72 h. Cells were then fixed with formol 8.0), 150 mM NaCl, 1 mM EDTA and Roche complete protease 10% and coloured with crystal violet 0.3% in methanol 20% inhibitor cocktail). Soluble protein complexes were purified by and lysis plaques were counted. incubating 300 mg of cleared cell lysate with 40 ml glutathione Functional annotation. DAVID functional annotation chart tool sepharose 4B beads (GE Healthcare). A total of 50 mg of cleared was used to perform enrichment analysis, in which an annotation cell lysate was separated on SDS–PAGE and transferred to term was considered as significant with a Benjamini–Hochberg nitrocellulose membrane. GST-tagged viral proteins and Myc or corrected P-value smaller than 0.05 [27]. 3XFlag-tagged cellular proteins were detected using standard immunoblotting techniques with anti-GST (Covance), anti-Myc or Supplementary information is available at EMBO reports online anti-Flag (Sigma) monoclonal antibodies. (http://www.emboreports.org). Yeast two-hybrid. Pairwise Y2H interactions were analysed by yeast mating, using Y187 and AH109 yeast strains (Clontech). ACKNOWLEDGEMENTS Prey and bait vectors were transformed into Y187 and AH109. The authors thank Liisa Holm for her technical help on Dali. Bait and prey strains were mated in an all-against-all array and This work was supported by the European Union’s 7th Program plated on a selective medium lacking histidine and supplemented (FP7/2007–2013) under grant agreement no. 267429 (SysPatho) and with increasing concentrations of 3-amino-triazole to test the European Community’s 7th Framework Program (FP7/2007–2013) interaction-dependent transactivation of HIS3 reporter gene. and Cosmetics Europe under grant agreement no.266753 (SCR&Tox). Author contributions: B.d.C conceived the method; B.d.C. and V.L. Some of the preys are membrane proteins but reported in the supervised the experiments; B.d.C., L.M.-S. and V.N. designed and literature to give positive results in Y2H (for example, BCL2L11, performed bioinformatic studies; A.A.-G., T.C. and B.d.C. designed BAD) or proteins whose signal peptide is likely to be masked and performed biological studies; P.A., A.A.-G., B.d.C., V.L. and (supplementary information online). Therefore, we did not L.M.-S. analysed the data; B.d.C., V.L. and L.M.-S. wrote the manuscript. exclude them from the Y2H. siRNA screening. Five picomoles of each siRNA (stealth select CONFLICT OF INTEREST RNAi, Invitrogen) was arrayed in 96 plates in 10 ml of OptiMEM The authors declare that they have no conflict of interest. (GibCo). Ten microliters per well of a mix lipofectamine RNAiMAX (Invitrogen)-OptiMEM (0.2 ml of lipofectamine RNAi- REFERENCES MAX in 10 ml of OptiMEM) was added. After 20 min room 1. Chatr-aryamontri A et al (2009) VirusMINT: a viral protein interaction temperature-incubation, siRNA–lipofectamine mixes were added database. Nucleic Acids Res 37: D669–D673 to 30 103 A549 suspension cells. Cells were incubated for 48 h 2. Navratil V, De Chassey B, Meyniel L, Delmotte S, Gautier C, Andre P, 1 Lotteau V, Rabourdin-Combe C (2009) VirHostNet: a knowledge base for at 37 C and 5% CO2 before influenza virus infection at the management and the analysis of proteome-wide virus-host interaction multiplicity of infection 0.5. Forty-eight hours after infection, networks. Nucleic Acids Res 37: D661–D668 supernatants were titrated. 3. de Chassey B et al (2008) Hepatitis C virus infection protein network. Mol Virus infection. siRNA-transfected cells were washed twice Syst Biol 4: 230 4. Jager S et al (2012) Global landscape of HIV-human protein complexes. with Dulbecco’s phosphate-buffered saline 1 and infected Nature 481: 365–370 with H1N1 influenza strain at multiplicity of infection 0.5 in 5. Braun P et al (2009) An experimentally derived confidence score for Dulbecco’s modified Eagle medium supplemented with binary protein-protein interactions. Nat Methods 6: 91–97 0.2 mgml 1 trypsin TPCK (infection medium). After 60 min at 6. Jessulat M et al (2011) Recent advances in protein-protein interaction 37 1C, the inoculum was discarded and cells were washed again prediction: experimental and computational methods. Expert Opin Drug 1 Discov 6: 921–935 and incubated for 48 h in infection medium at 37 C and 5% CO2. 7. Mukhopadhyay A, Maulik U, Bandyopadhyay S (2012) A novel Titre measure by neuraminidase activity. Influenza virus biclustering approach to association rule mining for predicting neuraminidase is able to cleave the methyl-umbelliferyl- HIV-1-human protein interactions. PLoS One 7: e32289 N-acetylneuraminic acid (4-MUNANA, Sigma) modifying its 8. Aloy P, Russell RB (2002) Interrogating protein interaction networks through structural biology. Proc Natl Acad Sci USA 99: 5896–5901 emission wavelength in a dose-dependent manner. In 96-black 9. Davis FP, Braberg H, Shen MY, Pieper U, Sali A, Madhusudhan MS plate, 25 ml infection supernatants were diluted in 25 ml (2006) Protein complex compositions predicted by structural similarity. Dulbecco’s phosphate-buffered saline 1 containing calcium Nucleic Acids Res 34: 2943–2952

6 EMBO reports &2013 EUROPEAN MOLECULAR BIOLOGY ORGANIZATION Prediction of virus–host protein interactions B. de Chassey et al scientificreport

10. Tuncbag N, Gursoy A, Nussinov R, Keskin O (2011) Predicting protein- 19. Lu Y, Wambach M, Katze MG, Krug RM (1995) Binding of the influenza protein interactions on a proteome scale by matching evolutionary and virus NS1 protein to double-stranded RNA inhibits the activation of the structural similarities at interfaces using PRISM. Nat Protoc 6: 1341–1354 protein kinase that phosphorylates the elF-2 translation initiation factor. 11. Murzin AG, Brenner SE, Hubbard T, Chothia C (1995) SCOP: a structural Virology 214: 222–228 classification of proteins database for the investigation of sequences and 20. Talon J, Horvath CM, Polley R, Basler CF, Muster T, Palese P, structures. J Mol Biol 247: 536–540 Garcia-Sastre A (2000) Activation of interferon regulatory factor 3 is 12. Bornholdt ZA, Prasad BV (2008) X-ray structure of NS1 from a highly inhibited by the influenza A virus NS1 protein. J Virol 74: 7989–7996 pathogenic H5N1 influenza virus. Nature 456: 985–988 21. Pichlmair A, Schulz O, Tan CP, Naslund TI, Liljestrom P, Weber F, 13. Holm L, Park J (2000) DaliLite workbench for protein structure Reis e Sousa C (2006) RIG-I-mediated antiviral responses to comparison. Bioinformatics 16: 566–567 single-stranded RNA bearing 5’-phosphates. Science 314: 14. Holm L, Kaariainen S, Wilton C, Plewczynski D (2006) Using Dali for 997–1001 structural comparison of proteins. Curr Protoc Bioinformatics Chapter 5 22. Gack MU et al (2009) Influenza A virus NS1 targets the ubiquitin ligase TRIM25 to evade recognition by the host viral RNA sensor RIG-I. Unit 5.5 Cell Host Microbe 5: 439–449 15. Schultz-Cherry S, Dybdahl-Sissoko N, Neumann G, Kawaoka Y, 23. Pleschka S (2008) RNA viruses and the mitogenic Raf/MEK/ERK signal Hinshaw VS (2001) Influenza virus ns1 protein induces apoptosis in transduction cascade. Biol Chem 389: 1273–1282 cultured cells. J Virol 75: 7875–7881 24. Meyniel-Schicklin L, de Chassey B, Andre P, Lotteau V (2012) Viruses 16. Zhirnov OP, Konakova TE, Wolff T, Klenk HD (2002) NS1 protein of and interactomes in translation. Mol Cell Proteomics 11: 014738 influenza A virus down-regulates apoptosis. JVirol76: 1617–1625 25. Cusick ME et al (2009) Literature-curated protein interaction datasets. 17. Ehrhardt C, Wolff T, Pleschka S, Planz O, Beermann W, Bode JG, Nat Methods 6: 39–46 Schmolke M, Ludwig S (2007) Influenza A virus NS1 protein activates the 26. Lamesch P et al (2007) hORFeome v3.1: a resource of human open PI3K/Akt pathway to mediate antiapoptotic signaling responses. J Virol reading frames representing over 10,000 human genes. Genomics 89: 81: 3058–3067 307–315 18. Lam WY, Tang JW, Yeung AC, Chiu LC, Sung JJ, Chan PK (2008) Avian 27. Huang da W, Sherman BT, Lempicki RA (2009) Bioinformatics influenza virus A/HK/483/97(H5N1) NS1 protein induces apoptosis in enrichment tools: paths toward the comprehensive functional analysis human airway epithelial cells. J Virol 82: 2741–2751 of large gene lists. Nucleic Acids Res 37: 1–13

&2013 EUROPEAN MOLECULAR BIOLOGY ORGANIZATION EMBO reports 7