Identification of cellular interaction partners of the influenza virus ribonucleoprotein complex and polymerase complex using proteomic-based approaches

Abstract

Cellular factors that associate with the influenza A viral ribonucleoprotein (vRNP) are presumed to play important roles in the viral life cycle. To date, interaction screens using individual vRNP components, such as the nucleoprotein or viral polymerase subunits, have revealed few cellular interaction partners. To improve this situation, we performed comprehensive, proteomics-based screens to identify cellular factors associated with the native vRNP and viral polymerase complexes. Reconstituted vRNPs were purified from human cells using Strep-tagged viral nucleoprotein (NP-Strep) as bait, and co-purified cellular factors were identified by mass spectrometry (MS). In parallel, reconstituted native influenza A polymerase complexes were isolated using tandem affinity purification (TAP)-tagged polymerase subunits as bait, and co-purified cellular factors were again identified by MS. Using these techniques, we identified 41 proteins that co-purified with NP-Strep-enriched vRNPs and four cellular proteins that co-purified with the viral polymerase complex. Two of the polymerase-associated factors, importin-beta3 and PARP-1, represent novel interaction partners. Most cellular proteins previously shown to interact with either viral NP and/or vRNP were also identified using our method, demonstrating its sensitivity. Co-immunoprecipitation studies in virus-infected cells using selected novel interaction partners, including nucleophosmin (NPM), confirmed their association with vRNP. Immunofluorescence analysis further revealed that NPM is recruited to sites of viral transcription and replication in infected cells. Additionally, overexpression of NPM resulted in increased viral polymerase activity, indicating its role in viral RNA synthesis. In summary, the proteomics-based approaches used in this study represent powerful tools to identify novel vRNP-associated cellular factors for further characterization.

Fig. 1. Purification of cellular interaction partners of influenza virus RNP and polymerasecomplex

A, Purification of NP-Strep from cells transfected with viral RNP components. Left panel: to reconstitute viral RNPs that can be purified by the Strep procedure, 293T cells were transfected with the indicated expression plasmids (pCA) encoding the viral nucleoprotein fused C-terminally fused to STREP-tag (NP-Strep), untagged NP and the polymerase subunits PA, PB1 and PB2 and a plasmid expressing influenza virus minigenome transcripts under the control of a polymerase-I (PolI) promoter. After 24 hours, when viral RNA replication and transcription had occurred (data not shown), a cell extract was prepared for Strep purification. Right panel: Native NP-Strep protein complex bound to Streptactin-beads was released by desthiobiotin, separated on a denaturing linear 4–20% SDS-PAGE and finally visualized by staining with Sypro-Ruby. Numbers on the right side indicate gel slices used for tryptic digestion with trypsin and analysis by MS. Position of NP-Strep and NP as well as PARP-1, DDB1 and NPM are indicated. B, Purification of the influenza virus polymerase complex. Left panel: Cartoon showing the procedure used to purify the viral polymerase complex using TAP-tagged PA. For this purpose, 293T cell were transiently transfected with the indicated expression constructs encoding for the TAP-tagged PA and untagged PB1 and PB2. After 24 hours, the native polymerase complex was purified by the TAP-method (right panel, lane 3). Right panel: HEK 293 T cells were transfected with plasmids encoding the indicated polymerase subunits in untagged or TAP-tagged form. Twenty-four hours after transfection a cell extract was prepared for the TAP-purification. Protein complexes were first bound to IgG-sepharose, released by cleavage with tobacco etch virus protease, and subsequently bound to calmodulin-binding agarose beads. In a final step bound protein complexes were eluted with Ca2+, concentrated, separated by 4–20% SDS-PAGE and visualized by silver staining. The position and names of proteins identified by mass spectrometry are indicated.

Fig. 2. Interaction of NPM and DDB1 with viral RNP in transfected cells and in infected cells

A, To confirm the interaction between DDB1 and reconstituted viral RNP by co-immunoprecipition experiments, 293T cells were transfected with expression plasmids encoding DDB1-Flag and the components of the polymerase complex (PB1-HA, PB2, PA, NP) and a plasmid expressing an influenza virus minigenome (lane 3). Cell extract was prepared 24 hours post transfection and used for immunprecipitation (IP) with HA-specific antibodies. The presence of DDB1-Flag, PB1-HA and NP in the precipitates (upper panels) as well as in the cytoplasmic extracts (lower panels) was analyzed by Western blotting using HA-, Flag and NP-specific antibodies. In control experiments only DDB1-Flag was transfected in 293T cells (lane 1) or all plasmids as in lane 3 with the exception that a plasmid expressing PB1 and not PB1-HA was used (lane 2). Approximately 3% of total cell extract used for IP was loaded. B, Co-immunoprecipitation experiments as in A expressing NPM-Flag. C, To test whether endogenous NPM and DDB1 interact with influenza RNPs during viral infection, MDCK cells were infected with influenza A/WSN/33 virus at a MOI of 0.5 or mock infected. Immunoprecipitation was carried out with total cell extract 12 hours post infection and anti-NP-specific antibodies. Precipitated cell extracts were analyzed by Western blot using antibodies specific for NPM (αNPM), NP (αNP) or PB1 (αPB1). Arrows indicate positions of NPM, NP, IgG heavy chain and PB1. Approximately 3% of total cell extract used for IP was loaded.

Fig. 3. Relocalization of NPM-Flag after influenza virus infection

A, MDCK cells were transiently transfected with an expression plasmid coding for NPM-Flag and subsequently infected with influenza A/WSN/33 virus (MOI 0.5). At the indicated time points post infection, cells were processed for immunofluorescence analysis using a Flag-specific monoclonal antibody to detect NMP-Flag and a rabbit polyclonal anti-NP-specific antiserum to visualize NP. B Magnification of the dotted region shown in A, panels J-L. Note that early in infection (3–4h) as well as in non-infected cells the majority of NPM-Flag localizes in the nucleolus, whereas late in infection (8h) NPM-Flag colocalizes with NP in the nucleus.

Fig. 4. Expression of NPM increases the influenza virus polymerase activity

HEK 293 T cells were transiently transfected with PB1-, PB2-, NP-, PA-expression plasmids and a PolI-Minigenome plasmid coding for the reporter protein firefly luciferase to monitor viral polymerase activity. The transfection mixture also contained a plasmid constitutively expressing renilla luciferase, which served to normalize variation in transfection efficiency. The firefly luciferase activity was determined 24 hour post transfection and normalized to the renilla values. Neg.: transfection reactions without PB2. Pos.: complete set of plasmids. The transfections with increasing amounts of plasmids (100, 500 and 1000 ng) encoding NPM, GFP and a NPM mutant lacking the C-terminal 134 amino acids are indicated. The error bars represent standard deviations.