EMCV Replication in BHK-21 in Vitro is Mediated by Caveolin-1, Actin, Dynamin-Dependent Endocytic Pathway

Background: Encephalomycarditis virus is a member of Cardiovirus, belongs to the family Picornaviridae, and can infect different domestic and wild animals. However, the endocytic pathway by which EMCV infected BHK-21 cells remains unclear. In this study, endocytic pathway used by EMCV replication in BHK-21 cells was elucidated. Methods: The function of numerous cellular key factors implicated in the various endocytic mechanisms known to date were systematic detected using chemical inhibitors. Furthermore, RNA interference (RNAi) silencing, the overexpression of dominant protein combined to virus infectivity assays, and confocal imaging to examine which cellular molecules involved in the infection process were also analyzed. Results: The results indicated that the EMCV replication was related to endocytosis. However, neither clathrin nor macropinocytosis pathway was involved in virus infection. QRT-PCR and WB analyses showed that caveolin-1 were signicantly up-regulated in EMCV infected BHK-21 cells. Immune-uorescent confocal microscopy analysis showed that caveolin-1 was temporally co-localized with EMCV VP1 at the early stage of EMCV infection. Overexpressed caveolin-1 or downregulated caveolin-1 expression inuenced the EMCV infection. Furthermore, EMCV infection was found to depend on dynamin and actin by chemical inhibitors resulted in diminished of virus infection. Conclusions: EMCV replication in BHK-21 cells via caveolin-1, dynamin, and actin-dependent endocytosis pathways.


Background
Encephalomycarditis virus (EMCV) is a member of the genus Cardiovirus in the Picornaviridae family and consists of positive single-stranded RNA genome [1,2]. The virus was rst isolated from a captive male gibbon [3] and now it has been isolated from wide range of animal species in the world [4][5][6][7]. EMCV is a well-recognized model to study the mechanisms of antiviral immunity, virus myocarditis, and insulin dependent diabetes mellitus [8,9]. However, the mechanisms involved in the internalization of cardioviruses are unclear [2,8].
The Picornaviridae family viruses enter by one or more pathway to infect cells. Pietia¨inen et al showed that EV1 entry to CV-1 cells via dynamin-dependent pathway [13]. PV infect different host cells using different endocytic pathway, it enter into Hela cells was clathrin-and caveolin-independent, whereas, it enter in brain microvascular endothelial cells was caveolin-and dynamin-dependent [14][15][16]. CBV3 enters in HeLa cells by clathrin-mediated endocytosis pathway [17] and coxsackie virus A9 enters into A549 cells is mediated by dynamin, β2-microglobulin, and Arf6 [12]. FMDV internalization is mediated by clathrin-dependent pathway [18][19][20]. On basis of the above mention facts, we hypothesized that, as a member of the family Picornaviridae, whether EMCV may also use the endocytic mechanisms to infect BHK-21 cells.
In this study, endocytic pathway used by EMCV replication in BHK-21 cells was elucidated. The function of numerous cellular key factors implicated in the various endocytic mechanisms known to date were systematic detected using chemical inhibitors. Furthermore, RNA interference (RNAi) silencing, the overexpression of dominant protein combined to virus infectivity assays, and confocal imaging to examine which cellular molecules involved in the entry process were also analyzed. At rst, we con rmed that the EMCV replication was related to endocytosis. Subsequently, we demonstrated that neither clathrin nor macropinocytosis pathway was involved in virus infection. We have shown that EMCV replication into the BHK-21 cells via caveolin-mediated and dynamin-dependent pathway. The EMCV strain used in the current study was the BHK-21 cells adapted EMCV (GenBank: X74312) and its titer was 10 6.0 TCID 50 ml -1 . The plaque-forming unit (PFU) was calculated according to previously literatures [21].

Gene overexpression and RNA interference
To evaluate the effect of Caveolin-1 in the infection as well as invasion of EMCV into BHK-21 cells, the replication-defective lentivirus system provided by Dr. Enqi Du (Northwest A&F University, China), was used to construct a recombinant plasmid to overexpress caveolin-1. Total RNA was extracted from BHK-21 cells and reverse transcribed into cDNA. The caveolin-1 gene was ampli ed by PCR based on the murine caveolin-1 sequence in GenBank (accession No. U07645.1). The ampli ed PCR products were digested by restriction endonuclease XbaI and BamH I (NEB, MA, USA) and inserted pTRIP-CMV-IRES-Puro. The recombinant plasmid named pTRIP-CAV1 using the primers listed in Table 1. As a control, EGFP has been constructed by using the methods and named pTRIP-EGFP. Lentivirus was produced with recombinant lentivirus vector pTRIP-CAV1 and pTRIP-EGFP according to the reference [23], the transfection cells were named BHK-CAV1 and BHK-EGFP, respectively.

Gene
Primer sequence 5´-3´ Size (bp) Caveolin-1 F:GCTCTAGAATGTCTGGGGGC AAATACGTGGACTC 537 R:CGGGATCCTCATATCTCTTTC TGCGTGCTGATGC Table 1 Caveolin-1 overexpresses primer sequences Moreover, we designed three individual small interfering RNA (siRNA) against caveolin-1 (Rebobio, Guangdong, China) and cells were transfected. The sequence of the siRNA strands has been showed in Table 2. The transfection of the siRNA was performed with Invitrogen Lipofectamine™ 2000 (ThermoFisher, MA, USA) by following the manufacturer's instructions. The silencing e ciencies were measured by qRT-PCR and WB analysis. After culturing for 2 days, the cells were infected with EMCV. At 9 hours post-infection (hpi), WB and virus infectivity assays were performed.

Chemical inhibitors and Cell viability Determination
DMEM supplemented with 10% FBS and one of the following chemical inhibitors: Nystatin, Pitstop, dynasore, mitmab, cytochalasin D, jasplakinolide and 1,1'-Dithiobis-2-naphthalenol (IPA-3) were purchased from Abcam. Methyl-β-cyclodextrin (MβCD) and NH 4 Cl were purchased from Sigma (Sigma, MO, USA), ba lomycin A1 and wortmannin were purchased from Solarbio (Solarbio, Beijing, China). The effects of inhibitors on cells were con rmed by the cell viability assay (see below). The inhibitors were present for 1 hr before and throughout the EMCV infection assay.
After RNA interference and chemical inhibitors treatment, the cell viability was conducted by the CellTiter 96 ® Aqueous Non-Radioactive Cell Proliferation Assay kit (Promega, WI, USA) according to the protocol.

Virus infectivity assays, Post-entry inhibitory effects and detection of virus internalization
For virus infectivity assays, cells were incubated with EMCV at 0.1 multiplicity of infection (MOI) for 1 hr at 37°C in serum-free medium and then washed three times with pre-warmed phosphate-buffered saline (PBS) and maintained in DMEM with 3% FBS. The virus replication was examined by virus yield titration, qRT-PCR and western blotting.
For post-entry inhibitory effects, cells were incubated with EMCV at 0.1 MOI. 2hrs later, the free viruses were washed by PBS and culture medium containing chemical inhibitors was added in cells. Cells incubated for 9 hpi were harvested for further analysis [24].
The ability of EMCV internalized to BHK-21 cells, BHK-Cav1 and BHK-EGFP were determined by measuring the quantity of infectious virus. For binding assay, cells were infected by 1 MOI EMCV at 37°C. 1hr later, cells were washed ve times and prepared by three cycles of freeze-thaw [25]. The last wash solution was collected and titrated on BHK-21 to determine the quantity of the virus present after washing [26].

IFA and Confocal microscopy
Cells were xed with ice cooling 75% ethanol at 4°C for 30 min. For Co-localization studies, cells were permeabilized with 0.1% Triton X-100 when needed. After washed slides in PBS, the suitable primary antibody was added and incubated at 37°C. 1 hr later, the slides were washed again and 100μl secondary urescen antibody was added to the samples. Then the sample was incubated for 1 hr. Finally, the samples counterstained with 4, 6-diamidino-2-phenylindole (DAPI) for 5 min at room temperature. An Olympas IX73 inverted uorescence microscope (Olympas, Japan) or confocal microscope ZEISS LSM 900 was used to mount and analyze (Zeiss, Oberkochen, Germany).

Western Blotting (WB)
Samples were lysed in NP-40 lysis buffer (Beyotime, Shanghai, China) and concentration was measured using the Pierce BCA Protein Assay Kit (Thermo Fisher Scienti c, A USA), and then heated 95°C 5 min.
Denature sampls run in 10% gel and transfered to PVDF membrane (Millipore, MA, USA). After incubated in 5% milk for 1hr, the membrane was incubated with the primary antibody overnight at 4°C and then treated with HRP-conjugated secondary antibody for 2 hrs at room temperature. The speci c bands were obtained with chemiluminescence (Cowin Bioscience, Beijing, China). Protein ladders were from YEASEN (Yeasen Biotech, Shanghai, China).

Statistical analyses
The results were analyzed by with one-way ANOVA using Graphpad PRISM Version 5.0 software. All data were shown as the means ± standard deviations (SD). Differences were considered statiscally signi cant if P-value was less than 0.05 (*P<0.05; **P<0.01; ***P<0.001).

The role of endocytosis in EMCV replication in BHK-21 cells
To elucidate whether the endocytic pathway correlated with EMCV replication in BHK-21 cells, endocytosis speci c inhibitors were used. Concentration and effects of the inhibitors applied in this study were described in Table 3. The suitable non-toxic concentration of inhibitors was measured by the CellTiter 96 ® Aqueous Non-Radioactive Cell Proliferation Assay kit (data not shown).  Table 3 Chemical inhibitors used in this study Previous research nds that NH 4 Cl can hamper the endosomal entry of viruses through preventing pHdependent activation of the fusion protein and blocking membrane fusion [27]. In this experiment, BHK-21 cells were rstly treated with NH 4 Cl and then incubated with the virus in the presence of NH 4 Cl.
Afterwards; the media was exchanged to remove unbound viruses. Infection process was sustained for 12 hrs and monitored with virus titer assay and qRT-PCR. As shown in Figure 1A and Figure 1B, we found that expression of EMCV-3D and virus titer was signi cantly decreased in EMCV positive cells compared to control cells in dose dependent manner. This indicated that EMCV infection is sensitive to inhibition of endosomal acidi cation. Next, Ba lomycin A1, another inhibitor of endocytic compartment acidi cation was tested [28]. The trend of ba lomycin A1 was as same as NH 4 Cl, also limited the EMCV infection.
Additionally, our results showed that 10nM ba lomycin A1 could not block virus replication at 10nM ( Consequently, to test whether EMCV enters into cells via endocytosis, the co-localization of viral particles with early endosome antigen 1 (EEA1) was examined at 9 hpi. As shown in Figure 2, co-localization of VP1 and EEA1 could be observed.

Clathrin-mediated endocytosis is not involved in EMCV replication in BHK-21 cells
As Clathrin-mediated endocytosis is often related to endosomal acidi cation [30] and is a classical pathway for most viruses to enter into host cells [11]. Therefore, we next detected whether EMCV enters

Caveolae is required for EMCV replication in BHK-21 cells.
Next, we investigated whether the caveolae-dependent pathway was involved in EMCV infection.
Caveolae is rich in cholesterol and sphingolipids and can be disrupted by nystatin or MβCD [29]. As shown in Figure 5, cells treated with 12.5 μg/ml and 25 μg/ml nystatin or MβCD before added virus inhibited EMCV proliferation. However, their effect was limited when added after infection (Supplementary Figure 1).

Caveolin-1 expression is positively correlated with the infection of EMCV
Caveolin-1 is a main structural protein of caveolae and correlated with many viruses to internalize into host cells [34]. In order to explore whether EMCV used caveolin-1 during its infection, the expression of caveolin-1 during EMCV infection was investigated. As shown in the results of qRT-PCR, caveolin-1 mRNA expression of cells infected with EMCV increased signi cantly compared to control cells ( Figure 6A). At protein level, we also found that caveolin-1 expression level of infected cells was increased in a timedependent manner, which was consistent with the expression of the EMCV VP1 protein ( Figure 6B).
To further investigate the impact of caveolin-1 expression changes on EMCV infection, caveolin-1 overexpression and siRNA assays were carried out. BHK-Cav1 and BHK-EGFP cells were cultured in medium without puromycin at least for 2 weeks prior to EMCV infection. Then BHK-Cav1 and BHK-EGFP cells were incubated with 0.1 MOI EMCV at 37°C for 1 hr. As shown in Figure 7A, the mRNA expression of EMCV-3D in the BHK-Cav1 cells was signi cantly higher compared to BHK-21 cells as well as BHK-EGFP cells infected with EMCV. WB analyses showed that the expression of VP1 was signi cantly higher in BHK-Cav1 cells compared to control cells ( Figure 7B). Consisted of WB results, BHK-Cav1 induced rapidly increase of virus titer ( Figure 7C).
Next, RNA interference inhibited the caveolin-1 expression in BHK-21 cells (Figure 8A), and correspondingly, the amounts of VP1 ( Figure 8A), titers of progeny virus ( Figure 8B) and virus copies ( Figure 8C) were also decreased. Above mention results veri ed that EMCV replication was directly affected by caveolin-1 siRNA. Additionally, we performed the same experiment using BHK-Cav1 cells. The results indicated that down-regulated expression of the caveolin-1 signi cantly inhibited the virus replication ( Figure 8D, 8E, 8F). To ensure the effect of caveolin-1, we examined the co-localization of virus particles with caveolin-1 by confocal imaging. As shown in Figure 9, after exposed to the EMCV for 120 min, EMCV-VP1 co-localized with caveolin-1 could be observed in infected BHK-21 cells.

Caveolin-1 is essential for EMCV infection by involving in internalization
To investigate the contribution of caveolin-1 to EMCV internalization, the involvement of caveolin-1 in EMCV endocytosis was studied by measuring virus proliferation in cells down-regulated caveolin-1 or BHK-Cav1. Figure 10A and 10B showed that BHK-Cav1 resulted in increasing EMCV internalization e ciency compared to BHK-EGFP or BHK-21 cells (P 0.05, P 0.001). Consistent with the results of caveolin-1 overexpressed, siRNAs which effectively restrain caveolin-1 expression inhibited the EMCV internalization (P<0.01) compared to control ( Figure 10C and 10D).

Dynamin is needed for EMCV replication in BHK-21 cells
Dynamin is a kind of large GTPase that can promote the split of endocytic membranes [12]. It was considered to participate in both clathrin-dependent endocytosis and several other endocytic pathways [35]. Therefore, we hypothesized whether dynamin played a role in EMCV replication. Two inhibitors which inhibited the GTPase activity (dynasore) [36,37] or blocked the lipid-binding (MiTMAB) were selected [38]. As shown in Figure 11, all these two inhibitors affected virus replication when incubated with BHK-21 before infection. However, their effects were limited when added into BHK-21 postinfection (Supplementary Figure 2).

Role of actin in EMCV infection in BHK-21 cells
The results suggested that EMCV enter into BHK-21 cells via caveolin-and dynamin-dependent endocytosis. Next, the role of the cytoskeleton during virus entry was examined by actin disrupting agent (cytochalasin D) and stabilizing compound (jasplakinolide) [39,40]. We found that both actin-stabilizing jasplakinolide and actin-disrupting agent cytochalasin D inhibited EMCV infection in a statistically signi cant manner ( Figure 12). Disrupting actin laments led to a reduction (P<0.05) in viral infection, lament stabilization also affected virus yield. However, the post-entry treatment showed that viral infection was not affected when inhibitors added after virus incubated (Supplementary Figure 3).

Discussion
Endocytosis is a key factor for viruses to infect permissive cells. First, we supposed whether replication of EMCV in BHK-21 cells is related to endocytosis. lysosomotropic agents sensitivity is considered a good evidence of endocytosis [41]; therefore, we pretreated cells with different inhibitors (NH 4 Cl or Ba omycin A1) of endosome acidi cation. Both reagents partly stop the virus infection (Figure 1 and Figure. 2), suggesting that endocytosis took part in EMCV infection.

As both enveloped and nonenveloped viruses, such as HIV [42], adenovirus [43], foot-and-mouth disease
virus [44], MRV [45] and BTV [46,47], enter into host cells through a clathrin-mediated pathway. In this study, the concentration of viral structural proteins and viral titers were not signi cantly decreased by treatment with clathrin speci c inhibitors ( Figure 3). These results suggested that clathrin-mediated endocytosis might be not an essential pathway for EMCV infection.
Next, we supposed whether macropinocytosis-dependent pathway may be involved in EMCV infection. First, cells were pretreated with macropinocytosis speci c inhibitor such as EIPA, IPA-3, wortmannin and then infected with the EMCV virus. Our results showed that these inhibitors did not affect EMCV replication ( Figure 4). These results indicated that macropinocytosis is not involved in EMCV infection.
Except clathrin-dependent endocytosis, lipid raft and caveolae-dependent endocytosis are alternative endocytic pathways [35,48]. It has been reported that MβCD could inhibit the CAV9 infection via lipid microdomains [49], we hypothesised that EMCV infected BHK-21 cells via caveolar/lipid rafts endocytosis. We found that EMCV infection signi cantly decreased in MβCD and nystatin treated cells compared to control cells before incubated with EMCV ( Figure 5), while their effect was limited when added after infection (Supplementary Figure 1). The results indicated that caveolae are essential during EMCV entry BHK-21 cells.
Entry into a permissive cell is an important stage in multiplication cycle of virus [11,48], different viruses exploit different cellular endocytic mechanisms to initiate internalization and infection [59]. We next examined whether caveolin-1 played any role in EMCV internalization. The EMCV internalization was strongly inhibited when caveolin-1 was downregulated ( Figure 10A and Figure. 10B) while EMCV internalization was increased by overexpression caveolin-1 ( Figure. 10C and Figure. 10D). These ndings, together with the results of co-localization ( Figure. 9) indicated that caveolin-1 is required for the early stage of EMCV infection.
Either overexpressed dominant-negative mutants of dynamin or disrupted actin assembly can block caveolae-mediated endocytosis [29,60]. Therefore, we supposed whether dynamin and actin involved in EMCV infection in BHK-21 cells. Two inhibitors of dynamin (dynasore and MiTMAB) were used. Both inhibitors affected virus replication when added before infection (Figure 11), it might be suggested that dynamin had a role in EMCV replication. However, neither dynasore nor MiTMAB blocked EMCV infection when added after infection (Supplementary Figure 2), this showed that there are some other possible pathways involved in virus entry process.
Lastly, we studied the role of the cytoskeleton during EMCV infection using Cytochalasin D and jasplakinolide. Cytochalasin D inhibits actin subunits the polymerization, whereas jasplakinolide inhibits the polymerization to stabilize laments [39, 61,62]. We discovered that virus replication was decreased in cells pretreated with cytochalasin D and jasplakinolide ( Figure 12) but their effect was limited when added after infection (Supplementary Figure 2). Our results suggested that both actin laments and actin reorganization are required for EMCV infection/entry.

Conclusion
In conclusion, our study disclosed for the rst time that EMCV replication in BHK-21 cells via Caveolin-1, dynamin, and actin-dependent endocytosis pathways. Both confocal image and biochemical assays endorsed it. Comprehensive understanding of the mechanism of interaction between the EMCV and host cell is important to understand the viral pathogenesis. Further Functional studies investigating mechanisms of EMCV binding and entry into host cell response to EMCV infection are needed. Declaration performed some of the experiments. Ying Ling and Rongxiu Liu prepared the samples. Data analysis was performed by Xiangrong Li and Haixia Zhang. The paper was written by Qiongyi Li.

Con icts of Interest
The authors declare that there is no con ict of interest.         EMCV co-localize with caveolin-1. Cells treated with virus were synchronized on ice for 60 min and incubated at 37 °C for 120 min before they were washed and xed. Localization was analyzed by confocal microscopy after performing a double immunofuorescence staining to visualize early caveolin-1 (left) and EMCV VP1 (middle). Right panel shows merged image with virus presented as green and caveolin-1 as red. Cell nuclei were shown in blue. Mock used as negative control. Scale bar, 10 µm. negative-control siRNA (NC) and two caveolin-1 siRNAs were infected with EMCV at 3 MOI, incubated in 37°C for 1 h, after which they were harvested and cell lysates were prepared by three cycles of freezethaw. Analysis by virus titre assay and qRT-PCR showed that the EMCV internalization was decreased when caveolin-1 was down-regulated. Cultures treated with medium were used as mock. Results are from three independent experiments, and data are expressed as mean values ± standard error. Statistical signi cance was analyzed by One-way ANOVA; *P<0.05;**P<0.01; ***P<0.001.

Figure 11
Dynamin inhibitors downregulated EMCV infection. (A) (C) Titers of progeny virus yield in the supernatants of culture at 9 hpi. Analysis by virus titre assay showed that the EMCV replication was decreased at 9 hpi when BHK-21 cells were treated with dynasore or mitmab. (B) (D) The expression of EMCV-3D genomes decreased at 9 hpi when BHK-21 cells were treated with dynasore or mitmab. EMCV-3D transcript level was determined by qRT-PCR. Cultures treated with medium were used as mock. Results are from three independent experiments, and data are expressed as mean values ± standard error.