m6A modification is associated with BPIV3 replication in infected cells

When paramyxoviruses infect cells, viral RNA (vRNA) and viral mRNAs are synthesized in inclusion bodies that serve as replication complexes [15]. To investigate whether vRNA undergoes m6A modification within these structures, we infected HeLa cells with recombinant bovine parainfluenza virus type 3 expressing enhanced green fluorescent protein (rBPIV3-EGFP) and analyzed fixed cells at 48 h post-infection (hpi). Co-immunostaining for double-stranded RNA (dsRNA), an intermediate in viral RNA synthesis, and the viral N protein revealed colocalization within cytoplasmic inclusion bodies, which is a characteristic feature of viral replication complexes (Fig 1A). When we co-stained with antibodies against m6A and the viral N protein, we observed that m6A exhibited a diffuse distribution in the cytoplasm of mock-infected cells, whereas, in infected cells, m6A distinctly colocalized with the N protein in the punctate dot structures representing inclusion bodies (Fig 1B). These observations were further quantified using Manders’ colocalization coefficients, which confirmed significant colocalization between N protein and dsRNA (Fig 1C) as well as between N protein and m6A (Fig 1D). To further validate these findings, we performed proximity ligation assay (PLA) using antibodies against m6A and N protein, followed by dsRNA immunostaining. PLA signals exhibited the strong colocalization with dsRNA-positive inclusion bodies, and quantitative analysis revealed a significantly higher number of PLA signals in infected cells compared with uninfected controls (S1A and S1B Fig). Moreover, Manders’ colocalization analysis demonstrated robust overlap between PLA signals and dsRNA (S1C Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. m6A modification of viral RNA during BPIV3 infection and enhancement of viral replication by METTL3.

HeLa cells were infected with rBPIV3-EGFP at an multiplicity of infection (MOI) of 1 and fixed at 48 h post-infection (hpi). The cells were costained with anti-BPIV3-N antibody (Ab) and anti-double-strand RNA (dsRNA) Ab (A) or m6A Ab (B). Representative images from three independent experiments are shown. White boxes in each panel indicate regions shown as magnified views in the corresponding lower-right insets. Cell nuclei were stained with DAPI. Manders’ colocalization coefficient for N with EGFP and with dsRNA (C), or with EGFP and with m6A (D), was calculated using the Coloc 2 plugin in Fiji/ImageJ software, based on analyses of eight cells for panel C and nine cells for panel D. The 293T cells were transfected with empty vector, METTL3, METTL14, ALKBH5, or FTO expression plasmids as m6A factors, and at 24 h post-transfection, the cells were infected with rBPIV3-EGFP at an MOI of 1. At 72 hpi, the culture supernatant was harvested, and the viral titer was determined by the TCID₅₀ method (E, F). METTL3-KD, METTL14-KD, or control-sh HeLa cells were infected with rBPIV3-EGFP at an MOI of 1. At 48 hpi, supernatants from infected cells were harvested, and viral titers in each cell line were determined using the TCID₅₀ method (G, H). HeLa cells were pre-treated with 7.5 µM STM2457 24 h before infection, and then the cells were infected with rBPIV3-EGFP at an MOI of 1 in the presence or absence (STM2457 + wash) of STM2457. At 72 hpi, supernatants from infected cells were harvested, and viral titers were determined in each cell line using the TCID₅₀ method (I, J). CelliterGlo was used to measure the cell viability of the STM2457-treated-infected cells (K). RLU: relative luminescence unit. Data are representative of three independent experiments (n = 3). Asterisks indicate significance (*p < 0.05); ns, not significant. Data represent the mean ± SD from n = 3 independent experiments. Schematic illustrations in panels E, G, and I were created with MS PowerPoint.

https://doi.org/10.1371/journal.ppat.1013755.g001

Given that m6A modification is primarily catalyzed by the methyltransferase METTL3, we investigated its role in viral replication. Overexpression of METTL3 prior to rBPIV3-EGFP infection significantly enhanced viral replication (Fig 1E and 1F). We also assessed the contribution of other m6A regulatory enzymes. Overexpression of the methyltransferase METTL14 similarly enhanced viral replication, whereas overexpression of the demethylases ALKBH5 and FTO showed no significant effect relative to empty vector controls (Fig 1E and 1F). Western blot analysis confirmed that the changes in viral replication correlated with alterations in N protein expression levels (S2A Fig). Conversely, knockdown of METTL3 and METTL14 using a specific shRNA markedly reduced viral replication (Fig 1G and 1H). Western blotting confirmed efficient reduction of endogenous METTL3 and METTL14 levels in shRNA-treated cells (S2B Fig). Furthermore, pretreatment of cells with STM2457, a selective METTL3 inhibitor, significantly attenuated rBPIV3-EGFP replication, without affecting cell viability (Fig 1I, 1J, and 1K). Collectively, these findings demonstrate that m6A modification occurs within viral inclusion bodies, and that METTL3 plays a critical role in BPIV3 replication.

BPIV3 infection causes recruitment of nuclear METTL3 to cytoplasmic inclusion bodies

METTL3 is predominantly localized in the nucleus, where it catalyzes m6A modification of host mRNA [27] while paramyxovirus RNA synthesis occurs exclusively in cytoplasmic inclusion bodies [20]. For paramyxovirus transcripts to undergo m6A modification, METTL3 must relocate to the cytoplasmic compartment. To test this hypothesis, we transfected the cells with METTL3 and subsequently infected them with BPIV3 to examine METTL3 localization. While METTL3 was exclusively nuclear in mock-infected cells, BPIV3 infection induced a striking translocation of METTL3 to the cytoplasm, where it was specifically recruited to inclusion bodies containing accumulated viral N proteins (Fig 2A). To further demonstrate the reproducibility of these observations, we also acquired low-magnification images (S3 Fig), which provide an overview of multiple A549 cells in the same field and clearly show that METTL3 localization differs between infected and uninfected cells. Quantification of METTL3 distribution confirmed the METTL3 localization shift from the nucleus to the cytoplasm upon infection (Fig 2B). Moreover, cytoplasmic FLAG-METTL3 in infected cells co-localized with dsRNA in the inclusion bodies (Fig 2C), suggesting its direct association with sites of viral RNA synthesis. We further examined the localization pattern of endogenous METTL3 upon BPIV3 infection and found that it predominantly resided within the nucleus in uninfected cells, whereas virus infection redirected endogenous METTL3 to cytoplasmic inclusion bodies (Fig 2D). To quantitatively assess these observations, we calculated Manders’ colocalization coefficient. The analysis demonstrated significant colocalization of FLAG-METTL3 with dsRNA (Fig 2E) as well as endogenous METTL3 with dsRNA (Fig 2F) within the inclusion bodies of infected cells. To provide biochemical support for these imaging data, we performed immunoprecipitation of viral RNA using an anti-dsRNA antibody from UV-crosslinked infected cell lysates. This analysis revealed that both viral N protein and METTL3 were co-precipitated with dsRNA (S4 Fig), further confirming that METTL3 is recruited to viral replication complexes containing dsRNA. This biochemical evidence complements our microscopy findings and strengthens the conclusion that METTL3 is directly associated with viral RNA-containing inclusion bodies. Collectively, these results demonstrate that METTL3 is actively recruited to viral replication complexes where viral transcripts are synthesized, positioning this m6A writer enzyme optimally for modification of nascent viral RNAs.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Cytoplasmic translocation of METTL3 during BPIV3 infection and its co-localization with viral dsRNA.

HeLa cells were transfected with FLAG-METTL3 expression plasmid, and at 24 h post-transfection (hpt), the cells were infected with rBPIV3-EGFP at an MOI of 1. At 48 h post-infection (hpi), the cells were fixed and costained with anti-FLAG antibody (Ab) for METTL3 and anti-BPIV3-N Ab (A). Arrowheads indicate regions where N and METTL3 fluorescence signals colocalize within the cells. The bar graph shows the distribution of METTL3 subcellular localization. For classification-based quantification, more than 30 METTL3 single- or METTL3/N double-positive cells per condition were randomly selected and classified into two localization patterns (“nucleus only” or “cytoplasmic or nucleus”). The number of cells displaying each localization pattern was counted for each condition from three independent experiments, and the results are presented as the percentage of total cells analyzed (B). HeLa cells were transfected with FLAG-METTL3 plasmid and, 24 hpt, infected with rBPIV3-EGFP at an MOI of 1. At 48 hpi, cells were fixed and co-stained with antibodies against dsRNA and FLAG-METTL3 (C) or dsRNA and endogenous METTL3 (endo METTL3) (D). Representative images from three independent experiments are shown. Manders’ colocalization coefficient between dsRNA and FLAG-METTL3 (E) or dsRNA and endogenous METTL3 (F) fluorescence signals in panels C and D was calculated using the Coloc 2 plugin in Fiji/ImageJ software. The coefficient was determined for each cell, and results represent analyses of 6–7 individual cells per condition from three independent experiments. Bars indicate the mean ± SD. White boxes in panels C and D indicate regions shown as magnified views in the corresponding lower-right insets. Data are representative of three independent experiments (n = 3). Asterisks indicate significance (*p < 0.05); ns, not significant.

https://doi.org/10.1371/journal.ppat.1013755.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Nucleocytoplasmic translocation of METTL3 during BPIV3 infection and its temporal co-localization with viral M protein.

HeLa cells were cotransfected with METTL3 and M expression plasmid or an empty vector. At 48 h post-transfection (hpt), the cells were fixed and costained with anti-BPIV3-M antibody (Ab) and anti-FLAG Ab for METTL3 (A). HeLa cells were transfected with METTL3 expression plasmid, and at 24 hpt, the cells were infected with rBPIV3-EGFP at an MOI of 1. At 48 h post-infection (hpi), the cells were fixed and costained with anti-FLAG antibody (Ab) for METTL3 and anti-BPIV3-M Ab (B). Cell nuclei were stained with DAPI. Insets represent an enlargement of the areas indicated by a small square. Arrowheads indicate cytoplasmic colocalization of M with METTL3. Representative images from three independent experiments are shown. For classification-based quantification, more than 20 METTL3 single-positive or METTL3/M double-positive cells per condition were randomly selected and categorized into two subcellular localization patterns (“nucleus only” or “cytoplasmic or nucleus”). The number of cells in each category was counted, and the data are presented as the percentage of total cells analyzed from n = 3 independent experiments. (C and D). Asterisks indicate statistically significant differences (*p < 0.05); ns, not significant.

https://doi.org/10.1371/journal.ppat.1013755.g003

Paramyxovirus matrix protein binds to METTL3 in the nucleus and mediates its translocation to the cytoplasm

The paramyxovirus M protein contains a bipartite nuclear localization signal (NLS) comprising two basic amino acid clusters that interact with importin to facilitate nuclear trafficking. The M protein also possesses a leucine-rich nuclear export signal (NES) that mediates cytoplasmic trafficking [16,28], however, the biological significance of this nucleocytoplasmic shuttling remains elusive. To investigate whether the M protein facilitates METTL3 translocation to the cytoplasm, we co-expressed the M protein and METTL3. We found that the expression of the M protein, but not the N protein, induced a dramatic shift in METTL3 localization from the nucleus to the cytoplasm (Figs 3A and S5). To further assess the temporal dynamics of this process, we co-expressed METTL3 with M protein and collected cells at multiple time points post-transfection. METTL3 was initially retained in the nucleus, but progressively relocalized to the cytoplasm over time, with a substantial cytoplasmic accumulation observed by 48–72 hours post-transfection (S6A and S6B Fig). These results clearly demonstrate that M protein-driven nuclear export of METTL3 is a gradual, time-dependent process.

We further examined the temporal dynamics of subcellular localization of METTL3 and M proteins in BPIV3-infected cells. Immediately after the viral challenge (0 h post-infection; 0 hpi), METTL3 was exclusively localized in the nucleus. By 24 hpi, METTL3 accumulated at the nuclear membrane, colocalizing with M protein. At 48 hpi, both METTL3 and M proteins were translocated to the cytoplasm, where they maintained their colocalization (Fig 3B). Quantitative analysis confirmed that METTL3 shifted from predominantly nuclear to cytoplasmic localization when co-expressed with M protein (Fig 3C) and during the course of viral infection (Fig 3D). These findings strongly suggest that the M protein mediates METTL3 nuclear export to the cytoplasm.

To determine whether the M protein-mediated cytoplasmic translocation of METTL3 is conserved among paramyxoviruses, we tested M proteins from Sendai virus (SeV), human parainfluenza virus type 3 (hPIV3), Nipah virus (NiV), and measles virus (MeV). All the M proteins examined demonstrated the ability to relocalize METTL3 to the cytoplasm (S7 Fig), although co-expression of N and P proteins was required to achieve this effect in hPIV3 and MeV. These results suggest that the ability to translocate METTL3 is a conserved function of the paramyxovirus M proteins.

Nucleocytoplasmic trafficking of METTL3 is mediated by Exportin-1-dependent nuclear export of paramyxovirus M protein

Previous studies demonstrated that substituting two leucine residues with alanine residues in the NES (M-L106A/L107A) inhibits M protein cytoplasmic trafficking, causing nuclear retention [16,28]. These studies also identified lysine 258 (K258) as a ubiquitination site required for nuclear export, with K258R mutation similarly blocking M protein nuclear export [16,28]. To further investigate this nucleocytoplasmic shuttling mechanism, we generated rBPIV3 M-L106A/L107A and M-K258R mutants (Fig 4A) and examined their effect on METTL3 trafficking. As expected, M protein of both rBPIV3 M-L106A/L107A and K258R mutants was retained in the nucleus and failed to facilitate METTL3 translocation to the cytoplasm (Fig 4B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Dependence of METTL3 nuclear export on exportin-1-mediated nuclear export of M protein.

(A) Schematic representation of BPIV3-M point mutants. BPIV3-M contains a nuclear export signal (NES) at L106 and L107 and a ubiquitination site (Ubi site) at K258. L106 and L107, or K258 in BPIV3-M, were substituted with alanine (M-L106A/L107A) or arginine (M-K258R) residues as indicated, respectively, and the substituted residue is shown as a red letter. (B) HeLa cells were cotransfected with expression plasmids of wt-BPIV3-M, M-L106A/L107A, and K258R along with FLAG-METTL3 expression plasmid. At 48 h post-transfection (hpt), the cells were costained with anti-M antibody (Ab) and anti-FLAG Ab for METTL3. (C) HeLa cells were transfected with the METTL3 expression plasmid. At 24 hpt, the cells were infected with rBPIV3-EGFP at an MOI of 1. At 4 h post-infection (hpi), DMSO or 5 ng/mL Leptomycin B (LMB) was added, and the cells were incubated for 48 h in the presence of LMB. Then, the cells were fixed, and METTL3 and BPIV3-N were costained. (D) HeLa cells were transfected with siRNA targeting Exportin-1 (si-Exportin-1) or control siRNA (si-Control). At 48 hpt, cells were lysed and subjected to immunoblotting with anti–Exportin-1 antibody to confirm knockdown efficiency. (E) HeLa cells were cotransfected with siRNA for Exportin-1 and METTL3 expression plasmid. At 24 hpt, the cells were infected with rBPIV3-EGFP at an MOI of 1. The cells were fixed at 48 hpi and costained for BPIV3-N and METTL3. Representative images from three independent experiments are shown. For classification-based quantification, more than 20 METTL3/M double-positive cells (F), METTL3/N double-positive cells (G), or METTL3/N double-positive cells under siRNA conditions (H) per condition were randomly selected and categorized into two subcellular localization patterns (“nucleus only” or “cytoplasmic or nucleus”). The number of cells in each category was counted, and the data are presented as the percentage of total cells analyzed from n = 3 independent experiments. Asterisks indicate statistically significant differences (*p < 0.05).

https://doi.org/10.1371/journal.ppat.1013755.g004

Exportin-1 is a critical transporter that facilitates cargo protein export from the nucleus to the cytoplasm by binding to NES and mediating transport through nuclear pore complexes [29,30]. To elucidate this mechanism further, we employed two experimental approaches. First, we used Leptomycin B (LMB), which specifically binds to the NES-binding site of exportin-1, thereby disrupting the export of NES-bearing proteins. When cells were pre-treated with LMB, both the M protein and METTL3 failed to translocate to the cytoplasm, even during BPIV3 infection (Fig 4C). Additionally, exportin-1 knockdown using siRNA, confirmed by western blotting (Fig 4D), similarly prevented METTL3 migration from the nucleus to the cytoplasm (Fig 4E). Classification-based quantification further demonstrated that both NES-mutant M proteins (L106A/L107A and K258R) (Fig 4F), LMB treatment (Fig 4G), and exportin-1 knockdown (Fig 4H) significantly reduced cytoplasmic redistribution of METTL3. Collectively, these findings demonstrate that M protein mediates METTL3 translocation through an exportin-1-dependent pathway.

Next, we investigated the molecular interactions between the M protein and METTL3. Co-immunoprecipitation experiments confirmed robust binding between the M protein and METTL3 (Fig 5A). METTL3 comprises two zinc finger domains (ZF1 and ZF2) and a methyltransferase domain (MTD) [31,32]. To precisely map the interaction regions, we generated domain-deletion mutants (Fig 5B). Co-immunoprecipitation experiments with these mutants revealed that the M protein binds to a region spanning amino acids 381–400 within the methyltransferase domain, which includes the catalytically critical DPPW motif. This was particularly evident from the significantly reduced interaction observed with the 1–380 mutant lacking the DPPW motif (Fig 5B).

Next, we examined the binding region of the viral M protein to METTL3. Since M protein deletion mutants are unstable and difficult to express in living cells, we fused a series of M deletion mutants to a GST tag and expressed them in a wheat cell-free protein production system (Fig 5C). Subsequent GST pull-down analysis revealed that M protein binds to METTL3 via an amino acids 132–184 region (Fig 5C). These results indicate that the central domain of viral M protein interacts with METTL3-MTD.

METTL3 catalyzes m6A modification and enhances BPIV3-N mRNA stability

METTL3 recognizes a specific RNA sequence motif (RRACH, where R = A/G and H = U/A/C) in the substrate RNAs for m6A modification. The S-adenosyl methionine (SAM)-binding domain of METTL3 serves as a methyl group donor to add a methyl group to adenosine within this consensus motif [32]. To predict the epitranscriptomic profile of BPIV3, we employed a Sequence-based RNA Adenosine Methylation site Predictor (SRAMP) algorithm, which predicts potential m6A modification sites. Our SRAMP analysis revealed high m6A probability scores in multiple regions, with particularly strong signals in both the N and P open reading frames (S8 Fig). Since the N protein is an essential structural component for nucleocapsid formation and is the most abundantly expressed viral protein during infection, we focused our subsequent analyses on the N gene transcripts as potential targets for m6A modification.

To determine whether the N gene is indeed regulated by METTL3-mediated epitranscriptomic modification, we first examined whether METTL3 physically interacts with BPIV3 N mRNA. We performed RNA immunoprecipitation (RIP) assays using cell lysates collected from FLAG-tagged METTL3-expressing cells infected with BPIV3. Following immunoprecipitation with anti-FLAG antibody, the amount of N mRNA bound to METTL3 was quantified by RT-PCR (Fig 6A). Compared to immunoprecipitates with control IgG, N mRNA substantially co-precipitated with METTL3 (Fig 6B), indicating a specific interaction between METTL3 and viral N transcripts.

Next, we mapped precise m6A sites within the N gene ORF using methylated RNA immunoprecipitation followed by quantitative reverse transcription PCR (MeRIP-qRT-PCR). Total RNA from BPIV3-infected cells was chemically fragmented into ≤ 100-nucleotide fragments and subjected to immunoprecipitation with a monoclonal antibody against m6A. The immunoprecipitated RNA fragments were analyzed by qRT-PCR using primer sets flanking potential m6A modification sites with high SRAMP scores: N872, N1146, N1372, N1427, and N1443 in BPIV3 N mRNA (Fig 6C and 6D). Our analysis revealed significant m6A enrichment in regions containing N872, N1146, and N1372-1443, whereas N327, which had a low SRAMP score, showed minimal precipitation with the m6A antibody (Fig 6E). These results provide experimental evidence that the computationally predicted m6A sites within N gene transcripts are indeed catalytically modified.

To assess the functional significance of these m6A modifications, we introduced point mutations to disrupt m6A consensus sequences in the N gene while maintaining amino acid coding through synonymous substitutions. We constructed three non-m6A mutants: one harboring mutations at positions N872 and N1146 (N-872/1146); another with mutations at positions N1372, N1427, and N1443 (N-1372/1427/1443); and a third containing mutations at all five m6A sites (N-all) (Fig 7A). These constructs were transfected into 293T cells and analyzed for mRNA decay kinetics to determine the transcript stability following transcriptional inhibition with actinomycin D (ActD). Our results demonstrated that both the N-1372/1427/1443 and N-all mutants exhibited significantly reduced mRNA stability compared to the wild-type (wt) N transcripts. Interestingly, the N-872/1146 mutant showed stability comparable to that of wt-N (Fig 7B), suggesting that m6A modifications at positions 1372, 1427, and 1443 play a critical role in maintaining N transcript stability. These differential effects on mRNA stability were also reflected at the protein level, as assessed by western blotting (Fig 7C), with reduced N protein expression observed in mutants with decreased mRNA stability.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Molecular interaction between M and methyltransferase domain of METTL3.

(A) 293T cells were cotransfected with M expression plasmid together with FLAG-METTL3 expression plasmid. For infected cells, the cells were transfected with FLAG-METTL3 expression plasmid and infected with rBPIV3-EGFP at an MOI of 1 at 24 h post-transfection (hpt). At 72 hpt, the cells were harvested and subjected to coimmunoprecipitation with anti-FLAG antibody (Ab). The precipitates were analyzed by western blotting using anti-BPIV3-M Ab. (B) Schematic representation of human METTL3. ZF1 and ZF2: two zinc finger domains, MTD: methyltransferase domain, DPPW: DPPW motif. The 1-400, 1-380, 1-200, and 161-580 deletion mutants lacking each domain are shown. 293T cells were cotransfected with the deletion mutants along with BPIV3-M. At 48 hpt, the cells were harvested and subjected to coimmunoprecipitation with anti-FLAG Ab as described in the legend of Fig 5A. (C) Schematic representation of FLAG- and GST-tagged deletion mutants of the BPIV3 M protein used to map the METTL3 interaction domain. Cells were transfected with a FLAG-METTL3 expression plasmid, and cells were harvested at 48–72 h post-transfection. Deletion mutants of the M protein fused with FLAG and GST tags were synthesized using a wheat germ cell-free expression system. Each FLAG-GST-tagged M deletion mutant was incubated with glutathione–sepharose beads at 4°C, followed by the addition of lysates containing FLAG-METTL3. After incubation, the beads were extensively washed, and bound proteins were analyzed by western blotting. Both METTL3 and M deletion mutants were detected using an anti-FLAG antibody. Each experiment was performed at least twice.

https://doi.org/10.1371/journal.ppat.1013755.g005

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Binding of METTL3 to BPIV3-N mRNA and m6A modification analysis of N mRNA at SRAMP-predicted sites.

(A) A schematic diagram of the RNA immunoprecipitation (RIP) assay used to detect METTL3-associated N mRNA. (B) 293T cells were transfected with METTL3 expression plasmids. At 24 h post-transfection (hpt), the cells were infected with rBPIV3-EGFP at an MOI of 1 for 48 h. The infected cells were lysed, and the extracts were subjected to RNA immunoprecipitation assay with anti-FLAG antibody (Ab). After immunoprecipitation, cDNA of BPIV3-N mRNA captured by METTL3 was synthesized and the amount of N mRNA was quantified by qPCR. Followed by normalization to input RNA values, relative enrichment values of N mRNA were expressed relative to control IgG values using the ΔΔCt method. (C) SRAMP was used to predict the m6A site of the N gene derived from the BPIV3 genome. The vertical axis shows the m6A score and the horizontal axis shows the position number of the base of the N gene where N6-methyladenosine is present within the m6A motif. A yellow bar indicates low-scoring m6A site, N327, and red bars indicate high-scoring m6A sites, N872, N1146, N1372, N1427, and N1443. (D) Schematic representation of the MeRIP (m6A RNA immunoprecipitation) assay. (i) The m6A-containing RNA fragments were bound by an m6A-specific Ab. (ii) Viral N mRNA containing m6A modifications was fragmented. (iii) Antibody–RNA complexes were captured using protein affinity beads. (iv) RNA was eluted, reverse transcribed into cDNA, and amplified by qPCR using primers specific for the indicated N gene regions (N327, N872, N1146, and N1372–1443). (E) 293T cells were infected with rBPIV3-EGFP at an MOI of 1. At 72 h post-infection, the infected cells were harvested, and only mRNA from total RNA was purified. The m6A-modified mRNA was immunoprecipitated with m6A-specific Ab and magnetic beads. Followed by cleavage of the captured mRNA with the cleavage enzyme, the m6A-specific Ab-captured mRNA was eluted from magnetic beads and purified. cDNA was synthesized by reverse transcription reaction, and the m6A sites in N mRNA were then identified by qPCR. The relative value of each to the mRNA value of input RNA was calculated using the ΔΔCt method. Data represent the mean ± SD from n = 3 independent experiments. Asterisks represent statistically significant differences (* p < 0.05). ns; not significant.

https://doi.org/10.1371/journal.ppat.1013755.g006

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. Role of N mRNA m6A modifications in RNA stability and efficient viral replication.

(A) Schematic representation of the N gene with mutations to substitute the critical A, C, or T in the m6A motif. The m6A motif, the putative N6 adenosine, and the mutation sites are indicated by yellow squares, cyan, and red characters, respectively. (B) 293T cells were transfected with wt-N, N-872-1146, N1372-1443, and N-all mutants expression plasmids. At 48 h post-transfection (hpt), the cells were treated with 5 μg/mL ActD for 9 h. After drug treatment, the total RNA was extracted, and cDNA of N mRNA was synthesized by reverse transcription reaction. The amount of N mRNA was quantified by qRT-PCR. For quantification, followed by normalization for the N mRNA values by the β-actin values, the stability of each N mRNA was expressed as a value relative to that of 0 h of ActD treatment. (C) 293T cells were transfected with wt-N, N-872-1146, N1372-1443, and N-all mutants expression plasmids. At 48 hpt, the cells were harvested and subjected to western blotting using anti-BPIV3-N Ab (n = 2). (D) Schematic representation of recombinant BPIV3 genome constructs used in this study. The wt BPIV3 full-length cDNA contains the N, P, M, F, HN, and L genes in the native order. The mutant BPIV3 full-length cDNA carries specific point mutations in the N gene (N-872/1146, N-1372/1427/1443, or N-all), indicated in pink. These constructs were used to generate recombinant viruses for subsequent infection experiments in cell culture. (E) Cells were infected with the wt or m6A mutant viruses at an MOI of 1. At various time points (48, 72, and 96 hpi), the culture supernatant was harvested, and the virus titers were measured by TCID₅₀ assay on MDBK cells. Data represent the mean ± SD from n = 3 independent experiments. Asterisks represent statistically significant differences (* p < 0.05). ns; not significant. Schematic illustration in panel D was created with MS PowerPoint.

https://doi.org/10.1371/journal.ppat.1013755.g007

To further validate these findings, we performed a minigenome assay using either wild-type N (wt-N) or N-1372/1427/1443 mutant expression plasmid (S9A Fig). Consistent with the reduced protein levels observed in Fig 7C, NanoLuc activity in the minigenome system was markedly decreased in the N-1372/1427/1443 mutant compared to wt-N, indicating impaired support for viral RNA synthesis.

Finally, to determine the impact of these m6A modifications on viral replication, we generated recombinant BPIV3 viruses harboring the same mutations (N-872/1146, N-1372/1427/1443, and N-all) within the viral genome sequence (Fig 7D). Consistent with our mRNA stability findings, replication of N-1372/1427/1443 mutant virus was significantly attenuated compared to that of the wt virus at 72 hpi (Fig 7E). To investigate whether the reduced viral replication observed with the N-1372/1427/1443 mutant resulted from decreased viral protein expression, we analyzed N and M protein levels in A549 and HeLa cells infected with either wild-type or N-1372/1427/1443 mutant viruses. Western blot analysis revealed comparable M protein levels between wild-type and mutant infections, whereas N protein expression was markedly reduced in cells infected with the N-1372/1427/1443 mutant (S9B Fig). These findings suggest that the impaired replication capacity of the mutant virus is closely linked to diminished N protein expression.

To test whether reduced N expression was responsible for the decreased viral titers, we performed a rescue experiment in which 293T cells were transfected with either an N expression plasmid or an empty vector prior to infection with wt or N-1372/1427/1443 mutant virus. Exogenous N supply significantly restored viral replication of the mutant virus compared to the empty vector control (S9C Fig). To investigate the molecular basis of the rescue effect, we extracted total RNA from infected cells and performed strand-specific RT-PCR followed by qPCR to quantify viral genomic and antigenomic RNA. Consistent with the observed recovery of viral titers, exogenous N protein expression significantly restored the synthesis of both viral genome (S9D Fig) and antigenome (S9E Fig) in cells infected with the N-1372/1427/1443 mutant virus compared to the empty vector control. However, neither the viral titers nor the genome/antigenome RNA levels fully returned to wild-type levels, even with exogenous N supplementation. This suggests that loss of the m6A sites within the N gene region of the genome/antigenome may also directly impair the stability of viral RNAs, in addition to reducing N protein expression.

Matrix protein attenuated host immune response by altering the m6A modification of interferon-β mRNA

While our above data demonstrated that the viral M protein facilitates m6A modification of viral transcripts by recruiting METTL3 to cytoplasmic replication complexes, we hypothesized that this relocalization might simultaneously deplete nuclear METTL3, thereby disrupting m6A modification of host mRNAs, particularly those involved in innate immune responses. To investigate the transcriptome-wide effects of M protein-induced nuclear METTL3 depletion, we transfected A549 cells with either M protein expression plasmids or empty vectors, stimulated them with polyinosinic-polycytidylic acid (poly(I:C)) to mimic viral infection, and performed RNA-sequencing (RNA-seq) analysis.

RNA-seq results revealed significant transcriptional alterations in response to M protein expression in poly(I:C)-stimulated A549 cells. Volcano plot analysis demonstrated a pronounced pattern of transcriptional suppression, with numerous genes significantly downregulated (blue dots) and relatively few genes upregulated (red dots) compared to the empty vector control (Fig 8A). This asymmetric distribution of differentially expressed genes suggested a broad suppressive effect of M protein on host gene expression.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 8. Reduction of m6A modification and mRNA expression of type I interferon genes by M expression.

A549 cells were transfected with an M expression plasmid or an empty vector and treated with 10 μg/ml poly(I:C) for 24 h. Total RNA was extracted and subjected to RNA sequencing. (A) Volcano plots illustrating differential gene expressions between cells transfected with the M protein expression plasmid (M.poly(I:C)) and those transfected with the empty vector (Empty.poly(I:C)). The x-axis represents log₂ fold changes, while the y-axis represents -log₁₀ adjusted p-values (q-values). Red and blue plots represent genes upregulated and downregulated by M expression, respectively. The inset zooms in on genes with moderate fold changes and highlights downregulated genes, with an enlarged view shown on the right. Genes with significant differential expression, especially under the control of IFN-β, are labeled. (B) Top 10 canonical pathways determined by IPA. The orange or blue nodes indicate predicted biological function activation or inhibition, respectively. The x-axis indicates the -log₁₀ p-values for each pathway, with the most significant terms related to interferon signaling. (C) IPA network depicting the pathways and gene interactions. Green nodes represent DEGs, and the edges indicate known interactions. Solid lines and dashed lines represent direct and indirect interactions, respectively. The degree of expression and prediction is reflected by the intensity of the color. (D) Total RNA was extracted from cells transfected with either M expression plasmid or empty vector after stimulation with poly(I:C) or DMSO addition, as prepared in A. The amount of IFN-β mRNA was quantified by qPCR using IFN-β gene-specific primer sets. (E) 293T cells were cotransfected with IFN-β + 3’UTR expression plasmid together with either M expression plasmid or empty vector. Total RNA extracted from the cells was subjected to m6A RNA immunoprecipitation using an m6A-specific antibody. IFN-β m6A modification efficiency was calculated by qPCR using primer sets designed to detect m6A sites within IFN-β mRNA, with enriched m6A-modified RNA as a template. Data represent the mean ± SD from n = 3 independent experiments (except for RNA-seq, which was performed once as a screening experiment). Asterisks represent statistically significant differences (* p < 0.05). ns; not significant.

https://doi.org/10.1371/journal.ppat.1013755.g008

Gene Ontology analysis of the differentially expressed genes revealed strong enrichment in interferon signaling and antiviral response pathways. The most significantly enriched functional terms were predominantly associated with interferon-mediated cellular responses (Fig 8B), indicating that M protein expression specifically impacts host antiviral defense mechanisms. To further characterize the regulatory relationships among these differentially expressed genes, we performed a pathway interaction network analysis. This analysis identified interferon-β (IFN-β) as a critical upstream regulator orchestrating the transcriptional response.

The network diagram revealed a hierarchical structure with IFN-β occupying a central node position and connecting multiple downstream effector genes involved in antiviral and immune responses (Fig 8C). This network architecture suggests that IFN-β functions as a master regulator of the transcriptional changes induced by M protein expression, controlling various aspects of the cellular immune response through its downstream targets. Given the central role of IFN-β in this regulatory network, we specifically quantified IFN-β mRNA levels in cells expressing M protein and subsequently stimulated them with poly(I:C). Consistent with our RNA-seq findings, qPCR analysis demonstrated that IFN-β mRNA expression was significantly reduced in M protein-expressing cells compared to that in the empty vector-transfected controls (Fig 8D).

To elucidate the mechanistic basis for this reduction, we examined the m6A modification status at two major m6A sites within the IFN-β gene, one located near the 3’ end of the coding sequence and the other in the 3’ untranslated region [33], using methylated RNA immunoprecipitation followed by qPCR analysis. Remarkably, the efficiency of m6A modification at both sites was substantially diminished in M protein-expressing cells (Fig 8E), which correlated with the reduced IFN-β expression observed earlier.

To further clarify the role of METTL3 in regulating IFN-β production, we first examined whether the M-K258R mutation affects the subcellular localization of METTL3. In cells infected with wild-type rBPIV3, both the M protein and METTL3 translocated from the nucleus to the cytoplasm at 48 hpi, where they colocalized. In contrast, infection with the M-K258R mutant virus failed to induce cytoplasmic translocation, and METTL3 remained restricted to the nucleus together with M-K258R (S10A Fig). Quantitative analysis confirmed this observation, demonstrating that cytoplasmic relocalization of METTL3 was abolished in M-K258R-infected cells (S10B Fig).

We next investigated the functional consequences of this altered localization on IFN-β expression. qPCR analysis revealed that IFN-β mRNA levels were significantly higher in cells infected with the M-K258R mutant compared to wild-type rBPIV3 (S10C Fig). To assess whether this increase was linked to changes in RNA modification, we performed MeRIP assays using an m6A-specific antibody. IFN-β transcripts from M-K258R-infected cells exhibited substantially higher m6A modification efficiency than those from wild-type-infected cells (S10D Fig).

Finally, to determine whether METTL3 expression itself directly influences IFN-β production, we compared METTL3-overexpressing and METTL3-knockdown cells. BPIV3 infection of METTL3-overexpressing cells led to markedly elevated IFN-β mRNA levels compared to empty vector controls (S10E Fig). Conversely, stable METTL3 knockdown significantly reduced IFN-β expression relative to control shRNA cells (S10F Fig).

Collectively, these data suggest a viral evasion strategy whereby M protein induces nuclear METTL3 depletion, resulting in impaired m6A modification of IFN-β mRNA and attenuation of the host anti-viral response.

Ethics statement

The human cell lines used in this study include commercially available resources from ATCC and RIKEN BRC. All experiments involving recombinant DNA were approved by the National Institute of Infectious Disease Committee (approval No. 2880) and the Ministry of Education, Culture, Sports, Science and Technology in Japan (approval no. 46 on April 9th, 2024) for Experiments using Recombinant DNA and Pathogens.

Cells and viruses

293T (ATCC: CRL-11268), HeLa (ATCC: CCL-2), A549 (ATCC: CCL-185), BHK/T7-9 (RIKEN BRC: RCB4942), and MDBK cells were maintained in Dulbecco’s modified Eagle medium supplemented with 10% fetal bovine serum. The MDBK cells were obtained from a collaborator; however, the original source was unclear. METTL3- and METTL14-knockdown (KD) HeLa cells were established by transfection of the gene-specific shRNA expression plasmids for METTL3 (Santa Cruz, Cat. #sc-92172-S) and METTL14 (Santa Cruz, Cat. # sc-89054-SH), and then selected with 1.5 μg/ml puromycin (InvivoGen, Cat. #ant-pr-1). Recombinant BPIV3 derived from the complete sequence of the BN-1 strain and expressing EGFP (rBPIV3-EGFP) was generated by the reverse genetics method in a previous report in our laboratory and used as a wt virus [50]

Antibodies and reagents

Rabbit anti-BPIV3-N and M polyclonal antibodies (pAbs) were generated as previously described [42]. The following antibodies were used as primary Abs: mouse anti-MeV-M monoclonal Ab (mAb) (clone E388) kindly provided by T. A. Sato, mouse anti-FLAG Ab (proteintech, Cat. #66008–7-Ig), rabbit-FLAG Ab (proteintech, Cat. #66008–4-Ig), mouse anti-HA monoclonal Ab (mAb) (proteintech, Cat. #66006–2-Ig), rabbit anti-HA tag pAb (MBL, Cat. #561), rabbit anti-CRM1 (exportin-1) pAb (proteintech, Cat. #27917–1-AP), mouse anti-double-stranded RNA Ab clone J2 (Nordic MUbio, Cat. #10010200), mouse anti-m6A mAb (Sigma, Cat. #MS1124), rabbit anti-METTL3 pAb (GeneTex, Cat. #GTX105037), rabbit anti-METTL14 pAb (Sigma-Aldrich, Cat. #HPA038002), and HRP-conjugated alpha tubulin mAb (proteintech, Cat. #HRP-66031). LMB, ActD, STM2457, and poly (I:C) were purchased from Sigma-Aldrich (Cat. #L2913), Fuji Film (Cat. #018–21264), MedChemExpress (Cat. #HY-134836), and Novus Biologicals (Cat. #NBP2–25288).

Plasmid constructions

The FLAG-tagged human METTL3 (Addgene, Cat. #53739), METTL14 (Addgene, Cat. #53740), ALKBH5 (Sino Biological, Cat. # HG24078-CF), and FTO (Sino Biological, Cat. # HG12125-NF) expression plasmids were purchased. The open reading frames of the BPIV3-M, -N, SeV-M, NiV-M, MeV-M, N, P, and hPIV3-M, -N, and -P genes were cloned into the eukaryotic expression plasmid pCAGGS or pcDNA. An HA tag was added to the N-terminus of SeV-M, NiV-M, MeV-M, and hPIV3-M. The BPIV3-M constructs with leucine substitution at residues L106 and L107 (L106A/L107A) in the NES and with lysine substitution at residues K258 (K258R) at the ubiquitination site of BPIV3-M, were generated by overlap PCR and were cloned into pCAGGS. The point mutants in NES and at the ubiquitination site of paramyxovirus M have been described previously [16]. The BPIV3-N constructs with the mutations for the disruption of m6A (non-m6A) sites at the nucleotide positions of 872 and 1146 (N872/1146), or 1372, 1427, and 1443 (N1372/1427/1443), and all positions (N-all) were also generated by overlap PCR and were cloned into pCAGGS. These M-L106A/L107A, M-K258R, and non-m6A mutations in the N gene were also inserted into the N gene in pUC57, which contained the complete genome of BPIV3 with the EGFP gene, as described previously [50]. The METTL3 genes lacking each domain were cloned into the pCAGGS vector. A BPIV3 minigenome plasmid encoding the NanoLuc luciferase reporter gene was constructed in the antigenomic orientation. The NanoLuc ORF was placed between the gene start signal derived from the BPIV3 N gene and the gene end signal from the L gene, and flanked by the BPIV3 leader and trailer sequences. The minigenome is set under the control of the T7 RNA polymerase promoter, and the transcript, expressed as a negative-sense RNA, is cleaved precisely at both ends by a hammerhead ribozyme and a hepatitis delta virus ribozyme. The construct was cloned into the pUC57 vector. The minigenome plasmid was used for BPIV3 minigenome assays in combination with the support plasmids expressing the BPIV3 N, P, and L proteins. For in vitro protein synthesis using a wheat germ cell-free expression system, a series of deletion mutants of the BPIV3 M gene were cloned into the pEU vector with N-terminal FLAG and GST tags.

Reverse genetics for the rescue of recombinant BPIV3

BPIV3 wt and mutants were generated using a reverse genetics system with viral genomic RNA expression plasmids and viral protein support plasmids. Briefly, BHK/T7-9 cells were transfected with the BPIV3 genomic expression plasmid and the support plasmids pCAGGS-BPIV3-N, -P, and -L using Lipofectamine LTX (Thermo Fisher Scientific, Cat. #15338100). At 72 hpt, the supernatant and cells were harvested and co-cultured with fresh MDBK cells. When the EGFP fluorescent signal spread throughout the cells, the supernatant was collected and stored at -80°C until use.

Virus titration

After infection with BPIV3 at a multiplicity of infection (MOI) of 1, cell culture supernatants were collected at various time points. To determine the viral titer, a 10-fold dilution series of the culture supernatant was prepared and MDBK cells seeded in 96-well plates were inoculated with the diluted viral solution. At 48–96 hpi, virus titers were calculated with TCID₅₀ (50% tissue culture infectious dose) using EGFP fluorescence as an indicator.

In vitro synthesis of recombinant M proteins using a wheat germ cell-free system

Recombinant M protein deletion mutants were synthesized using a wheat germ cell-free expression system (WEPRO7240G; CellFree Sciences) according to the manufacturer’s instructions. Plasmids encoding FLAG- and GST-tagged M protein deletion mutants cloned into the pEU vector were used as templates for in vitro transcription. A 20 µL transcription reaction mixture was prepared and incubated at 37°C for 6 h using SP6 RNA polymerase to generate capped mRNA. For translation, 20.8 µL of translation reaction mixture was prepared with WEPRO7240G and subjected to a bilayer translation reaction at 16°C for 20 h. After translation, the reaction mixture was gently mixed and collected. The synthesis of recombinant M proteins was confirmed by western blotting using an anti-FLAG Ab. The synthesized M proteins were subsequently used for a GST pull-down assay.

Western blotting and coimmunoprecipitation

293T, A549, and HeLa cells seeded in 6-well plates were infected with rBPIV3-EGFP or mutants at an MOI of 1–5 or co-transfected with the METTL3 and m6A factors expression plasmids and BPIV3-N or BPIV3-M expression plasmids using TransIT-LT1 (Mirus, Cat. #MIR2300). When the virus was superinfected, BPIV3 was infected with MOI 1–5 at 24 hpt. After 48 h of transfection, cells were harvested, lysed in RIPA buffer, and subjected to western blotting. For co-immunoprecipitation, a mouse anti-FLAG Ab was added to the cell supernatant, lysed in RIPA buffer, and rotated at 4°C for 1 h. Protein A/G beads (Santa Cruz Biotechnology, Cat. #sc-2003) were then added to the supernatant, which was rotated at 4°C for 1 h. After washing, the sample buffer was added to the beads and subjected to western blotting. GST pull-down assays were performed as follows. The cells were transfected with a FLAG-METTL3 expression plasmid and harvested at 48–72 hpt. Cells were lysed in RIPA buffer, and the clarified supernatant was collected. The recombinant FLAG-GST-tagged M protein or its deletion mutants synthesized in a wheat germ cell-free system were incubated with glutathione-sepharose beads at 4°C, followed by washing with RIPA buffer. The FLAG-METTL3 lysate was then added to the beads and incubated again at 4°C. After washing with RIPA buffer, bound proteins were subjected to western blotting. METTL3 and recombinant M proteins were detected using an anti-FLAG Ab.

Minigenome assay

293T cells were seeded in 6-well plates and transfected with 0.5 μg of BPIV3 minigenome plasmid, 0.4 μg of pCAGGS-BPIV3-N, 0.2 μg of pCAGGS-BPIV3-P, 0.4 μg of pCAGGS-BPIV3-L, and 1 μg of T7 polymerase expression plasmid using TransIT-LT1. At 48 h post-transfection, NanoLuc activity was measured using the Nano-Glo Luciferase Assay System (Promega, Cat. #N1120) on a GloMax Discovery System (Promega).

Indirect immunofluorescence

A549 and HeLa cells seeded on coverslips were co-transfected with BPIV3-N and BPIV3-M expression plasmids, along with METTL3 and METTL14 expression plasmids, using Lipofectamine LTX. For virus superinfection, BPIV3 was inoculated at an MOI of 1 at 24 hpt. At 24–48 hpi, the cells were fixed in 10% buffered formalin solution and permeabilized with 0.5% Triton-X 100. After blocking, the cells were incubated with primary Abs, followed by incubation with secondary Abs conjugated with Alexa Fluor 488, 594, and 647 (Molecular Probes). The following primary antibodies were used: mouse anti-double-strand RNA mAb, rabbit anti-METTL3 pAb, mouse anti-m6A mAb, mouse anti-FLAG mAb, rabbit anti-FLAG Ab, mouse MeV-M mAb, rabbit anti-BPIV3-N or M pAb, and mouse anti-HA tag mAb. Nuclear staining was performed with 4′,6′-diamidino-2-phenylindole (DAPI). Cells were observed with a laser scanning confocal microscope (Olympus, FV3000). Colocalization efficiency was analyzed using the ImageJ Fiji software. The Manders’ coefficients were calculated using the Coloc 2 plugin to quantify the degree of colocalization between two fluorescent signals. Analyses were performed on 6–9 infected cells per condition. To evaluate the subcellular localization of METTL3 during BPIV3 infection or M coexpression, quantitative analysis was performed based on immunofluorescence microscopy. For each condition, analysis was conducted across more than 90 randomly selected fields, focusing on METTL3-positive cells that were also positive for M or N proteins. Cells were categorized as either “nucleus only” or “cytoplasmic or nucleus” based on the distribution of METTL3 signal. The number of cells in each category was counted in three independent experiments, and the relative proportions were calculated and presented as percentages.

In situ PLA

An in-situ proximity ligation assay (PLA) was performed according to the manufacturer’s instructions (Duolink, Sigma-Aldrich). HeLa cells were infected with rBPIV3-EGFP at an MOI of 1 and fixed at 24–48 hpi, followed by permeabilization and blocking. Cells were incubated with mouse anti-m6A Ab and rabbit anti-N Ab for 1 h. After washing, species-specific PLA probes (anti-mouse PLUS and anti-rabbit MINUS) were applied. Phosphorylated connector oligonucleotides were hybridized to the PLA probes and ligated to form a circular DNA template when the probes were in close proximity (<40 nm). The template was subsequently amplified by rolling-circle amplification with DNA polymerase and detected with fluorescently labeled complementary oligonucleotides (Duolink Detection Kit Red). For dsRNA detection, mouse anti-dsRNA Ab was directly labeled using a FlexAble mouse IgG2a labeling kit (Proteintech). PLA signals were visualized by confocal microscopy, and the number of PLA foci was quantified using the Analyze Particle plug-in in ImageJ software. Colocalization of PLA signals with dsRNA or EGFP was quantified using Manders’ colocalization coefficient with the Coloc2 plugin in Fiji/ImageJ. Analyses were performed on 10 individual infected cells per condition from two independent experiments.

Drug treatments and gene knock-down by siRNA

For STM2457 drug treatment, HeLa cells were pre-treated with 7.5 μM STM2457 24 h before virus infection. The cells were cultured in the presence or absence of the drug during virus infection. Supernatants were collected at 72 hpi and TCID₅₀ was measured using MDBK. To examine the cytotoxicity of STM2457, cell viability after drug treatment was measured using CellTiter-Glo Luminescent Cell Viability Assay (Promega, Cat. #G7570). For treatment with LMB, HeLa cells were transfected with METTL3 expression plasmid. At 24 hpt, the cells were infected with BPIV3 at an MOI of 1, and 5 ng/mL of LMB was added at 4 hpi. At 48 hpi in the presence of LMB, the cells were fixed, and the subcellular localization of METTL3, and BPIV3-N was observed by confocal laser microscopy. For ActD treatment, 293T cells were transfected with BPIV3-N and non-m6A N mutant expression plasmids. At 48 hpt, the cells were treated with 5 μg/mL ActD for 9 h, and then, the cells were harvested and total RNA from the transfected cells was purified using NucleoSpin RNA (Takara Bio, Cat. #740955). For the knock-down of the Exportin-1 gene, HeLa cells were transfected with 50 nM siRNA for Exportin-1 (Santa Cruz Biotechnology, Cat. #sc-35116) and incubated for 48 h. The cells were harvested and subjected to western blotting and immunofluorescence.

RNA immunoprecipitation (RIP) assay

RIP assay was performed using the RiboCluster Profiler RIP-Assay kit (MBL, Cat. # RN1001) according to the manufacturer’s instructions. Briefly, 293T cells were transfected with METTL3 expression plasmid. At 24 hpt, the cells were infected with BPIV3 at an MOI of 1. At 48 hpi, the cells were harvested and subjected to RIP assay; METTL3-viral RNA and complexes were immunoprecipitated from cell extracts using FLAG antibody or Myc antibody as control IgG, and high-quality viral RNA was purified. After viral RNA purification, cDNA from viral mRNA was synthesized by reverse transcription reaction with oligo dT primer using PrimeScript High Fidelity RT-PCR Kit (TaKaRa Bio, Cat. #R022A). The amount of N mRNA was quantified by qPCR using THUNDERBIRD Next SYBR qPCR Mix (TOYOBO, Cat. # QPX-201). For quantification, N gene-specific primer set (For: 5’-GCCTATGCCAACCCAGAATTA -3’, Rev: 5’-CTTAACTACGAGCCCACCATAC -3’) was used to calculate relative values to input RNA using the ΔΔCt method.

Potential m6A prediction by sequence-based RNA adenosine methylation site predictor (SRAMP)

The m6A-modification sites in the N gene and other viral protein-coding genes of the BPIV3 BN-1 strain (Accession No. AB770484) were predicted via the SRAMP installed on a local computer [51]. The Full transcript mode was selected as a prediction mode. The score was evaluated by combining the binary score and the spectrum score.

m6A RNA immunoprecipitation (MeRIP)

For the analysis of m6A modification of the BPIV3-N gene, 293T cells were infected with BPIV3 at an MOI of 1. At 72 hpi, infected cells were harvested, and total RNA was purified according to the manufacturer’s instructions using NucleoSpin RNA. The 20 μg of total RNA was used to purify mRNA using OligotexdT30 < Super> mRNA Purification Kit (From Total RNA) (TaKaRa BIO, Cat. #9086). MeRIP was performed with EpiQuik CUT&RUN m6A RNA Enrichment (MeRIP) Kit (EPIGENTEK, Cat. #9086P-9018–24) according to the manufacturer’s instructions using 5 μg of purified mRNA. The m6A-modified mRNA was immunoprecipitated with m6A-specific antibodies and magnetic beads. Followed by cleavage of the captured mRNA with the Cleavage enzyme, captured m6A mRNA was eluted from magnetic beads and purified. After mRNA purification, cDNA was synthesized by reverse transcription reaction using PrimeScript RT reagent Kit (Perfect Real Time) (TaKaRa BIO, Cat. #RR037A). The m6A sites in N mRNA were then identified by qPCR using THUNDERBIRD Next SYBR qPCR Mix. Primer sets were designed to target nucleotide (nt) 872, nt 1164, nt 1372, nt 1427, and nt 1443, sites with predicted high m6A scores by SRAMP. The nt 327, a low-scoring non-m6A site, was also chosen for negative control. Each primer set for amplification of 80–120 bp amplicons was designed as follows: m6A N872, For:5’- TGCAGGTCTTGCCTCATTT -3’, Rev: 5’- CTATCAGTGCCTTGAGCCTATT 3’, m6A N1146, For: 5’- TGATGCTGAATCGCAGATGAG -3’, Rev: 5’- GGGTTTGTGGAAGGTTGTATCT -3’, m6A N1372-1443, For: 5’- AGGAGTGATACCCAGACTGAA -3’, Rev: 5’- GGTCGCTCTCTGTTTCCTCTTT -3’, non-m6A N327, For: 5’- GCCTATGCCAACCCAGAATTA -3’, Rev: 5’- CTTAACTACGAGCCCACCATAC -3’. The relative value of each to the value of the input RNA level was calculated using the ΔΔCt method.

For analysis of IFN-β m6A modification, 293T cells were transfected with IFN-β + 3’UTR expression plasmid. At 48 hpt, the cells were harvested, and total RNA was purified using the same NucleoSpin RNA kit. MeRIP was performed using the same EpiQuik CUT&RUN m6A RNA Enrichment Kit following identical procedures as described above. Primer sets were designed to flank the two m6A modification sites in IFN-β mRNA CDS as described previously [33] using the following primer set: For: 5’- GAAGGCCAAGGAGTACAGTC -3’, Rev: 5’- GGAGGTAACCTGTAAGTCTGTTAAT-3’.

N mRNA stability

293T cells were transfected with wt-N, N-872–1146, N1372-1443, and N-all mutant expression plasmids. At 48 hpt, the cells were treated with 5 μg/mL ActD for 9 h. After drug treatment, the total RNA was purified using NucleoSpin RNA and cDNA was synthesized by reverse transcription reaction using ReverTra Ace qPCR RT Master Mix with gDNA Remover (TOYOBO, Cat. #FSQ-201). The amount of N mRNA was quantified by qPCR using THUNDERBIRD Next SYBR qPCR Mix. For quantification, the N gene-specific primer set described above was used to calculate relative values to β-actin mRNA using the ΔΔCt method. Primer set for β-actin was designed as follows: For: -5’ ACCAACTGGGGACGACATGGGAGAAA -3’, Rev: 5’- TAGCACAGCCTGGATAGCAACGTA -3’.

RNA-Seq and data analysis

A549 cells were transfected with the M expression plasmid and an empty vector. At 48 hpt, the cells were treated with 10 μg/ml poly (I:C) for 24 h, and total RNA was extracted from the cells using MagExtractor -RNA- (TOYOBO, Cat. #NPK-601) and subjected to RNA sequencing (RNA-seq) and cDNA synthesis followed by qPCR of IFN-β mRNA. Total RNA was measured by Quantus Fluorometer and QuantiFluor RNA system (Promega, Cat. #E3310). RNA quality was then verified using a 5300 Fragment Analyzer System and an Agilent HS RNA Kit (Agilent Technologies, Cat. #DNF-472–0500). Subsequently, RNA libraries were constructed using the MGIEasy RNA Directional Library Prep Set (MGI Tech, Cat. #1000006386). Then, the concentration of the prepared library solutions was quantified using a Qubit 3.0 Fluorometer and the dsDNA High Sensitivity Assay Kit (Thermo Fisher Scientific, Cat. #Q32851). Furthermore, library quality was assessed with the Agilent 2100 Bioanalyzer and the High Sensitivity DNA Kit (Agilent Technologies, Cat. #5067–4626). Circularized DNA was generated from the prepared library using the MGIEasy Circularization Kit (MGI Tech, Cat. #1000005259). Afterward, DNA Nanoballs were produced using the DNB Rapid Make Reagent Kit (MGI Tech, Cat. #1000028453). Sequencing analysis was performed using DNBSEQ-G400 at 2x150 bp with the addition of Primer supplied with the High-Throughput Pair-End Sequencing Primer Kit (App-D, Cat. #1000020832). Normalization and identification of differentially expressed genes (DEGs) were performed using the R packages TCC (version 1.18.0) and DESeq (ver. 1.30.0), respectively. DEG visualization with volcano plots was performed on Prism 10. (version 10.1.1). The canonical pathway analysis and upstream analysis of DEGs were performed on the gene list from the QIAGEN Knowledgebase using Ingenuity Pathway Analysis (IPA) (Content version: 127006219, Release Date: 2024-11-17; Qiagen, Hilden, Germany).

Quantification of IFN-β mRNA by qPCR

A549 cells transfected with M expression plasmid or infected with BPIV3, and METTL3-KD HeLa cells infected with BPIV3, were harvested at 48 hours post-transfection or post-infection. Total RNA was extracted using the MagExtractor -RNA- kit (TOYOBO) according to the manufacturer’s instructions. The cDNA was synthesized from the extracted RNA using ReverTra Ace qPCR RT Master Mix with gDNA Remover. Quantitative PCR (qPCR) was then performed using THUNDERBIRD Next SYBR qPCR Mix (TOYOBO) to determine the relative levels of IFN-β mRNA. Primer sets for IFN-β quantification were as follows: For: -5’ CTTGGATTCCTACAAAGAAGCAGC-3’ and Rev: 5’- TCCTCCTTCTGGAACTGCTGCA-3’. Relative expression levels of IFN-β mRNA were calculated using the ΔΔCt method, normalized to β-actin mRNA.

Quantification of BPIV3 genome and antigenome RNA

293T cells were transfected with either an empty vector or an N protein expression plasmid. At 24 hours post-transfection, cells were infected with wt BPIV3 or the N1372/1427/1443 mutant virus at an MOI of 1. Total RNA was extracted at 48 hpi using the MagExtractor -RNA- kit (TOYOBO) according to the manufacturer’s protocol. Strand-specific primers used for cDNA synthesis were as follows: BPIV3 genome cDNA primer: 5’-AAGAGAAGAGACTTGTTTGGGAATA-3’ and BPIV3 antigenome cDNA primer: 5’-TATTCCCAAACAAGTCTCTTCTCTT-3’. RT-PCR was performed with the PrimeScript High Fidelity RT-PCR Kit (TaKaRa Bio) using these strand-specific primers to synthesize cDNA. qPCR targeting the N gene was then carried out using THUNDERBIRD Next SYBR qPCR Mix (TOYOBO) with the following primers: For: 5’-GCCTATGCCAACCCAGAATTA-3’ and Rev: 5’-CTTAACTACGAGCCCACCATAC-3’. Relative expression levels of genome and antigenome RNA were calculated using the ΔΔCt method with normalization to β-actin mRNA.

Statistical analysis

All statistical analyses were performed using GraphPad Prism 9 software (version 9.4.1). Data are presented as the mean ± standard deviation of the mean, as indicated. For comparisons between two groups, an unpaired two-tailed Student’s t-test was used. For multiple group comparisons, one-way or two-way analysis of variance (ANOVA) followed by Dunnett’s or Šídák’s multiple comparisons test was applied, depending on the experimental design. Each experiment was independently performed at least three times (biological replicates), and technical replicates were included as appropriate. A p-value < 0.05 was considered statistically significant. In DESeq, a difference with a q-value (false discovery rate, FDR) < 0.05 was considered statistically significant. In IPA, p-values were calculated using Fisher’s exact test, and values <0.05 were considered statistically significant.