YB-1 Functions as a Porter To Lead Influenza Virus Ribonucleoprotein Complexes to Microtubules

Abstract

De novo-synthesized RNAs are under the regulation of multiple posttranscriptional processes by a variety of RNA-binding proteins. The influenza virus genome consists of single-stranded RNAs and exists as viral ribonucleoprotein (vRNP) complexes. After the replication of vRNP in the nucleus, it is exported to the cytoplasm and then reaches the budding site beneath the cell surface in a process mediated by Rab11a-positive recycling endosomes along microtubules. However, the regulatory mechanisms of the postreplicational processes of vRNP are largely unknown. Here we identified, as a novel vRNP-interacting protein, Y-box-binding protein 1 (YB-1), a cellular protein that is involved in regulation of cellular transcription and translation. YB-1 translocated to the nucleus from the cytoplasm and accumulated in PML nuclear bodies in response to influenza virus infection. vRNP assembled into the exporting complexes with YB-1 at PML nuclear bodies. After nuclear export, using YB-1 knockdown cells and in vitro reconstituted systems, YB-1 was shown to be required for the interaction of vRNP exported from the nucleus with microtubules around the microtubule-organizing center (MTOC), where Rab11a-positive recycling endosomes were located. Further, we also found that YB-1 overexpression stimulates the production of progeny virions in an Rab11a-dependent manner. Taking these findings together, we propose that YB-1 is a porter that leads vRNP to microtubules from the nucleus and puts it into the vesicular trafficking system.

INTRODUCTION

In general, RNA transcripts form ribonucleoprotein (RNP) complexes with a number of RNA-binding proteins. The destiny of the RNP complexes in terms of localization, stability, and translational control is regulated by their protein constituents (16, 21, 33).

The genome of influenza type A viruses consists of eight-segmented and single-stranded RNAs of negative polarity (vRNA). vRNA exists as RNP complexes (designated vRNP) with viral RNA-dependent RNA polymerase consisting of three subunits (PB1, PB2, and PA) and nucleoprotein (NP). vRNA is transcribed into mRNA and replicated through cRNA (a full-size complementary copy of vRNA) into a large number of progeny vRNAs in the nucleus (reviewed in reference 49). The replicated vRNA is assembled into vRNP, and then the progeny vRNP interacts with M1. The vRNP-M1 complex is exported from the nucleus through the CRM1-dependent pathway mediated by the interaction of the vRNP-M1 complex with NS2 (also called NEP), which is a viral protein containing a nuclear export signal (NES) (19, 52, 54, 77). After the nuclear export, it is quite likely that the progeny vRNP accumulates in the microtubule-organizing center (MTOC) and then moves to the budding site beneath the cell surface along microtubules through Rab11a-dependent vesicular trafficking systems (28, 45). Finally, a set of eight segments of vRNA is incorporated into a progeny virion with other viral structural proteins (51, 53, 79).

The Rab11a-positive recycling endosome is important for the delivery of membranes and core polarity proteins to the lateral cell surface (reviewed in references 25, 42, and 74), leading to the construction of plasma membrane domains and epithelial cell polarity through binding to motor proteins along the cytoskeleton (75). The Rab11a-positive recycling endosome is typically located in close proximity to the nucleus and associated with the MTOC. Recent reports demonstrate that a number of viruses, including influenza virus (1, 17, 47), human cytomegalovirus (36), hantavirus (61), respiratory syncytial virus (6, 73), and Sendai virus (SeV) (9), employ the Rab11a-positive recycling endosomes during the egress. However, the targeting mechanism of cargo molecules, including influenza virus vRNP, to the Rab11a-positive recycling endosome is still poorly understood.

Since only one or two viral proteins are expressed from each segment, the virus has to hijack cellular functions/machineries consisting of numerous cellular proteins to achieve every infection process. Therefore, to understand the regulatory mechanism of the localization and intracellular transport of vRNP, identification and characterization of viral and cellular proteins involved in these processes are required. Here, we identified as a novel vRNP-interacting protein, Y-box-binding protein 1 (YB-1), a cellular protein that is involved in regulation of cellular transcription and translation (41). In the nucleus, YB-1 functions as a Y-box promoter element-binding transcription factor (34, 37, 41). However, a major portion of YB-1 localizes in the cytoplasm and regulates mRNA translation and degradation as a major component of cellular mRNA ribonucleoprotein (mRNP). A sudden translational arrest in response to a variety of stresses is accompanied by the formation of stress granules (SGs) and an increase in the number of mRNA-processing bodies (P-bodies) to reprogram gene expression posttranscriptionally (3). It is suggested that SGs are aggregates of translationally inactive mRNAs containing stalled translation initiation complexes, while P-bodies are mRNP aggregates with proteins involved in mRNA decay and translational repression (2, 21). YB-1 accumulates in these cytoplasmic structures (2) and acts as either a translational activator or inhibitor depending on its amount bound to the target mRNP (55). Therefore, it is proposed that YB-1 determines the fate of cellular mRNPs from their synthesis to destruction.

Here, we found that YB-1 translocates to the nucleus in response to influenza virus infection. The YB-1 imported into the nucleus accumulates in nuclear speckles (promyelocytic leukemia nuclear bodies [PML NBs]), together with vRNP, M1, and NS2 in the presence of leptomycin B (LMB), a potent inhibitor of CRM1, suggesting that YB-1 is associated with the vRNP export complexes in the nucleus. At late phases of infection, YB-1 was found in perinuclear granules with the newly synthesized vRNP and accumulated in Rab11a-positive recycling endosomes along microtubules. Further, we found that YB-1 mediates the interaction of vRNP with microtubules. Taking these findings together, we propose that YB-1 functions as a porter that facilitates the association of the progeny vRNP with microtubules, thereby leading the travel of vRNP onto Rab11a-positivie recycling endosomes.

MATERIALS AND METHODS

Biological materials.

vRNP was prepared from purified influenza A/Puerto Rico/8/34 (A/PR/8/34) virus as previously described (66). Rabbit polyclonal antibodies against YB-1, RAP55, RCK, PB1, PB2, PA, M1, and NP and rat antibody against NS2 were prepared as previously described (30, 31, 40, 50, 71, 72, 77). Rabbit polyclonal antibodies against TIAR, PML (Santa Cruz Biotechnology), and Rab11a (Invitrogen) and mouse monoclonal antibodies against α-tubulin (Sigma) and PML (Santa Cruz Biotechnology) were purchased. Anti-SeV antibody and anti-M1 monoclonal antibody were generous gifts from A. Kato (National Institute of Infectious Diseases, Japan) and S. Hongo and K. Sugawara (Yamagata University), respectively. MDCK, 293T, and HeLa cells were grown in minimal essential medium (MEM) containing 10% fetal bovine serum. For the construction of plasmids expressing His-YB-1 and FLAG-α-tubulin, the cDNAs were amplified with primers 5′-GGAATTCCATATGAGCAGCGAGGCCGAGACCCAGC-3′ and 5′-GAACCGCTCGAGCTCAGCCCCGCCCTGCTCAGCCTCGGGAG-3′ for YB-1 and with primers 5′-CGCCACCATGGACTACAAGGATGACGACGACAAGCATATGCGTGAGTGCATCTCC-3′ and 5′-CTATTAATACTCTTCACCCTCAT-3′ for α-tubulin, and cDNAs were reverse transcribed from HeLa total RNA using oligo(dT)₂₀ primer as a template. YB-1 cDNA was cloned into plasmid pET24b, and the resultant plasmid was then used as a template to amplify FLAG-YB-1 cDNA with primers 5′-GCCGCCACCATGGACTACAAGGATGACGACGACAAGCATATGAGCAGCGAGGCCGA-3′ and 5′-CCCGGATCCTATTACTCAGCCCCGCCCTG-3′. FLAG-YB-1 and FLAG-α-tubulin cDNAs were cloned into plasmid pCAGGS. The cDNA fragments of YB-1 deletion mutants were amplified with primers 5′-GCAGATATCATGAGCAGCGAGGCCGA-3′ and 5′-TGCGGATCCCTACCATGGCCCGCCGGCAGGC-3′ for the A/P domain (amino acids [aa]1 to 50), 5′-GCAGATATCATGAGCAGCGAGGCCGA-3′ and 5′-TGCGGATCCTTAACCAGGACCTGTAACATT-3′ for the ΔA/B domain (aa 1 to 129), 5′-GCAGATATCGACAAGAAGGTCATCGCAA-3′ and 5′-CCCGGATCCTATTACTCAGCCCCGCCCTG-3′ for the ΔA/P domain (aa 51 to 324), and 5′-GCACCATGGGATATCGGTGTTCCAGTTCAAGGC-3′ and 5′-CCCGGATCCTATTACTCAGCCCCGCCCTG-3′ for the A/B domain (aa 130 to 324) and then cloned into plasmid pGEX-6P. To establish HeLa cell lines constitutively expressing either FLAG-YB-1 or FLAG-α-tubulin, HeLa cells transfected with pSV2-Neo and either pCAGGS-FLAG-YB-1 or pCAGGS-FLAG-α-tubulin were selected by growth in the presence of 1 mg/ml G418 for 2 weeks, and then the G418-resistant colonies were isolated. Recombinant His-YB-1 and glutathione S-transferase (GST)-fused deletion mutants of YB-1 were purified according to the manufacturer's protocol. In addition, to remove the bacterial RNA possibly bound to YB-1, we treated recombinant YB-1 with RNase A before purification.

LC-MS analysis.

Molecular masses of trypsin-digested peptides from proteins coimmunoprecipitated with NP were calculated by liquid chromatography-coupled mass spectrometry (LC-MS) analysis (Thermo) after reduction and alkylation of cysteine residues. Assignment of observed ions was done with Mascot search software.

Cellular localization of viral RNAs and proteins.

Indirect immunofluorescence assays and fluorescence in situ hybridization (FISH) assays were carried out as previously described (28). Briefly, cells infected with A/PR/8/34 at multiplicity of infection (MOI) of 10 were fixed with 1% paraformaldehyde (PFA) for 5 min and then prepermeabilized with 0.01% digitonin in phosphate-buffered saline (PBS) for 5 min on ice. After being washed with PBS, cells were fixed in 4% PFA for 10 min and permeabilized with 0.5% Triton X-100 in PBS for 5 min on ice. After incubation in PBS containing 1% bovine serum albumin (BSA) for 1 h, coverslips were incubated with each antibody for 1 h and then with Alexa Fluor 488- or 568-conjugated anti-rabbit, -rat, and -mouse IgG antibodies (Invitrogen). After the indirect immunofluorescence assays, FISH assays were performed using RNA probes complementary to the segment 1 vRNA and cRNA/mRNA. Images were acquired by confocal laser scanning microscopy (Zeiss). Each micrograph is a confocal section taken at the same level of focus among samples, so that nuclei were fully observed with a maximum diameter.

Immunoprecipitation.

Infected cells cross-linked with 0.5% formaldehyde for 10 min at room temperature were lysed by sonication in a buffer containing 20 mM Tris-HCl (pH 7.9), 100 mM NaCl, 30 mM KCl, 0.1% NP-40, and 1 mM EDTA. The lysates were subjected to centrifugation at 12,000 × g, and the supernatant fractions were subjected to immunoprecipitation with antibodies where indicated. For detection of viral RNAs immunoprecipitated with YB-1 from infected cells, the immunoprecipitates were subjected to reverse cross-linking in a buffer containing 50 mM Tris-HCl (pH 7.9), 5 mM EDTA, 50 mM dithiothreitol (DTT), and 1% SDS for 45 min at 70°C. After reverse cross-linking, viral RNAs were purified by phenol-chloroform extraction followed by ethanol precipitation and then reverse transcribed with primers to determine the levels of vRNA (5′-GACGATGCAACGGCTGGTCTG-3′, which corresponds to the segment 5 cRNA between nucleotide sequence positions 424 and 444), cRNA (5′-AGTAGAAACAAGGGTATTTTTCTTTA-3′, which is complementary to the segment 5 cRNA between nucleotide sequence positions 1540 and 1565), and viral mRNA [oligo(dT)₂₀ for poly(A) tail]. The synthesized single-stranded cDNAs were subjected to quantitative real-time PCR analysis (Dice real-time thermal cycler system TP800; TaKaRa) with two specific primers, i.e., 5′-GACGATGCAACGGCTGGTCTG-3′, which corresponds to the segment 5 cRNA between nucleotide sequence positions 424 and 444, and 5′-AGCATTGTTCCAACTCCTTT-3′, which is complementary to the segment 5 cRNA between nucleotide sequence positions 595 and 614.

Gene silencing mediated by siRNA.

Short interfering RNAs (siRNAs) against the Rab11a and YB-1 genes were purchased from Invitrogen. Cells (5 × 10⁵) were transfected with 30 pmol of siRNA using Lipofectamine RNA interference (RNAi) Max (Invitrogen) according to the manufacturer's protocol.

Reconstruction of vRNP-microtubule complex mediated by YB-1.

Cellular tubulin and microtubule-associated proteins (MAPs) purified from bovine brain as previously described (67) were kindly provided by K. Mizumoto (Kitasato University). The purified tubulin proteins (40 μg) were assembled into microtubules by incubation at 37°C for 20 min in a buffer containing 50 mM piperazine-N,N′-bis(2-ethanesulfonic acid) (PIPES)-NaOH (pH 6.8), 1 mM EGTA, 5 mM MgCl₂, 20% glycerol, 1 mM GTP, and 20 μM paclitaxel (originally named taxol). The reconstituted microtubules were precipitated by centrifugation at 32,000 rpm at 20°C for 20 min in an SW55Ti rotor (Beckman) to remove the monomeric tubulin proteins. The precipitates were resuspended in a buffer containing 50 mM HEPES-NaOH (pH 7.9), 50 mM KCl, 20 μM paclitaxel, and 0.5% BSA and then incubated with either vRNP or YB-1-vRNP complexes. The vRNP complexes were immunoprecipitated with anti-NP antibody, and then the coprecipitated microtubules were detected by Western blotting assay with anti-α-tubulin antibody.

RESULTS

Identification of YB-1 as a novel vRNP-interacting protein.

To gain further insight into the regulatory mechanism of vRNP function, we tried to identify a novel cellular protein(s) that interacts with vRNP. At 8 h postinfection (hpi), infected cell lysates were subjected to immunoprecipitation using anti-NP antibody, and then the precipitated proteins were subjected to LC-MS analysis (Fig. 1A). We found that viral proteins PB1, PB2, PA, M1, and NS1 were coprecipitated with NP, as expected (32, 39, 56, 58, 59), suggesting that not only NP but also vRNP was immunoprecipitated in this experiment. In addition to these viral proteins, we also identified a number of cellular proteins interacting with vRNP, as previously reported (Table 1) (29, 43, 44, 46, 76). It is noted that YB-1, a major component of cellular mRNP, was also identified as a novel vRNP-interacting protein. YB-1 regulates the lifetime and the translational activity of cellular mRNP, depending on the amount of YB-1 on a target mRNA (22, 55). Further, it is proposed that YB-1-bound mRNP particles interact with microtubules, but a precise role(s) of YB-1 has not yet been uncovered (11, 12). Thus, we tried to examine whether YB-1 regulates the fate of vRNP.

Fig 1.

Identification of YB-1 as a novel vRNP-interacting proteins. (A) Identification of cellular and viral proteins interacting with vRNP complexes. HeLa cells infected with influenza virus at an MOI of 10 were subjected to immunoprecipitation assays with control IgG (lane 2) or anti-NP (lane 3) antibody-conjugated protein A-Sepharose. The coprecipitated proteins were eluted in 100 mM glycine (pH 2.8), separated through 10% SDS-PAGE, and visualized by silver staining. Molecular mass markers are also shown in lane 1. (B) Intracellular localization of YB-1 in infected cells. Infected MDCK cells were subjected to indirect immunofluorescence assays with anti-YB-1 antibody followed by FISH assays using a probe that hybridizes with segment 1 vRNA at 0, 4, 8, and 12 hpi. For LMB treatment, infected cells were incubated in culture medium containing 20 nM LMB at 7 hpi, and then the intracellular localization of vRNA and YB-1 was visualized by FISH and indirect immunofluorescence assays at 12 hpi. The result for SeV-infected cells at 12 hpi is also shown. Scale bars, 10 μm. (C) Intracellular localization of cellular proteins related to P-bodies and SGs. Mock-infected cells (left panels) or infected MDCK cells at 8 hpi (right panels) were subjected to indirect immunofluorescence assays with anti-RAP55 (upper panels, red), anti-RCK (middle panels, red), or anti-TIAR (lower panels, red) antibodies. Nuclear DNA was stained with TO-PRO-3 iodide (blue). Scale bars, 10 μm.

Table 1.

LC-MS analysis of vRNP-interacting proteins

Protein	No. of observed peptides	Mascot score	Sequence coverage (%)
Heterogenous nuclear ribonucleoprotein A1	6	85	9
Heterogeneous nuclear ribonucleoproteins C1/C2	11	64	21
Heterogeneous nuclear ribonucleoprotein D0	4	77	12
Heterogeneous nuclear ribonucleoprotein H	9	159	17
Heterogeneous nuclear ribonucleoprotein H2	8	111	17
Heterogeneous nuclear ribonucleoprotein L	7	98	17
ATP-dependent RNA helicase DDX3Y	15	80	17
ATP-dependent RNA helicase DDX5	7	81	10
ATP-dependent RNA helicase DDX17	10	87	12
Nucleolin	8	85	13
78-kDa glucose-regulated protein	11	157	15
Importin subunit alpha 7	3	65	8
Y-box-binding protein 1	3	91	10
Tubulin alpha-1C chain	5	70	12
Spliceosome RNA helicase UAP56	5	62	10
Heat shock protein 90 alpha	9	111	15
Heat shock protein 90 beta	9	73	14
Heat shock cognate 71-kDa protein	11	134	15
Heat shock 70-kDa protein 1	12	120	14

YB-1 localizes predominantly in the cytoplasm but translocates to the nucleus to regulate transcription in response to environmental stimuli such as DNA-damaging agents, UV irradiation, hyperthermia, and serum stimulation (34). More importantly, YB-1 is one of the components of SGs and P-bodies, which are cytoplasmic compartments possibly involved in the regulation of translation under stress conditions (78). Viral infection also gives a stressful environment to cells. To elucidate the biological significance of the interaction of vRNP with YB-1, we examined the intracellular localization of YB-1 in influenza virus-infected cells by indirect immunofluorescence assays using anti-YB-1 antibody (Fig. 1B). The vRNA was also counterstained by the fluorescence in situ hybridization (FISH) method (28). YB-1 localized at the cytoplasm in infected cells at 4 hpi, as it did in mock-infected cells. Along with the progression of infection, YB-1 was imported to the nucleus and accumulated in unknown nuclear speckles at 8 hpi. At 12 hpi, a portion of YB-1 was found in unknown cytoplasmic granules with the exported progeny vRNA. Since these cytoplasmic granules of YB-1 were not found in the presence of LMB, a potent inhibitor of CRM1, it is likely that YB-1 is exported with vRNA from the nucleus (Fig. 1B). Furthermore, the translocation of YB-1 upon influenza virus infection may not be involved in the innate immunity, since the localization of YB-1 was not changed in Sendai virus-infected cells (Fig. 1B). Next, we examined the intracellular localization of RAP55, RCK/p54, and TIAR proteins, which are components of cellular mRNP and accumulate in SGs and P-bodies (2). In contrast to the case for YB-1, we did not find any localization changes of these proteins in response to infection (Fig. 1C). Thus, the translocation of YB-1 found in infected cells may not be involved in a function as a component of SGs and P-bodies.

Interaction of YB-1 with vRNP-exporting complexes.

Progeny vRNP is exported to the cytoplasm from the nucleus through the CRM1-dependent pathway by assembling export complexes with viral proteins M1 and NS2 (19, 52, 54, 77). Since YB-1 may be exported to the cytoplasm together with vRNP, it is assumed that YB-1 interacts with vRNP export complexes in the nucleus. To address this, we examined the colocalization of YB-1 with M1 and NS2. We found that M1 but not NS2 accumulated in the nuclear speckles with YB-1 at 8 hpi (Fig. 2A). Since a major portion of the progeny vRNA and NS2 was localized in the cytoplasm (Fig. 1B and 2A), the newly synthesized vRNP may be exported to the cytoplasm immediately after formation of the export complex. Thus, we examined the localization of vRNA and NS2 in LMB-treated cells. Upon inhibition of the CRM1-dependent export pathway, vRNA accumulated in the nuclear speckles with YB-1, M1, and NS2 (Fig. 2B). Previously, it was reported that a small portion of M1 and NS2 may localize in PML NBs (62, 64). It has been shown that the overexpression of PML suppresses influenza virus proliferation (10, 27), suggesting that PML NBs may contribute to the cellular antiviral response. However, the role of PML NBs remains controversial, since influenza virus replicates normally in cells lacking the PML gene (20). As expected, in the presence of LMB, we found that vRNA and YB-1 were partially associated with PML NBs (Fig. 2C). Taking these findings together, it is possible that the YB-1 imported into the nucleus interacts with the vRNP export complexes in PML NBs and then YB-1 is subsequently exported from the nucleus with vRNP.

Fig 2.

Colocalization of YB-1 and vRNP export complexes in nuclear speckles. (A) Intracellular localization of YB-1, M1, and NS2. At 8 hpi, infected MDCK cells were subjected to indirect immunofluorescence assays with rabbit anti-YB-1 (red) and either mouse anti-M1 (upper panel, green) or rat anti-NS2 (lower panel, green) antibody. Scale bar, 10 μm. (B) Intracellular localization of vRNA, YB-1, M1, and NS2 in the presence of LMB. At 7 hpi, infected MDCK cells were incubated for 1 h in the presence of 20 nM LMB. Segment 1 vRNA (left panels, green), YB-1 (upper panel, red), M1 (middle panel, red), and NS2 (lower panel, red) were visualized by FISH and indirect immunofluorescence assays. Scale bars, 10 μm. (C) Accumulation of vRNA and YB-1 in PML NBs in the presence of LMB. After treatment of LMB as described for panel B, mock-infected (upper panel) and infected (middle and lower panels) MDCK cells were subjected to FISH assays using the probe that hybridizes with segment 1 vRNA (upper and middle panels, green) and to indirect immunofluorescence assays with rabbit (upper and middle panels, red) and mouse (lower panel, red) anti-PML and rabbit anti-YB-1 antibodies (lower panel, green). Scale bars, 10 μm.

Influenza virus produces three different RNAs, i.e., vRNA, cRNA, and viral mRNA. Both vRNA and cRNA form ribonucleoprotein complexes with the viral polymerase and NP, whereas vRNA, but not cRNA, is found in the cytoplasm (26) and packaged into the virions. In contrast, the viral mRNA interacts with cellular mRNA-binding proteins and is exported through the REF/Aly pathway generally used by cellular mRNAs (4, 57). To examine the specific interaction of YB-1 with viral RNAs, we visualized positive-sense RNAs (cRNA and viral mRNA) by FISH assays. The intracellular localization of positive-sense RNAs was not changed by LMB treatment, as previously reported (57), and the FISH signals were not colocalized with YB-1 in the absence or presence of LMB (Fig. 3A). Further, to show quantitative results, we performed immunoprecipitation assays with cell lysates prepared from cells constitutively expressing FLAG-YB-1 using anti-FLAG antibody as described in Materials and Methods. We found that vRNA interacted with YB-1, but neither cRNA nor viral mRNA did in infected cells (Fig. 3B). Thus, it is quite likely that YB-1 is involved in the functional regulation of vRNP but not in that of either cRNP or viral mRNP.

Fig 3.

Specific interaction of YB-1 with vRNA but not with either cRNA or viral mRNA. (A) Intracellular localization of YB-1 and viral mRNA/cRNA. At 8 hpi, infected MDCK cells were subjected to FISH assays using a probe that hybridizes with segment 1 cRNA and mRNA (green) and to indirect immunofluorescence assays with anti-YB-1 antibody (red) with or without 20 nM LMB treatment for 1 h. Nuclear DNA was stained with TO-PRO-3 iodide (blue). Scale bars, 10 μm. (B) Coimmunoprecipitation of YB-1 and viral RNA molecules. HeLa cells constitutively expressing FLAG-YB-1 were infected with influenza virus at an MOI of 10. After 8 hpi, cell lysates were prepared and subjected to immunoprecipitation assays in the presence of either control IgG or anti-FLAG antibody as described in Materials and Methods. The immunoprecipitated viral RNAs were eluted with 100 μg/ml FLAG peptide and then quantitatively analyzed by reverse transcription followed by real-time PCR with primers specific for segment 5 vRNA, cRNA, and NP mRNA. To quantitatively evaluate the data, 5% equivalents of mock-infected and infected samples were also observed.

Interaction of the YB-1-vRNP complex with Rab11a-positive recycling endosomes along microtubules.

At 12 hpi, YB-1 was partially colocalized with the vRNA exported from the nucleus in the cytoplasm (Fig. 1B). Recent studies have suggested that the progeny vRNP is transported to the plasma membrane along microtubules via Rab11a-positive recycling endosomes (1, 17, 47). Further, it has been reported that YB-1 interacts with microtubules (12). Based on these findings, we examined whether YB-1 accumulates in microtubules with vRNP (Fig. 4). We found that the cytoplasmic punctate signals of YB-1 in perinuclear regions were colocalized with α-tubulin (Fig. 4A). We then examined the interaction of YB-1 with microtubules, Rab11a, and vRNP by immunoprecipitation assays. YB-1 interacted with α-tubulin but hardly with Rab11a in mock-infected cells (Fig. 4B, lane 3). In contrast, the interaction of YB-1 with Rab11a was significantly increased in infected cells (Fig. 4B, lane 6). To address whether the vRNP-YB-1 complexes accumulate in microtubules with Rab11a, the proteins coprecipitated with YB-1 from infected lysates (Fig. 4B, lane 6) were eluted and then subjected to reimmunoprecipitation with anti-NP antibody (Fig. 4C). Since α-tubulin and Rab11a were immunoprecipitated with anti-NP antibody (Fig. 4C, lane 3), it is quite likely that the progeny vRNP is accumulated on Rab11a-positive recycling endosomes with YB-1 along microtubules.

Fig 4.

Accumulation of YB-1-vRNP complexes on microtubules with Rab11a-positive recycling endosomes in response to infection. (A) Colocalization of YB-1 and microtubules in cytoplasmic punctate signals. At 12 hpi, mock-infected (upper panels) and infected (lower panels) MDCK cells were subjected to indirect immunofluorescence assays with anti-YB-1 (red) and α-tubulin (green) antibodies. Scale bars, 10 μm. (B and C) Interaction of YB-1 with Rab11a-positive recycling endosomes on microtubules. HeLa cells constitutively expressing FLAG-YB-1 were infected with influenza virus at an MOI of 10. (B) Cell lysates were prepared and subjected to immunoprecipitation assays in the presence of either control IgG (lanes 2 and 5) or anti-FLAG antibody (lanes 3 and 6) at 12 hpi. Coprecipitated proteins were eluted with 100 μg/ml FLAG peptide and detected by Western blotting assays with anti-PB1, anti-Rab11a, anti-α-tubulin, and anti-FLAG antibodies. Ten percent equivalents of mock-infected (lane 1) and infected (lane 4) lysates were also subjected to Western blotting assays. (C) The eluate purified from infected lysate from panel B, lane 6, was reimmunoprecipitated with either control IgG (lane 2) or anti-NP antibody (lane 3), and then the eluate was subjected to Western blotting assays with anti-α-tubulin and anti-Rab11a antibodies. Lane 1, 30% equivalent of proteins immunopurified with anti-FLAG antibody from infected cell lysate.

YB-1 is a positive factor for Rab11a-dependent virus production.

It has been shown that Rab11a is required for the transport of vRNP to the apical plasma membrane and thereby affects the production of progeny virions (1, 17, 47). Since YB-1 binds to Rab11a together with vRNP (Fig. 4C), it is assumed that YB-1 is involved in production of progeny virions through the Rab11a-positive recycling endosome pathway. To address this, the effect of YB-1 overexpression on the virus titer was examined using siRNA-mediated Rab11a knockdown (KD) cells (Fig. 5). The transfection efficiency was approximately 60%. The amount of exogenously overexpressed FLAG-YB-1 was 3-fold higher than that of endogenous YB-1, and the expression level of Rab11a in KD cells decreased to approximately 30% of that in control cells transfected with the nontargeting siRNA (Fig. 5A). The expression level of viral proteins was found to be virtually unchanged by YB-1 overexpression (Fig. 5B). We found that the amount of infectious virions produced from cells overexpressing YB-1 was significantly increased compared to that produced from cells transfected with empty plasmid, whereas the virus titer was slightly enhanced by YB-1 overexpression in Rab11a KD cells (Fig. 5C). Further, the slopes of the lines in Fig. 5C were determined to compare the efficiencies of virus production (Fig. 5D). This result shows that the production of infectious viruses was increased 4.8-fold by the YB-1 overexpression, but the stimulatory activity of YB-1 was reduced 1.9-fold by Rab11a KD (Fig. 5D). Therefore, it is concluded that YB-1 stimulates the production of infectious viruses in an Rab11a-dependent manner. We also tried to measure the amount of infectious virions produced from YB-1 KD cells. However, YB-1 KD cells tend to die from influenza virus infection after 16 to 20 hpi (data not shown). Thus, it was difficult to demonstrate the effect of YB-1 on the virus titer using YB-1 siRNA.

Fig 5.

Effect of YB-1 overexpression on the production of infectious virions. (A) Expression levels of YB-1 and Rab11a proteins. 293T cells were transfected with either pCAGGS empty plasmid (lanes 1 and 3) or pCAGGS-FLAG-YB-1 (lanes 2 and 4) at 24 h after treatment of either nontargeting (control; lanes 1 and 2) or Rab11a (siRab11a; lanes 3 and 4) siRNA. At 24 h after transfection of expression vectors, the cell lysates were prepared and analyzed by SDS-PAGE followed by Western blotting assays with anti-YB-1, anti-Rab11a, and anti-β-actin antibodies. (B) Accumulation levels of viral proteins in cells overexpressing FLAG-YB-1. 293T cells were transfected with either pCAGGS or pCAGGS-FLAG-YB-1. At 24 h posttransfection, cells were infected with influenza virus at an MOI of 10. At 0, 2, 5, and 8 hpi, cell lysates were prepared and analyzed by Western blotting assays with anti-PB1, anti-NP, anti-M1, anti-NS2, and anti-α-tubulin antibodies. (C) Production of infectious virions. Control (open diamonds and filled squares) and Rab11a KD 293T (open triangles and open squares) cells transfected with either pCAGGS (open diamonds and open triangles) or pCAGGS-FLAG-YB-1 (filled and open squares) as described for panel A were infected with influenza virus at an MOI of 0.5. The culture supernatants collected at 3, 6, 12, 16, 20, and 24 hpi were subjected to plaque assays to examine the production of infectious virions in a single-round infection. The average titers and standard deviations determined from three independent experiments are shown. (D) Stimulatory activity of YB-1 on the virus titer in Rab11a KD cells. The slopes of the lines in panel C were determined by the least-squares method, and the ratio of the virus titer from cells overexpressing FLAG-YB-1 to that from cells transfected with pCAGGS is shown.

YB-1 functions as a porter for vRNP to direct it to microtubules.

Figures 4 and 5 suggested that YB-1 functions in the vRNP transport through the Rab11a-positive recycling endosome pathway along microtubules. Next, we tried to demonstrate whether YB-1 functions as a transporter of vRNP to microtubules using siRNA-mediated gene silencing (Fig. 6). At 48 h after transfection of YB-1 siRNA, the expression level of YB-1 in KD cells decreased to 25% of that in control cells (Fig. 6A). There were no differences found in the accumulation levels of viral proteins (Fig. 6B) and vRNA and viral mRNA (Fig. 6C) between control and YB-1 KD cells. Previous reports showed that the vRNP complexes exported from the nucleus accumulate at the MTOC around the perinucleus (45) with Rab11a-positive recycling endosomes (1, 17, 47). We carried out FISH assays to examine whether vRNP complexes localize around the perinucleus in YB-1 KD cells at 8 hpi (Fig. 6D and E). A major population of YB-1 KD cells (81.5% ± 7.2%) had vRNA in a diffusive pattern. Thus, it is quite likely that YB-1 stimulates the accumulation of vRNP at the MTOC. As shown in Fig. 2C, the replicated vRNA associated with PML NBs in the presence of LMB. Since vRNA accumulated in PML NBs even in YB-1 KD cells with LMB treatment (Fig. 6D), it is strongly suggested that YB-1 is not involved in the association between vRNP and PML NBs. It is reported that exogenously expressed M1 is localized in PML NBs (64), suggesting that vRNP might accumulate in PML NBs through the interaction between vRNP and M1. It is also suggested that YB-1 does not play a role in the vRNP export from the nucleus to the cytoplasm, since vRNA was found predominantly in the cytoplasm of YB-1 KD cells (Fig. 6E).

Fig 6.

Accumulation of vRNP on microtubules in YB-1 knockdown cells. (A) Expression level of YB-1 in YB-1 KD cells. HeLa cells transfected with nontargeting (control; lanes 1 to 4) or YB-1 (siYB-1; lanes 5 to 8) siRNA were lysed, and then the lysates (2.5 × 10³, 5 × 10³, 1 × 10⁴, and 2 × 10⁴ cells) were subjected to SDS-PAGE followed by Western blotting assays with anti-YB-1 antibody at 48 h posttransfection. (B and C) Accumulation levels of viral proteins and RNAs in YB-1 KD cells. At 48 posttransfection of siRNA, control and YB-1 KD cells were infected with influenza virus at an MOI of 10. At 0, 2, 5, and 8 hpi, cell lysates were prepared and analyzed by Western blotting assays with anti-PB1, anti-NP, anti-M1, anti-NS2, and anti-α-tubulin antibodies. Total RNAs purified from the cells at 0, 2, 5, and 8 hpi were subjected to reverse transcription followed by quantitative real-time PCR with primers specific for segment 5 vRNA and NP mRNA as described in Materials and Methods. (D and E) Intracellular localization of vRNA in YB-1 KD cells. At 8 hpi with or without LMB treatment for 1 h, infected control and YB-1 KD cells were subjected to FISH assays using a probe that hybridizes with segment 1 vRNA (panel D) (scale bars, 10 μm). Cells were counted, and the localization pattern of vRNA in the absence of LMB was determined (E). The number of cells showing each localization pattern was expressed as the percentage of the total cell number (n = 80) in panel E. The average percentages determined from three independent experiments are shown.

YB-1 interacts with vRNP (Fig. 1A and 4B), but it is unclear whether YB-1 interacts directly with one or more vRNP components, that is, viral polymerase complexes, NP, and vRNA. To test this, we performed pulldown assays with purified vRNP and His-YB-1 using Ni-nitrilotriacetic acid (NTA) resin (Fig. 7A). Not only viral polymerase subunits but also NP was coprecipitated with YB-1 (lane 6), demonstrating that YB-1 interacts directly with vRNP. Since YB-1 has a single-stranded RNA binding activity, it is possible that YB-1 binds to vRNA. To address this, vRNP treated with micrococcal nuclease (mnRNP) to deplete vRNA was also subjected to the pulldown assay with His-YB-1. We found that each viral polymerase subunit but not NP from mnRNP was coprecipitated with YB-1 (Fig. 7A, lane 9), suggesting that YB-1 interacts with viral polymerase complexes. YB-1 consists of three domains: the N-terminal domain, the cold shock domain (CSD), and the C-terminal tail domain. The CSD has the well-characterized RNA-binding motifs RNP-1 and RNP-2 and thereby functions as a nucleic acid-binding domain. The N-terminal domain is rich in alanine and proline (A/P domain), and the C-terminal domain contains alternating clusters of positively and negatively charged amino acid residues (A/B domain). All three domains are involved in interaction with a number of cellular proteins (reviewed in reference 18). To further characterize the interaction between YB-1 and vRNP, we carried out pulldown assays with GST-fused deletion mutants of YB-1 and vRNP (Fig, 7B). We found that PB1 is coprecipitated with the mutants harboring the A/B domain (lanes 5 and 6; ΔA/P and A/B), suggesting that YB-1 interacts with vRNP through the A/B domain. Next, we tried to demonstrate whether YB-1 recruits vRNP on microtubules in vitro. Reconstituted microtubules were incubated with either vRNP or YB-1-vRNP complex and then subjected to immunoprecipitation using anti-NP antibody (Fig. 7C). We found that microtubules were hardly coprecipitated with vRNP in the absence of YB-1 (lane 4). In sharp contrast, in the presence of YB-1, the interaction of vRNP with microtubules was increased by approximately 3-fold (lane 5). Finally, we carried out immunoprecipitation assays with cell lysates prepared from cells constitutively expressing FLAG-α-tubulin with or without YB-1 KD (Fig. 7D). We found that PB1 was immunoprecipitated with FLAG-α-tubulin from control lysates (lane 3) but slightly from YB-1 KD lysates (lane 6). Therefore, it could be concluded that YB-1 is required for the accumulation of vRNP on microtubules. Taking these findings together, we propose that YB-1 interacts directly with vRNP and functions as a porter that facilitates the binding of vRNP with microtubules, where vRNP is led to Rab11a-positivie recycling endosomes.

Fig 7.

YB-1 functions as a porter bringing the progeny vRNP to microtubules. (A) Direct interaction of YB-1 with viral polymerase complex. vRNP (lanes 4 to 6) or micrococcal nuclease-treated vRNP (mnRNP) (lanes 7 to 9) was incubated in the absence (lanes 2, 5, and 8) or presence (lanes 3, 6, and 9) of purified recombinant His-YB-1 protein at 30°C for 1 h. Complexes were purified using Ni-NTA resin, and then proteins were separated through SDS-PAGE and detected by Coomassie brilliant blue (CBB) staining and Western blotting assays with anti-PB1, anti-PB2, and anti-PA antibodies. Lanes 1, 4, and 7 represent 20% of input amounts. (B) Interaction of vRNP with deletion mutants of YB-1. Each GST-fused deletion mutant of YB-1 was incubated with vRNP at 30°C for 1 h. Complexes were purified using glutathione-Sepharose resins, and then proteins were separated through SDS-PAGE and detected by Western blotting assays using anti-PB1 antibody. Lane 1 represents 20% of input amounts. A schematic diagram of the deletion mutants of YB-1 is at the bottom. (C) Interaction between vRNP and microtubules mediated by YB-1. Reconstituted microtubules were incubated with either vRNP (lanes 2 and 4) or YB-1-vRNP complex (lanes 3 and 5), and then complexes were immunoprecipitated with either a nonspecific IgG (control; lanes 2 and 3) or anti-NP antibody (lanes 4 and 5). The immunoprecipitated proteins were separated through SDS-PAGE and subjected to Western blotting assays with anti-α-tubulin and anti-NP antibodies. Lane 1 represents 10% of input amount. (D) Interaction of vRNP with microtubules in YB-1 KD cells. HeLa cells constitutively expressing FLAG-α-tubulin were infected with influenza virus at an MOI of 10. At 8 hpi, cell lysates were prepared and subjected to immunoprecipitation assays with either control IgG or anti-FLAG antibody. The immunoprecipitated proteins eluted with 100 μg/ml FLAG peptide were separated through SDS-PAGE and then visualized by Western blotting assays using anti-PB1 and anti-α-tubulin antibodies. Five percent equivalents of control (lane 1) and YB-1 KD (lane 4) lysates are also shown.

DISCUSSION

We have identified YB-1, a cellular DNA/RNA-binding protein, as a vRNP-interacting protein (Fig. 1A). YB-1 was found to be relocalized to the nucleus from the cytoplasm and associated with PML NBs along with the progression of virus infection (Fig. 1B). Previous reports showed that the nuclear import of YB-1 requires phosphorylation by protein kinase C (PKC) (35) and Jak1 (15) and Akt (69) kinases. The other mechanism for nuclear import of YB-1 is thought to be triggered by a proteolytic cleavage by 20S proteasomes to separate the C-terminal fragment (aa 220 to 324) containing a cytoplasmic retention signal (68). When we performed immunofluorescence assays with an antibody recognizing either the N terminus (aa 1 to 13) (used in this study) or the C terminus (aa 307 to 324) (purchased from Sigma), the YB-1 nuclear import was found only when the antibody recognizing the N terminus was used (data not shown). However, we found that vRNP complexes interact with full-length YB-1 (Fig. 1A), suggesting that the 20S proteasome-mediated processing of YB-1 may not always be required in influenza virus-infected cells. Therefore, it is possible that the recognition specificity of our antibody is due to an unknown conformational change of YB-1 possibly induced by the phosphorylation and/or interaction with cellular and/or viral proteins. Identification of a signaling pathway and interacting proteins responsible for the nuclear import of YB-1 upon influenza virus infection is needed.

PML NBs are highly dynamic structures that are disrupted or changed in their morphology in response to environmental stimuli (38). The genomes of several DNA viruses have been shown to be localized and transcribed in the vicinity of PML NBs (13, 23, 24). In most cases, virus-encoded regulatory proteins localize at PML NBs and disrupt PML NBs for successful infection, suggesting a negative role for PML NBs in virus growth (13, 23, 24). In addition to the putative function of PML NBs during virus infection, it is postulated that PML NBs have functions in cellular gene transcription, tumor suppression, proteasomal degradation, cellular senescence, apoptosis, and DNA repair (14, 38). It is proposed that PML NBs might modulate chromatin architecture and transcription, since nascent RNA and several gene loci are found around PML NBs (5, 13, 65). Given this, it seems likely that viruses hijack this nuclear structure to set up a platform suitable for virus replication, although the functional relevance of the interaction between influenza virus and PML NBs is an open question. M1 and NS2 were shown to be associated with PML NBs (62, 64), although the functional significance of the colocalization remains to be determined. We found that the YB-1 imported into the nucleus accumulates in PML NBs, in which vRNA was present with M1 and NS2 when the vRNP export was inhibited (Fig. 2). Therefore, we speculate that the progeny vRNPs could be assembled into the export complexes at PML NBs and subsequently interact with YB-1. Additional experiments are needed to clarify a precise role(s) of PML NBs in the vRNP nuclear export as well as the mechanism for targeting YB-1 to PML NBs.

The accumulation of recycling endosome vesicles around the MTOC is disrupted in the presence of microtubule depolymerization reagents such as nocodazole (8). Thus, it is likely that the intact microtubule functions as a platform for the recycling endosome vesicle. We observed a direct binding of YB-1 with vRNP for the recruitment of vRNP to microtubules (Fig. 7). However, only a small portion of vRNP was colocalized with YB-1 in infected cells, suggesting that YB-1 transiently interacts with vRNP (Fig. 1). Further, we could not find YB-1 in the purified influenza virions (data not shown), suggesting that YB-1 might be dissociated from vRNP prior to packaging of vRNP into progeny virions. The association between YB-1 and enough microtubules is proposed to compete with the interaction of YB-1 with cellular mRNP in vitro (11). Thus, we speculate that vRNP-YB-1 complexes recruited on microtubules may be disassembled by the microtubule formation, and thereby vRNP could be loaded onto Rab11a-positive recycling endosomes bound to microtubules.

The Rab11a-mediated vesicular transport may be functionally important after arrival at the plasma membrane. For example, lipid rafts are required for the budding of influenza virus from the apical plasma membrane (63, 70), and cholesterol, an essential component of the lipid rafts, is enriched in the recycling endosome membrane (48). Interestingly, it has been shown that influenza virus particles are hardly pinched off from the plasma membrane in Rab11a KD cells (7). Thus, it is possible that the Rab11a-positive recycling endosome has an important role in the apical transport of proteins and/or membranes which are involved in budding process, including membrane scission (60). However, it is also possible that the trafficking of the viral genome is required to allow the efficient virus budding. The influenza virus genome consists of eight-segmented vRNA molecules. Since it is believed that the eight individual segments are packaged into a progeny virion (53), a hierarchical incorporation of each vRNA should be required. It is hypothesized that Rab11a-positive transport vesicles might be an assembly center of the eight segments of vRNA (17, 47). To further understand the mechanism of influenza virus egress, the dynamics of the recycling endocytic compartment are to be analyzed.