LMH ^OATP1A2 (LMH stably expressing OATP1A2) cell line was separately constructed and developed as previously described (Li et al., 2019). HepG2/C3A and HEK 293T cell lines were purchased from American Type Culture Collection. HEK 293T and LMH^OATP1A2 cells were propagated in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Grand Island, NY, United States) supplemented with 10% fetal bovine serum (FBS; Biological Industries, Kibbutz Beit Haemek, Israel), 100 U/mL penicillin (Life Technologies Corp., Grand Island, NY, United States), and 100 μg/mL streptomycin (Life Technologies Corp.). HepG2/C3A cells were cultured in Minimum Essential Medium (MEM; Gibco) supplemented with 10% FBS. All cells were cultured at 37°C with 5% CO₂.

HEV-3 strain Kernow C1/p6 (GenBank accession no. JQ679013) was generously provided by Suzanne U. Emmerson and propagated in HepG2/C3A cells (Shukla et al., 2012). Avian HEV stock was prepared from fecal suspensions obtained from SPF chickens intravenously inoculated with a clinical bile sample containing avian HEV from a 35-week-old breeder chicken in China (CaHEV, GenBank accession no. GU954430).

All reagents were purchased commercially unless otherwise indicated, as follows: anti-HA tag mouse monoclonal antibody (mAb), anti-His tag mAb, anti-tubulin mAb, and horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG antibody (TransGen Biotech, Beijing, China); anti-HEV ORF2 protein mAb (Clone no. 1E6) (Sigma-Aldrich, St. Louis, MO, United States); rabbit anti-MRCKβ polyclonal antibody (pAb) (Invitrogen, Carlsbad, CA, United States); rabbit anti-HA tag pAb, anti-Flag tag pAb, anti-RAC1 pAb, anti-CDC42 pAb, anti-RAC1 pAb, anti-RhoA pAb, anti-PAK1 pAb, anti-Arp3 pAb, anti-MYLK pAb, anti-MYH9 pAb, and anti-MLC 2V pAb (Protein Tech, Rosemont, IL, United States); HRP-conjugated goat anti-rabbit or anti-chicken IgG antibody and fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit IgG antibody (Jackson ImmunoResearch Laboratories, West Grove, PA, United States); ML141 and wiskostatin (Calbiochem, San Diego, CA, United States); ML7 (Abcam Cambridge, MA, United States); NSC23766 (Sigma Aldrich); IPA-3 and Fasudil (Selleck Chemicals, Houston, TX, United States); blebbistatin (TOCRIS Bioscience, Houston, TX, United States). The mAbs against HEV ORF2 protein (Clone nos. 1B5, 1H5, and 3E8) were produced as previously described (Dong et al., 2011; Liu et al., 2014; Li et al., 2018).

CDC42 gene (GenBank accession no. XM_0152968262) was amplified from liver mRNA from an SPF chicken. CDC42 genes were cloned into the pET-28a (+) (Novagen, Hornsby Westfield, Australia) using primers aCDC42-F/R to generate a fusion protein containing a His-tag. Plasmid pCAGEN-HA-CDC42 was constructed using primers aCDC42-F/R with the pCAGEN vector (Addgene, Watertown, MA, United States) to encode a fusion protein with a HA tag. Plasmid pLVX-ZsGreen1-CDC42 was constructed using primers lenti-aCDC42-F/pLVX-aCDC42-R with the pLVX-IRES-ZsGreen1 vector (TaKaRa) encoding a fusion protein with a Flag tag. Recombinant plasmids pCAGEN-ap237, pET-21b(+)-ap237, pET-21b(+)-ker239, pET-21b(+)-sar239, pET-21b(+)-rp239, and pET-21b(+)-sp239 were constructed as previously reported (Chen et al., 2018; Li et al., 2019). All positive plasmids were verified by sequencing conducted by Sango Biotech Co., Ltd. (Shanghai, China). Primers used in the study are listed in Table 1.

Procedures for inducing expression of ap237, sar239, ker239, rp239, and sp239 without any tags and His-tagged CDC42 (CDC42-His) were based on the modified method reported by Chen et al. (2018). Briefly, bacterial expression of proteins was induced by adding 1 mM IPTG followed by incubation for 6 h at 37°C. Then, bacterial cells were harvested and lysed by sonication. Proteins were dissolved in 8 M urea, filtered through a 0.22 μM filter, and then refolded during dialysis against a series of buffers consisting of 6 M urea, 4 M urea, and 2 M urea 0.01 M phosphate buffer (PB) at pH 7.5. Refolded proteins were purified using a Superdex 200 Increase 10/300 GL Column (GE Healthcare, New Jersey, United States) connected to an AKTA Purifier System (GE Healthcare). The recombinant VP2 protein of infectious bursal disease virus (IBDV VP2) served as the negative control and was expressed and purified according to the abovementioned procedures.

Plasmids pCAGEN-ap237 and pCAGEN-HA-CDC42 were co-transfected into cells, and then cells were collected at 48 h post-transfection (hpt). The cell pellet was lysed in NP-40 cell lysis buffer (Beyotime, Beijing, China) containing 1 × protease and phosphatase inhibitor cocktail (Beyotime) for 30 min on ice, and then the lysate was centrifuged at 13,000 × g at 4°C. Out of 600 μL of supernatant, 40 μL was set as the input group (namely, 6.7% of total proteins), and the additional supernatant was trisected and incubated with 1H5 mAb or anti-HA mAb and Dynabeads Protein G (Invitrogen) at 4°C for 6 h. To detect direct interactions among CDC42 and truncated capsid proteins, CDC42-His (10 μg) and each truncated capsid protein (ap237, ker239, sar239, sp239, or rp239, 10 μg/each) were co-incubated in PBS (pH 7.2) at 4°C overnight. Next, each mixture was divided into three equal parts based on volume, and the three portions were incubated at 4°C for 6 h with Dynabeads Protein G coated with either 3E8 mAb, anti-His mAb, or control Mouse IgG. Protein-bound beads were then collected and washed three times with PBS (pH 7.5) containing 0.02% Tween-20. Finally, all samples were analyzed by Western blotting.

To verify direct interactions among CDC42 and HEV capsid proteins, ELISAs were performed. Briefly, 96-well ELISA plates (Nunc Immunoplates, Nunc, Roskilde, Denmark) were coated with ap237 at various concentrations (10, 1, 0.1, or 0.01 μg/well) or CDC42-His (8, 4, 2, or 1 μg/well) and incubated at 4°C overnight. After washing wells with PBS containing 0.5% w/v Tween-20 (PBST) and blocking wells with PBST containing 1% BSA, CDC42-His, or HEV capsid proteins (1 μg/well) were added to plate wells, respectively. And then, plates were incubated for 1 h at 37°C. After washing three times, anti-His mAb (1:5,000) or 3E8 mAb (1:1,000) were added, respectively. Then, plates were incubated and washed again, as mentioned above. Finally, HRP conjugated goat anti-mouse IgG secondary antibody (1:5,000) was added into the wells. After another three washes, tetramethylbenzidine was added, the plates were incubated for 15 min then the colorimetric reaction was stopped by adding 3 M H₂SO₄. IBDV-VP2 protein was set as the negative control of CDC42-His protein because the vector and the tag of the two proteins were the same, while BSA was set as the negative control of the truncated capsid proteins of HEV because these proteins did not generate with the His-tag fusion. The OD_450nm value was read using an automated ELISA plate reader (Bio-Rad, CA, United States). Meanwhile, anti-avian HEV IgG antibodies in sera from challenged SPF chickens were detected using ELISA according to the method of Zhao et al. (2013).

For analyzing the co-localization of ap237 and CDC42, cells were fixed with 4% paraformaldehyde for 20 min at 37°C, permeabilized with 0.25% Triton X-100 for 15 min at 37°C, then blocked with PBS containing 1% BSA for 1 h after co-transfection for 48 h. Next, cells were separately incubated with rabbit anti-HA pAb and 3E8 mAb for 1 h at 37°C and followed by FITC-conjugated goat anti-rabbit IgG (green) and Alexa Fluor^® 555-conjugated goat anti-mouse IgG (red, Invitrogen) for an additional 1 h. Finally, cells were stained with Fluoroshield^TM with DAPI (Sigma-Aldrich) and observed with a Leica SP8 confocal system (Leica, Wetzlar, Germany). Co-localization of ap237 and CDC42 was confirmed after Pearson’s correlation (Rr) and Manders’ overlap coefficient (R) data analyses were conducted using Image J (National Institutes of Health, United States) and Image-pro Plus^® (Media Cybernetics, United States). Both coefficients indicated actual overlaps of fluorescence signals. For visualization of HEV-positive cells, 1E6 mAb was used as the primary antibody. All images were captured and processed using Leica Application Suite X (Version 1.0. Leica Microsystems).

The activation assay in vitro was carried out according to the manufacturer’s instructions of the CDC42 Pull-down Activation Assay Kit (Cytoskeleton, CO., United States). Briefly, 300 μg of CDC42-His was incubated with 50 μg of either ap237 or ker239, and then a 1/100 volume of GTPγS was immediately added to each group. Next, sample mixtures were incubated at room temperature for 15 min with gentle rotation. The reaction was stopped, and the samples were immediately added to PAK-PBD beads (10 μg of beads per sample) to pull down GTP-CDC42. Each sample with the beads was incubated at 4°C for 1 h, and the beads were washed twice and immunoblotted with rabbit anti-CDC42 pAb.

Since CDC42 proteins are generally activated very rapidly from 30 s to 30 min (Petermann et al., 2009; Quetglas et al., 2012), four-time points within 30 min were chosen to detect GTP-CDC42 in host cells during either ap237 (or ker239) incubation or CaHEV (or HEV-3) inoculation. Briefly, the cells were seeded into T25 flasks for 24 h before either ap237 (or ker239) or CaHEV (or HEV-3) were added into each flask. After washing, cells were collected at 0, 10, 20, and 30 min for GTP-CDC42 analysis. Briefly, 800 μg of cell lysis supernatant of each group were incubated with 10 μg of PAK-PBD beads at 4°C for 1 h, and the beads were washed twice and analyzed by Western blotting.

As previously described, 16 liver tissues from CaHEV-positive chickens and three liver tissues from HEV-negative SPF chickens were collected (Liu et al., 2017). Total RNA was extracted using TRIzol Reagent (TaKaRa, Tokyo, Japan). CaHEV and HEV-3 virus copy numbers were quantified as reported previously (Jothikumar et al., 2006; Troxler et al., 2011) using QuantiTect^® Probe RT-PCR kit (QIAGEN, Duesseldorf, Germany) with an Applied Biosystem StepOnePlus^TM Real-Time PCR System (Applied Biosystems, CA, United States). cDNAs were prepared from total RNA isolated from each group using random primers and a cDNA kit (TaKaRa, Tokyo, Japan) to measure mRNA abundance to detect relative numbers of virus particles. All specific primers used in the assay are shown in Table 1.

All siRNAs targeting CDC42, RAC1, RhoA, Arp3, PAK1, MYLK, and MYH9, respectively, and the siRNA-negative control (siNCtrl) were designed and synthesized by GenePharma (Shanghai, China). LMH^OATP1A2 cells were transfected with siRNAs using Lipofectamine^TM RNAiMAX (Invitrogen) for 24 h. Subsequently, the expression levels of the indicated protein in transfected cells were analyzed by qPCR and Western blotting, respectively. It is worth noting that 500 nM of ap237 were pre-incubated with LMH^OATP1A2 cells for Western blotting before transfection with siCDC42, siRAC1, and siRhoA, respectively. The siRNAs mentioned above are listed in Table 2.

Viral infection assays were performed according to the method of Oppliger et al. (2016). Briefly, after cells were cultured for 24 h in 12-well plates, they were transfected with CDC42-overexpressing plasmid, siCDC42-721, siRhoA, siRAC1, siPAK1, siMYLK, siNMHC, and siArp3, respectively, for another 24 h or were treated with different concentrations of the mentioned inhibitors of Rho GTPases family for 4 h. Then, the cells were inoculated with CaHEV (3.5 × 10⁶ copies/wells) or HEV-3 (6 × 10⁶ copies/well). After inoculation, the cells were washed thrice with PBS and were collected at 7 dpi for CaHEV and 5 dpi for HEV-3. Additionally, the negative-strand ORF1 RNA of CaHEV was also detected according to the method of Billam et al. (2008) using primer pairs EF1/ER1 and EF2/ER2 (Table 2) to confirm viral replication in LMH^OATP1A2 cells.

Non-enveloped HEV was obtained using methods reported previously by Nagashima et al. (2014) and Yin et al. (2016). First, each culture supernatant containing HEV virions was concentrated by ultracentrifugation at 100,000 g for 2 h at 4°C. Next, virions were treated with 0.1% w/v sodium deoxycholate and 0.1% w/v trypsin at 37°C for 4 h. Non- and quasi-enveloped virions were generated by equilibrium centrifugation in isopycnic gradient centrifugation at 160,000 g in an SW41i rotor for 16 h at 4°C (Yin et al., 2016). After gradients were fractionated, both virus density and load in each fraction were measured using refractometry and qPCR.

Cell viability was evaluated by the CCK-8 (Beyotime) assay described previously with modifications (Chen et al., 2017). LMH^OATP1A2 cells (5 × 10⁴ cells/well) and other cell lines (1 × 10⁴ cells/well) were seeded into 96-well plates and incubated with different concentrations of specific inhibitors of members of the Rho GTPase family. Next, the CCK-8 reagent (10 μL/well) was added and incubated for 1 h at 37°C. Finally, the OD_450nm value was read and used to calculate cell viability. Data are expressed as the percentage of the optical density of treated cells relative to that of the untreated control cells (controls defined as having 100% viability).

Each experiment was independently repeated at least three times. Data were presented as the mean ± standard deviation (SD). Statistical significance was determined by Student’s t-test or Kolmogorov-Smirnov test when two groups were compared using GraphPad (Graph-Pad Software Inc., San Diego, CA, United States). Pearson’s correlation coefficient (Rr) and overlap coefficient (R) was finished using Image-Pro Plus 6.0 software to assess the relationship between CDC42 and ap237. Moreover, Nair’s test was used to analyze outliers or stragglers of the data in the present study, and the outliers were removed, and the straggler was indicated as a hashtag. Asterisks indicate statistical significance as follows: ns, not significant; ^∗ and a, P < 0.05; ^∗∗ and aa, P < 0.01; ^∗∗∗ and aaa, P < 0.001.

The interaction between ap237 and CDC42 was confirmed by co-immunoprecipitation (Co-IP) assays and confocal microscopy-based observations. The results showed that CDC42-HA and ap237 proteins were specifically pulled down together without being pulled down by normal mouse IgG (MIgG) (Figure 1A). Subsequently, the interaction between endogenous CDC42 and ap237 was also detected in LMH^OATP1A2 cells. The endogenous CDC42 in the LMH^OATP1A2 cells was pulled down by ap237 (Figure 1B). Next, to examine whether CDC42 could directly interact with ap237, ap237, and CDC42 fused with a His-tag (CDC42-His) protein were expressed using the prokaryotic expression system. SDS-PAGE analysis showed that the two proteins were successfully expressed and purified (Supplementary Figure 1A). Then, using the purified two proteins for pull-down experiments and enzyme-linked immunosorbent assays (ELISAs), the results showed that ap237 bound to CDC42-His in solution (Figure 1C) as well as to solid phase-immobilized CDC42-His in a specific, dose-dependent manner (Figures 1D,E). Moreover, using confocal microscopy experiments to assess co-localization of ap237 and CDC42-HA in cells, the results revealed that ap237 (red) was present throughout the cytoplasm and co-localized with CDC42-HA (green) in HEK 293T cells (Supplementary Figure 1B, mean Rr: 0.895 ± 0.161; and R: 0.921 ± 0.252). The similar results were obtained in LMH^OATP1A2 cells (Figure 1F, mean Rr: 0.871 ± 0.102; R: 0.905 ± 0.150). Collectively, the abovementioned results indicated that CDC42 directly interacted with avian HEV capsid protein.

Expression levels of total CDC42 were analyzed after CaHEV infection in vitro and in vivo. It had been previously reported that CDC42 activated signaling pathways after its conversion from an inactive GDP-bound state (GDP-CDC42) to an active GTP-bound state (GTP-CDC42) (Quetglas et al., 2012). So, the relationship between GTP-CDC42 and avian HEV infection was also determined. The results showed that protein levels of GTP-CDC42 and mRNA levels of total CDC42 increased gradually after the LMH^OATP1A2cells were treated with ap237 and by CaHEV inoculation for 10, 20, and 30 min (Figures 2A–D). Avian HEV RNA was detected in the LMH^OATP1A2 cells, indicating that avian HEV successfully infected the cells (Figure 2E). In vivo, total CDC42 levels in livers from CaHEV-inoculated chickens also increased gradually at 2, 3, 4, and 5 day-post-inoculation (dpi) (Figure 2F), while total CDC42 levels in livers from infected chickens significantly exceeded levels observed in uninoculated specific-pathogen-free (SPF) chickens (Figure 2G). Furthermore, to analyze the relationship between GTP-CDC42 levels and interaction of CDC42 with ap237, CDC42 activation assays in vitro were performed. The results showed that more GTP-CDC42 was pulled down when ap237 was added into the mixture of CDC42 and GTPγS, suggesting that the interaction between CDC42 and avian HEV capsid protein promoted CDC42 binding to GTP (Figure 2H). Altogether, these results indicated that total CDC42 and GTP-CDC42 increased after host cells were infected with avian HEV.

To determine the relationship between the expression of CDC42 and avian HEV infection in the cells, CDC42 was overexpressed and knocked down in the LMH^OATP1A2 cells. After the LMH^OATP1A2 cells were transiently transfected with pLVX-ZsGreen-CDC42 plasmids, both Immunofluorescence Assay (IFA) and Western blotting showed that the expression levels of CDC42 increased (Supplementary Figures 2A,B). Next, knockdown assays were conducted using three synthesized small interfering RNAs (siRNAs designated siCDC42-323, siCDC42-578, and siCDC42-721) that target CDC42 mRNAs and a control siRNA (siNCtrl) at a concentration of 10 nM (Table 2). The results demonstrated that the CDC42 mRNA and protein levels in LMH^OATP1A2 cells were effectively knocked down, with the greatest effects observed in the siCDC42-721-transfected group (Supplementary Figures 2C,D). So, siCDC42-721 was used for further study, which working concentrations were determined as 10 and 40 nM (Supplementary Figures 2E,F). After CDC42 overexpression and knockdown assays in LMH^OATP1A2 cells were developed, viral infection assays were conducted. The results showed that avian HEV RNA increased in the cells with overexpression of CDC42 and decreased in the knockdown cells (Figure 3A), indicating that the expression levels of CDC42 were positively correlated with avian HEV infection in host cells. In addition, negative-strand CaHEV RNA was also detected in these treated cells, confirming that CaHEV successfully replicated in these cells (Supplementary Figure 3).

ML141, a known selective inhibitor of CDC42, had previously been shown to suppress CDC42 activation by blocking GTP binding to CDC42 (Figure 3B; Weinberg et al., 2014). The inhibitor was used to treat the LMH^OATP1A2 cells, and then CaHEV infected the treated cells. The maximal concentration of ML141 to treat cells was 10 μM (Supplementary Figure 4A), which was determined using a Cell Counting Kit-8 (CCK-8). Then, viral infection assays showed that the CaHEV RNA significantly decreased in the ML141-treated LMH^OATP1A2 cells with 1 and 10 μM (Figure 3C), indicating that ML141 inhibited CaHEV infection in a dose-dependent manner.

CDC42 is a key molecule within cellular cytoskeleton-associated Rho GTPase signaling pathways, and some viruses can exploit these pathways to participate in the life cycle of viral replication (Farhan and Hsu, 2016). Importantly, molecular pathways downstream of cytoskeleton-associated signaling pathways play diverse and complicated roles during viral infection by engaging in interactions with a high degree of overlap, crosstalk, and dynamic characteristics (Swaine and Dittmar, 2015; Trejo-Cerro et al., 2019). To confirm which pathways were involved in avian HEV infection, six well-characterized inhibitors of the Rho GTPases family were used, including Fasudil for Rho-associated protein kinase (ROCK), NSC23766 for RAC1, IPA-3 for p21-activated kinase 1 (PAK1), ML7 for myosin light chain kinase (MLCK), wiskostatin for neural Wiskott-Aldrich syndrome proteins (N-WASP), and blebbistatin for non-muscle myosin IIA (NMIIA) (Figure 3B). The optimized inhibitors concentrations were first determined to minimize cytotoxicity toward LMH^OATP1A2 cells (Supplementary Figure 4B). After optimal inhibitor concentrations were used to treat the cells, the viral infection assay was performed. The results showed that the amounts of CaHEV RNA in the LMH^OATP1A2 cells treated with inhibitors NSC23766, ML7, and blebbistatin were significantly less than the levels observed for controls and the cells treated with other inhibitors (Figure 4A). Next, the knockdown assays were conducted using six siRNAs that target the mRNAs of RhoA, RAC1, PAK1, MYLK, non-myosin heavy chain (NMHC), and Arp3, respectively (siRNAs designated siRhoA, siRAC1, siPAK1, siMYLK, siNMHC, and siArp3) (Table 2). The results showed that the six proteins in LMH^OATP1A2 cells were effectively knocked down (Supplementary Figure 5). Then, viral infection assays were carried out after the knockdown assay in LMH^OATP1A2 cells was developed. The result showed that avian HEV RNA significantly decreased in the LMH^OATP1A2 cells transfecting with siRAC1, siMYLK, and siNMHC in a dose-dependent manner (Figure 4B). These results indicated that CDC42/RAC1-MRCK-MYLK-NMIIA signaling pathway was involved in avian HEV infection in host cells. It has been reported that the upstream components of the Rho GTPases family activate its downstream components by binding with each other (Manser et al., 1994; Leung et al., 1998; Clayburgh et al., 2004; Vicente-Manzanares et al., 2009). To further confirm the involvement of this signaling pathway in CaHEV infection, the pull-down assays revealed that the members of the CDC42-MRCK-MYLK-NMIIA signaling pathway were pulled down by ap237, while RAC1 was not (Figure 4C). The results further confirmed that the CDC42-MRCK-MYLK-NMIIA signaling pathway was associated with avian HEV infection in the cells.

Sequences alignments showed that CDC42 is highly conserved among diverse species (Supplementary Figure 6). Genotype 1 and 3 human HEVs (Sar55 and Kernow-C1/p6, respectively), genotype 4 swine HEV (CHN-SD-sHEV), and genotype 3 rabbit HEV (CHN-SX-rHEV) were evaluated using assays as described above for avian HEV to determine whether CDC42 also participates in mammalian HEV infection. Four truncated capsid proteins of the abovementioned mammalian HEVs (sar239, ker239, sp239, and rp239) were used to determine the interactions with CDC42. The results of Co-IP and ELISA showed that all four capsid proteins are directly bound to CDC42 (Figure 5). These results suggested that the CDC42 may directly interact with mammalian HEV capsid proteins, as observed for ap237.

Although avian HEV capsid protein shares 48–50% amino acid identity with mammalian HEV, there is a high degree (>90%) of identity among human, swine, and rabbit HEV capsid proteins (Haqshenas et al., 2001). In addition, because only the HEV strain Kernow-C1/p6 (HEV-3) can be efficiently propagated in vitro, ker239 and Kernow-C1/p6 were selected to study mammalian HEV infection in host cells. Firstly, CDC42 activation assays in vitro demonstrated that the interaction between CDC42 and ker239 also facilitated CDC42 binding to more GTPs in vitro (Figure 6A). The protein levels of GTP-CDC42 and mRNA of total CDC42 also increased gradually if the HepG2/C3A cells were pretreated with ker239 for 10, 20, and 30 min (Figures 6B,C). For virus infection assay, non-enveloped HEV-3 virions were first produced by removing the viral envelope of HEV-3 virions from the HEV-positive culture supernatant and identified by iodixanol density gradient centrifugation based on previously mentioned information reported methods (Nagashima et al., 2014; Yin et al., 2016). The density of the viral particles was found to range from 1.177 to 1.261 g/mL, indicating that viral envelopes were removed successfully (Supplementary Figure 7A). After non-enveloped HEV-3 was inoculated, the protein levels of GTP-CDC42 and mRNA of total CDC42 also increased gradually in the HepG2/C3A cells (Figures 6D,E). Meanwhile, HEV-3 RNA was also increased, confirming that HEV-3 successfully infected the HepG2/C3A cells (Figure 6F). These results indicated that HEV infection could promote the up-regulation of CDC42 expression in the HepG2/C3A cells.

To further confirm the relationship between CDC42 expression and HEV-3 infection, overexpression of CDC42 following viral infection was performed. The results showed that the mRNA of CDC42 increased after the HepG2/C3A cells were transfected with the plasmids, indicating that the CDC42 was overexpressed (Figure 7A). The IFA results showed that the red fluorescence (HEV capsid proteins) significantly increased in the overexpressed cells (Figure 7B). In addition, after viral infection, the amounts of HEV RNA in the overexpressed-CDC42 group were more than that of the normal group (Figure 7C). These results indicated that overexpression of CDC42 in the HepG2/C3A cells could promote HEV-3 infection.

Moreover, the ML141 inhibitor was also used to analyze the relationship. The optimized concentrations of ML141 were firstly determined for the HepG2/C3A cells, and the maximal concentration was 20 μM (Supplementary Figure 7C). Then, the IFA and qPCR results of viral infection showed that the HEV protein and RNA were both decreased in a dose-dependent manner after the HepG2/C3A cells were treated with 5, 10, and 20 μM of ML141 (Figures 7D,E). Thus, these results collectively suggested that CDC42 is also potentially associated with the mammalian HEV infection.

There are two forms of natural HEV particles, non-enveloped HEV in bile and feces and quasi-enveloped HEV in blood and cell culture supernatants (Feng et al., 2013; Yin et al., 2016; Nagashima et al., 2017; Himmelsbach et al., 2018; Rivera-Serrano et al., 2019; Oechslin et al., 2020; Tallan and Feng, 2020). The other six inhibitors of the Rho GTPases family (Fasudil, NSC23766, IPA-3, ML7, wiskostatin, and blebbistatin) were used to demonstrate the correlation between the two forms of HEV infection and CDC42 downstream signaling pathways. Firstly, the concentrations of these inhibitors were optimized for treating HepG2/C3A cells (Supplementary Figure 7D). Then, three concentrations of each inhibitor showing no toxicity for the cells were selected. After the pretreated cells were inoculated with non-enveloped HEV, the qPCR results showed that the HEV RNA was decreased in a dose-dependent manner in the NSC23766, ML7, wiskostatin, and blebbistatin treated groups (Figure 8A). The IFA results also showed that the red fluorescence decreased in these groups (Figure 8B). These results indicated that the NSC23766, ML7, wiskostatin, and blebbistatin could inhibit non-enveloped HEV infection. Based on the signaling pathways inhibited by the inhibitors, we think that CDC42/RAC1-MRCK-MYLK-NMIIA and CDC42/RAC1-(N-) WASP-Arp2/3 signaling pathways are involved in non-enveloped HEV infection.

Previously, it was reported that the life cycles of non- and quasi-enveloped HEV infection in the host cells are different. Then, these inhibitors were also used to analyze quasi-enveloped HEV infection assays. Firstly, the quasi-enveloped HEV virions were also produced and confirmed that the density of quasi-enveloped HEV particles was range from 1.070 to 1.139 g/mL (Figure 9B), which was consistent with the records in other literature (Nagashima et al., 2014; Yin et al., 2016). After the cells were treated with three concentrations of each inhibitor and inoculated with the quasi-enveloped HEV, the results of qPCR and IFA showed that the HEV RNA and protein were decreased in a dose-dependent manner in the groups of ML141, NSC23766, IPA-3, wiskostatin, and blebbistatin (Figure 9). These results indicated that these inhibitors inhibited quasi-enveloped HEV infection in HepG2/C3A cells. Similarly, based on the signaling pathways inhibited by these inhibitors, we speculate that the CDC42/RAC1-PAK1-NMIIA/cofilin and CDC42/RAC1-(N-) WASP-Arp2/3 signaling pathways were involved in quasi-enveloped HEV infection.