Mutational analysis of hepatitis E virus ORF1 "Y-domain": Effects on RNA replication and virion infectivity

AIM To investigate the role of non-structural open reading frame 1 “Y-domain” sequences in the hepatitis E virus (HEV) life cycle. METHODS Sequences of human HEV Y-domain (amino acid sequences 216-442) and closely-related viruses were analyzed in silico. Site-directed mutagenesis of the Y-domain (HEV SAR55) was carried out and studied in the replicon-baculovirus-hepatoma cell model. In vitro transcribed mRNA (pSK-GFP) constructs were transfected into S10-3 cells and viral RNA replicating GFP-positive cells were scored by flow cytometry. Mutant virions’ infectivity was assayed on naïve HepG2/C3A cells. RESULTS In silico analysis identified a potential palmitoylation-site (C336C337) and an α-helix segment (L410Y411S412W413L414F415E416) in the HEV Y-domain. Molecular characterization of C336A, C337A and W413A mutants of the three universally conserved residues showed non-viability. Further, of the 10 consecutive saturation mutants covering the entire Y-domain nucleotide sequences (nts 650-1339), three constructs (nts 788-994) severely affected virus replication. This revealed the indispensability of the internal sequences but not of the up- or downstream sequences at the transcriptional level. Interestingly, the three mutated residues corresponded to the downstream codons that tolerated saturation mutation, indicating their post-translational functional/structural essentiality. In addition, RNA secondary structure prediction revealed formation of stable hairpins (nts 788-994) where saturation mutation drastically inhibited virion infectivity. CONCLUSION This is the first demonstration of the critical role of Y-domain sequences in HEV life cycle, which may involve gene regulation and/or membrane binding in intracellular replication complexes.


INTRODUCTION
The eukaryotic positive single-strand (+ss) RNA viruses share evolutionarily-conserved functional and putative domains, and even amino acid (aa) sequences in their nonstructural/replicase polyproteins [1] . In infected cells, one of the polyprotein proteolytic products, the methyltransferase (MTase), catalyzes 5' capping of viral mRNA and interacts with cytoplasmic membranes, essential for establishing replication complexes [2] . In addition to the MTase-domain, studies have also shown sequence and structural conservation of the downstream Y-domain in viral polyproteins [1,3,4] .
The hepatitis E virus (HEV), the only Hepevirus of the alphavirus-like superfamily, is the etiological agent of acute and chronic hepatitis E in humans [22,23] . The HEV +ssRNA genome (about 7.2 kb) contains three partially overlapping open reading frames (ORFs): ORF1, ORF2 and ORF3, flanked by 5' and 3' short untranslated-regions [24,25] . Of these, the largest gene, ORF1 (5109 bases), encodes the nonstructural polyprotein (1703 residues) essential for viral RNA replication in infected cells [26][27][28] . Homologous to the alphavirus polyprotein structural organization, the MTase-domain is followed by the Y-domain in HEV ORF1 ( Figure 1). While the 5' mRNA capping activity of the ORF1 MTase-domain (N-terminal) is well characterized and implicated in RNA replication [29,30] , the function of the Y-domain remains completely unexplored. Therefore, the present study was postulated to investigate a potential role of Y-domain sequences in HEV life cycle, using the repliconbaculovirus-hepatoma cell model.

In vitro capped mRNA synthesis and cell transfection
The replicon constructs (cDNA) were linearized with BglⅡ (Invitrogen) and in vitro transcribed in the presence of anti-reverse cap analog (ARCA; Ambion, Austin, TX, United States) essentially as described elsewhere [31] . The transcription mix was gel electrophoresed to check the size, integrity and quality of the capped-mRNA, followed by transfection into S10-3 cells. The transfected cultures were incubated at   presentative alphaviruses identified a potential palmitoylation-site homolog CC and an α-helix counterpart LYSWLFE in the Y-domain, predicted for cytoplasmic membrane binding. The residues C336C337 and their positions were highly conserved across the available sequences of all four human HEV genotypes ( Figure 2). In the predicted α-helix, while residue and positional conservation of L410, S412 and W413 were found among HEV and SFV, SINV and equine encephalitis virus (EEV) (Figure 3), the segment L410Y411S412W413L414F415E416 showed a high conservation within the HEV genotypes ( Figure 4).

Y-domain nts 788-994 are indispensable for virus replication
The saturation mutations introduced in the cDNA did not affect the gross yield of in vitro synthesized transcripts ( Figure 5A, left). Of the 10 consecutive mutant transcripts (pSK-GFP-Ydom1 to Ydom10), mutants of nts 788-994 (Ydom3, 4 and 5) drastically affected RNA replication by > 92% in S10-3 cells, whereas those of nts 650-787 (Ydom1 and 2) and nts 995-1339 (Ydom6, 7, 8, 9 and 10) had very mild or insignificant effect on viability compared to the wild-type (Ydom-WT) ( Figure 5B). This clearly demonstrated the indispensability of the internal sequences but not the up-or downstream sequences of the Y-domain at transcriptional level.

Universally conserved C336, C337 and W413 are critical for RNA replication
Similar to the saturation mutations, aa substitutions in the cDNA had no effect on the gross yield of in vitro synthesized RNA ( Figure 5A, right). Introduction of C336A, C337A and W413A substitutions within the predicted membrane binding motif of Y-domain completely abolished virus replication ( Figure 5B). Interestingly, the aa C336, C337 and W413 corresponded to codons (nts 1031-1033, 1034-1036 and 1213-1215, respectively) that were shown to be dispensable by saturation mutations. This very clearly indicated their post-translational functional/structural essentiality in virus replication, probably through membrane binding in intracellular replication complexes.

Effects of RNA hairpin/stem-loop structures (nts 788-994) on virion infectivity
As revealed by transfection results, Ydom3 (nts 788-856) had the most drastic effect on RNA replication, followed by Ydom5 (nts 926-994) and Ydom4 (nts 857-925). In line with this, while nts 788-856 formed the most stable RNA hairpin/ stem-loop compared to nts 926-994, nts 857-925 presented the least stable structure ( Figure 6A). This strongly supported the deleterious effects of saturation mutations that could completely unzip and destabilize the RNA secondary structures, critical for virus replication. Further, the three saturation 34.5 ℃ for 6 d to allow for optimal production of green fluorescent protein (GFP), indicating virus replication. S10-3 culture transfected with pSK-GFP-WT transcript showing green fluorescence served as positive control, while that receiving pSK-GFP-G816V, a lethal mutant [27] , served as negative control. All transfections were performed in duplicate and repeated for reproducibility.

RNA trans-encapsidation and virion infectivity assay
The mutant RNA encapsidation into viral ORF2 (capsid) proteins over-expressed by a recombinant baculovirus (vBacORF2) that could produce virus particles in S10-3 cells was performed, and tested for their infectivity on naïve HepG2/C3A cells [32,33] . In sum, RNA transfected S10-3 cells (in 12-well plates) were transduced with vBacORF2 on the following day. On day 6 posttransfection (i.e., day 5 post-transduction), lysates were prepared by vigorous vortexing of the cells in 10 × PBS, and normalized with sterile water. Lysates (inoculums) were cleared by centrifugation and overlaid on the HepG2/C3A cells (in 24-well plates), following 2.5 h incubation at 37 ℃. The inoculums were aspirated and complete medium was added, and the cells were incubated for 6 d to establish infection and GFP production. The assay was done in duplicate and repeated for reproducibility.

Fluorescence microscopy
The replication fitness of the mutant RNA was indirectly assessed by careful observation of GFP-positive S10-3 and HepG2/C3A cells under fluorescence microscope. The expression of capsid protein in S10-3 cells was confirmed on day 4 post-transduction [32] . Briefly, vBac-ORF2 transduced cells (in 8-chambered glass slides) were immune-stained with anti-ORF2 chimp sera and Alexa Fluor 488 goat anti-human IgG (Molecular Probes, Eugene, OR, United States). Following mounting with Vectashield (Vector Laboratories, Burlingame, CA, United States), slides were observed under FITC filter-aided indirect fluorescence microscope (H600L; Nikon, Tokyo, Japan).

Flow cytometry
On day 6, the duplicated transfected S10-3 cultures (24-well plates) were harvested by trypsinization [32] . In sum, each culture suspension (about 500 μL in cold PBS) was cleared at 4 ℃ and cell pellets were resuspended in 300 μL of cold PBS. The samples were immediately subjected to flow cytometry (10000 cells counted/sample), and data was analyzed for GFPpositive cells.

Mapping of potential palmitoylation-site and α -helix segment in the Y-domain
Multiple sequence analysis of HEV strains and re-  .    Figure 6B). This strongly supported the deleterious effects of saturation mutations that could completely unzip and destabilize the RNA secondary structures, critical for virus replication and infectivity.

DISCUSSION
The alphavirus-like superfamily includes pathogenic animal viruses like HEV, SFV and SNV, and plant viruses such as BMV, AMV, ToMV, TMV and CMV, having +ssRNA genomes. In infected cells, viral nonstructural/replicase proteins recruit and catalyze 5' capping of their genomic RNA onto cytoplasmic membranes, and synthesize the complementary strand that is sequestered in membrane-bound replication complexes [34] . In the present mutational study, the role of the ORF1 Y-domain was investigated, using an HEV replicon that expressed GFP (instead of the ORF2 protein) as a reporter of virus RNA replication and infectivity in hepatoma cells. Of the series of replicon constructs with saturation mutations in the Y-domain, the nts 788-994 mutants drastically affected RNA replication and particle infectivity, showing the critical role of the Y-domain internal sequences in the HEV life cycle.
A universal feature of +ssRNA viruses is the cytoplasmic membrane binding of nonstructural proteins that requires post-translational fatty acylation. Palmitoylation of cysteine residues has been widely studied in a number of cellular and viral proteins [35] .
In eukaryotic α-subunit of G [36] , Ras [37] , and neuronal GAP-43 [38] membrane binding proteins, a single cysteine in the only identified motif (MGC) undergoes palmitoylation. Similarly, in alphavirus-like superfamily nonstructural/replicase proteins, at least one cysteine is found in the homologous palmitoylation-sites. In SFV, the tight membrane association of nsP1 was shown due to increased palmitoylation of one of the highly conserved cysteines (C418C419C420)) within the     MTase core region, where C→A changes completely abolished palmitoylation [15,39] . These mutations were also shown to attenuate virus replication and pathogenicity in infected mice [8,9] . Likewise, the C420A mutation in SNV-nsP1 also had inhibitory effect on virus replication [8] . In ToMV, the palmitoylation-site counterpart contains three conserved but distantly located cysteines (C179/C186/C581), wherein C→S substitution strongly decreased membrane binding, 5' capping and RNA replication [10] . Akin to this, mutating the two conserved cysteines (C179/C186) at the same position aborted replication of BMV [11] and AMV [12] . Interestingly, mutation of the only C461 of CMV-1a also abrogated membrane binding and RNA replication [13] .

II
In accordance with this, C→A substitution of the highly conserved C336C337 residues in the predicted palmitoylation-site of HEV Y-domain also abolished RNA replication completely.
Moreover, in addition to cysteine palmitoylation, many of the viral polyproteins contain consensus hydrophobic sequences for tight membrane binding [40] . In SFV-nsP1, the amphipathic segment GSTLYTESRKLLRSWHLPSV (aa 245-264) that forms an α-helix, has been implicated in membrane binding [18,19] and RNA replication [21] . A mutational analysis of BMV and CMV-1a, wherein virus replication was abolished while affecting membrane binding and RNA recruitment, suggested the structural conservation of its amphipathic helix A [13] . Similarly, the poliovirus-2C [41] and hepatitis C virus (HCV)-NS5A [42] have been implicated in amphipathic helix-modulated interactions with intracellular membranes. In this report, a highly conserved segment LYSWLFE (aa 410-416) was mapped as the α-helix counterpart of the ORF1 Y-domain. Sequence alignment showed the universal conservation of L410, S412 and W413 among HEV and the alphavirus-like superfamily. In line with this and the previously reported deleterious effect of W259A change on SFV replication [21] , the specifically selected W413A completely abolished HEV RNA replication. Notably, tryptophan is a signature hydrophobic residue that is critical for α-helical protein folding for proteinprotein interactions. Of the several examples, W630 in the conserved motif KTXXXW of amphipathic helix of the G protein-coupled receptor Frizzled (C-terminal) has been shown crucial in intracellular protein     Probability ≥ 99% 99% > Probability ≥ 95% 95% > Probability ≥ 90% 90% > Probability ≥ 80% 80% > Probability ≥ 70% 70% > Probability ≥ 60% 60% > Probability ≥ 50% 50% > Probability A Parvez MK. Mutational analysis of HEV Y-domain 40 interactions [43] . Such functional/structural homology suggests the essentiality of the Y-domain (C-terminal) predicted α-helix in HEV replication that may embody common principles of viral nonstructural proteins in membrane interaction.
In conclusion, the present study shows the indispensability of highly conserved sequences (nts 788-994) of ORF1 Y-domain in HEV RNA replication and infectivity. Also, the universally conserved C336, C337 and W413 residues corresponding to the dispensable codons within the predicted membrane binding motif of Y-domain are critical for virus viability. Taken together, this is the first demonstration of the essentiality of Y-domain in the HEV life cycle, probably through gene regulation and/or membrane binding in replication complexes. Nevertheless, further molecular and biochemical studies are recommended to validate these findings.

Background
Hepatitis E virus (HEV) is the etiological agent of acute and chronic hepatitis in humans, worldwide. While there has been great progress in understanding the virus biology, the function of nonstructural open reading frame 1 "Y-domain" remains completely unexplored.

Research frontiers
The author has performed in silico analysis of closely related single-strand RNA virus sequences and mapped a potential palmitoylation-site and α-helix segment in the HEV nonstructural Y-domain. Mutational characterization of the viral replicon in S10-3 cells has shown the criticality of "Y-domain" residues C336C337 and W413 of the putative palmitoylation and helix, respectively. Further introduction of base mutations in the "Y-domain" severely affected RNA replication, revealing their post-transcriptional essentiality. Notably, the universally conserved C336C337 and W413 corresponded to non-conserved codons, indicating their post-translational indispensability. Moreover, RNA secondary structure prediction showed hairpin formations by the critical bases (nts 788-994) where mutations drastically affected virion infectivity of naïve HepG2 cells.

Innovations and breakthroughs
This is a novel study showing a critical role of the hitherto undefined Y-domain in HEV genomic RNA replication and infectious virion production in cultured liver cells.

Applications
The data warrants further biochemical and molecular studies of the Y-domain towards understanding HEV replication.

Terminology
The putative palmitoylation and α-helix motifs of the HEV Y-domain may be key to intracellular membrane-binding, essential for RNA replication and production of infectious virions.