The ORF3 protein of hepatitis E virus is a phosphoprotein that associates with the cytoskeleton.
ABSTRACT Hepatitis E virus (HEV) is a major human pathogen in the developing world. In the absence of an in vitro culture system, very little information exists on the basic biology of the virus. A small protein (approximately 13.5 kDa) of unknown function, pORF3, is encoded by the third open reading frame of HEV. We expressed pORF3 in transiently transfected COS-1 and Huh-7 cells and showed that it is a phosphoprotein which is modified at a serine residue(s). Deletion and site-directed mutants were created to establish Ser-80 as the phosphorylation site. This residue is present within a conserved primary sequence that showed consensus sites for phosphorylation by p34cdc2 kinase (cdc2K) and mitogen-activated protein kinase (MAPK). In vitro experiments with hexahistidine-tagged pORF3 expressed either in Escherichia coli or in COS-1 cells showed efficient phosphorylation with exogenously added MAPK. The pORF3 mutants also exhibited an in vitro phosphorylation profile with MAPK which was identical to that observed in vivo. In its primary sequence, pORF3 possesses two highly hydrophobic N-terminal domains. On subcellular fractionation, pORF3 was found to partition with the cytoskeletal fraction, and this association with the cytoskeleton was lost on deletion of hydrophobic domain I (amino acid residues 1 to 32). These results suggest that HEV pORF3 is a cytoskeleton-associated phosphoprotein and are discussed in terms of a possible function for pORF3 within the HEV replicative cycle.
- [show abstract] [hide abstract]
ABSTRACT: More than 50% of acute viral hepatitis occurring in some developing countries appears to be unrelated to infection by HAV or HBV and accumulating evidence suggests that a high proportion of this non-A, non-B hepatitis (NANB) is enterically transmitted. Epidemics or outbreaks of enterically-transmitted NANB (ET-NANB) have been documented in the Soviet Union, Nepal, Burma, Pakistan, India, Borneo, Somalia, Sudan, Ivory Coast, Algeria, and Mexico. These outbreaks primarily affect young to middle-age adults and are often associated with a high mortality rate in infected pregnant women, approaching 20% in most reported epidemics. Several investigators have reported finding 27 to 34 nm virus-like particles (VLPs) in stools of acutely infected cases. Stools containing these small, non-enveloped VLPs have been shown to cause NANB in experimentally infected cynomolgus macaques, African green monkeys, chimpanzees, rhesus monkeys, and Saguinus mystax monkeys (tamarins). Infected primates have been shown to seroconvert to 27-34 nm VLPs recovered from stools of cases occurring in the Soviet Union, India, Nepal, Burma, Pakistan and/or Mexico, suggesting that ET-NANB is caused by one virus or class of serologically related viruses. The morphological features and physicochemical properties of one candidate virus are very similar to those of some human caliciviruses, a group of viruses that is normally associated with outbreaks of severe diarrhoea.British Medical Bulletin 05/1990; 46(2):442-61. · 4.36 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: PD 098059 has been shown previously to inhibit the dephosphorylated form of mitogen-activated protein kinase kinase-1 (MAPKK1) and a mutant MAPKK1(S217E,S221E), which has low levels of constitutive activity (Dudley, D. T., Pang, L., Decker, S. J., Bridges, A. J., and Saltiel, A. R. (1995) Proc. Natl. Acad. Sci. U.S.A. 92, 7686-7689). Here we report that PD 098059 does not inhibit Raf-activated MAPKK1 but that it prevents the activation of MAPKK1 by Raf or MEK kinase in vitro at concentrations (IC50 = 2-7 microM) similar to those concentrations that inhibit dephosphorylated MAPKK1 or MAPKK1(S217E,S221E). PD 098059 inhibited the activation of MAPKK2 by Raf with a much higher IC50 value (50 microM) and did not inhibit the phosphorylation of other Raf or MEK kinase substrates, indicating that it exerts its effect by binding to the inactive form of MAPKK1. PD 098059 also acts as a specific inhibitor of the activation of MAPKK in Swiss 3T3 cells, suppressing by 80-90% its activation by a variety of agonists. The high degree of specificity of PD 098059 in vitro and in vivo is indicated by its failure to inhibit 18 protein Ser/Thr kinases (including two other MAPKK homologues) in vitro by its failure to inhibit the in vivo activation of MAPKK and MAP kinase homologues that participate in stress and interleukin-1-stimulated kinase cascades in KB and PC12 cells, and by lack of inhibition of the activation of p70 S6 kinase by insulin or epidermal growth factor in Swiss 3T3 cells. PD 098059 (50 microM) inhibited the activation of p42MAPK and isoforms of MAP kinase-activated protein kinase-1 in Swiss 3T3 cells, but the extent of inhibition depended on how potently c-Raf and MAPKK were activated by any particular agonist and demonstrated the enormous amplification potential of this kinase cascade. PD 098059 not only failed to inhibit the activation of Raf by platelet-derived growth factor, serum, insulin, and phorbol esters in Swiss 3T3 cells but actually enhanced Raf activity. The rate of activation of Raf by platelet-derived growth factor was increased 3-fold, and the subsequent inactivation that occurred after 10 min was prevented. These results indicate that the activation of Raf is suppressed and that its inactivation is accelerated by a downstream component(s) of the MAP kinase pathway.Journal of Biological Chemistry 12/1995; 270(46):27489-94. · 4.65 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: We demonstrate that members of the erk-encoded family of mitogen-activated protein (MAP) kinases (pp44/42mapk/erk) and members of the rsk-encoded protein kinases (RSKs or pp90rsk) are present in the cytoplasm and nucleus of HeLa cells. Addition of growth factors to serum-deprived cells results in increased tyrosine and threonine phosphorylation and in the activation of cytosolic and nuclear MAP kinases. Activated MAP kinases then phosphorylate (serine/threonine) and activate RSKs. Concurrently, a fraction of the activated MAP kinases and RSKs enter the nucleus. In addition, a distinct growth-regulated RSK-kinase activity (an enzyme[s] that phosphorylates recombinant RSK in vitro and that may be another member of the erk-encoded family of MAP kinases) was found associated with a postnuclear membrane fraction. Regulation of nuclear MAP kinase and RSK activities by growth factors and phorbol ester is coordinate with immediate-early gene expression. Indeed, in vitro, MAP kinase and/or RSK phosphorylates histone H3 and the recombinant c-Fos and c-Jun polypeptides, transcription factors phosphorylated in a variety of cells in response to growth stimuli. These in vitro studies raise the possibility that the MAP kinase/RSK signal transduction pathway represents a protein-Tyr/Ser/Thr phosphorylation cascade with the spatial distribution and temporal regulation that can account for the rapid transmission of growth-regulating information from the membrane, through the cytoplasm, and to the nucleus.Molecular and Cellular Biology 04/1992; 12(3):915-27. · 5.37 Impact Factor
JOURNAL OF VIROLOGY,
Copyright © 1997, American Society for Microbiology
Dec. 1997, p. 9045–9053Vol. 71, No. 12
The ORF3 Protein of Hepatitis E Virus Is a Phosphoprotein
That Associates with the Cytoskeleton
MOHAMMAD ZAFRULLAH,1MEHMET HAKAN OZDENER,1†
SUBRAT KUMAR PANDA,2AND SHAHID JAMEEL1*
Virology Group, International Centre for Genetic Engineering and Biotechnology,1
and Department of Pathology, All India Institute of
Medical Sciences,2New Delhi, India
Received 31 March 1997/Accepted 20 August 1997
Hepatitis E virus (HEV) is a major human pathogen in the developing world. In the absence of an in vitro
culture system, very little information exists on the basic biology of the virus. A small protein (?13.5 kDa) of
unknown function, pORF3, is encoded by the third open reading frame of HEV. We expressed pORF3 in
transiently transfected COS-1 and Huh-7 cells and showed that it is a phosphoprotein which is modified at a
serine residue(s). Deletion and site-directed mutants were created to establish Ser-80 as the phosphorylation
site. This residue is present within a conserved primary sequence that showed consensus sites for phosphor-
ylation by p34cdc2kinase (cdc2K) and mitogen-activated protein kinase (MAPK). In vitro experiments with
hexahistidine-tagged pORF3 expressed either in Escherichia coli or in COS-1 cells showed efficient phosphor-
ylation with exogenously added MAPK. The pORF3 mutants also exhibited an in vitro phosphorylation profile
with MAPK which was identical to that observed in vivo. In its primary sequence, pORF3 possesses two highly
hydrophobic N-terminal domains. On subcellular fractionation, pORF3 was found to partition with the
cytoskeletal fraction, and this association with the cytoskeleton was lost on deletion of hydrophobic domain I
(amino acid residues 1 to 32). These results suggest that HEV pORF3 is a cytoskeleton-associated phospho-
protein and are discussed in terms of a possible function for pORF3 within the HEV replicative cycle.
Hepatitis E virus (HEV), the causative agent of hepatitis E,
is a waterborne pathogen which is responsible for sporadic
infections as well as large epidemics of acute viral hepatitis in
developing countries (6, 22, 33, 41). In developed countries,
this disease is primarily seen in travellers visiting areas of
endemicity. Although largely a self-limited infection, it results
in significant morbidity and mortality, especially among preg-
nant women (23). In areas of endemicity, about 30% of acute
sporadic hepatitis and ?80% of epidemic hepatitis are due to
infection by HEV.
The HEV genome has been cloned and sequenced from at
least five distinct geographic isolates and shows a high degree
of nucleotide and amino acid sequence conservation (4, 5, 18,
open reading frames (ORFs) (38). Of these, the N-terminal
ORF, ORF1, of about 5 kb, is predicted to code for the puta-
tive nonstructural proteins, and domains homologous to those
dependent RNA polymerases are observed within this ORF (25).
The C-terminal structural region of the genome possesses two
ORFs: ORF2, encoding the major HEV structural protein, and
ORF3, encoding a small protein of undefined function.
HEV has provisionally been classified as a Hepevirus, a new
genus in the family Calciviridae (27), based on morphological
features and a broad similarity in genome organization. It has
also been suggested that HEV is a nonenveloped “alpha-like”
virus (25, 34). This suggestion is based on sequence similarities
between HEV ORF1 and the alphaviral junction sequences as
well as a 51-nucleotide conserved sequence present in the non-
structural region of the alphaviral genome which folds into stem-
loop structures and is known to bind cellular proteins (29). In its
genome sequences and codon usage, HEV most closely resem-
bles rubella virus, which is classified as a togavirus, and beet
necrotic yellow vein virus (14), a plant furovirus. It has been
proposed that the taxonomy of positive-stranded RNA viruses
should be reorganized to include rubella virus, beet necrotic
yellow vein virus, and HEV in separate but related families
(24). A conclusive classification of HEV awaits further knowl-
edge of its expression and replication strategy and of the na-
ture, processing, and properties of its component proteins.
In the absence of a reliable culture system for HEV, funda-
mental studies on its replication and expression strategy have
not been undertaken. In the past, we expressed the structural
region of HEV in animal cells and in vitro in a coupled tran-
scription-translation system to characterize pORF2 and
pORF3, proteins encoded by ORF2 and ORF3, respectively
(20). In that study, pORF2 was found to be an 88-kDa glyco-
protein which was expressed intracellularly as well as on the
cell surface and showed the potential to form homodimers.
pORF2 was also found to be synthesized as a precursor which
was processed through signal sequence cleavage into the ma-
ture protein. pORF3 was found to be a 13.5-kDa nonglycosy-
lated protein which was expressed intracellularly and showed
no major processing. We now show that pORF3 is a cytoskel-
eton-associated phosphoprotein and present results on map-
ping the phosphorylation site within its primary sequence.
These findings are discussed in terms of a possible function for
pORF3 within the HEV replicative cycle.
MATERIALS AND METHODS
Plasmid constructs. The expression vector pSG-ORF3 has already been de-
scribed (20). Plasmid pMT-ORF3 was constructed by subcloning a ?0.5-kb
SmaI-KpnI fragment from plasmid pSG-ORF3 into the same restriction sites in
* Corresponding author. Mailing address: Virology Group, ICGEB,
P.O. Box 10504, Aruna Asaf Ali Marg, New Delhi 110067, India.
Phone: 91-11-6176680. Fax: 91-11-6162316. E-mail: shahid@icgebnd
† Present address: Department of Biochemistry, Medical Faculty, 19
Mayis University, Samsun, Turkey.
plasmid pMT3 (37). This fragment contains the entire HEV ORF3 under the
control of the adenovirus major late promoter, tripartite leader, and intervening
sequence and the simian virus 40 (SV40) polyadenylation site. The vector also
carries the SV40 origin of replication, a dihydrofolate reductase marker gene,
and the adenovirus VAI sequences (37). Plasmid pSG-(His6)-ORF3 was con-
structed by subcloning a ?0.6-kb NdeI (blunt-ended)-HindIII fragment from
plasmid pRSET-ORF3 (30) into SmaI-HindIII-restricted plasmid pSGI (20).
Plasmid pMT-(His6)-ORF3 was constructed by subcloning a ?0.6-kb EcoRI
(blunt-ended)-XhoI fragment from plasmid pSG-(His6)-ORF3 into SmaI-XhoI-
restricted plasmid pMT3. Both pSG-(His6)-ORF3 and pMT-(His6)-ORF3 ex-
pression vectors contain the entire HEV ORF3 sequence fused in frame with a
42-amino-acid N-terminal peptide containing a hexahistidine tag and a phage T7
s10 protein epitope from plasmid pRSET-B (Invitrogen).
The ORF3 deletion mutants used in this study were generated by PCR am-
plification with the primers shown below. The PCR fragments were cloned into
either the pBS? (Stratagene) or the PCR-Script (Stratagene) vector, and their
sequences were confirmed. The mutant ORF3 was then subcloned into the pSGI
(20) and pMT3 (37) vectors with or without N-terminal hexahistidine and T7
s10 protein epitope fusions by means of relevant restriction sites on the
amplification primers or within the multiple cloning sites of the cloning
vector. The primers used for PCR amplification of various mutants were as
follows, with restriction sites (italic type) and start or stop codons (underlin-
ing) indicated. For the N-terminal deletion of 32 amino acid residues [ORF3
(?1-32)], primers 3-33(?), 5?-CGGAATTCCATGGGTCGCCACCGCCCTGT
CAGCCGTC-3?, and 3-123(?), 5?-CTAAAGCTTATTAGCGGCGCGGCCCC
AGCTGT-3?, were used. For the C-terminal deletions of various amino acid
residues, a common upstream primer and different downstream primers were
used. These primers were 3-1(?), 5?-CGGGATCCCATGGATAACATGTCTT
TTGCTGCG-3?; 3-274(?), 5?-CTAAAGCTTATTAGTTGGCGAACACGAG
GTCCAG-3? [ORF3, ?(92-123)]; 3-231(?), 5?-CTAAAGCTTATTATCGGGG
CGAAGGGGTTGGTTG-3? [ORF3, ?(78-123)]; and 3-186(?), 5?-CTAAAGC
TTATTAGAGAATCAACCCGGTCACCCC-3? [ORF3, ?(63-123)]. Details of
the exact amplification conditions and subcloning steps will be provided upon
Site-directed mutagenesis of pORF3 was carried out according to published
methods (43). The ORF3 mutagenic oligonucleotide used for annealing and
first-strand synthesis was 5?-CCCGATGGCACCGCTGCGGCC-3?, in which a
single T3G change was used to convert a serine codon (TCA) into an alanine
codon (GCA; underlined). Mutagenesis was carried out with the pSG-ORF3
plasmid, and a mutant clone, called pSG-ORF3(S80A), was identified by DNA
sequencing. Plasmid pMT-ORF3(S80A) was generated by subcloning a SmaI-
KpnI fragment of plasmid pSG-ORF3(S80A) into the same sites in plasmid
Transfection and labeling of cultured cells. COS-1 and Huh-7 cells were
maintained, transfected, and metabolically labeled as described earlier (20) with
minor modifications. Each 60-mm plate of cells was labeled in 1 ml of minimal
essential medium (GIBCO-BRL or Sigma) lacking cysteine and methionine with
100 to 150 ?Ci of either35S-Promix (?1,000 Ci/mmol; Amersham) or35S-
Translabel (?1,000 Ci/mmol; Dupont NEN). For phosphate labeling, at 40 to
44 h posttransfection, cells on 60-mm plates were washed once with phosphate-
deficient Dulbecco minimal essential medium (GIBCO-BRL) and incubated in 3
ml of deficient medium for 1 h. Following this step, each plate was labeled for 4 h
in a 37°C CO2incubator with 250 ?Ci of [32P]orthophosphate (Amersham or
Dupont NEN) in 1 ml of deficient medium. When protein kinase inhibitors were
used, these were diluted from a 100-fold stock in dimethyl sulfoxide and were
present throughout the 1-h prelabeling and 4-h labeling periods.
Immunoprecipitation. Transfected and phosphate-buffered saline-washed
cells were harvested in radioimmunoprecipitation (RIPA) buffer and immuno-
precipitated as described earlier (20). The antibodies used for immunoprecipi-
tations included a rabbit polyclonal antibody against Escherichia coli-expressed
and purified HEV pORF3 (30) and a commercially available T7 ? Tag monoclo-
nal antibody specific for the amino-terminal end of the phage T7 s10 gene
(Novagen). Immunoprecipitates were analyzed on sodium dodecyl sulfate
(SDS)–15% polyacrylamide gels as described earlier (20). Quantitation of auto-
radiographic signals was carried out following densitometric scanning with NIH
Phosphatase treatment. Lysates from cells transfected with the ORF3 expres-
sion vectors and labeled with either35S or32P were subjected to immuno-
precipitation with rabbit polyclonal anti-pORF3 as described above. The
immunoprecipitates were washed once in 250 ?l of lambda protein phospha-
tase (?-PPase) reaction buffer (50 mM Tris-HCl [pH 7.8], 5 mM dithiothreitol,
2 mM MnCl2, 100 ?g of bovine serum albumin per ml). The washed immuno-
precipitates were resuspended in 50 ?l of ?-PPase reaction buffer with or without
1 ?l of ?-PPase (400,000 U/ml; New England BioLabs) and incubated for 1 h at
37°C. The beads were centrifuged down, resuspended in SDS-polyacrylamide gel
electrophoresis (PAGE) loading buffer, and analyzed by SDS–15% PAGE fol-
lowed by autoradiography.
Phosphoamino acid analysis. Phosphoamino acids were analyzed according to
Duclos et al. (11). Lysates of COS-1 cells transfected with the pMT-ORF3 vector
were subjected to immunoprecipitation with rabbit polyclonal anti-pORF3 as
described earlier (20). Following autoradiography, pORF3 bands were excised
from the gels and hydrolyzed in 6 N HCl for 2 h at 110°C. Hydrolysates were
dried under vacuum and dissolved in pH 1.9 buffer (formic acid-glacial acetic
acid-water, 44:156:1,800) containing phosphoserine, phosphothreonine, and
phosphotyrosine (3 ?g each) as standards. Samples were applied to 20- by 20-cm
Polygram Cell300 thin-layer plates (Sigma-Aldrich) and electrophoresed in pH
1.9 buffer at 1,600 V for 80 min. After the plates were dried, a second dimension
of electrophoresis was performed in pH 3.5 buffer (pyridine-glacial acetic acid-
water, 10:100:1,890) at 1,200 V for 60 min. The plates were dried, sprayed with
0.25% ninhydrin (in acetone) to visualize the phosphoamino acid standards, and
exposed to X-ray film for autoradiography.
In vitro phosphorylation. For in vitro phosphorylation of pORF3, two ap-
proaches were taken. pORF3 was either immunoprecipitated from transfected
COS-1 cells or purified from an E. coli expression system (30) and subjected to
in vitro phosphorylation in the presence of [?-32P]ATP (3,000 Ci/mmol; Dupont
NEN) and an exogenously added protein kinase.
Transfected COS-1 cells were lysed in 1? lysis buffer (10 mM Tris-HCl [pH
8.0], 140 mM NaCl, 0.5% Triton X-100, 5 mM iodoacetamide, 2 mM phenyl-
methylsulfonyl fluoride [PMSF]), and immunoprecipitations were carried out as
described earlier (20). The beads were washed twice with 125 ?l each of 1?
kinase buffer (50 mM Tris-HCl [pH 7.5], 10 mM MgCl2, 1 mM EDTA, 2 mM
dithiothreitol, 0.05% Triton X-100, 2 mM PMSF, 1 mM Na3VO4, 5 mM NaF).
The washed beads were resuspended in 30 ?l of a mixture that contained 1?
kinase buffer, 10 ?Ci of [?-32P]ATP, and 1 ?l of either p34cdc2/cyclin B protein
kinase (cdc2K) (5,000 U/ml; New England BioLabs) or mitogen-activated pro-
tein kinase (MAPK) (50,000 U/ml; New England BioLabs). After incubation at
30°C for 1 h, the beads were washed twice each with 500 ?l of 1? lysis buffer
containing 1 mM ATP, resuspended in 40 ?l of SDS-PAGE loading buffer, and
analyzed on an SDS–15% polyacrylamide gel. Controls included immunopre-
cipitated pORF3 without any exogenously added kinase and immunoprecipitated
lysates from mock-transfected cells.
E. coli BL21(DE3) cells expressing wild-type or mutant forms of pORF3 or
control cells were grown and induced with isopropyl-?-D-thiogalactopyranoside
(IPTG) as described earlier (30). Cell pellets were resuspended at a concentra-
tion of 0.2 g/ml in 1? binding buffer (20 mM Tris-HCl [pH 7.9], 500 mM NaCl,
5 mM imidazole, 2 mM PMSF), and the cells were lysed by sonication. After
centrifugation for 10 min at 10,000 ? g and 4°C in a microcentrifuge, the pellets
were washed with 1? binding buffer containing 2 M urea for 30 min on ice. The
washed pellets were resuspended in 1? binding buffer containing 6 M urea and
incubated on ice for 30 min. One milliliter of the supernatant was incubated with
0.2 ml of settled volume of Ni-NTA resin (Qiagen) preequilibrated with 1?
binding buffer containing 6 M urea. Following binding at room temperature for
2 h with shaking, the unbound proteins were removed and the resin was washed
sequentially at room temperature for 15 min each time, twice with 1? binding
buffer containing 0.5% Triton X-100 and twice with 1? wash buffer (20 mM
Tris-HCl [pH 7.9], 500 mM NaCl, 20 mM imidazole, 2 mM PMSF) containing
0.5% Triton X-100. By SDS-PAGE and Coomassie blue staining, the washed
beads were estimated to contain approximately 1 to 2 mg of protein bound per
ml of resin. In vitro phosphorylation with cdc2K, MAPK, protein kinase A (PKA;
100 U/?l; Promega), or protein kinase C (PKC; 50 ?g/ml; Promega) or auto-
phosphorylation was carried out as described above with resin containing 5 to 10
?g of bound protein from pORF3-expressing or control E. coli BL21(DE3) cells.
For PKA and PKC, 1? phosphorylation buffer contained 20 mM HEPES (pH
7.5), 1 mM EGTA, 10 mM MgCl2, 2 mM PMSF, 1 mM Na3VO4, and 5 mM NaF.
Subcellular fractionation. COS-1 cells on 100-mm plates were transfected and
labeled as described earlier (20), except that the amounts of DNA, Lipofectin,
and the isotope as well as all volumes were increased three times over that used
for 60-mm plates. Subcellular fractionation of transfected cells was carried out
essentially as described elsewhere (28). Labeled cells were washed twice with
phosphate-buffered saline and incubated in 1 ml of hypotonic buffer (10 mM
HEPES [pH 7.5], 10 mM KCl) per 100-mm plate for 15 min on ice. Swollen cells
were scraped off the plates and disrupted by 20 strokes of Dounce homogeniza-
tion. This step resulted in approximately 80% cell breakage, which was confirmed
microscopically. The cell lysate was adjusted to 150 mM NaCl. Unbroken cells,
cell debris, and nuclei were removed by centrifugation for 20 min at 1,500 rpm
and 4°C in a microcentrifuge (Biofuge 17RS; Heraeus). The supernatant was
centrifuged for 60 min at 32,500 rpm (100,000 ? g) and 4°C in an SW60 rotor
(Beckman). The supernatant was removed and saved as the cytosolic fraction.
The pellet was resuspended in 1 ml of NTENT buffer (10 mM Tris-HCl [pH 8.0],
150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1% Nonidet P-40, 2 mM PMSF,
1 mM Na3VO4, 5 mM NaF) and ultracentrifuged again as described above. The
supernatant was saved as the membrane fraction. The cytoskeletal pellet was
washed once with 1 ml of NTENT buffer and resuspended in either 1 ml of RIPA
buffer for immunoprecipitation or 1 ml of NTENT buffer for enzymatic assays.
Alternatively, a total-cell lysate was prepared by directly adding RIPA buffer to
the cells and harvesting after 30 min on ice. All four fractions were immunopre-
cipitated with rabbit polyclonal anti-pORF3 as described earlier (20).
The various fractions were also characterized on the basis of the distribution
of marker proteins. The cytosolic enzyme lactate dehydrogenase (LDH) and the
membrane-associated enzyme 5? nucleotidase (5?ND) were used as markers for
these fractions and were assayed with commercially available kits (Sigma). Tu-
bulin was used as a marker for the skeletal fraction and was estimated by the
immunoprecipitation of a 55-kDa polypeptide from35S-labeled cells with an
antitubulin antibody (Sigma) as described above for pORF3.
9046ZAFRULLAH ET AL.J. VIROL.
HEV pORF3 is a phosphoprotein. The vectors pSG-ORF3
and pMT-ORF3 express HEV pORF3 from the SV40 control
elements and the adenovirus major late promoter, respectively.
These plasmids also contain the SV40 ori sequences for high-
copy-number maintenance in T-antigen-producing monkey
kidney COS-1 cells. When COS-1 cells transiently transfected
with either of these vectors were labeled with [32P]orthophos-
phate, a protein of about 13.5 kDa was immunoprecipitated
with rabbit anti-pORF3 antiserum (Fig. 1A, lanes 1 and 3).
That this protein was pORF3 was ascertained by its absence
from cells mock transfected with the pSGI vector (Fig. 1A, lane
2) as well as from cell lysates immunoprecipitated with preim-
mune rabbit serum (data not shown). Another species, of ?26
to 28 kDa, was also found to be labeled and immunoprecipi-
tated from ORF3-transfected cells. This protein was observed
earlier as well (20) and represents a highly stable dimeric form
of pORF3. As another control, the same [32P]orthophosphate
labeling experiment was also carried out on COS-1 cells trans-
fected with the ORF2 expression vector (20). On immunopre-
cipitation with specific antisera, pORF2 was found not to be
phosphorylated (data not shown). pORF3 was also found to be
expressed and phosphorylated in Huh-7 human hepatoma cells
(Fig. 1A, lanes 5 and 6). Since pORF3 expression levels were
consistently higher with the pMT-ORF3 vector in COS-1 cells
than with the pSG-ORF3 vector or in Huh-7 cells, the former
combination was used for most of the studies described in this
To further establish the phosphorylation of pORF3, lysates
of pMT-ORF3-transfected and labeled COS-1 cells were im-
FIG. 1. HEV pORF3 is a phosphoprotein. (A) Expression vectors pSG-
ORF3 (lanes 1 and 5) and pMT-ORF3 (lane 3) or one of the pSGI parent vectors
(lanes 2 and 6) was transiently transfected into COS-1 or Huh-7 cells. Trans-
fected cells were labeled with [32P]orthophosphate, lysed, and immunoprecipi-
tated with anti-pORF3 antiserum as described in the text. Washed immunopre-
cipitates were separated by SDS–15% PAGE and visualized by autoradiography.
Molecular weight (Mol Wt) markers (lane 4, bottom to top) are 14,400, 21,500,
30,000, 46,000, 69,000, 97,400, and 200,000. (B) COS-1 cells transfected with
pMT-ORF3 were labeled with either [35S]methionine-cysteine (lane 1) or
[32P]orthophosphate (lanes 2 and 3), lysed, and immunoprecipitated with anti-
pORF3 antiserum. The washed immunoprecipitates were treated with ?-PPase
(lanes 1 and 2) or buffer alone (lane 3) as described in the text, separated by
SDS–15% PAGE, and visualized by fluorography. Molecular size markers are
indicated (lane 4).
FIG. 2. Phosphoamino acid mapping of pORF3. COS-1 cells transfected with
pMT-ORF3 were labeled with [32P]orthophosphate, and immunoprecipitated
pORF3 was separated and subjected to acid hydrolysis as described in the text.
The32P-labeled acid hydrolysate and unlabeled phosphoamino acid standards
were spotted on a thin-layer chromatography plate and subjected to two-dimen-
sional (2D) electrophoresis as described in the text. Phosphoamino acid stan-
dards were visualized by staining with ninhydrin, and the32P-labeled amino acids
were visualized by autoradiography. The directions of electrophoresis and the
positions of the phosphoamino acid standards are indicated. 1D, one dimen-
VOL. 71, 1997 HEPATITIS E VIRUS ORF3 PROTEIN9047
FIG. 3. Mapping of the pORF3 phosphorylation site. (A) Structure of pORF3 showing the N-terminal hydrophobic domains (I and II) and the C-terminal
immunodominant region. The serine residues conserved in all five isolates of HEV are indicated with closed circles (F) and numbered as in the primary sequence of
pORF3 (38). The amino acid sequence around Ser-80 is also shown. The lower part of panel A shows the boundaries of the pORF3 deletion mutants, numbered
according to the residues deleted, as well as the Ser-803Ala mutant. The phosphorylation results shown for these mutants in panels B and C are also summarized.
(B) Expression and phosphorylation of pORF3 and its deletion mutants. COS-1 cells were transfected with pMT-ORF3 (lane 2), pMT-(His6)-ORF3 (lanes 6 and 11),
pMT-ORF3 [?(1-32)] (lane 3), pMT-(His6)-ORF3 [?(92-123)] (lane 7), pMT-(His6)-ORF3 [?(63-123)] (lane 8), pMT-(His6)-ORF3 [?(78-123)] (lane 12), or the pMT3
vector (lanes 4 and 9). Cells were labeled with [35S]methionine-cysteine (top panels) or [32P]immunoprecipitated with anti-pORF3 antiserum, separated, and visualized
9048 ZAFRULLAH ET AL.J. VIROL.
munoprecipitated, and the protein was subjected to treatment
with ?-PPase. The results are shown in Fig. 1B. While ?-PPase
completely removed the isotopic label from32P-labeled pORF3
(lane2),thelabelwasstillretainedafteranidentical treatment of
35S-labeled pORF3 (lane 1). As a control, mock incubation of
32P-labeled pORF3 under identical conditions but without ?-
PPase also did not result in any loss of the isotopic label (lane
3). This result clearly establishes that HEV pORF3 is a phos-
pORF3 is phosphorylated at a serine residue. To establish
the nature of pORF3 phosphorylation, the protein was sub-
jected to phosphoamino acid analysis as described in Materials
and Methods. Following limited acid hydrolysis, two-dimen-
sional electrophoresis, and comparison with standards, pORF3
was found to be phosphorylated at one or more serine residues
(Fig. 2). There was no evidence of phosphorylation at threo-
nine residues, and the protein was devoid of any tyrosine res-
In its primary sequence, pORF3 contains at least eight
serine residues that are completely conserved across all five
HEV isolates sequenced thus far. One serine residue, Ser-80, is
found in four of the five HEV isolates, being absent from the
most divergent Mexican isolate (18). To determine the serine
residue(s) in the primary sequence of pORF3 that was phos-
phorylated, N- and C-terminal ORF3 deletion mutants were
constructed in pSG as well as pMT vector backgrounds. These
are shown schematically in Fig. 3A. For all C-terminal pORF3
deletion mutants, the expression construct was designed to
contain an N-terminal fusion of a hexahistidine tag as well as a
phage T7 gene 10 epitope. This design was used to enable the
immunoprecipitation of C-terminally deleted versions of pORF3,
since our rabbit polyclonal antibody predominantly recognizes
the immunodominant region encompassing residues 90 to 123
and fails to either immunoprecipitate or Western blot C-ter-
minally deleted pORF3 (19). The results of COS-1 transient
transfection analysis of all the mutants are shown in Fig. 3B.
While all the mutants showed expression based on [35S]methi-
onine-cysteine labeling (Fig. 3B, top panels), only the N-ter-
minal deletion mutant ?(1-32) and the C-terminal deletion
mutant ?(92-123) showed any phosphorylation based on
[32P]orthophosphate labeling (Fig. 3B, bottom panels). The
C-terminal deletion mutants ?(63-123) and ?(78-123) showed
expression but not phosphorylation of pORF3 (Fig. 3B, lanes
8 and 12). This analysis of deletion mutants suggested that in
pORF3, the residue that is phosphorylated is Ser-80.
To conclusively establish this idea, site-directed mutagenesis
in which Ser-80 was changed to an alanine (Ala) residue was
carried out. When analyzed in COS-1 cells, this Ser3Ala mu-
tant showed expression but not phosphorylation (Fig. 3C). This
finding proved beyond a doubt that in pORF3, the only residue
that is phosphorylated is Ser-80. The wild-type and mutant
pORF3 phosphorylation results are summarized in Fig. 3A.
pORF3 phosphorylating kinase(s). To understand the path-
way(s) involved in pORF3 phosphorylation, various kinase in-
hibitors were used. COS-1 cells transfected with pMT-ORF3
were metabolically labeled with either [35S]methionine-cys-
teine (for protein expression) or [32P]orthophosphate (for
phosphorylation), and the expressed pORF3 was immunopre-
cipitated and quantified by densitometric scanning. Neither
1-(5-isoquinolinesulfonyl)-2-methylpiperazine (H7) nor Cal-
phostin, specific inhibitors of the PKC pathway, was able to
inhibit the phosphorylation of pORF3 in COS-1 cells (data not
shown). Staurosporine, a general inhibitor of the cyclic AMP-
dependent protein kinase (PKA), PKC, and protein tyrosine
kinase pathways, was also unable to inhibit pORF3 phosphor-
ylation up to a concentration of 100 nM (data not shown).
The phosphorylated residue, Ser-80, was found to be within
a sequence that also contains overlapping recognition motifs
for MAPK and cdc2K. To check if either of these kinases was
involved in pORF3 phosphorylation, the protein expressed in
E. coli was immobilized on Ni-NTA resin and subjected to in
vitro phosphorylation with purified kinases in the presence of
[?-32P]ATP. The results are shown in Fig. 4A. Both MAPK
(lanes 1 and 2) and cdc2K (lanes 3 and 4) phosphorylated
pORF3 in vitro. In the same experiment, pORF3 was found
not to possess any autophosphorylating activity (lanes 5 and 6).
Based on autoradiographic signals, MAPK was found to phos-
phorylate pORF3 about 80 to 100 times more efficiently than
cdc2K. When the input amount of pORF3 was reduced to 1 ?g
instead of the 10 ?g used in Fig. 4A, only MAPK was found to
phosphorylate it; no in vitro phosphorylation was observed
with cdc2K (data not shown).
In a similar in vitro experiment, two other kinases, PKA and
PKC, were found not to phosphorylate pORF3 (Fig. 4B). Both
of these kinases, however, phosphorylated other E. coli-de-
rived proteins present in the partially purified pORF3 prepa-
ration immobilized on Ni-NTA resin.
An identical in vitro phosphorylation experiment with [?-
32P]ATP and purified kinases was also performed with pORF3
expressed in COS-1 cells and immunoprecipitated with anti-
histidine tag (anti-His-tag) antibodies. The results presented in
Fig. 4C show that compared to a control of immunoprecipi-
tated material from mock-transfected cells, pORF3 was effi-
ciently phosphorylated in vitro by MAPK (lanes 4 and 5).
Although not visible in the exposure shown in Fig. 4B, a weak
signal was also observed with cdc2K on longer exposures. How-
ever, it was also present in control lanes, albeit at an intensity
lower than that observed with pORF3. This same band, moving
slightly faster than the differentially phosphorylated pORF3
band (lane 4), was also present in the control phosphorylated
with MAPK and therefore most likely represents nonspecific
phosphorylation of a cellular protein.
Other experiments were also carried out to further establish
MAPK as the pORF3-phosphorylating kinase. An inhibitor
of MEK1 (MAPK kinase) phosphorylation, the enzyme that
phosphorylates and thereby activates MAPK in cells (1, 12),
was used to inhibit pORF3 phosphorylation. Even at a 100 ?M
concentration, this inhibitor, PD098059 (New England Bio-
Labs), failed to inhibit pORF3 phosphorylation (data not
shown). However, when the status of MAPK phosphorylation
was evaluated by Western blotting with anti-phospho-MAPK
antibodies (New England BioLabs), 10 to 20% of residual
phosphorylated MAPK was still found in COS-1 cells treated
with 100 ?M PD098059 (data not shown).
In another set of experiments, all the pORF3 mutants de-
scribed in Fig. 3A were phosphorylated in vitro with [?-32P]
ATP in the presence of exogenously added MAPK. In com-
plete agreement with the immunoprecipitation results (Fig.
as described in the text. Only molecular weight (Mol Wt) markers of 30,000 14,400 are shown (lanes 1, 5, and 10). (C) Expression and phosphorylation of pORF3 and
its Ser-803Ala mutant. COS-1 cells were transfected with pMT-ORF3 (lane 2), three independent clones of pMT-ORF3(S80A) (lanes 3 to 5), or the pMT3 vector
(lane 6). Cells were labeled with [35S]methionine-cysteine (top panel) or [32P]orthophosphate (bottom panel) and lysed, and the lysates were immunoprecipitated with
anti-pORF3 antiserum, separated, and visualized as described in the text. Only molecular weight markers of 21,500 and 14,300 are shown (lane 1).
VOL. 71, 1997 HEPATITIS E VIRUS ORF3 PROTEIN9049
3B), only deletion mutant ?(92-123) was phosphorylated,
while mutants ?(78-123) and ?(63-123), lacking Ser-80, were
not phosphorylated (Fig. 5A). Similarly, in agreement with
results shown earlier (Fig. 3C), the Ser3Ala mutant of pORF3
was also not phosphorylated in vitro by MAPK (Fig. 5B).
These results strongly suggest that in vivo, pORF3 is a sub-
strate for MAPK.
pORF3 associates with the cytoskeleton. Subcellular frac-
tionation of COS-1 cells transfected with pMT-ORF3 was car-
ried out to localize pORF3 within the cells. As shown in Fig. 6,
while pORF3 was found to be distributed in the cytosolic (lane
2) and membrane (lane 3) fractions, a majority of it was
present in the cytoskeletal fraction (lane 4). However, when
the N-terminal deletion mutant ?(1-32), lacking hydrophobic
domain I (Fig. 3A), was similarly evaluated, most of the
pORF3 was found in the cytosolic fraction (lane 7), some was
found in the membrane fraction (lane 8), and none was found
in the cytoskeletal fraction (lane 9). Even a higher exposure of
the autoradiograph than that presented in Fig. 6 showed the
absence of ?(1-32) pORF3 from the cytoskeletal fraction. To
ascertain faithful subcellular fractionation into the cytosolic,
membrane, and cytoskeletal fractions, marker proteins were
evaluated. These included LDH (cytosolic marker), 5?ND
(membrane marker), and tubulin (cytoskeletal marker). The
results presented in Table 1 show no significant contamination
of the skeletal fraction with cytosolic or membrane proteins.
While some tubulin could also be immunoprecipitated from
the cytosolic fraction, a majority of it was still found in the
skeletal fraction. Neither pORF3 nor its N-terminal deletion
mutant could be immunoprecipitated with antitubulin antibod-
ies (data not shown), suggesting the lack of a stable interaction
between pORF3 and ?- or ?-tubulin. In summary, these results
show that pORF3 associates with the cytoskeleton and that this
association requires N-terminal hydrophobic domain I of
Because of the inability to reliably grow HEV in culture,
molecular studies on its expression and replication mecha-
nisms, constituent proteins, the nature of viral proteins, and so
forth have not been carried out. We have been using a sub-
genomic expression strategy to address some of these issues,
particularly the properties of the viral structural proteins (20).
The minor protein, pORF3, encoded by ORF3 within the
structural region of the viral genome is a 123-amino-acid pro-
tein of unknown function (36). We had demonstrated earlier
FIG. 4. In vitro phosphorylation of pORF3. (A) Hexahistidine-tagged pORF3
(lanes 1, 3, and 5) expressed in E. coli was partially purified by binding to
Ni-NTA–agarose beads and subsequent washing as described in the text. Lysates
of BL21(DE3) cells processed identically served as controls (lanes 2, 4, and 6).
Beads containing ?10 ?g of protein were subjected to in vitro phosphorylation
with [?-32P]ATP in the presence of MAPK (lanes 1 and 2), cdc2K (lanes 3 and
4), or no exogenously added kinase (lanes 5 and 6), and proteins were analyzed
by SDS–15% PAGE followed by autoradiography. Lanes 1 and 2 were exposed
for 30 min, while the rest of the gel was exposed overnight (?20 h). (B) Hexa-
histidine-tagged pORF3 (lanes 1, 3, and 5) expressed in E. coli or control
BL21(DE3) lysates (lanes 2 and 4) were subjected to in vitro phosphorylation as
described for panel A in the presence of MAPK (lane 1), PKA (lanes 2 and 3),
or PKC (lanes 4 and 5). Labeled proteins were analyzed by SDS–15% PAGE
followed by autoradiography for 15 min. The position of pORF3 is indicated by
an arrow. (C) Lysates from COS-1 cells expressing hexahistidine-tagged pORF3
(lanes 1 and 4) or control cells (lanes 2 and 5) were immunoprecipitated with
anti-His-tag antibodies as described in the text. Beads containing immunopre-
cipitates were subjected to in vitro phosphorylation with [?-32P]ATP in the
presence of cdc2K (lanes 1 and 2) or MAPK (lanes 4 and 5), and proteins were
analyzed by SDS–15% PAGE followed by autoradiography. The gel was exposed
overnight (?20 h). Molecular size markers (lane 3, bottom to top) are 14.4, 30,
46, 69, 97.4, and 200 kDa. The position of pORF3 is indicated by an arrow.
9050ZAFRULLAH ET AL.J. VIROL.
(20) that pORF3 was localized to the cytoplasm of expressing
cells and that it was not posttranslationally modified in a man-
ner that would change its size to any significant degree, e.g.,
through glycosylation. We now present evidence that pORF3 is
a phosphoprotein, provide characterization of its phosphory-
lation site, and show the possible involvement of cellular ki-
nases in this modification.
When cells expressing pORF3 were labeled with [32P]ortho-
phosphate, the isotope was found to be present on pORF3. The
32P label but not the35S label on pORF3 could be removed with
?-PPase, further supporting the phosphorylation of pORF3 when
expressed in COS-1 cells. That the phosphorylation was not an
artifact of expression in COS-1 cells, which are SV40 T-antigen-
transformed monkey kidney cells, is borne out by the phosphor-
ylation of pORF3 in Huh-7 cells as well. The latter are human
hepatoma cells and more closely mimic the natural target cell for
HEV infection. The signals were better in COS-1 cells than in
Huh-7 cells, as the former provide a replicating background for
the SV40 ori-containing transfected expression plasmid. There-
fore, subsequent characterizations were carried out in the COS-1
The phosphorylated residue on pORF3 was found to be
serine. Deletion mutants were used to map this residue to
Ser-80 in the primary sequence of pORF3 (or Ser-79 for the
Indian isolate) (30). It is interesting to note that this serine
residue falls within a region that is highly conserved among all
isolates of HEV (4, 5, 30, 36, 38), except for the Mexican
isolate (18). In all HEV isolates but the latter, this sequence
tion motifs for PKA (R-X-S/T) (21), PKC (K/R-X-S/T) (21),
MAPK (P-X-S/T-P) (9), and cdc2K (S/T-P-X-R/K) (21). In the
Mexican isolate, this sequence reads
does not contain recognition motifs for any of these kinases.
Two inhibitors specific for PKC, H7, with a Kiof ?6 ?M
(16), and Calphostin, with a Kiof ?0.05 ?M (39), were unable
to inhibit pORF3 phosphorylation, even at concentrations up
to 20 times their reported Kivalues. A general inhibitor of the
major protein kinases, Staurosporine, at 50% inhibitory con-
78P/R-M-S-P-L-R83and contains overlapping recogni-
FIG. 5. In vitro phosphorylation of pORF3 mutants. (A) Control BL21(DE3)
lysates (lane 1), hexahistidine-tagged pORF3 (lane 2), or its deletion mutants
(lanes 3 to 5) expressed in E. coli were subjected to in vitro phosphorylation in
the presence of MAPK. The mutants used were ?(92-123) (lane 3), ?(78-123)
(lane 4), and ?(63-123) (lane 5). Labeled proteins were analyzed by SDS–15%
PAGE followed by autoradiography for 30 min. (B) In vitro phosphorylation of
pORF3 and its Ser-803Ala mutant. COS-1 cells were transfected with pMT-
ORF3 (lane 2), three independent clones of pMT-ORF3(S80A) (lanes 3 to 5), or
the pMT3 vector (lane 6). Cells were harvested and lysed in 1? lysis buffer, and
the lysates were immunoprecipitated with anti-pORF3 antiserum. The immuno-
precipitated material was phosphorylated in vitro with [?-32P]ATP and exog-
enously added MAPK as described in the text. Labeled proteins were separated
by SDS–15% PAGE and visualized by autoradiography. Only molecular weight
(Mol Wt) markers of 21,500 14,300 are shown (lane 1, bottom to top).
FIG. 6. Cytoskeletal association of pORF3. COS-1 cells were transfected
with either pMT-ORF3 (lanes 1 to 4) or pMT-ORF3 [?(1-32)] (lanes 6 to 9),
labeled with [35S]methionine-cysteine, lysed, and fractionated as described in the
text. Total lysates (lanes 1 and 6), the cytosolic fractions (lanes 2 and 7), the
membrane fractions (lanes 3 and 8), or the cytoskeletal fractions (lanes 4 and 9)
were obtained as described in the text. The fractions were immunoprecipitated
with either anti-pORF3 antiserum (top panel) or antitubulin antibodies (bottom
panel) and analyzed by SDS–15% PAGE or SDS–12% PAGE, respectively,
followed by fluorography. Molecular weight (MW) markers (lane 5, bottom to
top) corresponding to 14,300 and 21,500 are shown in the top panel, and those
corresponding to 46,000 and 66,000 are shown in the bottom panel. The activities
of the marker enzymes LDH and 5?ND, estimated with equal volumes of the
three subcellular fractions, are given in Table 1.
VOL. 71, 1997 HEPATITIS E VIRUS ORF3 PROTEIN9051
centrations (IC50s) of 2.7 nM, 6.4 nM, and 8.2 nM for PKC,
protein tyrosine kinase, and PKA, respectively (39), was also
unable to inhibit pORF3 phosphorylation, even at a concen-
tration of 100 nM. Direct in vitro experiments were carried out
on partially purified pORF3 expressed in bacteria to demon-
strate that while PKA and PKC did not phosphorylate pORF3,
MAPK and, to some extent, cdc2K were capable of phosphor-
ylating this HEV protein. We were also able to demonstrate
specific in vitro phosphorylation of pORF3 expressed in COS-1
cells by MAPK but not by cdc2K. The in vitro reactions were
orders of magnitude more efficient with pORF3 expressed in
E. coli than with the protein expressed in COS-1 cells. This
could be because in some fraction of pORF3 molecules ex-
pressed in COS-1 cells, Ser-80 may already be blocked by in
vivo phosphorylation. Alternatively, in vitro phosphorylation of
the E. coli-expressed protein may also be artifactual due to a
relatively high input in the phosphorylation reaction and the
presence of the protein in a nonnative conformation. In fact, in
other experiments, when only 1 ?g or less of E. coli-expressed
pORF3 was used, no in vitro phosphorylation could be de-
tected with cdc2K under conditions in which MAPK phosphor-
ylated it efficiently (data not shown). There is, however, always
the question of what is happening in vivo. The inability of
inhibitors of PKA and PKC to inhibit phosphorylation sug-
gested that these pathways may not be involved in pORF3
phosphorylation. This suggestion was further supported by the
lack of direct in vitro phosphorylation of pORF3 with exog-
enously added PKA or PKC.
It was observed that, in vitro, MAPK phosphorylated
pORF3 with an efficiency that was about 100 times greater
than that of cdc2K. We propose that, in vivo as well, it is
MAPK and not cdc2K that phosphorylates pORF3. This is
because Staurosporine inhibits cdc2K at an IC50of 3 to 6 nM
(26), and we were unable to observe any inhibition of pORF3
phosphorylation by Staurosporine. Further, in COS-1 cells, in
which the levels of p53 are high, it is expected that the levels of
p21, a cyclin-dependent kinase inhibitor (13), will also be high
and will inhibit the p34cdc2/cyclin B phosphorylation pathway
(42). Further, as discussed above, pORF3 immunoprecipitated
from COS-1 cells as well as reduced amounts of E. coli-ex-
pressed pORF3 were only phosphorylated in vitro by MAPK
and not by cdc2K. Finally, the in vitro phosphorylation pattern
of pORF3 deletion and site-directed mutants with MAPK was
identical to the in vivo phosphorylation results for these mu-
The classical MAPK pathway consists of a protein kinase
cascade that links extracellular growth and differentiation sig-
nals to modulation of cellular gene expression (3, 10). Two
isoforms of MAPK, p44 and p42 (Erk1 and Erk2), are located
at the cytoplasmic end of this cascade (3, 10) and are activated
on phosphorylation by one of two isoforms of MAPK kinase
(MEK1 and MEK2) (1, 12) at neighboring threonine and ty-
rosine residues in the sequence T-E-Y (8, 31). On phosphor-
ylation of MAPK, part of its cytoplasmic pool is translocated to
the nucleus for downstream signalling (7, 8). The inhibitor
PD098059 binds to inactive forms of MEK1, preventing its
phosphorylation and activation by upstream activators, such as
c-Raf, at an IC50of 5 to 10 ?M (1). Our inability to inhibit
pORF3 phosphorylation and to quantitatively block MAPK
phosphorylation with PD098059 concentrations as high as 100
?M suggests that residual MAPK phosphorylation may be due
to the MEK2 isoform. PD098059 is not known to inhibit the
MAPK-like enzymes also facilitate the exit of cells from the
cell cycle and regulate diverse cellular processes, even in ter-
minally differentiated cells (32). How HEV pORF3 fits into
this sequence is not clear, but the evidence presented here
strongly implicates MAPK in its phosphorylation and raises the
interesting possibility that pORF3 is phosphorylated in re-
sponse to specific extracellular growth signals. Whether this is
a regulatory signal for HEV replication or nucleocapsid assem-
bly remains to be seen.
MAPK phosphorylates its target proteins at serine and/or
threonine residues within the motif P-X-S/T-P or P-X-X-S/T-P
(2, 9). Several proteins are thought to be candidate targets for
MAPK, including nuclear transcription factors (e.g., Jun, Myc,
Tal1, and Elk1), other protein kinases (e.g., RSK, MEK1,
Raf1, Lck, and EGF-R), structural proteins (e.g., caldesmon,
tau, and neurofilament) (3, 10), and cytoskeletal proteins, (e.g.,
myelin basic protein and microtubule-associated protein 2) (8,
17, 35). Interestingly, although there are no published reports
so far, it has been suggested that pORF3 also associates with
the cytoskeleton (15).
In its N-terminal half, pORF3 appears to contain two hy-
drophobic domains which have been proposed to constitute
either transmembrane segments or a signal sequence followed
by a transmembrane region (36). Earlier we had shown that
pORF3 does not undergo any signal processing (20). Here we
show that pORF3 is capable of associating with the cytoskel-
eton and that deletion of hydrophobic domain I results in a loss
of such an association. The exact nature of this association has
not yet been characterized, but preliminary coimmunoprecipi-
tation experiments suggest that pORF3 does not enter into a
stable complex with the tubulins. Perhaps some other cytoskel-
etal protein(s) is responsible for the localization of pORF3 in
this subcellular fraction. Other results (19) suggest that pORF3
enters into a complex with pORF2, the major structural pro-
tein of HEV. This would mean that one possible function of
pORF3 within the HEV life cycle is to serve as a cytoskeletal
anchor (site) at which pORF2 can assemble the viral nucleo-
On the basis of results presented here, pORF3 phosphory-
lation may regulate viral assembly in response to as-yet-un-
known signals in liver cells. Work is in progress to address the
pORF2-pORF3 interaction and the relevance of pORF3 phos-
phorylation in this process. It is also interesting to note that
while pORF3 of all other HEV isolates carries potential
MAPK and cdc2K phosphorylation sites, pORF3 of the most
divergent Mexican isolate lacks such a site. Whether this dif-
ference results in differences in disease pathogenesis is an
interesting aspect to study in the future.
We thank Randall Kaufman for plasmid pMT3, Yash Vaishnav for
suggestions on this work, and Seyed Hasnain and Sudhir Sopory for
critical reading of the manuscript. The help rendered by R. Kumar in
cell culture maintenance and propagation and Dipti Arora-Chugh in
proofreading the manuscript and the secretarial assistance of Pratibha
Chaturvedi are gratefully acknowledged.
This work was supported by internal funds from the ICGEB.
TABLE 1. Evaluation of marker proteins
Fraction LDH (U/ml)a
1,200 ? 40
770 ? 30
70 ? 15
56.6 ? 2.6
aReported as mean ? standard deviation.
9052ZAFRULLAH ET AL.J. VIROL.
1. Alessi, D. R., A. Cuenda, P. Cohen, D. T. Dudley, and A. R. Saltiel. 1995.
PD098059 is a specific inhibitor of the activation of mitogen-activated pro-
tein kinase kinase in vitro and in vivo. J. Biol. Chem. 270:27489–27494.
2. Alvarez, E., I. C. Northwood, F. A. Gonzalez, D. A. Latour, A. Seth, C. Abate,
T. Curran, and R. J. Davis. 1991. Pro-Leu-Ser/Thr-Pro is a consensus pri-
mary sequence for substrate protein phosphorylation. J. Biol. Chem. 266:
3. Avruch, J., X.-F. Zhang, and J. M. Kryiakis. 1994. Raf meets Ras: complet-
ing the framework of a signal transduction pathway. Trends Biochem. Sci.
4. Aye, T. T., T. Uchida, X.-Z. Ma, F. Iida, T. Shikata, H. Zhuang, and K. M.
Win. 1992. Complete nucleotide sequence of a hepatitis E virus isolated from
the Xinjiang epidemic (1986–1988) of China. Nucleic Acids Res. 20:3512.
5. Bi, S. L., M. A. Purdy, K. A. McCaustland, H. S. Margolis, and D. W.
Bradley. 1993. The sequence of hepatitis E virus isolated directly from a
single source during an outbreak in China. Virus Res. 28:233–247.
6. Bradley, D. W. 1990. Enterically-transmitted non-A, non-B hepatitis. Br.
Med. Bull. 46:442–461.
7. Chen, R. H., C. Sarnecki, and J. Blenis. 1992. Nuclear localization and
regulation of erk- and rsk-encoded protein kinases. Mol. Cell. Biol. 12:915–
8. Cicirelli, M. F., S. L. Pelech, and E. G. Krebs. 1988. Activation of multiple
protein kinases during the burst in protein phosphorylation that precedes the
first meiotic cell division in Xenopus oocytes. J. Biol. Chem. 263:2009–2019.
9. Clark-Lewis, I., J. S. Sanghera, and S. L. Pelech. 1991. Definition of a
consensus sequence for peptide substrate recognition by p44mpk, the meiosis-
activated myelin basic protein kinase. J. Biol. Chem. 266:15180–15184.
10. Davis, R. J. 1993. The mitogen-activated protein kinase signal transduction
pathway. J. Biol. Chem. 268:14553–14556.
11. Duclos, B., S. Marcandier, and A. J. Cozzone. 1991. Chemical properties and
separation of phosphoamino acids by thin-layer chromatography and/or elec-
trophoresis. Methods Enzymol. 201:10–21.
12. Dudley, D. T., L. Pang, S. J. Decker, A. J. Bridges, and A. R. Saltiel. 1995. A
synthetic inhibitor of the mitogen-activated protein kinase cascade. Proc.
Natl. Acad. Sci. USA 92:7686–7689.
13. El-Diery, W. S., T. Tokino, V. E. Velculescu, D. B. Levy, R. Parsons, J. M.
Trent, D. Lin, W. E. Mercer, K. W. Kinzler, and B. Vogelstein. 1993. WAF1,
a potential mediator of p53 tumor suppression. Cell 75:817–825.
14. Fry, K. E., A. W. Tam, M. M. Smith, J. P. Kim, K. C. Luk, L. M. Young, M.
Piatak, R. A. Feldman, K. Y. Yun, M. A. Purdy, K. A. McCaustland, D. W.
Bradley, and G. R. Reyes. 1992. Hepatitis E virus (HEV): strain variation in
the nonstructural gene region encoding consensus motifs for an RNA-de-
pendent RNA polymerase and an ATP/GTP binding site. Virus Genes 6:
15. Heller, T., R. H. Purcell, K. R. Klimpel, and S. U. Emerson. 1996. Assess-
ment of the function of the third open reading frame protein of the hepatitis
E virus, abstr. A111, p. 79. In IX Triennial International Symposium on Viral
Hepatitis and Liver Disease.
16. Hikada, H., M. Watanabe, and R. Kobayashi. 1991. Properties and use of
H-series compounds as protein kinase inhibitors. Methods Enzymol. 201:
17. Hoshi, M., E. Nishida, and H. Sakai. 1988. Activation of a Ca2?-inhibitable
protein kinase that phosphorylates microtubule-associated protein 2 in vitro
by growth factors, phorbol esters, and serum in quiescent cultured human
fibroblasts. J. Biol. Chem. 263:5396–5401.
18. Huang, C. C., D. Nguyen, J. Fernandez, K. Y. Yun, K. E. Fry, D. W. Bradley,
A. W. Tam, and G. R. Reyes. 1992. Molecular cloning and sequencing of the
Mexico isolate of hepatitis E virus (HEV). Virology 191:550–558.
19. Jameel, S. Unpublished data.
20. Jameel, S., M. Zafrullah, M. H. Ozdener, and S. K. Panda. 1996. Expression
in animal cells and characterization of the hepatitis E virus structural pro-
teins. J. Virol. 70:207–216.
21. Kennelly, P. J., and E. G. Krebs. 1991. Consensus sequences as substrate
specificity determinants for protein kinases and protein phosphatases.
J. Biol. Chem. 266:15555–15558.
22. Khuroo, M. S. 1980. Study of an epidemic of non-A, non-B hepatitis: pos-
sibility of another human hepatitis virus distinct from post-transfusion
non-A, non-B type. Am. J. Med. 68:818–823.
23. Khuroo, M. S., M. R. Teli, S. Skidmore, M. A. Sofi, and M. Khuroo. 1981.
Incidence and severity of viral hepatitis in pregnancy. Am. J. Med. 70:252–
24. Koonin, E. V., and V. V. Dolja. 1993. Evolution and taxonomy of positive-
strand RNA viruses: implications of comparative analysis of amino acid
sequences. Crit. Rev. Biochem. Mol. Biol. 28:375–390.
25. 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 do-
mains in the nonstructural polyprotein of hepatitis E virus: delineation of an
additional group of positive-stranded RNA plant and animal viruses. Proc.
Natl. Acad. Sci. USA 89:8259–8263.
26. Meijer, L. 1995. Chemical inhibitors of cyclin-dependent kinases, p. 351–363.
In L. Meijer, S. Guidet, and H. Y. L. Tung (ed.), Progress in cell cycle
research, vol. 1. Plenum Press, New York, N.Y.
27. Miller, M. J. 1995. Viral taxonomy. Clin. Infect. Dis. 21:280.
28. Niederman, T. M., W. R. Hastings, and L. Ratner. 1993. Myristoylation-
enhanced binding of the HIV-1 Nef protein to T cell skeletal matrix. Virol-
29. Niesters, H. G. M., and J. H. Strauss. 1990. Mutagenesis of the conserved
51-nucleotide region of Sindbis virus. J. Virol. 64:1639–1647.
30. Panda, S. K., S. K. Nanda, M. Zafrullah, I. H. Ansari, M. H. Ozdener, and
S. Jameel. 1995. An Indian strain of hepatitis E virus (HEV): cloning,
sequence, and expression of the structural region and antibody responses in
sera from individuals from an area of high-level HEV endemicity. J. Clin.
31. Payne, D. M., A. J. Rossomando, P. Martino, A. K. Erickson, J.-H. Her, J.
Shabanowitz, D. F. Hunt, M. J. Weber, and T. W. Sturgill. 1991. Identifica-
tion of the regulatory phosphorylation sites in pp42/mitogen-activated pro-
tein kinase (MAP kinase). EMBO J. 10:885–892.
32. Pelech, S. L., and D. L. Charest. 1995. MAP kinase-dependent pathways in
cell cycle control, p. 33–52. In L. Meijer, S. Guidet, and H. Y. L. Tung (ed.),
Progress in cell cycle research, vol. 1. Plenum Press, New York, N.Y.
33. Purcell, R. H., and J. R. Ticehurst. 1988. Enterically transmitted non-A,
non-B hepatitis: epidemiology and clinical characteristics, p. 131–137. In
A. J. Zuckerman (ed.), Viral hepatitis and liver disease. Alan R. Liss, Inc.,
New York, N.Y.
34. Purdy, M. A., A. W. Tam, C. C. Huang, P. O. Yarbough, and G. R. Reyes.
1993. Hepatitis E virus: a non-enveloped member of the “alpha-like” RNA
virus supergroup? Semin. Virol. 4:319–326.
35. Ray, L. B., and T. W. Sturgill. 1987. Rapid stimulation by insulin of a
serine/threonine kinase in 3T3-L1 adipocytes that phosphorylates microtu-
bule-associated protein 2 in vitro. Proc. Natl. Acad. Sci. USA 84:1502–1506.
36. Reyes, G. R., C. C. Huang, A. W. Tam, and M. A. Purdy. 1993. Molecular
organization and replication of hepatitis E virus (HEV). Arch. Virol. 7:15–
37. Swick, A. G., M. Janicot, T. Cheval-Kastelic, J. C. McLenithan, and M. D.
Lane. 1992. Promoter-cDNA-directed heterologous protein expression in
Xenopus laevis oocytes. Proc. Natl. Acad. Sci. USA 89:1812–1816.
38. 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.
39. Tamaoki, T. 1991. Use and specificity of Staurosporine, UCN-01 and Cal-
phostin as protein kinase inhibitors. Methods Enzymol. 201:340–347.
40. Tsarev, S. A., S. U. Emerson, G. R. Reyes, T. S. Tsareva, L. J. Letgers, I. A.
Malik, M. Iqbal, and R. H. Purcell. 1992. Characterization of a prototype
strain of hepatitis E virus. Proc. Natl. Acad. Sci. USA 89:559–563.
41. Wong, D. C., R. H. Purcell, M. A. Sreenivasan, S. R. Prasad, and K. M.
Pavri. 1980. Epidemic and endemic hepatitis in India: evidence for a non-A,
non-B virus etiology. Lancet ii:876–879.
42. Xiong, Y., G. J. Hannon, H. Zhang, D. Casso, R. Kobayashi, and D. Beach.
1993. p21 is a universal inhibitor of cyclin kinases. Nature 366:701–704.
43. Yuckenberg, P. D., F. Witney, J. Geisselsoder, and J. McClary. 1991. Site-
directed in vitro mutagenesis using uracil-containing DNA and phagemid
vectors, p. 27–48. In M. J. McPherson (ed.), Directed mutagenesis: a prac-
tical approach. IRL Press, Oxford, England.
VOL. 71, 1997HEPATITIS E VIRUS ORF3 PROTEIN 9053