ZNRD2 Mediated Nucleoprotein Aggregation Impairs Respiratory Syncytial Virus Replication
Haiwu Zhou, Mingbin He, Jinhong Du, Cong Liu, Weiwei Wang, Yuewen Han, Zhifei Li, Yali Qin, Mingzhou Chen
TL;DR
A host protein called ZNRD2 restricts respiratory syncytial virus replication by forming aggregates with viral nucleoprotein, but the virus can counteract this by preventing aggregate formation.
Contribution
This study reveals a dual antagonistic mechanism between RSV and ZNRD2, offering new insights into virus-host interactions and antiviral strategies.
Findings
ZNRD2 restricts RSV by promoting insoluble aggregation of the viral nucleoprotein.
RSV phosphoprotein inhibits ZNRD2-nucleoprotein aggregation to evade host restriction.
RSV infection causes dynamic changes in ZNRD2 solubility, indicating a complex host-virus interaction.
Abstract
Nucleoproteins (N) of negative‐sense RNA viruses exhibit an inherent tendency to oligomerize, forming a ribonucleoprotein complex that protects the viral genome. Here, immunoprecipitation coupled with mass spectrometry is used to identify zinc ribbon domain containing 2 (ZNRD2) as a host interactor for the respiratory syncytial virus (RSV) N protein. The results demonstrated that ZNRD2 functions as a restriction factor against RSV by enhancing the oligomerization and insolubility of N. Conversely, RSV N sequesters ZNRD2 into an insoluble aggregate, rendering it incapable of performing physiological functions required for the quality control of chaperonin assembly. Notably, RSV phosphoprotein (P) completely inhibited the formation of the insoluble ZNRD2‐RSV N complex by maintaining N in a monomeric conformation. RSV infection induces a fluctuation in the solubility of ZNRD2, indicating a…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6- —National Natural Science Foundation of China10.13039/501100001809
- —National Key R&D Program of China10.13039/501100012166
- —Natural Science Foundation of Wuhan
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsRespiratory viral infections research · interferon and immune responses · Virology and Viral Diseases
Introduction
1
The respiratory syncytial virus (RSV) is the primary cause of acute lower respiratory tract infections (ALRTIs) worldwide, particularly among infants, older adults, and immunocompromised individuals. Each year, RSV accounts for ≈33 million ALRTI cases globally, with over 100 000 deaths among children under 5 years of age.^[^ 1, 2 ^]^ Beyond its acute manifestations, RSV infection has been associated with the development of chronic respiratory conditions, including asthma and recurrent wheezing. Despite the recent approval of two RSV vaccines, their availability and adoption remain limited in low‐ and middle‐income countries, where accessibility and affordability persist as key barriers.^[^ 3, 4, 5 ^]^
RSV belongs to the order Mononegavirales, family Pneumoviridae, and genus Orthopneumovirus.^[^ 6 ^]^ The virions are enveloped and exhibit either a spherical or filamentous morphology.^[^ 7 ^]^ Its envelope hosts three glycoproteins: fusion (F), attachment (G), and small hydrophobic (SH) proteins.^[^ 8 ^]^ Internally, the matrix protein (M) lines the inner leaflet and acts as a structural bridge between the envelope and nucleocapsid.^[^ 9, 10 ^]^ The nucleocapsid consists of the nucleoprotein (N) that encapsulates a single‐stranded negative‐sense RNA genome comprising 15 222 nucleotides arranged in a left‐handed helical filament that acts as a template for transcription and replication.^[^ 11, 12, 13 ^]^ RNA synthesis, executed by the L‐P polymerase complex, is regulated by M2‐1 and M2‐2 proteins and occurs within viral factories (VFs, also called inclusion bodies), which are structures formed via liquid‐liquid phase separation by the N and P proteins.^[^ 14, 15, 16, 17 ^]^ VFs sequester host antiviral factors such as p65,^[^ 18 ^]^ MDA5, and MAVS (through N),^[^ 19 ^]^ while recruiting PABP (through M2‐1),^[^ 20, 21 ^]^ PP1α,^[^ 22 ^]^ and chaperones, HSP90 and HSP70 (through P),^[^ 23 ^]^ to promote viral replication.
The N proteins of negative‐sense RNA viruses (NSVs) consistently undergo oligomerization, which is essential for their biological functions. In non‐segmented NSVs (nsNSVs) such as RSV, Ebola virus (EBOV), vesicular stomatitis virus, parainfluenza virus 5, measles virus, and Hendra virus, the N protein assembles into a ring‐like ribonucleoprotein complex consisting of 10–14 protomers.^[^ 24, 25 ^]^ In contrast, segmented NSVs (sNSVs) display structural variability: influenza A and Lassa virus N form trimers, orthobunyaviruses form tetramers, Rift Valley fever virus forms hexamers, and arenaviruses form heptamers.^[^ 26, 27 ^]^ Structurally, the N protein contains N‐terminal (NTD) and C‐terminal (CTD) domains that form a positively charged groove for RNA binding, with oligomerization driven by the combined action of RNA binding and inter‐protomer interactions mediated by the flanking N‐ and C‐arms. In nsNSVs, N oligomerizes side by side to assemble compact nucleocapsid,^[^ 24, 28 ^]^ and its monomeric form (N^0^) is maintained by chaperones such as P or, in the case of EBOV, VP35.^[^ 13, 29 ^]^ In sNSVs, N oligomerization occurs via head‐to‐tail interaction, yielding more flexible structures,^[^ 26, 27 ^]^ and the monomeric state typically persists until viral RNA binding, a process often regulated by conformational changes or post‐translational modifications.^[^ 30, 31, 32 ^]^ In the case of RSV, N oligomerization may also be regulated by phosphorylation.^[^ 33 ^]^ Additionally, the newly synthesized N binds to P via N‐CTD: P‐NTD interactions, preventing unwanted RNA binding. P positions N near viral RNA for encapsidation and then dissociates to bind N oligomers via N‐NTD: P‐CTD interactions, facilitating VF formation and viral RNA synthesis. The N‐terminal 40 amino acids of P suffice to maintain N in a monomeric state in vitro and enable RNA encapsidation into decameric N‐RNA rings. Mutations that disrupt RNA binding (K170A + R185A, termed N^mono^) or oligomerization (N‐arm deletion) prevent oligomer formation and impair VF assembly and minigenome activity, underscoring the critical role of oligomerization in viral replication.^[^ 13, 29, 34, 35, 36, 37, 38, 39, 40, 41 ^]^
Protein aggregation is a fundamental pathological mechanism underlying numerous diseases, including neurodegenerative, metabolic, and ocular disorders, and is primarily driven by the accumulation of insoluble amyloid deposits. These aggregates, comprising proteins such as amyloid‐β, Tau, α‐synuclein, and transthyretin, originate from genetic mutations, environmental stressors, aging, or cross‐seeding events.^[^ 42, 43, 44, 45, 46 ^]^ Viral infections have also been implicated in promoting pathological aggregation, evidenced by prion‐mediated misfolding in Creutzfeldt‐Jakob disease^[^ 47 ^]^ and the facilitation of aggregate propagation by endogenous retroviruses.^[^ 48 ^]^ Consequently, the targeting of functional viral protein aggregates has emerged as a promising antiviral strategy.^[^ 49 ^]^ Thus, protein aggregation functions as both a pathological hallmark and a direct driver of pathogenicity.
RSV N adopts both soluble monomeric and insoluble oligomeric states during infection. The mechanisms via which host cells govern these distinct viral macromolecules and the fate of N outside VFs remain poorly defined. We showed that while RSV N resists host‐mediated degradation, further oligomerization and insoluble complex formation occur via aggregation with zinc ribbon domain containing 2 (ZNRD2), ultimately impairing its contribution to viral replication. RSV P inhibited the formation of insoluble ZNRD2‐RSV N aggregates by maintaining the monomeric state of the N protein, thereby counteracting this restriction. Furthermore, RSV N independently induces ZNRD2 to form insoluble aggregates outside the VFs, impairing ZNRD2's role in chaperonin assembly quality control.
Results
2
ZNRD2 Restricts RSV Replication via Direct N Interaction
2.1
To identify cellular proteins that interact with RSV N, we performed immunoprecipitation combined with mass spectrometry in HEK293T cells transiently expressing Myc‐tagged RSV N. Among the identified interactors, the four most abundant proteins were ZNRD2, E3 ubiquitin‐protein ligase, BRE1B (RNF40), DNA topoisomerase 2‐alpha (TOP2A), and La‐related protein 1 (LARP1) (Figure S1A and Table S1, Supporting Information). The interaction between endogenous ZNRD2 and RSV N was validated using co‐immunoprecipitation (co‐IP) (Figure 1A). Considering that RNF40, TOP2A, and LARP1 possess intrinsic nucleic acid‐binding properties, their association may be related to N's strong RNA‐binding abilities. Therefore, we focused on ZNRD2 (also known as Sjögren syndrome/scleroderma autoantigen 1 SSSCA1),^[^ 50 ^]^ the most highly enriched yet poorly characterized interactor. Previous bioinformatics analyses have predicted that ZNRD2 contains a zinc finger domain and acts as an adaptor protein that links the E3 ubiquitin ligase, HERC2, to the orphan subunits of the chaperonin CCT complex, thereby participating in assembly quality control.^[^ 51 ^]^
ZNRD2 restricts RSV replication via direct N interaction. A) Myc‐RSV N was transiently expressed in HEK293T cells and subjected to anti‐Myc co‐IP. B, C) RSV infection (MOI = 0.01, 0.1, 1) in HeLa cells stably expressing HA‐ZNRD2. Viral titers in supernatants were determined by TCID50 assay at 36 h post infection (hpi) (B). Viral protein expression (N, P, M2‐1) was analyzed by immunoblotting (C). D–H) Time course analysis of RSV infection (MOI = 0.1) in HeLa cells stably expressing HA‐ZNRD2. Viral mRNA levels (NS1, N, L) (D–F), genomic RNA levels (G), and protein expression (N, P, M2‐1) (H) were measured at 0, 12, 24, and 36 hpi. I–K) RSV infection (MOI = 0.1) in wild‐type, ZNRD2 knockout (KO), and ZNRD2 KO HeLa cells reconstituted with HA‐ZNRD2. Viral mRNA (NS1, N, L) (I), genomic RNA levels (J), and protein expression (N, P, M2‐1) (K) were analyzed at 36 hpi. Data were presented as mean ± SD (n = 3). Statistical analysis was performed using two‐way ANOVA with Šídák's multiple comparisons test (B, D–G, I) or ordinary one‐way ANOVA with Dunnett's multiple comparisons test (J).
To evaluate the functional effects of ZNRD2 on RSV replication, we established HeLa cell lines stably expressing HA‐tagged ZNRD2 or vector controls. Overexpression of ZNRD2 resulted in an approximately tenfold reduction in viral titers (TCID_50_) across multiple multiplicities of infection (MOIs: 0.01, 0.1, 1) compared to that in control cells (Figure 1B), along with a corresponding decrease in viral protein levels (Figure 1C). Time‐course analyses demonstrated the consistent suppression of viral mRNA, genomic RNA, and protein expression (Figure 1D–H). The progressive decline observed in NS1, N, and L transcripts aligns with the characteristic 3’‐to‐5’ transcription gradient of nsNSVs.^[^ 17, 52 ^]^ This inhibitory effect was independent of cell type, as HEp‐2 cells stably expressing ZNRD2 also exhibited significant reductions in viral RNA and protein accumulation (Figure S1B–D, Supporting Information).
Genetic depletion further substantiated the restrictive role of ZNRD2. Both shRNA‐mediated knockdown (Figure S1E, Supporting Information) and CRISPR‐Cas9 knockout of endogenous ZNRD2 in HeLa cells led to a significant enhancement in RSV replication compared to that observed in the scrambled control (SCR) or wild type cells (Figure 1I–K; Figure S1F–H, Supporting Information). The reintroduction of ZNRD2 into knockout cells restored the suppression of viral replication (Figure 1I–K). Notably, the inhibitory effect plateaued despite supraphysiological expression levels, an observation that has been investigated subsequently. Collectively, these gain‐ and loss‐of‐function experiments established ZNRD2 as a host factor that restricts RSV replication.
ZNRD2 Enhances the Insolubility of RSV N without Inducing Its Degradation
2.2
While validating the interaction between overexpressed ZNRD2 and RSV N, we unexpectedly observed a reduction in RSV N protein levels in whole cell lysates (Figure 2A; Figure S2A, Supporting Information). This reduction was not attributable to transfection artifacts because GFP expression remained stable under identical conditions (Figure S2B, Supporting Information). Moreover, ZNRD2 overexpression did not alter N protein levels of other nsNSVs, including human metapneumovirus, human parainfluenza virus type 3 (hPIV3), Nipah virus, and EBOV (Figure S2C, Supporting Information), ruling out the occurrence of nonspecific effects. Considering the known association between E3 ubiquitin ligases and ZNRD2, we initially hypothesized that RSV N was degraded. However, neither inhibition of protein synthesis with cycloheximide (Figure 2B) nor treatment with proteasome (MG132) or lysosome (chloroquine) inhibitors (Figure 2C) restored RSV N levels, thereby excluding canonical proteolytic degradation as the underlying mechanism.
ZNRD2 enhances the insolubility of RSV N without inducing its degradation. A) Myc‐RSV N and HA‐ZNRD2 were co‐expressed in HEK293T cells and subjected to anti‐HA co‐IP. B) Myc‐RSV N and HA‐ZNRD2 were co‐expressed in HEK293T cells and treated with cycloheximide (CHX). Cells were harvested at 0, 0.5, 1, 2, 4, and 8 h post‐treatment. C) Myc‐RSV N and HA‐ZNRD2 were co‐expressed in HEK293T cells and treated with MG132 or chloroquine (CQ) for 8 h, followed by immunoblotting analysis. D) Myc‐RSV N and HA‐ZNRD2 were co‐expressed in HEK293T cells and lysed using 1% SDS buffer, lysis buffer, or NP‐40 lysis buffer. E) Myc‐RSV N and HA‐ZNRD2 were co‐expressed in HEK293T cells and followed by solubility fractionation. F) Gradient expression of Myc‐RSV N was performed in HEK293T cells for solubility fractionation analysis. G) Myc‐RSV N or vector was expressed in HEK293T cells. Lysates from Myc‐RSV N‐expressing cells were first subjected to anti‐Myc co‐IP, followed by a second co‐IP using lysates from vector‐expressing cells. H) Myc‐tagged nucleoproteins from RSV, hMPV, hPIV3, NiV, or EBOV were expressed in HEK293T cells for solubility fractionation analysis.
RSV N is a cytosolic protein without membrane association and is not secreted in the absence of viral glycoproteins.^[^ 53 ^]^ Complete recovery of RSV N following sodium dodecyl sulfate (SDS)‐based lysis, which completely solubilized all cellular components, confirmed that the protein was not degraded (Figure 2D). Notably, progressively milder lysis conditions led to the diminished detection of both RSV N and ZNRD2 (Figure 2D), suggesting that the apparent loss reflected changes in solubility rather than degradation. To test this hypothesis, we performed solubility fractionation assays. Under NP‐40 lysis (soluble fraction) conditions, RSV N and ZNRD2 levels decreased simultaneously, whereas both accumulated in the SDS‐solubilized pellet fraction (insoluble fraction) (Figure 2E). This effect was reciprocal, as RSV N expression also reduced ZNRD2 levels (Figure S2D, Supporting Information), although high‐level overexpression of ZNRD2 masked this phenotype (Figure S2A, Supporting Information). Under native conditions, ZNRD2 remained predominantly soluble. However, the presence of RSV N induced a dose‐dependent shift of ZNRD2 to the insoluble fraction (Figure 2F). This phenomenon was consistent across cell types and was confirmed in HeLa cells (Figure S2E,F, Supporting Information).
Sequential co‐IP further demonstrated that the recovery of ZNRD2 was higher from lysates lacking RSV N, indicating that the “missing” ZNRD2 was trapped in an insoluble complex (Figure 2G). Among the nsNSV N proteins tested, only RSV N specifically induced the insolubility of ZNRD2 (Figure 2H). Next, we investigated whether the lack of functional effects of other viral N proteins stemmed from an absence of physical interaction. Unexpectedly, co‐IP assays revealed that ZNRD2 interacted with hPIV3 N (Figure S2G, Supporting Information). However, despite this interaction, hPIV3 N did not alter the solubility of ZNRD2 (Figure 2H); nor did ZNRD2 overexpression impact hPIV3 replication (Figure S2H, Supporting Information). Although its interaction with hPIV3 N was notable, its functional implications were not investigated further in this study. Considering the strong RNA‐binding capacity of RSV N, RNase treatment negligibly affected the ZNRD2 interaction and failed to reverse the insolubility (Figure S2I,J, Supporting Information), indicating that RNA is not essential for complex formation.
Oligomerization is Essential for Insoluble ZNRD2‐RSV N Complex Formation
2.3
RSV N inherently forms a large oligomeric complex upon RNA binding, a tendency so pronounced that its expression alone induces spontaneous insolubility.^[^ 35 ^]^ This insolubility extended to ZNRD2, implying that RSV N‐oligomerization contributed to the formation of an insoluble complex. To test this hypothesis, we used an oligomerization‐deficient RSV N mutant (N^mono^) harboring the K170A and R185A substitutions, which disrupted its RNA‐binding ability.^[^ 34 ^]^ Although endogenous ZNRD2 exists as a monomer without an intrinsic oligomerization tendency, RSV N, but not N^mono^, strongly triggered ZNRD2 oligomerization (Figure S3A, Supporting Information). This revealed that ZNRD2 possessed a latent capacity for oligomerization, which was activated and enhanced by RSV N (Figure 3A). Importantly, N^mono^ failed to co‐immunoprecipitate with ZNRD2 (Figure 3B) and did not induce the transition of ZNRD2 into the insoluble fraction (Figure 3C). These results demonstrated that RSV N oligomerization is essential for the formation of the insoluble ZNRD2‐RSV N complex.
Oligomerization is essential for insoluble ZNRD2‐RSV N complex formation. A) Myc‐RSV NWT or Myc‐RSV Nmono was co‐expressed with HA‐ZNRD2 in HEK293T cells and followed by treatment with disuccinimidyl suberate (DSS) for oligomerization analysis. B) Myc‐RSV NWT or Myc‐RSV Nmono was expressed in HEK293T cells and subjected to anti‐Myc co‐IP. C) Myc‐RSV NWT or Myc‐RSV Nmono was expressed in HEK293T cells for solubility fractionation analysis. D) HA‐ZNRD2WT, HA‐ZNRD2L180E, HA‐ZNRD2L183E, or HA‐ZNRD2I184E was co‐expressed with Myc‐RSV N in HEK293T cells and followed by DSS treatment for oligomerization analysis. E) HA‐ZNRD2WT, HA‐ZNRD2L180E, HA‐ZNRD2L183E, or HA‐ZNRD2I184E was co‐expressed with Myc‐RSV N in HEK293T cells and subjected to anti‐HA co‐IP. F) HA‐ZNRD2WT, HA‐ZNRD2L180E, HA‐ZNRD2L183E, or HA‐ZNRD2I184E was co‐expressed with Myc‐RSV N in HEK293T cells for solubility fractionation analysis. G–I) HeLa cell lines stably expressing HA‐ZNRD2WT, HA‐ZNRD2L180E, HA‐ZNRD2L183E, or HA‐ZNRD2I184E were infected with RSV (MOI = 0.1). Viral mRNA levels (NS1, N, L) (G), genomic RNA levels (H), and protein levels (N, P, M2‐1) (I) were measured at 36 hpi. Data were presented as mean ± SD (n = 3). Statistical analysis was performed using two‐way ANOVA with Šídák's multiple comparisons test (G) or ordinary one‐way ANOVA with Dunnett's multiple comparisons test (H).
Based on these findings, we hypothesized that ZNRD2's own oligomerization also contributes to complex formation. Despite its known cellular roles, the oligomerization properties of ZNRD2 have not yet been characterized. To investigate this, we conducted structure‐function analyses using systematic truncation and site‐directed mutagenesis. ZNRD2 consists of three regions: an N‐terminal domain (NTD), an intrinsically disordered region (IDR), and a C‐terminal domain (CTD). Deletion of the IDR had minimal impact on its interaction and insolubility with RSV N; however, removal of either the NTD or CTD abolished these effects entirely (Figure S3B–D, Supporting Information), indicating that both structured domains are critical. Further truncations within the NTD and CTD refined these regions; residues 14–85 were dispensable, whereas residues 148–177 contributed partially. Residues 2–13 and 178–187 were essential for insolubility (Figure S3E, Supporting Information). Attempts to delete residues 188–199 resulted in protein destabilization, precluding further analysis. Next, we introduced point mutations in the essential regions (residues 2–13 and 178–187) to disrupt oligomerization. The deletion of residues 2–13 significantly reduced oligomerization, with W13 emerging as a critical residue. The deletion of residues 178–187 completely abolished oligomerization, consistent with the effects of the L180E, L183E, and I184E substitutions in this segment. Mutations I178E, R185A, and A186E exhibited intermediate effects (Figure S3F,G, Supporting Information). Thus, L180, L183, and I184 are the core residues that mediate ZNRD2 oligomerization.
Oligomerization‐deficient ZNRD2 mutants lost the ability to promote RSV N oligomerization (Figure 3D). Among these, the L183E mutant retained weak binding to RSV N, whereas the L180E and I184E mutants completely lost this ability (Figure 3E). These mutations also restored soluble RSV N levels to those of the wild‐type ZNRD2 (Figure 3F). Critically, the stable expression of these mutants in HeLa cells did not suppress RSV replication (Figure 3G–I). Taken together, the formation of the insoluble ZNRD2‐RSV N complex requires oligomerization. Although insufficient, this necessity highlights oligomerization as a critical licensing step for complex assemblies.
Computational Modeling Suggests a Putative ZNRD2‐RSV N Interface
2.4
To characterize the molecular interface of the ZNRD2‐RSV N complex, we used AlphaFold3 for structural modeling.^[^ 54, 55 ^]^ The predicted model exhibited low global confidence (pTM = 0.34) (Figure 4A,B; Figure S4A,B, Supporting Information). The predicted alignment error (PAE) indicated weak interface signals between chain B (the N protomer proximal to ZNRD2) and chain K (ZNRD2), primarily involving the RSV N CTD and ZNRD2 NTD (Figure S4C, Supporting Information). Notably, AlphaFold3 correctly predicted the RSV N decamer as a ring‐like structure, with most residues displaying plDDT > 70 (Figure S4D, Supporting Information). The observed low pTM score may result from the high PAE values between spatially distant N protomers (Figure S4C, Supporting Information), similar to the effect exacerbated by intrinsically disordered regions (IDRs) and the large input in TM‐score calculations.^[^ 56, 57 ^]^
Computational modeling suggests a putative ZNRD2‐RSV N interface. A, B) The structural model of the ZNRD2‐RSV N complex was predicted using AlphaFold3. C) Electrostatic potential maps (EPM) of solvent‐accessible surface areas (SASA) for ZNRD2 and RSV N. The inset shows detailed interaction interfaces. D) SASA‐EPM of ZNRD2 molecular surface facing RSV N. E) HA‐ZNRD2WT or HA‐ZNRD2D32A was co‐expressed with Myc‐RSV N in HEK293T cells for solubility fractionation analysis. F) Myc‐RSV NWT, Myc‐RSV NY337A, or Myc‐RSV NK351A was expressed in HEK293T cells for solubility fractionation analysis. G) Myc‐RSV N mutants (Y337A, Y337F) were expressed in HEK293T cells and followed by treatment with DSS for oligomerization analysis. H) Myc‐RSV N mutants (Y337A, Y337F) were expressed in HEK293T cells and subjected to anti‐Myc co‐IP.
Five independent prediction runs (using different model seeds) revealed highly consistent structural features, which prompted us to consider the predicted structure potentially informative. In particular, as the position of the C‐terminus is erratic, the ZNRD2 N‐terminal α1 helix (residues 18–44) is consistently predicted to adopt a pose orthogonal to the RSV N α13 helix (residues 344–360) (Figure 4A,B; Figure S4A,B, Supporting Information). This configuration enabled the formation of a potential salt bridge between ZNRD2 D32 and RSV N K351, supported by the complementary distribution of electrostatic surface potentials at the interface (Figure 4C,D). Experimental mutagenesis confirmed that the D32A and K351A substitutions moderately impaired both interaction and insolubility (Figure 4E,F; Figure S4F,G, Supporting Information).
Using a 4 Å distance cutoff, we screened for putative interface residues (Figure S4E, Supporting Information). Unexpectedly, beyond D32/K351, the RSV N^Y337A^ mutant completely abolished ZNRD2's interaction and induction of insolubility (Figure 4F; Figure S4G, Supporting Information). In the published RSV nucleocapsid‐like structure (PDB: 8OOU), the aromatic ring of residue Y337 makes contact with the ribose phosphate backbone,^[^ 11, 58 ^]^ and a previous study has reported that the Y337 mutation impairs nucleocapsid assembly.^[^ 59 ^]^ Mutation of tyrosine to alanine (Y337A) impaired RSV N oligomerization, whereas mutation to phenylalanine (Y337F), which retains an aromatic side chain, did not impair oligomerization and maintained the interaction with ZNRD2. The minor loss of interaction observed with Y337F possibly resulted from a slight reduction in oligomerization levels (Figure 4G,H). Therefore, the Y337 site may not directly engage in ZNRD2 binding; rather, its mutation indirectly disrupts this interaction by impairing oligomerization.
These results indicated the limitations of AlphaFold3 in modeling large macromolecular complexes; however, the high reproducibility of AlphaFold3 prediction underscores the interface between RSV N CTD and ZNRD2 NTD, which is a potentially verifiable assumption. The predicted D32‐K351 salt bridge exerted a minor rather than decisive interaction. Collectively, these data reinforce the notion that oligomerization regulates ZNRD2‐RSV N complex formation via mechanisms beyond residue‐specific contacts.
RSV P Protein Prevents ZNRD2‐RSV N Aggregation
2.5
During RSV infection, N and P proteins assemble into micron‐scale viral factories (VFs) via liquid‐liquid phase separation, recruiting host factors such as PABP, eIF4G, PP1, and MDA5 via interactions with N, P, or M2‐1. Considering the established interaction between ZNRD2 and RSV N, ZNRD2 is expected to localize within VFs and inhibit viral replication and transcription. Instead, over the course of a 48‐h infection, ZNRD2 remained diffusely distributed in the cytoplasm and did not show any colocalization with VFs (Figure S5A, Supporting Information). Consistent with this, the RSV minigenome system revealed only a marginal reduction in reporter activity upon ZNRD2 overexpression (p = 0.0889; Figure S5B,C, Supporting Information). Moreover, although ZNRD2 insolubility increased in a dose‐dependent manner with increasing RSV N expression, during prolonged infection with increasing levels of N, ZNRD2 distribution fluctuated in both the soluble and insoluble fractions rather than showing progressive accumulation (Figure 5A). These findings suggest the presence of additional regulatory mechanisms governing the ZNRD2‐RSV N complex.
RSV P protein prevents ZNRD2‐RSV N aggregation. A) HeLa cells were infected with RSV (MOI = 0.1) and harvested at 0, 12, 24, 36, and 48 hpi for solubility fractionation analysis. B) Myc‐RSV N and HA‐ZNRD2 were co‐expressed in HeLa cells for confocal analysis. C) Myc‐RSV N and FLAG‐RSV P were co‐expressed with or without HA‐ZNRD2 in HeLa cells for confocal analysis. D) Myc‐RSV N and HA‐ZNRD2 were co‐expressed with FLAG‐RSV PWT or FLAG‐RSV PF241A in HEK293T cells and subjected to anti‐HA co‐IP. E) Myc‐RSV N was co‐expressed with FLAG‐RSV PWT or FLAG‐RSV PF241A in HEK293T cells for solubility fractionation analysis. Scale bar = 10 µm.
RSV infection is a complicated biological process that extends beyond the simple transfection of viral proteins and often involves the dynamic regulation of viral components. Limited insights into the physiological role of ZNRD2 have prompted studies on potential viral regulators, among which RSV P, a molecular chaperone that remains tightly associated with N and maintains it in a monomeric state, has emerged as a leading candidate. Notably, ZNRD2 and RSV N formed abundant cytoplasmic aggregates in the absence of P, consistent with their inherent tendency toward remaining insoluble (Figure 5B). However, co‐expression of P induced the formation of typical pseudo‐VF structures by the N‐P complex and completely prevented ZNRD2 colocalization (Figure 5C). This explains the lack of ZNRD2 localization within VFs during infection and suggests that P actively antagonizes ZNRD2‐N aggregation. To further dissect this mechanism, the F241A mutant of RSV P, known to be defective in N‐P interactions (Figure S5D, Supporting Information), was used.^[^ 29, 39 ^]^ Wild‐type P, but not P^F241A^, entirely abolished the ZNRD2‐RSV N interaction and reversed RSV N‐induced ZNRD2 insolubility (Figure 5D,E). The restoration of the reciprocal solubility shifts between ZNRD2 and RSV N was also P‐dependent (Figure S5E, Supporting Information).
RSV P actively prevents ZNRD2‐RSV N aggregation via a hierarchical molecular mechanism. RSV P maintains N in a soluble monomeric conformation by binding, thereby abolishing its ability to engage ZNRD2, as evidenced by the disrupted interaction and loss of colocalization. This interception of the primary interaction suppresses the nucleation of the insoluble complex. In the context of infection, physiological complexity generates spatiotemporal imbalances in N‐to‐P stoichiometry, leading to the generation of transient pools of P‐free N. These unbound N molecules become susceptible to ZNRD2 binding, resulting in oscillatory insolubility.
RSV N Inhibits the Physiological Functions of ZNRD2
2.6
Although ZNRD2 remains poorly characterized, it plays a well‐established physiological role as an adaptor linking the E3 ubiquitin ligase, HERC2, with the orphan subunits of the chaperonin CCT complex, facilitating their degradation to ensure precise chaperonin assembly and correct CCT stoichiometry.^[^ 51 ^]^ The finding that ZNRD2 forms an insoluble complex with RSV N to restrict viral replication raises the question of whether RSV N disrupts normal cellular functions of ZNRD2.
Ectopically expressed CCT subunits are recognized as orphans and are targeted for degradation via the HERC2‐ZNRD2 pathway.^[^ 51 ^]^ RSV N, but not the interaction‐deficient mutant RSV N^Y337A^, markedly increased the protein levels of CCT1, CCT3, and CCT4 but exerted modest effects on CCT2 and no significant effects on CCT5–CCT8 (Figure 6A–H). This pattern aligns with the previously described degradation mechanisms, especially for CCT4, which is the most reliably documented target of ZNRD2‐dependent degradation. Notably, RSV N rescued CCT4 from ZNRD2‐mediated degradation, even under conditions that promoted enhanced turnover (Figure S6A, Supporting Information). Control experiments using two degradation‐resistant CCT4 mutants, CCT4^A88E^ (defective in ZNRD2 binding) and CCT4^Δ103–107^ (degraded via a ZNRD2‐independent pathway), showed lack of stabilization by RSV N (Figure S6B,C, Supporting Information), confirming the specificity of this rescue. Genetic knockouts further demonstrated that the ability of RSV N to stabilize these CCT subunits is dependent on ZNRD2. RSV N failed to interact directly with any CCT subunit in the absence of ZNRD2; re‐expression of ZNRD2 restored binding to CCT1–CCT4 and CCT7 but not to CCT5, CCT6A, or CCT8 (Figure 6I), mirroring the observed stabilization pattern.
RSV N inhibits the physiological functions of ZNRD2. A–H) CCT‐3×FLAG was co‐expressed with Myc‐RSV NWT or Myc‐RSV NY337A in HEK293T cells. I) CCT‐3×FLAG, and Myc‐RSV N were co‐expressed with or without HA‐ZNRD2 in ZNRD2‐knockout HEK293T cells and subjected to sequential anti‐FLAG co‐IP. J) CCT4‐3×FLAG and Myc‐RSV N were co‐expressed with HA‐ZNRDWT, HA‐ZNRDK82A, HA‐ZNRDL42E, or HA‐ZNRDL43E in ZNRD2‐knockout HEK293T cells. K) CCT4‐3×FLAG and Myc‐RSV N were co‐expressed with HA‐ZNRD2WT, HA‐ZNRD2L180E, HA‐ZNRD2L183E, or HA‐ZNRD2I184E in ZNRD2‐knockout HEK293T cells.
Integration of these data with those of previous reports mapped the key residues in ZNRD2 required for interactions with HERC2 (K82), CCT4 (L42, L43), and RSV N (L180, L183, I184). Mutations disrupting the HERC2‐ZNRD2‐CCT4 axis (K82A, L42E, L43E) did not induce CCT4 degradation, while retaining normal RSV N binding and insoluble complex formation, and RSV N did not alter the degradation outcomes (Figure 6J; Figures S4F and S6D,E, Supporting Information). Conversely, oligomerization‐deficient ZNRD2 mutants (L180E, L183E, and I184E) lost RSV N binding and functional responses and unexpectedly impaired the ability to promote CCT4 degradation (Figure 6K), suggesting that these mutations may cause loss of ZNRD2 conformation or function.
In a high‐confidence standalone ZNRD2 structure predicted by AlphaFold3 (Figure S6F, Supporting Information), the two C‐terminal α‐helices fold into an angled alpha hairpin stabilized by a hydrophobic interface where residues L180, L183, and I184 are positioned (Figure S6G, Supporting Information). The substitution of these residues with hydrophilic glutamic acid (E) possibly destabilizes the interface and impairs oligomerization. In contrast, mutations in the smaller hydrophobic valine (V) showed that L180V retained oligomerization, whereas L183V and I184V partially impaired it, possibly due to a leverage effect caused by their location at the distal end of the hairpin turn of the interface. All three valine mutants maintained their interaction with RSV N (Figure S6H,I, Supporting Information). Consistent with this, while the L180E mutation partially impaired binding to CCT4 and HERC2, the L183E and I184E mutations abolished these interactions; nonetheless, all valine variants preserved their binding capabilities (Figure S6J, Supporting Information).
These results indicated that the hydrophobic residues, L180, L183, and I184, contribute to stabilization of the C‐terminal conformation of ZNRD2, possibly acting as a critical structural basis for oligomerization, and that perturbations in its interactions with CCT4 and HERC2 are also caused by oligomerization loss due to conformational changes. Consequently, physiologically latent oligomerization constitutes an essential structure endowing ZNRD2 with a competent conformation and function central to both cellular proteostasis and antiviral defense.
Discussion
3
Similar to that observed in other negative‐sense RNA viruses, NSV performs a critical role in protecting viral genomic RNA and facilitating viral replication. NSV's N encapsidates viral RNA and oligomerizes into compact helical nucleocapsids that shield the genome from innate immune surveillance and evade antiviral responses.^[^ 25, 26, 27, 28 ^]^ Within nsNSVs, N collaborates with phosphoproteins (e.g., RSV P or EBOV VP35) to drive VF formation, establishing shield compartments for viral RNA synthesis.^[^ 26, 35, 60, 61 ^]^ Despite extensive study of N within VFs, the mechanisms governing the biological fate of RSV N oligomers and their interactions with host factors outside these structures that affect the infection outcomes remain largely uncharted.
In the present study, we identified ZNRD2 as a previously unrecognized host factor that interacts with RSV N (Figure 1A; Figure S1A, Supporting Information). ZNRD2 exploits the intrinsic oligomerization propensity of RSV N to restrict viral replication. In particular, ZNRD2 promotes the formation of insoluble N aggregates by enhancing N oligomerization. This sequesters N into a non‐functional state rather than mediating its degradation, rendering it incompatible with productive infection, thereby suppressing viral gene expression and genome replication (Figures 1, 2, and 3A; Figures S1B–H and S2B–F, Supporting Information). Importantly, this process is independent of RNA binding but requires N oligomerization, as monomeric N fails to engage and form insoluble complexes with ZNRD2 (Figure 3B,C; Figure S2G,H, Supporting Information). These data indicate that ZNRD2 acts as a solubility modulator, leveraging the key biophysical vulnerabilities of viral nucleoproteins to inhibit infection.
Oligomerization is essential for the N function and is typically orchestrated by viral partners to promote productive infection.^[^ 26 ^]^ Our results revealed an additional regulatory layer mediated by the host factor, ZNRD2, which redirects RSV N oligomerization toward inert aggregates. This regulation distinguishes RSV from other nsNSVs, as evidenced by the preferential promotion of ZNRD2‐mediated RSV N's insolubility (Figure 2H). Although ZNRD2 exists in a monomeric state under native conditions, a hydrophobic motif (residues L180, L183, and I184) within its C‐terminal domain mediates dormant oligomerization, which is activated upon binding to oligomeric RSV N, acting as a nucleation platform, leading to the formation of an insoluble complex and inhibition of viral replication (Figure 3D–I; Figure S3, Supporting Information).
Although AlphaFold3 provided preliminary structural hypotheses for the ZNRD2‐RSV N complex, its predictions for intrinsically disordered proteins and large macromolecular assemblies exhibited low global confidence scores, possibly inherent to TM‐score calculation methodologies where increased input length disproportionately penalizes pTM values.^[^ 56, 57 ^]^ The considerably repeatable relative orientation between the α13 helix of RSV N and the α1 helix of ZNRD2 across multiple independent predictions renders these models testable hypotheses rather than definitive structures (Figure S4A,B, Supporting Information). The computationally proposed D32‐K351 salt bridge, although biophysically plausible, moderately contributed to the interaction upon mutagenesis (Figure 4C,D; Figure S4F,G, Supporting Information). The Y337A mutation phenocopied the oligomerization‐deficient N^mono^, redefining the causal relationships among oligomerization, binding, and complex formation (Figure 4F–H). Experimental data established oligomerization as a necessary condition for both binding and complex formation, whereas the minor impact of interface salt bridges demonstrated that oligomerization alone was insufficient. Thus, oligomerization and binding represent two interdependent, yet oligomerization‐dominant, determinants that collectively drive the formation of insoluble complexes. This parallels the viral RNP complex assembly, where both N protomer self‐oligomerization and RNA‐binding capacity are strictly co‐dependent and represent both necessary and sufficient conditions for functional RNP formation.^[^ 11, 12, 24, 25, 26, 27, 28 ^]^ The divergence between computational predictions and experimental validation underscores the significance of oligomerization that transcends residue‐specific contacts. Our approach demonstrates how low‐confidence AlphaFold predictions, when coupled with rigorous functional dissection, generate testable paradigms.
In the context of infection, the formation of the insoluble ZNRD2‐RSV N complex is modulated by viral phosphoprotein P. As a molecular chaperone, P maintains N in a monomeric state, while suppressing nonspecific host RNA binding to selectively promote viral RNA encapsidation.^[^ 29, 34, 35, 36, 37, 38, 39, 40, 41 ^]^ The interaction between N and P occurs via two distinct functional modes: one involving RNA‐free monomeric N (N^0^) binding via its CTD to the P NTD (residue 1–40), which induces a conformational change that blocks RNA access and inhibits N oligomerization; and another in which RNA‐bound N engages with the P CTD via a binding pocket in the N NTD, with residue F241 of P serving as a critical anchor point, forming a stable replication‐transcription complex within phase‐separated VFs. Critically, P‐bound N becomes incapable of engaging ZNRD2 as a consequence of an altered oligomerization status rather than direct competitive binding at the interaction interface, as evidenced by the significant enhancement of N solubility when accompanied by P (Figure 5D,E; Figure S5E, Supporting Information). This mechanism accounts for the inability of ZNRD2 to inhibit RSV minigenome activity under optimized transfection conditions (where P is abundant) (Figure S5B,C, Supporting Information), in contrast to that observed under authentic infection conditions. Consequently, ZNRD2 did not exert antiviral activity on the replication and transcription machinery, consistent with its exclusion from VFs (Figure S5A, Supporting Information). Thus, the functional target of ZNRD2 is the N protein, which exists outside the VFs and remains unbound to P. The fluctuations in ZNRD2 solubility during infection (Figure 5A) reflect the dynamic binding between N and P, highlighting the intricate adversarial nature of virus‐host interactions.
In addition to its antiviral role, ZNRD2's interaction with RSV N perturbs its physiological function in cellular proteostasis. ZNRD2 normally bridges the E3 ubiquitin ligase, HERC2, and orphan subunits of the chaperonin CCT complex, which assists in the folding of ≈10% of proteins,^[^ 62 ^]^ mediates the degradation of unassembled subunits to maintain proper stoichiometry, thereby ensuring precise chaperonin assembly.^[^ 51 ^]^ However, sequestration of ZNRD2 into insoluble aggregates by RSV N disrupts this quality control pathway, indirectly stabilizing orphaned CCT subunits that would otherwise undergo degradation (Figure 6). Consequently, overexpressed ZNRD2 forms insoluble aggregates that lose their physiological functions, leading to dominant‐negative effects (Figure S1I–K, Supporting Information).
Insolubility is the direct cause of bidirectional functional suppression between ZNRD2 and RSV N, with oligomerization acting as an essential prerequisite. The three hydrophobic residues (L180, L183, I184) in the C‐terminus of ZNRD2 stabilize the relative spatial configuration of two α‐helices (Figure S6F,G, Supporting Information), possibly constituting the structural framework underlying oligomerization that gives rise to functional interfaces capable of engaging with RSV N, CCT4, and HERC2 (Figure S6H–J, Supporting Information). These findings indicate that oligomerization is indispensable for maintaining ZNRD2's competent conformation and physiological functions, similar to the involvement of RSV N oligomerization in RNP and VF assembly. Therefore, oligomerization emerges as both the mechanistic basis for reciprocal functional suppression between ZNRD2 and RSV N and the architectural platform underpinning their native biological functions. This reveals an evolutionary trade‐off; wherein antiviral defense exploits the inherent vulnerabilities of viral proteins at the cost of partially compromising essential cellular proteostasis.
In summary, our study identified ZNRD2 as a novel antiviral host factor that restricts RSV replication by promoting N aggregation into an insoluble non‐functional complex. This mechanism uniquely exploits the intrinsic oligomerization propensity of RSV N, redirecting it from a replicative scaffold toward toxic aggregates, providing evidence of host‐directed conversion of viral oligomerization into an antiviral defense mechanism. RSV N reciprocally sequesters ZNRD2 into an insoluble complex, thereby disrupting its physiological function as a chaperonin assembly regulator. This functional suppression creates a transient proteostatic imbalance, potentially contributing to RSV pathogenesis. ZNRD2's dual role in orchestrating cellular proteostasis and viral antagonism revealed a sophisticated balance in host defense strategies. Our findings establish protein aggregation as a targeted antiviral paradigm against large viral ribonucleoproteins, reminiscent of the prion‐like phenomena observed in neurodegenerative diseases. Future studies should investigate whether similar mechanisms operate across mononegaviruses and explore therapeutic manipulation of these aggregation pathways to inhibit viral replication while preserving cellular homeostasis.
Experimental Section
4
Cells
HeLa, HEK293T, HEp‐2, BHK‐21, LLC‐MK2, and Vero cell lines were purchased from China Center for Type Culture Collection (CCTCC) and cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% FBS (WISENT) and 1% Pen‐Strep at 37 °C with 5% CO_2_. The ZNRD2 knockout cell lines were generated by cloning gRNA (5’‐ATGGGCGACTATCTGCTGCG‐3’) into the CRISPRv2 vector. Briefly, lentivirus was generated by transfecting HEK293T cells with the constructs of interest, pMD2.G and pSPAX2; the resulting viral particles were then used to infect wild‐type cells. At 48 h post‐infection, puromycin selection was employed (2 µg mL^−1^ for HeLa cells and 1 µg mL^−1^ for HEK293T cells). Cells were passaged under antibiotic pressure for 2–3 generations, followed by monoclonal cell screening. Single‐clone knockout cell lines were validated by Western blotting. Stable ZNRD2 knockdown cell lines were generated by cloning shRNA (#1, 5’‐CTCGACTCAGACGTGGATAAA‐3’, #2, 5’‐GACTCAGACGTGGATAAAGAT‐3’) into pLKO.1 puro vector, a scrambled shRNA (SCR) was used as a negative control. Lentivirus was similarly generated from HEK293T cells and used to infect HeLa cells. Puromycin selection was applied 48 h post‐infection. Stable cell lines expressing HA‐tagged ZNRD2 were generated by cloning the cDNA sequence into either the pHAGE vector (puromycin resistance) or the pCDH vector (blasticidin resistance). Lentivirus was similarly generated from HEK293T cells and used to infect wild‐type cells. Antibiotic selection was applied 48 h post‐infection (1 µg mL^−1^ puromycin for HEp‐2 cells and 10 µg mL^−1^ blasticidin for HeLa cells).
Plasmid transfections were performed using Lipofectamine 2000 (Invitrogen) for HeLa cells and Neofect (Neofect Biotech, Beijing) for HEK293T cells, following the manufacturers’ protocols.
Viruses
RSV (Strain A2) was propagated in HEp‐2 cells and concentrated following established protocols.^[^ 63 ^]^ hPIV3 were propagated in LLC‐MK2 cells. Viral titers were determined by TCID_50_ assay in Vero cells. Recombinant vTF7‐3 expressing T7 RNA polymerase was propagated in BHK‐21 cells. Viral infection experiments were performed in a Biosafety Level 2 (BSL‐2) laboratory at Wuhan University.
Plasmids
All plasmids were constructed using PCR‐based cloning methods and verified by sequencing. The cDNAs encoding viral proteins were generated by reverse transcription of RNA extracted from RSV‐infected HeLa cells. The N and P genes were fused with Myc and FLAG tags at their N‐terminus, respectively, and cloned into the pCAGGS vector. For the RSV minigenome system, the N, P, M2‐1, L genes, and minigenome were cloned without tags into the pGEM4 vector. The cDNAs encoding human proteins (ZNRD2, CCT1‐8) were generated by reverse transcription of RNA extracted from HEK293T cells. The ZNRD2 gene was fused with an HA tag at its N‐terminus, while the CCT gene was fused with a 3×FLAG tag at its C‐terminus, and all were cloned into the pHAGE vector.
Antibodies, Reagents, and Chemicals
Mouse anti‐ZNRD2 monoclonal antibody (OTI2F5) (Invitrogen, MA5‐25288); rabbit anti‐RSV N monoclonal antibody (GeneTex, GTX636711); mouse anti‐RSV P monoclonal antibody (abcam, ab94965); mouse anti‐RSV M2‐1 monoclonal antibody (abcam, ab94805); mouse anti‐FLAG monoclonal antibody (Sigma‐Aldrich, F3040); mouse anti‐HA monoclonal antibody (MBL, M180‐3); mouse anti‐Myc monoclonal antibody (MBL, M192‐3); mouse anti‐GAPDH monoclonal antibody (Abclonal, AC002,); Goat anti‐mouse IgG (H+L) secondary antibody, HRP (Invitrogen, 31430); Goat anti‐rabbit secondary antibody, HRP (Invitrogen, 31460); Goat anti‐Rabbit Secondary Antibody, Alexa Fluor 594 (Invitrogen, A‐11012); Goat anti‐mouse Secondary Antibody, Alexa Fluor 488 (Invitrogen, A‐11029).
Anti‐FLAG Affinity Gel (BioLegend, 651503), anti‐HA (MA‐103), and anti‐Myc (MA‐105) VHH magarose were purchased from ABMagic.
Inhibitors MG132 (HY‐13259), CQ (HY‐17589A), and CHX (HY‐12320) were purchased from MedChemExpress (MCE).
Co‐Immunoprecipitation (co‐IP)
HEK293T cells were seeded in 6‐well plates and transiently transfected with the plasmid of interest, At 48 h post‐transfection, cells were lysed in lysis buffer (50 mm Tris‐HCl, pH 7.4, 150 mm NaCl, 1% Triton X‐100, 0.1% SDS, 1× protease inhibitor cocktail) or Native IP buffer (50 mm HEPES‐KOH, 125 mm KOAc, 2 mm Mg(OAc)2, 1 mm DTT, 0.01% digitonin, 1× protease inhibitor cocktail, Figure S6J, Supporting Information).^[^ 51 ^]^ Lysates were centrifuged at 13 000 rpm for 15 min at 4 °C, and the supernatants were incubated with the corresponding beads overnight at 4 °C. The beads were washed five times with lysis buffer or Native IP buffer and resuspended in 1× SDS loading buffer for subsequent analysis.
IP‐MS
HEK293T cells were seeded in 10 cm dishes and transiently transfected with 10 µg of pCAGGS‐Myc‐RSV N. Lysate preparation and immunoprecipitation using anti‐Myc beads were performed as described above. After washing the beads five times, samples were sent to SpecAlly Life Technology (Wuhan, China) for interaction analysis by mass spectrometry.
Mass spectrometry analysis was performed using a timsTOF Pro ion mobility‐quadrupole‐time‐of‐flight mass spectrometer (Bruker), with sample injection and separation conducted via an online‐coupled UltiMate 3000 RSLCnano liquid chromatography system. Mass spectrometry data were processed using MaxQuant (version 2.0.1) with the Andromeda search algorithm. The search was performed against two protein reference databases: the Add database (version 23021405, containing 2 protein sequences) and the UniProt Human proteome database (release 2022‐03‐29, containing 20 377 protein sequences). Proteins with a fold change > 4 and a p < 0.05 between experimental and control samples were identified as potential interaction partners of the bait protein.
Solubility Fractionation Assay
Cells were lysed in NP‐40 lysis (50 mm Tris‐HCl, pH 7.4, 150 mm NaCl, 1% NP‐40, 1× protease inhibitor cocktail) on ice for 30 min. Lysates were spun at 20 000 g for 10 min at 4 °C. The supernatant was collected as the soluble fraction. The pellet was washed twice with cold PBS and resuspended in 1% SDS buffer (1% SDS and 0.1% Triton X‐100 in PBS) to obtain the insoluble fraction. Both fractions were mixed with SDS loading buffer and denatured at 95 °C for 10 min prior to western blotting analysis.
Oligomerization Assay
Cells were lysed in NP‐40 lysis buffer on ice for 30 min. Lysates were centrifuged at 3840 g for 15 min at 4 °C. The supernatant was collected as the monomeric fraction. The pellet was washed twice with cold PBS and resuspended in PBS containing 5 mm disuccinimidyl suberate (DSS) for crosslinking at 37 °C for 30 min. After crosslinking, samples were centrifuged at 3840 g for 5 min at room temperature. The supernatant was carefully removed, and the pellet, representing the oligomeric fraction, was resuspended in 1× SDS loading buffer and denatured at 60 °C for 15 min. Oligomerization was analyzed by Western blot.
Western Blot
Cells were lysed in lysis buffer followed by centrifugation or 1% SDS buffer without centrifugation (Figure 6; Figure S6, Supporting Information), then denatured with SDS loading buffer at 95 °C for 10 min. Samples were separated by Tris‐Glycine SDS‐PAGE and transferred to a 0.45 µm nitrocellulose membrane using electrotransfer. Membranes were blocked with 5% non‐fat dry milk in PBST (PBS containing 0.1% Tween‐20) at room temperature for 30 min and then incubated with the appropriate primary antibodies at room temperature for 2 h or at 4 °C overnight. After washing with PBST, membranes were incubated with HRP‐conjugated secondary antibodies for 1 h at room temperature, followed by extensive washing with PBST. Protein signals were detected using enhanced chemiluminescence (ECL) and visualized with a chemiluminescence imaging system.
Immunofluorescence (IF)
Cells were seeded on coverslips in 24‐well plates. At the indicated time points, coverslips were washed with cold PBS and fixed with 4% PFA at room temperature for 20 min. After fixation, coverslips were washed three times with PBS and permeabilized with 0.2% Triton X‐100 (in PBS) at room temperature for 20 min. Following permeabilization, coverslips were washed and blocked with 3% BSA (in PBS) at room temperature for 30 min. Primary antibodies were applied and incubated at room temperature for 2 h or at 4 °C overnight. After washing, coverslips were incubated with fluorophore‐conjugated secondary antibodies at room temperature for 1 h, followed by incubation with DAPI at room temperature for 10 min. After thorough washing, coverslips were sealed onto the glass slide using ProLong Diamond Antifade Mountant (Invitrogen). Images were acquired using a Leica SP8 confocal microscope.
RSV Minigenome Assay
HeLa cells seeded in 24‐well plates were co‐transfected with 31.25 ng pGEM4‐RSV N, 62.5 ng pGEM4‐RSV P, 15.625 ng pGEM4‐RSV M2‐1, 7.8125 ng pGEM4‐RSV L, 31.25 ng pGEM4‐RSV Minigenome (Firefly luciferase), and 125 ng pHAGE‐HA‐ZNRD2. The thymidine kinase promoter‐driven Renilla luciferase plasmid (pRL‐TK) was included as a transfection control. The minigenome system was driven by vTF7‐3 expressing T7 RNA polymerase. At 24 h post‐transfection, cells were lysed with Passive Lysis Buffer (Promega) at room temperature for 15 min. RSV minigenome activity was quantified using the Dual‐Luciferase Reporter Assay System (Promega). Subsequently, system protein expression was analyzed by Western blotting.
RNA Extraction and Quantitative PCR (qPCR)
Total RNA was extracted from cells using TRIzol reagent (Invitrogen) following the manufacturer's protocols. Strand‐specific reverse transcription of mRNA and viral genomic RNA was performed using the oligo d(T)23 and RSV N gene forward primer, respectively, with 1 µg of RNA and the ABScript II cDNA First‐Strand Synthesis Kit (Abclonal). The cDNA products were diluted 5‐fold and used as templates for qPCR with the 2× Universal SYBR Green Fast qPCR Mix (Abclonal) following the manufacturer's protocols. GAPDH was used as a cellular control gene for normalization. The following primers were used:
RSV NS1 forward: 5’‐GCTTTGGCTAAGGCAGTGAT‐3’
RSV NS1 reverse: 5’‐ACTGGCATTGTTGTGAAATTGGA‐3’
RSV N forward: 5’‐CTATGGTGCAGGGCAAGTGA‐3’
RSV N reverse: 5’‐GAATCCTGCTTCACCACCCA‐3’
RSV L forward: 5’‐TGTCTGTGATGCCGAATTGTC‐3’
RSV L reverse: 5’‐TCGCAGGACCTATTGTAAGGAC‐3’
ZNRD2 forward: 5’‐CCCTGAACGGAGCTGAAGTC‐3’
ZNRD2 reverse: 5’‐GGGCATTCAGAGCGGGATTA‐3’
GAPDH forward: 5’‐GTCAAGGCTGAGAACGGGAA‐3’
GAPDH reverse: 5’‐AAATGAGCCCCAGCCTTCTC‐3’
Structural Modeling and Analysis of the ZNRD2‐RSV N Complex
Five independent models were generated for the RSV N decamer‐ZNRD2 complex using AlphaFold3. All models exhibited consistent positioning of the ZNRD2 N‐terminal domain relative to the RSV N protomer interface, considering a convergent binding mode. Structural analysis and figure preparation were performed using ChimeraX (version 1.9). Residue pairs with spatial distances < 4 Å between ZNRD2 and RSV N were defined as core interaction sites.
Quantification and Statistical Analysis
Statistical analyses and graphical representations were performed using GraphPad Prism version 9.5. Data are presented as mean ± standard deviation (n = 3). Unpaired two‐tailed Student's t‐tests were used for comparisons between two groups with a single independent variable. One‐way ANOVA with Dunnett's test was applied for comparisons involving multiple groups with a single independent variable. Two‐way ANOVA with Šídák's multiple comparisons test was performed for comparisons between two groups with multiple independent variables. A p‐value of less than 0.05 was considered statistically significant.
Conflict of Interest
The authors declare no competing interests.
Author Contributions
H.Z. performed the experiments. J.D. performed the AlphaFold3 modeling. H.Z. and M.C. designed the experiments and wrote the paper. H.Z., M.H., J.D., C.L., W.W., Y.H., and Z.L. analyzed the results. Z.L. improved the language of the manuscript. Y.Q. and M.C. supervised the study. All authors discussed the results and approved the manuscript.
Supporting information
Supporting Information
Supplemental Table 1
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1N. I. Mazur , M. T. Caballero , M. C. Nunes , Lancet 2024, 404, 1143.39265587 10.1016/S 0140-6736(24)01716-1 · doi ↗ · pubmed ↗
- 2Y. Li , X. Wang , D. M. Blau , M. T. Caballero , D. R. Feikin , C. J. Gill , S. A. Madhi , S. B. Omer , E. A. F. Simões , H. Campbell , A. B. Pariente , D. Bardach , Q. Bassat , J.‐S. Casalegno , G. Chakhunashvili , N. Crawford , D. Danilenko , L A Ha Do , M. Echavarria , A. Gentile , A. Gordon , T. Heikkinen , Q. S. Huang , S. Jullien , A. Krishnan , E. L. Lopez , J. Markic , A. Mira‐Iglesias , H. C. Moore , J. Moyes , et al., Lancet 2022, 399, 2047.35598608 10.1016 · doi ↗ · pubmed ↗
- 3A. Mac Neil , M. Mc Morrow , Lancet Glob. Health 2024, 12, 1563.10.1016/S 2214-109X(24)00289-4PMC 1187406939216501 · doi ↗ · pubmed ↗
- 4S. M. Brunwasser , B. M. Snyder , A. J. Driscoll , D. B. Fell , D. A. Savitz , D. R. Feikin , B. Skidmore , N. Bhat , L. J. Bont , W. D. Dupont , P. Wu , T. Gebretsadik , P. G. Holt , H. J. Zar , J. R. Ortiz , T. V. Hartert , Lancet Respir. Med. 2022, 8, 795.10.1016/S 2213-2600(20)30109-0PMC 746459132763206 · doi ↗ · pubmed ↗
- 5M. N. Billard , L. J. Bont , Lancet 2023, 401, 1632.37086746 10.1016/S 0140-6736(23)00672-4 · doi ↗ · pubmed ↗
- 6B. Rima , P. Collins , A. Easton , R. Fouchier , G. Kurath , R. A. Lamb , B. Lee , A. Maisner , P. Rota , L. Wang , ICTV Report Consortium , J. Gen. Virol. 2017, 98, 2912.29087278 10.1099/jgv.0.000959 PMC 5775899 · doi ↗ · pubmed ↗
- 7L. Liljeroos , M. A. Krzyzaniak , A. Helenius , S. J. Butcher , Proc. Natl. Acad. Sci. USA 2013, 110, 11133.23776214 10.1073/pnas.1309070110 PMC 3703984 · doi ↗ · pubmed ↗
- 8M. B. Battles , J. S. Mc Lellan , Nat. Rev. Microbiol. 2019, 17, 233.30723301 10.1038/s 41579-019-0149-x PMC 7096974 · doi ↗ · pubmed ↗