Phosphorylation of Measles Virus Nucleoprotein Affects Viral Growth by Changing Gene Expression and Genomic RNA Stability

Abstract

The measles virus (MV) nucleoprotein associates with the viral RNA genome to form the N-RNA complex, providing a template for viral RNA synthesis. In our previous study, major phosphorylation sites of the nucleoprotein were identified as S479 and S510. However, the functions of these phosphorylation sites have not been clarified. In this study, we rescued recombinant MVs (rMVs) whose phosphorylation sites in the nucleoprotein were substituted (rMV-S479A, rMV-S510A, and rMV-S479A/S510A) by reverse genetics and used them in subsequent analyses. In a one-step growth experiment, rMVs showed rapid growth kinetics compared with wild-type MV, although the peak titer of the wild-type MV was the same as or slightly higher than those of the rMVs. Time course analysis of nucleoprotein accumulation also revealed that viral gene expression of rMV was enhanced during the early phase of infection. These findings suggest that nucleoprotein phosphorylation has an important role in controlling viral growth rate through the regulation of viral gene expression. Conversely, multistep growth curves revealed that nucleoprotein-phosphorylation intensity inversely correlated with viral titer at the plateau phase. Additionally, the phosphorylation intensity of the wild-type nucleoprotein in infected cells was significantly reduced through nucleoprotein-phosphoprotein binding. Excessive nucleoprotein-phosphorylation resulted in lower stability against RNase and faster turnover of viral genomic RNA. These results suggest that nucleoprotein-phosphorylation is also involved in viral genomic RNA stability.

INTRODUCTION

Measles virus (MV), a Morbillivirus within the Paramyxoviridae family, is a highly contagious human pathogen, especially in developing countries (1). MV is a negative-stranded RNA virus with a genome that comprises six structural genes encoding the nucleoprotein (N), phosphoprotein (P), matrix protein (M), fusion protein (F), hemagglutinin protein (H), and large protein (L). The N protein is the most abundant viral protein in infected cells (2). The L protein is an RNA-dependent RNA polymerase, and the P protein is a multifunctional protein that mainly functions as a cofactor of the L protein (3, 4). The M protein assists with virus assembly (5, 6), and the F and H proteins mediate membrane fusion and binding to host cellular receptors, respectively (7).

The N protein is most associated with encapsidation of viral genomic and antigenomic RNA (8), with N proteins tightly associating with the viral genome to form an N-RNA complex, which in turn forms a herringbone structure (8, 9). N proteins typically associate with every six nucleotide bases to completely cover the 15,894-nucleotide viral RNA genome (10). This tight association of N with the genome makes it resistant to RNases and small interfering RNAs (siRNAs) (11, 12). Viral transcription and replication are conducted on the N-RNA complex in association with the P and L proteins. During replication, nascent growing viral RNA molecules are immediately encapsidated by the N protein (8, 13, 14). Thus, the uncapped full-length viral antigenome is produced and serves as a template for genome replication (15).

It has been reported that the MV N protein undergoes phosphorylation, and in a previous report we identified the major phosphorylation sites of the MV N protein as S479 and S510 (16). We also showed that the major phosphorylation sites of the N protein are indirectly involved in the regulation of viral transcription by changing the phosphorylation status of the P protein (17). In the present study, we rescued recombinant MV (rMV) HL strains (18) of these N-phosphorylation sites (S479A, S510A, and S479A/S510A) using a reverse genetics system that we had previously established (19) and investigated their influence on virus growth.

MATERIALS AND METHODS

Cells, viruses, plasmids, and antibodies.

B95a cells (Epstein-Barr virus-transformed marmoset B-lymphoblastoid cells) (20, 21) were propagated in RPMI 1640 medium (Sigma) supplemented with 5% fetal bovine serum (JRH Bioscience), 2 mM l-glutamine, 100 U/ml penicillin, and 0.1 mg/ml streptomycin at 37°C/5% CO₂. HEK293 and 293SLAM cells (HEK293 cells stably expressing the signaling lymphocyte activation molecule [SLAM], which is a cellular receptor for MV) (22, 23) were maintained in Dulbecco's modified Eagle's medium (Sigma) with the same supplements as described above. The virulent MV-HL strain (18) and its recombinant viruses were propagated and titrated in B95a cells. The generation of the pCAGGS mammalian expression vectors (24) containing wild-type N (wt-N), N mutants (S479A, S510A, S479A/S510A, and 74–77A), and P and M genes from the MV-HL strain along with anti-N, -P, and -M polyclonal antibodies and anti-N protein monoclonal antibody 8G was performed as previously described (16, 25).

Rescue of rMVs.

Amplification of the MV N mutant genes (S479A, S510A, and S479A/S510A) was conducted by PCR using TaKaRa LA Taq (TaKaRa Bio) and primers F (5′-CGG CGG CCG CTT CAC GAT GGC CAC ACT TTT GAG G-3′; NotI site underlined) and R (5′-CAG GCC GGC CCT AGT CTA GAA GAT CTC T-3′; FseI site underlined). The amplicons were cloned into pCR2.1 TOPO (TOPO TA cloning kit; Invitrogen) vector according to the manufacturer's instructions. The resulting plasmids were transformed into Escherichia coli DH5α cells, grown, and then digested with NotI and FseI (TaKaRa Bio) and fragments inserted between the NotI and FseI sites of the MeV-HL full genomic plasmid (7+) (19) to replace the wt-N gene. Rescue of the recombinant viruses was conducted as previously reported (19). Briefly, vaccinia virus-infected HEK293 cells were transfected with N, P, and L protein expression plasmids and with the MV full-genome plasmid whose N protein was mutated at phosphorylation sites (pMDBI-wt, S479A, S510A, and S479A/S510A). B95a cells were added onto transfected HEK293 cells 2 days after transfection. When an adequate cytopathic effect was observed, infected cells were subjected to three freeze-thaw cycles followed by sonication. After removal of cellular debris, recombinant viruses in the supernatant were concentrated by centrifugation (25,000 rpm, 2 h, 4°C) in an Sw28 rotor (Beckman), and the viral pellet was dissolved in an appropriate volume of phosphate-buffered saline (PBS). Viral titers were expressed as the 50% tissue culture infectious dose per milliliter (TCID₅₀/ml) and were measured in B95a cells. Reverse transcription (RT) and direct sequencing using viral RNA extracted from infected cells of each recombinant virus were conducted to ensure that substitutions in the phosphorylation sites were contained in the N protein gene of the recombinant viruses.

Immunoprecipitation assay.

HEK293 cells were transfected with plasmids expressing wt-N and N mutants (S479A, S510A, S479A/S510A, and 74–77A), with or without the P plasmid. At 24 h posttransfection, 0.38 mCi/ml ³²P (Phosphorus-32 Radionuclide; PerkinElmer) or 0.06 mCi/ml ³⁵S (EasyTag Express Protein Labeling Mix; PerkinElmer) was added to the medium, and the reaction mixture was incubated at 37°C for 24 h. Cells were harvested and lysed using lysis buffer (0.5 mM EDTA–0.5% Triton X-100–PBS) supplemented with Protease Inhibitor Cocktail (BD Biosciences) and PhosSTOP Phosphatase Inhibitor Cocktail (Roche Applied Science) for 1 h at 4°C. Cell debris was removed from the lysate by centrifugation. Lysates were incubated with protein A Sepharose CL-4B (Amersham Biosciences) and anti-N or -P polyclonal antibodies for 16 h at 4°C. Beads were washed five times with PBS and suspended in 2× Laemmli's sodium dodecyl sulfate (SDS) sample buffer. Proteins were separated by SDS-polyacrylamide gel electrophoresis (PAGE), and radioactivity was detected using an Imaging Plate and FLA-5100 phosphorimager (Fujifilm). Experiments were independently performed two or three times. The principle of the Imaging Plate method is as follows: the photostimulable phosphor on the Imaging Plate stores radiation energy by the electrical change from the ground state to the excited state. The stored excitation energy is converted to chemiluminescence by laser scanning of the phosphorimager, and the data are saved in a digital format. This detection system enables quantitative analysis of phosphorylation within a linear range.

293SLAM cells (5 × 10⁵) were infected with wild-type or rMV-HL viruses. We then added ³²P (0.38 mCi/ml) and ³⁵S (0.06 mCi/ml) to the medium just after infection, and viruses were labeled until infected cells were harvested. At each time point from 16 h postinfection (hpi) to 6 days postinfection (dpi), cells were harvested and lysed. Following immunoprecipitation, SDS-PAGE and imaging were carried out as described above.

One-step and multistep growth kinetics.

We infected 293SLAM cells with wt or recombinant viruses at a multiplicity of infection (MOI) of 1.0 or 0.001, in duplicate. The infected cells and supernatant were harvested at 4, 8, 12, 16, 20, and 24 hpi (MOI = 1.0) and at 12, 24, 48, 72, 96, 120, and 144 hpi (MOI = 0.001), respectively. After three cycles of freeze-thawing and sonication of infected cells, lysates were centrifuged to remove cellular debris, and the supernatant was collected. The titer (TCID₅₀/ml) of each sample was measured using B95a cells.

Western blot analysis.

The 293SLAM cells infected with MVs were washed in PBS and lysed in buffer (0.5 mM EDTA–0.5% Triton X-100–PBS) containing Protease Inhibitor Cocktail at 8, 16, 20, and 24 hpi (MOI = 1.0) and at 2, 3, 4, 5, and 6 dpi (MOI = 0.001). The HEK293 cells transfected with expression plasmids using FuGENE 6 Transfection Reagent (Roche Applied Science) were lysed at 48 h posttransfection. Each lysate was separated by 8% or 10% SDS-PAGE and transferred onto an Immobilon-P Transfer Membrane (Millipore). Membranes were blocked with BlockACE (DS Pharma Biomedical) and then incubated with a primary antibody for 1 h at 37°C. The membrane was washed five times with PBS containing 0.05% Tween 20 and incubated with a polyclonal rabbit anti-mouse IgG conjugated to horseradish peroxidase (Dako Cytomation) (1:2,000 dilution) for 1 h at 37°C. Immunoreactive bands were visualized using ECL Western blotting detection reagents (GE Healthcare). Detection and scanning of chemiluminescence were performed with a luminescent image analyzer (LAS-1000UV mini system; Fujifilm).

Measurement of IFN-β and RANTES levels.

The 293SLAM cells were infected with wt-MV or rMVs at an MOI of 1.0. Total RNA was extracted with Isogen (Nippon Gene) at 24 hpi, and RT was conducted with a 1:1 mixture of oligo(dT) primers and random hexamers and with PrimeScript RTase (TaKaRa) according to the manufacturer's instruction. Quantitative PCR (qPCR) assays were conducted using beta interferon (IFN-β)-specific (5′-CAG CAG TTC CAG AAG GAG GA-3′ and 5′-AGC CAG GAG GTT CTC AAC AA-3′) and RANTES-specific (5′-TCA TCC TCA TTG CTA CTG CC-3′ and 5′-GGT GAC AAA GAC GAC TGC TG-3′) primers and SYBR Premix Ex Taq (TaKaRa). GAPDH (glyceraldehyde-3-phosphate dehydrogenase) was used as an internal control, and expression levels were determined with specific primers (5′-AAC ATC ATC CCT GCC TCT ACT G-3′ and 5′-GCT TCA CCA CCT TCT TGA TGT C-3′). All qPCR assays were conducted on a Rotor-Gene Q cycler (Qiagen).

Indirect immunofluorescence assays.

The HEK293 cells were transfected with expression vectors for wt-N, N protein mutants (S479A, S510A, S479A/S510A, and 74–77A), and the P protein using FuGENE 6 Transfection Reagent, according to the manufacturer's instructions. At 24 h posttransfection, cells were fixed with 3% paraformaldehyde–PBS and permeabilized with 0.5% Triton X-100–PBS. Cells were incubated with an anti-N protein polyclonal antibody followed by incubation with Alexa Fluor 488-conjugated goat anti-rabbit IgG (Invitrogen) (1:2,000) supplemented with Hoechst 33342 (Cambrex Bio Science). A confocal laser-scanning microscope (Fluoview FV1000-D system; Olympus) was used to visualize the fluorescence.

Complex formation assay.

We transfected HEK293 cells with plasmids expressing wt-N and N mutants and with P or M proteins. Cells were lysed at 48 h posttransfection, and the cell lysates of N, P, and M protein-transfected cells were mixed and incubated for 1 h at 4°C. N, P, and M protein complexes were immunoprecipitated with 8G anti-N protein monoclonal antibody. Coprecipitated N, P, and M proteins were separated by SDS-PAGE and detected by immunoblotting using anti-N, -P, and -M protein polyclonal antibodies.

Pulse-chase assay.

The 293SLAM cells infected with wt-MV or N protein rMVs (S479A, S510A, and S479A/S510A) at an MOI of 0.01 were labeled with [5,6-³H]uridine (20 μCi/ml, PerkinElmer) for 18 h between 6 and 24 hpi. The culture medium was replaced with medium containing excess UTP (Roche) (200 nM/ml) and ribavirin (250 μM) at 1 dpi to stop the incorporation of [5,6-³H]uridine. Infected cells were harvested and lysed at 0, 6, 24, and 48 h after the medium change. The released viruses were precipitated by centrifugation and lysed with infected cells. The nucleocapsid proteins in the lysate were separated by 25% to 40% sucrose gradient density centrifugation at 55,000 rpm for 1 h (Sw55Ti Rotor; Beckman). A Tri-Carb 2100TR Liquid Scintillation Analyzer (PerkinElmer) was used to measure radioactivity.

RNase protection assay.

We infected 293SLAM cells with wt-MV or rMVs (S479A, S510A, and S479A/S510A) at an MOI of 0.001. Infected cells were divided into two equal portions after 5 days of incubation. One portion was lysed with micrococcal nuclease (MNase) buffer (10 mM Tris-HCl [pH 7.5], 10 mM NaCl, 5 mM MgCl₂, 2 mM CaCl₂, 1% Triton X-100, 0.5% sodium deoxycholate, BaculoGold Protease Inhibitor Cocktail [BD Biosciences], and 100 U/ml DNase) and treated with TRIzol LS (Invitrogen) to extract RNA. The other portion was lysed with MNase buffer supplemented with 100 U/ml MNase for 1 h at 37°C followed by RNA extraction with TRIzol LS. From the extracted RNAs, the viral genomic RNA was quantified by RT-PCR to calculate the protection rate from MNase. The RT reaction was conducted with MV genome RT primer (5′-GGC CGT CAT GGT GGC GAA TCC TAC CTC GAC CTG TTG TTG AAT GAA GAG TTA GAA GAG TTC AC-3′) and PrimeScript RTase (TaKaRa), according to the manufacturer's instructions. Subsequently, qPCR assays were conducted with forward (5′-GGC CGT CAT GGT GGC GAA T-3′) and reverse (5′-TGT TCC ACG AAG ATC CT-3′) primers and SYBR Premix Ex Taq (TaKaRa) on a Rotor-Gene Q cycler. Genomic RNA quantities from MNase-treated and untreated samples were compared to calculate the protection rate.

RESULTS

N-P protein interaction alters the phosphorylation pattern of the N protein.

In our previous report, it was suggested that the interaction between the N and P proteins affects the N protein phosphorylation status (17). We first compared the phosphorylation statuses of wt-N and its phosphorylation site mutants (S479A, S510A, and S479A/S510A) in the presence or absence of the P protein. ³²P- or ³⁵S-labeled wt-N and N mutants with or without the P protein were immunoprecipitated with anti-N or anti-P protein antibodies, respectively. The N protein that was expressed alone showed a phosphorylation pattern different from that of the N protein that bound to the P protein (Fig. 1A and B), suggesting that the N-P protein interaction affects the phosphorylation status of the N protein. To further evaluate these changes, we rescued recombinant viruses whose N protein phosphorylation sites were substituted (rMV-S479A, rMV-S510A, and rMV-S479A/S510A) by reverse genetics and investigated the phosphorylation status of the N protein in cells infected with recombinant viruses at 2 dpi. The N protein in the rMV-infected 293SLAM cells showed the same phosphorylation pattern as the N protein in cotransfected cells (Fig. 1B and C). This implies that the N-phosphorylation of rMVs in the infected cells was also affected by the N-P protein association.

Fig 1.

Phosphorylation status of wt-N protein and phosphorylation mutants (S479A, S510A, and S479A/S510A) under various conditions. (A) HEK293 cells were transfected with pCAGGS expression vectors encoding wt or mutant N proteins. N proteins expressed alone in HEK293 cells were labeled with ³²P or ³⁵S and precipitated with an anti-N polyclonal antibody. (B) N proteins were transiently expressed with the P protein in HEK293 cells, labeled with ³²P or ³⁵S, and coimmunoprecipitated with the P protein using an anti-P polyclonal antibody and quantified. (C) 293SLAM cells were infected with wt or rMVs (rMV-S479A, rMV-S510A, and rMV-S479A/S510A) at an MOI of 0.001 and labeled with ³²P or ³⁵S. The bar graphs below each image represent relative phosphorylation levels (³²P-N/³⁵S-N) for each N protein.

The N protein expressed alone in cultured cells is imported into the nucleus, and the presence of the P protein retains the N protein in the cytoplasm (26). Therefore, we examined whether changes in N-phosphorylation through the N-P interaction depend upon the change in localization or on the binding of the N and P proteins. In our previous report, a nuclear localization signal was identified in the N protein; therefore, we generated an N protein mutant (74–77A, referred to as mN) that localized to the cytoplasm (27). We compared the phosphorylation levels of the wt-N and mN expressed alone or with the P protein. The phosphorylation level of the wt-N protein associated with the P protein was 37.4% of that of the free wt-N protein in HEK293 cells (Fig. 2A). The mN protein expressed with the P protein also had a reduced phosphorylation level (35.6%) compared with the free mN protein (Fig. 2B). This indicates that the N-phosphorylation level was directly downregulated by N-P protein binding and was not affected by the localization of the N protein.

Fig 2.

Phosphorylation of free N protein (N⁰) and P-associated N protein. (A) HEK293 cells transiently expressing wt-N protein, with or without the P protein, were labeled with ³²P or ³⁵S and immunoprecipitated (IP) with anti-N or anti-P antibodies (pAb) as appropriate. The bar graph shows the relative phosphorylation levels (³²P-N/³⁵S-N) of the N protein. Error bars indicate standard deviations. **, P < 0.01. The location of the wt-N protein was shown by immunofluorescence using the 8G anti-N monoclonal antibody (25), with nuclei stained with Hoechst 33342. (B) The same experiment was conducted using an N protein mutant (mN, 74–77A).

One- and multistep growth kinetics of rMVs.

To investigate the influence of N-phosphorylation on virus production, we compared the one- and multistep virus growth kinetics of wt-MV and rMVs (S479A, S510A, and S479A/S510A). Released and cell-associated viruses were harvested at 4 to 24 hpi (one-step growth) and 0.5 to 6 dpi (multistep growth) and their titers determined. In the one-step analysis, rMV-S479A/S510A rapidly replicated and reached the peak titer at 20 hpi, while the wt-MV showed slower growth kinetics until 20 hpi. The titer of wt-MV at 24 hpi was the same as or slightly higher than that of rMV-S479A/S510A. In one-step growth, rMV-S479A and rMV-A510A showed moderate replication levels; however, the peak titer for rMV-S479A was the lowest. The cell-free rMVs showed growth kinetics similar to those of the cell-associated viruses (Fig. 3A and B). Conversely, in multistep growth kinetics, rMV-S479A/S510A propagated most efficiently and showed higher peak titers. These titers remained high during the plateau phase of viral growth (around 3 to 6 dpi). The level of growth of rMV-S479A was lowest, while other viruses showed moderate growth levels (Fig. 3C and D). Although rMV-S479A replicated faster than wt-MV in one-step growth kinetics, it could not maintain a higher titer than the other viruses during the latter phases of multistep growth. The different characteristics in the early and late phases of growth suggested that N-phosphorylation has multiple roles in the virus life cycle.

Fig 3.

One- and multistep growth kinetics of rMVs. (A and B) 293SLAM cells were infected with wt-MV or rMVs (rMV-S479A, rMV-S510A, and rMV-S479A/S510A) at an MOI of 1.0. Cells were harvested at 4, 8, 12, 16, 20, and 24 hpi, and titrations (TCID₅₀/ml) of the cell-free virus (A) and cell-associated virus (B) were conducted. (C and D) 293SLAM cells infected with the rMVs at an MOI of 0.001 were harvested at 12 hpi, 1 dpi, 2 dpi, 3 dpi, 4 dpi, 5 dpi, and 6 dpi, and the titers of cell-free virus (C) and cell-associated virus (D) were measured.

To investigate the influences of N-phosphorylation on viral growth, we performed a time course analysis of N gene expression. Recombinant virus-infected 293SLAM cells were harvested at each time point and N proteins detected by immunoblotting. Accumulation of the N protein (Fig. 4A and B) was almost similar to the growth rates (Fig. 3) in the one-step and multistep growth analysis, suggesting that the differences in growth kinetics between wt-MV and rMVs were caused by changes in viral gene expression levels. It also suggests that N-phosphorylation is involved in the regulation of viral gene expression. It has been reported that transcription of the MV N protein induces expression of cytokines, such as RANTES, in a dose-dependent manner (28); therefore, we also evaluated whether these changes in viral gene expression affected induction of the host immune response. Induction of cytokines was significantly lower for wt-MV than for rMV-S479A/S510A and rMV-S510A (Fig. 4C and D). Infection with the rMV-S479A mutant also tended to induce higher cytokine levels than wt-MV infections. Therefore, it can be seen that upregulation of viral gene expression during the early phase of rMV infection was accompanied by high levels of cytokine induction.

Fig 4.

Viral N protein accumulation of wt-MV and rMVs and their phosphorylation statuses. (A) 293SLAM cells were infected with wt-MV and rMVs (rMV-S479A, rMV-S510A, and rMV-S479A/S510A) and lysed at 8, 16, 20, and 24 hpi (MOI = 1.0) or at 2, 3, 4, 5, and 6 dpi (MOI = 0.001). Lysates were subjected to 10% SDS-PAGE and immunoblotting with an anti-N polyclonal antibody. (B) Relative expression levels (nucleoprotein/GAPDH) of wt-N and phosphorylation mutants of the N protein. (C and D) IFN-β induction (C) and RANTES induction (D) by wt-MV and rMVs. *, P < 0.05; **, P < 0.01. (E) Relative phosphorylation levels of wt-N and N mutants in 293SLAM cells at each time point. (F and G) Correlation between phosphorylation level of the N protein and plateau-phase viral titers (3 to 6 dpi) during multistep growth kinetics (cell-free [G] and cell-associated [F] virus). R², coefficient of determination; P, P value.

To further evaluate the influence of N-phosphorylation on virus growth, we investigated the relative phosphorylation levels of the N protein at each time point. From 16 hpi to 6 dpi, the relative phosphorylation level of the N protein was constant in the 293SLAM cells (Fig. 4E). During the multistep growth analysis, viral titers at the plateau phase inversely correlated with the phosphorylation level of the N protein. Titers for the cell-free virus showed a high inverse correlation with N-phosphorylation intensity, with a coefficient of determination (R²) of 0.82 (P < 0.001) at 3 to 6 dpi (Fig. 4F). The R² for the cell-associated virus was 0.69 (P < 0.001) and was also significantly inversely correlated (Fig. 4G). These regression curves showed exponential trend lines. The intensity of N-phosphorylation seems to have a strong influence on the plateau-phase titer during multistep growth. Intensely phosphorylated N protein caused a reduction in the viral titer during the latter phase of infection. Conversely, N-phosphorylation intensity and early-phase viral titers did not show a significant correlation, suggesting that the difference in virus growth levels during the early phase of infection was not caused by different phosphorylation intensities of the N-protein. Taking the results together, N-phosphorylation status affects virus gene expression and N-phosphorylation intensity affects the titers selectively during the plateau phase of multistep growth kinetics.

Nucleocytoplasmic trafficking and N-M protein interaction are independent of N-phosphorylation.

The phosphorylation sites of the N protein are within the C-terminal tail region, and therefore we focused on the influence of N protein C-terminal functions, such as nucleocytoplasmic transport (27) and binding to the P protein (29) and M protein (30). Our previous research identified a nuclear export signal and a nuclear localizing signal for the MV N protein. These signal sequences were prerequisites for nucleocytoplasmic transport (27). The N protein is imported into the nucleus; however, its trafficking mechanism(s) and significance have not been fully clarified. Transiently expressed wt-N and its phosphorylation mutants in HEK293 cells were mainly localized in nuclear foci, with no significant difference in localization observed (Fig. 5A). This suggests that nuclear trafficking of the N protein is not under the regulation of N-phosphorylation.

Fig 5.

Nucleocytoplasmic trafficking and N-M protein interaction are not regulated by N-phosphorylation. (A) HEK293 cells were transfected with vectors encoding wt or N protein mutants. Locations of N proteins were visualized using an anti-N antibody. Nuclei were stained with Hoechst 33342. (B) Transiently expressed N protein and its phosphorylation mutants (S479A, S510A, and S479A/S510A) complexed with the M protein in HEK293 cells were coimmunoprecipitated with an anti-M polyclonal antibody. Precipitated N and M proteins were detected with anti-N and anti-M polyclonal antibodies, respectively. (C) HEK293 cells were transfected with plasmids encoding the N, P, and M proteins. Expressed proteins were precipitated with anti-N monoclonal antibody and detected by immunoblotting using anti-N, -P, and -M polyclonal antibodies.

It is apparent that the N-P protein interaction is not affected by the phosphorylation status of the N protein (Fig. 1B). The M protein also associates with the N protein tail region and assists in virus assembly (30). N protein mutants and the M protein were transiently coexpressed in HEK293 cells, and a coimmunoprecipitation assay was conducted. As a result, no significant changes in N-M association were observed among N protein mutants (Fig. 5B). Moreover, in the presence of the P protein, there was also no difference in the levels of coimmunoprecipitation of the wt-N and the N mutants or the P and M proteins (Fig. 5C). To investigate the influence of other viral proteins on N, P, and M protein binding, we conducted immunoprecipitation experiments using wt-MV- or rMV-infected cells. As with the transcription experiment, no differences were observed in N, P, and M protein interaction (data not shown). Taken together, these results indicate that nuclear trafficking of the N protein, N-P interactions, and N-M interactions are not regulated by N-phosphorylation. Alterations in growth of the rMVs were not caused by changes in these factors.

Influence of N-phosphorylation on viral RNA stability.

The viral genome tightly associates with the N protein to form the nucleocapsid and is protected from RNases and siRNAs (8, 11, 12). We evaluated turnover of the viral genomic and antigenomic RNAs by pulse-chase experiments to determine any differences in the encapsidation ability of the N protein. Encapsidated RNAs of rMV-S479A were rapidly degraded, while those of rMV-S479A/S510A were relatively stable compared with those of wt-MV or rMV-S510A (Fig. 6A). This suggested that the stability of an encapsidated RNA was affected by the phosphorylation status of the N protein. We further analyzed the relationship between the phosphorylation of the N protein and binding to the viral genome using a nuclease protection assay. In agreement with the results of the pulse-labeling experiments, the genomic RNA protection rate of S479A/S510A was highest, reaching 52.8%. In contrast, genomic RNAs coated with wt-N and S510A provided 30.2% and 33.4% protection, respectively. The protection rate of S479A was limited to 10.3% (Fig. 6B). The viral genome protection rate significantly correlated with viral titers during the plateau phase of multistep growth kinetics (for cell-associated virus, R² = 0.66, P < 0.001; for cell-free virus, R² = 0.91, P < 0.001) (Fig. 6C). The protection rate inversely correlated with the phosphorylation level of the N protein (R² = 0.96, P < 0.001) (Fig. 6D). These correlations also showed an exponential relationship similar to that shown in Fig. 4F and G, implying that the influence of N-phosphorylation on the plateau phase of multistep growth kinetics is a result of differences in the levels of stability of the viral genomic RNA. Collectively, the results show that the phosphorylation intensity of the N protein affects the stability of the viral genomic RNA and that the viral titer during late-phase infection strongly depends upon viral genome stability.

Fig 6.

Phosphorylation intensity of the N protein inversely correlated with viral genome stability. (A) 293SLAM cells were infected with wt-MV or rMVs (S479A, S510A, and S479A/S510A) at an MOI of 0.01. Infected cells were pulse-chased with tritium. Cells and supernatants were collected at 0, 6, 24, and 48 h after labeling, and radioactivity was measured. (B) 293SLAM cells were infected with wt-MV or rMVs, and each N-RNA complex in the cell lysate was treated with MNase. Viral genomic RNAs of MNase-treated samples or untreated samples were quantified by RT-PCR. Error bars indicate standard deviations. **, P < 0.01; n.s., not significant. (C) Correlation between viral genome protection rate and viral growth of wt or mutant MVs. (D) Viral genome protection rate and the phosphorylation level of N proteins.

DISCUSSION

We have demonstrated that phosphorylation of the N protein affects viral gene expression and viral genomic RNA stability. It has been reported that the P protein has two distinct binding sites for the free N protein and the N-RNA complex (31), and the P protein is therefore thought to be a carrier that binds free N proteins for delivery to the nascent replicating viral N-RNA complex for efficient encapsidation (4, 32). It has also been reported that the P protein reduces nonspecific binding of the N protein to cellular RNA (4). Thus, the P protein seems to play important roles in supporting viral genome encapsidation; changes in the phosphorylation status of the N protein via N-P interactions may also be required for appropriate formation of the nucleocapsid. In rabies virus, it has been reported that the phosphorylation level of the N protein associated with the P protein is kept low to strengthen its affinity for viral RNA (26, 33, 34). Consistent with these findings, we demonstrated significantly less phosphorylation of the wt-N protein associated with the P protein compared with that seen with the highly phosphorylated wt-N protein expressed on its own (Fig. 2). Hence, the P protein functions as a carrier of the N protein and seems to maintain the N protein at a lower phosphorylation state for effective encapsidation of the viral genome.

In the one-step growth experiments using rMVs, the phosphorylation of the N protein affected viral growth kinetics. The wt-MV showed delayed growth kinetics compared with the rMVs (Fig. 3A and B). The viral titer and N protein level for the wt-MV were similarly attenuated during the early growth phase until 24 hpi (Fig. 3A and B and 4A and B). This delay in wt-MV growth can be attributed to downregulation of viral gene expression. These results are not consistent with those we have previously reported. We demonstrated that mutations of the N-phosphorylation sites (S479A, S510A, and S479A/S510A) caused downregulation of viral transcription in a minigenome assay (16, 17). As with our data, some past reports that investigated the function of viral protein phosphorylations have suggested these kinds of discrepancies between viral transcriptional activity and growth; mutations at constitutive phosphorylation sites of the human respiratory syncytial virus (RSV) P protein (232A/237A and 116L/117R/119L/232A/237A) did not have a significant impact on viral transcription and replication in a minigenome assay, while recombinant viruses with mutated phosphorylation sites (rA2-PP2 and rA2-PP5) showed significantly reduced growth kinetics compared with wild-type virus (35). High-turnover phosphorylation at S54 was dispensable for viral RNA synthesis and formation of N-P and P-L protein complexes in transfection experiments (36), whereas an absence of S54 phosphorylation caused attenuation of viral titers because the S54 phosphorylation is involved in the viral uncoating step (37). Thus, in functional analysis of viral protein phosphorylation, since it affects various steps of viral life cycle, viral infection experiments sometime show different results from transfection experiments.

Past reports have demonstrated that strategic regulation of RNA synthesis is required for effective evasion of the host defense system (38, 39, 40, 41). Likewise, we hypothesized that wt-MV strategically downregulated viral gene expression to evade the host defense system during the early phases of infection. It has been reported that transcription of the MV N gene causes cytokine induction in a dose- and time-dependent manner (28). Additionally, inhibition of viral transcription by limited UV irradiation or ribavirin treatment reduced MV-induced expression of cytokines, such as RANTES (CCL5) (28, 42), which is thought to play an important role in response to virus infection by recruiting monocytes, macrophages, and virus-specific memory T cells (41, 43, 44, 45, 46). It was also reported that uncontrolled rapid viral RNA synthesis induces cell death and expression of cytokines such as beta interferon and interleukin-6 (39). Thus, a negative regulatory system of viral RNA production has been reported in some viruses (39, 47). We demonstrated that attenuation of wt-MV viral gene expression activity during early growth phases contributed to reduced host immune responses (Fig. 4C and D). This feature of wt-MV could be advantageous in the context of real virus infections.

On the other hand, the differences in the plateau phase of multistep growth kinetics were inversely correlated with the phosphorylation intensity of the N protein (Fig. 3C and D and 4E). Reductions in the phosphorylation level of the N protein enhanced stability of encapsidated viral genomic RNA (Fig. 6A and B). These data indicate that virus growth in the plateau phase is affected by the stability of the encapsidated viral genomic RNA, which is in contrast to the early phases of infection. It has been previously reported that dephosphorylation of the N protein increases viral RNA encapsidation of the N protein in rabies virus (48). Moreover, removal of the C-terminal tail of the N protein, which contains major phosphorylation sites, makes the nucleocapsid more rigid in MV (49). These data suggest that the negatively charged N protein and the viral genomic RNA repel each other and that this electrostatic repulsion weakens encapsidation of the viral genome. Therefore, excessive phosphorylation of the N protein interferes with the rapid assembly and/or strong association of the N-RNA complex. During one-step growth kinetics, the total phosphorylation intensity of the N protein and viral genome stability do not have a strong impact on virus production (Fig. 3A and B). During an MV infection, the genomic RNA mainly replicates at around 12 to 24 hpi and accumulates exponentially during this phase (50). Because an adequate amount of genomic RNA was supplied during this phase, viral RNA stability does not have a strong influence during the early phases of infection. In contrast, viral genome replication is restricted to the latter phases of infection. Hence, viral genome stability has a strong selective effect on the late phase of multistep growth kinetics.

In the present study, we demonstrated that N-phosphorylation affected viral growth by altering viral gene expression and genomic RNA stability. However, we also observed that nucleocytoplasmic transport of the N protein (Fig. 5A), N-P interactions (Fig. 1B and 5C), and N-M interactions (Fig. 5B and C) remained unaltered. During the early phases of infection, proper phosphorylation of the N protein controls viral gene expression, and this regulation was required for efficient evasion of the host immune response. The phosphorylation intensity of the N protein at plateau growth phases regulates viral genome stability, and these data provide new insights into the machinery regulating postreplication viral RNAs. The N protein may regulate its own phosphorylation (S479 and S510), thereby optimizing its function at each step of the viral life cycle for efficient viral propagation. Further analysis of the phosphorylation and dephosphorylation mechanisms of the N protein would help our understanding of the biological significance of N protein phosphorylation regulation.