Structural and molecular properties of mumps virus inclusion bodies Authors: Hiroshi Katoh, Ryuichi Kimura, Tsuyoshi Sekizuka, Kohei Matsuoka, Mika Hosogi, Yuki Kitai, Yukiko Akahori, Fumihiro Kato, Michiyo Kataoka, Hirotaka Kobayashi, Noriyo Nagata, Tadaki Suzuki, Yasuyuki Ohkawa, Shinya Oki, Makoto Takeda Published: 2024 ABSTRACT Viral RNA synthesis of mononegaviruses occurs in cytoplasmic membraneless organelles called inclusion bodies (IBs). Here, we report that IBs of mumps virus (MuV), which is the causative agent of mumps and belongs to the family Paramyxoviridae, displayed liquid organelle properties formed by liquid-liquid phase separation. Super-resolution microscopic analysis of MuV IBs demonstrated that nucleocapsid and phospho (P)-proteins formed a cage-like structure and that the viral polymerase adopted a reticular pattern and colocalized with viral RNAs. In addition, we characterized host RNAs localized in MuV IBs by a spatial transcriptome analysis, and found that RNAs containing G-quadruplex motif sequences (G4-RNAs) were concentrated. An in vitro phase separation assay showed that the G4-RNAs interacted with the P protein and enhanced condensation in P droplets. Together, our data show that MuV generates IBs with a characteristic cage-like structure and host G4-RNAs play an important role in forming MuV IBs. INTRODUCTION The order Mononegavirales, known as nonsegmented, negative-stranded RNA viruses, includes many human and animal pathogens, including the mumps virus (MuV), measles virus, Nipah virus, human respiratory syncytial virus (HRSV), human metapneumovirus (HMPV), rabies virus (RABV), and Ebola virus. MuV (Mumps orthorubulavirus) belongs to the family Paramyxoviridae and is the causative agent of mumps, which is a common childhood illness characterized by painful swelling of the parotid glands and is often accompanied by severe complications, such as orchitis, aseptic meningitis, pancreatitis, and deafness (1, 2). MuV genomic RNA encapsidated by the nucleocapsid (N) protein forms an active template for transcription and genome replication, a viral ribonucleoprotein (vRNP), with viral polymerases composed of the phospho-(P) and large (L) proteins (3). After infection, viral RNA synthesis occurs within viral factories in the cytoplasm of the infected cells, called inclusion bodies (IBs) (4). The formation of cytoplasmic IBs is a common feature of mononegavirus infections, except for the Borna disease virus (BoDV), whose viral factory forms in the nucleus and is called vSPOT (viral speckle of transcript) (5, 6). IBs of mononegavirus have been reported to have liquid organelle properties formed by liquid-liquid phase separation (LLPS) (7–12). The N and P proteins of mononegaviruses, including MuV, are the minimum essential viral components required to form IB-like structures in cells. Therefore, both proteins regulate LLPS as a structural platform for viral RNA synthesis (4, 7–9, 11, 13). LLPS has been widely observed in the formation of various membraneless organelles, such as the nucleolus, stress granules, P-bodies, Cajal bodies, and promyelocytic leukemia nuclear bodies, enabling space- and time-resolved regulation of various cellular processes, including RNA metabolism, ribosome biogenesis, gene expression, and the stress response (14). LLPS creates a spatiotemporal increase in the concentration of specific proteins and entails a localized increase in liquid viscosity (15), both of which markedly alter the reaction-diffusion system, thereby regulating cellular reactions. Moreover, the reactions can be switched on and off rapidly via the formation and dissolution of the LLPS (16). We previously identified that stress-induced heat shock protein 70 (Hsp72) is recruited to MuV IBs and that Hsp90 and the R2TP complex are involved in MuV RNA synthesis (4, 17, 18). Although the primary driving force of LLPS involves multivalent interactions among proteins and RNA molecules, the identity and function of host factors recruited to the MuV IBs remain largely unknown. Electron and confocal microscopy studies showed that MuV IBs are spherical membraneless organelles containing nucleocapsid-like structures and consist of the N, P, M, and V proteins and viral genomic RNA (4, 19). However, the precise distribution of each viral protein and RNA in these IBs remains unresolved. Recently, super-resolution microscopy was used to visualize the internal structures of viral factories of several mononegaviruses (20, 21). The vSPOT of BoDV displays porous cage-like structures composed mainly of the N protein, and the other viral proteins are localized within the vSPOT (20). In the case of HRSV, IBs are not homogeneous structures, with subcompartments called IB-associated granules (IBAGs) inside (21). The HSRV N, P, and L proteins are present throughout IBs except for the IBAGs, whereas the M2-1 protein and mRNAs are localized within IBAGs, suggesting that IBs spatially partition transcription and genome replication. Here, super-resolution microscopy was used to reveal the detailed distribution of viral proteins and RNAs in MuV IBs, showing that the N and P proteins form a reticular skeleton of IBs, within which positive- and negative-stranded viral RNA species occupy distinct regions. We also characterized host RNAs recruited to the IBs by a spatial transcriptome analysis, and in vitro LLPS assays showed that host RNAs containing G-quadruplex (G4) sequences play a role in molecular condensation in MuV IBs. RESULTS Characterization of IBs formed during MuV infection A previous study reported that MuV IBs in a persistent infection model exhibited properties of liquid organelles (22). In this study, we first examined whether MuV IBs in the early stage of lytic infection also exhibit the same properties. The shapes of IBs in MuV-infected Vero and A549 cells were observed sequentially by immunofluorescence with an anti-N antibody. At 8 hours postinfection (hpi), a limited number of spherical and small IBs corresponding to initial IBs formed during infection (Fig. 1A and fig. S1A). These structures grew with time (16 hpi) and occasionally lost their spherical features (24 hpi). In both Vero and A549 cells, viral mRNA transcription has already begun at 8 hpi, and genome replication is observed at 16 hpi (17, 23). As shown in Fig. 1A and fig. S1A, the formation of IBs progressed in parallel with the increase in viral RNA synthesis. At the later stage of infection, large inclusions were observed together with an explosion in the number of IBs of various sizes (48 hpi). Next, we observed MuV IBs by scanning electron microscopy (SEM) and a fluorescence image correlation approach (24). The IBs were not compartmentalized by membranes from the surrounding cytoplasm, and concentrated nucleocapsid-like structures were observed in regions of IBs (Fig. 1B). Next, characterization of MuV IBs by the time-lapse imaging and the fluorescence recovery after photobleaching (FRAP) was conducted at 24 hpi, at which time the release of infectious viral particles was confirmed and viral RNA synthesis peaked (17, 23). Cells expressing the N-terminal mCherry-fused P protein (mCherry-P) were infected with MuV to investigate the dynamics of MuV IBs. Live-cell imaging showed that an encounter of two inclusions yielded a fusion product, and the elongation of IBs resulted in two fission products whose new ends collapsed into a rounded surface. These properties indicate that the IBs are dynamic organelles subject to deformation, fusion, and fission (Fig. 1C, fig. S1B, and movies S1 and S2). Last, to examine the fluid nature of IBs, we performed the FRAP experiment. As shown in Fig. 1D and fig. S1C, a rapid fluorescence recovery was observed, indicating that the mCherry-P was interchanged in the IBs. These findings demonstrated that MuV IBs are liquid organelles like those in other mononegaviruses. Characterization of IBs formed during MuV infection. (A) Vero cells were infected with wild-type MuV at a multiplicity of infection (MOI) of 5.0. At 8, 16, 24, and 48 hpi, cells were fixed and stained with an anti-MuV N rabbit polyclonal antibody (PAb), and Alexa Fluor 488-conjugated goat antirabbit immunoglobulin G (IgG). 4′,6-Diamidino-2-phenylindole (DAPI) was used to stain nuclei. (B) The immunofluorescence image was superimposed on the SEM micrograph, and the IB region in the cell was visualized. (C) Vero cells transfected with pCAGGS-mCherry-P were infected with wild-type MuV at an MOI of 5.0. At 24 hpi, cells were imaged at the indicated times. Images were extracted from movie S1 and shown at 2.5-s intervals (upper row) and 5.0-s intervals (lower row). (D) Vero cells transfected with pCAGGS-mCherry-P were infected with wild-type MuV at an MOI of 5.0. At 24 hpi, the mCherry-P in living cells was photobleached, and fluorescence recovery was imaged by confocal microscopy. For the plots on the right, FRAP data were corrected for background, normalized, and represented as means from the recovery curve. Localization of viral proteins and RNAs in the IBs We used Airyscan confocal super-resolution microscopy, which achieves resolutions of ~90 nm, to characterize the organization of the IBs in detail. The N protein, detected by an anti-N antibody, was primarily detected at the periphery of the IBs in MuV-infected Vero cells (Fig. 2A). Because of the limited accessibility of epitopes in the N/P complexes within the IBs, immunostaining of N typically results in a stronger fluorescent signal at the periphery of the IBs (21, 25). In addition, the N protein formed web-like structures as well as the mCherry-P within IBs, but both proteins did not colocalize completely. Furthermore, small gaps were observed in the rim of IBs, as reported for the vSPOT of BoDV (20) (Fig. 2A, arrows and arrowheads). Localization of viral proteins and RNAs in the IBs. Vero cells transfected with pCAGGS-mCherry-P were infected with wild-type MuV at an MOI of 5.0 (A and C). Vero cells were infected with rOdate/L-mCherry at an MOI of 3.0 (B and D). [(A) and (B)] At 24 hpi, cells were fixed and stained with an anti-MuV N rabbit PAb and Alexa Fluor 488-conjugated goat antirabbit IgG. DAPI was used to stain nuclei. The arrows and arrowheads indicate the gaps in the rims of the N or P structure, respectively. [(C) and (D)] At 24 hpi, FISH was performed with specific probes to detect positive- and negative-stranded viral RNAs. The localization of the L protein in the IBs was visualized by generating a recombinant MuV, rOdate/L-mCherry, which expresses a C-terminal mCherry-fused L protein (L-mCherry; fig. S2A). The expression of L-mCherry was confirmed in Vero cells infected with rOdate/L-mCherry (fig. S2B). In Vero cells, rOdate/L-mCherry replicated less efficiently than the parental rOdate, but the virus titer reached ~107 plaque-forming units/ml at 72 hpi (fig. S2C). Two localization patterns of the L protein in the IBs were observed: one ring-shaped and the other more roughly reticulate than those of the N and P proteins (Fig. 2B). The L-mCherry was detected in infected cells from 16 hpi; however, ring-shaped and web-like localization patterns were observed throughout the experimental period (16 to 24 hpi), with no time-dependent localization changes observed (fig. S3). We next performed fluorescence in situ hybridization (FISH) experiments to determine the localization of viral RNAs in the IBs. We designed two strand-specific RNA probes—the FISH probe for negative-stranded RNA targeting the L coding region that recognizes genomic RNA and that for positive-stranded RNA targeting the N coding region that recognizes both N mRNA and antigenomic RNA. Two viral RNA species were present separately in the web-like structure voids formed by the mCherry-P. These were also observed near the IBs and were thought to be viral RNAs released from IBs (Fig. 2C). Positive-stranded viral RNAs colocalized with the L-mCherry, showing a reticulate but not ring-shape pattern. In contrast, genomic RNA was detected adjacent to the L-mCherry (Fig. 2D). These observations demonstrated that positive- and negative-stranded viral RNA species are separately distributed in different areas of the IBs in MuV-infected cells. Identification of RNAs localized in the MuV IBs by spatial transcriptome analysis Photoisolation chemistry (PIC) was used to characterize the host RNA profile consisting of MuV IBs. This method uses photocaged oligodeoxynucleotides to amplify cDNAs in response to photoirradiation. Therefore, only gene expression from photoirradiated regions was evaluated (Fig. 3A). MuV-infected A549 cells were seeded onto coverslips, and in situ reverse transcription (RT) was performed with caged oligodeoxynucleotides (ODNs). The cells were then immunostained with an anti-N antibody, and the resulting immunofluorescent images were binarized at an appropriate threshold to generate “masking sheets” to irradiate IBs (designated as IB). As controls for IBs, we created masking sheets to irradiate the cytoplasm, except for IBs and nuclei (IBexc) or the cytoplasm of mock-infected A549 cells (Mock). These masking sheets were loaded into the digital mirror device (DMD) system to precisely irradiate IBs and other control regions (Fig. 3B). Lysates were pooled from five replicates, and the library was prepared in accordance with the PIC method, as described previously (26). Sequencing was performed to obtain 3 × 106 reads per sample with a portion of the library. As a result, 16, 14, and 23% of reads were assigned to genes for IB, IBexc, and Mock samples, respectively. The average gene-assigned unique molecular identifier (UMI) counts were 3.8 × 104, 8.2 × 104, and 1.4 × 105 for IB, IBexc, and Mock samples, respectively (Fig. 3C). The average numbers of genes detected in IB, IBexc, and Mock samples were 7733, 11,660, and 11,915, respectively (Fig. 3D). Two-dimensional principal components analysis (PCA) revealed that the RNA profile of IB was distinct from IBexc and Mock (Fig. 3E). Cluster analysis of the differentially expressed genes showed that specific genes were enriched in the IB sample relative to the two control samples (Fig. 3F). PIC RNA sequencing for MuV IBs. (A) Schematic overview of RNA profiling with PIC. (B) A549 cells were infected with wild-type MuV at an MOI of 5.0. At 24 hpi, cells were fixed and subjected to PIC. Images of N localization (red), masking sheet (black/white), and merged images with nuclei (green) are shown. A representative image out of five replicates is shown. [(C) and (D)] The numbers of gene-assigned UMIs (C) and detected genes (D) are shown, with the horizontal lines indicating the average of replicates. (E) Two-dimensional PCA for the RNA profiles is shown. (F) Differentially expressed genes are clustered and shown in heatmaps. Characterization of host RNAs localized in the IBs of MuV-infected cells The adequacy of each sample data was initially assessed by analyzing the amounts of MuV transcribed. The results showed that the amount of MuV genes detected varied among IB samples (fig. S4, A and B). To more accurately evaluate host RNAs localized to the IB in MuV-infected cells, we selected three samples in which a large amount of the MuV genes was detected for further analysis. There were limited variations for the IBexc samples compared with the IB samples, but three samples with fewer MuV genes were used for the analysis. Three samples were randomly selected for the Mock sample. A comparative analysis of the sample groups was performed. Volcano plots showed that 468 and 155 genes were substantially concentrated in the IBs, with 75 of those genes in common (criteria: log2 fold change ≥ 2, P < 0.01; fig. S4C). Of the 75 genes, 19 (highlighted in pink in Fig. 4A) were detected in all IB samples and not in the other control samples. The 19 genes comprised 11 protein-coding RNAs, 4 pseudogene-coding RNAs, and 4 long noncoding RNAs (fig. S5). Although no common function was found among the encoded genes, the 19 genes had a higher guanine-cytosine (GC) content than that of the human genome (40.89%) (27). Also, they contained some G4–forming sequences, as predicted by the quadruplex-forming guanine-rich sequence (QGRS) mapper (28). For example, mitochondrial ribosomal protein S34 (MRPS34) mRNA, which was the most IB-enriched mRNA [95 transcripts per million (TPM) on average], had a GC content of 64.6% and nine putative G4-forming sequences (Fig. 4, B and C, and fig. S5). BCL2 binding component 3 (BBC3) mRNA, which was the third most IB-enriched mRNA (25 TPM on average), had a GC content of 64.8% and 26 putative G4-forming sequences (figs. S5 and S6). A G4 pull-down assay using the G4-specific ligand MultiTASQ (template-assembled synthetic G-quartets) probe (29) confirmed the presence of G4 structures in both mRNAs (fig. S7). Furthermore, the signal of BG4, a G4-specific antibody, was detected within the IBs of MuV-infected A549 cells (Fig. 4D). These data indicated that G4-containing RNAs (G4-RNAs), such as MRPS34 and BBC3 mRNAs, were recruited and concentrated in MuV IBs. Host RNAs enriched in MuV IBs. (A) List of 75 host RNAs enriched in MuV IBs. (B) The sequence of MRPS34 mRNA. G4 forming sequences predicted by the QGRS mapper are highlighted in yellow. (C) G4 forming sequences in MRPS34 mRNA and their G-scores predicted by the QGRS mapper are listed. (D) A549 cells were infected with wild-type MuV at an MOI of 5.0. At 24 hpi, cells were fixed and stained with an anti-DNA/RNA G-quadruplex mouse monoclonal antibody and an anti-MuV N rabbit PAb, and Alexa Fluor 488-conjugated goat antimouse IgG and Alexa Fluor 594-conjugated goat antirabbit IgG. DAPI was used to stain nuclei. Enhancement of the molecular concentration in P droplets by G4-RNA G4-RNAs facilitate the formation of various membraneless organelles, such as P-bodies, stress granules, paraspeckles, and their functions (30). We first investigated whether G4-RNA was involved in viral RNA synthesis in vivo. Knockdown of MRPS34 mRNA in A549 cells using small interfering RNA (siRNA) did not affect viral production (fig. S8). We next examined the role of MRPS34 and BBC3 RNAs in droplet formation in vitro. MuV N and P proteins (and those of other mononegaviruses) are the minimum requirements for forming the IB-like structure in eukaryotic cells (4). For using in vitro purified protein systems, the N and P proteins of MeV and HRSV are required for LLPS, whereas only the P protein of BoDV, HMPV, and RABV is required for forming liquid droplets. We first examined the minimum requirements of MuV proteins for phase separation. For this, N-terminal His-tagged N (His-N) and C-terminal His-tagged P (P-His) recombinant proteins were expressed in bacteria and purified (fig. S9A). An N-terminal AcGFP-fused and C-terminal His-tagged P protein (AcGFP-P-His) were also prepared. The absorbance at 260 nm (A260)/A280 ratios were <0.39, indicating that these proteins were nucleic acid-free. The in vitro LLPS assay revealed that only the P protein formed liquid droplets under a physiological salt concentration (150 mM NaCl) but not a high salt concentration (500 mM NaCl) (fig. S9, B and C). In contrast, the N protein formed aggregate-like structures but not liquid droplets, even at a physiological salt concentration (fig. S9B). The crowding agent polyethylene glycol, molecular weight 8000 (PEG-8000), which mimics cytosolic crowding conditions, was required for phase separation by the P protein (fig. S9C). Droplet formation depended on the P protein concentration (fig. S9D). In FRAP experiments, the fluorescence from fluorescent protein–fused P droplets was rapidly recovered after photobleaching (fig. S9E). These data demonstrated that the P protein alone undergoes LLPS in vitro in crowded environments. Last, we evaluated whether G4-RNAs influence P phase separation. Both MRPS34 and BBC3 RNAs were taken up in the P droplets, and the fluorescent signal was significantly enhanced in droplets containing both G4-RNAs (MRPS34 and BBC3 RNAs) than in droplets without RNA, even when the same concentration of AcGFP-P protein was added (Fig. 5, A and B, and fig. S10, A and B). To further investigate whether the G4 structures contributed to the condensation of the P protein in droplets, we synthesized both MRPS34 and BBC3 RNAs in which all the guanines presumed important for the G4 structure in the QGRS (blue-colored and underlined in Fig. 4C and fig. S6B) [MRPS34_G4(−) and BBC3_G4(−), respectively] were replaced by cytosines, and the in vitro LLPS assay was performed. As shown in Fig. 5 (A and B) and fig. S10 (A and B), both the MRPS34_G4(−) and BBC3_G4(−) RNAs and P protein incorporated into the droplets decreased significantly. The electrophoretic mobility shift assay (EMSA) results showed that the MuV P protein interacted directly with MRPS34 and BBC3 RNAs and preferentially bound the G4-containing MRPS34 and BBC3 RNAs compared with non-G4 RNA (Fig. 5C and fig. S10C). These observations indicated that G4-RNAs boost molecular condensation in P droplets interacting with proteins (Fig. 5D). Enhanced condensation of the P droplets by G4-RNA. (A) DIC and fluorescence images of P droplets in the presence of Cy3-labeled RNAs. (B) The mean fluorescence intensity of AcGFP-P in the P droplets (n = 100). Statistical significance was analyzed using unpaired Student’s t tests. (C) The indicated concentration of the P-His protein was mixed with either MRPS34 or MRPS34_4G(−) RNA and subjected to EMSA. (D) Schematic showing enhanced condensation of the P droplets by G4-RNAs. DISCUSSION One of the characteristics of mononegavirus infections is the formation of IBs. IBs are the site of viral RNA synthesis and have recently been shown to be liquid organelles formed by LLPS for some mononegaviruses (7–12). In this study, we confirmed that MuV IBs are liquid organelles based on three lines of evidence. First, the IBs were not surrounded by membranes and adopted spherical shapes during the early phase of infection. Second, the IBs underwent fusion and relaxation to form larger IBs. Third, the IB exhibited a dynamic exchange of components. Intrinsically disordered regions (IDRs) in proteins are well documented to confer multivalent weak protein-protein interactions and are critical for LLPS (14). The MuV N protein is divided into two domains, the N-terminal assembly domain and the C-terminal IDR region (N-tail) (31), whereas the MuV P protein consists of an N-terminal IDR region, middle oligomerization domain, and C-terminal region (32). Coexpressing both proteins with IDRs is sufficient to induce the assembly of IB-like structures in mammalian cells (4). In contrast, the MuV P protein alone was shown in this study to phase separate in an in vitro system, similar to BoDV, HMPV, and RABV (10, 33, 34). These findings suggest that the MuV P protein functions as a “scaffold” protein, which induces LLPS, whereas the MuV N protein is a “client” protein, which cannot undergo LLPS and is recruited into the scaffold to form liquid organelles. The difference in results between the cellular and in vitro systems suggests that cellular factors may inhibit the formation of P droplets, and the N protein may assist phase separation in crowded cellular conditions. Because there is still divergence between the cellular and in vitro systems, further development of experimental systems that more closely reflect cellular conditions is needed to understand the molecular mechanisms of MuV IB formation. Various spatial transcriptomic technologies have been developed (35) and used to survey the transcriptome in membraneless organelles such as stress granules and nuclear speckles (36–38). In this study, we performed a spatial transcriptome analysis using PIC to identify host RNAs recruited to MuV IBs. PIC enables the isolation of transcriptome information from locally defined areas of cells, including intracellular organelles, by photoirradiation (26). We irradiated N-positive droplets above a specific size with ultraviolet and obtained transcript data from MuV IBs. Viral mRNAs were also detected in IBexc samples, which may have occurred because of detection in very small droplets or because these mRNAs were being used for translation. Although it is necessary to consider that RNA without polyA cannot be detected because oligo dT RT primers were used, many IB specific host RNAs were identified by this method. Therefore, this PIC-based spatial transcriptome analysis should enable the identification of host and viral RNAs in viral factories of various viruses. Nucleic acids (DNA and RNA) are closely related to LLPS, and their structure and electronegativity provide platforms for forming phase-separated condensates. Specific RNAs have been reported to be incorporated into droplets of mononegaviruses through the interaction with viral N and/or P proteins and to influence the properties, such as shape and condensation, of the droplets (33, 34, 39). However, no studies have examined what RNA species are involved in forming IBs in infected cells. Our spatial transcriptome analysis found that host G4-RNAs were concentrated in MuV IBs and that G4-RNA enhanced the molecular condensation of P protein droplets in vitro. However, direct evidence for the contribution of G4-RNA in forming IBs in MuV-infected cells could not be clarified. Silencing of MRPS34 mRNA using siRNA did not affect viral production, suggesting that MRPS34 mRNA is only one of the G4-RNAs and other G4-RNAs compensate for its function even after MRPS34 mRNA was reduced. G4-RNA is a secondary structure of RNA that forms in guanine-rich sequences and functions as the scaffold for LLPS because of its multivalent interactions with RNA binding proteins (40). In contrast, G4-RNAs can also be involved in pathological LLPS by forming aggregates and triggering the phase transition of liquid droplets into aggregative droplet. Although the detailed function of G4-RNA in forming IBs in MuV-infected cells requires further investigation, G4-RNA may play an important role in concentrating viral and host factors necessary for efficient viral RNA synthesis because G4-RNA enriched the P protein in droplets through G4-RNA-P protein interactions. In IBs, viral polymerase synthesizes two types of viral RNAs (mRNA and genomic RNA). During the early stages of HRSV infection, a structural reorganization of polymerase proteins within the vRNP complex was observed, indicating a potential functional switch of the viral polymerase from primary transcription to genome replication (41). However, the spatial and temporal regulation of these two processes remains to be fully elucidated. In this study, we found that positive- and negative-stranded viral RNA species occupy distinct regions within the IBs of MuV-infected cells. Given that the synthesis of mRNA far exceeds that of antigenomic RNA, the FISH signal for positive-stranded viral RNA may be primarily indicative of N mRNA. However, it is also important to note that viral mRNA is certainly present outside of IBs during the early stages of infection, and the relative proportion within IBs has not yet been determined. The distinct distribution patterns of positive- and negative-stranded RNAs within IBs suggest the potential formation of subcompartments in MuV IBs, akin to those observed in HRSV IBs (21, 42). Our findings further reveal that the L protein exhibits two distinct localization patterns within the IBs of MuV-infected cells. One pattern is a ring-like structure that does not colocalize with any viral RNAs, while the other is a reticular pattern that colocalizes with positive-stranded viral RNA. The absence of colocalization between genomic RNA and the L protein may be due to the limitations of the FISH technique used in this study, which might not detect encapsidated genomic RNA, or it could be that genomic RNA rapidly migrates within the IBs after synthesis. Further investigation is required to better understand the spatial dynamics of transcription and genome replication mechanisms within IBs. The MuV N and P proteins formed a reticular structure within the IBs of the infected cells, filling their respective spaces and constituting the basic framework of the IBs. However, it is important to note that the immunostaining of N within the IBs may be affected by the limitation of antibody access (21, 25). These cage-like structures with N and P proteins of the IBs indicate that IBs function as molecular condensates and as a shelter from RNA degradation and antiviral reactions. For example, ribonuclease A (RNase A) treatment did not eliminate viral RNAs in IBs of RABV-infected cells (43). Furthermore, the cage-like structure contributes to evading immune sensing molecules, such as RIG-I and MDA5, by restricting the entry of factors to specific pores. In this study, the small gaps observed in the rim of MuV IBs are considered to correspond to these pores. Because membraneless organelles formed by LLPS do not restrict the exchange of substances from the outside, this structural property may be specific to the virus-induced condensates. Many membraneless organelles are not homogeneous and contain subcompartments induced and guided by RNAs (14, 44). G4-RNA concentrated in MuV IBs may facilitate the formation of cage-like structures characteristic of viral factories through interactions with N and/or P proteins. This study provides insight into the formation and function of IBs, which are the site of mononegavirus RNA synthesis. MATERIALS AND METHODS Cell culture, virus, antibody, and transfection Vero (African green monkey kidney) [American Type Culture Collection (ATCC) CCL-81] and A549 (human lung epithelial) (ATCC CCL-185) cells were purchased from the ATCC (Manassas, VA, USA) and cultured in Dulbecco’s modified Eagle’s medium (Sigma-Aldrich, St. Louis, MO, USA) supplemented with penicillin (100 U/ml), streptomycin (100 mg/ml), and 10% fetal bovine serum (FBS). Baby hamster kidney (BHK) cells constitutively expressing T7 RNA polymerase (BHK/T7-9 cells) (45) were maintained in Glasgow’s MEM (Gibco/Thermo Fisher Scientific Inc., Waltham, MA) with 10% FBS. This study used a wild-type MuV (Odate strain) (46), and infectious titers of the virus were determined by the plaque assay using Vero cells. An anti-MuV N rabbit polyclonal antibody (PAb) was generated as described previously (18). An anti-DNA/RNA G-quadruplex mouse monoclonal antibody (mAb) (BG4) was purchased from Absolute Antibody (Oxford, UK). Cells were transfected with plasmid DNAs using TransIT-LT1 Transfection Reagent (Mirus/TaKaRa Bio, Shiga, Japan) or with siRNA using Lipofectamine RNAiMAX (Invitrogen/Thermo Fisher Scientific Inc.) in accordance with the manufacturer’s instructions. Commercially available siRNA pools targeting MRPS34 mRNA and a control nontargeting siRNA pool (siGENOME SMARTpool) were purchased from Dharmacon/Horizon Discovery (Lafayette, CO, USA). Plasmids The cDNA of the mCherry gene was introduced into the 5′ terminal end of the P gene of pCAGGS-P (4) to yield the plasmid pCAGGS-mCherry-P. The cDNAs of the N, P, and mCherry-P were amplified from pCAGGS-N, pCAGGS-P, and pCAGGS-mCherry-P and cloned into pET15b or pET21b (Novagen/Merck, Darmstadt, Germany) to give plasmids pET15b-N, pET21b-P, and pET21b-mCherry-P, respectively, for expression in Escherichia coli. pMuV-Odate/L-mCherry was constructed by inserting the mCherry gene into the 3′ terminal end of the L gene of pMuV-Odate (47). Virus rescue The plasmid pMuV-Odate/L-mCherry (5 μg) was cotransfected with the helper plasmids, pCITE-Odate-N (250 ng), pCITE-Odate-P (100 ng), and pCITE-Odate-L (500 ng) in BHK/T7-9 cells (48). At 48 hours posttransfection, the transfected BHK/T7-9 cells were mixed with Vero cells at a 1:1 dilution. At the height of the cytopathic effect, usually 4 to 7 days posttransfection, the supernatants were collected, and the entire genome sequences of the rescued viruses were determined. Indirect fluorescent assay Cells were fixed in phosphate-buffered saline (PBS) containing 4% paraformaldehyde (PFA/PBS) for 15 min at room temperature. The cells were then permeabilized with PBS containing 0.2% Triton X-100 for 10 min, blocked with PBS containing 2% bovine serum albumin for 30 min at room temperature, and incubated with mouse anti-DNA/RNA G-quadruplex mAb and/or rabbit anti-N PAb and Alexa Fluor (AF)-conjugated antimouse and/or antirabbit immunoglobulin Gs (IgGs). Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI). Samples were examined under an LSM 900 confocal laser-scanning microscope with an Airyscan (Carl Zeiss AG, Oberkochen, Germany) equipped with a Plan-Apochromat 63× objective lens (numerical aperture = 1.4). The experiments were performed at least three times, and representative data are shown. Fluorescence in situ hybridization MuV RNAs were detected using the ViewRNA ISH Cell Assay Kit (Invitrogen/Thermo Fisher Scientific Inc.) and MuV probe sets that hybridize to nucleotides 279 to 1172 of the MuV antigenome for detecting positive-strand RNA and nucleotides 13,501 to 14,518 of the MuV genome for detecting negative-strand RNA (4). Vero cells infected with MuV were fixed in PFA/PBS for 30 min at room temperature. The cells were then permeabilized and hybridized according to the manufacturer’s protocol. Nuclei were stained with DAPI. Samples were examined under an LSM 900 confocal laser-scanning microscope with Airyscan. The experiments were performed at least three times, and representative data are shown. Time-lapse imaging Vero and A549 cells cultured in a glass-bottom dish were transfected with pCAGGS-mCherry-P and infected with MuV at a multiplicity of infection (MOI) of 5. After 24 hpi, the cells were observed by an LSM 900 confocal laser-scanning microscope equipped with an STX Stage Top Incubator (Tokai Hit, Shizuoka, Japan) heated at 37°C during imaging. Correlative light and electron microscopy Vero cells cultured in a 10-cm dish were infected with MuV at a MOI of 0.01. After 48 hpi, the cells were washed with PBS, fixed with PFA/PBS for 30 min at room temperature, and further fixed in the same fixative for 2 hours at 4°C. The specimens were dehydrated with a series of ethanol gradients and embedded in LR white resin. Thin sections were cut using an ultramicrotome to a thickness of 150 nm and mounted on glass slides. One of the sections was immunofluorescently stained to obtain the location of the MuV IB region by confocal microscopy. Adjacent sections were stained with 1% uranyl acetate and lead citrate to capture detailed morphology by electron microscopy. The target locations, which were detected by the immunofluorescence images, were then observed by ultrahigh-resolution field emission SEM (Regulus8220, Hitachi High-Tech, Tokyo, Japan) using the correlative light and electron microscopy (CLEM) method (MirrorCLEM, Astron, Tokyo, Japan). The SEM images were taken using the backscattered electrons with an acceleration voltage of 1.5 kV. Photoisolation chemistry Fixation and permeabilization A549 cells on coverslips were infected with MuV at an MOI of 5. At 24 hpi, the cells were washed with PBS and fixed with PFA/PBS for 15 min at room temperature. Specimens were permeabilized with 5% Triton X-100 in PBS treated with diethylpyrocarbonate (PBS-DEPC) for 3 min and then with 0.1 N HCl for 5 min, followed by neutralization with 1 M tris-HCl, pH 8.0 for 10 min at room temperature. In situ RT In situ RT for PIC RNA-seq was performed according to a previous report (49). Briefly, permeabilized specimens were incubated in PBS-DEPC for 5 min at 65°C and rapidly cooled in ice-cold PBS-DEPC. Primer mix {0.5 μl of caged ODNs (500 ng/μl) with a distinct barcode, 0.5 μl of 10 mM deoxynucleotide triphosphates (dNTPs) [New England Biolabs (NEB), Ipswich, MA, USA], and 5 μl of H2O} was also heated to 65°C and then rapidly cooled to 4°C, before combining with first-strand mix [2 μl of 5× First-Strand Buffer, 1 μl of 0.1 M dithiothreitol (DTT), 0.5 μl of RNase Out (40 U/ml) and 0.5 ml of Superscript II Reverse Transcriptase (200 U/μl), all from Invitrogen]. The RT reaction mix was applied to the specimens, which were then incubated at 42°C for 60 min, followed by heat inactivation in 70°C PBS for 10 min. Immunostaining Specimens were blocked with blocking solution [50% Blocking OneP (Nacalai Tesque, Kyoto, Japan) in TBST (tris-buffered saline with 0.1% Tween)] for 10 min at room temperature, incubated with rabbit anti-N PAb overnight at 4°C and incubated with an Alexa Fluor 488-conjugated antirabbit IgG antibody (Invitrogen; 1:1000; for 1 hour at room temperature). After nuclear staining with Nuclear Violet (AAT Bioquest, Pleasanton, CA, USA; 1:500 in TBST), specimens were mounted with SlowFade Diamond antifade mountant (Invitrogen). Photoirradiation Photoirradiation for uncaging was performed with a multipattern light-emitting diode (LED) illumination system (Opto-line, Saitama, Japan; LEOPARD2). The system was composed of a Leica DM6 B fluorescence microscope equipped with a DMD (Mightex Systems, Pleasanton, CA, USA; Polygon 1000), dichroic mirror (Semrock, Rochester, NY, USA; 442 nm; Di02-R442-25 × 36) and LED light source (Prizmatix, Holon, Israel; UHP-F-365 LED; 3 W). The light was illuminated to specimens through an objective lens [HC PL APO 63×/1.40-0.60 OIL] at a wavelength of 365 nm for 10 min. Masking sheets loaded on the DMDs were made by ImageJ Fiji (version 2.1.0). Cell lysis The lysis solution [0.1% Tween and Proteinase K (400 μg/ml) in PBS] was loaded onto specimens and incubated for 30 min at 55°C. Lysates of the replicates with distinct barcodes were combined into a single tube. cDNA:mRNA hybrids were purified with a Qiagen MinElute PCR Purification kit (Qiagen, Hilden, Germany) and eluted with H2O. Second-strand DNA synthesis The eluted cDNA:mRNA hybrids (15 μl) were combined with a second-strand mix [2 μl of 5× First-Strand Buffer (Invitrogen), 2.31 μl of Second Strand Buffer (Invitrogen), 0.23 μl of 10 mM dNTPs (NEB), 0.08 μl of E. coli DNA ligase (10 U/μl; Invitrogen), 0.3 μl of E. coli DNA polymerase (10 U/μl; Invitrogen) and 0.08 μl of E. coli RNaseH (2 U/μl; Invitrogen)] and incubated for 2 hours at 16°C. The double-stranded cDNAs were purified with AMpure XP beads (Beckman Coulter, Brea, CA, USA) and eluted with H2O. In vitro transcription The eluted double-stranded cDNAs (6.4 μl) were combined with the in vitro transcription mix [1.6 μl each of A/G/C/UTP solution, 1.6 μl of 10× T7 reaction buffer, and 1.6 μl of T7 enzyme from the MEGAscript T7 Transcription Kit (Invitrogen)] and incubated for 17 hours at 37°C. One microliter of TURBO DNase (2 U/μl; Invitrogen) was added and incubated for 15 min at 37°C. Six microliters of ExoSAP-IT PCR Product Cleanup Reagent (Invitrogen) was added and incubated for 15 min at 37°C, after which 5.5 μl of fragmentation buffer [200 mM tris-acetate (pH 8.1), 500 mM KOAc, and 150 mM MgOAc] was added and incubated for 3 min at 94°C. After adding 2.75 μl of 0.5 M EDTA (pH 8.0), the amplified RNAs (aRNAs) were purified using RNA Clean XP beads (Beckman Coulter) and eluted with H2O. Library preparation and sequencing A step-by-step protocol to construct a sequencing library was described elsewhere (49). The eluted aRNAs (5 μl) were combined with the RT primer mix [1 μl of randomhexRT primer (250 ng/μl) and 0.5 μl of 10 mM dNTPs (NEB)], incubated for 5 min at 65°C and then quickly cooled to 4°C. The RT reaction mix was then added [2 μl of 5× First-Strand Buffer, 1 μl of 0.1 M DTT, 0.5 μl of RNase Out (40 U/μl), and 0.5 μl of Superscript II Reverse Transcriptase (200 U/μl; all from Invitrogen)], incubated for 10 min at 25°C and further incubated for 1 hour at 42°C. The RT products (9 μl) were combined with polymerase chain reaction (PCR) mix [1.8 μl each of 10 μM RNA PCR primers 1, 2 and 22.5 μl of Phusion High-Fidelity PCR Master Mix and 9.9 μl of water] and amplified by PCR (98°C for 30 s, followed by 11 to 15 cycles of 98°C for 10 s, 60°C for 30 s, and 72°C for 30 s, with a final extension at 72°C for 10 min). Fragments of 250 to 500 bp were then purified and size selected with AMpure XP beads. The quality of the resulting cDNA library was assessed using high-sensitivity DNA chips on a Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA) and quantified using a Library Quantification Kit (Clontech/TaKaRa Bio) before sequencing on the Illumina NovaSeq 6000 platform. Data analysis The barcodes and UMIs in the reads were extracted using UMI-tools (version 1.0.0) with the following command: umi_tools extract -I r1.fastq --read2-in=r2.fastq --bc-pattern=NNNNNNCCCCCC --read2-stdout. The reads were mapped by the aligning software HISAT2 (version 2.1.0) to the reference genome (GRCh38). Read counts per gene were determined with featureCounts (version 1.6.4) and UMI-tools with the following command: umi_tools count --method=directional --per-gene --per-cell --gene-tag=XT. The duplicates in the UMIs were discarded. Using the resulting counts, differentially expressed genes (false discovery rate = 0.1) were extracted by the R library DESeq2 (version 1.20.0), which was also used to transform count data into regularized log data before clustering (heatmap3, version 1.1.6) and dimension reduction (UMAP, version 0.2.2.0). The TPM value was calculated with a TPMCalculator (PMID: 30379987) using the bam file of the HISAT2 mapping data. Statistical analysis of TPM values was performed using edgeR version 3.12.0 of the Bioconductor package (PMID: 19910308). Protein expression and purification N-terminal or C-terminal His-tagged recombinant proteins were expressed in Rosetta-gami 2(DE3) cells (Novagen) by induction with 0.2 mM isopropyl-β-d-thiogalactopyranoside at 37°C for 5 hours. The pelleted cells were resuspended in lysis buffer [20 mM tris-HCl (pH 7.5), 500 mM NaCl, 1 mM EDTA, 1% Triton X-100, 10 mM imidazole, and protease inhibitor cocktail (Complete Mini; Roche, Mannheim, Germany)]. After centrifugation at 16,000g for 20 min at 4°C, the supernatant was incubated with Ni Sepharose 6 Fast Flow (Cytiva, Tokyo, Japan) for 2 hours at 4°C. The resins were washed with lysis buffer three times, and the His-tagged proteins were eluted with elution buffer [20 mM tris-HCl (pH 7.5), 500 mM NaCl, 1 mM EDTA, and 500 mM imidazole]. The eluted proteins were desalted and concentrated with Amicon Ultra Centrifugal Filters (Merck). In vitro phase separation assay Proteins were diluted in phase separation buffer [20 mM tris-HCl (pH 7.5) and 1 mM EDTA] at a final NaCl concentration of 150 mM unless otherwise specified. Fluorescent labeled droplets were formed by mixing 5 μM P-His and 25 to 50 nM AcGFP-P-His. PEG-8000 (Sigma-Aldrich) was used as a molecular crowding agent at the final concentration of 5% (w/v). Cy3-RNA was used at a final concentration of 10 ng/μl. Samples were incubated at room temperature for 15 min and examined under an LSM 900 confocal laser-scanning microscope, and representative pictures were taken. The size and intensity of each droplet were calculated using ZEN software (Carl Zeiss AG). In vitro transcription DNA templates for in vitro transcription were amplified by PCR using primers containing the T7 promoter sequence at the 5′-end. The in vitro transcription reaction was performed using HiScribe T7 High Yield RNA Synthesis Kit (NEB, Ipswich, MA, USA) in accordance with the manufacturer’s instructions. For RNA labeling, Cy3-UTP (Enzo Life Sciences Inc., Farmingdale, NY, USA) was added to the in vitro transcription reaction at a final concentration of 0.05 mM. The transcribed RNA was purified using Monarch RNA Cleanup Kit (NEB) and resuspended in RNase-free water. Fluorescence recovery after photobleaching In vivo Vero and A549 cells cultured in a glass-bottom dish were transfected with pCAGGS-mCherry-P and infected with MuV at an MOI of 5. After 24 hpi, the cells were observed by an LSM 900 confocal laser-scanning microscope equipped with an STX Stage Top Incubator heated at 37°C. After acquiring five images (206 ms per frame), bleaching was performed in a circular region to bleach an IB subvolume using a 561-nm laser at 50% intensity. In vitro Proteins (5 mM P-His and 50 nM AcGFP-P-His) were diluted in phase separation buffer at final concentrations of 150 mM NaCl and 5% (w/v) PEG. Samples were incubated at room temperature for 15 min and examined under an LSM 900 confocal laser-scanning microscope. After acquiring five images (150 ms per frame), bleaching was performed in a circular region to bleach an IB subvolume using a 488-nm laser at 12% intensity. The experiments were performed at least five times, and representative data are shown. Electrophoretic mobility shift assay Proteins at the indicated concentrations were mixed with 2 μg of RNA, incubated for 30 min at room temperature, and then separated by 1% agarose gel electrophoresis. RNA was stained by ethidium bromide and visualized with a Printgraph CMOS I (ATTO, Tokyo, Japan). The experiments were performed three times, and representative data are shown. Pull-down assay BioTracker MultiTASQ G4 Cell Probe (Merck) (10 μM) was mixed with 2 μg of RNA for 1 hour at room temperature. RNA-probe complexes were captured with Dynabeads M-280 Streptavidin (Veritas/Thermo Fisher Scientific Inc.) for 30 min at room temperature, and then the beads were washed with tris buffer (20 mM tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA) three times. Pulled-down RNA was separated by 1% agarose gel electrophoresis, stained by ethidium bromide, and visualized with a Printgraph CMOS I. The experiments were performed three times, and representative data are shown. Statistical analysis Differences between groups were evaluated using unpaired Student’s t tests. Error bars indicate the SDs of triplicate measurements. Statistical significance was assumed at **P < 0.01.