Genome-wide CRISPR knockout screen identified host dependency factors for SADS-CoV infection

Several CRISPR screening approaches have been employed to identify host factors involved in SADS-CoV infection, leading to the identification of genes such as PLAC8 [16] and ZDHHC17 [20]. These screenings were conducted by infecting CRISPR knockout cell libraries with wild-type SADS-CoV, followed by next-generation sequencing (NGS) analysis of surviving cells to identify candidate genes. To enhance the efficiency of this screening process, we designed and rescued a recombinant SADS-CoV reporter virus expressing green fluorescent protein (GFP) to replace the wild-type virus for cell infection (Fig 1A). This approach, combined with fluorescence-activated cell sorting (FACS), enables the isolation of resistant cell populations post-knockout, which are subsequently subjected to NGS analysis for candidate gene identification. Validation experiments demonstrated that icSADS-CoV-GFP exhibited growth kinetics, peak titers (up to 10⁶ TCID₅₀/mL, TCID₅₀: 50% tissue culture infectious dose) (S1A–B Fig), and plaque morphology comparable to the wild-type virus (S1C Fig). It consistently expressed both fluorescent and N proteins in infected HeLa cells (S1D–E Fig), while transmission electron microscopy (TEM) imaging confirmed the assembly of intact virions with typical spike structures (S1F Fig). These results establish icSADS-CoV-GFP as a robust tool for genome-wide screening.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Genome-wide CRISPR-Cas9 screening identifies host factors using icSADS-CoV-GFP.

(A) Schematic diagram of the icSADS-CoV-GFP rescue. The GFP-2A sequence was inserted between the M and N genes of the full-length SADS-CoV genome. Created with BioRender.com. https://BioRender.com/cc4lsyi (B) Schematic of the screening procedure. A CRISPR sgRNA knockout library was generated in HeLa cells. The library was infected with icSADS-CoV-GFP, and GFP negative cells were collected by fluorescence-activated cell sorting (FACS) for genomic DNA extraction and sequencing analysis. Created with BioRender.com. https://BioRender.com/vszss7g (C) Flow cytometry analysis after three rounds of icSADS-CoV-GFP infection and sorting. (D) Enrichment of sgRNAs in surviving cells after icSADS-CoV-GFP infection, as determined by NGS. The fold change in sgRNA reads was calculated by comparing surviving cells with the uninfected library control. The top 11 enriched genes are labeled. (E) KEGG pathway analysis of 50 top-ranked genes from the screens in (D). (F) Polyclonal HeLa knockout cell lines were established for the top 11 candidate genes and infected with icSADS-CoV-GFP. Green signal indicates GFP fluorescence. The lower panel shows representative images of positive cells from three independent experiments. Scale bar, 1250 μm. (G) Relative fluorescence intensity was quantified to indicate the percentage of infected cells. Data are presented as means ± SD from three independent experiments. Statistical significance was analyzed by one-way ANOVA with Dunnett's test. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, ns, not significant.

https://doi.org/10.1371/journal.ppat.1014569.g001

To comprehensively identify novel host factors associated with SADS-CoV infection, we constructed a Human HeLa CRISPR GeCKO v2 library covering the whole genome. After infection with icSADS-CoV-GFP at an MOI (multiplicity of infection) of 0.01 for 48 h, GFP-negative cells were isolated by FACS and allowed to recover (Fig 1B–C). This process of viral infection and FACS-based isolation of GFP-negative cells was repeated for two additional rounds, resulting in a total of three rounds of selection. Genomic DNA was extracted from the GFP-negative cell population collected after the third infection, and NGS analysis was performed using primers specific to the GeCKO library to identify enriched sgRNAs. The screening results indicated that several host factors were highly enriched. The top-ranked host factors included PLAC8, SCAP, MBTPS2, Rab10, TMEM41B, VPS16, CACNB2, VPS39, LIME1, MANSC1, and THBS4 (Fig 1D). The top 50 candidates were predominantly enriched in viral and bacterial infection pathways, confirming the successful identification of a robust host gene signature associated with pathogen-host interactions (Fig 1E). Subsequently, we constructed lentiCRISPRv2 knockout vectors for each of these selected genes. After lentiviral transduction of the knockout sgRNAs and one week of selection, polyclonal knockout HeLa cell lines were established for each gene. At 24 h post-infection (hpi) with icSADS-CoV-GFP (MOI = 0.01), GFP fluorescence signals were observed under a fluorescence microscope. The results showed significant differences in GFP signal intensity among the various polyclonal knockout cell lines after infection. Calculating the infection rate based on fluorescence intensity revealed the top three candidate genes to be TMEM41B, SCAP, and Rab10 (Fig 1F–G). As the functions of the candidate genens TMEM41B [23] and SCAP [16] have been investigated in previous studies, we therefore selected Rab10 for further analysis, whose role in coronavirus infection warranted mechanistic exploration.

Rab10 is an important proviral host factor in SADS-CoV infection

To further investigate the role of Rab10 in SADS-CoV infection, we employed RNAi as an orthogonal approach to validate its function. HeLa cells were transfected with three independent Rab10-specific siRNAs and subsequently infected with icSADS-CoV-GFP for 24 h. Rab10-knockdown cells exhibited a marked reduction in the GFP-positive infection signal, with an average decrease of 78.39% compared with siNC-treated cells (Fig 2A–B). RT-qPCR analysis showed that the three siRNAs reduced Rab10 mRNA levels by an average of 87.28% (Fig 2C). Consistently, intracellular SADS-CoV RNA levels were reduced by an average of 75.32% (Fig 2D). Western blot analysis further confirmed efficient Rab10 depletion at the protein level and revealed a corresponding reduction in SADS-CoV N protein abundance (Fig 2E). Densitometric analysis indicated average reductions of 92.67% in Rab10 protein levels and 60.67% in viral N protein levels. Collectively, these results demonstrate that Rab10 is required for efficient SADS-CoV infection.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Rab10 is an important proviral host factor in pan-coronavirus infection.

(A–E) Effect of Rab10 knockdown on SADS-CoV infection in HeLa cells. HeLa cells were transfected with siRNAs targeting Rab10 (siRab10–1, siRab10–2, and siRab10–3) or a negative control (siNC) for 24 h, followed by infection with icSADS-CoV-GFP at an MOI of 0.1. Samples were harvested at 24 hpi. Representative immunofluorescence images of HeLa cells infected with icSADS-CoV-GFP (green) (A) and quantification of relative viral infection rates (B). Nuclei were stained with DAPI (blue). Scale bar, 275 μm. Relative Rab10 mRNA levels determined by RT-qPCR (C). Intracellular SADS-CoV genomic RNA levels determined by RT-qPCR (D). Western blot analysis of SADS-CoV N protein and Rab10 expression (E). (F–J) Effect of Rab10 knockdown on SADS-CoV infection in IPI-2I cells. Representative immunofluorescence images (F) and quantification of viral infection in IPI-2I cells (G). Nuclei were stained with DAPI (blue). Scale bar, 275 μm. Relative Rab10 mRNA levels determined by RT-qPCR (H). Intracellular SADS-CoV genomic RNA levels determined by RT-qPCR (I). Western blot analysis of SADS-CoV N protein and Rab10 expression (J). (K–O) Effect of Rab10 knockdown on swine enteric coronaviruses infection in LLC-PK1 cells. Relative Rab10 mRNA levels determined by RT-qPCR (K). Relative SADS-CoV mRNA levels determined by RT-qPCR (L). Relative mRNA levels of TGEV (M), PEDV (N), and PDCoV (O) in siRab10-treated cells compared to control. (P–S) Western blot analysis of viral N protein and Rab10 expression in siNC or siRab10–1 treated LLC-PK1 cells infected with SADS-CoV (P), TGEV (Q), PEDV (R) or PDCoV (S). Data points represent individual biological replicates (n = 3). Data are presented as means ± SD from three independent experiments. B–D, G–I, K, one-way ANOVA with Dunnett's test; L–O, unpaired t-test. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, ns, not significant.

https://doi.org/10.1371/journal.ppat.1014569.g002

Given that SADS-CoV infection primarily causes intestinal diarrhea in piglets by targeting the porcine small intestinal epithelium, we performed RNAi in the porcine ileal epithelial cell line IPI-2I. Cells were transfected with three porcine Rab10-specific siRNAs and subsequently infected with icSADS-CoV-GFP for 24 h. Rab10 depletion significantly reduced the GFP-positive infection signal, with an average decrease of 57.42% compared with siNC-treated cells (Fig 2F–G). The three siRNAs reduced Rab10 mRNA levels by 57.49%–84.78% (Fig 2H), while intracellular SADS-CoV RNA levels decreased by an average of 74.84% (Fig 2I). Western blot analysis further confirmed efficient depletion of Rab10 protein and a corresponding reduction in viral N protein abundance (Fig 2J). Densitometric analysis showed average reductions of 94.67% in Rab10 protein levels and 75.00% in SADS-CoV N protein levels. These results confirm that Rab10 is also required for SADS-CoV infection in porcine intestinal epithelial cells. To further assess the impact of Rab10 on the replication of diverse porcine coronaviruses, we employed swine LLC-PK1 cells, which are susceptible to SADS-CoV, TGEV, PEDV, and PDCoV. Following RNAi mediated knockdown of Rab10 in these cells, three distinct siRNAs reduced Rab10 mRNA levels by 46.71% to 79.54% (Fig 2K). At 24 hpi with the respective porcine coronaviruses, intracellular viral genome levels were quantified by RT-qPCR. Treatment with the most effective siRNA (siRab10–1) resulted in an average reduction of 88.18% for SADS-CoV (Fig 2L), 72.95% for TGEV (Fig 2M), 61.62% for PEDV (Fig 2N), and 45.72% for PDCoV replication (Fig 2O). Western blot further confirmed efficient Rab10 depletion and corresponding reductions in viral N protein levels (Fig 2P–S). These results suggest that Rab10 broadly promotes infection by multiple porcine coronaviruses.

To determine whether SADS-CoV regulates endogenous Rab10, we examined Rab10 expression and localization in infected HeLa cells. Rab10 mRNA levels remained unchanged from 12 to 48 hpi, while total Rab10 protein was not increased and declined at later time points (S2A–B Fig), suggesting that SADS-CoV primarily utilizes the pre-existing Rab10 pool rather than inducing its expression. Immunofluorescence analysis further showed that Rab10 displayed a dispersed punctate cytoplasmic pattern in mock-infected cells, whereas a subset of Rab10-positive structures overlapped with SADS-CoV S-positive puncta following infection, particularly near the cell membrane (S2C Fig). These results indicate that SADS-CoV alters the subcellular distribution of Rab10 without increasing its overall expression.

To further elucidate the function of Rab10 in SADS-CoV infection, we generated Rab10-knockout (KO) cell lines using the CRISPR-Cas9 genome-editing system. Sequencing revealed various deletions or insertions near the sgRNA target site in both alleles of two independent KO clones, resulting in complete loss of Rab10 protein (Fig 3A). Cell viability, assessed by the CCK-8 (Cell Counting Kit-8) assay, was 98.75% and 96.62% for the KO clones compared to wild-type (WT) cells, indicating that Rab10 knockout had a relatively small impact on the normal life activities of cells (Fig 3B). We then infected the KO clones (KO1 and KO2) and WT HeLa cells with icSADS-CoV-GFP in the presence of exogenous trypsin. At 24 hpi, TCID₅₀ assay of infectious virus particles in the supernatant revealed that KO cells produced significantly fewer infectious virions than WT cells (Fig 3C), and both GFP fluorescence and intracellular N protein level were markedly reduced in KO1 and KO2 cells compared to WT controls (Fig 3D–E). This reduction in viral infection was observed both with and without exogenous trypsin, suggesting that Rab10's role is independent of trypsin-mediated enhancement of viral entry. These results demonstrate that Rab10 knockout substantially impairs SADS-CoV infection. To determine whether Rab10 is essential for SADS-CoV infection, we performed a complementation assay by re-expressing Rab10 in the knockout cells. The stable overexpression of Rab10 via a doxycycline-inducible system in WT cells enhanced viral infection to some extent (Fig 3F). Importantly, restoring Rab10 expression in KO1 cells (rescue experiment) significantly recovered viral infectivity (Fig 3G–K). Collectively, these results confirm that Rab10 is an important proviral host factor for SADS-CoV infection and exhibits a similar facilitatory role in the replication of other porcine coronaviruses.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. SADS-CoV replication was inhibited in Rab10 knockout cells and restored by Rab10 reconstitution.

(A) Strategy for generating Rab10 knockout HeLa cell lines using CRISPR-Cas9. The schematic shows the sgRNA targeting exon 2 of the Rab10 gene. Sanger sequencing alignments confirm frameshift mutations (deletions/insertions) in two independent clones (KO1 and KO2) compared to WT sequences. (B) Cell viability of WT and Rab10 KO HeLa cells. Data show no significant cytotoxicity associated with Rab10 deletion. (C–E) Effect of Rab10 deficiency on SADS-CoV infection. WT and Rab10 KO cells were infected with icSADS-CoV-GFP (MOI = 0.1) in the presence or absence of trypsin. Viral titers in cell supernatants determined by TCID₅₀ assay (C). Representative fluorescence microscopy images of cells infected with icSADS-CoV-GFP (green) (D). Nuclei were stained with DAPI (blue). Scale bar, 275 μm. Immunoblot analysis of SADS-CoV N protein and Rab10 expression at 24 hpi (E). GAPDH served as a loading control. Band intensities were quantified and normalized to GAPDH (values shown below bands). (F) Schematic of the Doxycycline (Dox)-inducible Tet-On system (pLV2-TRE3GS-Rab10-TetOne-Blast) used for Rab10 expression. Binding of the rtTA protein to the TRE promoter in the presence of Dox triggers active gene expression. Created with BioRender.com. https://BioRender.com/i4o2s2d (G) Representative fluorescence microscopy images of icSADS-CoV-GFP (green) infection in HeLa WT cells, HeLa-WT-TetOne (empty vector), and HeLa-TetOne-Rab10 cells following induction with 2.5 μg/mL Dox. Nuclei were stained with DAPI (blue). Scale bar, 275 μm. (H) Reconstitution of Rab10 in Rab10^-/- KO1 cells. Rab10^-/- KO1 cells were transduced with TetOne-empty or TetOne-Rab10 and induced with Dox before infection with icSADS-CoV-GFP (green). Representative fluorescence microscopy images of the infected cells. Nuclei were stained with DAPI (blue). Scale bar, 275 μm. (I) Western blot verification of Rab10 rescue in Rab10^-/- KO1 cells. The levels of SADS-CoV N, Rab10, and GAPDH were detected following Dox induction and viral infection. GAPDH was used as a loading control. Band intensities were quantified using Image J and normalized to GAPDH. N/GAPDH ratios are provided. (J) Quantitative analysis of SADS-CoV vRNA (left axis, bars) and viral titers (right axis, dots) in HeLa WT, Rab10^-/- KO1, and Rab10-reconstituted cells. Data points represent individual biological replicates (n = 3). Data are presented as means ± SD from three independent experiments. B–C, J, one-way ANOVA with Dunnett's test. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, ns, not significant.

https://doi.org/10.1371/journal.ppat.1014569.g003

Rab10 is essential for the late-stage of SADS-CoV in a single-cycle infection

To determine the role of Rab10 in the SADS-CoV infection cycle, we first assessed whether the adsorption and internalization stages differ between HeLa WT and Rab10^-/- KO1 cells. For the adsorption assay, cells were incubated with SADS-CoV (MOI = 0.5) at 4°C for 2 h. For the internalization assay, cells were first allowed to adsorb virus at 4°C for 2 h and then shifted to 37°C for 1 h to permit entry. Viral genome RNA (gRNA) levels were quantified by Real-time quantitative PCR (RT-qPCR), revealing no significant differences between HeLa WT and Rab10^-/- KO1 cells in either adsorption or internalization (Fig 4A–B). These results indicate that Rab10 knockout is dispensable for viral particle adsorption or internalization. Confocal microscopy was performed to examine the spatial relationship between Rab10 and dsRNA-labeled viral RNA replication sites in HeLa WT cells at 6 hpi with SADS-CoV. The results revealed that Rab10 did not strongly colocalize with the viral RNA replication sites, which were marked by both the viral N protein and dsRNA (Fig 4C). This suggests that Rab10 is likely not involved in the viral RNA replication process. To pinpoint which stage of the infection cycle is impaired upon Rab10 depletion, we monitored viral RNA levels at 2 h intervals from 0 to 8 hpi. Intracellular vRNA was measured by relative quantification, while released viral RNA in the supernatant was measured by absolute quantification. Starting at 6 hpi, supernatant vRNA levels were significantly reduced in Rab10^-/- KO (KO1 and KO2) cells compared to WT cells (Fig 4D). Intracellular vRNA showed no difference at 6 h but became significantly reduced in KO cells at 8 h compared with WT cells (Fig 4E). A comparative analysis of the gRNA ratio between WT and KO cells revealed that WT cells released more viral gRNA into the supernatant by 6 h, and by 8 h the newly produced virions led to higher intracellular vRNA accumulation in WT cells (Fig 4F–G). These observations suggest that Rab10 is likely involved in the virus late-stage of the infection cycle. Furthermore, we performed long-term propagation assays (12–48 h) following icSADS-CoV-GFP infection. Analysis revealed that Rab10 deficiency significantly reduced intracellular gRNA levels, supernatant vRNA, and infectious virus titers compared to control cells (Fig 4H–J). Consistently, Rab10 knockout delayed SADS-CoV infection, as evidenced by the distinct differences in GFP fluorescence and N protein levels observed at 36 h (Fig 4K–L). Systematic dissection of the SADS-CoV life cycle reveals that Rab10 is dispensable for viral entry and primary replication, but essential for the efficient egress of progeny virions. Consequently, Rab10 acts as a critical proviral factor by facilitating late-stage viral release and subsequent spread.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Rab10 is essential for the late-stage assembly or egress of SADS-CoV in a single-cycle infection.

(A–B) Effect of Rab10 knockout on viral attachment and internalization. HeLa WT and Rab10^-/- KO1 cells were inoculated with SADS-CoV. Viral adsorption was determined by RT-qPCR of cell-bound viral RNA after incubation at 4°C (A). Viral internalization was assessed by quantifying intracellular viral RNA after incubation at 37°C and removal of surface virus (B). ns, not significant. (C) Confocal microscopy analysis of SADS-CoV replication complexes in HeLa WT cells. HeLa cells infected with SADS-CoV (MOI = 0.5) for 6 h were immunostained for N protein (green), dsRNA (red), and Rab10 (cyan). Nuclei were stained with DAPI (blue). Scale bar, 10 μm. (D–E) Single-cycle viral replication kinetics. SADS-CoV RNA copies in the cellular supernatant (D) and intracellular fold change (E) were measured from 2 to 8 hpi in HeLa WT, Rab10^-/- KO1, and Rab10^-/- KO2 cells. (F–G) Relative SADS-CoV RNA ratios comparing WT/KO2 (F) and WT/KO1 (G) in both intracellular and supernatant compartments across early time points (2–8 h). (H–I) Long-term viral growth kinetics. Viral RNA copies in the cellular supernatant (H) and intracellular vRNA fold change (I) were quantified via qRT-PCR from 12 to 48 hpi. (J) Determination of infectious progeny virus titers. Cellular supernatants from WT, KO1, and KO2 cells were collected at indicated time points, and titers were determined by TCID₅₀ assay. (K) Representative fluorescence images showing the progression of icSADS-CoV-GFP (green) infection (MOI = 0.1) in HeLa WT, Rab10^-/- KO1, and Rab10^-/- KO2 cells from 12 to 48 hpi. Nuclei were stained with DAPI (blue). Scale bar, 275 μm. (L) Western blot analysis of viral protein expression. HeLa WT and Rab10^-/- KO1 cells were infected with icSADS-CoV-GFP, and the levels of SADS-CoV N and Rab10 were detected at indicated time points. GAPDH served as a loading control. Relative N/GAPDH ratios are provided below the blots. Data are presented as means ± SD from three independent experiments. A–B, unpaired t-test; D–E, H–J, one-way ANOVA with Dunnett's test. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, ns, not significant.

https://doi.org/10.1371/journal.ppat.1014569.g004

Rab10 is co-released with SADS-CoV virions as a vesicular component to facilitate non-lytic egress

Numerous ultrastructural studies have demonstrated that coronaviruses co-opt the host endomembrane trafficking system for assembly and egress [24]. Coronavirus assembly occurs at the ERGIC, where the N, M, and E proteins coalesce and associate with the viral RNA to form virions [25]. This process leads to the formation of large virion-containing vacuoles (LVCVs), which are large, circular organelles that are thought to originate from Golgi cisternae and which are subsequently transported to the vicinity of the plasma membrane [26]. The TEM analysis of SADS-CoV-infected HeLa cells revealed intact virions sequestered within tubular membranous structures and large LVCVs, alongside characteristic double-membrane vesicles (DMVs) (Fig 5A). These ultrastructural features suggest a vesicle-mediated, non-lytic dissemination mechanism.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Rab10 is co-detecteded with SADS-CoV virions in extracellular vesicle enriched fractions as a vesicular component to facilitate non-lytic egress.

(A) TEM analysis of intracellular ultrastructures in SADS-CoV-infected HeLa cells. Notable structures identified include the Golgi apparatus (blue arrow), double-membrane vesicles (DMVs, orange arrow), individual virions (black triangles), and large virion-containing vacuoles (LVCVs, green arrow). Scale bar, 1 μm. (B) Representative fluorescence microscopy images of icSADS-CoV-GFP infection (MOI = 0.1) in HeLa WT and Rab10^-/- KO1 cells at 36 hpi. Nuclei were stained with DAPI (blue). Scale bar, 275 μm. (C) Lactate dehydrogenase (LDH) release in the culture supernatants of HeLa cells following mock treatment or icSADS-CoV-GFP infection. Supernatants were collected at the indicated time points. Data are presented as means ± SD from three independent experiments used unpaired t-test. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, ns, not significant. (D) Schematic workflow of the differential centrifugation and sucrose gradient ultracentrifugation protocol used to isolate and purify virions and extracellular vesicles from the cell culture supernatant. Created with BioRender.com. https://BioRender.com/ su1lhy7 (E) Negative-stain TEM image of purified SADS-CoV virions obtained from the sucrose gradient fractions. Scale bar, 200 nm. (F) Sucrose gradient fractionation analysis of extracellular components. Concentrated supernatants from uninfected (top) or icSADS-CoV-GFP infected (bottom) HeLa cells were fractionated through a 10–60% sucrose gradient. Collected fractions (F3–F49) were analyzed by western blot using antibodies against Rab10, CD9, SADS-CoV N protein, Alix, and Calnexin to determine the co-fractionation profile of the virus and host factors. (G) Representative confocal microscopy images of SADS-CoV-infected HeLa WT and Rab10^-/- KO1 cells stained for Rab10 (green), the viral nucleocapsid protein (SADS-CoV N, purple), and the viral spike protein (SADS-CoV S, red). Nuclei were stained with DAPI (blue). Scale bar, 5 μm. Boxed regions are enlarged in the right-hand panels, and arrowheads indicate representative areas of signal overlap or close spatial association. (H) Representative Confocal microscopy images images of SADS-CoV-infected HeLa WT and Rab10^-/- KO1 cells stained for Rab10 (green), LMAN1 (purple), and SADS-CoV S (red). Nuclei were stained with DAPI (blue). Scale bar, 5 μm. Boxed regions are enlarged, and arrowheads indicate representative areas of signal overlap or close spatial association.

https://doi.org/10.1371/journal.ppat.1014569.g005

To investigate how Rab10 participates in SADS-CoV release, we collected culture supernatants from icSADS-CoV-GFP infected HeLa WT and Rab10^-/- KO1 cells at 36 hpi (Fig 5B). An LDH-release assay showed that plasma membrane integrity was largely maintained at 36 hpi, the time point used for supernatant collection, whereas substantial LDH release was primarily observed at later stages of infection (Fig 5C), thus supernatants were carefully harvested before significant cell detachment or lysis to minimize contamination by intracellular components. To further characterize extracellular membrane-associated components, culture supernatants were subjected to sequential centrifugation followed by sucrose-density-gradient fractionation (Fig 5D). Transmission electron microscopy of the isolated material revealed virion-like particles (Fig 5E). Western blot analysis of the gradient fractions revealed that under basal conditions, secreted Rab10 was detected in extracellular fractions containing SADS-CoV N protein and the extracellular vesicle-associated markers Alix and CD9, whereas the intracellular contamination marker Calnexin was not detected (Fig 5F). Rab10 was primarily found in fractions F25–F31. Following SADS-CoV infection, both the amount of secreted Rab10 increased and its density profile shifted downward to fractions F27–F35, which coincided with the peak fractions for SADS-CoV N protein. This co-sedimentation suggests that Rab10 functions as a component of the transport vesicles encapsulating the virions.

We next used confocal microscopy to investigate whether Rab10 participates in the intracellular trafficking of SADS-CoV. To visualize putative mature virion-associated structures, cells were co-stained with antibodies against the viral N and S proteins. In wild-type cells, Rab10-positive puncta overlapped with intracellular structures positive for both SADS-CoV N and S proteins. In contrast, Rab10^-/- KO1 cells exhibited pronounced intracellular accumulation of N- and S-positive signals, suggesting impaired transport of virus-associated structures (Fig 5G). We further examined the localization of SADS-CoV relative to the ERGIC, the primary site of coronavirus assembly. In wild-type cells, Rab10 and SADS-CoV S-positive structures showed prominent punctate association with the ERGIC marker LMAN1. In Rab10^-/- KO1 cells, the viral protein- and LMAN1-positive signals accumulated within more restricted intracellular regions, with fewer dispersed punctate structures, further indicating defective trafficking of virus-associated compartments (Fig 5H).

We next investigated whether Rab10 participates in the post-assembly release of SADS-CoV particles from the ERGIC. Total ultracentrifugation-concentrated supernatants and corresponding cell lysates were analyzed by western blot for the viral proteins N and S, Rab10, and compartment-specific markers, including TGN46 for the trans-Golgi network, LAMP1 for lysosomes, and Alix and CD9 for exosomes/extracellular vesicles. In cell lysates, Rab10 knockout reduced the abundance of SADS-CoV N and S proteins, whereas the total cellular levels of TGN46, LAMP1, Alix, CD9, and Calnexin remained largely unchanged (Fig 6A). In concentrated supernatants from infected wild-type cells, SADS-CoV N and S proteins were detected together with Rab10, TGN46, LAMP1, Alix, and CD9. Infection increased the extracellular abundance of Rab10, TGN46, and LAMP1. In contrast, Rab10 knockout reduced the extracellular abundance of the viral proteins and markedly decreased extracellular TGN46 and LAMP1 levels (Fig 6B). These findings suggest that SADS-CoV may exploit Golgi- and lysosome-associated pathways for non-lytic egress.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Rab10 deficiency impairs the release of SADS-CoV through TGN/lysosome-associated trafficking pathway.

(A–B) Western blot analysis of viral and host protein levels in cell lysates (A) and concentrated culture supernatants (B). HeLa WT and Rab10^-/- KO1 cells were infected with SADS-CoV. Lysates and supernatants were harvested and immunoblotted for viral proteins (S, N) and host markers (Rab10, TGN46, LAMP1, Alix, CD9, Calnexin). GAPDH served as an internal loading control for cell lysates. For concentrated supernatant fractions, relative band intensities were expressed relative to the indicated control sample, which was set to 1.00. Calnexin was included as a marker of contamination by intracellular membranes or cellular contents. (C) Representative Confocal microscopy images images of SADS-CoV-infected HeLa WT and Rab10^-/- KO1 cells stained for Rab10 (green), TGN46 (purple), and SADS-CoV S (red). Nuclei were stained with DAPI (blue). Scale bar, 5 μm. Boxed regions are enlarged, and arrowheads indicate representative areas of signal overlap or close spatial association. (D) Representative Confocal microscopy images images of SADS-CoV-infected HeLa WT and Rab10^-/- KO1 cells stained for Rab10 (green), LAMP1 (purple), and SADS-CoV S (red). Nuclei were stained with DAPI (blue). Scale bar, 5 μm. Boxed regions are enlarged, and arrowheads indicate representative areas of signal overlap or close spatial association.

https://doi.org/10.1371/journal.ppat.1014569.g006

Confocal microscopy further revealed overlapping punctate signals among Rab10, SADS-CoV S, and TGN46 in wild-type cells (Fig 6C). Similarly, Rab10- and SADS-CoV S-positive structures overlapped with LAMP1-positive compartments (Fig 6D). In Rab10^-/- KO1 cells, SADS-CoV S-positive signals accumulated within compact TGN46- or LAMP1-positive intracellular regions, accompanied by a reduction in dispersed punctate structures. Collectively, these results suggest that SADS-CoV exploits the host endomembrane system for non-lytic egress and that Rab10 functions as a key trafficking factor during this process. Rab10 facilitates the transport of SADS-CoV-associated vesicular structures through an ERGIC–TGN/lysosome-associated pathway, thereby promoting the extracellular release of viral and vesicular components.

Rab10 promotes the release of coronavirus VLPs through TGN and lysosome associated vesicular trafficking

Given that coronavirus release involves complex endomembrane trafficking, we utilized a replication-incompetent virus-like particles (VLPs) system to further delineate the specific role of Rab10 during the assembly and release stages. In studies of coronavirus egress, VLPs serve as vital tools as they are genome-less, enveloped particles that mimic wild-type virus behavior without being replication-competent [27]. Although the relative importance of each structural protein for VLPs formation varies across SARS-CoV-2 studies, E and M proteins are generally considered the primary drivers of SARS-CoV-2 VLPs assembly and release [28]. We employed an improved VLPs assembly plasmid to simulate the assembly and release of coronavirus particles [29]. This plasmid encodes the N, M, and E protein sequences linked by T2A self-cleaving peptides and an IRES bicistronic element downstream of a CMV promoter, allowing a single plasmid to produce VLPs containing all three structural proteins (Fig 7A).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Rab10 is essential for the efficient secretion of coronavirus virus-like particles.

(A) Schematic representation of the VLPs expression plasmid (pcDNA3.1-CoV-N-T2A-M-IRES-E) containing a CMV promoter, SADS-CoV or SARS-CoV-2 N gene, a T2A self-cleaving peptide, the M gene, an Internal Ribosome Entry Site (IRES), and the E gene. Created with BioRender.com. https://BioRender.com/qse9jf6 (B) Experimental workflow for VLPs production and purification. HeLa cells were transfected with the VLPs plasmid for 48 h. Culture supernatants were collected and subjected to differential centrifugation (300 × g, 2,000 × g, and 10,000 × g) to remove cells and debris. VLPs were then concentrated by ultracentrifugation (100,000 × g) through a 20% sucrose cushion for western blot analysis. Created with BioRender.com. https://BioRender.com/3pit7so (C–L) Impact of Rab10 depletion on SADS-CoV and SARS-CoV-2 VLPs secretion. HeLa WT and Rab10^-/- KO1 cells were transfected with plasmids producing SADS-CoV VLPs (C) or SARS-CoV-2 VLPs (H). Western blot was performed to analyze the expression of viral N protein, Rab10, and specific organelle markers: LMAN1 (ERGIC), TGN46 (trans-Golgi), and LAMP1 (lysosomes) in both cell lysates and purified VLPs. GAPDH was used as the loading control for cell lysates. Quantitative densitometric analysis of SADS-CoV N (D), LMAN1 (E), TGN46 (F), and LAMP1 (G) expression in cell lysates transfected with SADS-CoV VLPs and their respective secretion levels in the medium. Data compare HeLa WT (white bars) and Rab10^-/- KO1 (red bars). Quantitative densitometric analysis of SARS-CoV-2 N (I), LMAN1 (J), TGN46 (K), and LAMP1 (L) expression in cell lysates transfected with SARS-CoV-2 VLPs and their respective secretion levels in the medium. Data compare HeLa WT (white bars) and Rab10^-/- KO1 (red bars). Data points represent individual biological replicates (n = 3). Data are presented as means ± SD from three independent experiments. Statistical significance was analyzed by unpaired t-test. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, ns, not significant.

https://doi.org/10.1371/journal.ppat.1014569.g007

After transfecting HeLa WT and Rab10^-/- KO1 cells with the VLPs plasmids for 48 h, we purified the SADS-CoV or SARS-CoV-2 VLPs via ultracentrifugation through a 20% sucrose cushion for western blot analysis (Fig 7B). Western blot analysis revealed that SADS-CoV N protein was clearly detectable in the supernatant of HeLa WT cells (Fig 7C). However, in Rab10^-/- KO1 cells, despite intracellular N protein expression levels showing no significant difference compared to WT cells, the level of N protein released into the supernatant was significantly reduced. Specifically, Rab10 deficiency led to an average reduction of 75% in SADS-CoV N protein secretion (Fig 7D). This phenomenon was highly consistent in the SARS-CoV-2 VLPs model, where N protein secretion decreased by an average of 90% (Fig 7H–I). These results strongly demonstrate that Rab10 exerts a critical proviral function during the particle egress of multiple coronaviruses, acting specifically at the assembly or release stages following viral protein synthesis.

Furthermore, we observed that the Rab10 protein itself is co-secreted into the extracellular space alongside the VLPs (Fig 7C, 7H). To further clarify the trafficking pathways involved, we monitored the release of major organelle markers during the coronavirus assembly and release process. The results showed that Rab10 depletion significantly inhibited the secretion levels of the trans-Golgi marker TGN46 and the lysosomal marker LAMP1 into the supernatant (Fig 7F–G, 7K–L), while the secretion of the ERGIC marker LMAN1 was not significantly affected (Fig 7E, 7J). This suggests that Rab10 likely serves as a component of the vesicular, mediating coronavirus release through either the classical post-Golgi secretory pathway or the lysosomal exocytosis pathway. In summary, using the VLPs model, we confirmed that Rab10 is a host factor essential for the efficient assembly or egress of SADS-CoV and SARS-CoV-2 virions. Additionally, Rab10 is a key regulator of the transport toward the TGN and lysosomal associated vesicular pathways after the virus leaves the ERGIC.

Rab10 interacts specifically with the SADS-CoV E protein but is dispensable for M-E mediated viral assembly

Coronavirus assembly is an orchestrated process requiring precise coordination between structural proteins, primarily driven by specific protein-protein interactions [30]. The M protein initiates virion assembly by recruiting the viral genome to budding sites through interactions with the C-terminus of the N protein. Simultaneously, M protein homodimerization and its interaction with the E protein drive membrane curvature and envelope formation, processes essential for viral budding and egress (Fig 8A).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 8. Rab10 interacts with SADS-CoV E protein but is dispensable for M-E interaction.

(A) Schematic representation of SADS-CoV virion structure and its assembly process at the ERGIC. Viral structural proteins, including Spike (S), Envelope (E), Membrane (M), and Nucleocapsid (N), are shown alongside the viral RNA and ribonucleocapsid helices. The C-terminal interaction between M and E proteins is highlighted. Created with BioRender.com. https://BioRender.com/wl71gh9 (B) Co-IP analysis of the interaction between Rab10 and viral proteins. HeLa cells were co-transfected with Myc-Rab10 and Flag-tagged N, M, or E proteins. Cell lysates were immunoprecipitated with anti-Flag beads and analyzed by western blot with indicated antibodies. (C) Construction and rescue of recombinant SADS-CoV expressing tagged E proteins. The schematic shows the organization of icSADS-CoV-GFP containing E-WT, E-Flag, or Flag-E. Fluorescence images depict the successful rescue of E-WT and E-Flag viruses at Passage 2. Scale bar, 275 μm. (D) Co-IP analysis of endogenous Rab10 and viral E protein interaction in infected cells. HeLa WT cells were infected with icSADS-CoV-GFP E-Flag and subjected to IP using anti-Flag or control IgG beads. (E) Confocal microscopy analysis of icSADS-CoV-GFP E-Flag infected HeLa cells. Cells were stained for Rab10, SADS-CoV N, LMAN1 or TGN46 (yellow), Flag (red), and DAPI (blue). Infected cells are identified by GFP expression (green). Scale bar, 5 μm. (F) Co-IP analysis of interactions between SADS-CoV structural proteins in the presence or absence of Rab10. HeLa WT and Rab10^-/- KO1 cells were co-transfected with M-Myc and E-Flag. Cell lysates were immunoprecipitated with anti-Myc antibodies as indicated, and immunoblotted to detect interacting partners. The numbers below the blots indicate the relative intensity of the co-immunoprecipitated protein bands compared to the WT control (set as 1.00). (G) Representative confocal microscopy images showing the co-localization of E-Flag (green) and M-Myc (red) in HeLa WT and Rab10^-/- KO1 cells. Nuclei were stained with DAPI (blue). Scale bars represent 5 μm. (H–I) Statistical quantification of the co-localization between SADS-CoV E and SADS-CoV M. Mander's (H) and Pearson's (I) correlation coefficients were calculated to compare HeLa WT and Rab10^-/- KO1 cells. Data are presented as means ± SD from three independent experiments. Statistical significance was analyzed by unpaired t-test. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, ns, not significant.

https://doi.org/10.1371/journal.ppat.1014569.g008

To identify the specific viral component interacting with Rab10, we co-transfected HeLa cells with Myc-tagged Rab10 and Flag-tagged SADS-CoV N, M, or E proteins. Co-immunoprecipitation (Co-IP) analysis revealed a specific interaction between Rab10 and the E protein, whereas no significant binding was detected with the N or M proteins (Fig 8B). To validate these interactions in a physiologically relevant context, we employed a reverse genetics system to construct recombinant viruses expressing Flag-tagged E protein. We designed infectious clones of icSADS-CoV-GFP with the Flag sequence inserted at either the N- or C-terminus of the E gene to track its expression and localization. We found that only the C-terminally tagged clone (icSADS-CoV-GFP E-Flag) was successfully rescued and capable of stable passage, while the N-terminally tagged clone (icSADS-CoV-GFP Flag-E) was unrescued (Fig 8C). This suggests that the N-terminus of the E protein is critical for membrane topology, ion channel activity, or early recognition of other structural proteins, and is thus intolerant to tagging. In icSADS-CoV-GFP E-Flag infected cells, endogenous Co-IP confirmed the physical association between Rab10 and E during active infection (Fig 8D). Confocal microscopy further showed that E-Flag-positive intracellular structures were spatially associated with Rab10 and the viral N protein. E-Flag positive signals were also detected in association with the LMAN1 and TGN46, consistent with the localization of the E protein within virus assembly and trafficking related membrane compartments (Fig 8E). These observations suggest that the E protein may contribute to the recruitment of Rab10 to SADS-CoV associated vesicular structures.

Since the M-E interaction is the core driver of virion morphogenesis, we next investigated whether Rab10 regulates the interaction between the E and M proteins. We co-expressed E-Flag and M-Myc in both HeLa WT and Rab10^-/- KO1 cells. Co-IP assays demonstrated that the binding efficiency between M and E remained unchanged in the absence of Rab10, with normalized interaction ratios of 1.00 and 1.02, respectively (Fig 8F). Confocal microscopy further showed that E and M proteins co-aggregated in dense perinuclear regions (putative ERGIC sites) regardless of Rab10 status. The colocalization pattern of M and E at these assembly sites remained unaffected in Rab10^-/- KO1 cells (Fig 8G). Quantitative analysis using Mander’s overlap coefficient and Pearson’s correlation coefficient revealed no significant difference between the two cell types (Fig 8H–I). Taken together, these results demonstrate that while the SADS-CoV E protein specifically recruits Rab10 during infection, Rab10 does not directly participate in the M-E interaction that drives spatial viral assembly. This implies that Rab10 recruitment likely facilitates the subsequent viral release phase rather than the initial structural assembly of the virion.

Rab10 mediated localization of the SADS-CoV E protein to the ERGIC and vesicular compartments

The ERGIC is a distinct membrane compartment that mediates cargo transport between the ER and the Golgi apparatus [31,32]. Coronaviruses assemble and bud at the ERGIC before exploiting host secretory pathways for subsequent transport and release, potentially through TGN or lysosome associated route [25].

We first investigated whether the ER-to-Golgi trafficking pathway is essential for SADS-CoV infection using Brefeldin A (BFA), which disrupts Golgi and block ER-to-Golgi transport (Fig 9A). Following icSADS-CoV-GFP infection, BFA was added at different time points, and all samples were collected at a common endpoint (Fig 9B). Secreted Gaussia luciferase (Gluc) was used to verify inhibition of the biosynthetic secretory pathway. BFA significantly reduced both Gaussia luciferase secretion and extracellular viral RNA after 6–8 h of treatment (Fig 9C–D). Furthermore, BFA displayed potent antiviral activity with an IC₅₀ of 0.027 μM and a CC₅₀ of 3.396 μM, corresponding to a selectivity index of 126 (Fig 9E–F). These results indicates that the ER-to-Golgi trafficking is required for efficient SADS-CoV propagation. However, these findings do not specifically establish the involvement of downstream lysosomal transport.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 9. SADS-CoV utilizes the conventional secretory pathway, and Rab10 is required for the proper localization of SADS-CoV E protein to the ERGIC transport vesicles.

(A) Schematic representation of BFA mechanism of action. BFA induces Golgi disassembly and fusion into the ER, thereby blocking the conventional protein secretory pathway. Created with BioRender.com. https://BioRender.com/xbn3zy9 (B) Experimental timeline for BFA treatment. HeLa cells were infected with SADS-CoV, and BFA was added at the indicated time points (2, 4, 6, 8 h) before samples were harvested at 12 h. (C–D) Effect of BFA treatment on viral secretion and replication. Gluc secretion (C) and extracellular viral RNA copies (D) were measured at the indicated time points. (E–F) Dose-response curve showing the inhibition of icSADS-CoV-GFP replication and cytotoxicity in HeLa cells treated with increasing concentrations of BFA. The IC₅₀ (E) and CC₅₀ (F) were determined in HeLa cells using inhibition assays and CCK-8 tests, respectively. (G–I) Representative confocal images showing the co-localization of transiently expressed E-HA (red) with endogenous Rab10 (green) and organelle markers LMAN1 (G), TGN46 (H) or LAMP1 (I) (purple) in HeLa WT cells. Nuclei were stained with DAPI (blue). White arrowheads indicate the punctate distribution of E protein at the ERGIC. Scale bar, 5 μm. (J) Localization of SADS-CoV E protein in the absence of Rab10. Representative confocal images showing the distribution of E-HA (green) relative to LMAN1 (red) in HeLa WT and Rab10^-/- KO1 cells. White arrowheads indicate the punctate distribution of E protein at the ERGIC. Scale bar, 5 μm. (K–L) Statistical quantification of the co-localization between SADS-CoV E and LMAN1. Mander's (K) and Pearson's (L) correlation coefficients were calculated to compare HeLa WT and Rab10^-/- KO1 cells. Data are presented as means ± SD from three independent experiments. Statistical significance was analyzed by unpaired t-test. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, ns, not significant.

https://doi.org/10.1371/journal.ppat.1014569.g009

We therefore evaluated the contribution of late endosomal/lysosomal trafficking. siRNA-mediated depletion of Arl8b, Rab7, or VPS39, confirmed by RT-qPCR, markedly reduced the GFP-positive infection signal and intracellular SADS-CoV RNA levels in wild-type and Rab10^-/- KO1 cells (S3A–G Fig.), indicating that these lysosomal trafficking regulators contribute to efficient SADS-CoV infection and dissemination. We further used CID-1067700 to interfere with lysosomal trafficking. Treatment with CID-1067700 markedly reduced the icSADS-CoV-GFP GFP-positive infection signal, similar to BFA treatment (S3H Fig.). To distinguish effects on viral release from those on viral entry or genome replication, we analyzed the release of SADS-CoV VLPs. CID-1067700 substantially reduced the amount of N protein detected in the culture medium, whereas intracellular N-protein abundance remained largely unchanged. The medium-to-cell lysate N-protein ratio decreased from 1.00 in DMSO-treated cells to 0.41 following CID-1067700 treatment, compared with 0.12 after BFA treatment (S3I Fig.). Consistently, time-of-addition experiments showed that prolonged treatment with either BFA or CID-1067700 reduced extracellular SADS-CoV RNA, with significant inhibition after 6–8 h of treatment (S3J–K Fig.). Together, these results support the involvement of late endosomal/lysosomal trafficking in SADS-CoV dissemination and VLP release.

To further determine whether Rab10 participates in the intracellular trafficking of the SADS-CoV E protein, HeLa cells were transfected with an E-HA expression plasmid and examined by confocal microscopy. Concurrently, E-HA expression induced the fragmentation of the compact ERGIC structure into numerous vesicular signals, within which Rab10 colocalized with the E-HA (Fig 9G), consistent with the localization observed in cells ectopically expressing HA-tagged E protein, where E-HA was also associated with Rab10-positive compartments containing the TGN marker TGN46 or the lysosomal marker LAMP1 (Fig 9H–I). Enlarged images revealed multiple overlapping punctate signals among E-HA, Rab10, and the respective organelle markers. These observations suggest that the SADS-CoV E protein is associated with Rab10-positive membrane compartments related to the ERGIC, TGN, and lysosomes.

Previous studies have suggested that the coronavirus E protein can be released in membrane associated vesicular structures, potentially through their membrane remodeling or ion channel [33]. We therefore compare the association of E-HA with LMAN1 in wild-type and Rab10^-/- KO1 cells. Interestingly, in Rab10-knockout cells, while the E protein still induced the formation of dispersed vesicles, it significantly failed to load onto LMAN1-positive ERGIC vesicles (Fig 9J). Quantitative analysis confirmed that both Mander's and Pearson's correlation coefficients for E protein colocalization with LMAN1 were significantly reduced in Rab10-deficient cells (Fig 9K–L). These findings suggest that Rab10 facilitates the proper association and trafficking of E protein within ERGIC vesicles for transport.

In summary, our results demonstrate that SADS-CoV infection depends on proper trafficking through the ER-to-Golgi/lysosome secretory pathway, with the E protein playing a central role in hijacking this host pathway. Furthermore, Rab10 acts as a critical guide that facilitates the trafficking of SADS-CoV associated vesicular structures through ERGIC-TGN/lysosomal related compartments, thereby promoting efficient viral dissemination and particle release.

The C-terminal 64–75 aa region of SADS-CoV E protein is essential for Rab10 recruitment and ERGIC-mediated trafficking

To further delineate the specific molecular motifs within the SADS-CoV E protein responsible for Rab10 recruitment, we designed and constructed seven alanine-substitution mutants (Emt1–Emt7) covering the full-length protein based on its structural domains (N-terminus, transmembrane domain, and C-terminus) (Fig 10A). Co-IP assays revealed a robust interaction between the wild-type E protein (E-WT) and Myc-tagged Rab10. Conversely, the Emt7 mutant, featuring mutations in residues 64–75 of the C-terminus, almost entirely abolished Rab10 binding (relative ratio = 0.05). Additionally, Emt6 (residues 53–64) exhibited substantially weakened binding affinity (ratio = 0.36) (Fig 10B).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 10. The C-terminal region of SADS-CoV E protein is essential for its interaction with Rab10 and its localization to the ERGIC.

(A) Schematic diagram of SADS-CoV E protein truncation mutants. The E protein consists of an N-terminal domain (residues 1–10), a transmembrane domain (residues 10–33), and a C-terminal domain (residues 33–75). Truncation mutants (mt1–mt7) were constructed by sequentially deleting residues from the C-terminus. (B) Co-IP analysis of the interaction between Rab10 and E protein mutants. HeLa cells were co-transfected with Myc-Rab10 and HA-tagged E-WT or mutants (mt1–mt7). Cell lysates were immunoprecipitated with anti-HA beads and analyzed by western blot with anti-Myc and anti-HA antibodies. GAPDH served as the loading control for input samples. The IP Myc/HA ratios, normalized to E-WT, are indicated below the blots. (C–D) Confocal microscopy analysis of the co-localization between Myc-Rab10 (green) and HA-tagged E protein mutants (red) in HeLa WT cells. Nuclei were stained with DAPI (blue). Scale bar, 5 μm. (E–F) Statistical quantification of co-localization using Mander's and Pearson's correlation coefficients. Panels (E) and (F) show the co-localization analysis between SADS-CoV E (WT, mt6, or mt7). (G) Measurement of Rab10 concentration in the cell culture supernatant. HeLa cells were transfected with HA-tagged E protein mutants, and the concentration of secreted Rab10 was quantified by ELISA. (H) Representative confocal images showing the co-localization of E-WT-HA or Emt7-HA (red) with the ERGIC marker LMAN1 (green) in HeLa WT cells. Nuclei were stained with DAPI (blue). Scale bar, 5 μm. (I–J) Statistical quantification of co-localization using Mander's and Pearson's correlation coefficients. Panels (I) and (J) show the co-localization analysis between E-HA (WT or mt7) and LMAN1. Data are presented as means ± SD from three independent experiments. E–F, G, one-way ANOVA with Dunnett's test. I–J, unpaired t-test. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, ns, not significant.

https://doi.org/10.1371/journal.ppat.1014569.g010

We next examined the intracellular spatial distribution of these mutants and Rab10 using confocal microscopy. While E-WT and the majority of mutants exhibited significant colocalization with Rab10, the spatial overlap between the Emt7 mutant and Rab10 was markedly reduced (Fig 10C–D). Quantitative analysis further confirmed that both Mander's and Pearson's correlation coefficients for Emt7 were significantly lower than those for E-WT and Emt6 (Fig 10E–F). In addition, Emt7 induced substantially less extracellular Rab10 detection than E-WT, suggesting that the extracellular association of Rab10 observed following E expression depends, at least in part, on the E–Rab10 interaction (Fig 10G).

We next examined whether the loss of Rab10 binding affected the subcellular position of the E protein relative to LMAN1-labeled vesicles. In HeLa WT cells, E-WT displayed a characteristic punctate distribution and localized to LMAN1 positive ERGIC-derived vesicles. In contrast, the Emt7 mutant exhibited a distinct diffuse cytoplasmic distribution and a significantly reduced overlap with LMAN1 labeled compartments (Fig 10H). Quantitative analysis showed a significant decrease in both Pearson's and Mander's coefficients for Emt7 compared to E-WT (Fig 10I–J). These findings suggest that the C-terminal Rab10-binding region contributes to the proper localization and trafficking of E within ERGIC-associated membrane compartments.

Finally, we utilized a reverse genetics system to evaluate the importance of this C-terminal region for viral fitness. Premature stop codons were introduced into the icSADS-CoV-GFP backbone to generate various C-terminal truncations (S4A Fig). Our results showed that the icSADS-CoV-GFP E Δ73–75aa, Δ69–75aa, and Δ65–75aa mutants could not be rescued at passage 1. In contrast, the wild-type virus was successfully recovered and exhibited robust fluorescence signaling (S4B Fig). This indicates that subcellular mislocalization resulting from C-terminal deletions is a key factor precluding viral rescue. Taken together, these data strongly demonstrate that the 64–75 aa region of the SADS-CoV E protein is a critical molecular determinant for recruiting Rab10, ensuring correct secretion/trafficking from the ERGIC, and maintaining the viral life cycle.

Disruption of the Rab10 binding site in the SADS-CoV E protein compromises the release process of progeny virions

To precisely map the functional motif within the SADS-CoV E protein responsible for Rab10 recruitment, we performed alanine-scanning mutagenesis across residues 64–75. Co-IP assays revealed that the C-terminal residues I73, N74, and V75 are critical determinants, as their substitution with alanine significantly abrogated the interaction with Rab10 (Fig 11A). This region corresponds to a putative PDZ-binding motif, which typically mediates specific interactions with host PDZ-domain proteins to target viral proteins to the ERGIC assembly sites. Sequence alignment across diverse coronaviruses showed that while residues I73 and N74 are poorly conserved, V75 is nearly universal among alpha- and delta-coronaviruses, with the exception of MHV, which possesses an isoleucine at the corresponding position (Fig 11B).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 11. The C-terminal V75 residue of the SADS-CoV E protein is critical for Rab10 interaction and viral egress.

(A) Alanine scanning mutagenesis of the C-terminal residues (64–75) of SADS-CoV E protein. HeLa cells were co-transfected with Myc-Rab10 and HA-tagged E-WT or alanine mutants. Cell lysates were immunoprecipitated with anti-HA beads and analyzed by western blot. (B) Sequence alignment of the C-terminal tails of E proteins from diverse coronaviruses, including alphacoronaviruses (SADS-CoV, TGEV, PEDV), deltacoronaviruses (PDCoV), and betacoronaviruses (SARS-CoV-2, MHV). (C) Co-IP analysis of Rab10 with E proteins from the indicated coronaviruses. HeLa cells were co-transfected with Myc-Rab10 and HA-tagged E from the indicated coronaviruses. Cell lysates were immunoprecipitated with anti-HA beads and analyzed by western blot. (D) Rescue of Rab10 interaction with MHV E protein. HeLa cells were co-transfected with Myc-Rab10 and HA-tagged MHV E-WT or the I83V mutant. (E) SADS-CoV VLPs secretion assay. HeLa cells were transfected with VLPs plasmids (N/M/E) containing E-WT, ΔE, E-mt7, or E-V75A. Cell lysates and concentrated medium samples were analyzed by western blot for SADS-CoV N. The ratio of N in medium versus lysate is shown below. (F) Schematic representation and rescue of recombinant icSADS-CoV-GFP E-WT and E-V75A. Fluorescence images show the successful rescue and propagation of viruses at Passage 2 in Vero E6 cells. Scale bar, 275 μm. (G) Comparison of viral protein expression and release. HeLa cells were infected with icSADS-CoV-GFP E-WT or E-V75A (MOI = 1) for 16 h. Cell lysates and concentrated culture supernatants (Medium) were analyzed by western blot for SADS-CoV S and N proteins. (H–K) Quantitative analysis of SADS-CoV egress for E-WT and E-V75A viruses. Metrics include total viral RNA (H), relative intracellular mRNA (I), released viral RNA in the supernatant (J), and released infectious viral titers measured by TCID₅₀ (K). Data points represent individual biological replicates (n = 3). Data are presented as means ± SD from three independent experiments. Statistical significance was analyzed by unpaired t-test. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, ns, not significant.

https://doi.org/10.1371/journal.ppat.1014569.g011

We further examined the interaction between Rab10 and E proteins from TGEV, PEDV, PDCoV, SARS-CoV-2, and MHV. Co-IP results demonstrated that the E proteins of TGEV, PEDV, PDCoV, and SARS-CoV-2 all interact with Rab10, whereas MHV E fails to do so (Fig 11C). To confirm the role of the terminal valine in this interaction, we introduced an I83V mutation into the MHV E C-terminus. This single substitution was sufficient to enhance its binding affinity for Rab10 (Fig 11D), identifying the C-terminal valine as a key molecular determinant for Rab10 engagement across multiple coronavirus genera.

The functional consequences of these mutations were first evaluated using the VLPs secretion system. Compared to the wild-type (E-WT), both the Emt7 (residues 64–75) and E-75A mutants markedly reduced the amount of the secretion of N protein into the culture supernatant, indicating impaired VLPs egress (Fig 11E). To validate these phenotypes during authentic infection, we generated a recombinant SADS-CoV mutant carrying the V75A substitution, designated icSADS-CoV-GFP E V75A using reverse genetics (Fig 11F). Although the V75A mutant remained viable and could be rescued, it exhibited a pronounced release defect in HeLa cells. Western blot analysis of infected HeLa WT cells showed that while intracellular levels of N and S proteins were comparable to or even higher than those of WT, their accumulation in the culture supernatant was markedly reduced (Fig 11G). Consistently, the V75A mutant showed slightly increased intracellular viral mRNA and approximately 2.3-fold higher intracellular viral RNA levels, but extracellular viral RNA and progeny infectious titers were reduced by approximately 10-fold and 100-fold, respectively (Fig 11H–K).

We further examined the V75A phenotype in porcine LLC-PK1 cells to confirm its physiological relevance (Fig 12A). Cells were initially infected with the indicated viruses during the first round of infection (P1). P1 supernatants were collected for the measurement of viral RNA and infectious virus titers when GFP fluorescence becomes uniformly bright and comparable between the two groups. However, when equal volumes of P1 culture supernatants were transferred to LLC-PK1 cells, the V75A mutant produced markedly weaker GFP signals during the second passage, indicating reduced production of extracellular infectious progeny (Fig 12B). Total viral RNA yields and intracellular viral RNA levels were not significantly different between the wild-type and V75A viruses (Fig 12C and D). In contrast, extracellular viral RNA was reduced by approximately one order of magnitude, and the released infectious titer decreased by approximately two orders of magnitude in the V75A mutant (Fig 12E and F). These results confirm that the V75A substitution primarily impairs viral egress rather than affecting intracellular viral RNA accumulation, and that this phenotype is conserved in a porcine cell line.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 12. The V75 residue of the SADS-CoV E protein is also required for viral release in LLC-PK1 cells and proposed model for Rab10-mediated SADS-CoV egress via the ERGIC-TGN-lysosome axis.

(A) Schematic representation of the two-round infection assay used to compare recombinant SADS-CoV carrying the wild-type E protein (icSADS-CoV-GFP E-WT) with the E-V75A mutant (icSADS-CoV-GFP E-V75A). Cells were initially infected with the indicated viruses during the first round of infection (P1). P1 supernatants were collected for the measurement of viral RNA and infectious virus titers when GFP fluorescence becomes uniformly bright and comparable between the two groups. Equal volumes of P1 supernatants were subsequently used to infect cells during the second round of infection (P2), and infection was evaluated by GFP expression. Created with BioRender.com. https://BioRender.com/voc0v0c (B) Representative fluorescence images showing GFP expression in cells infected with icSADS-CoV-GFP E-WT or icSADS-CoV-GFP E-V75A during P1 and P2. Comparable GFP signals were observed for the two viruses during P1, whereas GFP expression was markedly reduced in cells infected with the E-V75A mutant during P2. Scale bar, 275 μm. (C–F) Quantitative analysis of SADS-CoV egress for E-WT and E-V75A viruses. Metrics include total viral RNA (C), relative intracellular mRNA (D), released viral RNA in the supernatant (E), and released infectious viral titers measured by TCID₅₀ (F). Data points represent individual biological replicates (n = 3). Data are presented as means ± SD from three independent experiments. Statistical significance was analyzed by unpaired t-test. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, ns, not significant. (G) Proposed model for Rab10-mediated SADS-CoV egress. In wild-type cells, during a productive infection, the viral E protein recruits and binds Rab10, facilitating its loading onto LMAN1-positive vesicles at the ERGIC. Rab10 subsequently coordinates the trafficking of virion-containing vesicles through TGN46-positive post-TGN compartments and TGN46/LAMP1-positive exocytic vesicle/lysosome compartments, ultimately facilitating virion release by exocytosis. In Rab10-knockout cells, the E protein exhibits a loss of ability to recruit Rab10, the loading of viral cargo into LMAN1-positive vesicles is disrupted. Consequently, the trafficking of post-TGN vesicles is stalled, and the exit of viral particles from the TGN/lysosome compartments is significantly impaired, leading to reduced viral egress from the host cell. Created with BioRender.com. https://biorender.com/ovfgvpj.

https://doi.org/10.1371/journal.ppat.1014569.g012

In summary, these findings identify the C-terminal residues 73–75 (specifically V75) as a critical molecular determinant for Rab10 recruitment. Disruption of this motif leads to the intracellular sequestration of viral components and a concomitant failure in efficient virion exocytosis.

Generation of plasmid constructs

The following plasmids were constructed and used in this study: pCAGGS-hRab10-Myc for expressing human Rab10; a series of pCAGGS-based plasmids for expressing C-terminally HA-tagged E proteins from SADS-CoV (including mutants), PEDV, TGEV, PDCoV, SARS-CoV-2, and MHV (including I83V mutant), as well as C-terminal HA-tagged M and N proteins from SADS-CoV; pLV2-TRE3GS-Rab10-TetOne-Blast for doxycycline-inducible expression of Rab10; pcDNA3.1-based plasmids for co-expressing the N, M and E protein (separated by a T2A cleavage site) from either SARS-CoV-2 or SADS-CoV for VLPs production; and a set of pCAGGS plasmids for expressing Myc- or Flag-tagged SADS-CoV M, E, and N proteins. All constructs were verified by DNA sequencing.

Cells and viruses

All cell lines, including HeLa (ATCC, CCL-2), Vero E6 (ATCC, CRL-1586), HEK 293T (ATCC, CRL-3216), BHK21 (ATCC, CCL-10), IPI-2I (Porcine Intestinal) and LLC-PK1 (Porcine kidney, ATCC, CL-101), were cultured in Dulbecco's Modified Eagle Medium (DMEM, Gibco, 11965092) supplemented with 10% fetal bovine serum (FBS, Jin YuanKang, FBS-300) and 1% penicillin-streptomycin (Sigma, #P4333) at 37°C in a humidified incubator with 5% CO₂. All cell lines tested negative for mycoplasma contamination. SADS-CoV (GenBank: MF094681), the icSADS-CoV-GFP (in which the GFP-T2A sequence is substituted at the 5’ end of the N gene) reporter virus, PEDV (GenBank: GU372744), TGEV (GenBank: EU074218) and PDCoV (GenBank: KU981059) were stored in our laboratory. The SADS-CoV, icSADS-CoV-GFP and PEDV viral titers were determined by measuring the TCID₅₀ on Vero E6 cells, while TGEV and PDCoV were propagated and their titers were determined by TCID₅₀ assays in LLC-PK1 cells. All viral culture and infection experiments in this study were conducted in Biosafety Level 2 (BSL-2) laboratories and were approved by the institutional biosafety committee in accordance with national biosafety regulations.

Antibodies

The following primary antibodies were used in this study: rabbit anti-Rab10 (Abcam, ab237703, IFA: 1:100, WB: 1:1000); Alexa Fluor 488 Anti-Rab10 antibody MJF-R23 (Abcam, ab302654, IFA: 1:50); rabbit anti-SADS-CoV N; mouse anti-SADS-CoV N; mouse anti-SADS-CoV S were prepared and maintained in our laboratory; mouse anti-SARS-CoV-2 N (ABclonal, A20142, WB: 1:1000); rabbit anti-TGN46 (ABclonal, A19618, IFA: 1:100, WB: 1:1000); rabbit anti-LMAN1 (ABclonal, A4941, IFA: 1:100, WB: 1:1000); rabbit anti-LAMP1 (ABclonal, A22194, IFA: 1:100, WB: 1:1000); rabbit anti-Alix (ABclonal, A25326, WB: 1:1000); rabbit anti-CD9 (ABclonal, A1703, WB: 1:1000); rabbit anti-Caveolin-1 (ABclonal, A22417, WB: 1:1000); rabbit anti-Calnexin (Abcam, ab22595, WB: 1:1000); mouse anti-dsRNA (Scicons, #10010500, IFA: 1:100)rabbit anti-GAPDH (Proteintech, #81640–5-AP, WB: 1:10000); rabbit anti-Flag (Proteintech, #80801–2-RR, WB: 1:5000); mouse anti-Flag (Proteintech, #66008–4-Ig, IFA: 1:200, WB: 1:5000); rabbit anti-Myc (Proteintech, #10828–1-AP, WB: 1:4000); mouse anti-HA (ABclonal, AE008, IFA: 1:100, WB: 1:1000). The secondary antibodies used were as follows: Alexa Fluor 488 Goat Anti-Rabbit IgG H&L (Abcam, ab150077, IFA: 1:500); Alexa Fluor 488 Goat Anti-Mouse IgG H&L (Abcam, ab150113, IFA: 1:500); Alexa Fluor 594 Goat Anti-Rabbit IgG H&L (Abcam, ab150080, IFA: 1:500); Alexa Fluor 594 Goat Anti-Mouse IgG H&L (Abcam, ab150116, IFA: 1:500) and Alexa Fluor 647 Donkey Anti-Rabbit IgG H&L (Abcam, ab150075, IFA: 1:500). For immunoblotting, membranes were probed with the indicated primary antibodies, followed by incubation with the following secondary antibodies (LI-COR) diluted at 1:10,000: goat anti-rabbit IRDye 800CW (#926–32211), goat anti-rabbit IRDye 680RD (#926–68071), or goat anti-mouse IRDye 800CW (#926–32210). Blots were visualized and quantified using an Odyssey CLx Imaging System and Image Studio software (version 5.2).

Rapid reconstruction of SADS-CoV using the yeast homologous recombination platform

The full-length infectious clones pCC1BAC-SADS-CoV, pCC1BAC-SADS-CoV-GFP, and pCC1BAC-SADS-CoV E mutant were generated using a yeast homologous recombination platform. Specifically, the yeast replication and selection elements, consisting of the 2μ ori-TRP1 promoter-TRP1 sequence (2558 bp) from the pGBKT7 vector (Clontech, 630442), were integrated into the pCC1BAC vector. This modification enabled the plasmid to replicate in yeast. Concurrently, the cytomegalovirus (CMV) early promoter, the hepatitis delta virus (HDV) ribozyme sequence, and the bovine growth hormone (BGH) termination signal were inserted into the linearized pCC1BAC vector to create the intermediate plasmid pCC1BAC-CMV-HDVRz-BGH. For the construction of pCC1BAC-SADS-CoV-GFP, eight overlapping cDNA fragments were amplified and purified from SADS-CoV cDNA. A GFP-T2A sequence was introduced at the first start codon of the N gene coding region within the eighth fragment. All fragments possessed homologous arms approximately 50–100 bp in length to facilitate recombination. Fragment amplification was performed using PrimeSTAR Max DNA polymerase (Takara, R045A) according to the manufacturer's guidelines.

The linearized vector and the eight overlapping fragments were co-transformed into MaV203 yeast competent cells (Thermo, 11445012) via a standard yeast plasmid transformation protocol. The transformation mixture was subsequently plated on SD-Trp tryptophan selection agar plates and incubated at 30°C for 48 h to select for positive yeast colonies. Following screening and identification of positive yeast clones, the recombinant plasmids were propagated in DH10B electrocompetent cells (Thermo, 18290015) via electroporation (BioRad, Gene Pulser Xcel) and cultured in LB medium supplemented with 12.5 μg/mL chloramphenicol. The integrity of the obtained infectious clones was verified through full-length sequencing. These verified clones were then transfected into BHK21 cells using Lipofectamine 3000 (Thermo, L3000075) for 48 h, and the rescued virus could subsequently be passaged in Vero E6 cells and harvested for further study.

The recombinant icSADS-CoV-GFP E-Flag genome was generated by inserting a Gly-Ser linker and a FLAG epitope immediately upstream of the native stop codon of the SADS-CoV E gene. The inserted nucleotide sequence, 5′-GGTAGTGACTACAAA GACGATGACGACAAG-3′, encodes GS-DYKDDDDK and is followed by the native TAA stop codon. This C-terminal tagging strategy preserved the E open reading frame, the downstream E–M genomic junction, and the M open reading frame. To generate recombinant viruses carrying mutations in the E gene, the codon corresponding to the indicated amino acid was replaced with an alanine codon or a stop codon, as appropriate. The modified sequences were introduced into the full-length icSADS-CoV-GFP infectious clone, and the recombinant viruses were rescued using our established reverse-genetics system. The modified regions of the rescued viruses were verified by RT-PCR followed by Sanger sequencing.

Genome-wide CRISPR knockout screen

A human GeCKO CRISPR knockout pooled library containing 123,411 sgRNAs targeting 19,050 human genes was a gift from Feng Zhang (Addgene, #1000000048). The library was amplified in Endura competent cells (Lucigen) and purified using a Plasmid Maxi Kit (Qiagen, #12163). For lentiviral packaging, HEK 293T cells were co-transfected with the sgRNA library plasmid, psPAX2 (Addgene, #12260), and pMD2.G (Addgene, #12259) at a 4:3:2 mass ratio using PEI MAX 40K transfection reagent (Polysciences, 24765). Viral supernatants were collected 48 hpi, concentrated with Lenti-X Concentrator (Clontech, 631231), and stored at -80°C. For the CRISPR screen, HeLa cells were transduced with the packaged lentiviral sgRNA library at an MOI of 0.3. Transduced cells were selected with puromycin for 7 days to generate a stable pooled knockout library. Approximately 1 × 10⁸ library cells were seeded in T175 flasks and inoculated with icSADS-CoV-GFP at an MOI of 0.01. Infection proceeded until near complete cytopathic effect was observed. The medium was then replaced to allow the survival and outgrowth of resistant cell colonies. Surviving cells were harvested, re-plated, and subjected to two additional rounds of identical viral challenge. Following the third round of selection, the resistant cell pool was expanded. Genomic DNA was extracted from approximately 1 × 10⁷ resistant cells isolated by fluorescence-activated cell sorting (FACS). As a control, genomic DNA was also extracted from 1 × 10⁷ uninfected library cells. sgRNA-encoding regions were amplified from the genomic DNA and subjected to deep sequencing on an Illumina NovaSeq 6000 platform. Enriched sgRNA sequences and their corresponding target genes were subsequently identified through bioinformatic analysis (S1 Table).

RNA interference

All siRNAs were synthesized by Seven Biotech. The target sequences for all siRNAs are listed in Supplementary Table (S2 Table). For transfection, HeLa cells, IPI-2I cells or LLC-PK1 cells were seeded in 12-well plates one day prior to reach 30–50% confluence. Then, 50 nM siRNA was diluted in 100 μL of Opti-MEM medium. The diluted siRNA was mixed with 4 μL Lipofectamine RNAiMAX (Thermo, 13778150) reagent and incubated for 15 min at room temperature to allow complex formation. The resulting siRNA-lipid complexes were then added to the cells. After 36 h of incubation, the cells were subjected to subsequent experimental treatments.

Generation of Rab10 knockout HeLa cell lines

To generate Rab10-knockout HeLa cells, a human Rab10-targeting sgRNA (5′- catcatccgaaaaacgaaaa-3′) was cloned into the lentiCRISPR v2 vector and packaged into lentiviral particles. Wild-type HeLa cells were then transduced with the resulting lentivirus, followed by isolation of single-cell clones. Successful knockout of the Rab10 gene was confirmed by DNA sequencing and western blot.

Generation of Dox-inducible cell lines expressing Rab10

Briefly, the cDNA encoding wild-type Rab10 was cloned into the pLV2-TRE3GS-MCS-TetOne-Blast vector to generate lentiviral transfer plasmids. The lentiviruses were then produced by co-transfecting HEK 293T cells with the lentiviral packaging plasmids psPAX2, pMD2.G, and the respective transfer plasmid. The viral supernatant was collected 48 h post-transfection and concentrated 50-fold using a lentiviral concentrator. Rab10-knockout HeLa cells were infected with the concentrated lentiviruses carrying either wild-type Rab10. Transduced cells were selected with blasticidin (5 μg/mL). To induce the expression of wild-type Rab10, cells were treated with 2.5 μg/mL doxycycline for 24 h.

Cell viability assay

Cell viability was assessed using the Cell Counting Kit-8 (CCK-8, APExBiO, K1018) according to the manufacturer's instructions. Approximately 1 × 10⁴ control or Rab10-knockout HeLa cells were seeded into each well of opaque-walled 96-well plates. After 48 h, CCK-8 reagent was added to each well, followed by incubation for 2 h in a cell culture incubator. Absorbance at 450 nm was then measured using a BioTek ELx808 microplate reader.

RNA extraction, reverse transcription, and RT-qPCR

Total RNA was extracted from tissue or cell samples using RNAiso EASY (Takara, TCH020) according to the manufacturer's instructions. The extracted RNA was reverse-transcribed into cDNA using the PrimeScript RT Reagent Kit with gDNA Eraser (Takara, RR047A). RT-qPCR was then performed with TB Green Premix Ex Taq II (Takara, RR820A) on a QuantStudio 5 system (Applied Biosystems). Relative gene expression levels were calculated using the GAPDH gene for normalization.

For viral genomic RNA detection, viral RNA was purified from cell culture supernatants using the MiniBEST Viral RNA/DNA Extraction Kit (Takara, 9766). Quantification was conducted by one-step reverse transcription qPCR using the One Step TB Green PrimeScript PLUS RT-PCR Kit (Takara, RR096A) on the QuantStudio 5 system, with absolute quantification performed based on a standard curve. All primer sequences used for RT-qPCR are listed in the Supplementary Table (S2 table).

Transmission electron microscopy (TEM)

For ultrastructural examination, HeLa cells were either mock-infected or inoculated with SADS-CoV at an MOI of 1. At 24 hpi, cells were subjected to primary fixation with 2% glutaraldehyde for 30 min at room temperature. Following three washes in 0.1 M sodium cacodylate buffer, the samples underwent secondary fixation on ice using a solution of 1% (w/v) reduced osmium tetroxide, 0.1 M sodium cacodylate, and 1.5% (w/v) cyanoferrate. After thorough rinsing with buffer and distilled water, the specimens were dehydrated through a graded ethanol series and embedded in resin. Ultrathin sections (50–70 nm) were then prepared using a diamond knife and collected on Formvar-coated copper grids. For contrast enhancement, the grids were double-stained with 2% uranyl acetate (5 min) and lead citrate (2 min). Final imaging was performed at various magnifications (1,500 × to 12,000 ×) using a HITACHI H7650 transmission electron microscope.

VLPs purification

HeLa WT or HeLa Rab10^-/- KO1 cells were co-transfected with either pcDNA3.1-SARS-CoV-2-N-T2A-M-IRES-E or pcDNA3.1-SADS-CoV-N-T2A-M-IRES-E. At 48 h post-transfection, cell culture supernatants were harvested, filtered through 0.22 μm membranes, and clarified by centrifugation (10,000 × g, 30 min, 4°C). For the drug inhibition assay, DMSO, BFA (MCE, HY-16592; 5 μM), or CID-1067700 (MCE, HY-13452; 40 μM) were added to the culture medium at 12 h post-transfection, and supernatants were clooected at 48 h post-transfection. To concentrate VLPs, the clarified supernatants were subjected to ultracentrifugation (28,000 × g, 2 h, 4°C) over a 20% sucrose cushion. The resulting VLPs pellets were resuspended in phosphate-buffered saline (PBS) buffer for western blot.

Isopycnic gradient centrifugation and rate-zonal centrifugation in sucrose gradients

For differential centrifugation, the cell culture supernatant from SADS-CoV-infected cells was first clarified by centrifugation at 1,000 × g for 10 min at 4°C to remove cells and debris. The supernatant was further cleared by centrifugation at 10,000 × g for 30 min at 4°C. The clarified supernatant was then layered onto a 20% (w/v) sucrose cushion and pelleted by ultracentrifugation at 100,000 × g for 2 h. The resulting pellet was gently resuspended in PBS and incubated overnight at 4°C with mild agitation. The resuspended material was analyzed by western blot.

For rate-zonal centrifugation, the pellet obtained from differential centrifugation was resuspended and layered onto a continuous 10–60% (w/v) sucrose gradient. The gradient was centrifuged at 167,000 × g for 2 h at 4°C using a 55Ti rotor in an Optima XPN-100 ultracentrifuge (Beckman Coulter). Following centrifugation, approximately 50 fractions were collected sequentially from the top of the gradient. The distribution of viral and cellular marker proteins across the gradient was assessed by western blot analysis of individual fractions.

ELISA for detection of Rab10 in cell culture supernatant

The concentration of Rab10 protein in cell culture supernatant was quantified using a sandwich enzyme-linked immunosorbent assay (JL20386, Jonlnbio, China). Supernatants were collected 24 h post-transfection and corresponding treatments. The samples were centrifuged at 1,000 × g for 20 min, and the resulting supernatant was diluted 50-fold. Subsequently, 100 µL of the diluted sample was added to the ELISA microplate wells pre-coated with a Rab10 capture antibody and incubated at 37°C for 1 h. Following incubation and washing, 100 µL of biotin-labeled detection antibody was added to each well and incubated at 37°C for 1 h. After another wash step, 100 µL of horseradish peroxidase (HRP)-conjugated streptavidin was added and incubated at 37°C for 30 min. The plate was then thoroughly washed to remove unbound conjugates. For color development, 90 µL of tetramethylbenzidine (TMB) substrate was added to each well at 37°C for 15 min. The enzymatic reaction was stopped by adding 50 µL of stop solution. The absorbance was immediately measured at a wavelength of 450 nm using a microplate reader (BioTek ELx808, USA), and the sample concentration was calculated based on the standard curve.

Western blot and Co-IP assay

Cells were lysed in RIPA buffer (Beyotime, P0013B) containing PMSF. The lysates were mixed with SDS-PAGE sample loading buffer (Beyotime, P0015), boiled at 100°C for 10 min, separated by SDS-PAGE, and transferred onto nitrocellulose membranes. For the Co-IP assay, cell lysates were incubated with tag-specific antibodies at 4°C for 4 h, followed by incubation with Protein A/G beads (Thermo 88802) for 6 h. After washing with IP washing buffer, the beads were boiled in loading buffer and the proteins were resolved by SDS-PAGE. Membranes were blocked with 5% non-fat milk in TBST for 1 h at room temperature and then incubated with primary antibodies overnight at 4°C. After washing, the membranes were probed with appropriate IRDye-conjugated secondary antibodies for 1 h at room temperature. Signals were detected using an Odyssey CLx imaging system (LI-COR) at 680 and 800 nm, and band intensities were quantified with Image J. Information on specific antibodies can be found in the antibody section of Materials and methods.

IFA and confocal microscopy

HeLa WT or Rab10^-/- KO cells grown on 35-mm glass-bottom dishes (Cellvis, D35-10–1-N) were infected with SADS-CoV. For immunofluorescence, cells were fixed with 4% paraformaldehyde for 30 min (ensuring complete virus inactivation), permeabilized with 0.3% Triton X-100 for 10 min at 4°C, and blocked with 5% BSA for 1 h at room temperature. Subsequently, cells were incubated overnight with primary antibodies, followed by appropriate secondary antibodies for 1 h. Information on specific antibodies can be found in the antibody section of Materials and methods. Nuclei were stained with DAPI (Beyotime, C1002). Images were acquired using a Zeiss LSM880 confocal microscope equipped with a 60 × oil-immersion objective. For each experiment, 15–20 random fields were recorded. Maximum intensity projections were generated to analyze protein subcellular localization. Colocalization was quantified using Mander's correlation coefficient or Pearson's correlation coefficient with Zeiss analysis software.

Gaussia luciferase assay

Gaussia luciferase activity was measured using the Gaussia-Lumi Gaussia Luciferase Assay Kit (Beyotime, RS072S). HeLa cells were transfected with the pGluc Gaussia luciferase reporter plasmid (Beyotime, D2098). After treatment with DMSO or BFA, cell culture supernatants were collected, and Gaussia luciferase activity was analyzed according to the manufacturer's instructions. Briefly, 100 µL of supernatant was collected from treated wells into a 96-well plate, followed by the addition of 100 µL of Gaussia-Lumi detection working solution per well and thorough mixing. After incubation at room temperature for 10 min, chemiluminescence was measured using an Enspire multimode plate reader (PerkinElmer, USA).

Statistical analyses

Analysis of experimental data, standard deviation (SD) calculations, and graph plotting from the results of three independent experiments were performed using GraphPad Prism software (version 9.0). All statistical analyses were performed using either one-way ANOVA with Dunnett's test or an unpaired t-test. Significance levels are as follows: ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, ns, not significant.