Heat Shock Protein 70 Modulates Influenza A Virus Polymerase Activity

Background: It has been shown that heat shock protein 70 (Hsp70) plays a role in influenza A virus replication.

Results: A correlation between viral replication/transcription activities and nuclear/cytoplasmic shuttling of Hsp70 was observed.

Conclusion: Hsp70 modulates the influenza A virus polymerase activity.

Significance: This study, for the first time, suggests that Hsp70 may actually assist in influenza A virus replication.

Abstract

The role of heat shock protein 70 (Hsp70) in virus replication has been discussed for many viruses. The known suppressive role of Hsp70 in influenza virus replication is based on studies conducted in cells with various Hsp70 expression levels. In this study, we determined the role of Hsp70 in influenza virus replication in HeLa and HEK293T cells, which express Hsp70 constitutively. Co-immunoprecipitation and immunofluorescence studies revealed that Hsp70 interacted with PB2 or PB1 monomers and PB2/PB1 heterodimer but not with the PB1/PA heterodimer or PB2/PB1/PA heterotrimer and translocated into the nucleus with PB2 monomers or PB2/PB1 heterodimers. Knocking down Hsp70 resulted in reduced virus transcription and replication activities. Reporter gene assay, immunofluorescence assay, and Western blot analysis of nuclear and cytoplasmic fractions from infected cells demonstrated that the increase in viral polymerase activity during the heat shock phase was accompanied with an increase in Hsp70 and viral polymerases levels in the nuclei, where influenza virus replication takes place, whereas a reduction in viral polymerase activity was accompanied with an increase in cytoplasmic relocation of Hsp70 along with viral polymerases. Moreover, significantly higher levels of viral genomic RNA (vRNA) were observed during the heat shock phase than during the recovery phase. Overall, for the first time, these findings suggest that Hsp70 may act as a chaperone for influenza virus polymerase, and the modulatory effect of Hsp70 appears to be a sequel of shuttling of Hsp70 between nuclear and cytoplasmic compartments.

Introduction

Influenza A viruses are pleomorphic, enveloped RNA viruses belonging to the family Orthomyxoviridae. The genome of influenza A viruses consists of eight single-stranded RNA segments of negative polarity having partially complementary ends that form a closed structure. In a virus particle, each viral genomic RNA (vRNA)³ segment exists in association with multiple monomers of nucleoprotein (NP) and a single copy of a polymerase heterotrimer consisting of a polymerase basic protein 1 (PB1), polymerase basic protein 2 (PB2), and polymerase acidic protein (PA), thereby constituting a viral ribonucleoprotein complex (vRNP). These vRNPs are independent functional units capable of transcription and replication (1, 2). The PB1, which has motifs for binding to the vRNA and cRNA promoters, is the core subunit for RNA synthesis and responsible for the assembly of PB2 and PA into a multifunctional enzyme complex (3–5). The PB2 subunit is responsible for cap snatching from the host pre-mRNAs and initiation of transcription by providing these 5′-capped RNA fragments to copy the template (6–8). Multiple functions have been proposed for the PA subunit such as protease activity, endonuclease activity, promoter binding activity, and assembly of polymerase subunits into a functional polymerase complex (9–13).

Upon infection, virus enters the cell by receptor-mediated endocytosis; the vRNP complex is released into the cytoplasm and thence transported into the nucleus. Because virus transcription and replication take place in the nucleus and viral protein synthesis and virus assembly take place in the cytoplasm, the viral proteins have to be shuttled between the nucleus and cytoplasm. This entire cascade of events requires interaction of viral proteins with host cellular factors. Some of these host factors participate in the virus replication and transcription. Ran-binding protein 5 was reported to interact with PB1 alone or with PB1/PA heterodimers and transfer them into the nucleus (14). α-Importins have been reported to interact with influenza virus PB2 and NP in host-dependent manner (15, 16). Interaction of heat shock protein 90 (Hsp90) with PB2 has been reported to stimulate the viral polymerase activity (17). Hsp90 is also known to participate in assembly and nuclear transport of viral polymerase proteins by binding with PB2 monomers or PB2/PB1 heterodimers (18). Recently, host chaperonin CCT has been proposed to act as a molecular chaperone for PB2 protein by assisting its folding and incorporation into the heterotrimeric polymerase complex (19).

The heat shock protein 70 family comprises of highly related, stress-inducible or constitutively expressed, cytosolic, or compartment-specific, 66–78-kDa isoforms (20, 21). Of these, heat shock protein 70 (Hsp70) and heat shock cognate 70 protein (Hsc70) are two major cytoplasmic isoforms. All heat shock proteins share a highly conserved domain structure comprising (i) an N-terminal ATPase domain, (ii) a middle region with protease sensitive site, (iii) C-terminal substrate binding domain, and (iv) a G/P-rich C-terminal region containing an EEVD motif enabling it to bind co-chaperones and other heat shock proteins. Many housekeeping roles have been attributed to Hsp70 such as protein folding and prevention of aggregation, refolding of misfolded and aggregated proteins, signal transduction by controlling the conformation of proteins, anti-apoptotic effect, in tumorigenesis indicated by higher expression levels in many tumor cells, and membrane translocation of organellar and secretory proteins (20–25).

The role of Hsp70 in virus replication is quite broad, because it inhibits the replication of some viruses (26–28) and helps the replication of other viruses (29–33). Studies conducted to elucidate the role of Hsp70 in influenza virus replication have suggested that Hsp70 inhibits influenza virus replication either by preventing the nuclear export of the RNP complex (34) or by disrupting the binding of viral polymerase with viral RNA (35).

Generally, in nontransformed cells under normal conditions, Hsc70 is expressed abundantly, whereas Hsp70 is expressed at relatively low levels, but its expression increases considerably under various types of stress. Interestingly, elevated levels of Hsp70 have been reported in many tumors and transformed cell lines (36–38). Theodorakis and Morimoto (39) reported that HeLa and 293 cells constitutively expressed high levels of Hsp70. Many host factors interacting with influenza virus proteins have been identified in HEK293T and HeLa cells (16, 18, 40, 41). If Hsp70 had an inhibitory effect on influenza virus replication, it would have made HEK293T or HeLa cells unsuitable for influenza virus study rather than facilitating their large scale application. Keeping this assumption in mind, we decided to study the role of Hsp70 in cells constitutively expressing it.

In this study we found that Hsp70 interacted and translocated into the nucleus with PB2 monomers or PB2/PB1 heterodimers, presumably assisting the assembly of the viral polymerase complex. Knockdown of Hsp70 resulted in significant reductions in both virus transcription and replication. Moreover, an increase in viral polymerase activity was observed during the heat shock phase compared with the recovery phase, coinciding with the subcellular movement of Hsp70, suggesting that Hsp70 acted as chaperone for the viral polymerase complex.

EXPERIMENTAL PROCEDURES

Cells, Viruses, and Reagents

Human embryonic kidney 293T cells (HEK293T) and HeLa cells were grown in DMEM supplemented with 10% fetal calf serum and antibiotics at 37 °C in a 5% CO₂ atmosphere. Influenza virus strain A/Puerto Rico/8/1934 (H1N1) (PR8) was obtained from the virus repository of our laboratory.

Primary mAbs used in Western blot analysis and immunofluorescence assays included anti-PB2 (143/3), anti-PB1 (81/2) and anti-PA (58/1 and 65/4) (42–44), anti-FLAG® M5 (F4042; Sigma), anti-HA tag (12CA5, ab16918), anti-Hsp70 (C92F3A-5), anti-α-tubulin (DM1A, ab7291), anti-Hsp90 (16F1, ab13494), anti-β-actin (ab6276), and polyclonal anti-Lamin B1 (ab16048). All primary antibodies used were of mouse origin except anti-Lamin B1 and anti-Hsp90, which were rabbit polyclonal and rat monoclonal antibodies, respectively. HRP-conjugated anti-mouse, anti-rabbit, and anti-rat mAbs were used as secondary antibodies (Jackson ImmunoResearch) in Western blot analysis, whereas Alexa Fluor 488® goat anti-mouse IgG and Alexa Fluor 405® goat anti-mouse IgG antibodies (Invitrogen) were employed in immunofluorescence assays.

Prostaglandin A1 (PGA1) was purchased from Cayman Chemicals, and TNFα and cycloheximide (CHX) were purchased from Wako, Japan.

Construction of Plasmids

The plasmids containing influenza virus polymerase and NP genes of A/Hong Kong/483/1997 (H5N1) (HK483) (pCAGGS-HK483PB2, -PB1, -PA, and -NP), PR8 (pCAGGS-PR8-PB2, -PB1, -PA, and -NP) and A/Aichi/2/1968 (H3N2) (Aichi) (pHH21-AichiPB2, -PB1, -PA, and -NP) were kindly provided by Dr. Yoshihiro Kawaoka (Institute of Medical Science, University of Tokyo, Tokyo, Japan). For constructing plasmids expressing full-length PB2 and PB1 proteins of the HK483 influenza virus with N-terminal FLAG tags (FLAG-HK483PB2 and FLAG-HK483PB1), PCR products were amplified from pCAGGS-HK483PB2 and pCAGGS-HK483PB1 as templates using following primer pairs: FLAG-PB2-F/SacI (5′-TATTGAGCTCATGGATTACAAGGATGACGACGATAAGGGCGGCATGGAAAGAATAAAAGAACTACG-3′) and PB2-R/XhoI (5′-ATATCTCGAGTCACTAATTGATGGCCATCCGAATTC-3′) for FLAG-HK483PB2 and FLAG-PB1-F/Acc65I (5′-TATTGGTACCATGGATTACAAGGATGACGACGATAAGGGCGGCATGGATGTCAATCCGACTTTAC-3′) and PB1-R/SphI (5′-ATATGCATGCTCACTACTTCCCTTGCCGTCCGAGCTC-3′) for FLAG-HK483PB1. Then amplified FLAG-HK483PB2 and FLAG-HK483PB1 PCR products were digested with SacI/XhoI and Acc65I/SphI restriction enzymes, respectively, and cloned into the respective restriction sites of pCAGGS plasmid. The ORFs of PB2, PB1, and PA genes of Aichi were amplified from pHH21-AichiPB2, -PB1, and -PA as templates using following primer pairs: PB2–1F/KpnI (5′-GGGGTACCATGGAAAGAATAAAAGAACTACG-3′) and PB2-R/XhoI (5′-ATCTCGAGTTAATTGATGGCCATCCGAATTC-3′) for Aichi-PB2, PB1–1F/KpnI (5′-GGGGTACCATGGATGTCAATCCGAC-3′) and PB1-R/XhoI (5′-ATCTCGAGTCACTATTTTTGCCGTCTG-3′) for Aichi-PB1 and PA-1F/KpnI (5′-GGGGTACCATGGAAGATTTTGTAC-3′) and PA-R/XhoI (5′-ATCTCGAGCTATCTTAATGCATGTG-3′) for Aichi-PB1. The amplified PCR products were cloned into the respective restriction sites of pCAGGS plasmid.

For cloning human Hsp70 (NM-005345), total RNA was extracted from HEK293T cells using an RNeasy mini kit (Qiagen) and reverse transcribed using oligo(dT)₂₀ primer to generate cDNAs. These cDNAs were used to amplify the Hsp70 coding region using primer pair hHsp70-F/KpnI (5′-ATAAGGTACCATGGCCAAAGCCGCGGCGATC-3′) and hHsp70-R/XhoI (5′-ATATCTCGAGTCACTAATCTACCTCCTCAATGG-3′). The amplified PCR product was digested with KpnI and XhoI restriction enzymes and cloned into the corresponding restriction sites of the pCAGGS expression plasmid producing a pCAGGS-Hsp70 plasmid. A HA tag was inserted at the N terminus of the Hsp70 coding region by PCR using pCAGGS-Hsp70 as the template and primer pair HA-hHsp70-F/KpnI (5′-GGGGTACCACCATGGACTACCCATACGATGTTCCAGATTACGCTGCCAAAGCCGCGGCGATCGGCATCGAC-3′) and hHsp70-R/XhoI. The amplified PCR product was digested with KpnI and XhoI restriction enzymes and cloned into the respective sites of the pCAGGS expression plasmid yielding pCAGGS-HA-Hsp70.

The reporter plasmid pHW72-Luc2CP was constructed by substituting the open reading frame of enhanced GFP in pHW72-EGFP plasmid (kindly provided by R. Webby at St. Jude Children Research Hospital) with firefly luciferase gene (45). The pCAGGS-Luc2CP plasmid was constructed by amplifying the firefly luciferase (Luc2CP) gene using primer pair Luc2CP-1F/SacI (5′-ATTGAGCTCAATGGAAGATGCCA-3′) and Luc2CP-1830R/KpnI (5′-AATGGTACCCTATTAGACGTTGATCC-3′) and cloning the PCR product into the SacI and KpnI restriction sites of pCAGGS vector.

Co-immunoprecipitation Assays

HEK293T cells, grown in 10-cm tissue culture plates, were transfected with the plasmids indicated in the figures, using TransIT®-LT1 (Mirus). At 48 h post-transfection, cells were washed twice with cold PBS and collected by centrifugation. Cell pellets were resuspended in lysis buffer (50 mm Tris HCl, 280 mm NaCl, 0.5% Triton X-100, 0.2 mm EDTA, 2 mm EGTA, 10% glycerol, and 1 mm DTT; supplemented with a protease inhibitor mixture, Complete Mini EDTA free; Roche Applied Sciences), subjected to sonication, and clarified by centrifugation at 14,000 × g for 10 min. Then anti-FLAG affinity gel (anti-FLAG® M2-agarose gel; Sigma), washed three times with lysis buffer, was incubated with whole cell extracts overnight at 4 °C with gentle rotation. The gel was washed five times with lysis buffer, and the bound proteins were eluted using 3× FLAG peptide (F4799; Sigma) according to the manufacturer's instructions. The interaction between wild type PB2 protein and endogenous Hsp70 was evaluated similarly, except that cell lysate was mixed with anti-PB2 mAb (143/3), and immunoprecipitation was carried out using Protein G-Sepharose^TM 4 Fast Flow (GE Healthcare). The bound proteins were eluted using low pH buffer (0.1 m glycine HCl, pH 3.5). The immunoprecipitated proteins were identified by Western blotting using protein specific antibodies.

Reciprocal immunoprecipitation was carried out by transfecting HEK293T cells with HA-Hsp70 and viral polymerase expression plasmids as indicated in the figures. The immunoprecipitation procedure was the same as that described above except that whole cell extracts were mixed with anti-HA affinity gel (EZview^TM Red anti-HA affinity gel; Sigma) and bound proteins were eluted using influenza HA peptide (I2149; Sigma) according to the manufacturer's instructions.

Cell Fractionation

HEK293T cells were fractionated into cytosolic and nuclear fractions, with little modification, as described by Suzuki et al. (46). Briefly, cells grown in 10-cm tissue culture plates were infected with PR8 influenza virus at a multiplicity of infection (MOI) of 1 or mock infected. After 12 h of incubation, the cells were subjected to the heat shock or allowed to recover for indicated time points (see Fig. 8). The monolayers and resuspended cells were washed twice with ice-cold PBS. Then pelleted cells were resuspended in ice-cold 0.1% Nonidet P-40-PBS and lysed by pipetting up and down several times. A portion of the cell suspension was kept as whole cell lysate. The cell lysates were centrifuged at 14,000 × g for 1 min, and the supernatants were collected as “cytosolic fraction,” whereas the pellets (nuclei) were washed twice with ice-cold 0.1% Nonidet P-40-PBS. The harvested pellets were resuspended in Laemmli sample buffer, sonicated for 30 s, and collected as “nuclear fraction.” Equivalent proportions of two fractions were analyzed by SDS-PAGE and Western blotting. The purity of the fractions was assessed by detecting specific subcellular marker proteins such as α-tubulin as cytoplasmic protein and Lamin B1 as nuclear protein.

FIGURE 8.

Correlation between nuclear-cytoplasmic shuttling of Hsp70 and viral polymerase protein levels in subcellular fractions. A, HEK293T cells were infected with PR8 influenza virus (MOI 1) or mock infected. At 12 h post-infection, cells were treated as in Fig. 6A. An additional 24-h recovery phase sample was also included. The nuclear and cytoplasmic fractions from mock and PR8-infected cells were prepared and analyzed by Western blotting. Quality of fractions was assessed by blotting for Lamin B1 (nuclear fraction) and α-tubulin (cytoplasmic fraction). B, relative quantification of PB2, PB1, PA, NP, and Hsp70 in nuclear and cytoplasmic fractions. The levels of PB2, PB1, PA, NP, and Hsp70 in the nuclear fractions were normalized to the Lamin B1, whereas levels of PB2, PB1, PA, NP, and Hsp70 in the cytoplasmic fraction were normalized to the α-tubulin and were expressed relative to those during the pre-heat shock phase. The results are from two or three independent experiments. C, Western blot (bottom) and relative quantities of PB2, PB1 (top) in whole cell lysate from A. The PB2 and PB1 levels were normalized to the β-actin and expressed relative to their pre-heat shock levels. The values are from three independent experiments. The error bars represent means ± S.E.

siRNA Design and Knockdown of Hsp70

The siRNA targeting Hsp70 (siHsp70-1) was purchased from Santa Cruz Biotechnology Inc. (sc-29352). An additional Hsp70-specific siRNA (siHsp70-2, 5′-CGGUGGUGCAGUCGGACAUGA-3′) was designed using an online siRNA designing tool: Design for Small Interfering RNA (DSIR) (47). The siHsp70-2 and a nonsilencing control siRNA (5′-AAUUCUCCGAACGUGUCACGU-3′) (48) were purchased from Sigma Genosys siRNA Service. HEK293T and HeLa cells were transfected twice (first reverse transfection and second forward transfection) on alternative days with Hsp70-specific siRNA, control siRNA and transfection buffer alone (mock) using Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer's recommendations. HEK293T and HeLa cells were transfected with final concentrations of 80 and 40 nm for siHsp70-1, respectively, and 80 nm for siHsp70-2. Control siRNA concentrations were kept similar to those of Hsp70-specific siRNAs. Twenty-four hours after Hsp70 knockdown, cells were either infected with PR8 influenza virus at a MOI of 0.1 or transfected with HK483 or PR8 RNP expression plasmids for further analysis.

Luciferase Reporter Assays

Luciferase (firefly and Renilla luciferase) activities were measured with GloMax96 microplate Luminometer (Promega) using the Dual-Luciferase® assay system (Promega). All transfections in HEK293T or HeLa cells were performed using TransIT®-LT1 (Mirus) according to the manufacturer's recommendations. Luciferase activities were measured at 24 h post-transfection, except where otherwise indicated. Firefly luciferase activities were normalized to the transfection control Renilla luciferase activities and were expressed relative to that of mock treated cells, which were set to 1. Influenza virus polymerase-driven luciferase activities (viral polymerase activities) were measured by transfecting cells with the indicated vRNP expression plasmids, pCAGGS-NP (400 ng), -PB2, -PB1, and -PA (200 ng each), the pHW72-Luc2CP firefly luciferase reporter plasmid (100 ng), and pRL-CMV (Promega) Renilla luciferase transfection control reporter plasmid (50 ng), except where otherwise indicated.

To assess the effect of heat shock, PGA1 and plasmid-mediated overexpressed Hsp70 on NF-κB promoter activity, HEK293T and HeLa cells were transfected with pNFκB-Luc, an NF-κB promoter-dependent firefly luciferase reporter plasmid (Agilent Technologies), and pRL-CMV, a transfection control reporter plasmid alone or with pHA-Hsp70. At 24 h post-transfection, cells were treated with PGA1, heat shock, or TNFα alone or a combination of them as indicated in figures.

To assess the effect of heat shock on both luciferase protein activities, cells were co-transfected with pCAGGS-Luc2CP (100 ng) and pRL-CMV (50 ng) reporter plasmids. The growth medium was replaced with the one containing either CHX 100 μg/ml or ethyl alcohol (solvent for CHX) 30 min before starting the heat shock experiment as shown in Fig. 6A.

FIGURE 6.

Hsp70 enhances the viral polymerase activity during the heat shock phase. A, schematic diagram illustrating the experiment layout. B–H, HEK293T (B–D) and HeLa (F–H) cells were transfected with PR8 (B, C, F, and G) and HK483 (D and H) RNP expression plasmids along with reporter plasmids. B and F, after treating cells as in A, RNA was extracted and treated with DNaseI, and m-, c-, and vF.LucRNA levels of firefly luciferase reporter gene were quantified by real time PCR, normalized with GAPDH mRNA, and expressed relative to those of vF.LucRNA levels. The results are from three independent experiments, each performed in triplicate. C, D, G, and H, after treating cells as in A, cell lysates were prepared, and viral polymerase activities were measured by dual luciferase reporter assay. Cells maintained at 37 °C (pre-heat shock) served as controls. The results are from three independent experiments, each performed in triplicate. E, viral RNA expression in HEK293T cells infected with PR8 influenza virus. The cells were infected at a MOI of 1. After 12 h, cells were treated as in A; total RNA from cells was collected, and mRNA, cRNA, and vRNA levels of the NP gene were quantified by real time PCR. GAPDH was used as reference gene. The data are presented as fold change (C_T values) relative to cells infected at 37 °C. The results are based on three independent experiments performed in duplicate or triplicate. The data were analyzed by Student's t test. The error bars represent means ± S.E. *, p < 0.05; **, p < 0.01; ***, p < 0.001. luc., luciferase.

Overexpression of Hsp70 and Viral Transcription/Replication Activity

Hsp70 overexpression was achieved either by transfecting cells with the Hsp70 expression plasmid, by treating them with PGA1, or by subjecting them to heat shock. Plasmid-mediated Hsp70 overexpression was achieved by transfecting HeLa and HEK293T cells with increasing doses of the HA-Hsp70 expression plasmid along with HK483 RNP expression plasmids and reporter plasmids as indicated before. The total plasmid concentration was adjusted using the empty pCAGGS plasmid.

To induce Hsp70 overexpression by PGA1, cells transfected with vRNP expression plasmids and reporter plasmids were treated with PGA1 (20 μg/ml) for 3 h, and then growth medium containing PGA1 was replaced with fresh growth medium. Four hours later cell lysates were prepared, and luciferase activities were measured.

We also ascertained the relation between the subcellular localization of Hsp70 during different phases of the heat shock response and influenza virus transcription and replication activities either by transfecting cells with vRNP expression and reporter plasmids or by infecting cells with PR8 influenza virus at a MOI of 1. The schematic diagram of the experiment is shown in Fig. 6A. An additional group of cells transfected with all sets of plasmids except pCAGGS-PB1 was included as a background control. The heat shock response was divided into pre-heat shock, heat shock, and recovery (post-heat shock) phases. At 24 h post-transfection or 12 h postinfection, cells in one group were subjected to heat shock at 42 °C for 3 h, representing the heat shock phase. Another group of cells was subjected to heat shock at 42 °C for 3 h and then allowed to recover at 37 °C for 4 h. This group represented the recovery phase. Cells maintained at 37 °C throughout the course of the experiment served as untreated controls and represented the pre-heat shock phase. Viral polymerase activities in transfected cells were measured both by dual luciferase reporter assay and by measuring the influenza virus-like firefly luciferase m-, c-, and vRNA levels (mF.LucRNA, cF.LucRNA, and vF.LucRNA), whereas viral replication was determined by measuring viral NP gene vRNA, mRNA, and cRNA levels. We also determined the viral polymerase activities in mock, control siRNA-, and Hsp70-specific siRNA-treated HeLa cells during the pre-heat shock, heat shock, and recovery phases at indicated time points. The schematic diagram of the experiment is shown in Fig. 9A. The values of control siRNA- and siHsp70-treated cells were expressed relative to those of mock treated cells of the same group at that particular time point.

FIGURE 9.

Hsp70 induction restores viral polymerase activity during the heat shock phase in Hsp70 knocked down cells. A, schematic diagram illustrating the experiment layout. B, HeLa cells were treated with Hsp70 specific siRNA (siHsp70-2), control siRNA, or transfection reagent only (Mock) twice at alternate days. At 24 h post-knockdown, cells were transfected with HK483 RNP expression plasmids along with reporter plasmids as indicated in Fig. 3. After 24 h, cells were treated as in A, cell lysates were prepared, and viral polymerase activities were measured. Mock cells served as control. The values were normalized to the mock of each group at indicated time points. The results are based on three independent experiments, each performed in triplicate. The error bars indicate means ± S.E. C, Western blot showing changes in Hsp70 levels in mock, control siRNA-, and siHsp70-2-treated cells at different time points. β-Actin served as loading control.

Immunofluorescence Assay

All monoclonal antibodies, used for the co-localization studies of Hsp70 and polymerase subunits were of mouse origin. Therefore, double staining was achieved by labeling mouse anti-PB2, -PB1, and -PA mAbs with Alexa Flour 488 or Alexa Flour 594 dyes using Zenon® Alexa Fluor 488® labeling kit (Z25002) and Zenon® Alexa Fluor 594® labeling kit (Z25007). Prior to labeling, all primary antibodies were affinity-purified using an Affi-Prep MAPS II kit (Bio-Rad).

HEK293T cells were grown in 8-well Lab-Tek^TM chamber slides and were transfected with the indicated expression plasmids as shown in the figures. The slides were treated with poly-l-lysine (Cultrex) prior to the cell propagation according to the manufacturer's instructions. At 24 h post-transfection, cells were washed with sterile ice-cold PBS, fixed, and permeabilized with ice-cold methanol and blocked with 1.0% bovine serum albumin in PBS for 30 min. Later, cells were incubated for 30 min with primary antibodies (1:1000 diluted in PBS-BT; PBS containing 1% BSA and 0.05% Tween 20) specific for Hsp70 or HA tag. After incubation with primary antibodies, cells were washed with PBS and incubated with indicated Alexa Fluor-conjugated secondary antibodies (1:1000 diluted in PBS-BT) for a further 30 min and then again washed with PBS.

Single staining of the polymerase subunits was achieved by adding Alexa Fluor 594-labeled primary antibodies specific for PB2, PB1, or PA proteins. Double staining was achieved by adding the combination of Alexa Fluor 488- and Alexa Fluor 594-labeled primary antibodies specific for the PB2, PB1, or PA proteins. The unbound antibodies were removed by washing the cells with PBS. A second fixation was done by treating cells with 4% formaldehyde in PBS for 15 min. The cells were washed with PBS, and their nuclei were stained with DAPI.

After the final washing, mounting medium (Vectashield, Vector Laboratories) was added, and coverslips were applied to the slide. The cells were examined using the LSM780 confocal microscope (Carl Zeiss), and pictures were taken using Zen 2010D (Carl Zeiss) software.

RNA Extraction, Reverse Transcription, and Real Time PCR

Total RNA was extracted from infected or transfected cells using Isogen (Nippon Gene, Japan) following the manufacturer's protocol. RNA samples from transfected cells were treated with DNaseI (Invitrogen) according to the manufacturer's recommendations. RT was carried out using Superscript III reverse transcriptase kit (Invitrogen) in a 20-μl reaction mixture containing 300 ng of total RNA and specific primers. The NP gene- and sense-specific primers (PR8-NPcRNA, 5′-AGTAGAAACAAGGGTATTTTTC-3′), and (PR8-NPvRNA, 5′-AGCAAAAGCAGGGTAGATAATCACTCAC-3′) were used for cRNA and vRNA amplification from infected cells. Uni-12 and Uni-13 primers were used for amplification of cF.LucRNA and vF.LucRNA of firefly luciferase reporter gene from transfected cells. Oligo(dT)₂₀ primer was used for mRNA amplification from both infected and transfected cells. A GAPDH-specific primer was also included in the RT reaction mixtures for vRNA and cRNA estimation in infected and transfected cells. RT reaction was carried out according to the manufacturer's protocol. One microliter of RT mixture was used in real time PCR using SYBR® Premix Ex Taq^TM II (Tli RNase H Plus) kit (Takara, Japan) and gene specific primers. The reaction was performed at 95 °C for 10 s followed by 40 cycles at 95 °C for 5 s and at 60 °C for 30 s. All the reactions were carried out in replicates of two or three. The specificity of the primers was assessed by dissociation (melting) curve analysis. The levels of PCR products were monitored and analyzed with a CFX96^TM real time PCR detection system (Bio-Rad). The relative amounts of RNAs (mRNA, cRNA, and vRNA), expressed as threshold cycle (C_T) values, were normalized by the amount of GAPDH mRNA and expressed relative to an arbitrary value of 1 (49). Sequences of the gene-specific primers were NP-1186F (5′-ACCAATCAACAGAGGGCATC-3′) and NP-1333R (5′-TGATTTCGGTCCTCATGTCA-3′); F.Luc2CP-395F (5′-GGCTGCAAAAGATCCTCAAC-3′ and F.Luc2CP-514R (5′-AATGGGAAGTCACGAAGGTG-3′; and housekeeping gene GAPDH-556F (5′-TGCACCACCAACTGCTTAGC-3′) and GAPDH-642R (5′-GGCATGGACTGTGGTCATGAG-3′).

RESULTS

Hsp70 Interacts with Influenza Virus PB2 and PB1 Monomers as Well as PB2/PB1 Heterodimers but Not with PB2/PB1/PA Heterotrimers

Preliminary immunoprecipitation studies carried out using anti-PB2 mAb showed that interaction between Hsp70 and PB2 was specific. However, the efficacy of Hsp70 co-precipitation was not good, which could be due to steric hindrance offered by mAb (Fig. 1A). Therefore, keeping in view our observation and previous findings (35, 50), we decided to use tagged polymerase proteins or Hsp70. Iwai et al. (50) reported that addition of FLAG or HA tags to the N termini of influenza virus polymerase subunits did not interfere with their assembly into functional heterotrimeric polymerase complex. Therefore, we constructed expression plasmids encoding N-terminal FLAG-tagged viral polymerase proteins (FLAG-HK483PB2 and FLAG-HK483PB1) and N-terminal HA-tagged Hsp70 (HA-Hsp70). Co-immunoprecipitation experiments carried out using FLAG-HK483PB2andHA-Hsp70 showed that Hsp70 was successfully precipitated with PB2 and vice versa, indicating that addition of FLAG tag to the PB2 subunit or HA tag to Hsp70 did not affect the interaction between them (Fig. 1, A and B).

FIGURE 1.

Hsp70 interacts with PB2, PB1 monomers, and their dimers, but not with PB2/PB1/PA heterotrimer. A and B, effects of addition of HA and FLAG tags on the interaction of Hsp70 with PB2 of HK483 influenza virus. HEK293T cells were transfected with indicated plasmids, and immunoprecipitated proteins were identified by Western blotting using anti-HA tag, anti-FLAG tag, anti-Hsp70, and anti-PB2 mAbs. C, HEK293T cells were transfected with plasmids encoding HK483-FLAGPB2 alone or in combination with plasmids expressing the indicated polymerase subunits of HK483 influenza virus. Proteins were immunoprecipitated using anti-FLAG affinity gel and analyzed by SDS-PAGE followed by Western blotting. D, immunoprecipitation was carried out as in C except that HEK293T cells were transfected with plasmids encoding HK483-FLAGPB1 protein alone or in combination with plasmids expressing the indicated viral polymerase subunits. E, HEK293T cells were transfected with a plasmid encoding HA-Hsp70 alone or in combination with plasmids encoding the indicated polymerase subunits of PR8 influenza virus. Proteins were immunoprecipitated using anti-HA affinity gel and analyzed by SDS-PAGE followed by Western blotting. IP, immunoprecipitation.

It was shown that endogenous Hsp70 interacted with PB2 and PB1 monomers, (35), but it was not clear whether Hsp70 interacted with viral polymerase heterodimers or heterotrimers. To evaluate such interaction, HEK293T cells were transfected with FLAG-HK483PB2 or FLAG-HK483PB1 expression plasmids alone or in combination with other polymerase subunit expression plasmids and analyzed by immunoprecipitation studies as shown in Fig. 1 (C and D). We found that endogenous Hsp70 was co-immunoprecipitated with both FLAG-HK483PB2 and FLAG-HK483PB1 monomers as well as FLAG-HK483PB2/PB1 and FLAG-HK483PB1/PB2heterodimers but not with FLAG-HK483PB1/PA heterodimer and FLAG-HK483PB2/PB1/PA or FLAG-HK483PB1/PB2/PA heterotrimers (Fig. 1, C and D). Moreover, reciprocal immunoprecipitation using HA-Hsp70 and PR8 polymerase subunits confirmed that exogenous HA-Hsp70 also interacted mainly with the PB2 and PB1 monomers and to a lesser extent with the PA monomer (Fig. 1E). Similar results were obtained with Aichi (H3N2) polymerase subunit (data not shown), indicating that this interaction was not influenza virus strain-specific. These findings suggested that Hsp70 seemed to act as chaperone for PB2 and PB1 proteins and appeared to assemble PB2/PB1 dimer and was released from the complex upon introduction of the PA subunit.

Hsp70 Translocates and Co-localizes with Influenza Virus PB2 Protein

Although it has been shown that Hsp70 translocated into the nucleus upon infection with influenza virus (35), it was not clear whether Hsp70 translocated into the nucleus with the vRNP complex or with individual polymerase subunits. To address this issue, the subcellular localization of Hsp70 with each of the polymerase subunits was studied in HEK293T cells transfected with individual polymerase subunit expression plasmids. As already known, PB2 protein was mainly present in the nuclei of cells, whereas PB1 and PA proteins were present in both the cytoplasm and the nucleus (18). In mock transfected cells, endogenous Hsp70 was mainly located in the cytoplasm (Fig. 2A, panel a), and it only translocated into the nuclei of the cells expressing PB2 protein, but not in the cells expressing PB1 or PA (Fig. 2A, panels b–d). Similarly, Hsp70 translocated into the nuclei with only PB2 protein of PR8 influenza virus (data not shown). Because viral replication takes place in the nucleus, the polymerase subunits have to be transported into it to make a functional polymerase complex, which consists of PB2, PB1, and PA subunits; PB1 is in the center, having binding sites for PB2 and PA subunits, whereas PB2 and PA subunits do not interact directly (51). To explain the assembly of viral polymerase subunits into a functional polymerase complex while keeping in view the results for co-immunoprecipitation obtained in the present study, we transfected HEK293T cells with plasmids expressing HK483PB2/PB1, PB1/PA, and PB2/PA heterodimers and studied their co-localization patterns as well as effects on the localization of Hsp70. It was found that Hsp70 and PB1 translocated into the nuclei of the cells expressing PB2/PB1 heterodimers, suggesting that co-expression of PB2 with PB1 facilitated the nuclear translocation of the PB1 subunit (Fig. 2B, panel a). Although a trace amount of PB1 could be seen in the cytoplasm, its amount was quite small compared with the predominant cytoplasmic presence when cells expressed PB1 alone (Fig. 2A, panel c). In the cells expressing PB1/PA heterodimers, both PB1 and PA were mainly present in the nuclei of the cells with little cytoplasmic presence, and these cells showed little nuclear accumulation of Hsp70 (Fig. 2B, panel b). In the cells expressing PB2/PA heterodimers, the PB2 subunit localized in the nucleus along with Hsp70, whereas PA predominantly remained in the cytoplasm, indicating that there was no direct interaction between PB2 and PA subunits (Fig. 2B, panel c).

FIGURE 2.

Hsp70 translocates into the nucleus with PB2 monomer or PB2/PB1 heterodimer. Subcellular localization of Hsp70 with viral polymerase subunits was analyzed by confocal laser-scanning microscopy. A, HEK293T cells were transfected with the indicated plasmids of HK483 influenza virus or with empty plasmid (Mock). At 24 h post-transfection, the cells were fixed, blocked, and stained with mouse anti-Hsp70 and Alexa Fluor 488® goat anti-mouse IgG mAbs. Staining of viral polymerase proteins was done by using anti-PB2, -PB1, and -PA mAbs labeled with Alexa Flour 594® dye. The arrows in panel b indicate the nuclear localization of Hsp70. Cell nuclei were stained with DAPI. B, HEK293T cells were co-transfected with plasmids encoding viral polymerases (PB2, PB1, and PA) of HK483 influenza virus. At 24 h post-transfection, cells were fixed, blocked, and stained with mouse anti-Hsp70 and Alexa Fluor 405® goat anti-mouse IgG mAbs. Dual staining of polymerase proteins was achieved by labeling the anti-PB2, -PB1, and -PA mAbs with the Alexa Flour® dyes. Red, Alexa Flour 594®; green, Alexa Flour 488® dye). The arrows in panels a and c indicate the nuclear localization of Hsp70.

Knockdown of Hsp70 in HEK293T and HeLa Cells Resulted in Reduced Viral Transcription and Replication Activity

Because both HEK293T and HeLa cells express Hsp70 constitutively, we assessed the role of endogenous Hsp70 in influenza virus transcription and replication by knocking it down using Hsp70-specific siRNAs (siHsp70-1 and siHsp70-2). First, we determined the role of Hsp70 in virus transcription by luciferase reporter assay using siHsp70-1 siRNA in HEK293T and HeLa cells. The results showed 48 and 66% reduction in influenza virus polymerase activity in both cell lines (Fig. 3, A and C). Western blot analysis also confirmed 58% and more than 80% reduction in Hsp70 levels in HEK293T and HeLa cells, respectively (Fig. 3, B and D). It has been shown that siRNAs can produce unexpected or divergent results, partially because of off target effects (52, 53). Therefore, another siRNA (siHsp70-2) was designed. Hsp70 knockdown by siHsp70-2 also resulted in 53 and 46% reductions in the polymerase activities of both HK483 and PR8 vRNPs in HeLa cells, respectively (Fig. 3, E and F). We then determined the effect of Hsp70 knockdown on influenza virus replication at different time points. Hsp70 was knocked down in HeLa cells using siHsp70-2 as before. Then the cells were infected with the PR8 influenza virus at a MOI of 0.1. At 8, 24, 36, and 48 h postinfection, total RNA was isolated form the cells, and the mRNA, cRNA, and vRNA levels of the NP gene were quantified by real time RT-PCR. The results showed a significant reduction in all viral RNA species tested until 36 h postinfection compared with those in control siRNA-treated cells. In comparison to 36 h postinfection, 5–20% increase in all viral RNA species was observed at 48 h postinfection (Fig. 3G). Consistently, levels of PB2, PB1, and NP also decreased as determined at 12 h postinfection by Western blotting (Fig. 3, H and I). These results indicated the requirement of Hsp70 for efficient virus transcription and replication in HEK293T and HeLa cells.

FIGURE 3.

Knocking down Hsp70 decreases the virus transcription and replication. A and C, HEK293T (A) and HeLa (C) cells were transfected twice on alternate days with transfection reagent only (Mock), control siRNA, or Hsp70-specific siRNA (siHsp70-1). Twenty-four hours after knocking down the Hsp70, HEK293T (A) and HeLa (C) cells were transfected with HK483 RNP expression plasmids pCAGGS-NP (200 ng), -PB2, -PB1, and -PA (100 ng each), pHW72-Luc2CP firefly luciferase (50 ng), and pRL-CMV Renilla luciferase (25 ng) reporter plasmids. After 24 h, viral polymerase activities were measured. Nonsilenced cells served as controls (Mock). The results are based on three independent experiments, each performed in duplicate or triplicate. B and D, Western blot analysis (bottom panel) and its quantification (top panel) from A and C. The Hsp70 levels were normalized to β-actin, which served as a loading control, and were expressed relative to the mock treated cells. E and F, polymerase activities of HK483 and PR8 influenza viruses were measured in HeLa cells as in A and C, except that knockdown of Hsp70 was carried out using siHsp70-2 siRNA. The results are based on three independent experiments, each performed in triplicate. G, Hsp70 was knocked down as in E and F, and cells were infected with PR8 influenza virus at a MOI of 0.1. At 8, 24, 36, and 48 h postinfection, mRNA, cRNA, and vRNA levels of the NP gene were quantified by real time PCR, normalized with GAPDH mRNA, and expressed relative to those of control siRNA-treated cells. The results are based on two to three independent experiments, each performed in triplicate. H and I, Western blot analysis of viral proteins in Hsp70 knocked down HEK293T cells. Cells were infected as in G. At 12 h postinfection, cell lysates were analyzed by Western blotting. The indicated protein levels were normalized to the β-actin, which served as loading control, and were expressed relative to the mock treated cells. The data were analyzed by Student's t test. The error bars represent means ± S.E. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Increase in Viral Polymerase Activity during the Heat Shock Phase Is Correlated with Increased Intranuclear Accumulation of Hsp70 and Viral Polymerases

It is well known that following heat shock, there is a marked increase in the production and nuclear localization of Hsp70, which slowly returns to the cytosol during the recovery phase (54, 55). We also confirmed by immunofluorescence assay that Hsp70, after 3 h of heat shock at 42 °C, mainly localized in the nucleus, and most of it relocated into the cytoplasm after 4 h of recovery at 37 °C (Fig. 4). We hypothesized that if Hsp70 had an inhibitory effect on viral polymerase activity, it should have been more pronounced during the heat shock phase than the recovery phase, coinciding with the movement of Hsp70 into the nucleus. Moreover, increase or decrease in viral polymerase activities might be related to increase or decrease in viral polymerase proteins and Hsp70 levels in nuclear/cytoplasmic compartments during these phases of heat shock response. To test these hypotheses, we measured the viral transcription and replication activities as well as viral polymerase proteins and Hsp70 levels in infected cells by Western blotting.

FIGURE 4.

Subcellular localization of Hsp70 during different phases of the heat shock response. HEK293T (A) and HeLa (B) cells were subjected to heat shock or allowed to recover as shown in Fig. 6A. Treated cells were fixed, blocked, and stained with mouse anti-Hsp70 and Alexa Fluor 488® goat anti-mouse IgG mAbs. The nuclei were stained with DAPI.

To measure the viral polymerase activity during different phases of heat shock response, HEK293T and HeLa cells were transfected with PR8 vRNP expression plasmids and reporter plasmids, pRL-CMV and pHW72-Luc2CP. The ORF of firefly luciferase (Luc2CP) is cloned in minus sense with flanked noncoding regions of influenza virus M gene and human PolI promoter and the murine terminator sequences. The primary transcription of influenza virus-like firefly luciferase RNA (vF.LucRNA) from pHW72-Luc2CP is driven by human PolI enzyme, whereas the expression of mF.LucRNA and cF.LucRNA from the vF.LucRNA is driven by influenza virus polymerase complex. Therefore, at first we measured all three RNA species and then determined the change in m- and cF.LucRNA expression relative to vF.LucRNA levels. Interestingly, during the heat shock phase, vF.LucRNA levels remained unchanged in HEK293T cells, whereas more than 40% reduction in vF.LucRNA level was observed in HeLa cells. During the recovery phase, vF.LucRNA levels increased in both cells lines, suggesting a difference in human PolI activity between both cell lines during the heat shock phase (Fig. 5, A and B). Then we determined the changes in c- and mF.LucRNA levels relative to the vF.LucRNA. During the heat shock phase, more than 100 and 80% increase in mF.LucRNA levels and more than 50 and 30% increase in cF.LucRNA levels were observed in HEK293T and HeLa cells, respectively (Fig. 6, B and F). During the recovery phase, a trend in reduction in both mF.LucRNA and cF.LucRNA levels was observed in both cell lines. These findings suggested an increase in viral polymerase activity during the heat shock phase.

FIGURE 5.

Relations between m-, c-, and vF.LucRNA levels in HeLa and HEK293T cells during the pre-heat shock, heat shock, and recovery phases. HeLa (A) and HEK293T (B) were transfected with PR8 RNP expression plasmids along with reporter plasmids. After treating cells as in Fig. 6A, RNA was extracted, treated with DNaseI, m-, c-, and vF.LucRNA levels of firefly luciferase reporter gene were quantified by real time PCR, normalized with GAPDH mRNA, and expressed relative to pre-heat shock levels, which were set to 1. The results are from three independent experiments, each performed in triplicate. The error bars represent means ± S.E.

We also measured the transcriptional activity of viral polymerase by measuring the luciferase activity in transfected HEK293T and HeLa cells. Oza et al. (56) studied the influence of neuronal cell differentiation on the heat shock gene expression. They transfected the cells with Hsp70 promoter-dependent luciferase reporter plasmid along with Renilla expressing internal control plasmid, subjected the cells to heat shock for different durations, and determined the luciferase activities, therefore suggesting the applicability of luciferase reporter assay in this study. HeLa and HEK293T cells were transfected with the indicated plasmids. Cell lysates were prepared, and luciferase activities were measured during the pre-heat shock, heat shock, and recovery phases. Interestingly, the viral polymerase-driven luciferase activities of HK483 and PR8 vRNPs were significantly higher in cell lysates prepared immediately after heat shock than in cell lysates prepared after 4 h of recovery, thus coinciding with the movement of Hsp70 inside the cells (Fig. 6, C, D, G, and H). To rule out the possibility that increase in luciferase activity was not due to the stabilizing effect of Hsp70 on luciferase proteins, we determined the influence of heat shock at 42 °C and subsequent incubation at 37 °C (recovery) on firefly and Renilla luciferase activities. Cells were transfected with plasmids expressing firefly (Luc2CP) and Renilla luciferase proteins with or without CHX treatment. It was clear that luciferase activity of both luciferase proteins reduced during the heat shock phase and increased during the recovery phase (Fig. 7), suggesting that the increase in viral polymerase activity during the heat shock phase could be due to increased polymerase activity caused by increased accumulation of viral polymerase proteins.

FIGURE 7.

Effect of heat shock on firefly and Renilla luciferase activities. HeLa cells were co-transfected with firefly luciferase (pCAGGS-F.Luc2CP, 100 ng) and Renilla luciferase (pRL-CMV, 50 ng) expression plasmids. The concentration and ratio of expression plasmids was kept similar as in viral polymerase reporter assays. Thirty minutes before starting the treatment as outlined in Fig. 6A, the cells were treated with 100 μg/ml CHX (A) or ethyl alcohol (B, solvent for CHX). Firefly and Renilla luciferase activities were determined from cell lysates, normalized, and expressed relative to the pre-heat shock values. The results are from two independent experiments, performed in triplicate. The error bars represent means ± S.E. luc., luciferase.

We also evaluated the relations among the pre-heat shock, heat shock and recovery phases and the relative quantities of mRNA, cRNA, and vRNA of the NP gene of PR8 influenza virus in infected HEK293T cells. At 12 h postinfection, the infected cells were subjected to heat shock as described above, and total cellular RNA was isolated and reverse transcribed, and relative quantities of mRNA, cRNA, and vRNA of NP gene were determined. We found a more than 2-fold increase in the mRNA levels of the NP gene in HEK293T cells during the heat shock and recovery phases compared with the pre-heat shock phase. There was an ∼44% reduction in vRNA level during the recovery phase, whereas vRNA levels remained unchanged during the pre-heat shock and heat shock phases. Interestingly, cRNA levels also remained unchanged during the pre-heat shock, heat shock, and recovery phases (Fig. 6E).

Later, we measured viral polymerase proteins and Hsp70 quantities in the subcellular fractions during different phases of heat shock response to ascertain the possible relation with viral polymerase activities. HEK293T cells were infected with PR8 influenza virus at a MOI of 1. After 12 h, cells were treated as in Fig. 6A, and an additional 24-h recovery phase sample was also collected. Nuclear and cytoplasmic fractions were analyzed by Western blotting (Fig. 8A). It was quite interesting to find that there was a 3–4-fold increase in the viral PB2, PB1, and PA proteins and a 1.5-fold increase in NP protein levels in the nuclear fractions with reduction in their cytoplasmic levels during the heat shock phase. In case of Hsp70, ∼12-fold increases in the nuclear fraction and a slight reduction in the cytoplasmic fraction was seen. During the recovery phase at 4 h post-heat shock, a reduction in PB2, PB1, PA, and NP protein levels in the nuclear fractions with corresponding increase in cytoplasmic fractions was observed, and this nuclear to cytoplasmic shift continued even at 24 h post-heat shock. A similar but less pronounced reduction of Hsp70 levels in nuclear fractions and an increase in cytoplasmic fractions were observed (Fig. 8B). These findings suggested a possible correlation between Hsp70-assisted increase in viral polymerase proteins levels in the nucleus, resulting in increased viral polymerase activities. Interestingly, we did not detect Hsp90 in the nuclear fractions prepared from PR8-infected cells. Western blot analysis of whole cell lysates from Fig. 8A showed a reduction in overall quantity of viral proteins (Fig. 8C).

Heat Shock Restores Viral Polymerase Activity in Hsp70 Knocked Down Cells

We hypothesized that if Hsp70 had inhibitory effect on influenza virus polymerases, it should be observed in the Hsp70 knocked down cells during the heat shock phase, when Hsp70 expression is induced and Hsp70 translocates into the nucleus. Therefore, we determined the viral polymerase activities in mock, control siRNA-, and siHsp70-treated HeLa cells at 30, 90, and 180 min after heat shock as well as in the pre-heat shock and recovery (4- and 24-h post-heat shock) phases (Fig. 9A). Interestingly, an increase in the viral polymerase activity with the increase in heat shock duration (heat shock phase) and a subsequent reduction in viral polymerase activity with the increase in recovery time (recovery phase) were observed (Fig. 9B). Western blot analysis showed an increase in the Hsp70 levels in Hsp70 knocked down cells during the heat shock and recovery phases (Fig. 9C). These findings also confirmed that during the heat shock phase, Hsp70 translocated into the nucleus and increased the viral polymerase activity, whereas during the recovery phase, although Hsp70 levels increased, Hsp70 moved out into the nucleus leading to reduction in viral polymerase activity.

Plasmid-mediated Overexpression of Hsp70 Reduces the Viral Polymerase Activity

The knockdown of Hsp70 not only reduced the polymerase activity in Hsp70-depleted HEK293T and HeLa cells but also reduced the virus replication in infected cells. From these experiments, it was clear that Hsp70 at normal levels (refers to Hsp70 level in cells not treated to induce Hsp70 production) was required for influenza virus replication in both cell lines. However, what would be the effect on virus polymerase activities if Hsp70 is overexpressed in HEK293T and HeLa cells? To determine this, Hsp70 overexpression was achieved by transfecting cells with increasing doses of the Hsp70 expression plasmid.

Both HEK293T and HeLa cells were transfected with HK483 RNP expression plasmids, reporter plasmids, and the HA-Hsp70 expression plasmid in doses of 100, 200, 400, and 800 ng or empty pCAGGS to adjust the total amount of transfected DNA. Cell lysates were prepared 24 h post-transfection, and viral polymerase activities were measured. Compared with mock treated cells, a dose-dependent reduction in polymerase activity in both cell lines was observed (Fig. 10A). Interestingly, Western blot analyses showed no significant changes in the amounts of PB2 or PB1 proteins in mock and HA-Hsp70-transfected cells (Fig. 10B). Immunofluorescence staining of HA tag revealed that HA-Hsp70, like endogenous Hsp70, remained mainly in the cytoplasm of mock transfected cells or cells expressing PB1 and PA proteins (data not shown) but translocated into the nuclei of the cells expressing PB2 protein (Fig. 10C).

FIGURE 10.

Plasmid-mediated Hsp70 overexpression decreases the influenza virus polymerase activity. A, HEK293T and HeLa cells were transfected with 100, 200, 400, and 800 ng of the HA-Hsp70 expression plasmid or empty vector (Mock) along with HK483 RNP expression plasmids and reporter plasmids. After 24 h, viral polymerase activities were measured. The results are based on three independent experiments, each performed in triplicate. The Student's t test was used for statistical comparison of each of the HA-Hsp70 transfected group with mock. The error bars represent means ± S.E. *, p < 0.05; **, p < 0.01. B, representative Western blot of cell lysates from A. C, immunofluorescence staining of HEK293T cells transfected with the HA-Hsp70 expression plasmid alone (panel A) or co-transfected with HK483 PB2 expression plasmid (panel B). The expressed HA-Hsp70 was stained with anti-HA tag and Alexa Fluor 488® goat anti-mouse IgG (green) mAbs and PB2 with purified mouse anti-PB2 mAb labeled with Alexa Flour 594® dye (red). Cell nuclei were stained with DAPI.

Hsp70 Overexpression Suppresses the NF-κB Activity

So far, the results obtained showed that suppression of viral polymerase activity by plasmid-mediated Hsp70 overexpression or by heat shock during the recovery phase varied in terms of degree of suppression and viral protein expression. The results also suggested that these treatments might be affecting virus replication through different mechanisms. One of the common key factors affected is NF-κB. Heat shock, PGA1, and Hsp70 overexpression have been shown to inhibit NF-κB activity in a variety of cells (57–60), which is required for efficient influenza virus replication (61). Therefore, we also determined the effect of heat shock, PGA1, and plasmid-mediated Hsp70 overexpression on NF-κB activity in HEK293T and HeLa cells. The cells were transfected with the pNF-κB-luc reporter plasmid (which expresses the firefly luciferase protein under the influence of the NF-κB promoter) alone or with pHA-Hsp70 (800 ng), and pRL-CMV plasmid as a transfection control. We found that TNFα (10 ng/ml for 3 h) significantly increased the luciferase activity compared with the mock control, indicating the activation of NF-κB promoter. Both PGA1 and heat shock treatments significantly inhibited the NF-κB promoter activity, as indicated by a more than 80% reduction in luciferase activity. Plasmid-mediated Hsp70 overexpression caused only ∼25% suppression in NF-κB promoter activity (Fig. 11, A–D).

FIGURE 11.

Heat shock, PGA1, and plasmid-mediated overexpressed Hsp70 reduces NF-κB promoter activity. A–C, HEK293T (A) and HeLa (B and C) cells were transfected with pNFκB-Luc (1 μg) carrying an NF-κB promoter-dependent luciferase reporter construct and pRL-CMV (100 ng), a transfection control reporter plasmid. A and B, at 24 h post-transfection, cells were either treated with PGA1 (30 μg/ml), TNFα (10 ng/ml), or first with PGA1 for 3 h and then with TNFα for 4 h. Untreated cells served as mock control. After the indicated treatments, NF-κB promoter-driven luciferase activities were measured by dual luciferase reporter assay. C, at 24 h post-transfection, cells were either subjected to heat shock for 3 h, treated with TNFα (10 ng/ml) for 4 h, or first heat-shocked for 3 h and then treated with TNFα for 4 h. Untreated cells kept at 37 °C served as mock control. The results are from three independent experiments performed in triplicate. The error bars represent means ± S.E. D, HEK 293T cells were co-transfected with pNFκB-Luc (100 ng), pRL-CMV (50 ng), and pHA-Hsp70 (800 ng). In mock cells, empty pCAGGS plasmid was replaced with pHA-Hsp70. After 24 h, cells were stimulated with TNFα (5 ng/ml) for 4 h, and then NF-κB promoter-driven luciferase activities were measured by dual luciferase reporter assay. The results are from two independent experiments, each performed in triplicate. The error bars represent means ± S.E.

DISCUSSION

In this study, we investigated the role of Hsp70 in influenza virus transcription and replication in cells constitutively expressing Hsp70. Although previous studies (34, 35) have shown that Hsp70 inhibits influenza virus replication, those studies were conducted in cells expressing low levels of Hsp70 or in cells where higher levels of Hsp70 were induced by PGA1 or heat shock treatment. However, other cellular events contingent upon PGA1 or heat shock treatment were overlooked in those studies. In the present study, we found that Hsp70 had modulatory effect on influenza virus replication, presumably by facilitating the viral polymerase activity when present at normal levels and by suppressing the polymerase activity when its levels were increased above normal levels. Previously Li et al. (35) demonstrated that Hsp70 interacted with PB2 and PB1 proteins and translocated into the nuclei of A549 cells upon infection with influenza virus. We also found that Hsp70 interacted with the PB2 and PB1 monomers. Additionally, we found that Hsp70 co-precipitated with the PB2/PB1 heterodimer, and inclusion of the PA subunit in the PB1/PA heterodimer or PB2/PB1/PA heterotrimer resulted in separation of Hsp70 as indicated by its absence in immunoprecipitates (Fig. 1). Naito et al. (18) reported that Hsp90α interacted with influenza virus PB2, PB1 monomers, or PB2/PB1 heterodimers and translocated into the nucleus with PB2 monomers or PB1/PB2 heterodimers. They suggested that PB2 helped in the nuclear translocation of PB1. Hemerka et al. (62) also demonstrated the nuclear accumulation of PB1/PB2 heterodimers using bimolecular fluorescence complementation assay. In contrast, Fodor and Smith (63) have reported that the PB1 and PA subunits interact in the cytoplasm and are transported into the nucleus as a heterodimer, whereas PB2 is transported into the nucleus as a monomer. It is also reported that PB2 and PB1 are present in the nucleus, and PA remains in the cytoplasm during the early phase of infection (64). In general, current findings (Fig. 2) suggest that Hsp70 acted as a chaperone and translocated into the nucleus with PB2/PB1 dimer and that polymerase subunits assemble into a functional heterotrimeric complex in the nucleus.

The results obtained from immunoprecipitation and immunofluorescence studies suggested that Hsp70 played an important role in virus replication, and this role was confirmed by knocking down Hsp70 using Hsp70-specific siRNA. In contrast to Li et al. (35), our results suggested that knocking down Hsp70 caused a reduction in HK483 polymerase activity in HEK293T and HeLa cells (Fig. 3, A and C). Because our findings were not in agreement with those of Li et al., we thought that the reduction in polymerase activity could be an off target effect of siHsp70-1 siRNA. Therefore, another siRNA (siHsp70-2) targeting Hsp70 ORF was used. Interestingly, siHsp70-2 not only reduced the polymerase activities of both HK483 and PR8 RNPs (Fig. 3, E and F) but also reduced the virus replication as indicated by reductions in the mRNA, cRNA, and vRNA levels of the NP gene in infected cells. The reduction in all viral RNA species until 36 h postinfection with a subsequent increase observed at 48 h postinfection (Fig. 3G) coincided with decrease until 36 h postinfection and then increase in Hsp70 level at 48 h postinfection (data not shown).

Zeng et al. (55) demonstrated that there was a 2-fold increase in the nuclear import and a 3-fold decrease in the nuclear export of Hsp70 during the heat shock phase, resulting in the nuclear accumulation of Hsp70. In contrast, there was a decrease in the nuclear inflow and a marked increase in the nuclear outflow of Hsp70 during the recovery phase, resulting in relocation/accumulation of Hsp70 in the cytoplasm. They also found that Hsp70 interacted not only with protein aggregates but also with diffuse cytoplasmic and nucleoplasmic proteins, thereby extending its cytoprotective effect. In this study, measurements of viral polymerase-driven reporter gene m- and cF.LucRNA by real time PCR, luciferase reporter assays (Fig. 6) and nuclear/cytoplasmic localization analyses (Fig. 8) suggested a correlation between the subcellular location of Hsp70 and viral polymerase activity. Possibly, during the heat shock phase, nuclear translocation of Hsp70 caused an increased inflow/accumulation of viral polymerase subunits into the nucleus, which is the site of virus replication and might facilitate their assembly into functional polymerase complexes. It is also conceivable that enhanced viral polymerase activity might be due to the increased chaperone activity of Hsp70 translocated into the nucleus. Furthermore, during the recovery phase, increased nuclear outflow and reduced inflow of Hsp70 might reduce nuclear levels of viral polymerase subunits either by reducing their inflow, increasing their outflow, or both, leading to reduction in viral polymerase activity. These findings are in part supported by Lang et al. (65), who compared the effects of different incubation temperatures on the polymerase activities and growth of low pathogenic avian influenza viruses. They reported a significant increase in virus titers at 39 °C as well as a 20–60% increase in viral polymerase activities compared with at 35 °C. An increase in the mRNA levels and reduction in vRNA levels after heat shock are in line with the findings of Dalton et al. (66), who reported that Hsp70 was not responsible for down-regulation of viral RNA synthesis. They also reported that higher temperature affected the vRNA levels more than the cRNA levels, whereas mRNA levels either increased or remained unchanged. Previous studies (67–70) have shown that transcription of non-heat shock response genes by PolI and PolII enzymes and protein translation are repressed during heat shock phase, whereas they are significantly enhanced during recovery from heat shock. In agreement with the previous studies, a slight to significant reduction in vF.LucRNA levels, whose transcription is driven by PolI enzyme, during the heat shock phase and a significant increase during the recovery phase were observed. Transcription of mF.LucRNA and cF.LucRNA was influenza virus polymerase-driven, and more than 80 and 100% increases in their levels in both cell lines during the heat shock phase and thereafter reduction to the levels similar to the pre-heat shock ones strongly suggest that viral polymerase activity was increased during the heat shock phase (Fig. 6, B and F).

Interestingly, the mRNA levels remained unchanged the during heat shock and recovery phases in HEK293T cells infected with PR8 influenza virus, whereas polymerase activity determined by luciferase reporter assay showed a considerable reduction during the recovery phase than the heat shock phase (Fig. 6, compare C and E). This difference could be due to differences in the half-lives of mRNA and luciferase reporter proteins. The firefly luciferase (Luc2CP) used in this study contains two degradation sequences that reduce its half-life from 3 to 0.4 h (71). Nguyen et al. (72) and Pinto et al. (73) reported that firefly luciferase becomes rapidly inactivated during the heat shock and the activity recovers during the recovery form heat shock. We also found that not only firefly luciferase but also Renilla luciferase activity is reduced during the heat shock and recovers upon recovery form heat shock (Fig. 7). Therefore, the increase in luciferase activity during the heat shock phase cannot be attributed to the stabilization of luciferases by Hsp70.

Despite the increase in viral transcriptional activity and increased intranuclear accumulation of viral proteins during the heat shock phase, a reduction in total amount of viral proteins, in agreement with previous finding (35), was observed in heat-shocked or PGA1-treated cells (Figs. 8C and 12). Moreover, in agreement with Li et al. (35), we also observed no change in viral protein contents in cells transfected with increasing doses of HA-Hsp70 expression plasmid. This could be due to the difference in the mechanism of virus suppression by heat shock, PGA1 and plasmid-mediated overexpressed Hsp70. Both heat shock and PGA1 are well known stress response inducers to initiate global cell responses involving many cellular events such as induction of Hsp70 production, inhibition of the NF-κB activity (58, 59, 74), increase in the I-κBα production (75, 76), increase in ubiquitin expression (77, 78), translational and transcriptional arrest (79), and degradation of damaged or misfolded proteins (80), all of which possibly affect the virus transcription and replication through various mechanisms, thereby causing a reduction in total viral proteins. On the other hand, so far, the reported spectrum of plasmid-mediated overexpressed Hsp70 to interfere with other cellular events is quite narrow, such as inhibition of NF-κB (57, 60) and protein kinase C activities (81), etc. In fact, Hsp70 has been shown to interact and co-precipitate with p65, c-Rel, p50, and IκBα inhibitory protein, thus causing a delay in the activation of NF-κB complex (82).

FIGURE 12.

PGA1 reduces viral polymerase activity in cells constitutively expressing Hsp70. A and B, HEK293T (A) and HeLa (B) cells were transfected with HK483 and PR8 RNP expression plasmids and reporter plasmids. After 24 h, cells were treated with PGA1 (20 μg/ml) or an equal amount of ethanol (solvent for PGA1) for 3 h. The results are based on three independent experiments performed in triplicate (HK483) or two independent experiments done in quadruplicate (PR8). The error bars represent means ± S.E. *, p < 0.05; **, p < 0.01; ***, p < 0.001. C, Western blot analysis of viral proteins from A. D, quantification of viral proteins in C. The levels of indicated proteins were normalized to the β-actin levels and expressed relative to the mock treated cells.

NF-κB, despite acting as an immediate early mediator in the immune and inflammatory responses, is also involved in promoting many pathologic events such as the progression of AIDS by enhancing the transcription of human immunodeficiency virus-1 (83), increase in susceptibility of cells to influenza virus infection (61), and the efficiency of influenza virus production (84). Kumar et al. (85) showed that knockdown of p65 of NF-κB significantly reduced the vRNA levels in infected cells, and overexpression of p65 caused a significant increase in vRNA levels as determined by a cRNA luciferase-based reporter assay. Considering these findings, we sought to clarify whether plasmid-mediated Hsp70 overexpression, heat shock, or PGA1 affected the NF-κB activity in cells constitutively expressing Hsp70. Our findings, in agreement with previous ones, showed that all these treatments suppressed the NF-κB activity (Fig. 11). Interestingly, there appeared to be a relation between the suppression of the virus polymerase activity and reduction in NF-κB promoter activity. The inhibition of viral polymerase activity by plasmid-mediated overexpression of Hsp70 (approximately 35%) (Fig. 10) was ∼2-fold less than that achieved by cells treated with heat shock and PGA1 (60–80%) (Fig. 12). Similarly, plasmid-mediated overexpression of Hsp70 caused ∼25% suppression of the NF-κB promoter activity, whereas the heat shock or PGA1 treatment caused more than twice (>80%) reduction of the NF-κB promoter activity. However, despite these findings, the exact difference in mechanism of reduction in viral polymerase activity by plasmid-mediated Hsp70 overexpression, heat shock, or PGA1 is not clearly known and requires further investigation. Moreover, it was reported that some host factors interact in a host species-specific manner with viral polymerase proteins (15, 16). The results in the present study showed that there was no difference in the interaction of human Hsp70 (HEK293T) with avian (HEK483) or human origin (PR8 and Aichi) viral polymerases. However, further investigations using avian Hsp70 and viral polymerases of avian or mammalian origin are required to confirm the involvement of Hsp70 in host specificity.

In conclusion, the findings in the present study suggest that Hsp70, at normal levels, acts as a chaperone for viral polymerases in HEK293T and HeLa cells. The modulatory effect of Hsp70 on viral polymerase observed during different phases of heat shock response appears to be a consequence of directional movement of Hsp70 between cytoplasmic and nuclear compartments, whereas some other cellular pathways stimulated by the heat shock or PGA1 treatments might independently affect the viral polymerase activity and therefore obscure the effects of Hsp70.