Identification of claudin-3 as an entry factor for rat hepacivirus

Significance

Hepatitis C virus (HCV) is a major causative agent of chronic liver disease, but vaccine development remains challenging due to the lack of suitable small-animal models. Norway rat hepacivirus 1 (NRHV1), a hepacivirus related to HCV, has emerged as a promising surrogate; however, the absence of robust cell culture systems has hindered its virological characterization. Here, we establish rodent cell lines that are either susceptible or nonsusceptible to NRHV1, and show that NRHV1 utilizes rodent-specific entry factors, including claudin-3 (CLDN3), but not claudin-1. Notably, species-specific residues in CLDN3 determine the host range of NRHV1. These findings elucidate the entry mechanisms of nonhuman hepaciviruses and highlight their potential as surrogate models for studying HCV pathogenesis and immunity.

Abstract

Approximately 58 million people worldwide are believed to be infected with hepatitis C virus (HCV), a major causative agent of chronic liver diseases. Hepacivirus ratti strain rn-1, which was discovered from Rattus norvegicus (Norway rat) and designated Norway rat hepacivirus 1 (NRHV1), shares similar properties with HCV in terms of genetic homology, target cell tropism, pathogenicity, and the immune response. In vivo infection systems for NRHV1 will help overcome the challenges in HCV research for vaccine development. However, the virological characteristics of NRHV1, such as the mechanisms of cell entry, remain largely unexplored, in part owing to a paucity of cell culture systems for NRHV1. Here, we identified the host factors that facilitate NRHV1 entry by profiling the gene expression of two cell lines with different susceptibilities to NRHV1 infection. NRHV1 employs rodent orthologues of HCV entry factors, including scavenger receptor class BI, CD81, and occludin, and utilizes claudin-3 (CLDN3) but not claudin-1. The expression of rat and mouse, but not human, CLDN3 facilitates the entry of NRHV1 into murine cell lines that are nonsusceptible to NRHV1 infection. The host-specific cell entry of CLDN3 is determined by two amino acid residues, Ile⁴⁴ and Trp⁴⁶, in extracellular loop 1. These findings suggest that CLDN3 serves as an entry factor for rat hepacivirus.

Hepatitis C virus (HCV), a major causative agent of severe human liver diseases, is an enveloped positive-sense single-stranded RNA virus that belongs to the genus Hepacivirus of the family Flaviviridae. Approximately 58 million people are chronically infected with HCV, with an estimated 1.5 million new infections occurring each year (1, 2). The development of highly effective direct-acting antivirals (DAAs) has enabled the clearance of HCV in more than 95% of patients, but eliminating HCV worldwide is still challenging because of the high cost of DAAs and the absence of an effective vaccine. The development of the HCV replicon and infectious cell culture systems has substantially facilitated research on the life cycle of HCV and antiviral drugs (3). However, the paucity of versatile small animal models for HCV infection is a major obstacle to vaccine development and elucidation of the pathogenesis (4).

The genus Hepacivirus comprises at least 14 distinct species, each of which has a limited host range, although some species have been reported to be transmitted between different hosts. Nonprimate hepacivirus (NPHV) was initially identified in dogs, but NPHV has since been predominantly detected in horses and can be naturally transmitted to donkeys (5–7). Hepacivirus ratti strain rn-1 was identified in Norway rats (Rattus norvegicus) in New York and designated as Norway rat hepacivirus 1 (NRHV1). This virus has also been shown to be transmissible to mice (Mus musculus) (8–10). NRHV1 and NPHV possess binding sites for liver-specific microRNA-122 (miR-122) in their 5′-untranslated regions (UTRs), similar to those in the HCV 5′UTR. The replication efficiency of both viruses is enhanced through interaction with miR-122, similar to what is observed with HCV. However, unlike HCV, NPHV possesses a large stem-loop I structure at its 5′ end, which has been shown to support replication efficiency even in the absence of miR-122. (8, 11–13). NRHV1 can persistently infect laboratory rat strains and immunocompromised murine lines, but is typically cleared within 4 to 5 wk in immunocompetent mice (8, 10). Transient injection of an antibody to CD4 before infection or continuous injections of an antibody to CD8 before and after infection have been shown to result in persistent NRHV1 infection in immunocompetent mice (8, 14, 15), suggesting that effective CD4 T cell–dependent CD8 T cell stimulation and the development of virus-specific antibodies are needed for NRHV1 clearance, which is similar to the immune response for spontaneous clearance of HCV in humans (16–18). Thus, a rat or murine infection system using NRHV1 is a useful surrogate model for HCV infection.

The entry of HCV into hepatocytes occurs through a multistep process. Initially, HCV attaches to the cell surface through interactions with proteoglycans (19), low-density lipoprotein receptor (20), scavenger receptor BI (SR-BI) (21) and CD81 (22). The virion is then translocated close to tight junctions via epithelial growth factor receptor signaling and internalized into cells, where it cooperates with occludin (OCLN) and claudin-1 (CLDN1) (23–25). In vitro studies have shown that CLDN6 and CLDN9 can support HCV entry instead of CLDN1 (26–28). Several reports have suggested that SR-BI is also essential for NRHV1 entry and that GB virus-B (GBV-B) utilizes CLDN1 for entry (29–31). However, the other requirements in the entry step of hepaciviruses are still unclear. Recently, cell clones stably harboring subgenomic replicon (SGR) RNA of NRHV1 were established from the rat hepatoma cell line McA-RH7777 and the mouse hepatoma cell line Hepa1-6 (11). The cured cells of SGR-harboring McA-RH7777, McA-RH7777.hi, have been shown to permit infection with NRHV1 containing cell-adaptive mutations found in the replicon cell line (31), whereas other cell culture models for NRHV1 infection are still limited. Therefore, the NRHV1 life cycle and host–virus interactions have not yet been fully elucidated.

Here, we show that the cell line supporting the entire life cycle of NRHV1 is derived from the McA-RH7777 cell line but not from the murine hepatoma cell line AML12. Furthermore, our data suggest that NRHV1 primarily utilizes CLDN3, rather than CLDN1, for entry alongside the rodent orthologues of the HCV entry factors SR-BI, CD81, and OCLN. The AML12 cell line permits NRHV1 entry following the exogenous expression of host CLDN3. Furthermore, our findings show that two adjacent amino acid residues that form the molecular surface of the first extracellular loop (EL1) of CLDN3 are essential for NRHV1 entry. These data provide insights into the mechanism of hepacivirus entry and suggest the evolutionary process by which hepaciviruses adapt to their individual hosts.

Results

NRHV1 Replication in Rodent Cells and Antiviral Agents Against NRHV1.

To evaluate the capacity of rodent hepatoma cell lines to support NRHV1 replication, we electroporated multiple cell lines with either a replication-competent NRHV1 subgenomic replicon (SGR) RNA (adFEO-10) or a replication-deficient mutant harboring a GNN substitution in the NS5B coding region (11). Both RNA constructs encoded a fusion gene comprising firefly luciferase and neomycin phosphotransferase II. Transient replication was quantified via a luciferase reporter assay, while colony formation capacity was assessed through G418 selection. NRHV1 SGR exhibited greater luciferase activity than did the GNN mutant in rat McA-RH7777 and murine Hepa1-6 cells (Fig. 1A). NRHV1 SGR also yielded G418-resistant colonies in these cell lines (Fig. 1B). These findings were consistent with the data described by Wolfisberg et al. (11), although the luciferase activity evaluated in the NRHV1 SGR-transfected AML12 cells was similar to that in the NRHV1 SGR GNN-transfected cells (Fig. 1A). Furthermore, AML12 cells electroporated with NRHV1 SGR were eradicated by G418 treatment (Fig. 1B). The efficacy of colony formation is shown in Table 1. These findings suggest that AML12 cells are incapable of supporting NRHV1 replication.

Fig. 1.

Establishment of a rat cell line that produces infectious particles of NRHV1. (A) Transient replication of NRHV1 SGR in McA-RH7777, Hepa1-6, AML12, and AML/122 cells. Luciferase activities were measured after electroporation with NRHV1 SGR (WT) or the nonreplicative replicon (GNN) RNA. (B) Colony formation assay of NRHV1 SGR in rodent cell lines. The cultured cells were electroporated with NRHV1 SGR RNA, followed by selection with G418. (C and D) McA-RH7777-derived cured cell clones were inoculated with the NRHV1ser at 6 copies/cell and were passaged every 3 or 4 d. The intracellular NRHV1 RNA level (C) and the amount of supernatant NRHV1 RNA (D) were quantified via qRT–PCR. The dotted line indicates the limit of detection (LOD). (E) NRHV1ser-inoculated McA1.8 cells (42 dpi), naïve McA1.8 cells, and McA/SGR#1 cells were stained with an anti-NS5A antibody and DAPI.

Table 1.

Colony formation unit (CFU) of cells transfected with NRHV1 SGR RNA

Cell	Host	CFU/µg RNA
McA-RH7777	Rat	0.2 × 103
Hepa1-6	Mouse	1.4 × 103
AML12	Mouse	0 (< 10)
AML12/122	Mouse	5 × 104

Liver-specific miR-122 is a well-known host factor responsible for HCV replication, since exogenous expression of miR-122 can allow HCV replication in some cell lines that do not allow HCV replication (32, 33). Given that NRHV1 replication is also dependent on miR-122 (11), we transduced AML12 cells with a lentiviral vector expressing miR-122 (AML/122). The expression level of miR-122 in the AML/122 cells was comparable to that of the mouse liver, whereas miR-122 was undetectable in the parental cells (SI Appendix, Fig. S1A). Transfection of NRHV1 SGR into AML/122 cells induced luciferase activity and yielded numerous G418-resistant colonies (Fig. 1 A and B), suggesting that the nonpermissiveness of AML12 is due to a lack of miR-122. We isolated two NRHV1 SGR-harboring cell clones from each of the McA-RH7777, Hepa1-6, and AML/122 cell lines (SI Appendix, Fig. S1 B and C).

We next treated McA-RH7777 cells harboring NRHV1 SGR (McA/SGR) with several anti-HCV drugs and assessed the inhibitory effects on NRHV1 replication by a luciferase assay (SI Appendix, Fig. S1B). The second-generation HCV NS5A inhibitor pibrentasvir (PIB) inhibited replication in a dose-dependent manner (50% inhibitory concentration [IC₅₀] = 1.7 μM), whereas the NS5B inhibitor sofosbuvir (SOF) was not effective at a concentration of 10 μM, which was consistent with previous observations (11). In this study, we found that the first-generation NS5A inhibitor DCV also inhibited replication (IC₅₀ > 10 μM). To assess the antiviral efficacy of combination therapy, cells harboring SGR were treated with a 1:1 molar mixture of daclatasvir (DCV) and PIB across a range of concentrations. The combination regimen exhibited enhanced inhibitory activity relative to monotherapy with either agent. Notably, treatment with the DCV/PIB mixture at final concentrations of 1.8 μM and 3.2 μM resulted in approximately 50 and 90% inhibition of replicon activity, respectively (SI Appendix, Fig. S1B). Based on these findings, the combination treatment was employed to generate cured cell lines. Following a 2-wk exposure to 4 µM DCV/PIB mixture (2 μM DCV and 2 μM PIB), multiple surviving clones were isolated from each SGR-harboring cell line. The absence of NRHV1 SGR in these clones was confirmed by both luciferase reporter assay and RT–qPCR analysis (SI Appendix, Fig. S1 C and D). Furthermore, complete susceptibility to G418 selection validated the cured status of these clones.

Establishment of a Rat Cell Line That Produces Infectious Particles of NRHV1.

Given that the replication efficacy of AML/122 cells was greater than that of murine Hepa1-6 cells (Fig. 1 A and B), we investigated whether the cured cell clones derived from McA-RH7777- or AML/122-based replicon cells would permit NRHV1 infection. The cell clones were inoculated with the sera of the mice infected with NRHV1 (NRHV1ser) and then passaged every 3 or 4 d. The cells and supernatants were collected during serial passaging, and NRHV1 RNA was measured via qRT–PCR. The level of intracellular viral RNA gradually increased at 31 d postinoculation (dpi) in the McA1.8 cell clone but not in the McA1.5, McA4.2, or McA4.4 cell clones (Fig. 1C). Accordingly, the amount of NRHV1 RNA in the supernatant increased in a time-dependent manner in the McA1.8 cells (Fig. 1D). NS5A-positive foci were found in NRHV1ser-inoculated McA1.8 cells but not in naïve cells (Fig. 1E). The level of intracellular NRHV1 RNA was reduced by treatment with PIB, suggesting active replication of the genomic viral RNA in McA1.8 cells after entry (SI Appendix, Fig. S2A). We next determined whether cell culture-derived infectious NRHV1 (NRHV1cc) was released from the cells. The culture supernatant of McA1.8 cells infected with NRHV1ser at 35 dpi was inoculated into naïve McA1.8 cells. NS5A-positive cells were subsequently observed after inoculation (SI Appendix, Fig. S2B), whereas NS5A and core were detected in the inoculated cells by Western blotting (SI Appendix, Fig. S2C). Furthermore, the amount of intracellular and supernatant NRHV1 RNA logarithmically increased in inoculated McA1.8 cells in a time-dependent manner (SI Appendix, Fig. S2 D and E). These data suggest that the NRHV1 virus is included in the culture supernatant of the inoculated cells. Serial passages of NRHV1ser in McA1.8 cells improved viral fitness in vitro. We determined the differences in the nucleotide sequences of NRHV1ser and NRHV1cc and listed the nonsynonymous mutations observed in the NRHV genome in SI Appendix, Table S1 (Reference genome: KX905133).

Cured cells, which were prepared from replicon cells harboring HCV SGR following treatment with antivirals, can support HCV replication more effectively than can parental cells (34). McA1.8 and parental cells were electroporated with synthesized NRHV1 SGR. Compared with the parental cells, McA1.8 cells exhibited high luciferase activity at 24 to 48 h posttransfection (SI Appendix, Fig. S2F), while colony formation efficacy was also greater in the McA1.8 cells than in the parental cells (SI Appendix, Fig. S2G). The level of miR-122 was greater in the McA1.8 cells than in the parental cells (SI Appendix, Fig. S1A). These data suggest that the increased permissiveness for NRHV1 replication in McA1.8 cells facilitated viral propagation, mirroring observations in HCV study. In contrast, the level of intracellular NRHV1 RNA did not increase in any of the cured clones derived from AML/122 after 2 mo (SI Appendix, Fig. S2H). As shown in Fig. 1 A and B, AML/122 cells strongly supported NRHV1 replication. However, NRHV1 RNA remained undetectable in the cured cell lines after inoculation with cell culture-adapted NRHV1cc prepared from McA1.8 cells (SI Appendix, Fig. S2I). These data suggest that AML12-derived clones lack sensitivity to NRHV1 entry or disassembly.

Rodent Orthologues of HCV Entry Factors Required for NRHV1 Entry.

SR-BI is required for NRHV1 infection (31), but other host factors involved in NRHV1 entry have not yet been identified. Given that CD81, OCLN, CLDN1, and SR-BI are well characterized as the main entry factors for HCV infection (23, 25), we investigated the roles of these rodent orthologues in NRHV1 entry. We initially evaluated the expression levels of rodent orthologues in NRHV1-permissible (McA1.8) and nonpermissible cell lines (Hepa1-6 (11) and AML/122). Total RNA sample, which was extracted from murine liver, was used as a positive control because mouse hepatocytes permit NRHV1 infection (8). The expression levels of SR-BI and CD81 were comparable among all the rodent cells used in the present study, whereas OCLN was expressed in McA1.8 cells but not in the murine cell lines (Fig. 2A). CLDN1 was detected in mouse livers but in none of the cell lines (Fig. 2A), suggesting that CLDN1 is not necessary for NRHV1 entry. Next, we examined the role of SR-BI, CD81, and OCLN in NRHV1cc infection. SR-BI, CD81, or OCLN mRNA levels were significantly decreased at 24 h after transfection with the respective siRNAs (Fig. 2B). Compared with those in mock cells, the intracellular NRHV1 RNA levels were significantly lower at 3 dpi when NRHV1cc was inoculated into the cells 24 h after siRNA transfection (Fig. 2C). Transfection with si-CD81#2, which is siRNA targeting CD81 mRNA, did not affect NRHV1 RNA levels because of the low knockdown efficacy (Fig. 2 B and C). SR-BI knockout reportedly impairs NRHV1 infection both in vivo and in vitro (30, 31), although whether CD81 knockout or OCLN knockout affects NRHV1 infection is not known. We next established CD81- or OCLN-knockout cell lines from McA1.8 cells (SI Appendix, Fig. S3 A and B) to test the requirement of both proteins for NRHV1 infection. NRHV1 propagation was impaired, but not completely inhibited, in CD81-knockout cell clones (Fig. 2D). In contrast, OCLN-knockout cell lines were nonsusceptible to NRHV1cc (Fig. 2E). When CD81- and OCLN-knockdown clones were transduced with retroviral vectors encoding CD81 and OCLN, respectively, their susceptibility to NRHV1cc infection was restored (Fig. 2 F and G and SI Appendix, Fig. S3 C and D). The level of NRHV1 SGR in McA/SGR cells was not affected by any siRNA transfection (SI Appendix, Fig. S4A), suggesting that SR-BI, CD81, and OCLN are not involved in NRHV1 replication. These findings suggest that CD81, OCLN, and SR-BI, but not CLDN1, are essential host factors for NRHV1 entry.

Fig. 2.

Rodent orthologues of HCV entry factors required for NRHV1 entry. (A) Total RNA extracted from the indicated cells and mouse livers was subjected to qRT–PCR to target the indicated mRNAs. Undetectable values are indicated by number signs (#). (B and C) McA1.8 cells were transfected with 10 µM siRNA 24 h before inoculation with NRHV1cc at 5 copies/cell. The transfected cells were harvested at 0 dpi (B) or 3 dpi (C) and subjected to qRT–PCR. The levels of each mRNA or NRHV1 RNA were normalized to the levels in cells transfected with the control siRNA, si-NC. The asterisks indicate significant differences compared with the control, as estimated by the Tukey–Kramer multiple comparison test (*P < 0.05. **P < 0.01). (D and E) CD81-knockout (KO) or OCLN-KO cell lines were obtained from McA1.8 cells using two different sgRNAs for each gene. Empty vector-transfected cell lines (EVs) were used as controls. These cell lines were inoculated with NRHV1cc at 5 copies/cell and then harvested at 4 hpi or 4 dpi. Total RNA was subjected to qRT–PCR. The asterisks indicate significant differences compared with that in McA1.8 cells at 4 dpi, as estimated by the Tukey–Kramer multiple comparison test (**P < 0.01). (F) CD81KO1.1 and rescued (CD81KO1.1R) cells, along with parental and CD81-transduced McA1.8 cells, were inoculated with NRHV1cc at 5 copies/cell and then cultured for 3 d in the presence or absence of 2 μM PIB. Significant differences between the indicated groups were analyzed using the Bonferroni–Holm method (*P < 0.05. **P < 0.01). (G) OCLNKO1.1 and rescued (OCLNKO1.1R) cells, along with parental and OCLN-transduced McA1.8 cells, were inoculated with NRHV1cc at 5 copies/cell and then cultured for 3 d in the presence or absence of 2 μM PIB. Significant differences between the indicated groups were analyzed using the Bonferroni–Holm method (*P < 0.05. **P < 0.01).

CLDN3 Is Involved in NRHV1 Entry.

CLDN1 is an entry factor for HCV infection of human hepatocytes, while CLDN6 and CLDN9 have been reported to be capable of supporting HCV entry in vitro (26–28). Thus, we hypothesized that CLDN member(s) other than CLDN1 might be involved in NRHV1 entry. We selected CLDN family member expressed in hepatocytes based on single-cell RNA sequence data from mouse hepatocytes (35) and the NCBI database. The expression levels of these genes were surveyed in two rodent cell lines and mouse livers (Fig. 3A). The levels of CLDN2, CLDN6, CLDN12, CLDN15, and CLDN25 were comparable among the three groups. Rodent cell lines lacked the expression of CLDN5 and CLDN26. The levels of CLDN9, CLDN14, and CLDN24 were greater in murine cells and livers than in rat cells. Notably, CLDN3 and CLDN23 were expressed in NRHV1-permissible McA1.8 cells and mouse livers but not in nonpermissible AML/122 cells. McA1.8 cells were transfected with siRNAs targeting CLDNs prior to NRHV1cc infection (Fig. 3B). We silenced CLDN2 as a reference gene because it was expressed at a similar level in both cell lines. The level of intracellular NRHV1 RNA in NRHV1cc-infected McA1.8 cells was specifically reduced by CLDN3 knockdown (Fig. 3C). However, the level of NRHV1 SGR RNA was not affected by CLDN knockdown (SI Appendix, Fig. S4B). The phylogenetic analysis of CLDN family members suggested that CLDN3 is highly related to CLDN6 and CLDN9 and relatively different from CLDN1 (SI Appendix, Fig. S4C). We generated CLDN3-knockout McA1.8 (C3KO) cell lines using CRISPR-Cas9 to examine the involvement of CLDN3 in NRHV1 entry (Fig. 3D and SI Appendix, Fig. S4D). C3KO cells were completely resistant to NRHV1cc infection (Fig. 3E), and expression of sgRNA-resistant CLDN3 mutant rescued NRHV1cc infection (Fig. 3F and SI Appendix, Fig. S4E). Taken together, these data suggest that CLDN3 mediates NRHV1 entry.

Fig. 3.

Identification of CLDN members that facilitate NRHV1 infection. (A) Total RNA extracted from the indicated cell lines and mouse livers was subjected to qRT–PCR. The level of an individual claudin mRNA relative to that of GAPDH (ΔCt) was plotted. Undetectable values are indicated by number signs (#). (B) McA1.8 cells were transfected with 10 µM siRNAs 24 h before inoculation with NRHV1cc at 5 copies/cell. The transfected cells were harvested at 0 dpi (B) or 3 dpi (C) and subjected to qRT–PCR. The levels of an individual mRNA or NRHV1 RNA were normalized to the levels in cells transfected with si-NC. The asterisks indicate significant differences compared with the control, as estimated by the Tukey–Kramer multiple comparison test (*P < 0.05. **P < 0.01). (D) Western blot analysis of lysates extracted from the indicated cells. (E) The indicated cells were inoculated with NRHV1cc at 5 copies/cell and then harvested at 4 hpi or 4 dpi. The levels of viral RNA were evaluated via qRT–PCR. The asterisk indicates significant differences compared with that in McA1.8 cells at 4 hpi, as estimated by the Tukey–Kramer multiple comparison test (**P < 0.01). (F) McA1.8 and C3KO cells were transfected with a plasmid encoding sgRNA-resistant CLDN3 24 h before inoculation with NRHV1cc at 5 copies/cell. These cells were cultured in the presence or absence of 2 µM PIB and harvested at 3 dpi. The levels of viral RNA were evaluated via qRT–PCR. Undetectable values are indicated by number signs (#). Significant differences between the indicated groups were analyzed using the Bonferroni–Holm method (**P < 0.01).

Establishment of NRHV1-Susceptible Murine Cell Lines.

We hypothesized that NRHV1 does not infect AML/122 cells due to low or absent expression of CLDN3. We utilized a mitochondrial antiviral-signaling protein (MAVS)-based reporter system developed for real-time monitoring of HCV infection (36) to detect NRHV1 infection. NRHV1 NS3/4A, but not the protease-inactive mutant harboring the S138A mutation (37), was capable of cleaving mouse MAVS (SI Appendix, Fig. S5A), consistent with findings reported by Anggakusuma et al. (38). Then, we generated McA1.8 and AML/122 cell clones stably expressing the fluorescent reporter protein composed of mCherry, a nuclear localization signal (NLS), and the C-terminal region of mouse MAVS (mCnls-MAVS) (SI Appendix, Fig. S5B). Although mCnls-MAVS is primarily localized on the outer membrane of mitochondria, expression of NS3/4A, but not the S138A mutant, triggered its cleavage, resulting in nuclear translocation of mCherry-NLS (mCnls) (SI Appendix, Fig. S5C). Nuclear translocation of mCnls was observed in McA1.8-derived reporter cells following NRHV1cc infection (SI Appendix, Fig. S5D), suggesting that the fluorescent reporter system distinguishes NRHV1-infected cells from naïve cells. A similar nuclear mCnls signal was detected in AML/122-derived reporter cells transiently expressing NS3/4A (SI Appendix, Fig. S5E). To test whether exogenous expression of rat OCLN (rOCLN) and rat CLDN3 (rC3) can confer susceptibility to NRHV1, AML/122 cells were cotransfected with plasmids encoding both rOCLN and rC3. In NRHV1cc-inoculated AML/122 reporter cells, fluorescence remained cytoplasm, whereas nuclear translocation of mCnls was observed in cells transiently coexpressing rOCLN and rC3 (SI Appendix, Fig. S5F). AML/122 cells transfected with rC3, but not rOCLN vector, exhibited susceptibility to NRHV1 (Fig. 4A). Furthermore, AML/122 cell clones stably expressing rC3 (AML/122/rC3) also exhibited susceptibility to NRHV1cc (Fig. 4 B and C). The nuclear translocation of mCnls was observed in AML/122/rC3 cells, but not in AML/122 cells, following incubation with NRHV1cc (Fig. 4D). Quantification of nuclear mCherry fluorescence intensity using ImageJ enabled the determination of the proportion of infected cells. The results indicated that NRHV1 infectivity in AML/122 cells is strictly dependent on CLDN3 expression (SI Appendix, Fig. S5G). Serial passages of NRHV1cc-infected AML/122 cells demonstrated persistent NRHV1 replication, although NRHV1 RNA was undetectable in the supernatants (SI Appendix, Fig. S6 A and B).

Fig. 4.

NRHV1 susceptibility of AML12-derived cells expressing CLDN3. (A) AML/122 cells were transfected with the indicated expression plasmid 24 h before inoculation with NRHV1cc (5 copies/cell). The transfected cells were harvested at 3 dpi, and the NRHV1 RNA level was evaluated via qRT–PCR. The asterisks indicate significant differences compared with the EV control, as estimated by the Tukey–Kramer multiple comparison test (*P < 0.05. **P < 0.01). (B) Western blot analysis of lysates extracted from the indicated cell clones. The numbers indicate the number of cell clones. (C) These cell clones were incubated with NRHV1cc at 5 copies/cell for 5 h and washed with PBS to remove residual uninfected virus. The intracellular NRHV1 RNA level was quantified via qRT–PCR at the indicated points. (D) AML/122 and AML/122/rC3 cells expressing mCnls-MAVS were infected with mock or NRHV1cc (50 copies/cell) and then fixed and stained with DAPI at 3 dpi. (E) Naïve McA1.8/mCnls-MAVS cells were inoculated with cell extracts at 1 copy/cell. These cells were fixed at 5 dpi and stained with DAPI. (F) An individual cell clone was transfected with a plasmid encoding rOCLN or an empty vector (EV) 24 h before inoculation with NRHV1cc at 5 copies/cell. These cells were harvested at 3 dpi. Total RNA was subjected to qRT–PCR. *P < 0.05, **P < 0.01 by Student’s t test.

We next determined whether infectious NRHV1cc particles were assembled in AML/122/rC3 cells. The cell extracts from NRHV1cc-infected AML/122/rC3 and McA1.8 cells were collected after freeze–thawing cycles (SI Appendix, Fig. S6C) and inoculated into naïve McA1.8 cells according to the methods of Gastaminza et al. (39). The amount of NRHV1 RNA in the cell extracts approximately correlated with the level of intracellular NRHV1 RNA (SI Appendix, Fig. S6D). To normalize RNA input, the extract from the McA1.8 cells was diluted to match the NRHV1 RNA content of the AML/122/rC3 extract. The level of intracellular NRHV1 RNA increased in naïve AML/122/rC3 cells inoculated with the NRHV1-infected McA1.8 cell extract but not in those inoculated with the NRHV1cc-infected AML/122/rC3 cell extract (SI Appendix, Fig. S6E). The nuclear translocation of mCnls was observed in the McA1.8/mCnls-MAVS cells inoculated with NRHV1cc-infected McA1.8 cell extract, but not in those inoculated with the AML/122/rC3 cell extract (Fig. 4E), similar to the data shown in SI Appendix, Fig. S6E. Thus, while the replenishment of CLDN3 increases the susceptibility of AML/122 cells to NRHV1, the process of viral particle assembly is suggested to be disrupted in murine cell lines. In contrast, the replenishment of OCLN promoted NRHV1cc infection in AML/122/rC3 cells but not in AML/122 cells (Fig. 4F), suggesting that there is an unknown factor for the entry function instead of OCLN in AML/122 cells. We evaluated several commercially available antibodies targeting mouse/rat CLDN3 for their capacity to inhibit NRHV1 infection in McA1.8 cells. Despite their specificity for CLDN3, none of the antibodies tested significantly suppressed NRHV1 infection under our experimental conditions (SI Appendix, Fig. S6F). These findings suggest that the antibodies likely fail to recognize the extracellular domains of CLDN3 that are essential for NRHV1 entry.

Species-Specific and Cross-Species Residues of CLDN3 That Are Crucial for NRHV1 Entry.

Since mouse, rat, and hamster CLDN1, but not human CLDN3 (hCLDN3), can support HCV entry (24, 28, 40, 41), we investigated the specificity of CLDN3 as an entry factor for NRHV1. Transient expression of mouse CLDN3 (mCLDN3) and rat CLDN3 (rCLDN3), but not hCLDN3 or mCLDN1, conferred NRHV1 susceptibility to AML/122 cells (Fig. 5 A and B). CLDN3 is composed of four transmembrane regions with two extracellular loops (EL1 and EL2) and a C-terminal tail domain (CTD) (42). Species-specific amino acid residues are distributed throughout the sequence (SI Appendix, Fig. S7A). We constructed three chimeric proteins between hCLDN3 and rCLDN3 (Fig. 5C) to identify the region responsible for entry. NRHV1 susceptibility was restored only when rat EL1 domain was introduced into hCLDN3, but not when EL2 or CTD was substituted (Fig. 5D and SI Appendix, Fig. S7B), suggesting that the EL1 domain is responsible for species-specific receptor function. Given that five amino acid residues differ between the EL1 domains of rCLDN3 and hCLDN3, we individually substituted each residue in hCLDN3 with its rCLDN3 counterpart (SI Appendix, Fig. S7A). Among these, only the N44I mutant of hCLDN3 conferred NRHV1 susceptibility to AML/122 cells (Fig. 5E and SI Appendix, Fig. S7C). In contrast, the I44N mutant of rCLDN3 abolished susceptibility (Fig. 5F and SI Appendix, Fig. S7D). These findings suggest that Ile⁴⁴ in the EL1 domain contributes to the host-specific function of CLDN3 as an NRHV1 entry factor. AlphaFold2 prediction indicated that Ile⁴⁴ constituted the second β-strand and was exposed on the surface (SI Appendix, Fig. S8A). The introduction of the I44N mutation (substitution of Ile⁴⁴ with Asn) was predicted to result in the interaction between Asn⁴⁴ and Phe³⁴, which altered the surface structure of the EL1 domain (SI Appendix, Fig. S8B). Next, the residues proximal to Ile⁴⁴ on the molecular surface were substituted with alanine (SI Appendix, Fig. S8C). The W46A mutant (substitution of Trp⁴⁶ with Ala) did not confer susceptibility to NRHV1cc (Fig. 5G and SI Appendix, Fig. S8D). Trp⁴⁶ is conserved across species and is predicted to be adjacent to Ile⁴⁴ (SI Appendix, Figs. S7A and S8C). To further validate our findings, we quantified the nuclear translocation of mCnls in AML/122 cells transiently expressing the I44N and W46A mutants of rCLDN3 (Fig. 5H). Quantitative image analysis revealed no significant increase in the proportion of cells exhibiting mCnls nuclear translocation upon expression of either mutant (Fig. 5I).These data suggest that a species-dependent microstructure comprising Ile⁴⁴ and Trp⁴⁶ in the EL1 is essential for CLDN3-mediated NRHV1 entry.

Fig. 5.

Identification of CLDN3 sites required to facilitate NRHV1 entry. (A) Western blot analysis of lysates extracted from AML/122 cells transfected with the indicated plasmids. (B) AML/122 cells were transfected with the indicated plasmid 24 h before inoculation with NRHV1cc at 5 copies/cell. Each transfected cell group was harvested at 3 dpi. Total RNA was subjected to qRT–PCR to target the viral RNA. (C) Scheme of human-rat chimeric CLDN3 constructs. (D) AML/122 cells were transfected with the indicated plasmid 24 h before inoculation with NRHV1cc at 5 copies/cell. These cells were harvested at 3 dpi. Total RNA was subjected to qRT–PCR to target the viral RNA. (E–G) AML/122 cells were transfected with plasmids encoding hCLDN3 mutants (E), rCLDN3 mutants (F), or rCLDN3 surface residue mutants (G) 24 h before inoculation with NRHV1cc at 5 copies/cell. These cells were harvested at 3 dpi and then subjected to qRT–PCR to target the viral RNA. (H and I) AML/122 cells stably expressing the mCnls-MAVS reporter were transfected with the indicated plasmids. (H) Transient expression levels of wild-type (WT) rCLDN3 and its mutants were analyzed by Western blotting. (I) The cells were incubated with NRHV1cc (50 copy/cell) or mock-treated, fixed at 3 dpi, and stained with DAPI (Left panels). The proportion of infected cells was determined by calculating the percentage of nuclei exhibiting mCherry fluorescence above the defined threshold among total DAPI-stained nuclei. Statistical significance between groups was evaluated using Student’s t test (n.s., not significant; **P < 0.01).

As EL1 of CLDN1 is required for the internalization of the HCV E2–CD81 complex (23, 43), we examined whether a similar association occurs between NRHV1 E2, CD81, and CLDN3. We generated an expression construct encoding a single open reading frame for NRHV1 E1 and E2, in which Myc-tag was inserted into E1 region and FLAG-tag into E2 region (MF-E1E2). Upon expression, the polyprotein was cleaved to E1 and E2 and the insertion of epitope tags did not interfere with proteolytic procession (SI Appendix, Fig. S8E). Immunofluorescence analysis revealed partial colocalization of CD81 and NRHV1 E2 at the cell surface (Fig. 6A). Similarly, CD81 and CLDN3 also exhibited colocalization (Fig. 6B), suggesting that CLDN3 associates with the NRHV1 E2–CD81 complex during viral entry. To further assess the interaction between E2 and rCLDN3, we performed proximity ligation assays (PLA). A specific PLA signal was detected in cells coexpressing E2 and rCLDN3, indicating a close spatial association between the two proteins (Fig. 6C). Coimmunoprecipitation assays in AML/122 cells expressing MF-E1E2 together with either WT or mutant CLDN3 showed that E2 coprecipitated with WT rCLDN3, but not with a mutant lacking EL1 (Δ1) or with human CLDN3, indicating that the interaction is species-specific and dependent on EL1 (Fig. 6D).Furthermore, the I44N mutant coprecipitated with E2 at levels comparable to WT rCLDN3, whereas the W46A mutant exhibited reduced coprecipitation efficiency (Fig. 6E). E2 also coprecipitated with stably expressed rCLDN3 (Fig. 6F). These findings suggest that Trp46 is a critical residue for the interaction between NRHV1 E2 and CLDN3, and support a model in which CLDN3 facilitates recruitment of the E2–CD81 complex to promote viral internalization (Fig. 6G).

Fig. 6.

Interaction between NRHV1 envelope protein and rodent entry factors. (A and B) AML/122 cells were cotransfected with plasmids encoding V5-tagged rCD81 and MF-E1E2 (Myc-tagged E1 and FLAG-tagged E2) genes (A) or with plasmids encoding V5-tagged rCD81 and HA-tagged rCLDN3 genes (B). The cells were fixed at 24 h posttransfection and subjected to immunofluorescence analysis. (C) AML/122 cells were cotransfected with the plasmids encoding MF-E1E2 and HA-tagged rCLDN3. The cells were fixed at 24 h posttransfection and analyzed by proximal ligation assay (PLA) using rabbit anti-FLAG and mouse anti-HA antibodies. (D and E) AML/122 cells were cotransfected with MF-E1E2 and either WT or mutant CLDN3 constructs (R, WT rCLDN3; Δ1, EL1-deleted rCLDN3; H, human CLDN3). Lysates were divided and immunoprecipitated with control IgG or anti-FLAG antibody. Lysates and immunoprecipitates were analyzed by Western blotting. (F) AML/122/rC3 cells were transfected with the plasmid encoding MF-E1E2 or empty vector. Lysates were divided into two aliquots and immunoprecipitated with control IgG or anti-FLAG antibody, followed by Western blotting. (G) Schematic overview of NRHV1 permissiveness in different cell lines and the roles of host entry factors. The Left panel illustrates the susceptibility, replication capability, and particle producibility of various cell lines. The Right panel shows host factors required for NRHV1 entry and a structural model of rCLDN3 EL1. Factors labeled X and Y represent unidentified host components possibly involved in NRHV1 entry in McA1.8 and AML/122/rC3 cells, respectively.

Discussion

NRHV1 is known to infect immunocompetent rat strains, leading to chronic hepatitis, although no other hepaciviruses have been reported to infect laboratory rat and murine strains (8, 10). Researchers believe that in vivo and in vitro infection models of NRHV1 will serve as useful surrogate viral models for HCV research. However, the virological characteristics of NRHV1 remain poorly understood. An NRHV1-permissible cell line derived from rat hepatoma McA-RH7777 cells was established recently, whereas murine hepatocyte-derived permissive cell lines have not yet been developed (30, 31). In this study, we also successfully established an NRHV1-permissible cell line, McA1.8, from the rat McA-RH7777 cell line harboring NRHV1 SGR (Fig. 1). An increase in NRHV1 replication was observed in McA1.8 cells, similar to reports regarding the increased efficacy of HCV replication in HCV replicon-cured cells (34), which may be attributed to the increased expression of miR-122 (SI Appendix, Figs. S1 A and S2 F and G). Infection of McA1.8 cells with NRHV1ser, which was collected from immunodeficient mice persistently infected with WT NRHV1, resulted in the production of supernatant NRHV1 RNA during approximately 1 mo of serial passages (Fig. 1 C and D). The infectious viral particles were included in the supernatant, since viral passage was possible (SI Appendix, Fig. S2 B and C). The resulting viral particle was cell culture adaptive and designated NRHV1cc in this study. The viral genome of NRHV1cc contains multiple nonsynonymous mutations (SI Appendix, Table S1). Accumulated mutations in the nonstructural gene region may enhance viral replication in vitro, although further studies are needed to confirm this phenomenon. The NRHV1cc strain prepared from the McA-RH7777.hi cell line included several adaptive mutations in NS3 and NS5A, such as S1757A/T2373A and K2130Q/E2223K/D2374G, that promoted the propagation of NRHV1cc (31), while these mutations were not observed in our NRHV1cc strain. In addition, the L586S mutation in E2 was common to both NRHV1cc strains, which may contribute to increasing viral infectivity in cultured cells.

Since murine Hepa1-6 cells support NRHV1 replication (Fig. 1A) but do not permit NRHV1 propagation (11), AML12 cells were used in this study to establish a murine cell-derived culture model for NRHV1. Our data suggest that AML12 cells do not support NRHV1 replication and entry due to deficient expression of miR-122 and CLDN3, respectively (Figs. 1 and 4). Furthermore, AML/122/rC3 cells supported NRHV1cc entry and replication but not the formation of infectious particles (Fig. 4), suggesting that the AML12 cell line lacks some host factors necessary for infectious particle assembly. In contrast, previous reports have demonstrated that murine hepatoma cell lines such as AML12, MMH1-1, Hepa1-6, and Hep56.1D, but not rat McA-RH7777 cells, support low replication of HCV SGR without miR-122 replenishment (44, 45). The mechanism by which miR-122 facilitates hepacivirus replication has not been fully elucidated (46), and the dependency on miR-122 for viral replication seems to vary among hepacivirus species, as illustrated by NPHV, which achieves replication by using the stem loop I of the 5′UTR instead of miR-122 (12, 13). In the case of HCV, viral replication under miR-122 depletion or inactivation conditions induces the generation of a variant with a G28A mutation, which supports miR-122-independent replication, although miR-122 is required for sufficient HCV replication (46, 47). In addition, previous studies have shown that some HCV genotypes with adenine at position 28, such as genotypes 1b (Con1), 4a (ED43), and 5a (SA13), are resistant to miR-122 depletion in patients (2, 47). Our data suggest that the replication of the NRHV1 rn1 strain is strongly dependent on miR-122, in contrast to the relatively lower dependence observed in HCV and NPHV. Furthermore, murine AML12 and MMH1-1 cells do not secrete infectious JFH-1 particles following full-length JFH-1 RNA transfection (44). HCV assembly and release are associated with the very low-density lipoprotein (VLDL) biosynthesis pathway. AML12 cells expressing apolipoprotein E (apoE) can produce infectious particles in a trans-complementation assay employing AML12-derived SGR cells, suggesting that the lack of production of infectious HCV particles in AML12 cells is due to the loss of apoE (48). However, we cannot deny the possibility that factors other than apolipoprotein are required for infectious HCV particle formation in the AML12 cell line because compared with Huh7 cells, AML12 cells expressing apoE produce extremely low quantities of particles (48). The association of NRHV1 particle formation with VLDL synthesis has not yet been revealed and should be investigated in the future.

Recently, Wolfisberg and colleagues demonstrated that SR-BI is essential for NRHV1 entry (30, 31). The data in this study indicate that rodent CD81, OCLN, and CLDN3, in addition to SR-BI, are utilized in the entry of NRHV1 (Fig. 6G). The data deduced from single-cell RNA sequence analysis of the murine liver suggest that CLDN3, but not CLDN1, is a major member of the CLDN family in mouse hepatocytes (35). NRHV1 susceptibility was lost in McA1.8 cells deficient in OCLN and CLDN3 but was partially lost in CD81-knockout cells (Figs. 2 and 3). In HCV, CD81-mediated viral entry is critically dependent on the integrity of lipid rafts, and disruption of cholesterol-enriched membrane domains markedly impairs infection (49). Moreover, other tetraspanins, such as CD9 and CD151, have been shown to facilitate the entry of various enveloped viruses, including coronaviruses (50). These findings raise the possibility that NRHV1 may engage alternative tetraspanins, in addition to CD81, during the entry process. This mechanism could account for the residual infectivity observed in CD81-deficient cells.

These findings suggest that a host factor(s) compensates for the functions of CD81 under at least in vitro conditions. The expression of CLDN3, but not that of CLDN1, promoted the entry of NRHV1 into AML/122 cells (Fig. 4). Given that CLDN3 possesses structures homologous to those of CLDN1, NRHV1 may enter cells through a multistep process similar to HCV entry (23). Furthermore, the receptor function of CLDN3 corresponded with the host range of NRHV1 (Fig. 5B). In contrast, nonhost CLDN1 can facilitate HCV entry (24), although CLDN3 does not support HCV entry (27, 28). Although the role of CLDN1 in HCV entry is critical, our data indicate that CLDN3 plays a role in determining the host specificity of NRHV1. Determining which member of the CLDN family is predominant may contribute to the adaptive evolution of individual hepacivirus species. Interestingly, the AML/122 cell line, which is deficient in OCLN and CLDN3, allowed NRHV1 entry following CLDN3 replenishment alone and exhibited an increase in NRHV1 susceptibility by OCLN coexpression, but this increase was not notable (Fig. 4 A and F). OCLN expression confers weak HCV susceptibility to AML12 cells (24). AML/122 cells may express an alternative factor or OCLN variant that facilitates NRHV1 entry. Given that human CD81 and OCLN are essential for HCV entry, further research is needed to elucidate the specific entry mechanisms of NRHV1.

The CLDN family comprises more than 20 members and shares a structural topology of four transmembrane and two extracellular loop domains. The entry step of NRHV1 was shown to be mediated by EL1 of CLDN3, similar to CLDN1, in the case of the HCV entry step (Fig. 5 C and D), although GBV-B entry is suggested to be mediated by EL2 of CLDN1 (29). Complex formation of CD81 and CLDN1 through EL1 is indispensable for HCV entry (43). Since CD81 is dispensable for GBV-B entry (29), EL1 of CDLN3 may play an important role in the entry of NRHV1 through its interaction with CD81 or other equivalents. Our data indicated that Ile⁴⁴, a specific residue for rat/mouse CLDN3, and Trp⁴⁶, a conserved residue among rat/mouse and human CLDN3, are essential for NRHV1 entry (Fig. 5 F and G). Trp⁴⁶ of rat and murine CLDN3 is not included in the W29-GLW50-C53-C63 signature sequence conserved among EL1s of CLDN members (SI Appendix, Fig. S7A). The protein structure predicted by AlphaFold2 also underscores the importance of Ile⁴⁴ and Trp⁴⁶ in NRHV1 entry. According to the structure prediction, these residues are exposed on the molecular surface and are adjacent to each other. The amino acid substitution of Ile⁴⁴ with Asn in murine CLDN3 likely alters the microstructure conformation involving Ile⁴⁴ and Trp⁴⁶ by generating an interaction between Asn⁴⁴ and Phe³⁴. Previous studies have identified Ile³² and Glu⁴⁸ within the first extracellular loop (EL1) of human CLDN1 as critical determinants for HCV entry (51). Although these residues are conserved in CLDN3 across species (SI Appendix, Fig. S7A) and were therefore excluded from our mutagenesis analysis, it is plausible that Ile⁴⁴ and Trp⁴⁶ of CLDN3 fulfill a structurally analogous role to Ile³² and Glu⁴⁸ of CLDN1 in facilitating hepacivirus entry. Although the precise roles of CLDNs in the entry of hepaciviruses remain under investigation, previous findings on HCV entry suggest that CLDN1 functions at a postattachment step to facilitate endocytosis of the HCV–CD81 complex (23, 43). In fact, NRHV1 E2 and CD81, as well as CD81 and CLDN3, colocalized on the cell surface (Fig. 6 A and B), and a close spatial association between NRHV1 E2 and CLDN3 was also found (Fig. 6C). These findings suggest that CLDN3 may facilitate the recruitment of the NRHV1–CD81 complex to the tight junction region, similar to the role of CLDN1 in HCV entry. Furthermore, the interaction between NRHV1 E2 and CLDN3 required the EL1 of CLDN3 and specifically involved Trp⁴⁶ (Fig. 6 D–F). Ile⁴⁴ appeared to have little effect on the interaction, while it may be involved in interactions with other host factors. Further analysis is needed to elucidate the involvement of the microstructure containing two adjacent amino acid residues, Ile⁴⁴ and Trp⁴⁶, in the postattachment step.

In conclusion, we established rat McA1.8 cells and mouse AML/122/rCLDN3 cells as cellular models for NRHV1 infection. McA1.8 cells can support the entire life cycle of NRHV1, while AML/122/rCLDN3 cells permit NRHV1 entry and replication but not particle formation. Our data also suggest that NRHV1 requires CLDN3 but not CLDN1 for entry in addition to SRB1, CD81, and OCLN in a partially similar way to HCV entry. CLDN3 may play a role in NRHV1 entry in cooperation with CD81. EL1 of CLDN3 is considered essential for NRHV1 entry. Our data regarding point mutations in CLDN3 suggest that the host-specific microstructure on the surface of CLDN3 is essential for CLDN3-mediated NRHV1 entry. Understanding virus–host interactions will not only advance in the development of cell culture models but also provide insight into the host range, origin, and evolution of hepacivirus.

Materials and Methods

Detailed descriptions of the materials and experiments can be found in SI Appendix, Supplementary Materials and Methods.

Cells and Virus.

Rat hepatoma McA-RH7777, murine AML12, and Hepa1-6 cells were cultured in DMEM supplemented with 10% FBS and standard additives. NRHV1 subgenomic replicon (SGR)-harboring cells were established by electroporation with in vitro-transcribed SGR RNA, followed by G418 selection. To prepare serum-derived NRHV1 (NRHV1ser), sera collected from rats persistently infecting NRHV1 (rn1 strain) were intravenously injected into NOD-SCID mice to collect murine sera. All animal experiments were approved by the Institutional Committee of Laboratory Animal Experimentation (University of Yamanashi) and performed in accordance with the guidelines for the care and use of laboratory animals (University of Yamanashi). Cell culture-adapted NRHV1 (NRHV1cc) was collected from NRHV-permissible cell line. For details, see SI Appendix.

Plasmids.

The plasmid encoding the infectious clone of the NRHV1 rn1 strain, pRHV-rn1 (10), was kindly provided by Dr. Kapoor. The plasmid encoding NRHV1 SGR, RHV-SGR-adFEO-10 (11), was kindly provided by Dr. Scheel. The lentiviral vector encoding miR-122 was prepared previously (12, 47). For the preparation of expression vectors, see SI Appendix.

Gene Knockout, Knockdown, and qRT–PCR.

Gene knockout was mediated by CRISPR-Cas9 system and the oligos of sgRNA are listed in SI Appendix, Table S2. qRT-PCR was carried out using Fast SYBR Green Master Mix (Applied Biosystems, Waltham, MA) and a QuantStudio 3 Real-Time PCR System (Applied Biosystems). The primers are listed in SI Appendix, Table S4. The levels of viral RNAs and target mRNAs relative to GAPDH mRNA were calculated as delta Ct [ΔCt = Ct (target) − Ct (GAPDH)] and normalized to the level of the control group [ΔΔCt = ΔCt (experimental group) − ΔCt (control group)]. The relative levels were determined via the 2 − ΔCt or 2 − ΔΔCt method. The amount of NRHV1 RNA in the supernatant was quantified by developing a standard curve using known copy numbers of in vitro-transcribed NRHV1 RNA. For details, see SI Appendix.

Statistics.

The data used for the statistical analyses are expressed as the means ± SD. The significance of differences between two groups was determined by Student’s t test. Significant differences among multiple groups were estimated by the Tukey–Kramer or Bonferroni–Holm test. A P value of less than 0.05 was considered statistically significant (*P < 0.05; **P < 0.01).

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All study data are included in the article and/or SI Appendix.