Antiviral Chemistry & Chemotherapy 13:353–362
A cell-based model of HCV-negative-strand RNA
replication utilizing a chimeric hepatitis C
virus/reporter RNA template
Robert W King*, Marianne Zecher, Matthew W Jeffries, Denise R Carroll, Joseph M Parisi and
Claudio Pasquinelli
The Experimental Station, Bristol-Myers Squibb, Wilmington, Del., USA
*Corresponding author: Tel: +1 609 252 4323; Fax: +1 609 252 6012; E-mail: [email protected]
The inability of hepatitis C virus (HCV) to replicate
in cell culture has hindered the discovery of antiviral agents against this virus. One of the biggest
challenges has been to find a model that allows
one to easily and accurately quantify the level of
HCV RNA replication that is occurring inside the
cell. In an attempt to solve this problem, we have
created a plasmid pMJ050 that encodes a chimeric
‘HCV-like’ RNA that can act as a reporter for HCV
RNA replication. This RNA consists of an antisense
copy of the firefly luciferase sequence flanked by
the 5′ and 3′ untranslated regions of the negative
strand of the HCV RNA. If, in cells that contain
functional HCV proteins, the chimeric RNA is recognized as a substrate for the viral RNA-dependent RNA polymerase, the chimeric RNA will be
transcribed into the complementary strand. This
RNA has a 5′ HCV internal ribosome entry site and
the luciferase sequence in the coding orientation,
allowing translation of the RNA into biologically
active luciferase. When pMJ050 was transfected
into a cell line that is stably transfected with a
cDNA copy of the HCV 1b genome, luciferase was
produced in a manner that was dependent upon
the presence of at least a functional HCV RNAdependent RNA polymerase. In addition, we constructed a cell line, 293B4α that constitutively produced luciferase in response to the presence of
functional HCV proteins. This system permits the
accurate determination of the level of HCV RNA
replication by the quantification of luciferase.
Keywords: HCV replication, cell-based model,
RNA-dependent RNA polymerase
Introduction
Hepatitis C virus (HCV), a member of the Flaviviridae
family, is an enveloped, positive-stranded RNA virus that
causes both acute and chronic infection of the liver
(Houghton, 1996). Failure to clear the virus frequently
results in the progression to chronic liver disease, cirrhosis
and hepatocellular carcinoma. Because of this, HCV infection has become the leading cause for liver transplantation
in developed nations.
The HCV genomic RNA is approximately 9.6 kb in
length and consists of a 5′ untranslated region (UTR), a
single long open reading frame (ORF) that encodes the
viral polyprotein, and a 3′ UTR. The sequence of the 5′
UTR is highly conserved and contains an internal ribosome entry site (IRES) that allows cap-independent translation of the viral ORF (Tsukiyama-Kohara et al., 1992)
and regulatory elements that control positive-strand replication (Reigadas et al., 2001). The polyprotein that is
encoded by the ORF is approximately 3000 amino acids in
length and produces at least ten viral proteins from co- and
post-translational processing by cellular and viral proteinases. The structural proteins are located in the amino-
©2002 International Medical Press 0956-3202/02/$17.00
terminal one-third of the polyprotein and the non-structural proteins are found in the carboxy-terminal two-thirds.
The sequence of the HCV 3′ UTR consists of a poorly
conserved stretch of approximately 40 nucleotides, a poly
(U)/pyrimidine tract of variable length, and a highly conserved 98 nucleotide sequence (Blight & Rice, 1997).
Sequences and/or secondary structures in the 3′ UTR have
been shown to interact with the HCV NS3 protease/helicase and NS5B RNA-dependent RNA polymerase
(RdRp), viral proteins that are essential for virus replication
(Oh et al., 1999; Banergee & Dasgupta 2001; Friebe et al.,
2001; Reigadas et al., 2001).
Recently, several cell culture systems have been
described that model different aspects of the HCV replication cycle. Two tetracycline-responsive cell lines that were
stably transfected with either the nonstructural genes or the
entire HCV genome (Moradpour et al., 1998) have been
shown by Western blot analysis to model the processing of
the HCV polyprotein by the HCV-encoded NS3 serine
protease. Myung et al. recently have reported that a transient transfection system that utilizes a plasmid that
1
RW King et al.
encodes the entire HCV genome under the transcriptional
control of the T7 polymerase promoter and a replicationdeficient adenovirus that encodes the T7 polymerase can
act as a model for negative-strand HCV RNA synthesis
and virus particle assembly (Myung et al., 2001).
A selectable subgenomic HCV replicon system has been
described that efficiently models the replication of both the
plus and minus strands of HCV RNA, processing of the
HCV polyprotein, and internal ribosomal entry site
(IRES)-dependent translation of the selectable marker
(Lohmann et al., 1999). However, transfection of the
human hepatoblastoma cell line, HUH-7 with these replicons resulted in colony formation with efficiency in the
range of 0.0001–0.0005%. Continued passage of the transfected cells in selection media resulted in the accumulation
of adaptive mutations in the HCV genome that led to an
increase in the efficiency of colony formation of approximately 1000- to 20000-fold (Blight et al., 2000, Krieger et
al., 2001, Lohmann et al., 2001).
In the above system, the replication of the replicon RNA
is catalyzed by the HCV RdRp that is encoded by the
replicon RNA itself. Here we describe a cell-based system
in which the HCV RdRp utilizes an ‘HCV-like’ chimeric
RNA as a template for RNA synthesis. This template consists of the firefly luciferase gene in the antisense orientation that is flanked by the 5′ and 3′ UTRs of the HCVnegative strand. This RNA is recognized by the HCV
RNA-dependent RNA polymerase (RdRp) as a template
and is transcribed into the orientation of the positive
strand. This positive-strand-like RNA is translated in an
IRES-dependent fashion producing luciferase that can be
easily detected and quantified.
Materials and methods
Cell lines and culture conditions
293 (human embryonic kidney) cells were obtained from
the American Type Culture Collection and maintained in
Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 5% fetal bovine sera (FBS), penicillin (50
units per ml), and streptomycin (50 µg per ml) at 37°C and
5% CO2. Other cell lines were maintained in the above
medium with the addition of the following selection
agents: 293FL9 cells, G418 (500 µg per ml); 293B4α cells,
zeocin (500 µg per ml) and G418 (500 µg per ml);
293SVLuc cells, zeocin (500 µg per ml). HUH-7 (FCCC)
cells were a kind gift from Dr William Mason
(Philadelphia, Pa., USA). HUH-7 (Apath) cells were
licensed from Apath Corp. Hep G2 and HeLa cells were
purchased from the ATCC. HUH-7 (FCCC), HUH-7
(Apath), Hep G2 and HeLa cells were maintained in
DMEM supplemented with 10% FBS, minimal essential
amino acids, penicillin and streptomycin.
2
Cells (1×106 cells in 500 µl DMEM per transfection)
were transfected with 2 µg of plasmid DNA by electroporation (BioRad Corp., Hercules, Calif., USA, Gene Pulser;
1 pulse, 0.25 kV, and 960 µF). Cells were plated either in
35 mm plates or 100 mm plates in DMEM+FBS. Fortyeight hours post-transfection, the cells that were placed in
35 mm plates were assayed for luciferase activity as
described below; cells plated in 100 mm plates were selected for the stable integration of the plasmids by placing
them in DMEM, FBS and the appropriate selection
agents.
Plasmids
The HCV 1b genome was encoded by plasmid pCX-HCV
1b. The transcription of plus-strand viral RNA is under the
control of the cytomegalovirus immediate-early enhancer
and chicken β-actin promoter. Furthermore, the 3′ terminus of the RNA transcribed from this plasmid is polyadenylated by the hepatitis B virus poly-adenylation signal.
The reporter plasmid, pMJ050 was created by placing,
in order from 5′ to 3′, the HCV 1b 3′ UTR sequence (in
the antisense orientation), the luciferase coding sequence
(in the antisense orientation), the HCV 5′UTR including
the first 39 bases of the core gene (in the antisense orientation), and the hepatitis δ ribozyme (in the sense orientation) into plasmid pZeoSV, under the control of the SV40
promoter. The luciferase gene from pGL3 (Promega,
Madison, Wisc., USA) was placed under control of the
SV40 promoter in pZeoSV (Invitrogen, Carlsbad, Calif.,
USA) to create a control plasmid, pSVLuc. The sequences
of pMJ050 and pSVLuc were confirmed by automated
DNA sequencing on an ABI Biosystems Prism DNA
sequencer utilizing Rhodamine-labelled dideoxynucleotides and AmpliTaq polymerase FS.
The coding sequences of the HCV nonstructural proteins NS2, NS3 and NS5B were amplified by PCR from
pCX-HCV1b and cloned individually into pMCS5
(MoBiTec, Marco Island. Fla., USA) to created pRK2CA,
pRK470, and pRK401, respectively. The sequences for the
HCV NS2 protease, NS3 serine protease, NS3 helicase and
NS5B RdRp were mutated using the QuikChange
Mutagenesis Kit and the mutagenic primers N2CAS,
NS2CAA, 3PRMS, 3PRMA, 3HLMS, 3HLMA,
5BGAA, and 5BGAAS, respectively (Table 1) as directed
by the manufacturer (Stratagene, La Jolla., Calif., USA).
The mutated sequences were then cloned back into the
HCV genome in plasmid, pCX-HCV1b to create plasmids, pRK426 (NS2–), pRK475 (NS3Pr–), pRK480
(NS3Hl–), and pRK402 (NS5B–). The mutated sequences
in pRK426 (NS2–), pRK475 (NS3Pr–), pRK480 (NS3Hl–),
and pRK402 (NS5B–) were reverted back to wild type
using the mutagenic primers N2CASR, NS2CAAR,
3PRMRS, 3PRMRA, 3HLMRS, 3HLMRA, 5BGDDR,
©2002 International Medical Press
Cell-based model of HCV negative-strand RNA replication
Table 1. Sequence of oligonucleotide primers used for PCR mutagenesis
Oligo name
N2CAA
N2CAS
N2CAAR
N2CASR
5BGAAS
5BGAA
5BGDDR
5BGDDA
3PRMS
3PRMA
3PRMRS
3PRMRA
3HLMS
3HLMA
3HLMRS
3HLMRA
Oligonucleotide sequence (5′ to 3′)
CCG CAC CAA CAT GGC TGC GCG AAT GAG C
GCT CAT TCG CGC AGC CAT GTT GGT GCG G
CCG CAC CAA CAT GAC TGC GCG AAT GAG C
GCT CAT TCG CGC ATG CAT GTT GGT GCG G
CGA TGC TTG TGT GCG GCG CCG CCC TTG TCG TTA TCT G
CAG ATA ACG ACA AGG GCG GCG CCG CAC ACA AGC ATC G
CGA TGC TTG TGT GCG GCG ACG ACC TTG TCG TTA TCT G
CAG ATQT ACG ACA AGG TCG TCG CCG CAC ACA AGC ATC G
TAC TTG AAA GGA TGC GCG GGG GGT CCG
CGG ACC CCC CGC GAA TCC TTT CAA GTA
TAC TTG AAA GGG TGC TCG GGG GGT CCG
CGG ACC CCC CGA GCA CCC TTT CAA GTA
CAT AAT ATG TGA TGA GTC CCA CTC AAC TGA CTC
GAG TCA GTT GAG TGG GAC TCA TCA CAT ATT ATG
CAT AAT ATG TGA TGA GTG CCA CTC AAC TGA CTC
GAG TCA GTT GAG TGG CAC TCA TCA CAT ATT ATG
and 5BGDDA, respectively. The wild type sequences were
then cloned back into pRK426 (NS2–), pRK475 (NS3Pr–),
pRK480 (NS3Hl–), and pRK402 (NS5B–) to create
pRK426 (NS2–) (wt), pRK475 (NS3Pr–) (wt), pRK480
(NS3Hl–) (wt), and pRK402 (NS5B–) (wt), respectively. In
addition, the sequences for HCV NS3 and NS5b were
cloned into pIRESneo2 (Clontech, Palo Alto, Calif., USA)
to create pRKNS3 and pRKRdRp, respectively. Sequences
for all plasmids were confirmed by automatic sequencing.
Western Blot analysis
Cells (1×106 cells per well) were plated in 6-well tissue culture plates. Twenty-four hours later, the cells were lysed in
50 µl of lysis buffer (50mM Tris, pH 8.0; 1mM EDTA; 1%
NP-40). The lysate was clarified by microcentrifugation.
Fifteen microlitres of the clarified lysate was mixed with 5
µl of 4X loading buffer (500mM Tris-HCl, pH 6.8; 4%
SDS; 20% glycerol; 40mM DTT; 0.02% bromophenol
blue) and boiled briefly before loading onto a 4–12% gradient polyacrylamide gel. Proteins were separated by size and
transferred to a nitrocellulose filter.
HCV NS3 and NS5B proteins detected by Western blot
analysis using monoclonal antibodies (mAbs) that were
raised against Escherichia coli-derived recombinant HCV
NS3 and NS5B proteins that were then shown to react
specifically with these two proteins, respectively (data not
shown). HCV core protein was detected using human
patient sera from an HCV-infected patient. To determine
if this patient sample contained Abs that reacted with the
HCV core protein, we created cell lysates from two cell
lines, the 293 human embryonic kidney line and a 293derived line that constitutively expressed the HCV structural genes. We tested both lysates by Western blot and
found that a band that migrated to the area where we
would expect the HCV core to migrate was only present in
Antiviral Chemistry & Chemotherapy 13:6
the cell lysate that expressed the HCV structural genes
(data not shown).
Northern blot analysis
Total cellular RNA was isolated from transfected 293,
293FL9, and 293B4α cells using the RNAgents RNA
Isolation Kit (Promega) as directed by the manufacturer.
RNA
was
separated
by
size
on
a
1%
agarose/MOPS/formaldehyde gel and transferred to a
nitrocellulose filter. HCV RNA was detected with a 32Plabelled probe, which was prepared by random priming
(Amersham Megaprime DNA Labeling System,
Piscataway, NJ, USA) using the sequence for the HCV
NS3 as template. Control HCV RNA was generated by in
vitro T7 polymerase catalyzed transcription of Xho I linearized plasmids using a commercial T7 transcription kit
(Epicentre, Madison, Wisc., USA). The transcripts were
approximately 9.6 kilobases in size. RNA levels were quantified using the Bio-Rad phosphorimager and Quantity
One software package (Hercules, Calif., USA).
Luciferase assay
Luciferase levels were determined in transfected cells
(3×104 cells per well) using the Bright-Glo Luciferase
Assay System as directed by the manufacturer (Promega
Corp., Madison, Wisc., USA). Luciferase units were quantified using a TopCount NXT Microplate Scintillation and
Luminescence Counter (Packard, Downers Grove, Ill.,
USA). All luciferase levels were normalized for transfection
efficiency using plasmid pSV-β-galactosidase (Promega
Corp.).
RT-PCR analysis of HCV RNA
Total cytoplasmic RNA from cells (1×106 cells/sample) was
isolated using the RNAgents Total RNA Isolation Kit as
3
RW King et al.
directed by the manufacturer (Promega Corp.). The RTPCR mixture (50 µl total volume) consisted of 1 µl of purified RNA sample and 0.25 mM each of two oligonucleotides, one of which was complementary to the sense
sequence of the luciferase gene (LUCFOR:CCGAGTGTAGRAAACATTCC) and one that was complementary
to the antisense sequence (LUCREV:CTCGCATGCCAGAGATCC), which resulted in a RT-PCR product of
approximately 1000 bp. In addition, the mixture contained
reaction buffer, dNTPs and Taq polymerase, which were
supplied as components of the Access RT-PCR System
(Promega Corp.). RT-PCR was performed as directed by
the manufacturer.
Two-step, strand-specific PCR was performed by
adding only one of the oligonucleotides to the reaction
mixture and stopping the RT-PCR after the completion of
the RT step. The RT was inactivated at 95°C (15 min), the
other oligonucleotide and Taq polymerase were added, and
the PCR step was allowed to go to completion.
Actinomycin D treatment of the 293B4α cell
line
293B4α cells (1×106 cells) were treated with 5 µg/ml
actinomycin D treatment. Four hours after actinomycin D
treatment began, the cell were treated with 3H-uridine
(25 µCi/ml). Three hours after the addition of the
radioactively-labelled uridine, the cells were harvested
and total RNA extracted. RNA was separated by size on
a 1% agarose/MOPS/formaldehyde gel. The gel was dried
and radioactively-labelled RNA was detected by
autoradiography.
Figure 1. Western and Northern blot analysis of the
HCV proteins and RNA produced in the 293FL9 cell
line
(b)
(a)
Control
FL9
293
Control
293
(c)
FL9
(d)
293
FL9
(a) The HCV NS3 protein was detected with mouse monoclonal Ab
to the HCV NS3. (b) The HCV NS5B protein was detected with
mouse monoclonal Ab to NS5B. (c) The HCV core protein was
detected with antibodies from a serum sample from an HCV-positive patient, serum #38. Control: purified recombinant protein produced in E. coli; 293: cell lysate from 293 cells; and FL9: cell lysate
from 293FL9 cells. (d) The HCV RNA was detected by Northern blot
analysis using a 32P-labelled probe that was complementary to the
HCV NS3 positive-strand sequence. Control: in vitro transcribed full
length positive-strand HCV RNA; 293: cytoplasmic RNA isolated
from 293 cells; FL9: cytoplasmic RNA isolated from 293 FL9 cells;
and B4a: cytoplasmic isolated from 293 B4a cells.
Figure 2. Schematic diagram showing the events
leading to luciferase production in the transfected
293 cell lines
Results
(a)
SV40
Expression of the HCV proteins in 293 cells
Plasmid pCX-HCV1b, which contains a cDNA copy of
the HCV 1b genome under the control of the chicken
β-actin promoter, was transfected by electroporation into
293 cells. Colonies capable of growing in 300 µM G418
were isolated and tested for the production of HCV
proteins. From several stably transfected clones, we
selected one – designated 293FL9 – which constitutively
produced the HCV positive-strand RNA and proteins
(HCV NS3, NS5B and core proteins, Figure 1a–c; HCV
RNA, Figure 1d).
To create a system in which the production of a
reporter protein was dependent upon the reporter’s RNA
being transcribed by the HCV replicase complex, we
placed an antisense copy of the firefly luciferase gene
between the 5′ and 3′ UTRs of the HCV-negative strand
(Figure 2). In addition, the sequence of the hepatitis δ
ribozyme was placed 3′ of the 3′ UTR-negative-strand
sequence. This construct then was placed under the
4
3’UTR(as)
Luciferase(as)
5’UTR/IRES(as)
Hepatitis δ ribozyme
DNA-dependent RNA polymerase
(provided by the cell)
(b)
(–) RNA
Ribozyme cis-cleavage
(c)
(–) RNA
HCV RNA-dependent RNA polymerase
(provided by the HCV genome)
(d)
(+) RNA
5’UTR/IRES
Luciferase
3’UTR
IRES-mediated
translation
Luciferase activity
(a) The genetic construct in pMJ050. The HCV 3′ and 5′ UTRs (from
the genomic orientation) and luciferase gene are in the antisense
orientation; whereas the hepatitis δ ribozyme is in the sense orientation. (b) The reporter RNA as transcribed by the cellular DNAdependent RNA polymerase. (c) The reporter mRNA after processing by the hepatitis δ ribozyme. (d) The luciferase coding strand of
RNA as transcribed by the HCV RDRP.
©2002 International Medical Press
Cell-based model of HCV negative-strand RNA replication
Figure 3. Production of luciferase in transientlytransfected 293 cells
(a)
Luc activity (light units)
100 000
10 000
1000
50
z
J0
-5
3F
29
29
3F
L9
L9
/p
G
/p
M
EM
J0
pM
3/
29
29
3/
pG
EM
-5
50
z
100
(b)
Luciferase production is dependent upon HCV
RNA replication enzymes
Luc activity (light units)
1 000 000
100 000
10 000
1000
pC
50
M
J0
J0
50
KN
S3
/p
pM
pR
X-
H
CV
1b
/p
M
J0
50
pC
XH
pR
CV
KR
1b
dR
p/
pM
J0
50
100
(a) 293 or 293FL9 were transfected with either pGEM-5z or pMJ050.
(b) 293 cells were co-transfected with pCX-HBV 1b and pMJ050,
pRKRdRp+pMJ050, and pRKNS3+pMJ050. Forty-eight hours posttransfection, cells were lysed and intracellular luciferase levels were
determined. Luciferase levels were corrected for transfection efficiency using pSV-β-galactosidase.
Antiviral Chemistry & Chemotherapy 13:6
control of the SV40 promoter in pZeoSV to create
plasmid, pMJ050.
Our hypothesis was that in 293FL9 cells transfected
with pMJ050, the SV40 promoter would drive transcription of an mRNA consisting of the 5′UTR (HCV-negative
strand) – firefly luciferase in the antisense orientation – 3′
UTR (HCV-negative strand) – hepatitis δ ribozyme. The
hepatitis δ ribozyme then would cleave itself from the
RNA leaving a native HCV-negative strand 3′ terminus.
The HCV replication complex would recognize this RNA
as a template for the HCV RdRp and transcribe it into the
complementary strand. The resultant RNA would consist
of the positive-strand 5′ UTR (including the HCV IRES),
the luciferase gene in the sense orientation and the positive
strand 3′ UTR. Translation of the luciferase gene would be
driven by the HCV IRES resulting in the production of
detectable and quantifiable luciferase.
To test this hypothesis, pMJ050 was transfected into
293 and 293FL9 cells. Forty-eight hours post-transfection
the cells were lysed and luciferase was quantified (Figure
3a). 293FL9 cells that were transfected with pMJ050 produced significantly more luciferase than 293 cells transfected with pMJ050, or 293 and 293FL9 cells transfected with
pGEM-5z (a control plasmid that does not contain the
luciferase gene).
Once we established that luciferase was produced in
293FL9 cells transfected with pMJ050, we wanted to
determine (1) if the HCV RdRp alone was sufficient to
produce luciferase, and (2) if not, what HCV proteins were
required for luciferase production. 293 cells that were cotransfected with pCX-HCV1b and pMJ050 produced
luciferase; however it was approximately 20% less than the
level that was produced in 293FL9 cells that had been
transfected with pMJ050 (compare Figure 3b and 3a).
When 293 cells were co-transfected with pMJ050 and
pRKRdRp, a plasmid that encoded HCV NS5b alone,
luciferase not was produced.
To determine if additional HCV proteins were necessary
for recognition of the chimeric template, we created plasmids that encoded the full length HCV 1b genome with
mutations in the NS2 protease, NS3 protease, NS3 helicase
and NS5B RdRp sequences. Previously, these mutations
had been shown to inactivate enzyme function (Grakoui et
al., 1993, Lohmann et al., 2001). Plasmids pCX-HCV1b,
pRK426 (NS2–), pRK475 (NS3Pr–), pRK480 (NS3Hel–),
and pRK402 (NS5b–) were co-transfected with pMJ050
into 293 cells. Forty-eight hours post-transfection, the cells
were lysed and tested for luciferase.
293 cells that were co-transfected with pMJ050 and
pRK475 (NS3Pr–), pRK480 (NS3Hel–), or pRK402
5
RW King et al.
Figure 4. Expression of luciferase in cells that are
transiently-transfected with pMJ050 and plasmids
encoding HCV genome mutants and their revertants
Figure 5. Luciferase production in 293, 293FL9, and
293B4α cell lines
1 000 000
100
100 000
Luc activity (light units)
Normalized luc activity (light units)
125
75
50
25
10 000
1000
e
as
S3 e
-p ( w
ro t )
-p
te
ro
as
te
e
a
N se
S
(
3w
N
S3
he t)
-h
lic
el
as
ic
e
as
N e
(
S
5b wt
N
S5
)
b- - Rd
Rp
Rd
Rp
(w
t)
as
293 cells were co-transfected with pMJ050 plus pCX-HCV 1b,
pRK426 (NS2–), pRK426 (NS2–) (wt), pRK475 (NS3Pr–), pRK475
(NS3Pr–) (wt), pRK480 (NS3Hel–), pRK480 (NS3Hel–) (wt), pRK402
(NS5b–), and pRK402 (NS5b–) (wt). Forty-eight hours after transfection, cells were lysed and luciferase levels were determined.
Luciferase levels were corrected for transfection efficiency using
pSV-b-galactosidase.
(NS5b–) did not produce luciferase (Figure 4). Plasmid
pRK426 (NS2–), when transfected with pMJ050 into 293
cells produced approximately 30% of the level of luciferase
as cells transfected with pCX-HCV1b plus pMJ050. In
addition, 293 cells transfected with pMJ050, pCXHCV1b, pRK426 (NS2–), pRK475 (NS3Pr–), pRK480
(NS3Hel–), or pRK402 (NS5b–) individually did not produce detectable levels of luciferase (data not shown).
To show that the mutations in these four plasmids were
responsible for the reduction in luciferase levels, the mutations were repaired by site-directed mutagenesis, creating
plasmids pRK426 (NS2–) (wt), pRK475 (NS3Pr–) (wt),
pRK480 (NS3Hel–) (wt), and pRK402 (NS5b–) (wt).
These plasmids co-transfected into 293 cells with pMJ050
showed levels of luciferase that were equal to that contained
in cells co-transfected with pCX-HCV1b and pMJ050.
Creation of a stably-transfected cell line that
produces luciferase in an HCV RdRp-dependent
fashion
293FL9 cells were transfected with pMJ050 and grown in
medium containing the antibiotic zeocin. Thirty days later,
6
29
3B
4
α
100
29
3/
pG
EM
-5
z
29
3/
pM
29
J0
3F
50
L9
/p
G
EM
29
-5
3F
z
L9
/p
M
J0
50
N
S3
S2
N
N
-p
ro
te
-p
ro
te
m
as
pl
S2
N
CV
H
o
N
W
ild
-t
yp
e
H
C
V
id
0
293 and 293FL9 cells were co-transfected with pGEM-5z or
pMJ050, and assayed for luciferase activity 48 h post-transfection.
A similar number of 293B4α cells were lysed and assayed for
luciferase activity.
48 zeocin-resistant clones were isolated, expanded, and
tested for luciferase production (data not shown). Of these,
11 clones produced luciferase. One clone, which was designated as 293B4α, consistently and reproducibly produced
high levels of luciferase (Figure 5). The 293B4α cell line
also constitutively produced the HCV core, NS3 protease
and NS5b RdRp, as well as the firefly luciferase proteins
(data not shown).
Production of luciferase RNA in the 293B4α cell
line
To establish if luciferase RNA was produced in both sense
and antisense orientations (equivalent to the positive and
negative strands of HCV RNA, respectively) in the
293B4α cell line, we isolated total cytoplasmic RNA from
the 293B4α, 293FL9 and 293SVLuc cell lines. By using
non-quantitative, strand-specific RT-PCR we determined
that the 293B4α cell line contained luciferase RNA in both
the sense and antisense orientations (Figure 6a, lanes 1 and
2; and Figure 6b, lanes 2 and 3). 293FL9 cells, whose
genome does not contain the luciferase gene, did not contain luciferase RNA (Figure 6b, lane 7) and the 293SVLuc
©2002 International Medical Press
Cell-based model of HCV negative-strand RNA replication
–
–
+
S
+
–
+
A
+
–
+
S
–
+
+
A
–
+
+
S
–
–
–
A
–
–
–
S
A
S
Strand polarity
A
S
S
A
293VLuc
–
–
+
A
293FL9
RNase treatment
DNase treatment
Reverse transcriptase
Strand polarity*
luciferase plasmid RNA
B4α cells
293B4α
luciferase plasmid RNA
Figure 6. RT-PCR analysis of the RNA produced in the 293B4α cells
S/A
S/A
S
A
1000 bp PCR
product from
luciferase gene
Total cytoplasmic RNA was isolated from 293B4α, 293FL9 and 293SVLuc cells. RNA was analyzed in a two-step RT-PCR, which would allow the
differentiation of sense and antisense RNA, and a single-step RT-PCR. Lanes labeled ‘RNase (+)’ and ‘DNase (+)’ were ones in which the RNA
sample was treated with either RNase or DNase prior to the RT-PCR. ‘RT(–)’ reactions were ones in which the RT step was omitted. S: uncoupled, two-step reactions that would detect the sense sequence of the luciferase gene; A: uncoupled, one-step reactions that would detect the
antisense sequence of the luciferase gene, equivalent to the HCV minus-strand; S/A: single-step RT-PCR that does not differentiate between
sense and antisense strands.
cells, which are stably transfected with the luciferase gene
under the control of the SV40 promoter, only contained
luciferase RNA in the sense orientation (Figure 6b, lanes 8
and 9).
To establish that this RT-PCR product was actually
derived from RNA and not contaminating host or plasmid
DNA , the RNA samples were first treated overnight with
either DNase or RNase prior to the RT-PCR step. DNase
treatment of the RNA samples had no effect on the quantity or quality of the product of either the antisense or sense
strand-specific RT-PCR (Figure 6a, lanes 5 and 6); however, RNase treatment completely eliminated the ability of
both of the strand-specific RT-PCRs to produce a product
(Figure 6a, lanes 3 and 4).
Since DNase treatment is rarely complete, sometimes
leaving behind enough DNA to produce a product, we performed the PCR amplification step without including the
RT portion. We found that upon elimination of the RT
step, no product was seen (Figure 6a, lanes 7 and 8).
Since the HCV RdRp is resistant to the anti-transcriptional effects of actinomycin D, we treated 293FL9 and
293B4α cells with actinomycin D and determined its effect
on luciferase production. Eight hour treatment of 293B4α
Antiviral Chemistry & Chemotherapy 13:6
cells, with 5 µg/ml actinomycin D resulted in a 50% reduction in luciferase expression (data not shown). However,
since we could not distinguish between the inhibition of
the production of the sense-strand luciferase RNA or the
transcription of HCV and antisense-strand luciferase
RNAs, we treated 293B4α and 293FL9 cells with actinomycin D for 4 h and then added 3H-uridine for another 3
h. Total cellular RNA was separated by size and radioactively-labelled RNA was detected by autoradiograghy (data
not shown). We found that a radioactively-labelled species
of RNA of the correct size (approximately 2300 bases) was
detected in the 293B4α sample that was absent in the
293FL9 sample.
Production of luciferase in other human cell
lines
It has been reported that the ‘HCV-like’ RNA in the
HCV replicon system replicates only in a certain subclone
of HUH-7 cells (Lohmann et al., 1999; Bright et al.,
2000). To determine if our system was also cell specific, we
co-transfected three cell lines of human liver origin,
HUH-7 (FCCC), HUH-7 (Apath) and Hep G2, and a
human cervical carcinoma cell line, HeLa, with pCX-
7
RW King et al.
HCV1b and pMJ050 and determined if luciferase was
produced (Figure 7).
We found that luciferase was produced only in the 293
cell line and in the HUH-7 cell line that was used to create
the replicon system, HUH-7 (Apath). Luciferase was not
detected in the other HUH-7 subclone, HUH-7 (FCCC)
or in HepG2 and HeLa cells.
Figure 7. Production of luciferase in transientlytransfected 293, HUH-7 (FCCC), HUH-7 (Apath), Hep
G2 and HeLa cells
100 000
8
10 000
1000
1b
CV
/H
H
C)
-7
U
H
H
U
H
U
H
-7
(F
CC
29
29
3/
3/
H
CV
CV
1b
/p
G
EM
/p 5z
H (FC
7( C 1 b MJ
H Apa C)/H / p G 050
U
E
t
C
H
7( h)/H V1 M A
b
5
C
pa
V1 /pM z
t
H h)/ b/p J05
ep H
G 0
em
G CV
2
1
H /HC b/p -5z
ep
V1 M
G
J0
b
2
H /H /pG 50
eL CV
em
a/
-5
H 1b/
C
H
eL V1 pM z
a/
b/ J0
H
CV pGE 50
1b M
-5
/p
z
G
EM
-5
z
100
H
The lack of a suitable cell culture system that allows the
replication of HCV has hindered the understanding of the
molecular biology of the HCV replication cycle and the
development of new therapies for HCV infection.
Although several new cell-based systems recently have
been described as models for HCV replication, they only
model a portion of the virus’ replication cycle (GoobarLarsson, 2001, Lohmann et al., 1999, Moradpour et al.,
1998, Myung et al., 2001, Pietschmann et al., 2001). In
addition, detecting and quantifying the level of virus replication is difficult in these systems relying on either Western
or Northern blot analysis, or quantitative RT-PCR. In this
study, we described a cell culture system that models the
production and processing of the HCV polyprotein, as well
as replication and IRES-dependent translation of a
chimeric ‘HCV-like’ RNA.
The advantage of this system is that these three
processes can be quickly and easily monitored and quantified. This is accomplished by quantifying the level of
luciferase, whose production is dependent upon the replication of the negative-strand ‘HCV-like’ RNA into the
positive-strand (or coding) orientation so that it can be
translated in an IRES-dependent fashion. In the 293B4α
cell line, an integrated cDNA copy of the HCV 1b
genome provides the necessary viral components for the
replication of this RNA.
This reporter construct consisted of the 5′ UTR of the
HCV-negative strand, the firefly luciferase gene in the
antisense orientation, the 3′ UTR of the HCV negative
strand, and the hepatitis δ ribozyme. The strategy was that
the DNA-dependent RNA polymerase that is provided by
the cell would transcribe this sequence into RNA. After
transcription, the hepatitis δ ribozyme would be removed
through a cis cleavage event, resulting in an ‘HCV-like’
RNA whose 3′ terminal sequence is exactly the same as that
found in the wild-type virus.
The critical question was if the HCV RdRp in the
293FL9 cell line would recognize the reporter RNA as a
template for RNA replication. Others have shown previously that in an in vitro enzyme system, the HCV RdRp
can utilize a variety of templates including ones that are
and are not derived from the HCV genome (Kolykhalov,
1996; Tanaka et al., 1996; Luo et al., 2000). Moreover, the
Luc activity (light units)
Discussion
Cells were transfected with pCX-HCV1b and pGEM-5z or pMJ050.
Forty-eight hours post-transfection, cells were lysed and intracellular luciferase levels were determined. Luciferase levels were corrected for transfection efficiency using pSV-β-galactosidase.
HCV RdRp is able to utilize poly(C) and poly(A) templates if the appropriate primers are supplied
(DeFranscesco et al., 1996).
Recently, it was shown that in an intracellular environment the HCV RdRp could recognize a series of templates
(Goobar-Larsson et al., 2001). The templates were not
restricted to ones that had HCV-specific sequences but
activity was higher with those that contained the positiveand negative-strand 3′ UTR sequences. Interestingly, the
HCV RdRp was active even when still part of the HCV
polyprotein; however, activity was diminished. Moreover,
cleavage of the RdRp from the polyprotein had no effect on
template usage or specificity. Finally, RdRp activity was
greatest when the RdRp was in the presence of the other
HCV nonstructural proteins.
To determine if the HCV RdRp produced in the
293FL9 cell line could use an ‘HCV-like’ template, we
transfected pMJ050 into the cells and assayed for
©2002 International Medical Press
Cell-based model of HCV negative-strand RNA replication
luciferase. We found that luciferase was produced in transfected 293FL9 cells but not in transfected 293 cells. Cotransfection of 293 cells with a plasmid that encodes the
entire HCV-1 1b genome and pMJ050 resulted in
luciferase production; whereas co-transfection of 293 cells
with a plasmid that encodes only the HCV NS5b and
pMJ050 did not. Moreover, co-transfection of 293FL9
cells with pMJ050 and plasmids that encoded cDNA
copies of the HCV genome that contained mutations
which inactivated the HCV NS2 protease, NS3 protease,
NS3 helicase and NS5B RDRP, respectively, resulted in
significantly reduced or background levels of lucifease.
These data showed that in this system the HCV RdRp is
essential but not sufficient for utilizing the chimeric ‘HCVlike’ RNA as a template for negative- to positive-strand
replication. This is in disagreement with the observations
of Goobar-Larsson et al. (2001).
It had been shown previously that the mutations that we
had engineered into the NS2 sequence had completely
inhibited the activity of the protease (Grakoui 1993);
therefore, it was unexpected that the NS2– HCV 1b
genome could still drive production of luciferase. We speculate that in the cellular environment, NS3 in the
uncleaved NS2/NS3 state may still be functional but at a
reduced level.
The HCV replicon systems that have been described
previously have identified several mutations that have been
selected for by continuous passage of the cells (Blight et al.,
2000; Krieger et al., 2001; Lohmann et al., 2001). These
mutations increased HCV RNA replication and enhanced
the efficiency of colony formation. A survey of the HCV 1b
sequence that we used to create the 293FL9 cell line
showed that this genome lacked any of the mutations that
conferred cell culture adaptation in the replicon systems
(data not shown). We hypothesize that these adaptive
mutations may not be necessary in our system because we
are not looking at (1) replication of the HCV genome itself
but instead at an ‘HCV-like’ RNA template that is approximately 1/5 the size of the HCV genome, and (2) we only
are looking at the replication of a continuously-produced,
negative-strand-like RNA to a positive-strand-like RNA.
Our system does not require the continuous synthesis of
both negative- and positive-strand RNA from an input
strand of RNA as in the replicon system.
Using the zeocin-resistance gene that was encoded by
pMJ050, we isolated several 293FL9 clones that contained
integrated copies of the antisense luciferase reporter construct. These clones were assayed for luciferase activity and
the 293B4α cell line was selected based on its high levels of
luciferase production. Western blot analysis showed that
this cell line synthesized the HCV core, NS3 protease and
NS5B RDRP and the firefly luciferase proteins. In addition, RT-PCR demonstrated that both the noncoding and
Antiviral Chemistry & Chemotherapy 13:6
coding RNA transcripts of the luciferase gene (representative of the negative and positive species of HCV RNA)
were produced.
It has been difficult to validate that luciferase expression
in the 293B4α cell line is dependent upon the HCV proteins and is not the result of the reporter cassette inserting
into the cellular genome in such a way that transcription of
the luciferase (and antisense luciferase) RNA is being driven by cellular promoters. Since this is a stably-transfected
system, it is not possible to do the type of genetic experiments that we did with the transient-transfection system.
One of the classic methods of showing that RNA synthesis is occurring through RdRp activity is to show that transcription occurs in the presence of actinomycin D since
viral RdRps are resistant to the anti-transcriptional activity
of actinomycin D. However, since transcription of the
HCV and the antisense luciferase RNAs are driven by
eukaryotic promoters, actinomycin D would inhibit transcription of these RNAs, leading to an overall reduction in
luciferase expression. To try to circumvent this problem,
293FL9 and 293B4α cells were treated with actinomycin D
and then labelled with 3H-uridine. A radioactively-labelled
RNA species of the expected size was detected in 293B4α
but not 293FL9 cells.
The greatest utility of this system lies in the fact that the
HCV RdRp can recognize and reproduce an ‘HCV-like’
template in the environment of the cell. If the template is a
reporter gene, as in this example, then HCV replication can
be detected and quantified easily, accurately and reproducibly. This ability to easily quantify HCV replication will
be essential if we are to gain a better understanding of the
molecular biology of the HCV replication cycle. Moreover,
this system would be quite amenable for use in the identification of chemical compounds that contain antiviral
activity against HCV. In fact, we have formatted the
293B4α cell line into a medium-throughput cell-based
assay that has allowed us to identify the anti-HCV activity
of nucleoside analogues, interferon-α and -β (King et al.,
2002), as well as low molecular weight, non-nucleoside
inhibitors of the HCV RdRp, NS2 protease and NS3 protease (data not shown).
One of the drawbacks to this system is that it can not be
used to study the production of HCV-negative-strand
RNA from positive-strand since the ‘negative-strand’ of the
‘HCV-like’ reporter RNA is constitutively transcribed from
DNA under the control of the SV40 promoter. This promoter is such a strong promoter that the large amount of
‘negative-strand-like’ RNA produced from this transcriptional process would overshadow the small quantity of ‘negative-strand-like’ RNA produced from the use of the positive-strand-like RNA by the HCV RdRp. Moreover, it is
not possible to study the production of negative-strand
RNA from the positive-strand template using the produc-
9
RW King et al.
tion of HCV negative-strand RNA from HCV positivestrand in this system because the HCV RNA encoded by
pCX-HCV1b is poly-adenylated and does not serve as a
template for the HCV RdRp in our system.
Finally, as has been reported for the replicon system,
there also is cell specificity in our system. The 293 and
HUH-7 (Apath) cell lines, which supported the production of luciferase in an HCV protein-dependent manner,
must contain certain factors that are necessary for the replication of the chimeric ‘HCV-like’ RNA, which are absent
in the HUH-7 (FCCC), Hep G2, and HeLa cell lines.
These factors have yet to be identified.
Acknowledgements
King RW, Zecher M & Jeffries MW (2002) Inhibition of the replication of a hepatitis C virus-like RNA template by interferon and
3′-deoxycytidine. Antiviral Chemistry & Chemotherapy 13:363–370.
Kolykhalov AA, Feinstone SM & Rice CM (1996) Identification of
a highly conserved sequence element at the 3′ terminus of hepatitis
C virus genome RNA. Journal of Virology 70:3363–3371.
Krieger N, Lohmann V& Bartenschlager R (2001) Enhancement of
hepatitis C virus RNA replication by cell culture-adaptive mutations. Journal of Virology 75:4614–4624.
Luo G, Hamatake RK, Mathis DM, Racela J, Rigat KL, Lemm J &
Colonno RJ (2000) De novo initiation of RNA synthesis by RNAdependent RNA polymerase (NS5B) of hepatitis C virus. Journal
of Virology 74:851–863.
Lohmann V, Korner F, Dobierzewska A & Bartenschlager R (2001)
Mutations in hepatitis C virus RNAs conferring cell culture adaptation. Journal of Virology 75:1437–1449.
Lohmann V, Korner F, Koch JO, Herian U, Theilmann L &
Bartenschlager R (1999) Replication of subgenomic hepatitis C
virus RNAs in a hepatoma cell line. Science 285:110–113.
We thank Leah Breath for providing the HCV NS3 and
NS5B antibodies, and Bruce Korant, Heather Foley and
Michelle Kimberland for critically reading the manuscript.
DRC’s participation was as part of the DuPont
Pharmaceuticals Company Summer Internship Program.
Lohmann V, Korner F, Herian U & Bartenschlager R (1997)
Biochemical properties of hepatitis C virus NS5B RNA-dependent RNA polymerase and identification of amino acid sequence
motifs essential for enzymatic activity. Journal of Virology
71:8416–8428.
References
Moradpour D, Kary P, Rice CM & Blum HE (1998) Continuous
human cell lines inducibly expressing hepatitis C virus structural
and nonstructural proteins. Hepatology 28:192–201.
Banergee R & Dasgupta A (2001) Specific interaction of hepatitis C
virus protease/helicase NS3 with the 3′-terminal sequence of the
viral positive- and negative-strand RNA. Journal of Virology
75:1708–1721.
Blight KJ, Kolykhalov AA & Rice CM (2000) Efficient initiation of
HCV RNA replication in cell culture. Science 290:1972–1974.
Blight KJ & Rice CM (1997) Secondary structure determination of
the conserved 98-base sequence at the 3′ terminus of hepatitis C
virus genome RNA. Journal of Virology 71:7345–7352.
DeFrancesco R, Behrens SE, Tomei L, Altamura S & Jiricny J
(1996). RNA-dependent RNA polymerase of hepatitis C virus. In
Methods in Enzymology:Viral Polymerases and Related Proteins.
Edited by LC Kuo, DB Olsen & SS Olsen. San Diego: Academic
Press.
Friebe P, Lohmann V, Krieger N & Bartenschlager R (2001)
Sequences in the 5′ nontranslated region of hepatitis C virus
required for RNA replication. Journal of Virology 75:12047–12057.
Goobar-Larsson L, Wiklund L & Schwartz S (2001) Intracellular
hepatitis C virus RNA-dependent RNA polymerase activity.
Archives of Virology 146:1553–1570.
Grakoui A, McCourt DW, Wychowski X, Feinstone SM & Rice
CM (1993) A second hepatitis C virus-encoded proteinase.
Proceedings of the National Academy of Science USA 90:10583–10587.
Houghton M (1996) Hepatitis C Viruses. In Virology, 3rd edn.; pp.
1035–1058. Edited by BN Field, DM Knipe & PM Howley.
Philadelphia: Lippincott-Raven.
Myung J, Khalap N, Kalkeri G, Garry R & Dash S (2001) Inducible
model to study negative strand RNA synthesis and assembly of
hepatitis C virus from a full-length cDNA clone. Journal of
Virological Methods 94:55–67.
Oh J-W, Ito T & Lai MM (1999) A recombinant hepatitis C virus
RNA-dependent RNA polymerase capable of copying the fulllength viral RNA. Journal of Virology 73:7694–7702.
Pietschmann T, Lohmann V, Rutter G, Kurpanek K &
Bartenschlager R (2001) Characterization of cell lines carrying
self-replicating hepatitis C virus RNAs. Journal of Virology
75:1252–1264.
Reigadas S, Ventura M, Sarih-Cottin L, Castroviego M, Livak S &
Astier-Gin T (2001) HCV RNA-dependent RNA polymerase
replicates in vitro the 3′ terminal region of the minus-strand viral
RNA more efficiently than the 3′ terminal region of the plus
strand. European Journal of Biochemistry 268:5857–5867.
Tanaka T, Kato N, Cho M-JE, Sugiyama K & Shimotohno K (1996)
Structure of the 3′ terminus of the hepatitis C virus genome.
Journal of Virology 70:3307–3312.
Tsukiyama-Kohara K, Iizuka N, Kohara M & Nomoto A (1992)
Internal ribosome entry site within hepatitis C virus RNA. Journal
of Virology 66:1476–1483.
Yanagi M, Claire MS, Shapiro M, Emerson SU, Purcell RH &
Bukh J (1998) Transcripts of a chimeric cDNA of hepatitis C virus
genotype 1b are infectious in vivo. Virology 224:161–172.
Received 8 October 2002; accepted 31 December 2002
10
©2002 International Medical Press