Chicken Organic Anion-Transporting Polypeptide 1A2, a Novel Avian Hepatitis E Virus (HEV) ORF2-Interacting Protein, Is Involved in Avian HEV Infection

The process of viral infection is centered around the interaction between the virus and host cells. Due to the lack of a highly effective cell culture system in vitro, there is little understanding about the interaction between avian HEV and its host cells. In this study, a total of seven host proteins were screened in chicken liver cells by a truncated avian HEV capsid protein (ap237) in which the host protein OATP1A2 interacted with ap237. Overexpression of OATP1A2 in the cells can promote ap237 adsorption as well as avian HEV adsorption and infection of the cells. When the function of OATP1A2 in cells was inhibited by substrates or inhibitors, attachment and infection by avian HEV significantly decreased. The distribution of OATP1A2 in different chicken tissues corresponded with that in tissues during avian HEV infection. This is the first finding that OATP1A2 is involved in viral infection of host cells.

ABSTRACT

Avian hepatitis E virus (HEV) is the main causative agent of big liver and spleen disease in chickens. Due to the absence of a highly effective cell culture system, there are few reports about the interaction between avian HEV and host cells. In this study, organic anion-transporting polypeptide 1A2 (OATP1A2) from chicken liver cells was identified to interact with ap237, a truncated avian HEV capsid protein spanning amino acids 313 to 549, by a glutathione S-transferase (GST) pulldown assay. GST pulldown and indirect enzyme-linked immunosorbent assays (ELISAs) further confirmed that the extracellular domain of OATP1A2 directly binds with ap237. The expression levels of OATP1A2 in host cells are positively correlated with the amounts of ap237 attachment and virus infection. The distribution of OATP1A2 in different tissues is consistent with avian HEV infection in vivo. Finally, when the functions of OATP1A2 in cells are inhibited by its substrates or an inhibitor or blocked by ap237 or anti-OATP1A2 sera, attachment to and infection of host cells by avian HEV are significantly reduced. Collectively, these results displayed for the first time that OATP1A2 interacts with the avian HEV capsid protein and can influence viral infection in host cells. The present study provides new insight to understand the process of avian HEV infection of host cells.

IMPORTANCE The process of viral infection is centered around the interaction between the virus and host cells. Due to the lack of a highly effective cell culture system in vitro, there is little understanding about the interaction between avian HEV and its host cells. In this study, a total of seven host proteins were screened in chicken liver cells by a truncated avian HEV capsid protein (ap237) in which the host protein OATP1A2 interacted with ap237. Overexpression of OATP1A2 in the cells can promote ap237 adsorption as well as avian HEV adsorption and infection of the cells. When the function of OATP1A2 in cells was inhibited by substrates or inhibitors, attachment and infection by avian HEV significantly decreased. The distribution of OATP1A2 in different chicken tissues corresponded with that in tissues during avian HEV infection. This is the first finding that OATP1A2 is involved in viral infection of host cells.

INTRODUCTION

Hepatitis E virus (HEV) belongs to the Hepeviridae family, which contains the genera Orthohepevirus and Piscihepevirus (1). All four major HEV genotypes that infect humans, genotypes 5 and 6 isolated from wild boar, and genotypes 7 and 8 from camels are assigned to the species Orthohepevirus A. Avian HEV, the second known animal strain of HEV, belongs to the species Orthohepevirus B (1, 2). It was identified from chickens with big liver and spleen disease, also known as hepatitis-splenomegaly syndrome (3), which can cause slightly increased mortality (1% to 4%) and decreased egg production (10% to 40%) in broiler breeders and laying hens aged 30 to 72 weeks (4–6). In addition, avian HEV RNA has also been detected in healthy chickens (7). To date, both the fecal-oral transmission route and vertical transmission of avian HEV have been demonstrated (8, 9).

Until now, five genotypes (genotypes 1 to 5) and a single serotype of avian HEV from chickens have been identified (10–15). The avian HEV genome is a positive-sense single-stranded RNA of approximately 6.6 kb, which consists of three open reading frames (ORFs): ORF1, ORF2, and ORF3 (16). Of these, ORF2 encodes the virus capsid protein, including 606 amino acids (aa) (16). Some previous studies indicated that the capsid protein is closely related to viral infection of host cells and induction of the immune response (17–21). Over the last decade, the major focus of research was on the antigen properties of the capsid protein (18–20, 22), but less effort has been directed toward its function in virus infection.

In regard to human HEV, it was documented that the truncated ORF2 protein named p239 (amino acids 368 to 606), a self-assembling viruslike particle that covers the complete P domain (23), can bind to HepG2 cells and serve as a material replacing the natural viral particle to research the interaction between the virus and host cells (24). Next, utilizing p239 as a bait protein, the host factors GRP78/Bip, α-tubulin, heat shock protein 90 (HSP90), cytochrome P4502C8, and retinol-binding protein 4 were screened and specifically interacted with the HEV ORF2 protein (25, 26). In addition, using another truncated ORF2 protein expressed in insect cells as a bait protein (amino acids 112 to 606), several membrane proteins, such as heparin surface proteoglycans (27), asialoglycoproteins ASGR1 and ASGR2 (28), and transmembrane protein 134 (29), were identified. The functions of these host factors in virus infection are different. For example, both heparin surface proteoglycans and asialoglycoproteins mainly mediate viral binding and entry, while transmembrane protein 134 (located in the endoplasmic reticulum) negatively regulates ORF2-mediated inhibition of the NF-κB signaling pathway.

In this study, based on alignments of the amino acids between human and avian HEV ORF2 proteins, the region spanning aa 313 to 549 of the avian HEV ORF2 protein (named ap237) was selected as the bait protein. This region corresponded with the amino acid region of the human HEV p239 protein. In some previous studies, the results showed that ap237 contains most of the antigenic epitopes of avian HEV (18–20) and the key domain (aa 471 to 507) for binding to LMH cells (30) derived from chicken hepatocellular carcinoma epithelial cells (31), which support avian HEV replication (32). Next, ap237 was employed as a bait protein to target the host factors in chicken liver tissue. A total of seven host proteins were pulled from chicken liver cells by ap237, and of these host proteins, organic anion-transporting polypeptide 1A2 (OATP1A2), a multiple-transmembrane protein localizing on the cell membrane and expressed in the liver, was chosen for subsequent research. First, direct binding between ap237 and the ectodomain of OATP1A2 was determined. Following this, the functions of OATP1A2 during avian HEV attachment and infection were analyzed using an LMH cell line lacking endogenous OATP1A2 and LMH cells stably expressing OATP1A2. Finally, the correlations of OATP1A2 expression and avian HEV infection in different tissues were determined. The results of the present study indicate that OATP1A2 is a cofactor involved in avian HEV infection of host cells.

RESULTS

Design, expression, and purification of GST-ap237.

In a previous study, it was documented that the region from aa 368 to 606 of the human HEV ORF2 protein (named p239) was expressed by a bacterial system and can form polymers (33). p239 can enter the host cells by mimicking the natural HEV particle (24). Through an alignment of human and avian HEV ORF2 amino acids, it was observed that the region spanning aa 313 to 549 of the avian HEV ORF2 protein corresponded to the p239 region of the human HEV ORF2 protein, and this region was selected (Fig. 1A). In addition, three-dimensional (3D) modeling of the avian HEV ORF2 protein showed that ap237 consisted of part of a middle (M) domain (aa 313 to 400) and a complete protruding (P) domain (aa 401 to 549) (Fig. 1B), which was predicted based on the 3D structure of the human capsid protein (23).

FIG 1.

Designation, prediction, expression, and identification of soluble GST-ap237 in bacteria. (A) Amino acid alignments of genotype 3 human and avian HEV truncated capsid proteins. The alignments of capsid proteins from genotype 3 HEV and CaHEV isolates were performed using the Clustal W module of the MegAlign program of Lasergene 7.1 (DNASTAR, Inc.). (B) 3D structure of the CaHEV capsid protein predicted by using SWISS-MODEL and visualized by using CHIMERA software. The backbone of ap237 (aa 313 to 549) is indicated in green in the pentamer (left). The P domain (blue), M domain (dark blue), S domain (purple), and proline-rich hinge (green) of the monomer are highlighted with different colors. (C) SDS-PAGE of GST-ap237 protein produced in E. coli Transetta(DE3) cells. (D) Western blot analysis of the expression of recombinant GST-ap237 using anti-GST antibody. A GST tag expressed by empty vector pGEX-6P-1 was used as a negative control. (E) Western blot analysis of the expression of recombinant GST-ap237 using the 3E8 mAb. The vertical lines in panels D and E indicate marker (M) images; the Western blot results were taken separately and joined together.

According to the above-described alignments and prediction, ap237 with an N-terminal glutathione S-transferase (GST) tag was designed and expressed as a bait protein for the GST pulldown assay. The soluble recombinant protein GST-ap237 (55 kDa) was successfully expressed in the supernatant of the cell lysate (Fig. 1C) and purified using the glutathione-Sepharose 4 Fast Flow system (Fig. 1C). Western blot analysis revealed that GST-ap237 was bound with both anti-GST antibodies (Fig. 1D) and the 3E8 monoclonal antibody (mAb) specific for ap237 (Fig. 1E).

Screening of host proteins specifically pulled by ap237.

To identify cellular proteins from chicken liver cells that interact with ap237, a GST pulldown assay and mass spectrometry (MS) were performed. The silver staining results of SDS-PAGE showed that there were at least five specific bands in GST-ap237 lanes that were not observed in the lanes with only GST (Fig. 2A). These five bands were cut and mixed for analysis via MS. The MS results showed that 7 proteins were identified (Table 1). Of these seven proteins, the protein under UniProt accession no. A0A1D5PMA0 (solute carrier organic anion transporter family member 1A2 [OATP1A2]) was predicted to be a cell membrane protein, while the other proteins were located in the cytoplasm or nucleus. To confirm the cellular location of OATP1A2, primary cultured chicken embryo hepatocytes (CEH) were probed with anti-GST-1A2^ecto (ectodomain of OATP1A2) mouse sera and detected using confocal immunofluorescence and Western blot assays. The results showed that the OATP1A2 was located mainly in the cell membrane (Fig. 2B and C).

FIG 2.

SDS-PAGE analysis of host proteins by a GST pulldown assay with GST-ap237 and identification of OATP1A2 in chicken embryo hepatocytes (CEH). (A) GST pulldown assay. GST protein was used to exclude the proteins that bind with the GST tag, and GST or GST-ap237 without incubation of the chicken liver cell lysate was set as a blank control. The eluted protein complex was resolved by 10% SDS-PAGE, followed by silver staining. Five visible bands specifically pulled by the ap237 protein are indicated by red arrows. (B) Immunofluorescence assay of OATP1A2 protein in CEH using murine anti-GST-1A2^ecto sera and TRITC-goat anti-mouse IgG. Bar, 25 μm. (C) Western blot analysis of OATP1A2 protein using murine anti-GST-1A2^ecto sera and HRP-goat anti-mouse IgG. WCL, CEH whole-cell lysates; LMH, LMH cell lysates; Cytos, CEH cytosolic proteins; SMPs & MAPs, solubilized membrane and membrane-associated proteins from CEH; MFs, CEH membrane fractions.

TABLE 1.

Proteins present in five specific bands observed in the GST-ap237 lane of Fig. 2A identified by MS

UniProt accession no.	Protein (short name, MW [kDa])a	Sequence coverage (%)	No. of unique peptides	Location(s)
A0A1D5PMA0	Solute carrier organic anion transporter family member (OATP1A2, 72.542)	6.2	4	Cell membrane
P23668	16-kDa beta-galactoside-binding lectin (galectin CG-16, 14.9)	23.88	3	Extracellular region or secreted
A0A1D5PDV6	Ribosomal protein S19 (RPS19, 15.4)	36.23	5	Nucleus
E1BRU7	Hexokinase domain-containing 1 (HKDC1, 102.2)	27.32	4	Cytosol and mitochondrion
F1N8M4	Dynamin-like 120-kDa protein, mitochondrial (OPA1, 113.3)	5.42	5	Mitochondrion
P84172	Elongation factor Tu, mitochondrial (TUFM, 38.3)	43.43		Mitochondrion
P0CG62	Polyubiquitin B (UBB, 34.4)	44.59	3	Nucleus

^aMW, molecular weight.

OATP1A2 interaction and colocalization with ap237.

Upon review of MS results of the silver-stained bands and MS results for the protein complex, one of the host proteins, a membrane protein, OATP1A2, was screened. In human, OATP1A2 is a transmembrane transporter and expressed on the cell membrane in liver (34). It is well known that the chicken liver is the major tissue targeted by avian HEV infection (9), so OATP1A2 was selected for this study. First, to further confirm ap237 interaction with OATP1A2, a coimmunoprecipitation (co-IP) assay was performed by transient coexpression of OATP1A2-3×Flag and hemagglutinin (HA)-ap237 in HEK 293T cells. The results showed that OATP1A2-3×Flag can be pulled by HA-ap237 using anti-HA mAb but not by HA-ORF3 (Fig. 3A). In turn, HA-ap237, and not HA-ORF3, can also be pulled by OATP1A2-3×Flag using an anti-Flag mAb (Fig. 3B). In addition, to examine the localization of ap237 and OATP1A2 in cells, HEK 293T cells were transiently cotransfected with plasmids as described above. After 48 h of transfection, the subcellular localization of OATP1A2-3×Flag and HA-ap237 was examined by confocal microscopy. Both of the proteins were distributed throughout the cytoplasm, and OATP1A2 colocalized extensively with the ap237 protein with Manders’ overlap coefficient of 0.89 (Fig. 3C).

FIG 3.

Interaction and colocalization of avian HEV capsid protein and chicken OATP1A2. (A) Immunoprecipitation with protein G-anti-HA mAb. HEK 293T cells were cotransfected with the recombinant plasmids, and the cell lysates obtained at 48 h posttransfection were immunoprecipitated with anti-HA mAb. The cell lysate and protein G-antibody-antigen complexes were detected by Western blot analysis. (B) Immunoprecipitation with protein G-anti-Flag mAb. (C) Colocalization of OATP1A2 with ap237. HEK 293T cells were cotransfected with recombinant plasmids, and cells were then fixed and subjected to indirect immunofluorescence analysis using mouse anti-HA mAb and rabbit anti-Flag polyclonal antibody. The nucleus is indicated by DAPI (blue) staining in the images.

Direct binding of the OATP1A2 ectodomain with ap237.

To identify the part of OATP1A2 directly interacting with ap237, OATP1A2 and ap237 without any tags were separately expressed by a bacterial system. For OATP1A2, the results showed that the protein was a multiple-transmembrane protein with 9 transmembrane domains, as predicted (Fig. 4). Meanwhile, the 3D structure of OATP1A2 from chickens was predicted using the Phyre2 web portal with a confidence of 100% and a coverage of 66%; this also exhibited the transmembrane structure (Fig. 5A). Next, the direct interaction of 1A2^ecto (ectodomain of OATP1A2) and ap237 was determined by a GST pulldown assay and an indirect enzyme-linked immunosorbent assay (ELISA). First, using the bacterial system, 1A2^ecto with an N-terminal GST tag was successfully expressed as a soluble form, as shown by SDS-PAGE (Fig. 5B, left), and reacted with anti-GST antibodies, as demonstrated by Western blot analyses (Fig. 5B, right); ap237 without any tag was also successfully expressed by a bacterial system (Fig. 5C). Under nondenaturing conditions, a part of the protein formed a multimeric structure. Next, by the GST pulldown assay, Western blot analysis showed that ap237 was pulled by GST-1A2^ecto as the bait protein, but no pulling of ap237 was observed with only GST (Fig. 5D). Furthermore, the sp239 control protein from swine HEV was not pulled by GST-1A2^ecto. In addition to the GST pulldown assay, an indirect ELISA was also used to confirm the direct binding of 1A2^ecto with ap237. The optical density at 450 nm (OD₄₅₀) values revealed that GST-1A2^ecto can bind in a specific and dose-dependent manner with ap237 at a GST-1A2^ecto concentration of 20 μg/ml, but this does not occur with GST protein (Fig. 5E). Meanwhile, indirect ELISA results also indicated that chicken OATP1A2 (GST-1A2^ecto) specifically binds with ap237 but not with other capsid proteins from mammalian HEVs, namely, sp239, r239, sar239, and ker239, and porcine reproductive and respiratory syndrome virus (PRRSV) N protein (Fig. 5F). Furthermore, chicken OATP1A2 (GST-1A2^ecto) also specifically binds with the soluble capsid proteins in avian HEV-infected sera (Fig. 5F).

FIG 4.

Prediction of extracellular domains of chicken OATP1A2 protein. Prediction of the transmembrane helices of chicken OATP1A2 was performed by using the TMHMM v.2.0 server. (A) List of the locations of the predicted transmembrane helices (TMHs). (B) Posterior probabilities of inside/outside/transmembrane helix. (C) Amino acid sequence of 1A2^ecto.

FIG 5.

The OATP1A2 ectodomain directly interacts with ap237. (A) OATP1A2 structure predicted using the Phyre2 Web portal. (B) SDS-PAGE (left) and Western blot analysis (right) of GST-1A2^ecto produced in E. coli Transetta(DE3) cells. (C) SDS-PAGE analysis of the expression of ap237 protein without any tag. The purified ap237 protein was treated with loading buffer containing 2-mercaptoethanol and boiled for 10 min (lane 1) and treated with loading buffer without 2-mercaptoethanol (lane 2). (D) GST pulldown assay to test 1A2^ecto binding with ap237. Beads conjugated to GST (lane 1) or GST-1A2^ecto (lane 3) were incubated with ap237. Incubation of the beads conjugated to GST-1A2^ecto with sp239 (lane 2) was used as a control. After washing, proteins eluted from beads were analyzed via SDS-PAGE and then detected by anti-GST antibodies and the 3E8 mAb. (E) Detection of chicken OATP1A2 binding with ap237 by an indirect ELISA. ELISA plates were coated with different concentrations of ap237 and separately incubated with different concentrations of GST or GST-1A2^ecto. The binding of proteins was detected by anti-GST antibodies. (F) Evaluation of OATP1A2 binding with different HEV capsid proteins by an indirect ELISA. Plates were coated with the proteins ap237, sp239, r239, sar239, and ker239 (4 μg/well) and incubated with GST or GST-1A2^ecto. PRRSV N protein was used as a negative control.

LMH^1A2-GFP cell lines with stable expression of OATP1A2 and knockdown treatment.

To investigate the relationship of OATP1A2 with ap237 in vitro, the LMH^1A2-GFP cell line, with stable expression of OATP1A2 fused with green fluorescent protein (GFP), was generated by lentivirus infection. The LMH^GFP cell line, expressing only GFP, was also generated as the control. Green fluorescence from GFP in LMH^1A2-GFP cells was observed under a fluorescence microscope (Fig. 6A, left), and the expression of OATP1A2-GFP was also detected by anti-GST-1A2^etco sera with red fluorescence (Fig. 6A, right). In addition, quantitative PCR (qPCR) showed that the relative mRNA level of OATP1A2 in LMH^1A2-GFP cells was significantly (P < 0.001) higher than those in LMH cells, LMH^GFP cells, DF-1 cells, and CEH (Fig. 6B). Western blot analysis further showed that OATP1A2 fused with GFP was expressed mainly in solubilized membrane proteins (SMPs), membrane-associated proteins (MAPs), and membrane fractions (MFs) and in small amounts in the cytoplasm (Cytos) (Fig. 6C).

FIG 6.

Identification of the LMH^1A2-GFP cell line and knockdown of OATP1A2 expression. (A) Immunofluorescence assay of OATP1A2-GFP expression in the LMH^1A2-GFP cell line. The recombinant protein OATP1A2-GFP was detected using mouse anti-GST-1A2^ecto sera and TRITC-conjugated goat anti-mouse IgG. The nucleus is indicated by DAPI (blue) staining in the images. (B) Relative expression of OATP1A2 mRNA in LMH, LMH^GFP, LMH^1A2-GFP, DF-1, and CEH cells. LMH cells were used to normalize the relative expression levels in other cells. (C) Subcellular localization of OATP1A2 in LMH^1A2-GFP cells. A total of 5 × 10⁶ LMH^1A2-GFP cells (or LMH^GFP cells) were used to extract the membrane protein. Three products, cytosolic proteins (Cytos), solubilized membrane and membrane-associated proteins (SMPs & MAPs), and membrane fractions (MFs), were subjected to Western blot analysis using the anti-GFP antibody. (D) Knockdown of OATP1A2 expression in LMH^1A2-GFP cells. LMH^1A2-GFP cells were transfected with three OATP1A2-specific siRNAs (si1A2-1, si1A2-2, and si1A2-3) or a control small interfering RNA (siNCtrl). After transfection for 48 h, the cells were collected, and total RNAs were obtained to detect the expression of OATP1A2 RNA. Western blotting was performed using rabbit anti-GFP antibodies. (E) qPCR and Western blotting for detection of OATP1A2 expression in LMH^1A2-GFP cells treated with different concentrations of si1A2-1. Error bars indicate the standard errors of the means (SEM). ***, P < 0.001; ****, P < 0.0001; ns, not significant.

To knock down the expression of OATP1A2 in LMH^1A2-GFP cells, three small interfering RNAs (siRNAs) (si1A2-1, si1A2-2, and si1A2-3) targeting the mRNA encoding OATP1A2 were transfected into the cells. All three siRNAs effectively (P < 0.001) knocked down the expression of OATP1A2 mRNA and protein in LMH^1A2-GFP cells, at a concentration of 50 nM (Fig. 6D). Comparing the OATP1A2 expression levels in LMH^1A2-GFP cells transfected with the three siRNAs, si1A2-1 was selected for the following experiments (Fig. 6D), since OATP1A2 mRNA was significantly decreased with various concentrations of si1A2-1 (10 nM, 20 nM, 30 nM, 50 nM, and 100 nM) (Fig. 6E, top). Western blotting showed that, except for a small amount of expression at 10 nM, OATP1A2 was not detected at other concentrations in LMH^1A2-GFP cells (Fig. 6E, bottom).

OATP1A2 enhances ap237 attachment to LMH cells.

A previous study showed that ap237 can mimic natural avian HEV attachment to host cells (30). In this study, ap237 was used to analyze the relationship between ap237 attachment and the expression of OATP1A2 in cells. The Western blot results showed that the amounts of ap237 binding with LMH^1A2-GFP cells transfected with control siRNA (siNCtrl) were significantly larger than those with LMH and LMH^GFP cells (Fig. 7A), which had similar amounts of bound ap237. Meanwhile, when the expression levels of OATP1A2 in LMH^1A2-GFP cells were reduced by si1A2-1 (50 nM) interference, the amount of ap237 attachment to these cells markedly decreased (Fig. 7A). Additionally, a flow cytometry assay revealed that ap237 attached to 17.2% of LMH^GFP cells and 59.8% of LMH^1A2-GFP cells, which is a significant increase. Furthermore, ap237 attachment decreased (34.3%) in LMH^1A2-GFP cells transfected with si1A2-1 (Fig. 7B). Collectively, both the Western blot and flow cytometry results suggested that high expression levels of OATP1A2 in the cells enhance ap237 attachment.

FIG 7.

High levels of OATP1A2 expression enhance ap237 binding to LMH cells. (A) Western blot analysis of ap237 binding to the cell lines LMH, LMH^GFP, and LMH^1A2-GFP transfected with siNCtrl or si1A2-1. The binding of ap237 to cells was analyzed by Western blotting using the 3E8 mAb. (B) Flow cytometry analysis of ap237 binding to LMH, LMH^GFP, and LMH^1A2-GFP cells transfected with siNCtrl or si1A2-1. After incubation with ap237, cells were dissociated with enzyme-free cell dissociation buffer and fixed with 4% paraformaldehyde for use in the flow cytometry assay. The mAb 3E8 was used to detect ap237.

OATP1A2 enhances avian HEV attachment to and infection of LMH cells.

To determine avian HEV infection in LMH^1A2-GFP cells, the presence of replicative intermediates of avian HEV RNA in the cells was detected by negative-strand-specific reverse transcription-PCR (RT-PCR), and avian HEV loads in the cell lysates over time were detected by TaqMan real-time RT-PCR. CEH were used as positive controls. The results showed that in CEH and LMH^1A2-GFP cells, the negative strands of avian HEV isolated in China (CaHEV) were detected starting at 2 days postinfection, while in LMH and LMH^GFP cells, they were detected starting at 4 days postinfection (Fig. 8A). The quantification of viral ORF3 RNA revealed that the CaHEV loads increased over time in all four cell lines, and the numbers of ORF3 RNA copies in LMH^1A2-GFP cells were significantly higher than those in the other three cell lines (Fig. 8A), which were correlated with the expression levels of OATP1A2 RNAs in these cells (Fig. 6B). These results indicated that CaHEV can effectively replicate and propagate in LMH^1A2-GFP cells.

FIG 8.

High expression levels of OATP1A2 enhance CaHEV attachment to and infection of LMH cells. (A) Determination of CaHEV propagation in CEH, LMH, LMH^GFP, and LMH^1A2GFP cells. The infected cells were collected, and total RNA was extracted to detect the negative-strand (top) and positive-strand (bottom) CaHEV ORF3 genes using TaqMan real-time RT-PCR and negative-strand-specific RT-PCR, respectively. The presence and absence of negative-strand CaHEV RNA are indicated by “+” and “−,” respectively. (B and C) CaHEV attachment (B) and infection (C) assays with LMH, LMH^1A2-GFP, LMH^GFP, and LMH^1A2-GFP cells treated with siRNA. CaHEV ORF3 RNA was detected by TaqMan real-time RT-PCR. Error bars indicate the SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant.

To clarify the relationship between OATP1A2 expression levels and natural avian HEV infection, viral attachment and infection assays using the differently treated cell lines (LMH, LMH^GFP, LMH^1A2-GFP, and LMH^1A2-GFP transfected with si1A2-1) were performed in vitro. First, the amounts of CaHEV attachment to and infection of LMH^1A2-GFP cells were markedly (P < 0.001) larger than those in LMH and LMH^GFP cells (Fig. 8B and C). Second, when different concentrations (10 nM, 20 nM, 30 nM, 50 nM, and 100 nM) of si1A2-1 targeting OATP1A2 were used to knock down the expression levels of OATP1A2, as shown in Fig. 6E, the amounts of CaHEV attachment to and infection of these cells were significantly decreased (Fig. 8B and C). In cells transfected with siNCtrl, OATP1A2 expression and viral attachment and infection showed no significant change (P > 0.05) (Fig. 8B and C).

Inhibition of OATP1A2 reduces avian HEV attachment to and infection of LMH cells.

OATP1A2 is an important member of the OATP transporter family and participates in the absorption and transportation of drugs (e.g., imatinib) and some endogenous substrates (e.g., chenodeoxycholic acid [CDCA] and sodium cholate [SC]). To explore whether or not the substrates or inhibitor of OATP1A2 could specifically inhibit CaHEV attachment to and infection of LMH^1A2-GFP cells, LMH^1A2-GFP cells were preincubated with the substrates CDCA, SC, and imatinib or the inhibitor carvedilol before CaHEV inoculation. First, the cytotoxicity of these reagents to LMH cells was determined by separately treating the cells with different concentrations of the reagents (data not shown). The results showed that the maximum working concentrations of CDCA, SC, imatinib, and carvedilol were 50 μM, 50 μM, 10 μM, and 10 μM, respectively (Fig. 9A). Under the maximum concentrations of the substrates and the inhibitor, the expression of OATP1A2 at both the mRNA and protein levels was significantly decreased to various extents (Fig. 9B). In the CaHEV attachment and infection assays, the amount of CaHEV ORF3 RNA significantly (P < 0.05) decreased in LMH^1A2-GFP cells preincubated with CDCA, SC, carvedilol, and imatinib at the maximum concentrations, and these inhibitions were dose dependent according to the concentrations of the substrates (Fig. 9C).

FIG 9.

Inhibition of avian HEV attachment and infection of LMH^1A2-GFP and CEH cells. (A) Cytotoxicity analysis of the substrates (or inhibitors) at their maximum working concentrations (50 μM, 50 μM, 10 μM, and 10 μM for CDCA, SC, carvedilol, and imatinib, respectively) for LMH^1A2-GFP cells. (B) Detection of mRNA and protein levels of OATP1A2 expression in LMH^1A2-GFP cells treated with different substrates or an inhibitor. Anti-GFP antibody was used to detect the OATP1A2 fusions for Western blotting. (C) Inhibition of CaHEV attachment to and infection of LMH^1A2-GFP cells by the substrates or inhibitor. After preincubation with CDCA, SC, carvedilol, and imatinib, CaHEV attachment (left) and infection (right) assays were evaluated using LMH^1A2-GFP cells. CaHEV ORF3 RNA was quantified by TaqMan real-time RT-PCR. (D) Blocking CaHEV attachment (left) and infection (right) in LMH^1A2-GFP cells by ap237. (E) Anti-GST-1A2^ecto mouse sera blocked CaHEV attachment to (left) and infection of (right) LMH^1A2-GFP cells. sp239 protein and negative serum were used as controls. CaHEV ORF3 RNA was quantified by TaqMan real-time RT-PCR. Error bars indicate the SEM. (F) Inhibition of CaHEV infection of CEH evaluated using imatinib, ap237, and murine anti-GST-1A2^ecto along with the corresponding dimethyl sulfoxide (DMSO), sp239, and negative mouse sera (**, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant).

Additionally, in the OATP1A2 blocking assay, ap237 was used as the blocking agent and blocked CaHEV attachment to and infection of LMH^1A2-GFP cells in a dose-dependent manner compared to the control protein sp239 (Fig. 9D). When anti-GST-1A2^etco mouse sera were used as a blocking agent, the results also showed that CaHEV attachment to and infection of the cells were blocked when the sera were diluted 1:5 and 1:50 (Fig. 9E). Moreover, CaHEV infection of CEH was significantly inhibited by imatinib, ap237, and mouse anti-GST-1A2^etco sera compared to the corresponding control reagents (Fig. 9F).

Correlation of OATP1A2 expression and avian HEV infection in vivo.

To define the distribution of OATP1A2 in different tissues from chickens inoculated with CaHEV, the OATP1A2 protein was detected by an immunohistochemistry (IHC) assay using mouse anti-GST-1A2^ecto serum along with phosphate-buffered saline (PBS) and negative mouse serum. Positive immunoreactive signals were detected in chicken tissues of liver, brain, spinal cord, kidney, and testis. As shown in Fig. 10A, in chicken liver, a positive signal of OATP1A2 can be seen in cholangiocytes and Kupffer cells (c and f). In the spinal cord, OATP1A2 expression was observed in axons in white matter (Fig. 10Ai) and nerve cell bodies in gray matter (l). In the chicken brain, OATP1A2 was almost exclusively expressed in brain capillary cells (Fig. 10Ao). The localization of OATP1A2 expression in the chicken kidney appeared to be in the apical domain of distal nephrons (Fig. 10Ar). Significant expression of OATP1A2 was also detected in spermatocytes in chicken testis (Fig. 10Au).

FIG 10.

Correlation of OATP1A2 expression and avian HEV infection in vivo. (A) Localization of OATP1A2 protein in chicken tissues detected by an IHC assay. OATP1A2 protein in chicken liver (a to f), spinal cord (g to l), brain (m to o), kidney (p to r), and testis (s to u) was detected using mouse polyclonal antiserum against 1A2^ecto. Positive immunoreactive signals are indicated by arrowheads. PBS and mouse anti-GST serum and were used as negative controls. (B) Relative expression of OATP1A2 mRNA in chicken tissues. Chicken tissues (0.2 g) were homogenized in RNAiso Plus reagent, and total RNA was extracted. The total RNA (2 μl) was used as the template to perform qPCR for OATP1A2 mRNA. The GAPDH gene was used as a reference gene. (C) CaHEV ORF3 RNAs in infected chicken tissues and fecal samples were detected by TaqMan real-time RT-PCR, and the negative-strand CaHEV ORF3 gene was detected by negative-strand-specific RT-PCR. The y axis shows the number of CaHEV ORF3 RNA copies per gram of chicken tissue. The presence and absence of negative-strand CaHEV RNA are indicated by “+” and “−,” respectively.

In addition, to evaluate the agreement of the expression levels of OATP1A2 and CaHEV infection in these chicken tissues, qPCR and TaqMan real-time RT-PCR were performed. First, CaHEV positive-strand RNA was detected in the fecal samples of chickens (Fig. 10C), indicating that the chickens were successfully infected by CaHEV. The mRNA of OATP1A2 showed that the relative amounts in the liver, pancreas, bile duct, testis, kidney, brain, and spinal cord were larger than the ones in the crops, glandular stomach, muscle stomach, duodenum, jejunum, ileum, cecum, rectum, lung, spleen, ovary, thymus, and nerve tissue (Fig. 10B). Meanwhile, in regard to CaHEV RNA, the results showed that the amount of viral RNA was also larger in the liver, kidney, testis, brain, and spinal cord than in the crops, jejunum, rectum, and heart (Fig. 10C). In the tissues positive for positive-strand CaHEV ORF3 RNA, except for crop, heart, and muscle, all other tissues were positive for the negative-strand CaHEV ORF3 RNA. Comparison of the above-described results implied that the expression of OATP1A2 may also be associated with avian HEV infection in vivo.

DISCUSSION

Due to the lack of an efficient cell culture system and animal model, there are few studies about the interaction of HEV with host cells. However, for human HEV, the capsid protein expressed by different systems as the bait protein was used to screen the interaction with host proteins (25, 27–29). For example, p239, a truncated human HEV capsid protein, can form polymers and was used as the bait protein to research the interactions of HEV with host cells (25, 27–29). Following those previous studies about human HEV, ap237 (a truncated capsid protein of CaHEV that corresponds to the amino acid region of human HEV p239) was designed and used as a bait protein to screen the interaction with host proteins from chicken liver cells by a GST pulldown assay in this study.

Among the screened cellular proteins, OATP1A2 (a multiple-transmembrane protein localizing on the cell membrane) is a member of an important superfamily of solute carriers, which is highly expressed in liver tissue based on research on human OATP1A2 (34). In the present study, OATP1A2 was mainly localized in the cell membrane preparations from CEH and LMH^1A2-GFP cells, which contain endogenous OATP1A2 and stably express OATP1A2 protein (Fig. 2C and Fig. 6C). The interaction between ap237 and OATP1A2 proteins was determined in HEK 293T cells (Fig. 3C), and colocalization of OATP1A2 with ap237 was seen in the cytoplasm rather than the membrane, which may be because of the transient expression of these proteins in the cytoplasm. Furthermore, this protein is involved in the biological transportation of endogenous substrates, including bile acid and bile salt, among others (35, 36). In regard to avian HEV, some previous studies suggested that chicken liver is the main target tissue upon viral infection (4, 37), and bile samples of infected chickens contain large numbers of viral RNA copies (38). Based on those previous findings, it was speculated that OATP1A2 may have some relationship with avian HEV infection. Following this, the host protein OATP1A2 was selected for research in the present study. The results of co-IP and indirect ELISAs confirmed that the extracellular region of OATP1A2 directly interacts with the avian HEV capsid protein, and high expression levels of OATP1A2 in LMH cells can enhance avian HEV attachment and infection. It was also observed that the expression levels of OATP1A2 in different chicken tissues exhibit a positive correlation with the amounts of avian HEV RNA. These results confirmed our speculation that OATP1A2 is involved in avian HEV infection of host cells. To our knowledge, this is the first report that OATP1A2 is involved in viral infection.

Heparin surface proteoglycans and asialoglycoproteins are involved in and facilitate HEV infection by binding to ORF2 (27, 28). To date, no cell-specific receptor mediating HEV entry has been identified. Usually, attachment factors and receptors are often difficult to classify in practice because both of them contribute to effective infection. To invade target cells, many viruses use more than one attachment factor and receptor. As a typical example, hepatitis C virus utilizes more than 10 molecules for cell entry (39). In the present study, the fact that the substrates and inhibitor of OATP1A2, ap237, si1A2, and anti-1A2 sera could not completely block avian HEV infection of cells (Fig. 9) suggests that OATP1A2 is not the only essential factor for avian HEV infection. More research is necessary to determine the exact role that OATP1A2 plays in avian HEV infection.

It has been noted that the SLCO1A2 gene (encoding the OATP1A2 protein) is conserved in chickens, humans, chimpanzees, cows, and mice, among others, in the GenBank database (40), but the sequences of OATP1A2 are diverse among different species. We found that the OATP1A2 protein of chickens shared only about 50% amino acid identity with human OATP1A2 in the UniProt database. In this study, the results of the indirect ELISA also showed that OATP1A2 can specifically bind with the avian HEV capsid protein but not with the human, pig, or rabbit HEV capsid protein. Thus, further experiments are necessary to determine whether the OATP1A2 proteins of different species bind with their respective HEV capsid proteins. Furthermore, if they interact among different species, another area of research would be to understand if the binding of OATP1A2 is a result of species tropism or cross-species infection by HEV. In human HEV, soluble ORF2 proteins circulate in HEV-infected patients and are the major antigens in patient sera (41–43). In the present study, the positive reaction of OATP1A2 with avian HEV-positive sera (Fig. 5F) indicates that soluble avian HEV ORF2 proteins also exist in avian HEV-infected chicken sera and that OATP1A2 interacts with natural ORF2s.

In the present study, IHC results showed that the distributions of chicken OATP1A2 in chicken tissues are consistent with those of human OATP1A2 in human tissues (34, 44, 45). For human HEV, some previous studies documented that HEV infection favors Kupffer cells, cholangiocytes, and interstitial lymphocytes in the liver; the cytoplasm of neurons in the gray matter; the perivascular area in the white matter of the spinal cord; and the perivascular area in the brain (46). These results indicate that the distribution of OATP1A2 in human cells correlates with human cells that are targeted by HEV. Moreover, the results of the present study also confirmed that the distribution of chicken OATP1A2 in different tissues is most consistent with tissues that avian HEV infects. However, the negative and positive RNAs of avian HEV were also detected in tissues not expressing OATP1A2, such as jejunum, ileum, and rectum, which may be explained by the finding that the transmembrane protein OATP1A2 is not the only attachment factor or receptor for avian HEV infection, and therefore, further studies are needed to elucidate the OATP1A2 functions and identify other factors required for avian HEV infection.

Substrates and inhibitors of OATP1A2 can significantly reduce CaHEV attachment to and infection of LMH^1A2-GFP cells. One possibility is that the substrates or inhibitors reduce the expression of OATP1A2 (e.g., imatinib and carvedilol) in cells and that this causes a decrease in virus binding with OATP1A2. Another possibility is that the substrates and inhibitors can directly bind to OATP1A2 on the cell surface, which results in a decrease of CaHEV attachment and infection; specific mechanisms need to be further explored to determine this. Notably, these results implied that the substrates or inhibitors of OATP1A2 can be used as therapeutic drugs to prevent or cure HEV infection.

In conclusion, this is the first study to document that the host protein OATP1A2 can directly interact with the avian HEV capsid protein. Also, the expression levels of OATP1A2 in host cells can influence avian HEV attachment to and infection of cells in vitro. Additionally, the distribution of OATP1A2 in chicken tissues positively corresponds with tissues targeted by avian HEV in vivo. In addition, the substrates and inhibitors of OATP1A2 as well as blocking of OATP1A2 with ap237 and anti-1A2^ecto sera can significantly reduce avian HEV attachment and infection. These findings offer new clues to understand the mechanism of avian HEV infection and may provide a theoretical basis and guidance for the prevention of HEV infection, its treatment, and vaccine development.

MATERIALS AND METHODS

Virus and cells.

An avian HEV infectious stock was produced by intravenously inoculating four 8-week-old specific-pathogen-free (SPF) chickens with 200 μl of a clinical bile sample containing avian HEV isolated in China (CaHEV) (GenBank accession no. GU954430) from a 35-week-old broiler breeder chicken in China (14). At 21 days postinoculation, virus-positive fecal samples were collected and suspended in PBS (10 mM; pH 7.2). After centrifugation at 6,200 × g at 4°C for 10 min, the supernatant was incubated with 8% polyethylene glycol 8000 (PEG 8000; Sigma Chemical Co., St. Louis, MO, USA) at 4°C overnight. Following this, the mixture was centrifuged at 13,000 × g for 1 h again, and the pellet was suspended with PBS and filtered with a 0.22-μm filter. This virus stock was stored at −80°C.

Human embryo kidney HEK 293T cells, DF-1 cells, and the LMH cell line were purchased from the American Type Culture Collection (ATCC) and grown in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, Carlsbad, CA, USA), 100 U/ml penicillin (Life Technologies Corp., Grand Island, NY, USA), and 100 μg/ml streptomycin (Life Technologies Corp., Grand Island, NY, USA). All of the cell lines were cultured at 37°C in a humidified atmosphere with 5% CO₂.

Primary cultured chicken embryo hepatocytes (CEH) were isolated from a 19-day-old SPF chicken embryo purchased from Beijing Merial Vital Laboratory Animal Technology Company. Liver tissue was immersed in PBS, minced, washed three times to remove red blood cells, and then trypsinized for 10 to 15 min. After centrifugation, the cell pellet containing chicken hepatocytes was resuspended in growth medium consisting of M199 medium (Gibco, Carlsbad, CA, USA) supplemented with 10% FBS, 5 μg/ml transferrin (Thermo Fisher Scientific, CA, USA), 10 ng/ml epidermal growth factor (EGF; Thermo Fisher Scientific, CA, USA), 40 ng/ml dexamethasone (Acros Organics, Belgium, WI, USA), 3 μg/ml insulin (Gibco, Carlsbad, CA, USA), 2 mM l-glutamine (Gibco, Carlsbad, CA, USA), 100 U/ml penicillin, and 100 μg/ml streptomycin. The cells were counted and plated on collagen-coated cell culture dishes or plates. Cells were maintained in a humidified incubator at 37°C with 5% CO₂. After plating for 12 h, the culture medium was changed, and the regular medium was then changed every 2 to 3 days.

For GST pulldown assays, CEH cells were pelleted by centrifugation at 600 × g at 4°C. Next, the cell pellet was resuspended with PBS, cells were then counted, and tubes were packed with 1 × 10⁷ cells per tube. After centrifugation, cell pellets were resuspended and lysed with NP-40 buffer (Beyotime, Shanghai, China) on ice for 30 min, and the supernatant was then collected by centrifugation and used to perform GST pulldown assays.

Plasmid construction.

To produce the bait protein for the GST pulldown assay, a recombinant plasmid containing the truncated ORF2 gene encoding ap237 was constructed with the pGEX-6P-1 vector (GE Healthcare, NJ, USA) with an N-terminal GST tag as the backbone. Briefly, the truncated ORF2 gene encoding ap237 was amplified with the primer pair GST-ap237-F/GST-ap237-R and cloned into the vector. The positive recombinant plasmid was designated pGEX-6P-1-ap237. In addition, in order to express another form of ap237 without the tag protein, the truncated gene was also amplified with primer pair Ap237-F/Ap237-R and cloned into the pET-21b vector (Novagen, Darmstadt, Germany) for bacterial expression. The recombinant plasmid was named pET-21b-ap237. To generate the eukaryotic expression vector for the coimmunoprecipitation (co-IP) assay, the truncated ORF2 gene was amplified with the primer pair HA-ap237-F/HA-ap237-R and cloned into the pCAGEN vector (47) with an N-terminal HA tag. The positive plasmid was named pCAGEN-HA-ap237.

For interaction with the host protein OATP1A2, two different eukaryotic expression vectors were constructed. First, based on the coding sequence of OATP1A2 (GenBank accession no. XM_416421.6), the primer pair OATP1A2-F/OATP1A2-R was designed and used to amplify the complete coding sequence from the chicken liver cDNA library. Next, for the co-IP assay, the gene was cloned into the p×3Flag-CMV-14 expression vector (Sigma-Aldrich, St. Louis, MO, USA) with C-terminal 3×Flag tags, and the recombinant plasmid was named pCMV-1A2-3Flag. To construct the LMH cell lines with stable expression of OATP1A2, the gene was inserted into the pTrip-puro lentivirus vector. To facilitate the screening of positive cells, OATP1A2 was fused with a GFP coding sequence with a linker sequence (5ʹ-TCCGGCCGGACTCAGATCTCGAGCTCAAGCTTCGAATTCAAA-3ʹ), including 14 amino acids (48), using primers pairs 1A2-F/1A2-linker-R and linker-GFP-F/GFP-R, by overlapping PCR. The PCR product was inserted into the vector, and the positive plasmid was named pTrip-1A2-GFP-puro or pTrip-GFP-puro. For the construction of pTrip-GFP-puro, the GFP sequence was amplified with the primer pair GFP-F/GFP-R and cloned into the pTrip-puro vector. In addition, in order to produce antiserum against OATP1A2 and evaluate the direct interaction between ap237 and OATP1A2, the transmembrane helices of OATP1A2 were predicted by using the TMHMM server (v.2.0) (49). The gene sequences encoding extracellular domains were connected and synthesized by Genewiz, Inc. (Suzhou, China), and ligated into the pGEX-6P-1 vector for expression. The transmembrane region of OATP1A2 was designated 1A2^ecto, and the positive recombinant plasmid was named pGEX-6P-1-1A2^ecto.

All primers used in the study are listed in Table 2.

TABLE 2.

Primers used in this study

^aThe nucleotides of the linker sequence are underlined, and the T7 promoter sequence is in boldface type.

Protein expression in bacteria.

To express GST-ap237 and GST-1A2^ecto in a bacterial system, recombinant plasmids pGEX-6P-1-ap237 and pGEX-6P-1-1A2^ecto were separately transformed into Escherichia coli Transetta(DE3) (TransGen Biotech, Beijing, China). Next, the two proteins were induced with 0.2 mM isopropyl-β-d-thiogalactoside (IPTG) for 20 h at 16°C. After that, the bacteria were sonicated and clarified. The two soluble proteins were filtered with a 0.22-μm filter and purified with the glutathione-Sepharose 4 Fast Flow system (GE Healthcare, Uppsala, Sweden). The expression and purification of the proteins were analyzed by SDS-PAGE and Western blotting using anti-GST antibodies and the 3E8 monoclonal antibody (mAb) (19, 22) against the avian HEV ORF2 protein as the primary antibody for ap237 detection.

The procedure for inducing the expression of ap237 without any tags was based on modified methods described previously by Chen et al. (50). Briefly, the expression of ap237 was induced by adding 1.0 mM IPTG for 6 to 8 h at 37°C. Bacterial cells were harvested and lysed by sonication. The cell pellets were dissolved in 8 M urea, filtered with a 0.22-μm filter, refolded with gradient buffer (6 M urea, 4 M urea, and 2 M urea in PBS), and purified using a Superdex 200 Increase 10/300 GL column (GE Healthcare, NJ, USA) connected to an Äkta purifier (GE Healthcare, NJ, USA). Next, the expression and purification of ap237 were analyzed by SDS-PAGE and Western blotting using the 3E8 mAb as the primary antibody.

The control proteins used in this study were truncated capsid proteins (aa 368 to 606), which included sar239, ker239, r239, and sp239; these were from genotype 1 human HEV (Sar-55; GenBank accession no. AF444002), genotype 3 human HEV (Kernow-C1; GenBank accession no. JQ679013), genotype 3 rabbit HEV (CHN-SX-rHEV; GenBank accession no. KX227751), and genotype 4 swine HEV (CHN-SD-sHEV; GenBank accession no. KE176351), respectively, and were all expressed and purified as previously described by Chen et al. (50). Another negative-control protein, the porcine reproductive and respiratory syndrome virus (PRRSV) N protein, was also expressed and purified according to the same procedures as the ones described above.

GST pulldown assay and mass spectrometry assay.

For the GST pulldown assays, GST or GST-ap237 protein was separately conjugated to glutathione-Sepharose beads and then blocked with 5% bovine serum albumin (BSA) for 1 h. After washing three times with PBS, the two beads were separately incubated for 8 to 12 h at 4°C with the chicken liver cell lysate treated with NP-40 buffer. After washing three times again, the two beads were transferred to a new tube. Finally, the binding proteins were eluted with elution buffer (50 mM Tris-HCl containing 10 mM reduced glutathione [pH 8.0]) and analyzed by SDS-PAGE with silver staining. The specific bands for GST-ap237 were cut, mixed together, and subsequently analyzed by mass spectrometry (MS). MS was performed using the Q Exactive HF Orbitrap liquid chromatography-tandem MS (LC-MS/MS) system at the Research Center for Proteome Analysis, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, China. After acquisition of the raw data, Mascot 2.2 (Matrix Science, MA, USA) (51) was used to retrieve the Gallus gallus UniProt database (52, 53), and the score, coverage, peptides, and peptide spectrum match (PSM) values of all the proteins were obtained.

In addition, a GST pulldown assay was also performed to confirm a direct interaction between OATP1A2 and ap237. The GST or GST-1A2^ecto protein was separately conjugated to glutathione-Sepharose beads and then incubated for 8 to 12 h at 4°C with ap237. After washing, the eluted proteins were analyzed by Western blotting using the 3E8 mAb and anti-GST antibodies as the primary antibodies. The sp239 protein was used as the negative control.

Production of antisera against GST-1A2^ecto.

Five 7-week-old BALB/c mice were immunized intraperitoneally with purified GST-1A2^ecto protein at a dose of 100 μg per mouse, which was emulsified with Freund’s complete adjuvant (Sigma-Aldrich, St. Louis, MO, USA). The mice were boosted in 14-day intervals with the same dose of antigen in Freund’s incomplete adjuvant. To obtain negative serum eliminating the interference of the GST tag antibody, another five mice were immunized with only the GST tag protein according to the same method. Seven days after the second boost, blood was collected at the tail vein. Following this, the titers of anti-GST-1A2^ecto antibodies in the sera were determined with an indirect ELISA.

Coimmunoprecipitation and confocal immunofluorescence assay.

To confirm the interaction of ap237 and OATP1A2, a cell-based co-IP assay was performed as follows. Briefly, the pCAGEN-HA-ap237 and pCMV-1A2-3Flag plasmids were cotransfected into HEK 293T cells for 48 h. Next, the cells were lysed with NP-40 lysis buffer containing 1 mM phenylmethylsulfonyl fluoride (Sigma Chemical Co., St. Louis, MO, USA) on ice for 30 min and clarified by centrifugation at 12,000 × g for 15 min at 4°C. After centrifugation, the clarified supernatant was subjected to immunoprecipitation using Dynabeads protein G (Thermo Fisher Scientific, CA, USA) bound to the anti-HA mAb (ProteinTech, Wuhan, China) or the anti-Flag mAb (Sigma-Aldrich, St. Louis, MO). The immunoprecipitates were then resolved by 10% SDS-PAGE. The bound proteins were detected by Western blotting using the anti-Flag and the anti-HA mAbs. The recombinant plasmid pCAGEN-HA-ORF3 and two empty vectors, pCAGEN-HA and pCMV-3Flag, were used as controls.

In addition, cotransfected HEK 293T cells were also used to perform a confocal immunofluorescence assay for analyzing the colocation of ap237 and OATP1A2 in the cells. After cotransfection for 48 h, washed cells were fixed with 4% paraformaldehyde (Solarbio Life Science, Beijing, China) for 15 min at 37°C and permeabilized with 0.25% Triton X-100 for 15 min at 37°C. After washing three times with PBS buffer, the cells were incubated with 1% bovine serum albumin in PBS for 1 h at 37°C for blocking. Next, the cells were incubated with rabbit anti-Flag polyclonal antibodies (ProteinTech, Chicago, IL, USA) and the anti-HA mAb for 1 h at 37°C. After washing again, the cells were incubated for 1 h with Alexa Fluor 488-conjugated AffiniPure goat anti-mouse IgG(H+L) and Cy3-conjugated goat anti-rabbit IgG(H+L) (Jackson ImmunoResearch, West Grove, PA, USA). Finally, the cells were fixed with Fluoroshield with 4′,6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich, St. Louis, MO, USA) and observed with a Leica SP8 confocal system (Leica, Wetzlar, Germany). The colocalization of ap237 and OATP1A2 was evaluated by the determination of Manders’ overlap coefficient using Image-plus pro software. Manders’ overlap coefficient indicates an actual overlap of the fluorescence signals (54). This value, ranging from 0 to 1.0, implies 0% to 100% colocalization of both selected channels.

Indirect ELISA.

To test the direct interaction of the avian HEV capsid protein and chicken OATP1A2, an indirect ELISA was performed as follows. Ninety-six-well ELISA plates were coated with different dosages of ap237 (8 μg/well, 4 μg/well, 2 μg/well, and 1 μg/well) dissolved in PBS (pH 7.2) and incubated for 12 h at 4°C. After washing, the plates were incubated with blocking buffer consisting of 2.5% dry milk in 0.05% Tween 20 in PBS (PBST) for 1 h at 37°C. Plates were then incubated with different concentrations of GST-1A2^ecto (200 μg/ml, 20 μg/ml, 2 μg/ml, and 0.2 μg/ml) and GST (100 μg/ml, 10 μg/ml, 1 μg/ml, and 0.1 μg/ml) with 100 μl/well for 1 h at 37°C. After washing, the plates were treated with blocking buffer and incubated for 1 h at 37°C. Following this, the anti-GST antibodies at a dilution of 1:1,000 were added to the wells. After incubation for 1 h at 37°C, horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG antibody (Jackson ImmunoResearch, West Grove, PA, USA) was added to the wells at a dilution of 1:5,000. The colorimetric reaction was triggered by the addition of tetramethylbenzidine (Sigma Chemical Co., St. Louis, MO, USA) for 15 min and stopped by the addition of 3 M H₂SO₄. The OD₄₅₀ value was read using an automated ELISA plate reader (Bio-Rad, CA, USA). In order to analyze the interaction of chicken OATP1A2 with capsid proteins of HEVs of the other species, 96-well ELISA plate wells were also coated with sar239, ker239, r239, and sp239 and then incubated with GST-1A2^ecto (20 μg/ml) and GST (10 μg/ml). In addition, the wells were also coated with the PRRSV N protein as the negative-control protein.

To identify the interaction of OATP1A2 with the soluble ORF2 proteins in avian HEV-infected chicken serum, 300 μl positive sera was added to the ELISA plate coated with mAb 3E8 (800 ng/well). After incubation and washing, GST-1A2^ecto and the GST control were added, followed by washing and incubation with rabbit anti-GST antibody. HRP-goat anti-rabbit IgG antibody (Jackson ImmunoResearch, West Grove, PA, USA) was used as the secondary antibody.

Quantitative real-time PCR for quantifying OATP1A2 RNA.

To quantify the expression of OATP1A2 RNA in chicken tissues and different cell lines, the total RNA of each tissue or cell line was extracted using the High Pure RNA isolation kit (Roche, Manheim, Germany). The mRNA levels were quantified using the StepOnePlus real-time PCR system (Applied Biosystems, Foster City, CA, USA) and FastStart universal SYBR green master mix (ROX) (Roche, Manheim, Germany). Specific primers used were designed as previously reported (55) and are shown in Table 2. An endogenous gene, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), was used as the reference gene with the primers listed in Table 2. The mRNA levels were calculated using the 2^−ΔΔC_T method with GAPDH as an endogenous reference.

Establishment of cell lines with stable expression of OATP1A2.

The lentivirus gene transfer system was applied to generate cell lines constitutively expressing recombinant OATP1A2-GFP. A cell line with stable expression of GFP was also established as the control. Briefly, to produce replication-defective lentivirus stocks, HEK 293T cells were cotransfected with 2 μg pTrip-1A2-GFP-puro or 2 μg pTrip-GFP-puro with packing plasmids pFIV-34N (3 μg) (System Biosciences, CA, USA) and pVSV-G (3 μg) (System Biosciences, CA, USA). Three days after transfection, cell supernatants containing recombinant lentivirus were harvested and used to infect target LMH cells. At 48 h postinfection, LMH cells with stable expression of OATP1A2-GFP and those expressing only GFP were selected with 4 μg/ml puromycin. OATP1A2 expression was detected by an indirect fluorescence assay using mouse anti-GST-1A2^ecto sera and tetramethylrhodamine isothiocyanate (TRITC)-conjugated goat anti-mouse IgG (Jackson ImmunoResearch, West Grove, PA, USA). Puromycin-resistant and green cell clones were examined by qPCR to determine the presence of OATP1A2 and GFP. Next, the established cell lines were separately named LMH^1A2-GFP and LMH^GFP. In addition, to determine the expression location of OATP1A2-GFP in the LMH^1A2-GFP cell line, the membrane proteins were extracted for Western blotting to detect the target OATP1A2 using the Mem-PER Plus membrane protein extraction kit (Thermo Scientific, Waltham, MA, USA) according to the manufacturer’s instructions. In addition, the expression and location of OATP1A2 in CEH were also determined with mouse anti-GST-1A2^ecto sera using confocal immunofluorescence and Western blotting.

RNA interference to knock down the expression of OATP1A2.

To knock down the expression of OATP1A2 in the cell line LMH^1A2-GFP, three siRNAs targeting OATP1A2 and a negative control were designed and synthesized by RiboBio Company (Guangzhou, China). The siRNA target sequences were CACAGATAGAGAAGCAATT (si1A2-1), CCAAACAGAGCATGTATGA (si1A2-2), and TGAAAGGAATGTCTCATTA (si1A2-3). LMH^1A2-GFP cells were transfected with the siRNAs using Lipofectamine RNAiMAX reagent (Invitrogen, CA, USA). After transfection, qPCR and Western blot assays were used to analyze the expression of OATP1A2 in transfected cells.

ap237 attachment to different cell lines.

In a previous study, it was documented that ap237 can imitate natural avian HEV particles and attach to LMH cells (30). In this study, the variations of ap237 attachment to differently treated LMH cell lines expressing different levels of OATP1A2 were first analyzed. The different cell lines (LMH^1A2-GFP, LMH, LMH^GFP, and LMH^1A2-GFP transfected with siRNA) were seeded on six-well plates at a concentration of 1.6 × 10⁶ cells/well. After culturing for 24 h (with a confluence of 70% to 90%), the cells were washed gently with cold PBS and incubated with 500 nM ap237 (or 5 μM for the flow cytometry assay) diluted with DMEM for 1 h at 4°C. Finally, the amount of ap237 attached to the cells was determined by Western blot and flow cytometry assays. The mAb 3E8 was used as the primary antibody to target ap237 in the assays.

TaqMan real-time RT-PCR for quantifying avian HEV RNA.

For determining the amount of CaHEV, TaqMan real-time RT-PCR was developed and performed with a 20-μl reaction mixture on the StepOnePlus real-time PCR system (Applied Biosystems, Foster City, CA, USA) using the QuantiTect probe RT-PCR kit (Qiagen, Hilden, Germany). The primers and probe targeting ORF3 were designed as reported previously (56) and are shown in Table 2. The cDNA from RNA was synthesized via reverse transcription for 30 min at 50°C, followed by an PCR initial activation step for 15 min at 95°C. Next, targeting ORF3 was amplified for 45 cycles at 94°C for 15 s and 60°C for 60 s. In addition, to generate the internal control RNA, a 174-bp RT-PCR product was amplified using the primers Taqman-T7f and Taqman-R (Table 2) and then cloned into the pMD19-T vector (TaKaRa Biotech Corporation, Dalian, China). The vectors were linearized using the PstI restriction enzyme (TaKaRa Biotech Corporation, Dalian, China) and purified by using the Wizard DNA clean-up system (Promega, Madison, WI, USA), according to the manufacturer’s protocol. Next, the purified plasmid DNA was transcribed in vitro using a RiboMAX T7 large-scale RNA production system (Promega, Madison, WI, USA). The concentration of RNA was determined at least four times by using a spectrophotometer (BioTek Instruments, Inc., Belgium, WI, USA), and the numbers of copies per microliter were calculated using the mean values and the formula [(grams/microliter RNA)/(length × 340)] × 6.022 × 10²³, where length is the number of nucleotides.

Avian HEV attachment to and infection of differently treated cell lines.

Although there are no highly effective cell culture systems for avian HEV infection, the virus can attach to and infect LMH cells at low levels (32, 57). The relationship between avian HEV infection and OATP1A2 expression levels was also evaluated in the differently treated LMH cell lines; however, the amount of avian HEV was detected only by TaqMan real-time RT-PCR. For the attachment assay, LMH, LMH^1A2-GFP, LMH^GFP, and LMH^1A2-GFP cells transfected with siRNA were seeded on 12-well plates at a concentration of 8 × 10⁵ cells/well and cultured at 37°C with 5% CO₂. When cell confluence was approximately 80%, the cells were inoculated with CaHEV (1 × 10⁶ HEV RNA copies) diluted with DMEM and 2% (wt/vol) FBS. After 2 h of incubation at 4°C, the cells were washed three times with cold PBS and collected using the RNAiso Plus reagent (TaKaRa Biotech Corporation, Dalian, China) to extract the total RNA. For the infection assay, the procedures were similar to those for the attachment assay, except that after the cells were inoculated with CaHEV diluted with DMEM, the cells were cultured at 37°C, and half of the maintenance medium was changed every 2 days. To determine viral replication in the cells, infected cells were collected at various days postinfection. Following this, the total RNA was extracted for quantifying the viral RNA by TaqMan real-time RT-PCR. The negative-strand viral ORF3 RNA was also detected using negative-strand-specific RT-PCR developed previously by Billam et al. (58) with the primer pairs EF1/ER1 and EF2/ER2 (Table 2) for further identification of viral replication in the cells.

Next, the variations in avian HEV attachment to and infection of LMH^1A2-GFP cells were evaluated when the cells were treated with different substrates and an inhibitor of OATP1A2. The inhibitor carvedilol (Adamas, Shanghai, China) and the substrates, including chenodeoxycholic acid (Acros Organics, Belgium, WI, USA), sodium cholate, and imatinib (Sigma Chemical Co., St. Louis, MO, USA), were preincubated with LMH^1A2-GFP cells at 37°C with 5% CO₂ for 24 h. Subsequently, the treated LMH^1A2-GFP cells were inoculated with CaHEV, and the total RNA was extracted in the same manner as described above.

In addition, avian HEV attachment to and infection of LMH^1A2-GFP cells were also analyzed when OATP1A2 in the cells was blocked with ap237 and anti-GST-1A2^ecto mouse sera. The LMH^1A2-GFP cells were first incubated with ap237 and anti-GST-1A2^ecto mouse sera at 37°C with 5% CO₂ for 24 h and then inoculated with CaHEV as described above. sp239 and negative mouse sera were used as the negative-control protein and negative-control sera.

To further confirm that OATP1A2 is involved in avian HEV infection of host cells, CEH cells were treated with imatinib, ap237, and anti-GST-1A2^ecto mouse sera and then inoculated with CaHEV. The total RNA was extracted and used as the template to detect viral RNA copies.

Avian HEV infection of chickens.

Six 8-week-old SPF chickens were purchased from Beijing Merial Vital Laboratory Animal Technology Company. All chickens were negative for avian HEV antibodies and RNA. The chickens were housed together and intravenously inoculated with the same dose of the viral stock (1 × 10⁶ CaHEV RNA copies). At 3 weeks postinoculation, different tissues, including tongue, esophagus, crop, glandular stomach, muscle stomach, duodenum, jejunum, ileum, cecum, rectum, liver, gallbladder, bile duct, pancreas, spleen, heart, lung, kidney, testis (or ovary), brain, spinal cord, nerve, thymus, and muscle, were collected to detect CaHEV and OATP1A2. In addition, fecal samples were collected to detect CaHEV RNA in order to confirm successful infection.

Immunohistochemistry assay.

Healthy paraffin-embedded sections from CaHEV-infected chicken tissues (3 μm) were prepared by Yangling Yike Company. Sections were deparaffinized by heating in an oven at 60°C and rehydrated, and endogenous peroxidase activity was then eliminated using 3% H₂O₂ in deionized water. Sections were immersed in 10 mM sodium citrate (pH 6.0) to heat induce antigen retrieval. Immunochemical staining was performed according to the instructions of the Ready-To-Use SAB-POD (mouse IgG) kit (Boster, Wuhan, China). Briefly, after blocking for 30 min at 37°C with 5% BSA, mouse anti-GST-1A2^ecto serum diluted in blocking buffer (1:100) was added and incubated overnight at 4°C; PBS and mouse anti-GST serum were used as negative controls. Subsequently, sections were washed three times and incubated with biotinylated goat anti-mouse IgG for 30 min at 37°C, followed by rinsing and incubation with SABCs (streptavidin-biotin complexes) at 37°C for 30 min. After washing, the immune reaction was visualized using 3,3'-diaminbenzidine tetrahydrochloride substrate, and nuclei were counterstained with hematoxylin. Finally, dehydration and transparency of sections were performed, and sections were mounted for observation under a microscope.

Computer analysis.

To predict the spatial locations of ap237, the 3D model of the avian HEV capsid protein was built by using SWISS-MODEL (59), using advanced remote-homology detection methods based on the established crystal structure of the HEV capsid protein, deposited under accession no. 2ZTN (23) within the Research Collaboratory for Structural Bioinformatics Protein Data Bank. Following this, the areas of interest were visually highlighted with different colors using CHIMERA software (60). In addition to this, the OATP1A2 structure was predicted using the Phyre2 Web portal (61), based on the d1pw4a template.

Statistical analyses.

The data from qPCR, TaqMan real-time RT-PCR, and cell viability assays were evaluated using GraphPad Prism, version 6 (GraphPad Software, Inc., La Jolla, CA, USA). Statistical differences among the groups were calculated using one-way analysis of variance (ANOVA). P values are indicated in the figures.

Ethics statement.

All animal experiments were performed according to guidance for experimental animal welfare and ethical treatment by the Ministry of Science and Technology of China (www.most.gov.cn/fggw/zfwj/zfwj2006/zf06yw/zf06qt/200612/t20061226_39235.htm). Experimental procedures and animal use and care protocols were carried out in accordance with the guidelines of the Northwest A&F University Institutional Animal Use and Care Committee and were approved by the Committee on Ethical Use of Animals of Northwest A&F University.