Ethics statement

All animal experiments were conducted in strict accordance with the guidelines outlined in the Regulations on the Administration of Laboratory Animals by the State Scientific and Technological Commission of the People’s Republic of China. The Yangzhou University Experimental Animal Ethics Committee (Permission number: 202302083) approved all of the animal studies according to the Administrative Measures of Jiangsu Province on laboratory Animals of Jiangsu Administrative Committee of Laboratory Animals.

Viruses, animals, cells, and antibodies

NDV strains including genotype II (La Sota), IV (Herts/33), and IX (F48E8) were provided by our laboratory. Specific pathogen-free (SPF) chicken embryos and chickens were purchased from Beijing Merial Company (Beijing, China) and Jinan Speifului Poultry Technology Co., Ltd (Shandong, China), respectively. The biological characteristics of NDV strains were shown in S1 Table. The viruses were propagated in the allantoic cavity of 10-day-old SPF chicken embryos. Following a 48-hour (h) incubation period, the allantoic fluid was collected and preserved at −70°C for subsequent analysis and characterization. Human lung epithelial (A549) cells, chicken macrophage (HD11) cells, human cervical cancer epithelial (HeLa) cells, chicken tracheal epithelial (CTE) primary cells, and chicken embryo fibroblasts (CEFs) were preserved and cultured in our laboratory. The half-maximal tissue culture infective dose (TCID₅₀) of NDV was subsequently determined. To achieve inefficient viral replication, viral suspensions were irradiated with ultraviolet (UV) light as described previously. Viral inactivation was then confirmed by validating the lack of replication in 10-day-old chicken embryonated eggs and in CEF monolayer cultures. The antibodies used in this study were as follows: mouse anti-NDV NP (provided by our laboratory), rabbit anti-OCLN, rabbit anti-ZO-1, rabbit anti-CLDN1 (CST, USA), rabbit anti-MLC, rabbit anti-p-MLC (Affinity, China), mouse anti-vimentin (Proteintech, China), Actin-Tracker Red-594 (Beyotime, China), anti-mouse or anti-rabbit HRP IgG (CWBIO, China), anti-mouse or anti-rabbit FITC (TransGen, China) or AF594 IgG (Bioss, China).

Cytopathic observation and cell viability

Cytopathic effects were observed utilizing a cellular morphology assay under light microscopy. Briefly, cells seeded in 6-well plates were either mock-infected or inoculated with NDVs at a multiplicity of infection (MOI) of 1. The morphological changes in each treatment group were monitored at specified time points using a microscope at 100 × magnification. Cell viability was assessed using the Cell Counting Kit-8 (CCK-8, Beyotime Biotech, China) according to the manufacturer’s instructions. For this purpose, cells were cultured in 96-well plates and infected with NDVs at an MOI of 1 for defined durations. Post-infection, the cells were maintained at 37°C for 1 h after the addition of 10 µL of CCK-8 reagent. Subsequently, the absorbance at 450 nm for each well was quantified using an enzyme-linked immunosorbent assay (ELISA) reader.

Transwell assay

Cell migration was assessed using a 24-well Transwell assay (Corning Incorporation, Corning, NY, USA). Cells seeded in 6-well plates were infected with NDV at an MOI of 0.1 for either 6 or 24 h. Subsequently, the infected cells were detached using trypsin. A suspension of quantitatively prepared cells (detailed preparation method described in the respective figure legend) in 200 µL of serum-free medium was then added to the upper chamber of the Transwell insert (pore size, 8 µm). The lower chamber was filled with 500 µL of complete culture medium supplemented with 20% fetal bovine serum (FBS) to create a chemoattractant gradient. After a 36-hour incubation at 37°C, the membrane was washed with PBS to remove non-migratory cells remaining on the upper surface. Migratory cells that had transmigrated to the underside of the membrane were fixed with 4% paraformaldehyde. Following fixation, these migrated cells were stained with crystal violet. For quantification, five randomly selected fields per membrane were imaged under an inverted light microscope, and the number of migratory cells in each field was counted [33].

Transepithelial electrical resistance (TEER) assay

A549 cells were cultured to 90% confluence and dissociated into a single-cell suspension using trypsin. The suspension was plated at 2 × 10⁵ cells/500 μL in the upper compartment of a transwell insert, with 1.5 mL of fresh medium in the lower compartment. Cells were allowed to adhere and form a monolayer barrier. NDV was then added at 0.1 MOI, with a control group untreated. The TEER measurement (Millicell, Germany) was taken at specific time points post-infection to assess barrier function. Resistance per unit area was calculated as “Resistance per unit area = Resistance (Ω) × Effective membrane area (cm²)”.

Western blot assay

Cells were lysed in RIPA buffer (Beyotime Biotech, China) supplemented with the protease inhibitor PMSF (Beyotime Biotech, China) at the indicated time. The protein content of the lysates was quantified utilizing the BCA Protein Assay Kit (Beyotime Biotech, China). Subsequently, the proteins were denatured and fractionated via 10% SDS-PAGE, followed by transferring onto polyvinylidene difluoride (PVDF) membranes. After blocking, the membranes were subjected to incubation with appropriately diluted primary and secondary antibodies. Protein detection was achieved by treating the membranes with a chemiluminescent substrate and imaging using a chemiluminescence imager (Tanon, China) in a darkroom environment. Lastly, the optical density of the protein bands was quantified using ImageJ version 1.48v software (National Institutes of Health, USA).

Plasmid construction and transfection

To generate plasmids expressing His-tagged viral NP, P, V, W, M, F, and HN proteins, as well as a Flag-tagged L protein, the coding sequences for each gene were amplified using gene-specific primer pairs and cDNA derived from the F48E8 strain (GenBank Accession Number: FJ436302) as the template, with sequences provided in S2 Table. These amplified fragments were subsequently cloned into the XbaI/BamHI or EcoRI/XbaI restriction sites of the pcDNA3.1-His/Flag vector using the Basic Seamless Cloning and Assembly Kit (TransGen, China), following the manufacturer’s protocol. Successful plasmid construction was confirmed via DNA sequencing. All plasmids were named pcDNA3.1-gene name-tag, including pcDNA3.1-NP-His, pcDNA3.1-P-His, pcDNA3.1-M-His, pcDNA3.1-F-His, pcDNA3.1-HN-His, pcDNA3.1-V-His, pcDNA3.1-W-His, and pcDNA3.1-L-Flag.

For transfection experiments, cells were seeded to achieve 70% confluency prior to transfection with the respective plasmids, employing the TransIntro EL Transfection Reagent (TransGen, China) according to the manufacturer’s instructions.

RNA interference and pharmacological treatment

Small interfering RNAs (siRNAs) targeting TJ genes was synthesized by GenePharma (China), with sequences provided in S3 Table. A non-targeting siRNA (siRNA-NC) was employed as a negative control. Cells cultured in 24-well plates were transfected with 60 pmol of siRNA using 2 μL of EL Transfection Reagent (TransGen Biotech, China), following the transfection protocol provided by the manufacturer. The efficiency of siRNA transfection was confirmed by western blot analysis at 24 hours post-transfection (hpt). Thereafter, cells exhibiting successful gene knockdown were selected for subsequent experimental investigations.

Cells were infected with NDV at the indicated dose at 37°C. Following a 1-h adsorption period, the cells were rinsed with PBS and subsequently cultured in DMEM supplemented with 1.5% 3,3’-iminodipropionitrile (IDPN) (Sigma-Aldrich, USA), 20 μM Z-VAD-FMK, 20 μM MG132 (Beyotime, China), and 1 μM Bafilomycin A1 (TOPSCIENCE, China) for the duration indicated. Furthermore, cells were treated with 10 μM ajoene (GLPBIO, USA) for 12 h prior to infection with NDV at the indicated dose for the specified duration. Subsequent to the incubation period, the cells were prepared for further analysis.

Confocal microscopy assay

Cells grown on 14-mm coverslips were treated with siRNAs or pharmacological agents prior to infection with NDV at the indicated dose for the specified durations. Subsequent to the infection period, the cells were washed with PBS, fixed with 4% paraformaldehyde, permeabilized using Triton X-100 (Beyotime Biotech, China), and blocked with 5% bovine serum albumin (BSA). The primary antibody was then added and incubated overnight at 4°C, followed by a 1-h incubation at 37°C with either FITC-conjugated or AF594-conjugated secondary antibodies. Afterward, nuclear staining was performed with DAPI (Beyotime Biotech, China), and the coverslips were mounted using PBS supplemented with 50% glycerol. The prepared samples were finally visualized using a confocal microscope (Leica, Germany).

Transmission electron microscopy (TEM) assay

Cells were infected with NDV at an MOI of 0.1 for 6 h. Following infection, the cells were removed from the culture medium, subjected to a PBS wash, and then incubated with electron microscopy fixative at room temperature in the dark for a duration of 5 min. The cells were gently detached in a unidirectional manner using a cell scraper and aspirated into a centrifuge tube. Following centrifugation at 2000 rpm for 5 min, the cell sediment was resuspended in fresh electron microscopy fixative. An additional 30-min fixation was conducted at room temperature in the dark, after which the sample was transferred to 4°C for preservation. The prepared cells were subsequently sent to Servicebio (Wuhan, China) for TEM examination and imaging.

Animal experiments

NDVs were inoculated into six 4-week-old SPF chickens at a dose of 10⁵ 50% egg infectious dose (EID₅₀)/0.1mL per chicken via the intranasal and intraocular routes, respectively. Control chickens received an equivalent volume of PBS. At 4 days post-inoculation (dpi), chickens from each group were humanely euthanized, and lung tissue was harvested to assess histopathological alterations. The lung samples were fixed with neutral formalin solution, processed through a series of dehydration and paraffin embedding steps, sectioned, baked, dewaxed, and stained with hematoxylin and eosin. Following the preparation of histological sections, the histopathological modifications were examined and documented using a light microscope. Furthermore, 0.3 g of lung tissue was homogenized in RIPA, followed by centrifugation. The supernatant was then subjected to determine the protein expression of OCLN and ZO-1. Additionally, oropharyngeal and cloacal swabs were gathered daily from the infected chickens to evaluate virus shedding. The swab samples were meticulously scrubbed and subjected to centrifugal force. Subsequently, they were used to inoculate 9-day-old SPF chicken embryos. At 96 hpi, allantoic fluid was harvested and evaluated using the hemagglutination (HA) assay for analysis.

RNA extraction and real‑time quantitative PCR (RT‑qPCR)

For RNA extraction and subsequent cDNA synthesis, TransZol Up Reagent (TransGen Biotech, China) and PrimeScript RT Reagent (Takara, China) were employed, respectively. The mRNA expression was quantified using reverse transcription quantitative polymerase chain reaction (RT-qPCR). To assess the amplification of target genes, 2 μL of cDNA from each sample was amplified using the SYBR Green dye-based method (Vazyme, China), adhering to the manufacturer’s instructions. Relative gene expression was determined by the 2^−ΔΔCT method. The results are presented as the fold change (2^−ΔΔCT) in log10 relative to the control group.

To determine the expression of TJ genes, cells in 12-well plates were subjected to mock infection or inoculated with 1 MOI of NDV for 24 h. To measure the levels of inflammatory cytokines, cells treated with IDPN or ajoene were subjected to NDV infection for 24 h, and 0.3g of above-mentioned lung tissue was harvested at 4 dpi. Subsequently, RNA was extracted from the treated cells and lung tissue, followed by reverse transcription to cDNA. The mRNA expression levels were then assessed using qPCR. The TJ and inflammatory genes encompassed OCLN, ZO-1, IL-1β, IL-6, IL-18, and TNF-α, with the primer sequences detailed in S4 Table.

Immunofluorescence assay (IFA)

Cells treated with specific siRNAs or pharmacological agents were exposed to NDVs at the indicated dose for the specified duration. Post-infection, the cells underwent a series of processing steps including PBS washing, fixation with 4% paraformaldehyde, permeabilization, and blocking with 5% BSA. Subsequently, the cells were subjected to an overnight incubation at 4°C with an anti-NP primary antibody, followed by a 1-hour incubation at 37°C with a secondary antibody. Fluorescence microscopy was employed to visualize and photograph the cells, with nuclear staining performed using DAPI. For quantitative analysis, fifteen random lesions per treatment group were chosen, and the extent of viral spread was determined by counting the number of cells within each lesion.

Plaque assay

Cells were infected with NDV at a low dose (0.00001 MOI) at 37°C for 1 h. Subsequently, the cells were rinsed with PBS to eliminate any unattached viruses and overlaid with a 3% low-melt agarose gel. Once the overlay agar had solidified, the plate was inverted and maintained at 37°C for incubation. Following 60 h of incubation, the cells were subjected to staining with crystal violet (Beyotime, China). The relative plaque size was quantified using ImageJ software.

Statistical analysis

The data are presented as the means ± standard deviations of three independent replicates derived from a representative experiment. Either one-way or two-way analysis of variance (ANOVA) was utilized to evaluate the statistically significant differences, with a significance threshold set at p < 0.05. The statistical analyses were conducted with GraphPad Prism 7.00 (San Diego, USA).

NDV disrupts the TJ homeostasis with the reduced expression of TJ-associated proteins

We first established an NDV infection model in the A549 cell line and monitored cellular morphology at various time intervals post-infection. The control and avirulent NDV LaSota groups maintained characteristic cell morphology throughout the observation period. In contrast, infection with the virulent NDV strains Herts/33 and F48E8 resulted in pronounced cellular damage, characterized by enhanced intercellular spaces, vacuolation, cell rounding, aggregation, and even disintegration. Notably, the F48E8 strain elicited the most pronounced degree of cellular damage across all infected groups, with the significant cellular lesion observed at 12 hpi (Fig 1A). Tight junction is the main mode of connection between epithelial cells, and the findings of cellular damage imply that NDV infection may affect intercellular tight junction integrity. In general, impairment of tight junction integrity can enhance cellular migration [11]. Here, we employed a transwell assay to determine the cell migration ability induced by NDV (Fig 1B). The results showed that both the virulent NDV strains Herts/33 and F48E8 significantly enhanced cell migration (Herts/33: 180 cells on average; F48E8: 242 cells on average) compared to the La Sota and mock groups (La Sota: 108 cells on average; Mock: 87 cells on average) at 6 hpi (Fig 1C). Interestingly, the migratory cell count for these virulent strains had notably decreased at the later stage of infection (24hpi, Herts/33: 36 cells on average; F48E8: 21 cells on average; La Sota: 124 cells on average; Mock: 94 cells on average) (S1 Fig). This observation could be attributed to the fact that compromised TJs in the early phase of infection facilitated cellular migration, whereas the extensive cell death occurring in the later stages of infection leads to a diminished migratory rate. Similar to the results of cellular pathology, the F48E8 strain demonstrated the most substantial induction of cell migration at the early stages of infection. TJs serve as the primary structural components responsible for forming the intercellular barrier. To evaluate the impact of NDV on lung epithelial barrier permeability, an in vitro model was constructed using A549 cells. TEER readings reflect the permeability barrier formed by tight junctions in cultured cells. A decrease in TEER values points to possible disruption of these cell-cell connections [34]. Here, TEER values showed a significant decrease from 6 to 24 hpi following infection exclusively with the virulent NDV, particularly the F48E8 strain, when compared to controls, indicating that the virulent NDV infection exacerbates barrier injury (Fig 1D). Therefore, these findings suggest that NDV infection can disrupt the intercellular TJ integrity.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. NDV disrupts the TJ integrity and diminishes the expression levels of TJ-associated proteins.

(A) The cellular morphology was monitored in NDV- and mock-infected A549 cells at an MOI of 1 from 6 to 24 hpi under light microscopy. Scale bar: 20 μm. (B) Transwell assay was employed to determine the cell migration. The transwell model was established that the cells in serum-free medium were introduced into the upper chamber, and the lower chamber was then supplemented with complete culture medium supplemented with 20% fetal bovine serum to serve as a migratory stimulus. Created in BioRender. Lu, D. (2025) https://BioRender.com/hx6j612. (C) Transwell assay was employed to determine the cell migration in NDV- and mock-infected A549 cells at an MOI of 0.1 at 6 hpi. A suspension containing 4 × 10⁴ cells was utilized for the transwell assay. Scale bar: 50 μm. The quantitative assessment of the migratory cell population. The infected groups were compared to the mock group, and statistical analysis was carried out. (D) The detection of barrier permeability following NDV infection. In vitro model of the lung epithelial barrier infected with NDVs. NDVs infected the lung epithelial barrier for 6h, 12 h, 18 h, and 24 h, and the TEER values were determined. Created in BioRender. Lu, D. (2025) https://BioRender.com/o7k6m61. (E, F) The expression levels of TJ-associated proteins OCLN and ZO-1 were assessed by western blotting across a range of NDV doses (0, 0.1, 1, and 5 MOI at 24 hpi) and at various time points post-infection (1 MOI at 12 and 24 hpi). The gray value of each protein was quantified by Image J and normalized to ACTB. The gray values of 0 MOI group and control group were considered as “1”, respectively. (G) A549 cells were infected with NDV at 1 MOI and then treated with Z-VAD-FMK (20 µM), Bafilomycin A1 (1 μM), MG132 (20 µM). At 24 hpi, the cells were harvested for western blotting to detect OCLN and ZO-1 protein expression. The gray value of each protein was quantified by Image J and normalized to ACTB. The gray value of control group was considered as “1”. (H) A549 cells were transfected with eukaryotic expression plasmids encoding each viral protein. At 36 hpt, the expression levels of OCLN, ZO-1, and the viral proteins were assessed. The gray value of each protein was quantified by Image J and normalized to ACTB. The gray values of mock group were considered as “1”. Red boxes indicate the positions of the respective viral proteins in the electrophoresis gel. ** P < 0.01; **** P < 0.0001.

https://doi.org/10.1371/journal.ppat.1013458.g001

To confirm the direct impact of NDV infection on TJs, we assessed expression levels of TJ-associated proteins. Here, we observed that both virulent NDV strains significantly reduced the protein expression levels of ZO-1 and OCLN in A549 cells, whereas no appreciable changes were detected in the La Sota-vaccinated and mock-infected control groups. Notably, this downregulation of ZO-1 and OCLN by virulent NDV strains exhibited both time- and dose-dependent characteristics, with the F48E8 strain demonstrating a more pronounced effect compared to the other virulent strain (Fig 1E and 1F). We further evaluated the broad-spectrum regulatory effects of NDV on TJ proteins across multiple cell lines (HeLa, HD11, CTE) by establishing infection models. Consistent with findings in A549 cells, only the virulent F48E8 strain induced significant cellular damage, whereas avirulent NDV La Sota and controls did not (S2A–S2C Fig). Western blot assays revealed stable expression of ZO-1 and OCLN in all cell lines tested. Upon infection, virulent F48E8 markedly reduced TJ protein levels in HD11 and CTE cells compared to La Sota and controls. Notably, NDV exhibited a significantly weaker ability to downregulate ZO-1 and OCLN in HeLa cells (S2D–S2F Fig). CLDN1 is a classic TJ protein, frequently used alongside ZO-1 and OCLN as markers in TJ research. In this study, we detected CLDN1 expression in these cells and found it was only present in A549 and CTE cells. Interestingly, CLDN1 was undetectable in both HeLa and HD11 cells. This finding aligns with previous reports confirming the absence of CLDN1 in HeLa cells [35], and notably, our study provides the first evidence that HD11 cells also lack CLDN1 expression. Furthermore, our results indicated that virulent NDV significantly downregulated CLDN1 expression in A549 and CTE cells, whereas avirulent NDV did not exert a significant effect (S3 Fig). Based on these results, we selected stably expressed ZO-1 and OCLN, and stable cell lines A549 and HD11 (from different species), for further study. Intriguingly, qPCR analysis showed that NDV infection did not decrease the mRNA levels of these TJ genes, and infection with virulent strains even upregulated them (S4 Fig). These findings suggest that NDV-mediated suppression of TJ protein expression operates at the translational level, not the transcriptional level. Caspase-hydrolase, autophagy-lysosome, and ubiquitin-proteasome constitute the principal protein degradation pathways [36]. To investigate the pathway by which NDV degrades TJ proteins, A549 cells were infected with the virulent NDV and subsequently treated with Z-VAD-FMK (a caspase inhibitor), Bafilomycin A1 (an autophagy inhibitor), and MG132 (a ubiquitin-proteasome inhibitor). Western blot analysis revealed that treatment with all three inhibitors—Z-VAD-FMK, Bafilomycin A1, and MG132—led to a restoration of ZO-1 protein levels; however, only Z-VAD-FMK treatment could significantly restore the expression of OCLN. These findings suggest that infection of the virulent NDV can modulate the degradation of TJ protein via diverse pathways, with the caspase-hydrolase pathway exerting a predominant influence (Fig 1G). To assess the impact of NDV viral proteins on TJs, we examined the expression levels of TJ proteins, ZO-1 and OCLN, following transfection. We initially constructed plasmids expressing eight NDV proteins: six structural (NP, P, M, F, HN, L) and two non-structural (V, W). Western Blot analysis confirmed successful transfection and expression of all viral proteins. Notably, distinct viral proteins exerted differential effects on ZO-1 and OCLN. Specifically, NP protein significantly downregulated OCLN expression, whereas F and HN proteins significantly reduced ZO-1 levels. These findings suggest that NDV’s regulation of TJs is not mediated by a single protein but involves a complex, synergistic interplay among multiple viral proteins (Fig 1H). Therefore, virulent NDV disrupts TJs by targeting OCLN and ZO-1, a process involving complex degradation pathways and interactions between viral proteins.

Reduced expression of TJ proteins facilitates both viral replication and spread of NDV

NDV reduces the expression of TJ proteins, a process previously implicated in the replication cycles of diverse viruses [37,38]. To confirm whether the TJ proteins exerted a regulatory influence on viral replication, we next assessed the effect of TJ protein knockdown on NDV replication. Considering the substantial impact of the virulent F48E8 strain on the TJ homeostasis, we were poised to utilize F48E8 for subsequent investigation. We employed siRNA-OCLN and siRNA-ZO-1 to generate A549/OCLN-KD and A549/ZO-1-KD cells, respectively. Western blot assay confirmed significant knockdown of these TJ genes (Fig 2A and 2B). Concurrently, the CCK-8 assay was conducted to verify that the reduction of OCLN and ZO-1 levels did not affect cell viability (Fig 2C and 2D). To examine the impact of OCLN and ZO-1 on viral replication, A549/OCLN-KD and A549/ZO-1-KD cells were infected with NDV. The viral load in cells and supernatants was quantified using western blot, IFA, and TCID₅₀ assays, respectively. Western blot (Fig 2E and 2F) and IFA (Fig 2G and 2H) assays collectively showed higher viral NP protein in OCLN or ZO-1 knock-down infected cells than in control infected cells. TCID₅₀ assays revealed that viral titers in the supernatant of cells with OCLN or ZO-1 knockdown were markedly elevated compared to the control infection group at 12 and 24 hpi. However, no significant difference in titer was observed at 36 hpi, potentially due to the saturation of viral replication during the late stage of infection (Fig 2I and 2J). Macrophages are one of the susceptible target cells for NDV, and we also the impact of TJ proteins on viral replication in avian HD11 cells. We employed siRNA-OCLN and siRNA-ZO-1 to generate HD11/OCLN-KD and HD11/ZO-1-KD cells, respectively. Western blot confirmed substantial knockdown of TJ genes, and the CCK-8 assay confirmed that decreased OCLN and ZO-1 levels did not impact cell viability (S5A and S5B Fig). To examine the impact of OCLN and ZO-1 on viral replication, HD11/OCLN-KD and HD11/ZO-1-KD cells were infected with NDV. The viral load in cells and supernatants was quantified using western blot and TCID₅₀ assays, respectively. Western blot and TCID₅₀ assays indicated significantly enhanced viral replication in NDV-infected HD11/OCLN-KD and HD11/ZO-1-KD cells versus HD11/ctrl cells (S5C and S5D Fig).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Downregulation of TJ proteins facilitates viral replication and spread.

siRNA-OCLN and siRNA-ZO-1 (60 pmol) were employed to generate A549/OCLN-KD and A549/ZO-1-KD cells, respectively. siRNA-NC transfected cells were recognized as the control A549/ctrl. (A, B) Western blot assay determined the protein levels of OCLN and ZO-1 following siRNA transfection. (C, D) CCK-8 assay determined the cell viability following siRNA transfection. The transfected groups were compared to the control group, and statistical analysis was carried out. (E, F) A549/ctrl, A549/OCLN-KD and A549/ZO-1-KD cells were infected with NDV at 1 MOI for 24 h. The viral replication in cells was evaluated by western blotting. The gray value of each protein was quantified by Image J and normalized to ACTB. The gray value of NDV-infected A549/ctrl group was considered as “1”. (G, H) A549/ctrl, A549/OCLN-KD and A549/ZO-1-KD cells were infected with NDV at 0.1 MOI for 24 h. The viral replication in cells was evaluated by IFA. Fluorescent intensities were quantified using Image J. Scale bar: 50 μm. (I, J) A549/ctrl, A549/OCLN-KD and A549/ZO-1-KD cells were infected with NDV at 0.1 MOI for 12, 24, and 36 h. The viral load in supernatants was detected by TCID₅₀. (K) A549/ctrl, A549/OCLN-KD and A549/ZO-1-KD cells were infected with NDV at a low dose (MOI = 0.00001) for 24 and 36 h, and the foci in cells were evaluated by IFA. Fifteen random lesions per treatment group were chosen, and the extent of viral dissemination was determined by counting the number of cells within each lesion. Scale bar: 50 μm. (L) A549/ctrl, A549/OCLN-KD and A549/ZO-1-KD cells were infected with NDV at a low dose (MOI = 0.00001) within the agarose gel for 60 h. The relative plaque size was quantified using ImageJ software. * P < 0.05; ** P < 0.01; *** P < 0.001; **** P < 0.0001; ns, no significant difference.

https://doi.org/10.1371/journal.ppat.1013458.g002

TJ proteins have been demonstrated to be involved in the spread of virus via cell-to-cell contact [39]. Here, we hypothesized that TJ proteins enhanced NDV replication by facilitating its spread. To verify this hypothesis, we observed foci and plaques forming separately after NDV infection at a low MOI. The extent of cellular foci formation serves as an important indicator of viral spread leading to cytopathic effect, as observed with various viruses including hazara orthonairovirus (HAZV) [30], human parainfluenza virus (HPV) [31], etc. Foci formation results showed that NDV-infected A549/ctrl cells were observed as foci consisting of multiple cells at 24 hpi, and they subsequently expanded at 36 hpi. By contrast, viral foci were observed in A549/OCLN-KD and A549/ZO-1-KD cells significantly at 24 and 36 hpi. The size of viral foci was greater than those observed in A549/ctrl cells. The cell numbers within the viral foci observed in these cells were counted, when the remarkable difference was seen at 24 and 36 hpi. By 24 hpi, cell counts within viral foci in A549/OCLN-KD and A549/ZO-1-KD cells (each averaging 12 cells) was markedly elevated over A549/ctrl cells (averaging 4 cells). At 36 hpi, the viral foci in A549/OCLN-KD (averaging 37 cells) and A549/ZO-1-KD cells (averaging 29 cells) contained significantly more cells than in A549/ctrl cells (averaging 17 cells) (Fig 2K). Furthermore, the impacts of OCLN and ZO-1 on NDV spread between cells were measured by plaque assay and revealed that NDV infection in A549/OCLN-KD (averaging 1.88 mm) and A549/ZO-1-KD (averaging 1.64 mm) cells resulted in big-sized plaques compared to those formed in A549/ctrl cells (averaging 0.73 mm) (Fig 2L). Additionally, we conducted confirmatory experiments on HD11 cells. Foci formation results showed that the size of viral foci in HD11/OCLN-KD or HD11/ZO-1-KD cells was greater than those observed in HD11/ctrl cells. Viral foci in HD11/OCLN-KD cells (averaging 65 cells) and HD11/ZO1-KD cells (averaging 57 cells) contained more cells than in HD11/ctrl cells (averaging 11 cells), suggesting increased viral spread (S5E and S5F Fig). Therefore, these findings indicate that TJ proteins exert an inhibitory effect on NDV replication and spread. In other words, NDV facilitates its own replication and spread by downregulating the expression of these proteins.

NDV induces redistribution of TJ structure, accompanied by F-actin fiber rearrangement and cage formation

We observed that the expression levels of OCLN and ZO-1 were decreased in the NDV-infected cells. OCLN and ZO-1 are essential constituents of the cellular TJ framework [40]. Therefore, we hypothesized that NDV infection could disrupt TJ structure. To test this hypothesis, we firstly examined alternations in the junction integrity in vivo following NDV infection. We successfully established an in vivo NDV infection model in chickens (Fig 3A). In this model, chickens infected with NDV exhibited rapid viral shedding, especially those inoculated with the virulent strain (S5 Table). In addition, in vivo studies observed decreased OCLN and ZO-1 expression in chicken lungs post-infection with NDV (Fig 3B). H&E staining results demonstrated that the alveolar walls in the control chicken lung tissue exhibited tight junction and a mesh-like arrangement, appearing hyalinized. In contrast, the alveolar architecture was compromised, with a loss of characteristic morphology, disrupted junctions, extensive cell death, infiltration of inflammatory cells, and hemorrhage following NDV infection (Fig 3C). Observing extensive inflammatory cell infiltration on H&E staining, we subsequently quantified inflammatory cytokine levels in response to NDV infection. The qPCR results revealed that NDV infection induced a significant upregulation of mRNA levels of inflammatory cytokines, particularly for IL-1β, IL-6, and TNF-α (S6 Fig). In contrast, the avirulent NDV strain failed to elicit significant alterations in the expression of TJ proteins nor did it induce appreciable lung lesions (S7A and S7B Fig). Thus, these results indicate that NDV induces TJ damage in chicken lungs, which represents an important factor contributing to lung injury.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. NDV disrupts the TJ structure, accompanied by F-actin structural redistribution.

(A) 4-week-old SPF chickens were infected with 10⁵ EID₅₀ NDV via the intranasal and intraocular route for 4 days. PBS group was considered as the control group. The oropharyngeal and cloacal swabs were gathered daily from the infected chickens to evaluate virus shedding. The lung tissue was harvested to assess histopathological alterations and TJ-associated protein expression at 4 dpi. Created in BioRender. Lu, D. (2025) https://BioRender.com/ebond7q. (B) The OCLN and ZO-1 protein levels in lung tissue were detected by western blotting. The gray value of each protein was quantified by Image J and normalized to ACTB. The gray value of mock-infected group was considered as “1”. (C) The pathological changes of lung tissue were observed by H&E staining (40× and 100×). The red triangle represents disrupted junctions, the green triangle represents necrotic cells, the red circle represents the infiltration of inflammatory cells, and the red box represents the hemorrhage. (D) The intercellular TJ structure was observed in NDV-infected and mock-infected A549 cells at 6 hpi by TEM assay. The solid black unidirectional arrow denotes a standard tight junction, whereas the solid black bidirectional arrow signifies an enlarged tight junction. Scale bar: 1 μm. (E) The localization of OCLN and ZO-1 was observed in NDV- (MOI = 0.1) and mock-infected A549 cells at 18 hpi by confocal microscopy. Scale bar: 20 μm. (G) The confocal microscopy was employed to assess the localization of F-actin and compared its distribution with that of OCLN and ZO-1 in NDV- (MOI = 0.1) and mock-infected A549 cells. The white arrow denotes the aggregation of NDV NP in the F-actin cage, whereas the dashed box highlights the aggregation of OCLN in the F-actin cage. Scale bar: 10 μm.

https://doi.org/10.1371/journal.ppat.1013458.g003

To vividly visualize the TJ structure, we utilized TEM and confocal microscopy to clearly observe intercellular TJ structure. TJ complexes were easily detectable at the intercellular junctions of neighboring cells in mock-infected group, but in NDV-infected cells, many cell-cell junctions were lost, and increased gaps between neighboring membranes were observed (Fig 3D). The confocal microscopy results showed that both OCLN and ZO-1 were distributed in a typical linear pattern along the layer of cell-to-cell contact in mock-infected cells. However, cellular localization of OCLN and ZO-1 was severely disrupted and a fragmented and clumpy pattern in NDV-infected cells, especially the phenomenon of clustering in OCLN (Fig 3E). F-actin can form TJ complexes with TJ proteins, which are crucial for the maintenance of cellular TJ homeostasis [7]. We further observed the localization of F-actin and compared its distribution with that of OCLN and ZO-1. In mock-infected cells, F-actin exhibited a typical fiber network structure throughout the cytoplasm. However, the F-actin fiber network was significantly disrupted, with evident fiber aggregation and cage-like formation observed. Notably, NP and OCLN aggregates were encased within F-actin cages, whereas fluorescence outside these cages was weak. This finding suggests their potential roles as hubs for viral replication and TJ redistribution. Furthermore, F-actin exhibited certain cellular co-localization with OCLN and ZO-1 in mock-infected cells, but the co-localization was compromised following NDV infection (Fig 3F). In contrast, the avirulent NDV strain failed to induce the obvious alternations of OCLN, ZO-1, and F-actin structure (S8A Fig). Hence, these findings suggest that NDV induces TJ redistribution, coupled with F-actin reorganization and cage formation.

Vimentin serves as a critical molecular switch in the TJ injury during NDV infection

Vimentin structure serves as a critical mediator that promotes epithelial barrier dysfunction by modulating TJs, thereby playing a pivotal role in pathogen infection [32]. In addition, we have previously demonstrated that vimentin rearrangement contributes to NDV infection [27]. Here, we hypothesized that vimentin structure served as a pivotal regulatory factor in the TJ injury during NDV infection. To validate this hypothesis, we examined the structural changes of vimentin following NDV infection. Moreover, we employed non-cytotoxic IDPN to selectively inhibit vimentin rearrangement, subsequently evaluating the effect of weakened vimentin rearrangement on TJ homeostasis. The cytotoxicity of varying IDPN concentrations on A549 cells was evaluated using a CCK-8 assay. IDPN concentrations below 1.5% were found to be non-toxic (Fig 4A). In mock-infected and IDPN-treated cells, vimentin filaments formed a widespread, elongated network throughout the cytoplasm. Conversely, in virulent NDV-infected cells, these filaments appeared dense, shortened, and extensively accumulated around the nucleus. Notably, the virulent NDV induced vimentin cage formation, encapsulating NDV NP aggregates (Fig 4B). NP is a crucial NDV ribonucleoprotein (RNP) component, essential for viral genome synthesis along with P and L proteins [21]. Previous studies have confirmed that vimentin cages are sites for virus replication [41,42]. Here, the vimentin cage may also promote NDV replication by influencing NP function. In contrast, the avirulent NDV failed to induce significant alterations in vimentin structure, with its morphology resembling that of mock-infected cells (S8B Fig). Subsequently, confocal microscopy, western blotting and TCID₅₀ assays demonstrated that IDPN treatment remarkedly impaired viral replication both within the cells and in the supernatant (Fig 4B–4D). Of note, confocal microscopy analysis revealed a significant association between the redistribution vimentin, F-actin, and OCLN following NDV infection, indicating a tight spatial correlation within these cellular structures. Briefly, NDV infection elicited the formation of a vimentin cage, which was found to be colocalized with an F-actin cage, as indicated by the white arrowheads. Echoing the phenomenon of colocalization observed between the F-actin cage and OCLN, the vimentin cage wrapped around the NDV-induced OCLN aggregations, as indicated by white dashed boxes. In contrast, IDPN treatment significantly inhibited NDV-induced vimentin rearrangement and cage formation, which was consistent with our previous results [27]. Additionally, IDPN treatment successfully inhibited the NDV-induced redistribution of ZO-1, OCLN, and F-actin, as compared to untreated NDV-infected cells. Vimentin/F-actin cage formation was also inhibited (Fig 4E). These findings indicate that vimentin functions as a critical molecular switch in mediating TJ injury induced by NDV.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. NDV-induced vimentin rearrangement contributes to TJ injury and facilitates viral replication and spread.

(A) Chemical formula of the IDPN pharmaceutical compound. Cytotoxicity of IDPN was assessed using the CCK8 assay. (B) Structural examination of vimentin in A549 cells infected with NDV at 0.1 MOI was conducted, with a comparative analysis of the fibrous structures in the presence and absence of IDPN (1.5%). The dashed box highlights the aggregation of NDV NP in the vimentin cage. Scale bar: 10 μm. (C) The replication levels of NDV (MOI = 0.1) were quantified by western blotting at 24 hpi, comparing NP levels in the presence and absence of IDPN (1.5%). (D) The replication levels of NDV (MOI = 0.1) were quantified using the TCID₅₀ assay at 12 and 24 hpi, comparing the viral titers in the presence and absence of IDPN (1.5%). (E) The co-localization of vimentin with F-actin, ZO-1, and OCLN was monitored by confocal microscopy in NDV- (MOI = 0.1) and mock-infected A549 cells at 18 hpi, which was compared in the presence and absence of IDPN (1.5%). The white arrow indicates the rearranged cage-like structure of vimentin and F-actin, whereas the dashed box highlights the aggregation of OCLN in the vimentin cage. Scale bar: 10 μm. (F) The cell migration by NDV (MOI = 0.1) were assessed at 6 hpi using the transwell assay, comparing the amounts of migrated cells in the presence and absence of IDPN (1.5%). A suspension containing 4 × 10⁴ cells was utilized for the transwell assay. Scale bar: 50 μm. (G) The plaque formation was assessed in NDV-infected A549 cells (MOI = 0.00001) using the plaque assay, comparing the plaque size in the presence and absence of IDPN (1.5%). The relative plaque size was quantified using ImageJ software. (H) The foci formation was measured in NDV-infected A549 cells (MOI = 0.00001) by IFA, comparing the number of cells within foci in the presence and absence of IDPN (1.5%). Scale bar: 50 μm. ** P < 0.01; *** P < 0.001; **** P < 0.0001.

https://doi.org/10.1371/journal.ppat.1013458.g004

To elucidate the impact of NDV replication activity on TJ proteins, we inactivated the virus using UV treatment and employed chicken embryos to confirm the lack of replication in the inactivated virus. After inactivation, we analyzed the expression levels of OCLN and ZO-1 proteins through western blotting. Subsequently, we employed confocal microscopy to observe the structure of OCLN, ZO-1, F-actin, and vimentin within the cells. Western blot analysis revealed that UV-inactivated NDV did not diminish the expression levels of OCLN and ZO-1 proteins (S9A Fig). Moreover, confocal microscopy observations indicated that UV-inactivated NDV failed to induce structural rearrangements of OCLN, ZO-1, F-actin, and vimentin, resembling the structure observed in mock-infected cells (S9B Fig). Thus, the effective viral replication is required for the modulation of TJ homeostasis.

Vimentin rearrangement is necessary for TJ injury-mediated enhancement of viral spread and inflammation

Our previous work showed TJ integrity affected virus spread and vimentin rearrangement impacted TJ integrity. This study extended that to examine vimentin’s role in NDV spread. Here, we performed a transwell assay to determine the impact of vimentin rearrangement on cell migration induced by NDV. The results showed that NDV triggered massive cell migration (averaging 233 cells), while IDPN treatment significantly inhibited this characteristic of NDV (averaging 107 cells) (Fig 4F). The result indicated that vimentin rearrangement was required for cell migration by NDV. Subsequently, we further confirmed the effect of vimentin rearrangement on NDV spread by measuring plaque and lesion formation. The plaque analysis showed that the plaque size induced by NDV in IDPN-treated cells (averaging 0.87 mm) was significantly smaller than that of the untreated infection group (averaging 2.57 mm) (Fig 4G). Furthermore, lesion formation results showed that NDV-infected cells were observed as massive foci consisting of multiple cells, whereas IDPN treatment impaired the foci formation. The cell numbers within the viral foci observed in these cells were counted, when the remarkable difference was seen between the normal infection group and IDPN-treated infection group. The cell number within the viral foci in the normal infection group (averaging 22 cells) was significantly more than those in IDPN-treated infection group (averaging 11 cells) (Fig 4H).

To elucidate whether the vimentin rearrangement can directly influence the TJ structure, we further used ajoene, a vimentin rearrangement activator, for subsequent analysis. Ajoene, a phytochemical derived from garlic, induces rearrangement and condensation of the vimentin filament network by binding to vimentin at Cys-328, ultimately leading to its disruption [43]. The CCK8 assay showed that ajoene concentrations below 10 μM were found to be non-toxic (Fig 5A). Confocal microscopy confirmed ajoene as an effective inducer of vimentin rearrangement under physiological conditions (Fig 5B). Notably, ajoene-induced vimentin rearrangement led to changes in the distribution of F-actin, ZO-1, and OCLN (indicated by white arrows), suggesting that vimentin structure was vital for maintaining TJ structural integrity (Fig 5B–5D). This finding verified that the TJ structure could be directly governed by the vimentin rearrangement. Moreover, we evaluated the effects of ajoene-induced vimentin rearrangement on viral replication and spread. Western blot and TCID₅₀ analyses demonstrated enhanced replication of NDV in cells treated with ajoene (Fig 5E and 5F). The transwell assay revealed that NDV induced cell migration (averaging 118 cells), and ajoene treatment significantly potentiated NDV-induced migration (averaging 215 cells) (Fig 5G). IFA results showed NDV-infected cells formed multi-cellular foci, which were augmented by ajoene treatment. Counting the cells within viral foci revealed a significant difference between the control infection and ajoene-treated groups. The ajoene-treated group had a notably higher cell count within viral foci (averaging 41 cells) compared to the control group (averaging 16 cells) (Fig 5H). Additionally, we measured cytokine levels after IDPN or ajoene treatments. The qPCR results showed that IDPN treatment inhibited NDV-induced inflammatory cytokine activation, whereas ajoene treatment markedly elevated cytokine expression levels (S10 Fig).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Activated vimentin rearrangement directly drives TJ injury, facilitating viral replication and spread.

(A) Chemical formula of the ajoene pharmaceutical compound. Cytotoxicity of ajoene was assessed using the CCK8 assay. (B) Structural examination of vimentin and F-actin in A549 cells treated with ajoene (10 μM) was conducted. Scale bar: 10 μm. (C) Structural examination of vimentin and OCLN in A549 cells treated with ajoene (10 μM) was conducted. The white arrow indicates the rearranged vimentin and OCLN. Scale bar: 10 μm. (D) Structural examination of vimentin and ZO-1 in A549 cells treated with ajoene (10 μM) was conducted. The white arrow indicates the rearranged vimentin and ZO-1. Scale bar: 10 μm. (E) The NP protein levels were examined in NDV-infected (MOI = 1) A549 cells treated with or without ajoene (10 μM) by western blotting. The gray value of each protein was quantified by Image J and normalized to ACTB. The gray value of NDV-infected untreated group was considered as “1”. (F) The replication levels of NDV (MOI = 0.1) were quantified using the TCID₅₀ assay in A549 cells, comparing the viral titers in the presence and absence of ajoene (10 μM). (G) The cell migration by NDV (MOI = 0.1) were assessed at 6 hpi using the transwell assay in A549 cells, comparing the amount of migrated cells in the presence and absence of ajoene (10 μM). A suspension containing 2 × 10⁴ cells was utilized for the transwell assay. Scale bar: 50 μm. (H) The cell foci were measured in NDV-infected A549 cells (MOI = 0.00001) by IFA, comparing the number of cells within foci in the presence and absence of ajoene (10 μM). Scale bar: 100 μm. ** P < 0.001; *** P < 0.001; **** P < 0.0001.

https://doi.org/10.1371/journal.ppat.1013458.g005

Furthermore, we confirmed that vimentin rearrangement was essential for increased viral spread and inflammation in avian cells due to TJ injury. As stable avian epithelial cells were unavailable for our study, we evaluated TJ-like structure formation in the avian macrophage cell line HD11, which is susceptible to NDV. Using confocal microscopy, we examined the localization of TJ proteins in mock-infected HD11 cells and observed that both OCLN and ZO-1 proteins formed distinct fluorescent structures at cell contacts, arranged in a honeycomb-like pattern, indicative of a typical TJ-like structure (S11 Fig). This finding aligns with previous reports documenting TJ-like structures in certain non-epithelial cells [8,9,44]. Given that HD11 cells stably express TJ proteins and form these TJ-like structures, we considered them a relevant avian cell model for our subsequent validation studies. We utilized confocal microscopy to investigate the structural changes in OCLN, ZO-1, cytoskeletal vimentin, and F-actin following NDV infection. In NDV-infected cells, these proteins showed severe alterations, appearing fragmented and clumped, similar to changes observed in A549 cells. OCLN, vimentin, and F-actin accumulated significantly around the nucleus, forming cage-like structures (marked by white arrowheads), while the ZO-1 structure appeared degraded. Notably, NDV caused vimentin and F-actin to form cages around OCLN and NDV NP aggregations (S13A Fig). We subsequently tested vimentin’s role in TJ injury in HD11 cells by assessing NDV-induced vimentin rearrangement. IDPN significantly inhibited NDV-induced rearrangement while demonstrating no non-toxic to cells at concentrations ≤1.5% and no influence on normal vimentin structure (S12A and S12B Fig). In contrast, IDPN treatment successfully prevented these NDV-induced redistributions (S13A Fig). IDPN significantly reduced NDV replication in cells and supernatant, as shown by western blotting and TCID₅₀ assays (S13B and S13C Fig). NDV promoted cell migration (average 153 cells), which IDPN markedly inhibited (average 102 cells) (S13D Fig). NDV caused infected cells to form large foci, but IDPN treatment impaired foci formation. Cells within foci were significantly fewer in the IDPN-treated group (average 9 cells) compared to the NDV-only group (average 38 cells) (S13E Fig). These results demonstrate the role of vimentin in NDV spread within HD11 cells. Additionally, we used ajoene to examine vimentin rearrangement’s direct effects on TJs and viral infection in HD11 cells. Confocal microscopy confirmed ajoene induced vimentin rearrangement without significant toxicity below 10 μM (S14A Fig). The rearrangement coincided with F-actin, OCLN, and ZO-1 clustering, showing vimentin directly influences the TJ-like structure in HD11 cells. We also assessed ajoene’s effect via vimentin rearrangement. This led to increased NDV replication (S14B and S14C Fig), enhanced cell migration (Ajoene-NDV: 217 cells vs NDV: 118 cells) (S14D Fig), and larger viral foci (Ajoene-NDV: 86 cells vs NDV: 33 cells) (S14E Fig). Consistent with A549 findings, ajoene boosted inflammation, whereas IDPN reduced it (S15 Fig). These findings indicate that vimentin rearrangement promotes viral replication and spread through the induction of TJ injury.

MLC/p-MLC activation facilitates vimentin-mediated TJ injury, thereby promoting NDV replication and spread

MLC/p-MLC activation is known to induce actomyosin contraction, which can lead to TJ disruption [5]. Here, we examined MLC phosphorylation in A549 cells following NDV infection. Cells were either mock infected or infected with NDV at 1 MOI. Western blot analysis revealed increased p-MLC levels in NDV-infected cells compared to mock-infected controls (Fig 6A), indicating that NDV infection indeed induces MLC phosphorylation. To assess the functional significance of MLC/p-MLC activation in NDV-induced TJ injury and viral infection, we employed the MLC inhibitor blebbistatin. Prior to infection, cells were treated with blebbistatin. Concentrations below 20 μM were determined to be noncytotoxic via CCK8 assay (Fig 6B). Blebbistatin treatment significantly mitigated NDV-induced downregulation of TJ proteins levels and viral replication (Fig 6C and 6D). We next investigated the effect of MLC/p-MLC activation on the TJ complex structure using confocal microscopy. We examined the localization of vimentin, F-actin, OCLN, and ZO-1. Compared to NDV-infected cells, blebbistatin treatment prevented the NDV-induced redistribution of these cytoskeletal and TJ proteins, causing their localization patterns to more closely resemble those of mock-infected cells (Fig 6E). This further supports MLC/p-MLC involvement in TJ rearrangement upon NDV infection. Given MLC/p-MLC’s role in TJ redistribution and its impact on viral replication, we next explored its contribution to NDV spread. Both transwell migration and foci formation assays demonstrated that MLC/p-MLC activation promotes NDV dissemination. Specifically, blebbistatin treatment significantly inhibited the migration of NDV-infected cells (Blebbistatin-NDV: 36 cells vs NDV: 102 cells) and the formation of viral foci (Blebbistatin-NDV: 11 cells vs NDV: 28 cells) (Fig 6F and 6G). Collectively, these findings suggest that MLC/p-MLC activation facilitates NDV-induced TJ injury, thereby promoting viral replication and spread.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. MLC/p-MLC activation drives vimentin-mediated TJ injury, boosting NDV replication and spread.

(A) A549 cells were mock infected or NDV-infected at 1 MOI for 24 h, and the MLC and p-MLC protein levels were assessed by western blotting. The gray value of each protein was quantified by Image J and normalized to ACTB. The gray value of mock-infected group was considered as “1”. (B) Cytotoxicity of the MLC inhibitor blebbistatin was assessed using the CCK8 assay. (C) The OCLN, ZO-1, and NP protein levels were examined in NDV-infected (MOI = 1) A549 cells treated with or without blebbistatin (20 μM) by western blotting. The gray value of each protein was quantified by Image J and normalized to ACTB. Gray values for OCLN and ZO-1 in the mock-infected group were normalized to “1”, and NP values in the NDV-infected untreated group were similarly normalized to “1”. NT indicates not test. (D) The replication levels of NDV (MOI = 0.1) were quantified using the TCID₅₀ assay in A549 cells, comparing the viral titers in the presence and absence of blebbistatin (20 μM). (E) Structural analysis of vimentin, F-actin, OCLN, and ZO-1 in NDV-infected (0.1 MOI) A549 cells (treated with or without blebbistatin, 20 μM) was performed, with mock-infected cells serving as a control group. Scale bar: 10 μm. (F) The cell migration by NDV (MOI = 0.1) were assessed at 6 hpi using the transwell assay in A549 cells. Migration levels were compared between blebbistatin (20 μM)-treated and untreated cells, with mock-infected cells as controls. A suspension containing 2 × 10⁴ cells was utilized for the transwell assay. Scale bar: 10 μm. (G) The cell foci were measured in NDV-infected A549 cells (MOI = 0.00001) by IFA. The number of cells within foci was compared between blebbistatin (20 μM)-treated and untreated cells. Scale bar: 10 μm. ** P < 0.01; **** P < 0.0001.

https://doi.org/10.1371/journal.ppat.1013458.g006