Binding affinity of PEDV S protein for pAPN revealed by docking structure and SPR

The three-dimensional structure of the CoV S protein plays a key role in receptor recognition [21]. In this study, we first performed a structural alignment of the PEDV S protein with those of TGEV and SADS-CoV. Since the structure of the TGEV S protein remains unreported, we used the I-TASSER platform to predict its three-dimensional structure before alignment. Alignment results showed that the C-terminal domain (CTD) of the S1 proteins of PEDV and TGEV contain two layers of β-sheets, whereas the CTD of SADS-CoV S1 protein only has one twisted β-sheet (Fig 1A). Due to gene duplication, both the S1 proteins of PEDV and TGEV contain a second N-terminal domain (NTD), while SADS-CoV S1 protein contains only one NTD. In contrast to the NTD and CTD of S1 protein having unique structural features, the S2 proteins of the three viruses were structurally conserved and aligned well. Moreover, we further aligned the amino acid sequences of the S protein from the three viruses and performed a homology analysis (S1A Fig). The results showed that the amino acid homology between the S proteins of PEDV and TGEV was 31.5%, while the homology between the S proteins of PEDV and SADS-CoV was only 18.8% (S1B Fig). Therefore, these results indicate that the structural and sequence basis of the PEDV S protein is more similar to that of TGEV than to SADS-CoV.

Download:

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

Fig 1. Binding affinity of pAPN with the S proteins of PEDV, TGEV, and SADS-CoV revealed by docking structure and SPR.

(A) Structural alignment of the S proteins from PEDV, TGEV and SADS-CoV. (B) Docking structure and atomic details illustrating the interaction between PEDV S protein (green) and pAPN (pink). (C) Docking structure and atomic details illustrating the interaction between TGEV S protein (purple) and pAPN (pink). (D) Docking structure and atomic details illustrating the interaction between SADS-CoV S protein (yellow) and pAPN (pink). Contacting residues are represented as sticks, with hydrogen bonds shown as red lines and salt bridges as yellow lines. (E) Binding affinities of pAPN to the S1 proteins of PEDV, TGEV, and SADS-CoV measured by SPR.

https://doi.org/10.1371/journal.ppat.1013317.g001

To investigate the binding of pAPN with PEDV S protein, we first conducted the molecular docking analysis of pAPN with the S proteins of PEDV, TGEV, and SADS-CoV, respectively. Hydrogen bonds and salt bridges between key contact residues were discovered in the docking structures of pAPN with the S proteins of three viruses. Within a distance cutoff of 5 Å, the docking structure analysis identified four residues (T724, R490, K633 and S691) in the PEDV S protein that interact with pAPN (Fig 1B), while six residues (G541, K408, Y402, T645, K538 and T549) in the TGEV S protein were found to interact with pAPN (Fig 1C). Similarly, residues interacting through hydrogen bonds or salt bridges (K367, Y365, F294, V357, N358) also exhibited in the docking structure between pAPN and SADS-CoV S protein (Fig 1D). We further predicted the binding affinity between pAPN and the S proteins of the three viruses using the Rosetta Interface-Analyzer. Typically, docking structures of proteins with a free binding energy below -15 kcal/mol are considered to exhibit a binding propensity [22]. It showed that the free binding energies of pAPN with the S proteins of PEDV, TGEV and SADS-CoV were -32.406, -48.748, and -7.892 kcal/mol, respectively (S1C Fig). Additionally, the binding affinity between pAPN and PDCoV S protein was used to validate the accuracy of molecular docking (S2 Fig), with a free binding energy is -37.840 kcal/mol, indicating that pAPN exhibits a strong binding affinity for PDCoV, consistent with previous reports. Lower free binding energy indicating more stable binding and stronger affinity, although the free binding energy of PEDV with pAPN (-32.406 kcal/mol) is higher than those of TGEV (-48.748 kcal/mol) and PDCoV (-37.840 kcal/mol), PEDV shows a binding propensity for pAPN, especially in comparison to SADS-CoV (-7.892 kcal/mol). Although molecular docking provides valuable structural insights and preliminary predictions of potential interactions, it does not allow for quantitative assessment of binding strength. We then performed surface plasmon resonance (SPR) to directly evaluate the binding affinities between pAPN and the S1 proteins of PEDV, TGEV, and SADS-CoV. The SPR results showed that pAPN binds to PEDV S1 protein with a lower affinity (KD = 47.46 ± 6.44 μM) compared to TGEV S1 protein (K_D = 391.88 ± 37.37 nM), while no detectable binding was observed between pAPN and SADS-CoV S1 protein (Fig 1E). The binding energy differences between pAPN-PEDV and pAPN-TGEV, derived from the dissociation constants obtained by SPR, were in good agreement with those predicted by molecular docking analyses. This consistency supports the validity of the proposed interaction models of molecular docking. The binding affinity between PEDV and pAPN suggests that pAPN may play an important role in PEDV infection, despite not being the primary receptor for the virus.

pAPN promotes PEDV replication in susceptible but not nonsusceptible cells

Next, we investigated whether the binding affinity between PEDV and pAPN, revealed by molecular docking and SPR, ultimately promotes PEDV infection in various cells. Here, we conducted an immunofluorescence assay (IFA) to visualize the effect of pAPN on PEDV, TGEV, and SADS-CoV infection in IPI-2I-APN^-/- cells lacking endogenous APN and intestinal porcine epithelial cells (IPECs) expressing endogenous APN. Fluorescence images showed that in IPI-2I-APN^-/- cells, PEDV infection was reduced but not abolished, whereas TGEV infection was completely inhibited. Notably, re-expression of pAPN in the knockout cells significantly enhanced PEDV replication and fully restored TGEV infection. Moreover, pAPN also significantly promoted the infection of PEDV and TGEV in IPECs, suggesting that pAPN promotes PEDV replication in susceptible cells, even in IPI-2I-APN^-/- cells lacking endogenous pAPN (Fig 2A and 2B). In addition, pAPN had no effect on SADS-CoV infection in either IPECs or IPI-2I-APN^-/- cells, consistent with SADS-CoV not utilizing pAPN as a receptor for entry (Fig 2A and 2B). We then performed RT-qPCR and western blotting assays in IPI-2I-APN^-/- cells at 24 hours post-infection (hpi). RT-qPCR results showed that pAPN enhanced PEDV RNA levels by approximately 7.8-fold. In contrast, TGEV replication in IPI-2I-APN^-/- cells was exhibited a profound dependence on pAPN, with re-expression of pAPN resulting in a more than 2000-fold increase in RNA levels relative to cells lacking pAPN. Consistent with the IFA results, pAPN had no effect on SADS-CoV infection (Fig 2C). Western blotting analysis further confirmed that pAPN significantly promoted PEDV replication in IPI-2I-APN^-/- cells, despite PEDV being able to infect cells even without pAPN (Fig 2D and 2E). Moreover, pAPN rescued TGEV infection in IPI-2I-APN^-/- cells, while the presence or absence of pAPN had no effect on the susceptibility to SADS-CoV. Importantly, we further evaluated the role of pAPN in PEDV infection in Caco2 cells, and the results showed that PEDV failed to infect these cells, even with pAPN overexpression (S3A Fig). However, pAPN overexpression allowed TGEV to infect Caco2 cells (S3B Fig). Overall, these findings suggest that pAPN facilitates PEDV replication in susceptible cells but could not rescue PEDV infection in non-susceptible cells.

Download:

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

Fig 2. The role of pAPN in the infection of PEDV, TGEV and SADS-CoV.

(A) Effect of pAPN on PEDV, TGEV, and SADS-CoV infection in IPI-2I-APN^-/- cells and IPECs. Cells were infected with PEDV, TGEV and SADS-CoV (MOI = 0.1) in the presence of 5 μg/mL trypsin. The N proteins of PEDV, TGEV and SADS-CoV were stained using IFA at 24 hpi. Scale bar, 50 μm. (B) Mean intensity of the N protein from (A) was quantified (n = 5 cells; mean ± SD). (C) IPI-2I-APN^-/- cells re-expressing pAPN were infected with PEDV, TGEV and SADS-CoV (MOI = 0.1), and relative viral RNA levels were quantified by RT-qPCR at 24 hpi (n = 3 independent experiments; mean ± SD). (D) Western blotting of N proteins in IPI-2I-APN^-/- cells re-expressing pAPN after infection with PEDV, TGEV and SADS-CoV (MOI = 0.1) at 24 hpi. (E) Relative density of the N proteins from (D) (n = 3 independent experiments; mean ± SD). Non-transfected infected cells served as controls in all independent experiments. Two-tailed P-values were calculated by unpaired Student’s t test. P < 0.05 was considered significant (ns P ≥ 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

https://doi.org/10.1371/journal.ppat.1013317.g002

pAPN facilitates PEDV infection via internalization not attachment

To investigate whether pAPN mediates PEDV infection by promoting viral entry, we conducted comparative experiments involving PEDV, TGEV, and SADS-CoV. IPECs with or without pAPN overexpression were infected with a 10 multiplicity of infection (MOI) of each virus at 4 °C for 30 min to allow attachment, followed by incubation at 37 °C for 1 h to internalize. The RT-qPCR results showed that pAPN overexpression did not significantly affect the attachment of PEDV, TGEV or SADS-CoV (Fig 3A, 3C and 3E). However, pAPN overexpression significantly enhanced PEDV internalization by 1.2-fold (Fig 3B) and TGEV internalization by 7-fold (Fig 3D) compared to the controls, whereas no significant effect was observed on SADS-CoV internalization (Fig 3F). These findings emphasize that pAPN promotes PEDV and TGEV infection through internalization rather than attachment. To visualize pAPN-mediated virus internalization, we labeled the lipid envelopes of PEDV, TGEV, SADS-CoV, and PDCoV with the lipophilic dye 1,1’-Dioctadecyl-3,3,3’,3’-tetramethylindodicarbocyanine (DiD). IFA confirmed that the virions were successfully labeled with DiD, with no effect on infectivity [15,22], indicating that this method effectively visualizes virions (S4 Fig). We then used confocal microscopy to visualize the colocalization of DiD-PEDV, DiD-TGEV and DiD-SADS-CoV with pAPN during the attachment and internalization stages. Fluorescence images showed low colocalization between pAPN and the three viruses during the attachment stage (Fig 3G). At 60 min post-infection (mpi), DiD-PEDV, DiD-TGEV and DiD-SADS-CoV were successfully internalized, and the fluorescence intensity profiles showed the obvious colocalization of DiD-PEDV and DiD-TGEV with pAPN, while DiD-SADS-CoV exhibited low colocalization with pAPN (Fig 3H). Statistical analysis revealed that the colocalization ratio of pAPN with PEDV and TGEV was significantly higher during the internalization stage than during the attachment stage, whereas SADS-CoV exhibited low colocalization with pAPN at both stages (Fig 3I). Moreover, pAPN significantly increased the number of PEDV and TGEV particles during the internalization stage, but not during the attachment stage. In contrast, pAPN did not affect the number of SADS-CoV particles at either the attachment or internalization stage (Fig 3J and 3K). These results reveal the unique role of pAPN in promoting the internalization rather than the attachment of PEDV, which is consistent with TGEV, suggesting similarity in pAPN-mediated entry among these viruses.

Download:

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

Fig 3. pAPN promotes PEDV and TGEV internalization rather than attachment.

(A-F) The roles of pAPN in the attachment (0 mpi) and internalization (60 mpi) of PEDV, TGEV, and SADS-CoV were assessed using RT-qPCR (n = 4 experiments; mean ± SD). Control represents the viral RNA levels in cells without pAPN transfection. (G and H) Colocalization fluorescence imaging of DiD-PEDV, DiD-TGEV and DiD-SADS-CoV with pAPN during the attachment and internalization stages. Scale bar, 10 µm. (I) Colocalization analysis of DiD-PEDV, DiD-TGEV and DiD-SADS-CoV with pAPN in (G and H) (three experiments; n = 5 cells; mean ± SD). (J and K) The number of particles of the three viruses per cell during the attachment and internalization stages. Two-tailed P-values were calculated using an unpaired Student’s t test. P < 0.05 was considered significant (ns P ≥ 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

https://doi.org/10.1371/journal.ppat.1013317.g003

pAPN initiates PEDV internalization pathways resembling a receptor

Activation of endocytic signaling is caused by receptor binding, and both CME and CavME are extensively used by enveloped viruses for internalization. Previous studies have demonstrated that PEDV, TGEV, and SADS-CoV enter cells through CME and CavME [16,23,24]. To determine whether pAPN is involved in initiating CME and CavME of PEDV, we evaluated the colocalization of DiD-PEDV with pAPN and clathrin (Cla)/caveolin-1 (Cav1) in IPECs at 60 mpi. Fluorescence images showed that DiD-PEDV colocalized with pAPN, as well as with clathrin-coated pits (CCPs) or caveolae, which was confirmed by triple-color colocalization analysis (Fig 4A-4D). We then used single-virus tracking to uncover the dynamics of CME and CavME initiated by PEDV upon binding to pAPN. Representative time-lapse images and kymographs (Fig 4E and 4F, S1 Movie) showed that DiD-PEDV first binds pAPN at the attachment sites, followed by the recruitment of Cla/Cav1, leading to the de novo formation of CCPs/caveolae. Subsequently, DiD-PEDV was rapidly endocytosed, as revealed by the fluorescence intensities and trajectories (Fig 4G, 4H, 4J and 4K). pAPN-mediated PEDV internalization exhibited two stages: a binding stage with restricted movement and a moving stage with directed movement (Fig 4I and 4L). Similarly, we observed that TGEV also colocalized with pAPN and Cla/Cav1 during internalization (Fig 4M–4P), and we then visualized the dynamics of pAPN in the CME and CavME of TGEV via single-virus tracking. As shown in the time-lapse images and kymographs (Fig 4Q and 4R, S2 Movie), pAPN binding to TGEV activated the de novo formation of CCPs/caveolae for rapid internalization, consistent with the results observed for PEDV. DiD-TGEV colocalized with pAPN and CCPs/caveolae at the viral binding site, initially displaying slow movement before undergoing accelerated internalization, indicating that pAPN initiated CME and CavME of TGEV, which also exhibited two distinct stages (Fig 4S and 4X). Moreover, fluorescence images showed that SADS-CoV colocalized with Cla/Cav1 during internalization, but did not colocalize with pAPN (S5A-S5D Fig, S3 Movie). Single-virus tracking also confirmed that SADS-CoV could undergo both CME and CavME independently of pAPN (S5E–S5N Fig), suggesting that pAPN is not essential for SADS-CoV internalization. These results indicate that pAPN can initiate CME and CavME to facilitate PEDV internalization upon binding to PEDV. Although pAPN is not a receptor for PEDV, it initiates PEDV internalization pathways resembling a receptor.

Download:

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

Fig 4. pAPN mediates PEDV and TGEV internalization via CME and CavME.

(A-D) Colocalization images and triple-color colocalization ratio of PEDV, pAPN and Cla/Cav1 at 60 mpi. Scale bar, 10 µm. (E and F) Time-lapse images and kymographs (DiD-PEDV: red; pAPN: green; Cla/Cav1: white) showing PEDV internalization via CME and CavME initiated by pAPN. Scale bar, 2 µm. (G) Fluorescence intensities (green and orange lines) and velocities (blue line) of the circled PEDV in (E). (H) Trajectories of the circled PEDV in (E), showing the binding stage (green) and moving stage (orange). Scale bar, 2 µm. (I) MSD plots of the circled PEDV during the binding stage and moving stage in (E). (J) Fluorescence intensities (green and orange lines) and velocities (blue line) of the circled PEDV in (F). (K) Trajectories of the circled PEDV in (F), showing the binding stage (green) and moving stage (orange). Scale bar, 2 µm. (L) MSD plots of the circled PEDV during the binding stage and moving stage in (F). (M-P) Colocalization images and triple-color colocalization ratio of TGEV, pAPN and Cla/Cav1 at 60 mpi. Scale bar, 10 µm. (Q and R) Time-lapse images and kymographs (DiD-TGEV: red; pAPN: green; Cla/Cav1: white) showing TGEV internalization via CME and CavME initiated by pAPN. Scale bar, 2 µm. (S) Fluorescence intensities (green and orange lines) and velocities (blue line) of the circled TGEV in (Q). (T) Trajectories of the circled TGEV in (Q), showing the binding stage (green) and moving stage (orange). Scale bar, 2 µm. (U) MSD plots of the circled TGEV during the binding stage and moving stage in (Q). (V) Fluorescence intensities (green and orange lines) and velocities (blue line) of the circled TGEV in (R). (W) Trajectories of the circled TGEV in (R), showing the binding stage (green) and moving stage (orange). Scale bar, 2 µm. (X) MSD plots of the circled TGEV during the binding stage and moving stage in (R).

https://doi.org/10.1371/journal.ppat.1013317.g004

pAPN initiates PEDV internalization less efficiently than TGEV

To further explore the dynamics of PEDV internalization mediated by pAPN, we employed single-virus tracking to observe the internalization of PEDV, TGEV, and SADS-CoV in IPECs overexpressing pAPN. In the representative time-lapse images (Fig 5A, S4A Movie), pAPN gradually accumulated around DiD-PEDV at the binding site, consistent with the results observed in the kymographs and fluorescence intensity analysis (Fig 5B). Moreover, the trajectory, velocity, and mean squared displacement (MSD) revealed two stages of PEDV internalization: a binding stage with restricted movement and a moving stage with directional movement (Fig 5C–5E). Similarly, we also observed TGEV internalization events mediated by pAPN (S4B Movie). Time-lapse images and kymographs showed that TGEV could also recruit pAPN to the binding site and undergo rapid internalization after colocalization (Fig 5F), further supported by fluorescence intensity analysis (Fig 5G). Viral movement modes also exhibited that TGEV colocalized with pAPN at a low speed, followed by entry into cells at a rapid speed and directed movement (Fig 5H–5J). Additionally, we visualized the internalization dynamics of PDCoV, which also uses pAPN as a receptor, and the results showed that the internalization dynamics of PDCoV initiated by pAPN were similar to TGEV (S6 Fig, S4C Movie). In contrast, SADS-CoV internalized successfully without recruiting pAPN (Fig 5K and 5L), following the same two stages of binding stage and moving stage (Fig 5M–5O), without pAPN involvement (S4D Movie). Additional time-lapse images and kymographs are shown in S7 Fig and S5 Movie.

Download:

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

Fig 5. Dynamics of PEDV, TGEV, and SADS-CoV internalization.

(A) Time-lapse images and kymographs of PEDV internalization mediated by pAPN. (B) pAPN fluorescence intensities (green line) and velocities (orange line) of the circled PEDV in (A). (C) Trajectories of the circled PEDV in (A), showing the binding stage (blue) and moving stage (red). (D and E) MSD plots of the circled PEDV during the binding stage and the moving stage in (A). (F) Time-lapse images and kymographs of TGEV internalization mediated by pAPN. (G) pAPN fluorescence intensities (green line) and velocities (orange line) of the circled TGEV in (F). (H) Trajectories of the circled TGEV in (F), showing the binding stage (blue) and moving stage (red). (I and J) MSD plots of the circled TGEV during the binding stage and the moving stage in (F). (K) Time-lapse images and kymographs of SADS-CoV internalization. (L) pAPN fluorescence intensities (green line) and velocities (orange line) of the circled SADS-CoV in (K). (M) Trajectories of the circled SADS-CoV in (K), showing the binding stage (blue) and moving stage (red). (N and O) MSD plots of the circled SADS-CoV during the binding stage and the moving stage in (K). Scale bar, 2 µm.

https://doi.org/10.1371/journal.ppat.1013317.g005

According to these two stages of viral internalization, we analyzed the kinetics of their pAPN-mediated internalization movements. The zero time point of the internalization events is defined as the moment when virions attachment to the cell membrane begins to recruit pAPN. The internalization duration was calculated as the sum of the binding and moving durations. The binding duration was defined as the time elapsed from the beginning of pAPN recruitment to the last frame before the rapid and directed movement of virions. The moving duration was defined as the time elapsed from the beginning of the rapid and directed movement of virions to the last frame before the virions speed slowed. Statistical analysis showed that the internalization durations of PEDV, TGEV, and PDCoV mediated by pAPN were 220.3 ± 23.6 s, 157.1 ± 38.6 s, and 154.0 ± 21.4 s, respectively (Fig 6A). The binding durations of PEDV, TGEV, and PDCoV internalization mediated by pAPN were 198.7 ± 24.0 s, 134.1 ± 38.6 s, and 132.1 ± 22.9 s, respectively (Fig 6B). The moving durations of PEDV, TGEV, and PDCoV internalization mediated by pAPN were 21.3 ± 2.8 s, 22.3 ± 3.9 s, and 21.9 ± 4.5 s, respectively (Fig 6C). Among all internalization events of PEDV, TGEV, PDCoV, and SADS-CoV, pAPN mediated 35% of PEDV internalization events, 80% of TGEV internalization events, and 75% of PDCoV internalization events, but did not participate in SADS-CoV internalization (Fig 6D-6G). These results indicate that the binding duration of pAPN-mediated PEDV internalization is significantly longer than that of TGEV and PDCoV, while the moving duration is similar among the three viruses. Therefore, although pAPN supports PEDV internalization, the process is less efficient in initiating this process compare to its role in TGEV and PDCoV internalization.

Download:

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

Fig 6. Duration and proportion of internalization events mediated by pAPN.

(A) Internalization durations of PEDV (n = 15), TGEV (n = 16) and PDCoV (n = 15) mediated by pAPN. (B) Binding durations of PEDV, TGEV and PDCoV mediated by pAPN. (C) Moving durations of PEDV, TGEV, and PDCoV mediated by pAPN. (D-G) Proportion of pAPN-dependent (PEDV: n = 15; TGEV: n = 16; PDCoV: n = 15; SADS-CoV: n = 0) and pAPN-independent (PEDV: n = 28; TGEV: n = 4; PDCoV: n = 5; SADS-CoV: n = 10) internalization events. Two-tailed P-values were calculated by unpaired Student’s t test. P < 0.05 was considered significant.

https://doi.org/10.1371/journal.ppat.1013317.g006

To determine the impact of pAPN-facilitated internalization on downstream viral membrane fusion and replication kinetics, we first used DiD-PEDV and DiD-TGEV to infect IPECs overexpressing pAPN at a MOI of 100. When lipophilic dyes label the viral membrane, they spread into a larger area of the cell membrane upon fusion, resulting in increased fluorescence [25]. The fluorescence intensity of a single particle can thus serve as a reporter for membrane fusion. The impact of pAPN-facilitated internalization on PEDV fusion was quantitatively assessed by analyzing the intensity changes of DiD signals from individual viral particles at 30, 60, and 90 mpi. The results showed that DiD-TGEV exhibited significantly higher fluorescence intensity at all three time points compared to DiD-PEDV (S8A and S8B Fig). These findings suggest that the faster internalization of TGEV, mediated by pAPN, results in earlier and more efficient membrane fusion than PEDV. We next investigated whether the increased membrane fusion efficiency resulting from pAPN-facilitated internalization ultimately impacts productive infection. IPECs with or without pAPN transfection were infected at an MOI of 1, and viral titers were measured at multiple time points post-infection. The results showed that pAPN significantly enhanced PEDV titers at 6 and 12 hpi, whereas for TGEV, pAPN enhanced infectious virus production at all measured time points from 6 to 60 hpi (S8C and S8D Fig). For PEDV, pAPN may primarily function as an entry co-factor that facilitates initial internalization and membrane fusion, thereby conferring an early kinetic advantage. However, as the infection progresses, the involvement of other host factors or alternative entry pathways may reduce PEDV’s dependence on pAPN. In contrast, as a functional receptor for TGEV, pAPN is essential for productive infection, and TGEV remains highly reliant on pAPN throughout its entire replication cycle. These results underscore the biological relevance of entry kinetics and support the view that pAPN-mediated internalization not only affects entry efficiency but also shapes the temporal dynamics of infection establishment. The acceleration of early entry steps facilitated by pAPN enables viruses to initiate replication sooner, thereby shortening its overall replication cycle.

Cell culture

IPECs, Vero cells, ST cells, LLC-PK1 cells, and Caco2 cells were obtained from the American Type Culture Collection (ATCC). The IPI-2I-APN^-/- cell line was kindly provided by Dr. Shaobo Xiao (Huazhong Agricultural University, China) [48]. All cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with high-glucose (Vivacell, China), 10% fetal bovine serum (FBS, Vivacell, China) and 100 U/mL penicillin and streptomycin. All cells used in this study were maintained in a humidified incubator at 37 °C with 5% CO₂.

Virus propagation, purification, and labeling

The PEDV CV777 strain (GenBank accession No. AF353511.1), PEDV ZJU strain (GenBank accession No. KU558701.1) and SADS-CoV GDWT strain (GenBank accession No. MG557844) used in this study were propagated in Vero cells. The TGEV H16 strain (GenBank accession No. FJ755618.2) was propagated in ST cells. The PDCoV CZ2020 strain (GenBank accession No. OK546242) was propagated in LLC-PK1 cells. All viruses were treated with trypsin (5 μg/mL, Sigma, USA) at 37 °C for 72 h. After incubation, the viral supernatant was harvested and centrifuged at 12,000 × g for 30 min at 4 °C to remove cell debris. The viruses were then concentrated by ultracentrifugation at 100,000 × g for 2 h at 4 °C. The viruses were further purified using differential centrifugation and 10% - 60% (wt/vol) sucrose gradient ultracentrifugation, and stored at -80 °C. The centrifuge assay was performed using a Beckman SW32Ti rotor (Beckman Coulter, USA). For single-virus tracking, the lipid membranes of viruses were labeled using the lipophilic dye DiD (Thermo Fisher, USA). The purified viruses were incubated with 20 μM DiD dye at room temperature for 2 h with gentle vortex. Excess dye was removed using a NAP-10 filtration column (GE Healthcare, USA). The DiD-labeled viruses were then stored at -80 °C for further use in single-virus tracking experiments.

Plasmid construction and transfection

mCherry-pAPN and GFP-pAPN were constructed by cloning pAPN from IPECs into the pmCherry-C1 and pEGFP-C1 plasmids. GFP-Cla, mCherry-Cla, GFP-Cav1 and mCherry-Cav1 were constructed in our previous study (23). For plasmid transfection, cells were grown in 35-mm petri dishes (Nest Biotechnology, China) until approximately 80% confluence. A transfection mix was prepared by diluting 2 μL of Lipofectamine 3000 reagent (Thermo Fisher, USA) and 1 μg of plasmid DNA in DMEM (Vivacell, China) to a final volume of 100 μL. The mixture was incubated for 20 min at room temperature before being added to the cells. The transfected cells were then incubated at 37 °C with 5% CO₂ for 24 h.

Immunofluorescence assay (IFA)

IPECs and IPI-2I-APN^-/- cells were infected with viruses at an MOI of 0.1 and incubated at 37 °C for 24 h. After infection, the cells were fixed with 4% paraformaldehyde (Beyotime, China) for 15 min, followed by permeabilization with 0.2% Triton X-100 for 10 min. The cells were then blocked with QuickBlock blocking buffer (Beyotime, China) at room temperature for 15 min. For the detection of viral proteins, the cells were incubated overnight at 4 °C with primary antibodies, including a mouse anti-SADS-CoV N IgG (1:500 dilution, Blue City Biology, China), a mouse anti-PEDV N IgG (1:500 dilution, prepared in our laboratory) or a mouse anti-TGEV N IgG (1:1000 dilution, Blue City Biology, China). The cells were then incubated with a FITC-conjugated goat anti-mouse IgG secondary antibody (1:500 dilution; Biodraon, China) at room temperature for 1 h. Finally, cellular nuclei were stained with Hoechst 33342 (Beyotime, China) and the cells were imaged using a Nikon A1 confocal microscopy (Nikon, Japan).

Attachment and internalization assays

For attachment assay, IPECs grown to 70% - 90% confluence in 12-well cell culture plates were infected with PEDV, TGEV, or SADS-CoV at an MOI of 10. The cells were incubated with the viruses at 4 °C for 30 min to allow for attachment. After incubation, unbound virions were removed by washing the cells three times with phosphate-buffered saline (PBS). The cells were then lysed using VeZol Reagent (Vazyme Biotech, China) to extract RNA, and the viral RNA levels were quantified by RT-qPCR. For internalization assay, cells were washed three times after attachment, and then cells were incubated in pre-warmed DMEM (Vivacell, China) for 60 min at 37 °C to allow internalization. After incubation, residual plasma membrane-bound virions were removed by treating the cells with proteinase K (0.5 μg/mL) for 120 min at 4 °C. The cells were then washed three more times with PBS, lysed in VeZol Reagent (Vazyme Biotech, China), and the internalized viral RNA levels were also quantified by RT-qPCR.

Reverse Transcription Quantitative real-time PCR (RT-qPCR)

Total RNA was extracted from cells using the VeZol Reagent (Vazyme Biotech, China) according to the manufacturer’s instruction. First-strand synthesis was performed using a HiScript II 1st Strand cDNA Synthesis Kit (Vazyme Biotech, China). RT-qPCR was conducted on a LightCycler 96 system (Roche LifeScience, Switzerland) using the AceQ qPCR SYBR Green Master Mix (Vazyme Biotech, China). Positive and negative controls were included in each experiment, and all results were normalized to the expression level of GAPDH. Quantification of data was performed using the 2^-ΔΔCT method. The primers used for RT-qPCR were as follows: TGEV N gene primers: 5’-CATGGTGAAGGGCCAACGTA-3’ and 5’-GGCAACCCAGACAACTCCAT-3’, SADS-CoV N gene primers: 5’-CTAAAACTAGCCCCACAGGTC-3’ and 5’-TGATTGCGAGAACGAGACTG-3’, PEDV N gene primers: 5’-GCACTTATTGGCAGGCTTTGT-3’ and 5’-CCATTGAGAAAAGAAAGTGTCGTAG-3’, and GAPDH gene primers: 5’-CCTTCCGTGTCCCTACTGCCAAC-3’ and 5’-GACGCCTGCTTCACCACCTT

CT-3’.

Western blotting

Cells were washed three times with PBS (pH 7.2) and lysed in RIPA lysis buffer (Vazyme Biotech, China) containing 1 mM protease and phosphatase inhibitors (Beyotime, China). After centrifugation at 13,000 × g for 10 min at 4 °C, the supernatants were collected and the protein content was normalized using a bicinchoninic acid (BCA) protein assay kit (Beyotime, China). Equal amounts of protein were separated by 10% Bis-Tris sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and transferred onto 0.2-μm nitrocellulose blotting membranes (GE Healthcare, USA). The membranes were blocked with 5% nonfat milk for 1 h at room temperature to prevent non-specific binding. After blocking, the membranes were incubated with primary antibodies specific for each target protein overnight at 4 °C. The next day, the membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies for 1 h at room temperature. The protein bands were then visualized using enhanced chemiluminescence (ECL) buffer (Vazyme Biotech, China) and quantified using Fiji/ImageJ software.

Single-virus tracking

Cells were initially seeded onto 35-mm petri dishes (Nest Biotechnology, China), and transfected with plasmids for 24 h at 37 °C. The cells were then incubated with DiD-labeled virions (MOI = 100) at 4 °C for 30 min to facilitate virus adsorption. After adsorption, the cells were washed three times with PBS to remove unbound virions. To prepare for live-cell imaging, the culture medium was then replaced with phenol red-free DMEM (Procell Life, China) containing 2% ProLong live antifade reagent (Thermo Fisher, USA), which includes naturally occurring E. coli plasma membrane enzymes. For single-virus tracking, the cells were maintained in a stage-top incubator (Tokai Hit, Japan) at 37 °C with 5% CO₂. Imaging was performed using a commercial Nikon A1 plus confocal microscopy (Nikon, Japan) equipped with an Apo 60 × /1.40 numerical aperture (NA) oil immersion objective lens (Nikon, Japan). The fluorescence signals were captured using an EMCCD camera (Andor, UK) and excited with lasers at wavelengths of 488 nm (Melles Griot, USA), 561 nm (Coherent, USA), and 640 nm (Coherent, USA). Emission filters used included 447/60 nm, 527/55 nm, 615/70 nm, and 707/90 nm (Coherent, USA). Three-dimensional (3D) fluorescence images were obtained using Z-stacks with a step size of 1 μm. More than 200 viral particles were tracked per group across three independent experiments. From these, representative internalization events were selected for quantitative analysis. These events were clearly identified as pAPN-mediated based on defined trajectories, velocity, colocalization with pAPN, and DiD fluorescence signal intensity.

Phylogenetic analysis and molecular docking

Alignment of multiple sequences and homology analysis of PEDV S, TGEV S and SADS-CoV S were performed using the MegAlign 6.0 (DNASTAR, USA). Amino acid sequences of PEDV, TGEV, and SADS-CoV were visualized based on the online program ESPript algorithm (http://www.ipbs.fr/ESPript) and aligned with the online program CLUSTAL-W (https://www.genome.jp/tools-bin/clustalw). The structure of TGEV S was predicted using I-TASSER (USA), and the structures of PEDV S and SADS-CoV S were downloaded from RCSB PDB database (https://www.rcsb.org/). TGEV S, PEDV S, and SADS-CoV S were docked with pAPN using the online program Cluspro (https://cluspro.org/help.php) and RosettaDock (https://docs.rosettacommons.org). Finally, PyMOL 2.4.0 software (USA) was used for protein dehydration, hydrogenation, and conformational display. The free binding energies obtained from molecular docking were converted to dissociation constants to facilitate comparison with SPR-derived affinities, using the following equation: ΔG = RT ln K_D.

SPR assay

The S1 proteins of PEDV, TGEV, and SADS-CoV, each carrying a C-terminal His tag, were expressed and purified using a HEK293T cell expression system. pAPN fused with a hFc tag was also expressed and purified from HEK293T cells. SPR experiments were performed using the Biacore 8K system at room temperature. pAPN was immobilized on the CM5 sensor chips (GE Healthcare, USA) via amine coupling. Serially diluted PEDV S1, TGEV S1, and SADS-CoV S1 in HEPES buffer (Genscript, China) were followed over the chips. All the data was processed using the Biacore 8K Evaluation Software version 5.0.

Data analysis

Time-lapse images, kymographs, and fluorescence intensity were obtained using NIS-Elements AR 4.51.00 (Nikon, Japan) and deconvoluted with a Gaussian filter to remove background noise and enhance image clarity. Fluorescence intensity data was normalized using the equation: x’ = (x-min)/(max-min). Here, x’ represents the normalized data, x is the raw data, and min and max are the minimum and maximum fluorescence intensity values used in the normalization process. The trajectories and velocities were analyzed using Fiji/ImageJ software. MSD was calculated using MATLAB program (MathWorks, USA), and movement modes were determined by binomially fitting MSD using Origin 2022 (USA). Colocalization analysis was evaluated by Pearson correlation coefficient (PCC). Tricolor colocalization was used to assess the extent to which viral particles simultaneously associate with pAPN and Cla/Cav1 during internalization. Specifically, we calculated two ratios: (1) the proportion of viral particles colocalized with both pAPN and Cla/Cav1 relative to those colocalized with Cla/Cav1 alone, and (2) the proportion of viral particles colocalized with both pAPN and Cla/Cav1 relative to those colocalized with pAPN alone. Colocalization analysis was performed using threshold-based segmentation and object-based quantification in Fiji/ImageJ.

Statistical analysis

Significance was determined using an unpaired Student’s t-test, conducted with Origin 2022 (USA). Data are presented as mean ± standard deviation (SD) and were derived from three independent experiments, unless otherwise indicated. For multiple comparisons, Bonferroni correction was applied, and the reported P values are adjusted accordingly. P-values < 0.05 were considered as statistically significant, with levels set as ns (no significant difference), P ≥ 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.