Characterization of the Quasi-Enveloped Hepatitis E Virus Particles Released by the Cellular Exosomal Pathway

ABSTRACT

Our previous studies demonstrated that membrane-associated hepatitis E virus (HEV) particles—now considered “quasi-enveloped particles”—are present in the multivesicular body with intraluminal vesicles (exosomes) in infected cells and that the release of HEV virions is related to the exosomal pathway. In this study, we characterized exosomes purified from the culture supernatants of HEV-infected PLC/PRF/5 cells. Purified CD63-, CD9-, or CD81-positive exosomes derived from the culture supernatants of HEV-infected cells that had been cultivated in serum-free medium were found to contain HEV RNA and the viral capsid (ORF2) and ORF3 proteins, as determined by reverse transcription-PCR (RT-PCR) and Western blotting, respectively. Furthermore, immunoelectron microscopy, with or without prior detergent and protease treatment, revealed the presence of virus-like particles in the exosome fraction. These particles were 39.6 ± 1.0 nm in diameter and were covered with a lipid membrane. After treatment with detergent and protease, the diameter of these virus-like particles was 26.9 ± 0.9 nm, and the treated particles became accessible with an anti-HEV ORF2 monoclonal antibody (MAb). The HEV particles in the exosome fraction were capable of infecting naive PLC/PRF/5 cells but were not neutralized by an anti-HEV ORF2 MAb which efficiently neutralizes nonenveloped HEV particles in cell culture. These results indicate that the membrane-wrapped HEV particles released by the exosomal pathway are copurified with the exosomes in the exosome fraction and suggest that the capsids of HEV particles are individually covered by lipid membranes resembling those of exosomes, similar to enveloped viruses.

IMPORTANCE Hepatitis E, caused by HEV, is an important infectious disease that is spreading worldwide. HEV infection can cause acute or fulminant hepatitis and can become chronic in immunocompromised hosts, including patients after organ transplantation. The HEV particles present in feces and bile are nonenveloped, while those in circulating blood and culture supernatants are covered with a cellular membrane, similar to enveloped viruses. Furthermore, these membrane-associated and -unassociated HEV particles can be propagated in cultured cells. The significance of our research is that the capsids of HEV particles are individually covered by a lipid membrane that resembles the membrane of exosomes, similar to enveloped viruses, and are released from infected cells via the exosomal pathway. These data will help to elucidate the entry mechanisms and receptors for HEV infection in the future. This is the first report to characterize the detailed morphological features of membrane-associated HEV particles.

INTRODUCTION

Hepatitis E virus (HEV), a member of the genus Orthohepevirus in the family Hepeviridae (1), is the causative agent of acute or fulminant hepatitis E, which occurs in many parts of the world, principally as a waterborne infection in developing countries and a zoonotic infection in industrialized countries (2–7). Recently, chronic hepatitis E has become a significant clinical problem in immunosuppressed individuals, especially in solid organ transplant recipients (4, 8). The HEV genome is a positive-sense, single-stranded RNA composed of approximately 7,200 nucleotides (nt) and is capped and polyadenylated (9, 10). The genome consists of a 5′ untranslated region (UTR), three open reading frames (ORFs; ORF1, ORF2, and ORF3), and a 3′ UTR with a poly(A) tail (11). ORF1 encodes nonstructural proteins, including methyltransferase, papain-like cysteine protease, helicase, and RNA-dependent RNA polymerase (12, 13). ORF2 and ORF3 overlap, and the ORF2 and ORF3 proteins are translated from a bicistronic subgenomic RNA of 2.2 kb (14, 15). The ORF2 protein is the viral capsid protein, while the ORF3 protein is a small protein of only 113 or 114 amino acids that is suggested to act as an adapter to link the intracellular transduction pathways, reduce the host inflammatory response, and protect virus-infected cells (2). In addition, the ORF3 protein was found to play an important role in the egress of virions from infected cells (16–18) and is a functional ion channel (as a viroporin) (19).

Four major genotypes (genotypes 1 to 4) of HEV have been identified in humans. HEV genotypes 1 and 2 have been found only in humans and are associated with epidemics in developing countries, whereas HEV genotypes 3 and 4 are zoonotic and are responsible for sporadic or clustered cases of the disease worldwide (20). Various animal strains of HEV have also been identified in increasing numbers of animal species, including chickens, pigs, wild boars, deer, mongooses, rabbits, rats, ferrets, and bats (21).

Exosomes are lipid bilayer extracellular vesicles (30 to 100 nm) that are secreted from various cell types and act as mediators of intercellular communication by delivering function proteins, mRNA, and microRNA (miRNA) to recipient cells (22, 23). They originate as intraluminal vesicles (ILVs) during the process of multivesicular body (MVB) formation. Recent studies on human immunodeficiency virus, human T-cell lymphotropic virus, and dengue virus have demonstrated that the exosomes that are released from infected cells harbor and deliver many regulatory factors, including viral RNA and proteins, viral and cellular miRNAs, and other host functional genetic elements, to the neighboring cells, which helps in the establishment of productive infections and the modulation of cellular responses (23).

It was recently shown that hepatitis A virus (HAV) is able to escape humoral immunity by cloaking itself in cellular membranes on release from host cells (24). These virus-containing microvesicles—designated “quasi-enveloped” HAV (eHAV)—resemble exosomes and were shown to protect virions from antibody-mediated neutralization. Similar results were reported for hepatitis C virus (HCV). Exosomes isolated from HCV-infected human hepatoma Huh7.5.1 cells were shown to contain HCV particles (25). Furthermore, hepatic exosomes can transmit productive HCV infection in vitro and are partially resistant to antibody neutralization. In addition, Liu et al. (26) demonstrated that HCV occurred in both exosome-free and exosome-associated forms, as reported for HAV. The exosome-associated HCV was infectious and was resistant to neutralization by an anti-HCV neutralizing antibody.

The HEV particles that are present in feces and bile are nonenveloped, while those in circulating blood and culture supernatants have been found to be covered with a cellular membrane, similar to enveloped viruses (18, 27, 28). Recently, these membrane-associated HEV virions were designated “quasi-enveloped” HEV (eHEV), similar to eHAV, because of the absence of peplomers (29). In this report, for simplicity, nonenveloped HEV is abbreviated neHEV. Our previous studies demonstrated that a PSAP motif in the ORF3 protein of HEV is necessary for virions to be released from infected cells (17) and that tumor susceptibility gene 101 (Tsg101) and the enzymatic activities of vacuolar protein sorting protein 4 (Vps4A and Vps4B) are involved in the release of eHEV virions, with a specific interaction between the ORF3 protein and Tsg101, indicating that HEV utilizes the MVB pathway to release eHEV particles (30). Furthermore, it was found that eHEV particles are abundantly present in the lysates of infected cells (31). We recently revealed that eHEV particles are present in MVBs with ILVs and that the egress of HEV depends on the exosomal pathway with secretory ILVs through the MVBs (32). Thus, secretion of eHEV particles is associated with exosomes. However, the similarities between exosomes and eHEV have not yet been characterized.

In this study, we characterized exosomes purified from the culture supernatants of HEV-infected PLC/PRF/5 cells. The results indicate that eHEV particles released by the exosomal pathway are copurified with exosomes in the exosome fraction. Furthermore, transmission electron microscopic (TEM) images suggest that the capsids of HEV particles are individually covered by lipid membranes resembling those of exosomes, similar to enveloped viruses.

RESULTS

Characterization of exosomes isolated from culture supernatants of uninfected or HEV-infected PLC/PRF/5 cells.

To characterize the exosomes in the culture supernatants of genotype 3 HEV (strain JE03-1760F)-infected cells, uninfected or HEV-infected PLC/PRF/5 cells were grown in serum-free medium, and culture supernatants were collected. The exosomes derived from uninfected or HEV-infected cells were pelleted by ultracentrifugation. In a previous study (32), we detected CD81, one of the exosome marker proteins (33), in the pellets of culture supernatants of both HEV-infected and uninfected cells and indicated that the exosomes isolated from HEV-infected cells contained detectable levels of viral ORF2 and ORF3 proteins. In the present study, we isolated CD63-, CD9-, and CD81-positive exosomes from the purified exosomes by immunoprecipitation. Western blotting of the precipitates confirmed the presence of CD63, CD9, and CD81 in both exosome fractions (Fig. 1A, upper panels). In addition, exosomes isolated from HEV-infected cells contained detectable levels of viral ORF2 and ORF3 proteins (Fig. 1A, middle and lower panels). Furthermore, a long-distance reverse transcription-PCR (RT-PCR) showed that the exosome fraction of HEV-infected cells contained nearly full-length HEV genomes (approximately 6.4 kbp, corresponding to the sequence from nt 19 to 6403) (Fig. 1B). As determined by direct sequencing, the nucleotide sequence of the amplified products was identical to that of the HEV inoculum (strain JE03-1760F). Similarly, real-time RT-PCR showed that exosomes purified from the culture supernatant of HEV-infected cells contained high levels of HEV RNA. Ultracentrifugation using the culture supernatant of HEV-infected cells, containing 6.0 × 10⁸ copies of HEV RNA, resulted in the recovery of 4.2 × 10⁸ copies of HEV RNA from the exosome fraction (Fig. 1C). The recovery rate was 69.2% under the same conditions as those used for exosome purification. Understandably, the exosomes separated from the culture supernatant of uninfected cells did not contain the viral proteins and HEV RNA (Fig. 1A, middle and lower panels, and B).

FIG 1.

Exosomes purified from culture supernatants of HEV-infected PLC/PRF/5 cells harbor viral proteins and RNA. (A) Western blotting of CD63-, CD9-, or CD81-positive exosomes purified from the culture supernatant of HEV-infected or uninfected PLC/PRF/5 cells. The cells were incubated in serum-free medium, and the exosomes were purified from culture supernatants by ultracentrifugation and then immunoprecipitated (IP) with anti-CD63, anti-CD9, or anti-CD81 antibody. The precipitated exosomes were subjected to Western blotting with anti-CD63, anti-CD9, or anti-CD81 antibody (upper panels). Viral ORF2 and ORF3 proteins were detected with an anti-ORF2 MAb (middle panels) and an anti-ORF3 MAb (lower panels), respectively. (B) Amplification of HEV RNA from exosomes purified from the culture supernatant of HEV-infected or uninfected cells by long-distance RT-PCR. (C) Quantitative RT-PCR of HEV RNA in exosomes purified from the culture supernatant of HEV-infected cells. The culture supernatant of HEV-infected cells, containing 6.0 × 10⁸ copies of HEV RNA (input), was ultracentrifuged under the conditions used for exosome purification. A total of 4.2 × 10⁸ copies of HEV RNA were recovered in the exosome fraction by ultracentrifugation. The data are presented as means ± SD for two independent experiments.

Morphological analysis by TEM of exosomes isolated from culture supernatants of HEV-infected cells.

Subsequently, to analyze the exosomes morphologically, exosomes purified from culture supernatants of HEV-infected PLC/PRF/5 cells were subjected to TEM. The purified exosomes were observed by negative staining. A large number of exosomes were observed in the exosome fraction purified from the culture supernatant of uninfected cells (Fig. 2A). Similarly, exosomes were visible in the exosome fraction isolated from the culture supernatant of HEV-infected cells (Fig. 2B). In addition, virus-like particles with a uniform diameter of ∼40 nm were observed in the exosome fraction derived only from HEV-infected cells, and these particles were found to maintain their round shape under the conditions of observation (Fig. 2B and C). On the other hand, exosomes with multiple virus-like particles, such as those seen for membrane-associated HAV or HCV (24, 25), were not observed in the exosome fraction purified from culture supernatants of HEV-infected cells. These TEM images suggested that the exosome fraction derived only from HEV-infected cells contained virus-like particles.

FIG 2.

Transmission electron microscopic images of exosomes purified from the culture supernatant of uninfected (A) or HEV-infected (B) PLC/PRF/5 cells, with negative staining. Bars, 100 nm. (C) High-magnification image of virus-like particles found in the exosome fraction purified from the culture supernatant of HEV-infected cells. Bars, 50 nm. The results of one of three experiments are shown.

Characterization of virus-like particles in the exosome fraction.

To characterize the eHEV particles in the exosome fraction, the exosome fraction was treated with digitonin, which effectively water-solubilizes lipids. After treatment with 1.5% digitonin, the exosome fraction was subjected to equilibrium centrifugation in a sucrose density gradient (Fig. 3). The eHEV particles in the exosome fraction exhibited a peak density of 1.11 g/ml, while the digitonin-treated HEV particles banded at 1.19 g/ml in sucrose.

FIG 3.

Sucrose density gradient fractionation of exosomes purified from the culture supernatant of HEV-infected cells with or without prior treatment with 1.5% digitonin.

To analyze whether the digitonin-treated HEV particles in the exosome fraction could be captured by monoclonal antibody (MAb) TA1708, which binds specifically to the membrane on the surfaces of eHEV particles (31), the peak fraction (1.19 g/ml) obtained by sucrose density gradient centrifugation was subjected to an immunocapture RT-PCR assay using anti-ORF2 MAb (H6225) (34), anti-ORF3 MAb (TA0536) (27), or MAb TA1708 (Table 1). The HEV particles without digitonin treatment (1.11 g/ml) were efficiently captured by MAb TA1708 (47.1%) but not by anti-ORF2 and anti-ORF3 MAbs (8.0% and 12.8%, respectively). In contrast, the higher-density (1.19 g/ml) digitonin-treated HEV particles were trapped by both anti-ORF2 and anti-ORF3 MAbs (40.4% and 57.0%, respectively), while the capture efficiency with MAb TA1708 was reduced to 7.6% (Table 1). These results indicate that HEV particles in the exosome fraction are associated with the lipid membrane.

TABLE 1.

Reactivity of MAb TA1708 with HEV particles in the exosome fraction, with or without prior treatment with 1.5% digitonin, as evaluated using immunocapture RT-PCR

Treatment of HEV particles in the exosome fraction	% captured HEV particles in the total HEV per well
	MAb TA1708 (anti-eHEV)	MAb H6225 (anti-ORF2)	MAb TA0536 (anti-ORF3)
No treatment (1.11 g/ml)a	47.1	8.0	12.8
1.5% digitonin (1.19 g/ml)b	7.6	40.4	57.0

^aHEV particles in the exosome fraction purified from the culture supernatant of HEV-infected cells (strain JE03-1760F) that had been cultured in serum-free medium were directly subjected to sucrose density gradient centrifugation, and the peak fraction (1.11 g/ml) was used for immunocapture RT-PCR.

^bHEV particles in the exosome fraction purified from the culture supernatant of HEV-infected cells (strain JE03-1760F) that had been cultured in serum-free medium were mixed with 1.5% digitonin and incubated at room temperature for 15 h. The treated HEV particles were subjected to sucrose density gradient centrifugation, and then the peak fraction (1.19 g/ml) was used for immunocapture RT-PCR.

Comparative analyses of membrane-associated (eHEV) and -unassociated (neHEV) particles by EM.

To clarify the morphological features of HEV particles, we performed immunogold labeling of membrane-unassociated neHEV particles that had been treated with detergent and protease. First, we observed neHEV particles generated from eHEV particles in culture supernatants after treatment with sodium deoxycholate (DOC-Na) and trypsin. A large number of virus-like particles of ∼30 nm in diameter were found in the purified preparations after treatment with detergent and protease (Fig. 4A). To further demonstrate that these particles were not membrane associated, additional observation was performed using immunogold labeling. The virus-like particles were clearly coated with anti-HEV ORF2 MAb (Fig. 4B) but did not bind with anti-HEV ORF3 MAb (Fig. 4C). Similarly, the virus-like particles in the exosome fraction were treated with detergent and protease, and then the treated particles were observed by negative staining. Before treatment, these fractions contained membrane-associated HEV-like (eHEV-like) particles of ∼40 nm in diameter (Fig. 2B and C and 4D, upper panel). In contrast, virus-like particles of ∼30 nm in diameter were observed after treatment with detergent and protease (Fig. 4D, lower panel), while eHEV-like particles were not observed by TEM. The size of the virus-like particles in the exosome fraction treated with detergent and protease was identical to that of neHEV particles generated from eHEV particles in culture supernatants after treatment with detergent and protease. Furthermore, these virus-like particles treated with detergent and protease were coated with anti-HEV ORF2 MAb (Fig. 4E). In addition, the virus-like particles in the exosome fraction were treated with digitonin, and then the treated particles were observed by immunogold labeling with anti-HEV ORF3 MAb. The virus-like particles treated with digitonin were coated with anti-HEV ORF3 MAb (Fig. 4F). The size of the virus-like particles treated with digitonin was identical to that of neHEV particles generated from eHEV particles in culture supernatants after treatment with detergent and protease.

FIG 4.

Transmission electron microscopic images of negatively stained eHEV or neHEV particles. (A) neHEV particles generated from eHEV particles in culture supernatant treated with DOC-Na and trypsin, shown at two different magnifications. (B) Immunogold labeling with a mouse anti-ORF2 MAb (H6225; IgG), shown at two different magnifications. neHEV particles were labeled with the anti-ORF2 MAb and 12-nm colloidal gold-conjugated goat anti-mouse IgG. (C) Immunogold labeling with a mouse anti-ORF3 MAb (TA0536; IgG), shown at two different magnifications. neHEV particles were incubated with the anti-ORF3 MAb and a 12-nm colloidal gold-conjugated goat anti-mouse IgG. (D) Virus-like particles in purified exosomes without (upper panel) or with (lower panel) prior treatment with detergent and protease. (E) Immunogold labeling with anti-ORF2 MAb (H6225), shown at two different magnifications. eHEV particles treated with detergent and protease were labeled with the anti-ORF2 MAb and a 12-nm colloidal gold-conjugated goat anti-mouse IgG. (F) Immunogold labeling with anti-ORF3 MAb (TA0536), shown at two different magnifications. eHEV particles treated with digitonin were labeled with the anti-ORF3 MAb and a 12-nm colloidal gold-conjugated goat anti-mouse IgG. Bars, 50 nm. Results representative of one of three experiments are shown.

To compare the morphological features between eHEV and neHEV particles in detail, the membrane-associated particles in the exosome fraction and membrane-unassociated particles were mixed and then subjected to TEM imaging. Electron microscopic imaging confirmed the presence of HEV particles of two distinct sizes (Fig. 5A). eHEV particles were 39.6 ± 1.0 nm in diameter (n = 200), while neHEV particles were 26.9 ± 0.9 nm in diameter (n = 200). Furthermore, immunogold labeling with anti-ORF2 MAb showed that the neHEV particles were clearly coated with anti-HEV ORF2 MAb (Fig. 5B, left particle). On the other hand, anti-HEV ORF2 and ORF3 MAbs did not bind to the eHEV particles (Fig. 5B, right particle, and C, respectively). In contrast, immunogold labeling with MAb TA1708 against eHEV particles showed that eHEV particles were coated with MAb TA1708 (Fig. 5D, left panel). Similarly, the purified exosomes were clearly coated with MAb TA1708 (Fig. 5D, right panel). Similar results were obtained using the culture supernatants of genotype 4 HEV (strain HE-JF5/15F)-infected PLC/PRF/5 cells. Before treatment with detergent and protease, the exosome fraction contained eHEV particles of ∼40 nm in diameter (Fig. 5E and F, open arrows). Furthermore, neHEV particles of ∼30 nm in diameter were observed after treatment with detergent and protease (Fig. 5F, closed arrows, and G). Taken together, these results provided direct evidence that the exosome fraction contained the eHEV particles. Furthermore, the EM observations indicated that the capsids of the HEV particles were individually covered by the lipid membranes, similar to enveloped viruses.

FIG 5.

Transmission electron microscopic images of negatively stained genotype 3 (A to D) or 4 (E to G) eHEV and neHEV particles. (A) Genotype 3 eHEV particles in the exosome fraction (open arrows) and neHEV particles treated with detergent and protease (closed allows) were mixed and then observed by TEM at two different magnifications. (B) Immunogold labeling with anti-ORF2 MAb (H6225). Genotype 3 neHEV and eHEV particles were mixed and then labeled with the anti-ORF2 MAb and a 12-nm colloidal gold-conjugated goat anti-mouse IgG. (C) Immunogold labeling with anti-ORF3 MAb (TA0536). neHEV particles were not labeled with the anti-ORF3 MAb and a 12-nm colloidal gold-conjugated goat anti-mouse IgG. (D) Immunogold labeling with mouse MAb TA1708 (IgM) against eHEV particles. The purified exosomes separated from the culture supernatants of HEV-infected cells were labeled with MAb TA1708 and a 12-nm colloidal gold-conjugated goat anti-mouse IgM. (E) Genotype 4 eHEV particles in the exosome fraction, shown at two different magnifications. (F) eHEV particles in the exosome fraction (open arrows) and neHEV particles treated with detergent and protease (closed allows) were mixed and then observed by TEM at two different magnifications. (G) neHEV particles generated from genotype 4 eHEV particles in culture supernatant treated with DOC-Na and trypsin, shown at two different magnifications. Bars, 50 nm. Results representative of one of three experiments are shown.

Association between eHEV particles in culture supernatants and the exosome secretion pathway.

The present study revealed that the exosome fraction contained eHEV particles (Fig. 2, 4, and 5). However, with the exosome purification method that was used, the rate of recovery of HEV particles from the culture supernatant of HEV-infected cells was approximately 70% (Fig. 1C). In a previous study, we reported that almost all of the HEV particles that were released into the culture supernatant were associated with the cellular lipid membrane (31). To investigate whether the secretion of eHEV particles into culture supernatants is associated exclusively with exosomes, we carried out immunoprecipitation and real-time RT-PCR by using eHEV particles in the culture supernatant after removal of the exosome fraction. The eHEV particles in the exosome fraction obtained by centrifugation of the culture supernatant of HEV-infected cells were captured by use of anti-CD63, anti-CD9, anti-CD81, and anti-epithelial cell adhesion molecule (anti-EpCAM) MAbs and phosphatidylserine (PS) binding protein (19.9%, 24.3%, 38.8%, 4.9%, and 78.5%, respectively) (Table 2). Similar results were obtained when the eHEV particles in the residual supernatant fraction after removal of the exosome fraction were collected by use of anti-CD63, anti-CD9, anti-CD81, and anti-EpCAM MAbs and PS binding protein (21.6%, 26.1%, 33.5%, 5.9%, and 76.0%, respectively). In addition, nearly 90% of the HEV particles from the exosome fraction and the residual supernatant fraction after removal of the exosome fraction were captured by sequential immunoprecipitations using the four MAbs and PS binding protein (Table 2). In contrast, almost none of the neHEV particles were captured by the four MAbs and PS binding protein or by sequential immunoprecipitations using the four MAbs and PS binding protein (0.6%, 0.8%, 1.6%, 0.1%, 3.7%, and 4.2%, respectively). These findings support the specificity of these immunoprecipitation assays and indicate the presence of eHEV particles with a membrane containing exosome-specific molecules in the residual supernatant after removal of the exosome fraction, suggesting that at least 90% of eHEV particles are secreted as exosomes. In addition, genotype 4 eHEV particles in the exosome fraction obtained by centrifugation of the culture supernatant of infected cells were captured by use of anti-CD63, anti-CD9, anti-CD81, and anti-EpCAM MAbs and PS binding protein (16.1%, 23.5%, 34.6%, 2.4%, and 75.8%, respectively) (Table 3). Similar results were obtained when the eHEV particles in the residual supernatant fraction after removal of the exosome fraction were collected by use of anti-CD63, anti-CD9, anti-CD81, and anti-EpCAM MAbs and PS binding protein (17.3%, 25.4%, 39.9%, 4.6%, and 74.2%, respectively) (Table 3). These results suggest that eHEV particles are released by the cellular exosomal pathway regardless of genotype.

TABLE 2.

Reactivities of anti-CD63, anti-CD9, anti-CD81, and anti-EpCAM MAbs and PS binding protein with genotype 3 eHEV or neHEV particles as evaluated using immunoprecipitation and real-time RT-PCR

Genotype 3 HEV particles	% captured HEV particles in the total HEV per tube
	Anti-CD63 MAb	Anti-CD9 MAb	Anti-CD81 MAb	Anti-EpCAM MAb	PS binding protein	Anti-CD63 MAb, anti-CD9 MAb, anti-CD81 MAb, anti-EpCAM MAb, and PS binding proteind
eHEV particles in exosome fractiona	19.9	24.3	38.8	4.9	78.5	90.2
eHEV particles in residual supernatant after centrifugationb	21.6	26.1	33.5	5.9	76.0	89.1
neHEV particlesc	0.6	0.8	1.6	0.1	3.7	4.2

^aeHEV particles in the exosome fraction obtained by centrifugation from the culture supernatant of genotype 3 HEV-infected cells that had been cultured in serum-free medium were subjected to immunoprecipitation with the indicated MAb or PS binding protein, and the precipitate was used for real-time RT-PCR.

^beHEV particles in the residual supernatant fraction after removal of the exosome fraction were subjected to immunoprecipitation, and the precipitate was used for real-time RT-PCR.

^cneHEV particles generated from the culture supernatant of genotype 3 HEV-infected cells after treatment with 0.1% DOC-Na and 0.1% trypsin at 37°C for 3 h were subjected to immunoprecipitation, and the precipitate was used for real-time RT-PCR.

^dGenotype 3 HEV particles were subjected to sequential immunoprecipitations with anti-CD63 MAb, anti-CD9 MAb, anti-CD81 MAb, anti-EpCAM MAb, and PS binding protein, and the precipitate was used for real-time RT-PCR.

TABLE 3.

Reactivities of anti-CD63, anti-CD9, anti-CD81, and anti-EpCAM MAbs and PS binding protein with genotype 4 eHEV or neHEV particles as evaluated using immunoprecipitation and real-time RT-PCR

Genotype 4 HEV particles	% captured HEV particles in the total HEV per tube
	Anti-CD63 MAb	Anti-CD9 MAb	Anti-CD81 MAb	Anti-EpCAM MAb	PS binding protein
eHEV particles in exosome fractiona	16.1	23.5	34.6	2.4	75.8
eHEV particles in residual supernatant after centrifugationb	17.3	25.4	39.9	4.6	74.2
neHEV particlesc	1.6	0.9	1.2	0.1	3.3

^aeHEV particles in the exosome fraction obtained by centrifugation from the culture supernatant of genotype 4 HEV-infected cells that had been cultured in serum-free medium were subjected to immunoprecipitation with the indicated MAb or PS binding protein, and the precipitate was used for real-time RT-PCR.

^beHEV particles in the residual supernatant fraction after removal of the exosome fraction were subjected to immunoprecipitation, and the precipitate was used for real-time RT-PCR.

^cneHEV particles generated from the culture supernatant of genotype 4 HEV-infected cells after treatment with 0.1% DOC-Na and 0.1% trypsin at 37°C for 3 h were subjected to immunoprecipitation, and the precipitate was used for real-time RT-PCR.

Infectivity of eHEV particles in the exosome fraction of cultured cells.

To investigate whether eHEV particles in the exosome fraction are infectious, purified exosomes were inoculated into naive PLC/PRF/5 cells. First, the HEV RNA titer of the collected culture supernatant was monitored by real-time RT-PCR. The HEV RNA level in the culture supernatant of cells infected with exosomes purified from HEV-infected cells increased from 2 days postinoculation, reaching 3.7 × 10⁸ copies/ml at 20 days (Fig. 6A).

FIG 6.

Inoculation of exosomes purified from culture supernatants of HEV-infected PLC/PRF/5 cells by ultracentrifugation. (A) Quantitation of HEV RNA in culture supernatants of PLC/PRF/5 cells that were inoculated with purified exosomes or in culture supernatants from primary infection for cells cultured for up to 20 days. The data presented are means ± SD for three wells each. (B) Immunofluorescence staining of PLC/PRF/5 cells inoculated with exosomes purified from culture supernatants of HEV-infected cells. At 10 days postinoculation, the cells were incubated with an anti-ORF2 MAb (H6225) (upper panels) or anti-ORF3 MAb (TA0536) (lower panels) and then stained with Alexa Fluor 488-conjugated anti-mouse IgG. The nuclei were stained with DAPI. (C) Expression of viral proteins in culture supernatants of cells inoculated with exosomes purified from HEV-infected or uninfected cells. At 16 days postinoculation, ORF2 or ORF3 protein in the culture supernatant was detected by Western blotting with an anti-ORF2 MAb (H6210) (left) or anti-ORF3 MAb (TA0536) (right), respectively.

Next, we performed an immunofluorescence assay and Western blotting to detect the expression of viral ORF2 and ORF3 proteins. Viral proteins in cells inoculated with exosomes purified from the culture supernatant of HEV-infected cells were detected by an immunofluorescence assay on day 10 (Fig. 6B). Similarly, Western blotting revealed that the culture supernatant of exosome-inoculated cells contained detectable levels of viral ORF2 and ORF3 proteins at 16 days postinoculation (Fig. 6C). Furthermore, the culture supernatants of cells inoculated with HEV-positive exosomes were able to establish a secondary infection of HEV in naive cells (Fig. 6A). These results indicate that eHEV particles in the exosome fraction isolated from the culture supernatant of HEV-infected cells are capable of transmitting infection and establishing productive infection in naive PLC/PRF/5 cells.

Neutralization of eHEV particles in the exosome fraction by antibodies against the HEV capsid protein.

Next, we investigated the exosome-mediated transmission of HEV infection in the presence of neutralizing antibodies. We previously reported that MAb H6225 against the HEV capsid protein can efficiently neutralize the nonenveloped particles in fecal specimens (34). In fact, membrane-unassociated neHEV particles, which were derived from the culture supernatant of HEV-infected cells following treatment with DOC-Na and trypsin, were neutralized by MAb H6225 (Fig. 7A). In contrast, the infectivity of these HEV particles was not affected by a negative-control MAb (MAb 905) (35). On the other hand, the HEV particles in the exosome fraction treated with MAb H6225 generated viruses with an efficiency similar to that for particles treated with a negative-control MAb (Fig. 7B). These results indicate that the HEV particles in the exosome fraction were not neutralized by the anti-ORF2 MAb because these particles were covered with a cellular lipid membrane, similar to enveloped viruses.

FIG 7.

Quantitation of HEV RNA in the culture supernatant of PLC/PRF/5 cells inoculated with neHEV particles (A) or eHEV particles included in the exosome fraction (B) that had been mixed with anti-ORF2 MAb (H6225; 1 mg/ml) or a negative-control MAb (905; 1 mg/ml) and cultured for up to 28 days. The data presented are means ± SD for three wells each.

DISCUSSION

Our previous study showed that eHEV particles are released together with ILVs through MVBs by the cellular exosomal pathway (32). For this reason, it is likely that the membrane structure of eHEV particles closely resembles that of exosomes. In this study, we characterized the exosomes purified from culture supernatants of HEV-infected cells. Electron microscopic images indicated that eHEV particles are 39.6 ± 1.0 nm in diameter and that the capsids of HEV particles are individually covered by lipid membranes, similar to enveloped viruses (Fig. 2, 4, and 5). This is the first report on the morphological features of eHEV and neHEV particles as demonstrated by immunoelectron microscopy (IEM).

HAV is known to be a nonenveloped virus. Recently, Feng et al. (24) reported that HAV released from cells is cloaked in host-derived membranes and that these enveloped viruses (eHAV) resemble exosomes. The culture supernatant of HAV-infected hepatoma cells contains two populations of virus particles in iodixanol gradients. The low-density population is consistent with membrane-associated HAV particles. EM of this fraction revealed numerous virus-like particles enclosed in membranes. These membrane-bound structures ranged from 50 to 110 nm in diameter, which is similar to the size of exosomes, and contained one to four virus-like particles, indicating that eHAV was protected from antibody-mediated neutralization (24). Similarly, it has been reported that the HCV particles in purified exosomes occur in both exosome-free and exosome-associated forms (26). The exosome-free fraction in the iodixanol working solutions contained only HCV virions that were mostly 50 to 65 nm in diameter, and occasionally up to 100 nm. In contrast, the exosome-associated fraction contained exosomes of ≥100 nm, with one or more HCV virions detected inside. Furthermore, it has been documented that exosome-associated HCV is infectious and resistant to neutralization by an anti-HCV neutralizing antibody. We previously demonstrated that membrane-associated HEV particles (eHEV, by analogy with eHAV) in culture supernatants and circulating blood are covered with cellular lipid membranes and can replicate efficiently in cultured cells despite the presence of neutralizing antibodies (28). Interestingly, the viral particles in the culture supernatants of HEV-infected cells exhibited a monophasic pattern (1.15 to 1.16 g/ml) on sucrose density gradient centrifugation (17). These observations suggest that almost all HEV particles in the culture supernatant are associated with the cellular membrane. EM images revealed that the capsids of HEV particles were individually covered by lipid membranes that resembled the lipid membranes of exosomes, similar to enveloped viruses (Fig. 4 and 5). Note, however, that exosomes with multiple virus-like particles, as with membrane-associated HAV or HCV, were not observed in the exosome fraction purified from culture supernatants of HEV-infected cells. Early EM studies revealed that neHEV particles in the feces and bile were 27 to 32 nm in diameter (36–38). In the present study, when we observed eHEV particles in HEV-infected cells by EM, the diameter of eHEV particles was estimated to be 39.6 ± 1.0 nm due to the thickness of the lipid membrane (3 to 8 nm) (32). After the treatment of eHEV particles in the exosome fraction and culture supernatant with detergent and protease, the diameter of virus-like particles was changed to 26.9 ± 0.9 nm (Fig. 4A and D). These findings support the hypothesis that HEV buds individually into the MVB to acquire the envelope structure and utilizes the cellular exosomal pathway for release.

MVBs or late endosomes are components of the endocytic pathway that range from 250 to 1,000 nm in diameter. ILVs are vesicles of 30 to 100 nm in diameter that are located within MVBs (39). MVBs can either be degraded or fuse with the plasma membrane, releasing the ILVs into the extracellular space. The ILVs are called exosomes following their release from the MVB (39). The formation of ILVs within the MVB and the budding of HEV share many features (30, 32). Although it has been suggested that the egress of HEV is dependent on the cellular endosomal sorting complexes required for transport (ESCRT) machinery (30), the origin of the eHEV membrane has been unclear. Exosomes contain proteins such as heat shock proteins (Hsp70 and Hsp90), membrane transport and fusion proteins (GTPases, annexins, and flotillin), proteins involved in MVB biogenesis (Alix and Tsg101), tetraspanins (CD9, CD63, CD81, and CD82), lipid-related proteins, and phospholipases (40). In contrast to heat shock proteins and membrane transport and fusion proteins, parts of tetraspanins, EpCAM, and PS are exposed on the surfaces of exosomes (40). Thus, in the present study, we isolated CD63-, CD9-, CD81-, EpCAM-, or PS-positive exosomes from the purified exosome fraction by immunoprecipitation. Western blotting of the precipitates confirmed the presence of CD63, CD9, and CD81 in exosome fractions derived from the culture supernatants of both HEV-infected and uninfected cells (Fig. 1A, upper panels). Furthermore, the exosomes purified from the culture supernatant of HEV-infected cells contained detectable levels of viral ORF2 and ORF3 proteins (Fig. 1A, middle and lower panels). In addition, approximately 90% of eHEV particles in the culture supernatant of HEV-infected cells were trapped by continuous immunocapture with anti-CD63, anti-CD9, anti-CD81, and anti-EpCAM MAbs and with PS binding protein (Table 2). When the exosome purification method was used, the rate of recovery of eHEV particles from the culture supernatant of HEV-infected cells was 69.2% (Fig. 1C); however, the recovery rate was increased to 85.9% and 90.8% when centrifugation for exosome purification was repeated two and three times, respectively (data not shown). These results indicate that the membrane components of eHEV released by the exosomal pathway closely resemble those of exosomes.

In a previous study, we generated MAb TA1708 against eHEV particles and reported that MAb TA1708 was capable of capturing eHEV particles, but not neHEV particles, in an immunocapture RT-PCR (31). This antibody specifically recognizes trans-Golgi network protein 2 (TGOLN2), an intracellular antigen derived from the trans-Golgi network. In this study, we found that MAb TA1708 also captured HEV particles in the exosome fraction in the immunocapture RT-PCR (Table 1). These results indicate that the eHEV particles present in the exosome fraction are associated with the lipid membrane and contain TGOLN2 on their surfaces. We used immunogold labeling to identify the eHEV particles in the exosome fraction. IEM imaging revealed that eHEV particles and exosomes were bound to MAb TA1708 (Fig. 5D). Western blotting of the exosomes isolated from the culture supernatant of HEV-infected or uninfected cells indicated the presence of an ∼37-kDa protein detectable by MAb TA1708 and anti-TGOLN2 MAb (data not shown), confirming that exosomes and eHEV particles possess TGOLN2 on the surface. In addition, immunoprecipitation and real-time RT-PCR revealed that eHEV particles retained the antigenicity of CD63 (33), one of the MVB marker proteins, on their surfaces (Table 2). These results support the hypothesis that membrane components are common among eHEV particles and exosomes and suggest that eHEV membranes are derived from the membranes of MVBs.

Feng et al. (29) proposed the presence of a “quasi-enveloped” virus, such as enveloped HAV (eHAV), which is distinguished from other enveloped viruses by the lack of viral glycoproteins (peplomers) in the surrounding lipid bilayer. Membrane-associated eHEV particles seem to be completely covered with a lipid membrane, as these particles cannot be captured by an anti-ORF2 or anti-ORF3 MAb upon immunocapture RT-PCR (17, 31). Interestingly, eHEV particles can be propagated in cultured PLC/PRF/5 cells as efficiently as neHEV particles can be propagated in fecal suspensions. Indeed, eHEV particles in circulating blood (28), culture supernatant (34), and the exosome fraction of the supernatant (Fig. 7B) were not neutralized by an anti-ORF2 MAb with neutralizing activity. Recent studies have shown that two forms of HEV particles (eHEV and neHEV) use different mechanisms to enter target cells (41, 42). At present, however, the entry receptor(s) for both eHEV and neHEV particles has not been well characterized. In the present study, EM indicated that the lipid membrane individually covered the capsids of HEV particles (Fig. 4 and 5). It is therefore very likely that the ORF2 protein of eHEV particles cannot bind to the entry receptor, suggesting that eHEV and neHEV utilize different types of entry receptors. On the other hand, Ramakrishnaiah et al. (25) analyzed the role of HCV entry receptors in the transmission of exosome-mediated HCV. Some inhibition of HCV RNA transfer was observed for the CD81, scavenger receptor class B member 1 (SR-BI), and claudin 1 entry receptors. These findings suggest that entry receptors may partly contribute to the uptake of exosomes even in the absence of a viral envelope or core proteins. In HAV entry, an antibody against the HAV receptor, TIM-1 (HAVCR-1), inhibited both eHAV and nonenveloped HAV (24). On the other hand, Feng et al. (24) suggested that eHAV bound to TIM-1 cannot undergo constitutive endocytosis or trafficking to late endosomes and lysosomes because eHAV membranes are stable at acidic pH levels. Further research is necessary to determine the relationships among the entry receptors, exosomes, and eHEV virions.

In this study, we showed that exosomes derived from HEV-infected PLC/PRF/5 cells can transmit productive HEV infection in vitro and are resistant to antibody neutralization (Fig. 6 and 7). We have reported that eHEV virions in blood circulation can replicate efficiency in cultured cells (28). In addition, we have demonstrated that immune sera have no ability to neutralize even HEV particles of the same strain in serum (28). As all the HEV vaccines are ORF2 based and have been shown to confer cross-genotypic protection (43–45), the role of eHEV and neHEV in causing infection that necessitates non-ORF2-based antibodies for protection emerges as an important concern. Future studies on agents targeting the entry of HEV into cells are warranted to protect against infection with eHEV virions.

In conclusion, the present study revealed that eHEV particles released by the cellular exosomal pathway are copurified with exosomes in the exosome fraction and that tetraspanins (CD63, CD9, and CD81), EpCAM, and PS, as well as TGOLN2, are present on the surfaces of eHEV virions, indicating that membrane components are common among eHEV particles and exosomes. Interestingly, EM images suggest that the capsids of HEV particles are individually covered by lipid membranes that resemble the lipid membranes of exosomes, similar to enveloped viruses. Furthermore, the eHEV particles in the exosome fraction are infectious and resistant to neutralization by anti-HEV ORF2 neutralizing antibody. Future studies on the entry receptors may help to elucidate the relationship between exosomes and exosome-like eHEV particles.

MATERIALS AND METHODS

Cell culture.

PLC/PRF/5 cells (ATCC CRL-8024; American Type Culture Collection) were grown in Dulbecco's modified Eagle medium (DMEM; Thermo Fisher Scientific) supplemented with 10% (vol/vol) heat-inactivated fetal bovine serum (FBS; Equitech-Bio), 100 U/ml penicillin, 100 μg/ml streptomycin, and 2.5 μg/ml amphotericin B (growth medium) at 37°C in a humidified 5% CO₂ atmosphere, as described previously (46).

Viruses.

Culture supernatants containing a cell culture-adapted genotype 3 JE03-1760F strain (passage 26; 4.3 × 10⁷ copies/ml) (47) or genotype 4 HE-JF5/15F strain (passage 6; 1.5 × 10⁸ copies/ml) (48) were used for virus inoculation. Membrane-unassociated neHEV particles were generated from eHEV particles in the culture supernatant or in the purified exosome fraction after treatment with 0.1% DOC-Na and 0.1% trypsin at 37°C for 3 h to remove the lipid membrane and the viral ORF3 protein on the surface.

Isolation of exosomes.

Exosomes from cell culture supernatants were isolated by ultracentrifugation according to a previously described method (49), with several modifications. In brief, the growth medium for culturing HEV-infected or uninfected PLC/PRF/5 cells was replaced by serum-free medium (VP-SFM; Gibco/Invitrogen), and cells were then cultured at 35.5°C. Every other day, half of the culture medium was replaced with serum-free medium. The collected culture supernatants were filtered through a 0.22-μm membrane, and then the filtrates were centrifuged at 100,000 × g for 70 min at 4°C, using an SW40 rotor (Beckman Coulter Instruments). The exosome pellet was washed with phosphate-buffered saline without Mg²⁺ and Ca²⁺ [PBS(−)] and then resuspended in PBS(−) and stored at 4°C until use.

Isolation of CD63-, CD9-, CD81-, EpCAM-, and PS-positive exosomes.

CD63-, CD9-, CD81-, EpCAM-, or PS-positive exosomes in the culture supernatant or in its exosome fraction were isolated by immunoprecipitation using antibody-coated beads and exosome-human CD63 isolation/detection reagent (from cell culture medium), exosome-human CD9 isolation reagent (from cell culture), or exosome-human CD81 isolation reagent (from cell culture) (Thermo Fisher Scientific), an ExoCap EpCAM kit for serum plasma (JSR Life Sciences), or PS binding protein-coated beads and a MagCapture exosome isolation kit PS (Wako), respectively, according to the manufacturers' instructions.

Western blotting.

The bead-bound exosomes were lysed in lysis buffer (50 mM Tris-HCl [pH 8.0], 1% NP-40, 150 mM NaCl, and protease inhibitor cocktail [Sigma-Aldrich]), and the exosome lysates were separated by SDS-PAGE. The proteins were blotted onto polyvinylidene difluoride membranes (0.45 μm; Millipore), immunodetected with anti-CD63 MAb (Sigma-Aldrich), anti-CD9 MAb (Sigma-Aldrich), anti-CD81 polyclonal antibody (PAb) (System Biosciences), anti-HEV ORF2 MAb (H6210) (34), or anti-HEV ORF3 MAb (TA0536) (27), and then visualized by chemiluminescence assay, using an ImageQuant LAS500 system (GE Healthcare) as described previously (18).

Amplification and nucleotide sequencing of a nearly full-length HEV genome.

To examine whether the full-length HEV genome was incorporated into the particles in exosome fractions, a long-distance RT-PCR method capable of amplifying nearly full-length HEV RNA was performed. In brief, total RNA was extracted from exosome fractions by use of TRIzol-LS reagent (Thermo Fisher Scientific) and then used as a template to synthesize cDNA by use of a 41-mer oligonucleotide containing a T₁₅ sequence (SSP-T; 5′-AAGGATCCGTCGACATCGATAATACG[T]₁₅-3′) and SuperScript IV reverse transcriptase (Thermo Fisher Scientific). The cDNA was then amplified by a nested PCR with KOD FX Neo (Toyobo), using primers HE017 (5′-GTGGTCGATGCCATGGAGGCCCA-3′ [plus-strand sequence from nt 14 to 36 of the HEV genome]) and 166 (5′-AAGGATCCGTCGACATCGAT-3′ [outer sequence of SSP-T]) in the first round (94°C for 2 min followed by 40 cycles of 98°C for 10 s, 53°C for 20 s, and 68°C for 4 min) and primers HE019 (5′-CGATGCCATGGAGGCCCAYCAGTT-3′ [plus-strand sequence from nt 19 to 42 of the HEV genome]) and HE041 (5′-GCCAATGGCGAGCYGACWGTKAA-3′ [Y = T/C, W = A/T, and K = G/T; minus-strand sequence from nt 6381 to 6403 of the HEV genome]) in the second round (94°C for 2 min followed by 30 cycles of 98°C for 10 s, 55°C for 20 s, and 68°C for 4 min). The amplification product was sequenced directly on both strands by use of a BigDye Terminator v3.1 cycle sequencing kit (Thermo Fisher Scientific) on an ABI Prism 3130xl genetic analyzer (Thermo Fisher Scientific).

Quantification of HEV RNA.

Total RNA was extracted from the exosome fraction or culture supernatant by use of TRIzol-LS reagent. The quantification of HEV RNA was performed by real-time RT-PCR, using a LightCycler apparatus (Roche), a QuantiTect Probe RT-PCR kit (Qiagen), and primer sets and a probe targeting the ORF2 and ORF3 overlapping region, as described previously (34).

Electron microscopy.

Exosomes isolated from the culture supernatants of HEV-infected or uninfected cells by ultracentrifugation or purified membrane-unassociated neHEV particles were resuspended in PBS(−). neHEV particles were generated as described above and then pelleted through a 30% sucrose cushion at 270,000 × g for 4 h at 4°C. Exosomes or viral particles were allowed to adsorb for 1 min onto Formvar-coated EM grids. Excess liquid was absorbed with filter paper. The grids were negatively stained by use of uranyl acetate for 1 min and then observed by TEM (model HT-7600; Hitachi) at an acceleration voltage of 80 kV. Two hundred particles were used for the calculations to measure the viral particle size (reported as the mean ± standard deviation [SD]).

Digitonin treatment and sucrose density gradient centrifugation.

The exosome fraction containing eHEV particles (1 × 10⁶ copies), collected from the culture supernatant of PLC/PRF/5 cells inoculated with HEV, was treated with or without 1.5% digitonin (Nacalai Tesque) at room temperature for 15 h. The digitonin-treated exosome fraction was subjected to equilibrium centrifugation in a sucrose density gradient as described previously (27). The gradients were fractionated, and the density of each fraction was measured by refractometry.

Immunocapture RT-PCR.

Immunocapture RT-PCR analysis of the peak fractions from sucrose density gradient centrifugation was performed as described previously (27). The following antibodies were used in the present study: an anti-HEV ORF2 MAb (H6225) (34), an anti-HEV ORF3 MAb (TA0536) (27), and MAb TA1708, which is capable of capturing eHEV particles (31).

Immunoelectron microscopy.

Immunogold labeling was performed according to the method described by Zhang et al. (50), with some modifications. Viral suspensions or exosomes purified from the culture supernatant of HEV-infected cells by ultracentrifugation were fixed for 5 min with 2% paraformaldehyde, and then Formvar-coated nickel grids were floated on drops of virus suspension for 20 min. The grids were treated with PBS(−) containing 1% (vol/vol) bovine serum albumin (BSA) for 20 min, and the first antibody (anti-HEV ORF2 MAb [H6225; mouse IgG], anti-HEV ORF3 MAb [TA0536; mouse IgG], or MAb TA1708 [mouse IgM]) was applied and incubated for 1 h [1/100 dilution in PBS(−) containing 0.2% BSA]. After washing five times with 0.2% BSA in PBS(−), colloidal gold-conjugated secondary antibodies were incubated with the grids for 2 h (1/20 dilution of goat anti-mouse IgG or IgM conjugated to 12-nm colloidal gold; Jackson ImmunoResearch Laboratories). The grids were washed five times with 0.2% BSA in PBS(−) and rinsed three times with PBS(−). Finally, the grids were negatively stained and observed as described above.

Virus inoculation.

Monolayers of PLC/PRF/5 cells in 6-well plates were inoculated with 1.0 × 10⁶ copies of eHEV particles in the exosome fraction of JE03-1760F-infected cells pelleted by ultracentrifugation. After incubation at room temperature for 1 h, the cells were washed with PBS(−), 2 ml of growth medium was added to each well, and the cells were incubated at 35.5°C. Every other day, half of the culture medium (1 ml) of the HEV-infected cells was replaced with growth medium. The collected culture medium was centrifuged at 1,300 × g at room temperature for 2 min, and the supernatant was stored at −80°C until use.

Immunofluorescence assay.

PLC/PRF/5 cells inoculated with the exosome fractions purified from supernatants of HEV-infected or uninfected cells at 10 days postinoculation in an 8-well chamber slide (Thermo Fisher Scientific) were subjected to immunofluorescence staining using a MAb against the viral ORF2 protein (H6225) (34) or ORF3 protein (TA0536) as the primary antibody, followed by Alexa Fluor 488-conjugated anti-mouse IgG (Thermo Fisher Scientific) as the secondary antibody, as described previously (18). The nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI; Roche). The slide glasses were mounted with Fluoromount/Plus medium (Diagnostic BioSystems) and then viewed under an FV1000 confocal laser microscope (Olympus).

Neutralization assay.

One hundred microliters of membrane-unassociated neHEV particles or membrane-associated eHEV particles in the exosome fraction pelleted by ultracentrifugation, containing 1.0 × 10⁵ copies of HEV RNA, was mixed with an equal volume of anti-ORF2 MAb (H6225; 1 mg/ml) or a MAb against hepatitis B virus e antigen as a negative control (MAb 905) (35) and kept at room temperature for 1 h. Monolayers of PLC/PRF/5 cells in 6-well plates were inoculated with the mixture. After incubation at room temperature for 1 h, the cells were washed with PBS(−), and 2 ml of growth medium was added to each well. The cells were then incubated at 35.5°C. Every other day, half of the culture medium (1 ml) of the HEV-infected cells was replaced with growth medium. The collected culture medium was centrifuged at 1,300 × g at room temperature for 2 min, and the supernatant was stored at −80°C until use.