A Bicistronic Subgenomic mRNA Encodes both the ORF2 and ORF3 Proteins of Hepatitis E Virus
http://www.100md.com
《病菌学杂志》
Molecular Hepatitis Section, Laboratory of Infectious Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland 20892
ABSTRACT
Hepatitis E virus replicons containing the neomycin resistance gene expressed from open reading frames (ORFs) 2 and 3 were transfected into Huh-7 cells, and stable cell lines containing functional replicons were selected by constant exposure to G418 sulfate. Northern blot analyses detected full-length replicon RNA and a single subgenomic RNA. This subgenomic RNA, which was capped, initiated at nucleotide 5122 downstream of the first two methionine codons in ORF3 and was bicistronic; two closely spaced methionine codons in different reading frames were used for the initiation of ORF3 and ORF2 translation.
INTRODUCTION
Hepatitis E virus (HEV) was discovered in 1983 (3), but molecular characterization did not begin until the first full-length genomic sequence was obtained by Tam et al. (19). However, lack of an efficient cell culture system for this virus has greatly hampered detailed analysis of the viral replication cycle. Therefore, many important questions about this virus remain unanswered.
HEV is the sole member of the Hepeviridae family and of the genus Hepevirus (5). It is a human pathogen that causes hepatitis E, an acute self-limiting disease that does not progress to chronicity. There are four recognized genotypes that infect humans (18): genotypes 1 and 2 are thought to infect humans and nonhuman primates exclusively, whereas genotypes 3 and 4 also infect swine (2, 4). It is thought that hepatitis E may be a zoonosis, but the extent of transmission between animals and humans remains to be determined (14).
The virion is 27 to 30 nm in diameter and does not possess an envelope (16). It most likely is icosahedral and is believed to be composed of a single capsid protein. The genome is a single-stranded, positive-sense RNA molecule of approximately 7.2 kb and is capped. The coding region is preceded by a short noncoding region of 25 nucleotides (nt) and is followed by a noncoding region of 65 nt and a poly(A) tract. The coding region consists of three partially overlapping open reading frames (ORFs). ORF1, consisting of approximately 5 kb, is located at the 5' end and encodes nonstructural proteins involved in RNA synthesis; these include guanylyl transferase, methyl transferase (13), and an RNA-dependent RNA polymerase (1, 9). ORF2, approximately 2 kb, occupies the 3' end of the coding region and encodes the capsid protein. ORF3 is a small reading frame of only 372 bases, with a 5' end that overlaps ORF1 by 4 nt and a 3' end that overlaps ORF2 by 331 nt; ORF3 could encode a protein with a maximum of 123 amino acids. The function(s) of ORF3 has not been fully defined, but it is postulated to interact with the ORF2 protein (22) and with cellular proteins involved in cell signaling (10, 23).
Since HEV does not infect cultured cells efficiently, it has been difficult to determine how expression of the various viral proteins is regulated. Northern blot analyses of liver tissue from infected cynomolgus macaques detected genome-length RNA and two 3'-coterminal RNAs of 2 and 3.7 kb (19). Subsequently, two subgenomic RNAs were also reported to exist in cultured cells infected with a strain of HEV isolated in China (25). Since transfected recombinant full-length genomes are infectious, it is thought that ORF1 of the genomic RNA is translated immediately upon entry into cells to produce the enzymes responsible for viral RNA synthesis. It has been shown that production of ORF2 and ORF3 proteins following transfection of full-length genomes requires a functional viral polymerase, presumably for the synthesis of the subgenomic RNAs that encode them (6). However, the sequences and specificities of these putative RNAs have not been described. Compared to genomes of genotypes 1, 2 and 3, genomes of genotype 4 contain a nucleotide insertion in ORF3 which changes the downstream reading frames so that different methionine codons are believed to initiate translation in both ORF2 and ORF3, and this frameshift is predicted to lengthen the ORF2 protein by 14 amino acids and shorten the ORF3 protein by 9 amino acids (24).
We have recently isolated a number of subclones of Huh-7 cells that permit transfected HEV recombinant genomes to replicate relatively efficiently (S. U. Emerson, unpublished data). Since these transfected cells produce infectious HEV (6), the viral replication cycle in these cells is assumed to approximate the normal in vivo cycle. Therefore, we have used these cells as a model system in which to examine the synthesis of subgenomic RNA.
MATERIALS AND METHODS
Constructs. The infectious cDNA clone of the HEV strain Sar 55, pSK-HEV-2 (GenBank accession no. AF 444002), was used as the parental clone for all mutants. The replicon pHEV/2Neo was constructed by replacing nt 5148 to 5816 of pSK-HEV-2 with the neomycin resistance (neo) gene, utilizing the start codon of HEV ORF2. The replicon expresses HEV ORF1 proteins and neo, but HEV ORF3 is truncated and the remaining partial fragment of HEV ORF2 is out of frame. The neo gene was amplified by PCR from the plasmid pcDNA3.1(+) (Invitrogen) by using a primer pair specific to the 5' and 3' ends of the gene, including a 3'-terminal EcoRI restriction site. The 5' end of the neo gene was extended with nt 3963 to 5147 of pSK-HEV-2, generated by fusion PCR, including a 5'-terminal SfiI restriction site. The resulting fused PCR product was digested with SfiI and EcoRI and substituted into pSK-HEV-2 to yield pHEV/2Neo. The addition of the neo gene increased the genome size to 7.3 kb. The construct pHEV/3Neo, in which the neo gene is placed in frame with ORF3 of HEV, was constructed by replacing nt 5134 to 5816 of pSK-HEV-2 with the neo gene, using the same strategy as that used for pHEV/2Neo. The pHEV/3Neo construct was further modified to contain an extended poly(A) tail of 32 adenosines. Construct pHEV/T+ represents plasmid pSK-HEV-2 mutated in ORF3 by site-directed mutagenesis to insert an extra T after nt 5116. This insertion caused a frameshift which would force translation of ORF3 to initiate only from nt 5131. Construct pHEV5131Ala represents plasmid pSK-HEV-2 mutated at nt 5131 to 5133 by site-directed mutagenesis to change the ATG codon to GCA (Ala). This mutation eliminated the third AUG in ORF3 as an initiation codon for translation. Plasmid pCMV5122 contains a PCR-generated HEV fragment corresponding to nt 5122 to the 3' poly(A) tail of the genome cloned into the directional TOPO expression vector pcDNA3.1D/V5-His-Topo (Invitrogen) according to the manufacturer's instructions. The construct contains the human cytomegalovirus immediate-early promoter. Construct pT75007 contains a PCR-generated HEV fragment corresponding to nt 5007 to the 3' poly(A) tail of pSK-HEV-2 cloned into pCR8/GW/TOPO (Invitrogen) as specified by the manufacturer. The construct contains a T7 promoter prior to the HEV sequences and a unique BglII site following the poly(A) sequences. It was used to produce a size marker for Northern blot analyses. Plasmid pHEV-ORF3/His contains coding sequences for a His tag of six histidine residues inserted into pSK-HEV-2 prior to the stop codon of ORF3.
The entire HEV sequence of each construct was determined after mutagenesis to verify that unwanted mutations had not been introduced.
Cells. S10-3 cells, a subclone of the human hepatoma cell line Huh-7 (15), were derived by limiting dilution; they permit relatively efficient replication of HEV (S. U. Emerson, unpublished data). Cell monolayers of S10-3 cells were grown in Dulbecco's modified Eagle's medium supplemented with 2 mM L-glutamine and 9% fetal bovine serum in a CO2 incubator at 37°C. Transfected cells were maintained in a CO2 incubator at 34.5°C.
Transcription in vitro and transfection of cultured cells. Plasmids were linearized at a unique BglII site located immediately downstream of the poly(A) tract of the HEV sequence. Capped transcripts were synthesized with the T7 Riboprobe in vitro transcription system (Promega) from 4 μg linearized plasmid in the presence of 0.5 mM anti-reverse cap analog [3'-O-methyl-m7G(5')pppG; Ambion], 0.05 mM GTP, 0.5 mM (each) ATP, CTP, and UTP, 10 mM dithiothreitol, 1.6 U of RNasin/μl, and 0.8 U of T7 RNA polymerase/μl. The mixtures were incubated at 37°C for 2.5 h, with an additional 0.4 U of polymerase/μl added after the first 90 min. Transcription was terminated by the addition of 1.25 U of RNase-free RQ DNase/μg of plasmid DNA, followed by incubation for 30 min at 37°C. The RNA was extracted using the RNeasy clean-up protocol (QIAGEN). The concentration was determined by measurement of the optical density at 260 nm, and the integrity of the transcripts was determined by electrophoresis on a nondenaturing agarose gel. Transcribed RNA (1.5 to 3 μg) was mixed with a liposome mixture of 18 μl DMRIE-C (Invitrogen) and 900 μl OptiMEM (Invitrogen) for the transfection of S10-3 cells seeded to 80% confluence in one well of a six-well plate. When DNA was transfected, 1.5 μg of DNA was diluted in 0.5 ml OptiMEM and added to 0.5 ml OptiMEM containing 10 μl DMRIE-C. This transfection mixture was gently mixed and added to S10-3 cells seeded in one well of a six-well plate. In both cases, the cells were incubated at 34.5°C for 5 h, after which the transfection mixture was replaced with Dulbecco's modified Eagle's medium supplemented with 2 mM L-glutamine and 9% fetal bovine serum and the incubation was continued at 34.5°C.
Stable cell line selection. S10-3 cells were grown in the continuous presence of 1 mg/ml G418 sulfate starting 2 days after transfection with capped HEV/Neo transcripts synthesized in vitro. The medium was replenished every 2 to 3 days. Cells were split weekly once the cells surviving selection formed a confluent monolayer. The presence of replicon RNA was determined at 4 weeks posttransfection by reverse transcriptase (RT) PCR with and without reverse transcriptase, using primer sets designed to amplify a region overlapping HEV and neo sequences.
Preparation of biotinylated RNA probes. Plasmid pHEV/2Neo was used as template for PCR to generate DNA fragments that were agarose gel purified using the QIAquick gel extraction kit (QIAGEN). Primers included a T7 RNA polymerase promoter sequence (in capital letters below) to allow in vitro strand-specific transcription of RNA probes. Two DNA fragments for the synthesis of RNA probes detecting plus-strand HEV/2Neo RNA were amplified using primer sets E/neof6583 and E/neor6767T7 (gtagttattcaggattatgacaac and TAATACGACTCACTATAGgtcacagagtcagagacatagac) and E/neof5049 and E/neor5334T7 (ctgagtcagtgaagccagtgcttg and TAATACGACTCACTATAGGcctcgtcctgcagttcattcag). Two micrograms of the purified DNA fragment was transcribed in vitro with the T7 Riboprobe system (Promega) as recommended by the manufacturer. Digestion of the DNA template with RQ DNase I and clean-up of the RNA transcript on RNeasy columns (QIAGEN) followed. The RNA concentration was determined by measurement of the optical density at 260 nm.
RNA transcripts (500 ng) were biotinylated with the Bright Star psoralen-biotin nonisotopic labeling kit (Ambion) according to the instructions of the manufacturer.
Preparation of total RNA and Northern blot analysis. Total cellular RNA was extracted with the RNeasy kit (QIAGEN), including DNase treatment, according to the manufacturer's instructions. Northern blot analysis was conducted with the reagents and conditions specified in the NorthernMax kit (Ambion). Total RNA (10 μg) was denatured with formaldehyde loading dye solution for 15 min at 65°C and separated on a 1% agarose formaldehyde gel. The RNA was transferred to a nylon membrane (Bright Star Plus; Ambion) by capillary blotting. Following transfer, the RNA was cross-linked to the membrane (UV Stratalinker 1800; Stratagene), and the membrane was prehybridized in ultrahybridization solution (NorthernMax; Ambion) at 68°C for 1 h. Hybridization followed at 68°C overnight with the HEV-specific biotinylated riboprobe added to the ultrahybridization solution. The membrane was washed, and the bound biotinylated probe was detected with the Bright Star BioDetect kit (Ambion) as recommended by the manufacturer. The membrane was exposed to BioMax MS film (Kodak).
Determination of the 5'-terminal sequence of subgenomic RNA. Total RNA extracted from the HEV/2Neo or HEV/3Neo replicon-transfected cells was used to amplify the 5'-terminal sequence of the subgenomic RNA with the FirstChoice RNA ligase-mediated rapid amplification of cDNA ends (RLM-RACE) kit (Ambion) according to the manufacturer's instructions. Briefly, 10 μg of total RNA was treated with calf intestine alkaline phosphatase to remove 5'-terminal phosphates. Following RNA purification with TRIzol LS reagent (Invitrogen), one-half of the starting material was treated with tobacco acid pyrophosphatase (TAP) for 1 h at 37°C. One-fifth of this reaction mixture or 1 μl of RNA without TAP treatment was incubated with a 45-nt-long RNA adapter oligonucleotide and T4 RNA ligase at 37°C for 1 h. One-fifth of this reaction mixture was used in a reverse transcription reaction with random primer and M-MLV RT at 42°C for 1 h, followed by HEV-specific nested PCR with Super Taq polymerase (Ambion) and forward primer corresponding to the adapter sequence and reverse primer specific to HEV replicon HEV/2Neo or HEV/3Neo sequences. HEV/2Neo-specific reverse outer primer E/neor5460 (TACTTTCTCGGCAGGAGCAAGGTG) and reverse inner primer E/neor5334 (CCTCGTCCTGCAGTTCATTCAG) or E/neor5172 (CTGCGTGCAATCCATCTTGTTCAATCATGGTCGCGAACCCATGGGC) were used to amplify the 5'-terminal region of subgenomic RNA. The 5'-terminal sequence of the genome-length RNA was determined in parallel as a positive control and was amplified with HEV-specific outer primer Er343 (CAGCGGTGGACCACATTAGGATTG) and the adapter-specific outer primer, followed by inner HEV primer Er194 (AACAAGCTGGCGAGGTTGCATTAG) and the adapter-specific inner primer. The PCR product was gel purified using the QIAquick gel extraction kit (QIAGEN), and the consensus sequence of the purified PCR product was determined. In addition, the PCR product was cloned with the TOPO PCR cloning kit (Invitrogen), and several clones were sequenced.
Western blot analysis. Transfected S10-3 cells were lysed with cell lysis buffer (Promega) containing HALT protease inhibitor cocktail (EDTA free; Pierce) or with sodium dodecyl sulfate (SDS) gel loading buffer with -mercaptoethanol and proteinase inhibitor mix (Complete; Roche) and sonicated on ice three times for 1 min. Proteins were separated by SDS-polyacrylamide gel electrophoresis (PAGE) in a 10 to 20% or 16% Novex Tricine gel or 7% NuPAGE Tris-acetate gel (Invitrogen). Proteins were diluted in the appropriate sample buffer and NuPAGE sample reducing agent (Invitrogen), denatured for 10 min at 90°C, and electrophoretically transferred onto a polyvinylidene difluoride (PVDF) membrane (0.2 μm; Invitrogen) or nitrocellulose membrane (0.45 μm) for analysis by Western blotting. The PVDF membrane was pretreated for 3 min in methanol and 5 min in Novex Tris-glycine transfer buffer (Invitrogen). PVDF membranes were then blocked with 3% nonfat milk and 0.25% Tween 20 in Tris-buffered saline, and nitrocellulose membranes were blocked with StartingBlock blocking buffer (Pierce) with 0.5% Tween 20 overnight at 4°C, incubated with the indicated antibody for 1 h at room temperature, followed by an anti-rabbit or anti-human horseradish peroxidase-conjugated secondary antibody (1:50,000; Jackson ImmunoResearch), and visualized by a Visualizer Western blot detection kit (Upstate, Lake Placid, NY) as specified by the manufacturer.
Immunofluorescence microscopy. Transfected cells were briefly washed in 1x phosphate-buffered saline and then fixed with acetone for indirect immunofluorescence microscopy as described previously (6). Following fixation, the cells were stained with a mixture of an ORF2-specific convalescence-phase serum collected from an HEV-infected chimpanzee (Ch1313) (6) and rabbit anti-ORF3 polyclonal antibody (8) as primary antibody, followed by incubation with a mixture of the specific secondary antibodies Alexa 488-conjugated goat anti-human immunoglobulin G (Molecular Probes) and Alexa 568-conjugated goat anti-rabbit immunoglobulin G (Molecular Probes). Fluorescent cells were visualized with a Zeiss Axioscope 2 Plus fluorescence photomicroscope.
RESULTS
Establishment of a cell line persistently infected with an HEV replicon. Initial attempts to identify genomic or subgenomic RNA in cells transfected with infectious RNA genomes transcribed in vitro were frustrated by a combination of an insufficient number of transfected cells, inefficient viral replication, and a high background of residual input RNA. In order to overcome these obstacles, we sought to establish cell cultures in which every cell contained replicating viral genomes and excess input viral RNA was eliminated. This was accomplished by transfecting cells with an HEV replicon (HEV/2Neo) consisting of a replication-competent genome in which the ORF2 capsid protein gene was replaced by the neomycin resistance gene (neo). Exposure of the transfected cells to G418 sulfate killed all those cells not expressing neo; the expansion of surviving cells in the continuous presence of G418 sulfate diluted out excess input RNA and ensured that all cells contained a functioning replicon.
Identification of genomic and subgenomic viral RNA. If the ORF2 protein is encoded by a subgenomic RNA, as has been suggested (19), the neo gene occupying ORF2 of the replicon should also be expressed from a subgenomic RNA. To determine whether subgenomic RNAs were indeed present in these cells, we first performed Northern blot analyses of RNA extracted from three independently expanded G418 sulfate-resistant cell lines (s2N1, s2N2, and s2N5). The probe used was specific for the 3'-terminal region of the genome (HEV/2Neo, nt 6583 to 6767) and should detect HEV/2Neo genomic RNA (7.3 kb) and any 3'-terminal subgenomic RNAs (Fig. 1A). For each of the three cell lines, two RNA species of approximately 7.3 kb and 2.2 kb were detected (Fig. 1B). No other RNA species were detected even after prolonged exposure of the blot. Two independent G418 sulfate-resistant cell lines expanded from another transfection (s2N8-1 and s2N8-2) were blotted against another probe specific for the overlap region of the three viral ORFs (HEV/2Neo, nt 5049 to 5334). This probe also detected only two RNA species, and they appeared to be identical in size to those detected by the first probe (Fig. 1C). The identification of a subgenomic RNA in each of the five cell lines supported the contention that the ORF2 gene was translated from a subgenomic RNA, but the results failed to support the conclusion that a second subgenomic RNA encodes ORF3 products.
Attempts to detect a minus strand of either genomic or subgenomic RNA failed, most likely because the concentration of the minus strand was too low.
The 5' terminus of the subgenomic RNA is downstream of the proposed translation initiation codon of ORF3. Since the viral genome is capped, it seemed probable that subgenomic RNAs would also be capped. Therefore, in order to determine the 5' sequence of the subgenomic RNA, 5' RACE was performed, utilizing a procedure specific for capped RNA. In this procedure, total RNA is treated with calf intestine alkaline phosphatase to remove 5'-terminal phosphates and thus prevent ligation to an adapter molecule. A 5' monophosphate is then exposed by cleaving the GpppX cap with TAP, and this 5' phosphate is then ligated to an oligonucleotide adapter. Finally, RT-PCR is performed with a forward primer specific for the adapter and a reverse primer specific for a selected HEV sequence.
As positive controls, a reverse primer complementary to nt 171 to 194 (Er194) of HEV was chosen to amplify the 5' sequence of full-length genomic RNA from two of the cell lines (s2N5 and s2N8-1) as positive controls. A single product of the expected size (about 200 bp) was detected for each cell line following agarose gel electrophoresis (data not shown). The consensus sequence of this product was identical to that of the 5' terminus of the original cDNA clone of the replicon, thus validating the assay (data not shown).
In order to maximize the probability of identifying two subgenomic RNAs of slightly different size or sequence, we designed two different reverse primers corresponding to two regions (Er5172, nt 5127 to 5172; and Er5334, nt 5316 to 5334) 181 nt apart and downstream of the first methionine codon in ORF3. Again, the results from analyses of the two cell lines (s2N5 and s2N8-1) were identical. Primer Er5172 amplified a PCR product of about 100 bp, whereas primer Er5334 amplified a product of about 250 bp (Fig. 2A). The negative controls corresponding to each sample were performed with the same extracted total RNA, however without TAP treatment. We obtained PCR products that differed in size from one cell line to the other (Fig. 2A) and from the uniform PCR product obtained by including TAP. The PCR products of the negative controls most likely reflect products amplified from randomly nicked RNA.
The 5' sequence obtained for either the approximately 100-bp or the approximately 250-bp products originated at nt 5122 of the HEV genome (Fig. 2B, examples shown for cell lines s2N8-1), 25 nt upstream of the predicted translation initiation codon for the ORF2-encoded protein but 18 nt downstream of that predicted for the ORF3-encoded protein. As Fig. 2B demonstrates, there were no nucleotide insertions or deletions or minor species. The same sequence was obtained when total RNA from another cell line, Huh-7.5, stably transfected with HEV/2Neo replicon transcripts was analyzed by 5' RLM-RACE (data not shown). These results suggested that there was only a single subgenomic RNA, that it was indeed capped, and that it encoded ORF2-expressed proteins. However, ORF3 protein expression could not be accounted for if it involved initiation at either of the first two methionine codons in ORF3.
The first in-frame AUG of ORF3 is not used to initiate translation. We attempted to determine the amino terminus of the ORF3 protein by direct chemical sequencing. RNA transcribed from a mutant replicon encoding an NH2-ORF3/6His-COOH fusion protein (HEV-ORF3/His) was transfected into S10-3 cells, and the fusion protein was purified from cell lysate by nickel-agarose chromatography. Although a band corresponding to the ORF3 protein was detected by staining with Simply Blue reagent (Invitrogen) or by Western blotting following SDS-PAGE and transfer to PVDF membrane, two independent attempts to determine the N-terminal sequence by Edman degradation failed, suggesting that the amino terminus was blocked (data not shown). Therefore, we resorted to an indirect method to determine which methionine codon might be used to initiate ORF3 protein production.
An additional T residue was inserted after nt 5116 of our infectious cDNA clone of a genotype 1 strain to generate pHEV/T+ and mimic the gene structure in this region of genotype 4 strains. The resulting frameshift should force translation of the ORF3 protein to initiate at the third AUG codon. Capped in vitro transcripts of this plasmid were infectious for S10-3 cells, and both ORF2 and ORF3 proteins were detected by immunofluorescence microscopy (Fig. 3A). Infectious virus was recovered from cells transfected with HEV/T+ (S. U. Emerson, unpublished data), as would be predicted based on the viability of genotype 4 strains.
ORF3 protein produced by the HEV/T+ mutant should contain 114 amino acids compared to the 123 predicted if initiation occurred at the first in-frame AUG of the genotype 1 wild-type parent. Western blot analysis was performed to compare the sizes of the ORF3 proteins produced following transfection of wild-type and HEV/T+ genomes. The ORF3 protein extended by six histidine residues was used as a size marker. The ORF3 proteins produced by the wild type and the HEV/T+ mutant comigrated on SDS-PAGE and migrated slightly faster than the ORF3/His fusion protein (Fig. 3B). These results suggested that the first AUG codon previously assumed to serve as the initiation codon for ORF3 protein synthesis was not used for this purpose in the genotype 1 strain but rather that the third in-frame AUG codon was used.
Subgenomic RNAs expressing ORF2- and ORF3-encoded proteins have the same 5' terminus. Although a single subgenomic RNA was produced when neo was expressed from ORF2, it could be argued that a subgenomic RNA expressing ORF3 protein would not have been under the same selection pressure and might have been lost. In order to examine this possibility, we constructed a second replicon almost identical to the first, except that neo was inserted into ORF3. Cell clones were selected as described before, and cap-dependent 5' RACE was performed with the same outer primers used for the ORF2 analysis and with inner primer Er5334. Not only did the PCR product display a mobility similar to those shown in Fig. 2A (data not shown), but sequence analysis demonstrated that the 5' sequence of the subgenomic RNA was identical to that obtained previously (Fig. 4). Therefore, regardless of whether the cell lines were selected by pressure on ORF2 or ORF3 gene expression, the subgenomic RNA initiated at nt 5122 of the parent genome.
Mutation of the third AUG in the 5' terminus of ORF3 abolishes ORF3 protein synthesis. Identification of nt 5122 as the 5' terminus of the subgenomic RNA encoding neo in frame with either ORF2 or ORF3 implied that the third AUG in ORF3 was the only start codon available to initiate translation of the ORF3 protein. In order to confirm these findings, we mutated the third methionine codon in ORF3 of our infectious cDNA clone to an alanine codon. Capped in vitro transcripts of the pHEV5131Ala construct transfected into S10-3 cells expressed HEV ORF2 proteins (Fig. 5a). However, indirect immunofluorescence microscopy of the HEV5131Ala-transfected cells failed to detect HEV ORF3-encoded proteins (Fig. 5b). As a positive control, cells transfected in parallel with transcripts of the parental infectious HEV cDNA contained both ORF2 and ORF3 proteins as previously reported (6; data not shown). The lack of ORF3 protein expression in HEV5131Ala-transfected cells confirmed that the third AUG in ORF3 of the HEV genome is responsible for ORF3 protein synthesis.
A recombinant mRNA lacking the first two methionine codons in ORF3 encodes both ORF2 and ORF3 proteins. As formal confirmation that ORF2 and ORF3 proteins both could be translated from a single mRNA, a cytomegalovirus promoter-driven plasmid (pCMV5122) was constructed to express a subgenomic RNA [nt 5122 to the poly(A) tail] analogous to the HEV/2Neo and HEV/3Neo subgenomic RNAs but encoding authentic HEV proteins. As expected, transient expression of this plasmid following transfection of S10-3 cells resulted in the synthesis of both HEV ORF2 and HEV ORF3 proteins within the same cell, as detected by immunofluorescence microscopy (Fig. 6A). Western blot analyses with an ORF2 protein-specific polyclonal serum (Fig. 6B) and with a rabbit anti-ORF3 protein polyclonal antibody (Fig. 6C) (8) were performed on cell lysates. Both ORF2 and ORF3 proteins were detected, and their mobility on SDS-PAGE was indistinguishable from that of the two proteins expressed from the full-length HEV genomes in other experiments.
DISCUSSION
The existence of the two 3'-coterminal subgenomic viral RNAs identified in the liver of macaques infected with HEV (19) has been difficult to confirm in the absence of an efficient cell culture system. Although transfected HEV recombinant genomes are able to replicate and produce infectious virus in Huh-7 cells (6), the level of replication has not been sufficient to permit detection of de novo-synthesized viral RNA. We have overcome this problem by generating stable transformed cell lines that were selected for their ability to survive continuous treatment with G418 sulfate due to the expression of the neomycin resistance gene (neo) from either ORF2 or ORF3 of an HEV replicon. It should be emphasized that the Huh-7 cell transfection system is as close to in vivo macaque models as is presently possible. The Huh-7 cells are liver cells of human origin, and production of ORF2 and ORF3 proteins in these cells requires a functional RNA-dependent RNA polymerase, which most likely reflects the need for synthesis of subgenomic mRNA. Also, authentic HEV, which is infectious for macaques, has been isolated from these cells following transfection (6).
Analyses of the viral RNA in cultures of multiple cell lines demonstrated that the full-length replicon transfected into the cells was maintained during expansion of the culture and, therefore, was replicating and that a single subgenomic RNA was also produced and maintained. These results are consistent with the report of an approximately 2-kb subgenomic RNA found in the liver of infected macaques. However, we found absolutely no indication of the 3.7-kb, second subgenomic RNA that was reported to be in the infected liver (19) or that was detected in cells infected with a Chinese strain of HEV (25). At the moment, we can only speculate as to the origin and function of this larger subgenomic RNA, but one intriguing explanation is that it is a defective virus similar to those commonly produced by RNA viruses.
Our results demonstrated that the "different" strategy proposed for translation of ORF3 proteins of genotype 4 viruses compared to that of all other genotypes (24) actually describes the process also used by genotype 1 viruses; the presumed difference was based on the assumption that the first methionine codon in ORF3 was used to initiate translation. However, our data demonstrated that neither the first nor the second AUG of ORF3 was even present in the subgenomic RNA expressing the ORF3 protein and that substitution of the third methionine codon in ORF3 by an alanine codon abolished the expression of the ORF3 protein. Although we have examined a genotype 1 strain only, it is not unreasonable to assume that genotype 2 and 3 strains will utilize a similar expression strategy and that each mammalian strain of HEV will produce ORF2 and ORF3 proteins very similar in size to those of any other genotype.
In the scanning model proposed by Kozak (11), translation of capped RNAs involves binding of the 40S ribosomal subunit to the 5' end of mRNA, followed by scanning in the 3' direction until an AUG codon in the appropriate context for initiation is reached. In most cases, if the first in-frame AUG is not in the optimal context, a proportion of the scanning complexes may fail to initiate here and continue scanning. Our studies provide compelling evidence for a bicistronic RNA in which two different reading frames are utilized, depending on which of the two closely spaced AUG codons are selected by the ribosomes for the initiation of translation.
Although apparently rare, a similar phenomenon involving two closely spaced AUG codons was described previously for the S1 gene of reovirus (17). In that case, the choice of a particular initiation codon was thought to reflect the degree of homology to the Kozak consensus sequence as well as the distance from the cap. It was also proposed that the use of a bicistronic mRNA enabled reovirus to regulate the expression of the cognate proteins at the level of elongation (7).
The discovery that the HEV subgenomic RNA initiates at nt 5122 eliminates a number of potential inconsistencies. First, it greatly reduces the likelihood that ORF2 and ORF3 proteins of genotype 4 strains differ in size from those of the other three genotypes. Also, it provides an explanation for why the first two AUG codons in ORF3 are in an unfavorable context for initiation according to the Kozak model (data not shown): they are not used for translation since they are excluded from the subgenomic RNA.
It is interesting to note that 12 nucleotides at the start of ORF3 are almost universally conserved. Evidence was previously presented that this sequence represented a cis-acting replication element (8). Mutations, even silent ones, introduced into this region of ORF3 unexpectedly abolished the synthesis of the ORF2 protein as well as that of the ORF3 protein in cells transfected with the full-length genome. This result, coupled with the current demonstration that both ORF2 and ORF3 proteins are produced from a single subgenomic RNA that begins downstream of this region, suggests that this highly conserved sequence is part of, or is, the promoter for subgenomic RNA synthesis. Significantly, this sequence is totally conserved across all four mammalian genotypes.
It is intriguing that the translation initiation codon for the ORF3-encoded protein is before that of the ORF2 protein in a bicistronic mRNA. Since the first methionine codon of capped RNAs is the one most often used for initiation, one might expect that this placement evolved to ensure a high level of ORF3 protein production. Observation of the sequence indicated that the Kozak consensus sequences for the initiation codons in ORF2 and ORF3 were equally optimal; therefore, ORF3 protein might be translated as efficiently as ORF2 protein. Knowledge of the molar ratios of ORF2 protein to ORF3 protein in infected cells could prove informative as to what function(s) ORF3 protein performs. Unfortunately, HEV replication in our cell lines is still too low for us to quantify ORF2 and ORF3 proteins in cells.
Since ORF2 protein is the major, if not only, capsid protein, it must be relatively abundant. It is difficult to imagine a comparable need for the ORF3 protein since there is no evidence that it is a major component of the virion; virus-like particles are made in its absence in a baculovirus expression system (12, 21), and antibody to it does not neutralize virus infectivity (S. U. Emerson, unpublished data) (20). Resolution must await more-efficient cell culture systems that will allow the functions of ORF2 and ORF3 proteins to be more fully defined. In the meantime, the discovery of a unique strategy for the synthesis of ORF2 and ORF3 proteins from a single subgenomic RNA provides further evidence that HEV belongs in its own family rather than in the Caliciviridae family, where it was originally classified. It also demonstrates that there is much more to learn about this interesting human pathogen.
ACKNOWLEDGMENTS
This work was supported by the Intramural Research Program of the NIH, NIAID.
We thank Mark Garfield of the Research Technologies Branch of NIH, NIAID, for performing the Edman degradation. We are grateful to Charles M. Rice (Rockefeller University) for providing the Huh-7.5 cell line.
REFERENCES
Agrawal, S., D. Gupta, and S. K. Panda. 2001. The 3' end of hepatitis E virus (HEV) genome binds specifically to the viral RNA-dependent RNA polymerase (RdRp). Virology 282:87-101.
Arankalle, V. A., L. P. Chobe, M. V. Joshi, M. S. Chadha, B. Kundu, and A. M. Walimbe. 2002. Human and swine hepatitis E viruses from Western India belong to different genotypes. J. Hepatol. 36:417-425.
Balayan, M. S., A. G. Andjaparidze, S. S. Savinskaya, E. S. Ketiladze, D. M. Braginsky, A. P. Savinov, and V. F. Poleschuk. 1983. Evidence for a virus in non-A, non-B hepatitis transmitted via the fecal-oral route. Intervirology 20:23-31.
Cooper, K., F. F. Huang, L. Batista, C. D. Rayo, J. C. Bezanilla, T. E. Toth, and X. J. Meng. 2005. Identification of genotype 3 hepatitis E virus (HEV) in serum and fecal samples from pigs in Thailand and Mexico, where genotype 1 and 2 HEV strains are prevalent in the respective human populations. J. Clin. Microbiol. 43:1684-1688.
Emerson, S. U., D. Anderson, A. Arankalle, X. J. Meng, M. Purdy, G. G. Schlauder, and S. Tsarev. 2004. Hepevirus. Elsevier/Academic Press, London, United Kingdom.
Emerson, S. U., H. Nguyen, J. Graff, D. A. Stephany, A. Brockington, and R. H. Purcell. 2004. In vitro replication of hepatitis E virus (HEV) genomes and of an HEV replicon expressing green fluorescent protein. J. Virol. 78:4838-4846.
Fajardo, J. E., and A. J. Shatkin. 1990. Translation of bicistronic viral mRNA in transfected cells: regulation at the level of elongation. Proc. Natl. Acad. Sci. USA 87:328-332.
Graff, J., H. Nguyen, C. Yu, W. R. Elkins, M. St. Claire, R. H. Purcell, and S. U. Emerson. 2005. The open reading frame 3 gene of hepatitis E virus contains a cis-reactive element and encodes a protein required for infection of macaques. J. Virol. 79:6680-6689.
Koonin, E. V., A. E. Gorbalenya, M. A. Purdy, M. N. Rozanov, G. R. Reyes, and D. W. Bradley. 1992. Computer-assisted assignment of functional domains in the nonstructural polyprotein of hepatitis E virus: delineation of an additional group of positive-strand RNA plant and animal viruses. Proc. Natl. Acad. Sci. USA 89:8259-8263.
Korkaya, H., S. Jameel, D. Gupta, S. Tyagi, R. Kumar, M. Zafrullah, M. Mazumdar, S. K. Lal, L. Xiaofang, D. Sehgal, S. R. Das, and D. Sahal. 2001. The ORF3 protein of hepatitis E virus binds to Src homology 3 domains and activates MAPK. J. Biol. Chem. 276:42389-42400.
Kozak, M. 1989. The scanning model for translation: an update. J. Cell Biol. 108:229-241.
Li, T. C., Y. Yamakawa, K. Suzuki, M. Tatsumi, M. A. Razak, T. Uchida, N. Takeda, and T. Miyamura. 1997. Expression and self-assembly of empty virus-like particles of hepatitis E virus. J. Virol. 71:7207-7213.
Magden, J., N. Takeda, T. Li, P. Auvinen, T. Ahola, T. Miyamura, A. Merits, and L. Kaariainen. 2001. Virus-specific mRNA capping enzyme encoded by hepatitis E virus. J. Virol. 75:6249-6255.
Meng, X. J. 2000. Novel strains of hepatitis E virus identified from humans and other animal species: is hepatitis E a zoonosis J. Hepatol. 33:842-845.
Nakabayashi, H., K. Taketa, K. Miyano, T. Yamane, and J. Sato. 1982. Growth of human hepatoma cell lines with differentiated functions in chemically defined medium. Cancer Res. 42:3858-3863.
Purcell, R. H., and S. U. Emerson. 2001. Hepatitis E virus, p. 3051-3061. In D. Knipe, P. Howley, D. Griffin, R. Lamb, M. Martin, B. Roizman, and S. Straus (ed.), Fields virology, 4th ed. Lippincott Williams and Wilkins, Philadelphia, Pa.
Sarkar, G., J. Pelletier, R. Bassel-Duby, A. Jayasuriya, B. N. Fields, and N. Sonenberg. 1985. Identification of a new polypeptide coded by reovirus gene S1. J. Virol. 54:720-725.
Schlauder, G. G., and I. K. Mushahwar. 2001. Genetic heterogeneity of hepatitis E virus. J. Med. Virol. 65:282-292.
Tam, A. W., M. M. Smith, M. E. Guerra, C. C. Huang, D. W. Bradley, K. E. Fry, and G. R. Reyes. 1991. Hepatitis E virus (HEV): molecular cloning and sequencing of the full-length viral genome. Virology 185:120-131.
Tam, A. W., R. White, P. O. Yarbough, B. J. Murphy, C. P. McAtee, R. E. Lanford, and T. R. Fuerst. 1997. In vitro infection and replication of hepatitis E virus in primary cynomolgus macaque hepatocytes. Virology 238:94-102.
Tsarev, S. A., T. S. Tsareva, S. U. Emerson, A. Z. Kapikian, J. Ticehurst, W. London, and R. H. Purcell. 1993. ELISA for antibody to hepatitis E virus (HEV) based on complete open-reading frame-2 protein expressed in insect cells: identification of HEV infection in primates. J. Infect. Dis. 168:369-378.
Tyagi, S., H. Korkaya, M. Zafrullah, S. Jameel, and S. K. Lal. 2002. The phosphorylated form of the ORF3 protein of hepatitis E virus interacts with its non-glycosylated form of the major capsid protein, ORF2. J. Biol. Chem. 277:22759-22767.
Tyagi, S., M. Surjit, A. K. Roy, S. Jameel, and S. K. Lal. 2004. The ORF3 protein of hepatitis E virus interacts with liver-specific alpha1-microglobulin and its precursor alpha1-microglobulin/bikunin precursor (AMBP) and expedites their export from the hepatocyte. J. Biol. Chem. 279:29308-29319.
Wang, Y., H. Zhang, R. Ling, H. Li, and T. J. Harrison. 2000. The complete sequence of hepatitis E virus genotype 4 reveals an alternative strategy for translation of open reading frames 2 and 3. J. Gen. Virol. 81:1675-1686.
Xia, X., R. Huang, and D. Li. 2000. [Studies on the subgenomic RNAs of hepatitis E virus.] Wei Sheng Wu Xue Bao 41:622-627.(Judith Graff, Udana Toria)
ABSTRACT
Hepatitis E virus replicons containing the neomycin resistance gene expressed from open reading frames (ORFs) 2 and 3 were transfected into Huh-7 cells, and stable cell lines containing functional replicons were selected by constant exposure to G418 sulfate. Northern blot analyses detected full-length replicon RNA and a single subgenomic RNA. This subgenomic RNA, which was capped, initiated at nucleotide 5122 downstream of the first two methionine codons in ORF3 and was bicistronic; two closely spaced methionine codons in different reading frames were used for the initiation of ORF3 and ORF2 translation.
INTRODUCTION
Hepatitis E virus (HEV) was discovered in 1983 (3), but molecular characterization did not begin until the first full-length genomic sequence was obtained by Tam et al. (19). However, lack of an efficient cell culture system for this virus has greatly hampered detailed analysis of the viral replication cycle. Therefore, many important questions about this virus remain unanswered.
HEV is the sole member of the Hepeviridae family and of the genus Hepevirus (5). It is a human pathogen that causes hepatitis E, an acute self-limiting disease that does not progress to chronicity. There are four recognized genotypes that infect humans (18): genotypes 1 and 2 are thought to infect humans and nonhuman primates exclusively, whereas genotypes 3 and 4 also infect swine (2, 4). It is thought that hepatitis E may be a zoonosis, but the extent of transmission between animals and humans remains to be determined (14).
The virion is 27 to 30 nm in diameter and does not possess an envelope (16). It most likely is icosahedral and is believed to be composed of a single capsid protein. The genome is a single-stranded, positive-sense RNA molecule of approximately 7.2 kb and is capped. The coding region is preceded by a short noncoding region of 25 nucleotides (nt) and is followed by a noncoding region of 65 nt and a poly(A) tract. The coding region consists of three partially overlapping open reading frames (ORFs). ORF1, consisting of approximately 5 kb, is located at the 5' end and encodes nonstructural proteins involved in RNA synthesis; these include guanylyl transferase, methyl transferase (13), and an RNA-dependent RNA polymerase (1, 9). ORF2, approximately 2 kb, occupies the 3' end of the coding region and encodes the capsid protein. ORF3 is a small reading frame of only 372 bases, with a 5' end that overlaps ORF1 by 4 nt and a 3' end that overlaps ORF2 by 331 nt; ORF3 could encode a protein with a maximum of 123 amino acids. The function(s) of ORF3 has not been fully defined, but it is postulated to interact with the ORF2 protein (22) and with cellular proteins involved in cell signaling (10, 23).
Since HEV does not infect cultured cells efficiently, it has been difficult to determine how expression of the various viral proteins is regulated. Northern blot analyses of liver tissue from infected cynomolgus macaques detected genome-length RNA and two 3'-coterminal RNAs of 2 and 3.7 kb (19). Subsequently, two subgenomic RNAs were also reported to exist in cultured cells infected with a strain of HEV isolated in China (25). Since transfected recombinant full-length genomes are infectious, it is thought that ORF1 of the genomic RNA is translated immediately upon entry into cells to produce the enzymes responsible for viral RNA synthesis. It has been shown that production of ORF2 and ORF3 proteins following transfection of full-length genomes requires a functional viral polymerase, presumably for the synthesis of the subgenomic RNAs that encode them (6). However, the sequences and specificities of these putative RNAs have not been described. Compared to genomes of genotypes 1, 2 and 3, genomes of genotype 4 contain a nucleotide insertion in ORF3 which changes the downstream reading frames so that different methionine codons are believed to initiate translation in both ORF2 and ORF3, and this frameshift is predicted to lengthen the ORF2 protein by 14 amino acids and shorten the ORF3 protein by 9 amino acids (24).
We have recently isolated a number of subclones of Huh-7 cells that permit transfected HEV recombinant genomes to replicate relatively efficiently (S. U. Emerson, unpublished data). Since these transfected cells produce infectious HEV (6), the viral replication cycle in these cells is assumed to approximate the normal in vivo cycle. Therefore, we have used these cells as a model system in which to examine the synthesis of subgenomic RNA.
MATERIALS AND METHODS
Constructs. The infectious cDNA clone of the HEV strain Sar 55, pSK-HEV-2 (GenBank accession no. AF 444002), was used as the parental clone for all mutants. The replicon pHEV/2Neo was constructed by replacing nt 5148 to 5816 of pSK-HEV-2 with the neomycin resistance (neo) gene, utilizing the start codon of HEV ORF2. The replicon expresses HEV ORF1 proteins and neo, but HEV ORF3 is truncated and the remaining partial fragment of HEV ORF2 is out of frame. The neo gene was amplified by PCR from the plasmid pcDNA3.1(+) (Invitrogen) by using a primer pair specific to the 5' and 3' ends of the gene, including a 3'-terminal EcoRI restriction site. The 5' end of the neo gene was extended with nt 3963 to 5147 of pSK-HEV-2, generated by fusion PCR, including a 5'-terminal SfiI restriction site. The resulting fused PCR product was digested with SfiI and EcoRI and substituted into pSK-HEV-2 to yield pHEV/2Neo. The addition of the neo gene increased the genome size to 7.3 kb. The construct pHEV/3Neo, in which the neo gene is placed in frame with ORF3 of HEV, was constructed by replacing nt 5134 to 5816 of pSK-HEV-2 with the neo gene, using the same strategy as that used for pHEV/2Neo. The pHEV/3Neo construct was further modified to contain an extended poly(A) tail of 32 adenosines. Construct pHEV/T+ represents plasmid pSK-HEV-2 mutated in ORF3 by site-directed mutagenesis to insert an extra T after nt 5116. This insertion caused a frameshift which would force translation of ORF3 to initiate only from nt 5131. Construct pHEV5131Ala represents plasmid pSK-HEV-2 mutated at nt 5131 to 5133 by site-directed mutagenesis to change the ATG codon to GCA (Ala). This mutation eliminated the third AUG in ORF3 as an initiation codon for translation. Plasmid pCMV5122 contains a PCR-generated HEV fragment corresponding to nt 5122 to the 3' poly(A) tail of the genome cloned into the directional TOPO expression vector pcDNA3.1D/V5-His-Topo (Invitrogen) according to the manufacturer's instructions. The construct contains the human cytomegalovirus immediate-early promoter. Construct pT75007 contains a PCR-generated HEV fragment corresponding to nt 5007 to the 3' poly(A) tail of pSK-HEV-2 cloned into pCR8/GW/TOPO (Invitrogen) as specified by the manufacturer. The construct contains a T7 promoter prior to the HEV sequences and a unique BglII site following the poly(A) sequences. It was used to produce a size marker for Northern blot analyses. Plasmid pHEV-ORF3/His contains coding sequences for a His tag of six histidine residues inserted into pSK-HEV-2 prior to the stop codon of ORF3.
The entire HEV sequence of each construct was determined after mutagenesis to verify that unwanted mutations had not been introduced.
Cells. S10-3 cells, a subclone of the human hepatoma cell line Huh-7 (15), were derived by limiting dilution; they permit relatively efficient replication of HEV (S. U. Emerson, unpublished data). Cell monolayers of S10-3 cells were grown in Dulbecco's modified Eagle's medium supplemented with 2 mM L-glutamine and 9% fetal bovine serum in a CO2 incubator at 37°C. Transfected cells were maintained in a CO2 incubator at 34.5°C.
Transcription in vitro and transfection of cultured cells. Plasmids were linearized at a unique BglII site located immediately downstream of the poly(A) tract of the HEV sequence. Capped transcripts were synthesized with the T7 Riboprobe in vitro transcription system (Promega) from 4 μg linearized plasmid in the presence of 0.5 mM anti-reverse cap analog [3'-O-methyl-m7G(5')pppG; Ambion], 0.05 mM GTP, 0.5 mM (each) ATP, CTP, and UTP, 10 mM dithiothreitol, 1.6 U of RNasin/μl, and 0.8 U of T7 RNA polymerase/μl. The mixtures were incubated at 37°C for 2.5 h, with an additional 0.4 U of polymerase/μl added after the first 90 min. Transcription was terminated by the addition of 1.25 U of RNase-free RQ DNase/μg of plasmid DNA, followed by incubation for 30 min at 37°C. The RNA was extracted using the RNeasy clean-up protocol (QIAGEN). The concentration was determined by measurement of the optical density at 260 nm, and the integrity of the transcripts was determined by electrophoresis on a nondenaturing agarose gel. Transcribed RNA (1.5 to 3 μg) was mixed with a liposome mixture of 18 μl DMRIE-C (Invitrogen) and 900 μl OptiMEM (Invitrogen) for the transfection of S10-3 cells seeded to 80% confluence in one well of a six-well plate. When DNA was transfected, 1.5 μg of DNA was diluted in 0.5 ml OptiMEM and added to 0.5 ml OptiMEM containing 10 μl DMRIE-C. This transfection mixture was gently mixed and added to S10-3 cells seeded in one well of a six-well plate. In both cases, the cells were incubated at 34.5°C for 5 h, after which the transfection mixture was replaced with Dulbecco's modified Eagle's medium supplemented with 2 mM L-glutamine and 9% fetal bovine serum and the incubation was continued at 34.5°C.
Stable cell line selection. S10-3 cells were grown in the continuous presence of 1 mg/ml G418 sulfate starting 2 days after transfection with capped HEV/Neo transcripts synthesized in vitro. The medium was replenished every 2 to 3 days. Cells were split weekly once the cells surviving selection formed a confluent monolayer. The presence of replicon RNA was determined at 4 weeks posttransfection by reverse transcriptase (RT) PCR with and without reverse transcriptase, using primer sets designed to amplify a region overlapping HEV and neo sequences.
Preparation of biotinylated RNA probes. Plasmid pHEV/2Neo was used as template for PCR to generate DNA fragments that were agarose gel purified using the QIAquick gel extraction kit (QIAGEN). Primers included a T7 RNA polymerase promoter sequence (in capital letters below) to allow in vitro strand-specific transcription of RNA probes. Two DNA fragments for the synthesis of RNA probes detecting plus-strand HEV/2Neo RNA were amplified using primer sets E/neof6583 and E/neor6767T7 (gtagttattcaggattatgacaac and TAATACGACTCACTATAGgtcacagagtcagagacatagac) and E/neof5049 and E/neor5334T7 (ctgagtcagtgaagccagtgcttg and TAATACGACTCACTATAGGcctcgtcctgcagttcattcag). Two micrograms of the purified DNA fragment was transcribed in vitro with the T7 Riboprobe system (Promega) as recommended by the manufacturer. Digestion of the DNA template with RQ DNase I and clean-up of the RNA transcript on RNeasy columns (QIAGEN) followed. The RNA concentration was determined by measurement of the optical density at 260 nm.
RNA transcripts (500 ng) were biotinylated with the Bright Star psoralen-biotin nonisotopic labeling kit (Ambion) according to the instructions of the manufacturer.
Preparation of total RNA and Northern blot analysis. Total cellular RNA was extracted with the RNeasy kit (QIAGEN), including DNase treatment, according to the manufacturer's instructions. Northern blot analysis was conducted with the reagents and conditions specified in the NorthernMax kit (Ambion). Total RNA (10 μg) was denatured with formaldehyde loading dye solution for 15 min at 65°C and separated on a 1% agarose formaldehyde gel. The RNA was transferred to a nylon membrane (Bright Star Plus; Ambion) by capillary blotting. Following transfer, the RNA was cross-linked to the membrane (UV Stratalinker 1800; Stratagene), and the membrane was prehybridized in ultrahybridization solution (NorthernMax; Ambion) at 68°C for 1 h. Hybridization followed at 68°C overnight with the HEV-specific biotinylated riboprobe added to the ultrahybridization solution. The membrane was washed, and the bound biotinylated probe was detected with the Bright Star BioDetect kit (Ambion) as recommended by the manufacturer. The membrane was exposed to BioMax MS film (Kodak).
Determination of the 5'-terminal sequence of subgenomic RNA. Total RNA extracted from the HEV/2Neo or HEV/3Neo replicon-transfected cells was used to amplify the 5'-terminal sequence of the subgenomic RNA with the FirstChoice RNA ligase-mediated rapid amplification of cDNA ends (RLM-RACE) kit (Ambion) according to the manufacturer's instructions. Briefly, 10 μg of total RNA was treated with calf intestine alkaline phosphatase to remove 5'-terminal phosphates. Following RNA purification with TRIzol LS reagent (Invitrogen), one-half of the starting material was treated with tobacco acid pyrophosphatase (TAP) for 1 h at 37°C. One-fifth of this reaction mixture or 1 μl of RNA without TAP treatment was incubated with a 45-nt-long RNA adapter oligonucleotide and T4 RNA ligase at 37°C for 1 h. One-fifth of this reaction mixture was used in a reverse transcription reaction with random primer and M-MLV RT at 42°C for 1 h, followed by HEV-specific nested PCR with Super Taq polymerase (Ambion) and forward primer corresponding to the adapter sequence and reverse primer specific to HEV replicon HEV/2Neo or HEV/3Neo sequences. HEV/2Neo-specific reverse outer primer E/neor5460 (TACTTTCTCGGCAGGAGCAAGGTG) and reverse inner primer E/neor5334 (CCTCGTCCTGCAGTTCATTCAG) or E/neor5172 (CTGCGTGCAATCCATCTTGTTCAATCATGGTCGCGAACCCATGGGC) were used to amplify the 5'-terminal region of subgenomic RNA. The 5'-terminal sequence of the genome-length RNA was determined in parallel as a positive control and was amplified with HEV-specific outer primer Er343 (CAGCGGTGGACCACATTAGGATTG) and the adapter-specific outer primer, followed by inner HEV primer Er194 (AACAAGCTGGCGAGGTTGCATTAG) and the adapter-specific inner primer. The PCR product was gel purified using the QIAquick gel extraction kit (QIAGEN), and the consensus sequence of the purified PCR product was determined. In addition, the PCR product was cloned with the TOPO PCR cloning kit (Invitrogen), and several clones were sequenced.
Western blot analysis. Transfected S10-3 cells were lysed with cell lysis buffer (Promega) containing HALT protease inhibitor cocktail (EDTA free; Pierce) or with sodium dodecyl sulfate (SDS) gel loading buffer with -mercaptoethanol and proteinase inhibitor mix (Complete; Roche) and sonicated on ice three times for 1 min. Proteins were separated by SDS-polyacrylamide gel electrophoresis (PAGE) in a 10 to 20% or 16% Novex Tricine gel or 7% NuPAGE Tris-acetate gel (Invitrogen). Proteins were diluted in the appropriate sample buffer and NuPAGE sample reducing agent (Invitrogen), denatured for 10 min at 90°C, and electrophoretically transferred onto a polyvinylidene difluoride (PVDF) membrane (0.2 μm; Invitrogen) or nitrocellulose membrane (0.45 μm) for analysis by Western blotting. The PVDF membrane was pretreated for 3 min in methanol and 5 min in Novex Tris-glycine transfer buffer (Invitrogen). PVDF membranes were then blocked with 3% nonfat milk and 0.25% Tween 20 in Tris-buffered saline, and nitrocellulose membranes were blocked with StartingBlock blocking buffer (Pierce) with 0.5% Tween 20 overnight at 4°C, incubated with the indicated antibody for 1 h at room temperature, followed by an anti-rabbit or anti-human horseradish peroxidase-conjugated secondary antibody (1:50,000; Jackson ImmunoResearch), and visualized by a Visualizer Western blot detection kit (Upstate, Lake Placid, NY) as specified by the manufacturer.
Immunofluorescence microscopy. Transfected cells were briefly washed in 1x phosphate-buffered saline and then fixed with acetone for indirect immunofluorescence microscopy as described previously (6). Following fixation, the cells were stained with a mixture of an ORF2-specific convalescence-phase serum collected from an HEV-infected chimpanzee (Ch1313) (6) and rabbit anti-ORF3 polyclonal antibody (8) as primary antibody, followed by incubation with a mixture of the specific secondary antibodies Alexa 488-conjugated goat anti-human immunoglobulin G (Molecular Probes) and Alexa 568-conjugated goat anti-rabbit immunoglobulin G (Molecular Probes). Fluorescent cells were visualized with a Zeiss Axioscope 2 Plus fluorescence photomicroscope.
RESULTS
Establishment of a cell line persistently infected with an HEV replicon. Initial attempts to identify genomic or subgenomic RNA in cells transfected with infectious RNA genomes transcribed in vitro were frustrated by a combination of an insufficient number of transfected cells, inefficient viral replication, and a high background of residual input RNA. In order to overcome these obstacles, we sought to establish cell cultures in which every cell contained replicating viral genomes and excess input viral RNA was eliminated. This was accomplished by transfecting cells with an HEV replicon (HEV/2Neo) consisting of a replication-competent genome in which the ORF2 capsid protein gene was replaced by the neomycin resistance gene (neo). Exposure of the transfected cells to G418 sulfate killed all those cells not expressing neo; the expansion of surviving cells in the continuous presence of G418 sulfate diluted out excess input RNA and ensured that all cells contained a functioning replicon.
Identification of genomic and subgenomic viral RNA. If the ORF2 protein is encoded by a subgenomic RNA, as has been suggested (19), the neo gene occupying ORF2 of the replicon should also be expressed from a subgenomic RNA. To determine whether subgenomic RNAs were indeed present in these cells, we first performed Northern blot analyses of RNA extracted from three independently expanded G418 sulfate-resistant cell lines (s2N1, s2N2, and s2N5). The probe used was specific for the 3'-terminal region of the genome (HEV/2Neo, nt 6583 to 6767) and should detect HEV/2Neo genomic RNA (7.3 kb) and any 3'-terminal subgenomic RNAs (Fig. 1A). For each of the three cell lines, two RNA species of approximately 7.3 kb and 2.2 kb were detected (Fig. 1B). No other RNA species were detected even after prolonged exposure of the blot. Two independent G418 sulfate-resistant cell lines expanded from another transfection (s2N8-1 and s2N8-2) were blotted against another probe specific for the overlap region of the three viral ORFs (HEV/2Neo, nt 5049 to 5334). This probe also detected only two RNA species, and they appeared to be identical in size to those detected by the first probe (Fig. 1C). The identification of a subgenomic RNA in each of the five cell lines supported the contention that the ORF2 gene was translated from a subgenomic RNA, but the results failed to support the conclusion that a second subgenomic RNA encodes ORF3 products.
Attempts to detect a minus strand of either genomic or subgenomic RNA failed, most likely because the concentration of the minus strand was too low.
The 5' terminus of the subgenomic RNA is downstream of the proposed translation initiation codon of ORF3. Since the viral genome is capped, it seemed probable that subgenomic RNAs would also be capped. Therefore, in order to determine the 5' sequence of the subgenomic RNA, 5' RACE was performed, utilizing a procedure specific for capped RNA. In this procedure, total RNA is treated with calf intestine alkaline phosphatase to remove 5'-terminal phosphates and thus prevent ligation to an adapter molecule. A 5' monophosphate is then exposed by cleaving the GpppX cap with TAP, and this 5' phosphate is then ligated to an oligonucleotide adapter. Finally, RT-PCR is performed with a forward primer specific for the adapter and a reverse primer specific for a selected HEV sequence.
As positive controls, a reverse primer complementary to nt 171 to 194 (Er194) of HEV was chosen to amplify the 5' sequence of full-length genomic RNA from two of the cell lines (s2N5 and s2N8-1) as positive controls. A single product of the expected size (about 200 bp) was detected for each cell line following agarose gel electrophoresis (data not shown). The consensus sequence of this product was identical to that of the 5' terminus of the original cDNA clone of the replicon, thus validating the assay (data not shown).
In order to maximize the probability of identifying two subgenomic RNAs of slightly different size or sequence, we designed two different reverse primers corresponding to two regions (Er5172, nt 5127 to 5172; and Er5334, nt 5316 to 5334) 181 nt apart and downstream of the first methionine codon in ORF3. Again, the results from analyses of the two cell lines (s2N5 and s2N8-1) were identical. Primer Er5172 amplified a PCR product of about 100 bp, whereas primer Er5334 amplified a product of about 250 bp (Fig. 2A). The negative controls corresponding to each sample were performed with the same extracted total RNA, however without TAP treatment. We obtained PCR products that differed in size from one cell line to the other (Fig. 2A) and from the uniform PCR product obtained by including TAP. The PCR products of the negative controls most likely reflect products amplified from randomly nicked RNA.
The 5' sequence obtained for either the approximately 100-bp or the approximately 250-bp products originated at nt 5122 of the HEV genome (Fig. 2B, examples shown for cell lines s2N8-1), 25 nt upstream of the predicted translation initiation codon for the ORF2-encoded protein but 18 nt downstream of that predicted for the ORF3-encoded protein. As Fig. 2B demonstrates, there were no nucleotide insertions or deletions or minor species. The same sequence was obtained when total RNA from another cell line, Huh-7.5, stably transfected with HEV/2Neo replicon transcripts was analyzed by 5' RLM-RACE (data not shown). These results suggested that there was only a single subgenomic RNA, that it was indeed capped, and that it encoded ORF2-expressed proteins. However, ORF3 protein expression could not be accounted for if it involved initiation at either of the first two methionine codons in ORF3.
The first in-frame AUG of ORF3 is not used to initiate translation. We attempted to determine the amino terminus of the ORF3 protein by direct chemical sequencing. RNA transcribed from a mutant replicon encoding an NH2-ORF3/6His-COOH fusion protein (HEV-ORF3/His) was transfected into S10-3 cells, and the fusion protein was purified from cell lysate by nickel-agarose chromatography. Although a band corresponding to the ORF3 protein was detected by staining with Simply Blue reagent (Invitrogen) or by Western blotting following SDS-PAGE and transfer to PVDF membrane, two independent attempts to determine the N-terminal sequence by Edman degradation failed, suggesting that the amino terminus was blocked (data not shown). Therefore, we resorted to an indirect method to determine which methionine codon might be used to initiate ORF3 protein production.
An additional T residue was inserted after nt 5116 of our infectious cDNA clone of a genotype 1 strain to generate pHEV/T+ and mimic the gene structure in this region of genotype 4 strains. The resulting frameshift should force translation of the ORF3 protein to initiate at the third AUG codon. Capped in vitro transcripts of this plasmid were infectious for S10-3 cells, and both ORF2 and ORF3 proteins were detected by immunofluorescence microscopy (Fig. 3A). Infectious virus was recovered from cells transfected with HEV/T+ (S. U. Emerson, unpublished data), as would be predicted based on the viability of genotype 4 strains.
ORF3 protein produced by the HEV/T+ mutant should contain 114 amino acids compared to the 123 predicted if initiation occurred at the first in-frame AUG of the genotype 1 wild-type parent. Western blot analysis was performed to compare the sizes of the ORF3 proteins produced following transfection of wild-type and HEV/T+ genomes. The ORF3 protein extended by six histidine residues was used as a size marker. The ORF3 proteins produced by the wild type and the HEV/T+ mutant comigrated on SDS-PAGE and migrated slightly faster than the ORF3/His fusion protein (Fig. 3B). These results suggested that the first AUG codon previously assumed to serve as the initiation codon for ORF3 protein synthesis was not used for this purpose in the genotype 1 strain but rather that the third in-frame AUG codon was used.
Subgenomic RNAs expressing ORF2- and ORF3-encoded proteins have the same 5' terminus. Although a single subgenomic RNA was produced when neo was expressed from ORF2, it could be argued that a subgenomic RNA expressing ORF3 protein would not have been under the same selection pressure and might have been lost. In order to examine this possibility, we constructed a second replicon almost identical to the first, except that neo was inserted into ORF3. Cell clones were selected as described before, and cap-dependent 5' RACE was performed with the same outer primers used for the ORF2 analysis and with inner primer Er5334. Not only did the PCR product display a mobility similar to those shown in Fig. 2A (data not shown), but sequence analysis demonstrated that the 5' sequence of the subgenomic RNA was identical to that obtained previously (Fig. 4). Therefore, regardless of whether the cell lines were selected by pressure on ORF2 or ORF3 gene expression, the subgenomic RNA initiated at nt 5122 of the parent genome.
Mutation of the third AUG in the 5' terminus of ORF3 abolishes ORF3 protein synthesis. Identification of nt 5122 as the 5' terminus of the subgenomic RNA encoding neo in frame with either ORF2 or ORF3 implied that the third AUG in ORF3 was the only start codon available to initiate translation of the ORF3 protein. In order to confirm these findings, we mutated the third methionine codon in ORF3 of our infectious cDNA clone to an alanine codon. Capped in vitro transcripts of the pHEV5131Ala construct transfected into S10-3 cells expressed HEV ORF2 proteins (Fig. 5a). However, indirect immunofluorescence microscopy of the HEV5131Ala-transfected cells failed to detect HEV ORF3-encoded proteins (Fig. 5b). As a positive control, cells transfected in parallel with transcripts of the parental infectious HEV cDNA contained both ORF2 and ORF3 proteins as previously reported (6; data not shown). The lack of ORF3 protein expression in HEV5131Ala-transfected cells confirmed that the third AUG in ORF3 of the HEV genome is responsible for ORF3 protein synthesis.
A recombinant mRNA lacking the first two methionine codons in ORF3 encodes both ORF2 and ORF3 proteins. As formal confirmation that ORF2 and ORF3 proteins both could be translated from a single mRNA, a cytomegalovirus promoter-driven plasmid (pCMV5122) was constructed to express a subgenomic RNA [nt 5122 to the poly(A) tail] analogous to the HEV/2Neo and HEV/3Neo subgenomic RNAs but encoding authentic HEV proteins. As expected, transient expression of this plasmid following transfection of S10-3 cells resulted in the synthesis of both HEV ORF2 and HEV ORF3 proteins within the same cell, as detected by immunofluorescence microscopy (Fig. 6A). Western blot analyses with an ORF2 protein-specific polyclonal serum (Fig. 6B) and with a rabbit anti-ORF3 protein polyclonal antibody (Fig. 6C) (8) were performed on cell lysates. Both ORF2 and ORF3 proteins were detected, and their mobility on SDS-PAGE was indistinguishable from that of the two proteins expressed from the full-length HEV genomes in other experiments.
DISCUSSION
The existence of the two 3'-coterminal subgenomic viral RNAs identified in the liver of macaques infected with HEV (19) has been difficult to confirm in the absence of an efficient cell culture system. Although transfected HEV recombinant genomes are able to replicate and produce infectious virus in Huh-7 cells (6), the level of replication has not been sufficient to permit detection of de novo-synthesized viral RNA. We have overcome this problem by generating stable transformed cell lines that were selected for their ability to survive continuous treatment with G418 sulfate due to the expression of the neomycin resistance gene (neo) from either ORF2 or ORF3 of an HEV replicon. It should be emphasized that the Huh-7 cell transfection system is as close to in vivo macaque models as is presently possible. The Huh-7 cells are liver cells of human origin, and production of ORF2 and ORF3 proteins in these cells requires a functional RNA-dependent RNA polymerase, which most likely reflects the need for synthesis of subgenomic mRNA. Also, authentic HEV, which is infectious for macaques, has been isolated from these cells following transfection (6).
Analyses of the viral RNA in cultures of multiple cell lines demonstrated that the full-length replicon transfected into the cells was maintained during expansion of the culture and, therefore, was replicating and that a single subgenomic RNA was also produced and maintained. These results are consistent with the report of an approximately 2-kb subgenomic RNA found in the liver of infected macaques. However, we found absolutely no indication of the 3.7-kb, second subgenomic RNA that was reported to be in the infected liver (19) or that was detected in cells infected with a Chinese strain of HEV (25). At the moment, we can only speculate as to the origin and function of this larger subgenomic RNA, but one intriguing explanation is that it is a defective virus similar to those commonly produced by RNA viruses.
Our results demonstrated that the "different" strategy proposed for translation of ORF3 proteins of genotype 4 viruses compared to that of all other genotypes (24) actually describes the process also used by genotype 1 viruses; the presumed difference was based on the assumption that the first methionine codon in ORF3 was used to initiate translation. However, our data demonstrated that neither the first nor the second AUG of ORF3 was even present in the subgenomic RNA expressing the ORF3 protein and that substitution of the third methionine codon in ORF3 by an alanine codon abolished the expression of the ORF3 protein. Although we have examined a genotype 1 strain only, it is not unreasonable to assume that genotype 2 and 3 strains will utilize a similar expression strategy and that each mammalian strain of HEV will produce ORF2 and ORF3 proteins very similar in size to those of any other genotype.
In the scanning model proposed by Kozak (11), translation of capped RNAs involves binding of the 40S ribosomal subunit to the 5' end of mRNA, followed by scanning in the 3' direction until an AUG codon in the appropriate context for initiation is reached. In most cases, if the first in-frame AUG is not in the optimal context, a proportion of the scanning complexes may fail to initiate here and continue scanning. Our studies provide compelling evidence for a bicistronic RNA in which two different reading frames are utilized, depending on which of the two closely spaced AUG codons are selected by the ribosomes for the initiation of translation.
Although apparently rare, a similar phenomenon involving two closely spaced AUG codons was described previously for the S1 gene of reovirus (17). In that case, the choice of a particular initiation codon was thought to reflect the degree of homology to the Kozak consensus sequence as well as the distance from the cap. It was also proposed that the use of a bicistronic mRNA enabled reovirus to regulate the expression of the cognate proteins at the level of elongation (7).
The discovery that the HEV subgenomic RNA initiates at nt 5122 eliminates a number of potential inconsistencies. First, it greatly reduces the likelihood that ORF2 and ORF3 proteins of genotype 4 strains differ in size from those of the other three genotypes. Also, it provides an explanation for why the first two AUG codons in ORF3 are in an unfavorable context for initiation according to the Kozak model (data not shown): they are not used for translation since they are excluded from the subgenomic RNA.
It is interesting to note that 12 nucleotides at the start of ORF3 are almost universally conserved. Evidence was previously presented that this sequence represented a cis-acting replication element (8). Mutations, even silent ones, introduced into this region of ORF3 unexpectedly abolished the synthesis of the ORF2 protein as well as that of the ORF3 protein in cells transfected with the full-length genome. This result, coupled with the current demonstration that both ORF2 and ORF3 proteins are produced from a single subgenomic RNA that begins downstream of this region, suggests that this highly conserved sequence is part of, or is, the promoter for subgenomic RNA synthesis. Significantly, this sequence is totally conserved across all four mammalian genotypes.
It is intriguing that the translation initiation codon for the ORF3-encoded protein is before that of the ORF2 protein in a bicistronic mRNA. Since the first methionine codon of capped RNAs is the one most often used for initiation, one might expect that this placement evolved to ensure a high level of ORF3 protein production. Observation of the sequence indicated that the Kozak consensus sequences for the initiation codons in ORF2 and ORF3 were equally optimal; therefore, ORF3 protein might be translated as efficiently as ORF2 protein. Knowledge of the molar ratios of ORF2 protein to ORF3 protein in infected cells could prove informative as to what function(s) ORF3 protein performs. Unfortunately, HEV replication in our cell lines is still too low for us to quantify ORF2 and ORF3 proteins in cells.
Since ORF2 protein is the major, if not only, capsid protein, it must be relatively abundant. It is difficult to imagine a comparable need for the ORF3 protein since there is no evidence that it is a major component of the virion; virus-like particles are made in its absence in a baculovirus expression system (12, 21), and antibody to it does not neutralize virus infectivity (S. U. Emerson, unpublished data) (20). Resolution must await more-efficient cell culture systems that will allow the functions of ORF2 and ORF3 proteins to be more fully defined. In the meantime, the discovery of a unique strategy for the synthesis of ORF2 and ORF3 proteins from a single subgenomic RNA provides further evidence that HEV belongs in its own family rather than in the Caliciviridae family, where it was originally classified. It also demonstrates that there is much more to learn about this interesting human pathogen.
ACKNOWLEDGMENTS
This work was supported by the Intramural Research Program of the NIH, NIAID.
We thank Mark Garfield of the Research Technologies Branch of NIH, NIAID, for performing the Edman degradation. We are grateful to Charles M. Rice (Rockefeller University) for providing the Huh-7.5 cell line.
REFERENCES
Agrawal, S., D. Gupta, and S. K. Panda. 2001. The 3' end of hepatitis E virus (HEV) genome binds specifically to the viral RNA-dependent RNA polymerase (RdRp). Virology 282:87-101.
Arankalle, V. A., L. P. Chobe, M. V. Joshi, M. S. Chadha, B. Kundu, and A. M. Walimbe. 2002. Human and swine hepatitis E viruses from Western India belong to different genotypes. J. Hepatol. 36:417-425.
Balayan, M. S., A. G. Andjaparidze, S. S. Savinskaya, E. S. Ketiladze, D. M. Braginsky, A. P. Savinov, and V. F. Poleschuk. 1983. Evidence for a virus in non-A, non-B hepatitis transmitted via the fecal-oral route. Intervirology 20:23-31.
Cooper, K., F. F. Huang, L. Batista, C. D. Rayo, J. C. Bezanilla, T. E. Toth, and X. J. Meng. 2005. Identification of genotype 3 hepatitis E virus (HEV) in serum and fecal samples from pigs in Thailand and Mexico, where genotype 1 and 2 HEV strains are prevalent in the respective human populations. J. Clin. Microbiol. 43:1684-1688.
Emerson, S. U., D. Anderson, A. Arankalle, X. J. Meng, M. Purdy, G. G. Schlauder, and S. Tsarev. 2004. Hepevirus. Elsevier/Academic Press, London, United Kingdom.
Emerson, S. U., H. Nguyen, J. Graff, D. A. Stephany, A. Brockington, and R. H. Purcell. 2004. In vitro replication of hepatitis E virus (HEV) genomes and of an HEV replicon expressing green fluorescent protein. J. Virol. 78:4838-4846.
Fajardo, J. E., and A. J. Shatkin. 1990. Translation of bicistronic viral mRNA in transfected cells: regulation at the level of elongation. Proc. Natl. Acad. Sci. USA 87:328-332.
Graff, J., H. Nguyen, C. Yu, W. R. Elkins, M. St. Claire, R. H. Purcell, and S. U. Emerson. 2005. The open reading frame 3 gene of hepatitis E virus contains a cis-reactive element and encodes a protein required for infection of macaques. J. Virol. 79:6680-6689.
Koonin, E. V., A. E. Gorbalenya, M. A. Purdy, M. N. Rozanov, G. R. Reyes, and D. W. Bradley. 1992. Computer-assisted assignment of functional domains in the nonstructural polyprotein of hepatitis E virus: delineation of an additional group of positive-strand RNA plant and animal viruses. Proc. Natl. Acad. Sci. USA 89:8259-8263.
Korkaya, H., S. Jameel, D. Gupta, S. Tyagi, R. Kumar, M. Zafrullah, M. Mazumdar, S. K. Lal, L. Xiaofang, D. Sehgal, S. R. Das, and D. Sahal. 2001. The ORF3 protein of hepatitis E virus binds to Src homology 3 domains and activates MAPK. J. Biol. Chem. 276:42389-42400.
Kozak, M. 1989. The scanning model for translation: an update. J. Cell Biol. 108:229-241.
Li, T. C., Y. Yamakawa, K. Suzuki, M. Tatsumi, M. A. Razak, T. Uchida, N. Takeda, and T. Miyamura. 1997. Expression and self-assembly of empty virus-like particles of hepatitis E virus. J. Virol. 71:7207-7213.
Magden, J., N. Takeda, T. Li, P. Auvinen, T. Ahola, T. Miyamura, A. Merits, and L. Kaariainen. 2001. Virus-specific mRNA capping enzyme encoded by hepatitis E virus. J. Virol. 75:6249-6255.
Meng, X. J. 2000. Novel strains of hepatitis E virus identified from humans and other animal species: is hepatitis E a zoonosis J. Hepatol. 33:842-845.
Nakabayashi, H., K. Taketa, K. Miyano, T. Yamane, and J. Sato. 1982. Growth of human hepatoma cell lines with differentiated functions in chemically defined medium. Cancer Res. 42:3858-3863.
Purcell, R. H., and S. U. Emerson. 2001. Hepatitis E virus, p. 3051-3061. In D. Knipe, P. Howley, D. Griffin, R. Lamb, M. Martin, B. Roizman, and S. Straus (ed.), Fields virology, 4th ed. Lippincott Williams and Wilkins, Philadelphia, Pa.
Sarkar, G., J. Pelletier, R. Bassel-Duby, A. Jayasuriya, B. N. Fields, and N. Sonenberg. 1985. Identification of a new polypeptide coded by reovirus gene S1. J. Virol. 54:720-725.
Schlauder, G. G., and I. K. Mushahwar. 2001. Genetic heterogeneity of hepatitis E virus. J. Med. Virol. 65:282-292.
Tam, A. W., M. M. Smith, M. E. Guerra, C. C. Huang, D. W. Bradley, K. E. Fry, and G. R. Reyes. 1991. Hepatitis E virus (HEV): molecular cloning and sequencing of the full-length viral genome. Virology 185:120-131.
Tam, A. W., R. White, P. O. Yarbough, B. J. Murphy, C. P. McAtee, R. E. Lanford, and T. R. Fuerst. 1997. In vitro infection and replication of hepatitis E virus in primary cynomolgus macaque hepatocytes. Virology 238:94-102.
Tsarev, S. A., T. S. Tsareva, S. U. Emerson, A. Z. Kapikian, J. Ticehurst, W. London, and R. H. Purcell. 1993. ELISA for antibody to hepatitis E virus (HEV) based on complete open-reading frame-2 protein expressed in insect cells: identification of HEV infection in primates. J. Infect. Dis. 168:369-378.
Tyagi, S., H. Korkaya, M. Zafrullah, S. Jameel, and S. K. Lal. 2002. The phosphorylated form of the ORF3 protein of hepatitis E virus interacts with its non-glycosylated form of the major capsid protein, ORF2. J. Biol. Chem. 277:22759-22767.
Tyagi, S., M. Surjit, A. K. Roy, S. Jameel, and S. K. Lal. 2004. The ORF3 protein of hepatitis E virus interacts with liver-specific alpha1-microglobulin and its precursor alpha1-microglobulin/bikunin precursor (AMBP) and expedites their export from the hepatocyte. J. Biol. Chem. 279:29308-29319.
Wang, Y., H. Zhang, R. Ling, H. Li, and T. J. Harrison. 2000. The complete sequence of hepatitis E virus genotype 4 reveals an alternative strategy for translation of open reading frames 2 and 3. J. Gen. Virol. 81:1675-1686.
Xia, X., R. Huang, and D. Li. 2000. [Studies on the subgenomic RNAs of hepatitis E virus.] Wei Sheng Wu Xue Bao 41:622-627.(Judith Graff, Udana Toria)