Template requirements and binding of hepatitis C virus
NS5B polymerase during in vitro RNA synthesis from the
3¢-end of virus minus-strand RNA
The
´
re
`
se Astier-Gin, Pantxika Bellecave, Simon Litvak and Michel Ventura
UMR-5097 CNRS, Universite
´
Victor Segalen Bordeaux 2, Bordeaux, France
Hepatitis C virus (HCV) is the major causative agent
of non-A, non-B hepatitis [1]. This virus has a posit-
ive-stranded RNA genome and belongs to the Flavivir-
idae family. The RNA contains a large open reading
frame that encodes a polyprotein which is cleaved into
10 viral proteins: C, E1, E2, p7, NS2, NS3, NS4A,
NS4B, NS5A and NS5B [2]. Recently, a frame shift
product of HCV core encoding sequence, the F pro-
tein, was described [3,4]. This protein has no known
functions. The large open reading frame is flanked by
two untranslated regions (UTR). The 341-nucleotide
(nt) 5¢UTR in association with the first nucleotides of
the core protein contains an internal ribosome entry
site (IRES) that directs cap-independent translation of
the viral RNA [5,6]. The 3¢UTR is composed of a
short variable region, a polypyrimidine tract (poly
U-UC) of variable length and a highly conserved
98-nucleotide segment (3¢X). The two latter domains
are essential for viral infectivity in vivo [7] and RNA
replication of HCV in the HCV replicon system [8,9].

(Received 23 March 2005, revised 24 May
2005, accepted 3 June 2005)
doi:10.1111/j.1742-4658.2005.04804.x
In our attempt to obtain further information on the replication mechanism
of the hepatitis C virus (HCV), we have studied the role of sequences at
the 3¢-end of HCV minus-strand RNA in the initiation of synthesis of the
viral genome by viral RNA-dependent RNA polymerase (RdRp). In this
report, we investigated the template and binding properties of mutated and
deleted RNA fragments of the 3¢-end of the minus-strand HCV RNA in
the presence of viral polymerase. These mutants were designed following
the newly established secondary structure of this viral RNA fragment. We
showed that deletion of the 3¢-SL-A1 stem loop significantly reduced the
level of RNA synthesis whereas modifications performed in the SL-B1 stem
loop increased RNA synthesis. Study of the region encompassing the 341
nucleotides of the 3¢-end of the minus-strand RNA shows that these two
hairpins play a very limited role in binding to the viral polymerase. On the
contrary, deletions of sequences in the 5¢-end of this fragment greatly
impaired both RNA synthesis and RNA binding. Our results strongly sug-
gest that several domains of the 341 nucleotide region of the minus-strand
3¢-end interact with HCV RdRp during in vitro RNA synthesis, in parti-
cular the region located between nucleotides 219 and 239.
Abbreviations
HCV, hepatitis C virus; IRES, internal ribosome entry site; nt, nucleotide; RdRp, RNA-dependent RNA polymerase; TCA, trichloroacetic acid;
UTR, untranslated region.
3872 FEBS Journal 272 (2005) 3872–3886 ª 2005 FEBS
plus-strand RNA has been determined [10,11] and the
involvement of the three stem loops of the 3¢X has
been extensively studied both in vitro, in the replicon
system and in vivo [7–9,12]. The secondary structure of
the 3¢-end of the minus-strand RNA has been estab-

by recombinant HCV NS5B using deletion mutants of
the 3¢ terminus of the minus-strand RNA but deletions
were made on the basis of the structure of the 5¢UTR of
the plus-RNA, the structure of the 3¢-end of the minus-
strand RNA being at that time unavailable.
In the present study we investigated the involvement
of sequences and ⁄ or structures in RNA synthesis direc-
ted in vitro by HCV NS5B by using new mutants of
the 3¢-end of minus-strand RNA.
Results
Effect of mutations in the 3¢- or the 5¢-end of the
(–)IRES HCV RNA template on RNA synthesis
The secondary structure of the 3¢-terminal nucleotide
region of the HCV minus-strand RNA is illustrated in
Fig. 1. This fragment contains two domains: domains
I and II. Structures of both domain I (A), as deter-
mined by Schuster et al. [13] and by Smith et al. [14],
and domain II, as determined, respectively, by Smith
et al. [14] (B), Schuster et al. [13] (C) or predicted by
RNA Draw software (D) are represented. Nucleotides
are numbered increasingly from the 3¢-end of the
RNA; the five first stem-loops (A) are named as repor-
ted by Schuster et al. [13]. The 228 nt of domain I fold
into five stable stem-loops (A). It has been found to
display the same secondary structure in the three mod-
els presented here: the fragment containing 365 nucleo-
tides described in [14], the 416 nt fragment described
by [13] and the 341 nt fragment, used in this work. On
the contrary, the 5¢-end of the different RNA frag-
ments (137 nt in Fig. 1B, 188 nt in Fig. 1C, and 113 nt

RdRp assay and the levels of RNA synthesis were
compared with that of the wild-type (–)IRES. Very
different results were obtained when changing either
stem-loop. As shown in Table 1, the deletion of the
SL-A1 stem loop in the (–)IRES DSL-A1 mutant
reduced the RNA synthesis by 39%. These results were
T. Astier-Gin et al. Binding and replication of 3¢-end of HCV minus RNA
FEBS Journal 272 (2005) 3872–3886 ª 2005 FEBS 3873
in accordance with those of Kashiwagi et al. [20],
which showed that deletion of SL-A1 reduced the
RNA synthesis by 25%.
We then performed different deletions or site-direc-
ted mutagenesis in the sequence of the hairpin SL-B1
(Fig. 2). We carried out the following mutations of
SL-B1: (a) [(–)IRESD91-97)] that contains a deletion of
nt 91–97 corresponding to a bulge where the ODN7
antisense hybridized; (b) [(–)IRES Dhp2] that contains
a deletion of the 39 nt corresponding to the apical part
of SL-B1; (c) [(–)IRES hp2b] with a change of four nt
that induces a dissociation of the stem at the base of
SL-B1; (d) [(–)IRES LDH2] with a complete deletion
of SL-B1 In contrast to the results obtained when
changes were introduced in SL-A1, none of the modifi-
cations in SL-B1 reduced RNA synthesis (Table 1). In
all cases RNA synthesis was increased.
Altogether, these results indicated that while the
presence of the SL-A1 domain is necessary for efficient
RNA synthesis, the SL-B1 region does not contain
Fig. 1. Secondary structure of the 3¢ ter-
minal sequences of HCV minus-strand RNA.

of (–)IRES RNA leading to (–)IRES 239 reduced
RNA synthesis by 51%. These results are in agreement
with those obtained by Oh et al. [18]. Further deletion
of the 5¢-end by 20 nt to give (–)IRES 219 showed a
striking reduction of RNA synthesis to only 19% of
that obtained with the (–)IRES wild-type RNA. Struc-
ture prediction by computer analysis showed that the
four bases at the 3¢-end of the (–)IRES 239 and
(–)IRES 219 were unannealed as in the wild-type
Fig. 2. Secondary structure of RNA mutated
in the SL-A1 and SL-B1 domains. Only the
secondary structure of the 151 nt from the
3¢-end of the minus-strand RNA is shown.
The computer predicted structure at 25 °C
of the same domain is shown for each mut-
ant RNA. The DG of the wild-type (–)IRES
RNA and those of the five mutant RNA are
indicated. The arrows or the curly bracket
showed the location of deletions or muta-
tions, respectively.
T. Astier-Gin et al. Binding and replication of 3¢-end of HCV minus RNA
FEBS Journal 272 (2005) 3872–3886 ª 2005 FEBS 3875
(–)IRES RNA and that the secondary structures of the
stem loops SL-A1, SL-B1, SL-C1 and SL-D1 were
unmodified. The 5¢-SL-E1 stem loop was unchanged in
(–)IRES 239 whereas in (–)IRES 219, the base of the
stem was replaced by a six nt bulge (Fig. 3). We also
performed a deletion of 237 nt at the 5¢-end to give
the (–)IRES 104. This deletion preserved the structure
of the SL-A1 and of the SL-B1 hairpin and the 4 nt at

ively, the elimination of this less stably structured
domain decreased RNA synthesis by increasing the
relative amount of structured regions giving rise to
templates poorly replicated by the NS5B.
Analysis of RNAs synthesized using wild-type
and mutant (–)IRES RNA as templates
To examine the products synthesized in the presence of
the mutated (–)IRES, an RdRp assay was performed
with the NS5B 1a (Fig. 4). The same amount of each
product (833 Bq) was loaded onto a 6% denaturing
polyacrylamide gel. All templates with mutations or
deletions in the SL-A1 or the SL-B1 stem loop allowed
synthesis of a major RNA product with the size of
the input template (Fig. 4A). No arrest bands were
observed during the synthesis with the exception of
RNA (–)IRES DSL-A1 that gave a small amount of a
product about 39 nt shorter than the template
(Fig. 4A). The relative quantity of this short product
was variable in different experiments. In all cases,
slower migrating bands were also observed. The major
one migrated to a position corresponding to an RNA
two times larger than the template. For the wild-type
(–)IRES RNA we have previously shown that this
product corresponds to two successive copies of the
template [17]. Products of higher molecular weight in
very low amounts were also visible. They may corres-
pond to three (or more) successive copies of the tem-
plate.
When RNA synthesis was performed with (–)IRES
RNA templates that have deletions in the 5¢-end, a

50% shifting of
32
P RNA. Results correspond to mean values of
3–4 independent experiments for each RNA. N.D., not determined.
RNA
RNA synthesis percentage
(–)IRES
K
d
(nM)
(–)-heparin (+)-heparin
(–)IRES 341 100 100 340 ± 40
(–)IRES DSL-A1 61 ± 7 59 ± 7 370 ± 10
(–)IRESD91-97 164 ± 20 154 ± 28 ND
(–)IRESDhp2 159 ± 15 165 ± 17 ND
(–)IRES hp2b 186 ± 14 173 ± 18 ND
(–)IRES LDH2 131 ± 16 101 ± 5 ND
(–)IRES 239 49 ± 8 40 ± 2 376 ± 31
(–)IRES 219 19 ± 4 13 ± 3 290 ± 30
(–)IRES 104 24 ± 3 25 ± 2 330 ± 30
(–)IRES 2 0 ND No binding
Binding and replication of 3¢-end of HCV minus RNA T. Astier-Gin et al.
3876 FEBS Journal 272 (2005) 3872–3886 ª 2005 FEBS
addition to the product twice the template size, the
(–)IRES 239 RNA gave a product of about 375 nt
indicated by a star (Fig. 4A). Again no prominent
arrest of RNA synthesis was observed when (–)IRES
RNAs harboring 5¢ deletions were used as templates.
Data presented in Fig. 4A were obtained by using
the NS5B D21 from HCV H77 genotype 1a. To exam-

) did not significantly modify
the level of RNA synthesis, suggesting that the amount
of heparin was sufficient to sequester all enzyme mole-
cules. These data were confirmed by performing a
kinetic experiment in the presence of heparin at
200 lgÆmL
)1
. As shown in Fig. 5B, the amount of syn-
thesized RNA greatly increased in the first 10 min of
the reaction to reach a plateau after 30 min. Analysis
of the RNA products on polyacrylamide gel showed
that the size of the products did not increase, indica-
ting that the elongation step was achieved (data not
shown).
We then compared the total level of RNA synthes-
ized in the presence or absence of heparin at
200 lgÆmL
)1
, using the various templates. Results
reported in Table 1 show that the level of RNA syn-
thesis obtained with the mutated templates (compared
Fig. 3. Predicted structure of the (–)IRES RNA with deletions of the 5’-end. The computer predicted structure at 25 °C and the DG values
are shown for three 5’-deleted mutants.
T. Astier-Gin et al. Binding and replication of 3¢-end of HCV minus RNA
FEBS Journal 272 (2005) 3872–3886 ª 2005 FEBS 3877
to wild-type RNA) was very similar under both condi-
tions. The only exception was observed in the case of
the (–)IRES LDH2 template, where the entire SL-B1
stem loop has been deleted. Results with this construc-
tion displayed a slightly but significantly reduced RNA

case, in addition to the template size product, a short
RNA product was present in relatively high amounts
suggesting that under these conditions HCV NS5B
often released from the 104 nt RNA template after ini-
tiation of RNA synthesis.
In addition to RNA synthesized from the 3¢-end of
the HCV minus-strand RNA, we have previously des-
cribed the products using the 3¢-end of the plus-strand
HCV RNA as template [17]. To assess whether the
high molecular weight RNA produced from one of
these plus-strand RNA fragments called (+)
3¢UTRNDX also disappeared, we performed experi-
ments in the presence of heparin. The (+) 3¢UTRNDX
template corresponded to 150 nt 3¢ of the NS5B cod-
ing sequence plus the 3¢UTR sequence deleted from
BA
Fig. 4. RNA synthesized by NS5B in the
presence of wild-type and mutated RNAs.
Wild-type and mutated (–)IRES RNAs were
used in RdRp assays. An aliquot of the reac-
tion products was precipitated by 10% TCA
and the radioactivity incorporated in newly
synthesized RNA determined as described
in Experimental procedures section. The
remaining of the products was purified by
phenol ⁄ chloroform extraction (1 : 1, v ⁄ v) and
precipitated by one volume of isopropanol in
the presence of 0.5
M ammonium acetate.
32

to HCV NS5B
Data described above indicate that deletions of the
stem loop SL-A1 at the 3¢-end or of 102, 122 or 237
nucleotides at the 5¢-end of the (–)IRES 341 nt RNA
diminished in vitro RNA synthesis directed by the
HCV NS5B. To assess whether the low level of RNA
synthesis was related to the binding of these templates
to the HCV RdRp, we performed gel shift assays.
Results of such an experiment are presented in Fig. 7.
They showed that the binding of both wild-type and
239 (–)IRES RNA was complete at 1 lm NS5B. In a
native polyacrylamide gel, the two RNA migrated as
one species (Fig. 7A); however, in some experiments
a slowly migrating band of RNA was observed
(see below). In our experimental conditions the
[
32
P]RNA ⁄ NS5B complex remained at the top of the
gel. The K
d
values for the wild-type and deleted RNAs
were determined from curves obtained as in Fig. 7B.
As shown in Table 1, with the unique exception of the
(–)IRES20 RNA that did not bind the NS5B, all four
mutated RNAs bound the viral polymerase with the
same affinity as the wild-type (–)IRES RNA.
Competition experiments were then performed with
all mutated RNAs and the wild-type (–)IRES RNA
for binding to the enzyme. NS5B (500 nm) and wild-
type [

RdRp reaction mixture without ATP and UTP. Various concentra-
tions of heparin were then added followed by ATP and 3H-UTP.
The reaction mixture was further incubated at 25 °C for 2 h. The
amount of radioactivity incorporated into the nucleic acids was
measured after TCA precipitation and plotted against heparin con-
centration. (B) An RdRp assay using 32P-UTP as labeled nucleotide
was performed as above in the presence of heparin (200 lgÆmL
)1
)
with wild-type (–)IRES RNA as template. Twenty microliters were
removed from the reaction mixture at different times. The amount
of radioactivity incorporated into the nucleic acids was measured
after TCA precipitation and plotted against the incubation time in
minutes.
T. Astier-Gin et al. Binding and replication of 3¢-end of HCV minus RNA
FEBS Journal 272 (2005) 3872–3886 ª 2005 FEBS 3879
to 97 nm for (–)IRES hp2b. On the contrary, higher
amounts of 5 ¢ deleted RNAs were needed to dissociate
the NS5B ⁄ [
32
P](–)IRES RNA complex (Table 2 and
Fig. 8B). As expected no competition was observed
between the wild-type (–)IRES and the (–)IRES 20
RNAs (data not shown). These results strongly suggest
that multiple domains of the 341 nt of the minus-
strand RNA 3¢-end are involved in the binding to
NS5B in particular the region located between nt 219
and 239. Consequently, one can hypothesize that the
5¢ deleted RNA did not bind NS5B in the same man-
ner as the wild-type RNA and could not efficiently dis-

)1
)
was then added followed by ATP and
[
32
P]UTP. The reaction mixture was further
incubated at 25 °C for 2 h. [
32
P]RNA prod-
ucts were quantified after TCA precipitation
of an aliquot of the reaction mixture as des-
cribed in Experimental procedures section.
32
P-labeled reaction products (833 Bq each)
were denatured and loaded onto a 6%
denaturing polyacrylamide gel. (A) RNA
products synthesized without or with hep-
arin (200 lgÆmL
)1
) by HCV or GBV-B NS5B.
The templates used corresponded to 3’
domains of plus or minus-strand HCV RNA.
(B) RNA products synthesized by HCV NS5B
in the presence of heparin with wild-type or
5’ deleted (–)IRES RNA.
Binding and replication of 3¢-end of HCV minus RNA T. Astier-Gin et al.
3880 FEBS Journal 272 (2005) 3872–3886 ª 2005 FEBS
synthesis by NS5B. It looks like, when increasing the
relative amount of unstructured domains in the tem-
plate, the RNA synthesis by NS5B is enhanced. How-

deleted mutants (–)IRES 219 and (–)IRES 239, con-
firmed these results. In both cases, these two templates
can sustain efficient elongation without polymerization
arrest and NS5B release from the template. The level
of RNA synthesized from (–)IRES 219 is in the range
of that observed with (–)IRES 104 but the addition of
20 nucleotides at the 3¢-end to give (–)IRES 239
enhances this process by doubling the amount of RNA
product (Table 1).
A striking observation when analyzing the products
obtained with our recombinant NS5B-1a is the pres-
ence of high molecular weight RNA two to three times
the size of the template. These products correspond to
successive copies of the (–)IRES RNA [17]. In this
study, we showed that these high molecular weight
products were also synthesized with the recombinant
RdRp of an HCV of genotype 1b (Fig. 4B) but not
with the NS5B of the highly related GBV-B virus
(Fig. 6A). These products were not specifically pro-
duced from templates derived from the 3¢-end of HCV
minus-strand RNA as they were also present when
RNA fragments of the 3¢UTR were used as templates.
Data from RdRp assays performed in the presence of
heparin indicated that these products occurred after a
reinitiation event on a different template except in the
case of a product of about 375 nt when using the
(–)IRES 239 as template. The nature of the latter
375 nt RNA remains to be elucidated.
A
B

nt 247 and 313 from the 3¢-end were present in their
templates derived from the 3¢-end of the minus-strand
RNA. This differed from our data that showed that
short products were only visible with (–)IRES 104. This
discrepancy could be explained by differences in RdRp
assays, particularly the divalent cation used. Kashiwagi
et al. [20] used manganese whereas we use magnesium
which is assumed to be the divalent cation involved in
the viral polymerase activity in infected cells.
With the exception of the (–)IRES 20, all our RNA
mutants bound the HCV NS5B but they compete with
the wild-type sequence with varying efficiencies. The five
RNAs carrying deletions or mutations in SL-A1 or
SL-B1 domains (DSL-A1, D91-97, Dhp2, hp2b and
LDH2) competed with (–)IRES RNA for binding on
NS5B with a slightly lower efficiency than the wild-type
RNA. On the contrary, the 5¢ deleted mutants were poor
competitors. Indeed, the amount of (–)IRES 239, 219
and 104 needed to displace labeled (–)IRES RNA were
four-, eight- and 16-fold higher than that of wild-type
RNA, respectively. The strong effect of the deletion of
nt 219–239 on RNA binding correlated with the effect
of this deletion on RNA synthesis, suggesting an
important interaction between this domain and the
AB
Fig. 8. Competition gel shift assay with wild-type and deleted (–)IRES RNA.
32
P-labeled (–)IRES RNA was incubated with NS5B (500 nM)and
different amounts of unlabelled RNAs in a 10 lL reaction mixture. RNA was renatured as described in material and methods section. After
20-min incubation at 25 °C, the products were analyzed in non denaturing polyacrylamide gels.

that this deletion affected the hinge between stem loops
SL-E1 and SL-AII. Notably, the apical loop and the
upper part of the stem of SL-E1 has a primary sequence
homology with the SL-II stem loop of the 3¢UTR
[13,14]. It has been previously shown that a recombinant
NS5B interacts with the 3¢X sequence of the 3¢UTR by
binding the stem of the SL-II stem loop and the hinge
between SL-I and SL-II [12]. More precise analysis by
site directed mutagenesis in the context of the 341 nt
RNA fragment is needed to identify the role of this stem
loop in NS5B binding and RNA synthesis from the
HCV minus-strand RNA.
Altogether these results suggest that several domains
of the RNA fragment comprising the 341 nt from the 3¢-
end of the minus-strand RNA are able to interact with
NS5B. This is reminiscent of the observation by Friebe
et al. [15] showing that the 341 nt of the 5¢UTR are nee-
ded for efficient replication of HCV RNA in the repl-
icon system. Their study showed that the first 125 nt of
the 5¢UTR are sufficient for a low level of RNA replica-
tion in agreement with our results. However, our data
differ from theirs since they observed that a deletion
between nucleotides 72–96 and 61–104 abolished RNA
synthesis whereas in our experiments deletion in this
domain had no such effect. This discrepancy may be
explained by the fact that we have used different systems
to study RNA synthesis and also that sequences identi-
fied in the replicon system may be involved in replica-
tion step(s) other than the synthesis of plus-strand RNA
from the minus-strand. In our case, we used the soluble

amplified from the sequence coding for the GBV-B-NS5B
kindly provided by A. Martin (Institut Pasteur, Paris,
France). The primers used were VB1: 5¢-AAACATATGA
GCATGAGCTACCACCTGGACC-3¢ and VB3: 5¢-CTCG
AGCTTCACAAGAAACTTCTGC-3¢. The PCR fragment
was cleaved by the restriction enzymes Nde1 and Xho1 and
inserted between the Nde1 and the Xho1 sites of the
pET21b. The HCV NS5B-1b and the GBV-B NS5B were
purified following the same procedure as for the HCV
NS5B-1a except that for the GBV-B NS5B a desalting col-
umn (HiTrap 5 mL desalting, Amersham Pharmacia Bio-
tech) was used instead of the monoS column after IMAC.
RNA templates
RNA (–)IRES corresponding to the 3¢-end of the minus-
strand RNA was synthesized by in vitro transcription of DNA
obtained by PCR amplification from the pGEM9Zf(–)
containing the 341 nucleotides of the 5¢UTR of HCV-H77
(pCU-UTRu). The PCR primers were designed to introduce
a T7 RNA polymerase promoter in the correct orientation
(Table 3). PCR was performed with the AmpliTaq gold
DNA polymerase kit (Applied Biosystem, Branchburg,
USA). RNAs were synthesized using the MEGAscript kit
(Ambion, Austin, TX, USA). DNA templates were digested
with DNase for 15 min. After phenol ⁄ chloroform extraction,
the RNAs were precipitated with isopropanol. The purity
and the integrity of RNAs were determined by analysis on a
6% polyacrylamide gel containing 7 m urea in TBE buffer
(90 mm Tris ⁄ borate pH 8.0, 1 mm EDTA). All RNA
mutants were obtained by modification of the 5¢UTR
sequence contained in the pCV-UTR4. The RNA (–)IRES

RdRp assay
The assay was performed in a total volume of 20 lL
containing 20 mm Tris ⁄ HCl pH 7.5, 1 mm DTT, 5 mm
MgCl
2
,40mm NaCl, 17 U RNasin (Promega, Madi-
son, WI, USA), 0.5 mm each of the 3 NTP (ATP, CTP,
GTP), 86 nm of RNA template, 150 nm of purified NS5B
and either 10 lCi [
32
P]UTP[a P] (3000 CiÆmmol
)1
, Amer-
sham Pharmacia Biotech) and 2 lm UTP or 2 lCi [
3
H]UTP
(46 CiÆmmol
)1
). The reaction mixture was incubated for 2 h
at 25 °C and stopped by the addition of 10% (v ⁄ v) TCA.
The radioactivity incorporated into newly synthesized RNA
was then determined. To quantify and analyze the
32
P-labe-
led RNA, the synthesis was stopped by adding 6.25 mm
EDTA, 10 mm Tris ⁄ HCl pH 7.5 and 0.125% (w ⁄ v) SDS.
An aliquot of the reaction products was precipitated by
10% (v ⁄ v) TCA and the radioactivity incorporated in newly
synthesized RNA determined as above. The remaining of
the products was purified by phenol ⁄ chloroform extraction

(–)IRES DSLA1 GCCAGACACTCCACCATGAATCACTCCCCTGTGAGGAACTACTGTCTTCACG
DSLA1 5’341T7
TAATACGACTCACTATAGGGTGCACGGTCTACGAGACCT
(–)IRES DHp2s TTTGCGGCCGCGCCAGCCCCCTGATGGGGGCGACACTCCACCATGAATTCT
Dhp2 AGCCATGGTTT
DHp2r AAACCATGGCTAGAATTCATGGTGGAGTGTCGCCCCCATCAGGGGGCTGGC
GCGGCCGCAAA
(–)IRES D91–97a GGCTGCACGACACTCCGCCATGGCTAGACGCTTTC
D91–97 D91–97b CGTCTAGCCATGGCGGAGTGTCGTGCAGCCTCCAGG
(–)IRES hp2bs CCCCTGATGGGGGCGTATTTCCACCATGAATCACTCCCC
hp2b hp2br GTGATTCATGGTGGAAATACGCCCCCATCAGGGGGCTGG
(–)IRES LDH2s TTTTGCGGCCGCGCCAGCCCCCTGATGGGGGCGCAGCCTCCAGGA
LDH2 CCCCCCCTCCCGGGAGAGCCATAGTGGTCTGCGGAACCGGTTTT
LDH2r AAAAACCGGTTCCGCAGACCACTATGGCTCTCCCGGGAGGGGGGG
TCCTGGAGGCTGCGCCCCCATCAGGGGGCTGGCGCGGCCGCAAAA
(–)IRES 239r
TAATACGACTCACTATAGGGGCACGCCCAAATCTC
239 5’S2 GCCAGCCCCCTGATGGGGGCGA
(–)IRES 219r AAA
TAATACGACTCACTATAGGCATTGAGCGGGTTTATCC
219 5’S2 GCCAGCCCCCTGATGGGGGCGA
(–)IRES 104r AAA
TAATACGACTCACTATAGACACTCATACTAACGCCATG
104 5’S2 GCCAGCCCCCTGATGGGGGCGA
+UTR3’ UTR 3a2 AAA
TAATACGACTCACTATAGCCGGCTGGACTTGTCCGG
NDX UTR3d GGAGCCACCATTAAAGAAGGG
Binding and replication of 3¢-end of HCV minus RNA T. Astier-Gin et al.
3884 FEBS Journal 272 (2005) 3872–3886 ª 2005 FEBS
Gel shift assay

amount of [
32
P]RNA released from the NS5B protein was
calculated from scanning analyses and plotted against the
concentration of unlabelled RNA. The 0 and 100%
released [
32
P]RNA corresponded to the values obtained in
the absence of unlabelled RNA with enzyme and without
enzyme, respectively. The dissociation constant was estima-
ted from the concentration of the unlabelled RNA resulting
in 50% dissociation of [
32
P]RNA.
Acknowledgements
We thank Laura Tarrago-Litvak for helpful discus-
sions and critical reading of the manuscript. This work
was supported by the Agence Nationale de Recherche
contre le Sida (ANRS), the Centre National de la
Recherche Scientifique (CNRS), the Institut National
de la Sante
´
et de la Recherche Me
´
dicale (INSERM),
The University Victor Segalen Bordeaux 2, the Ligue
contre le Cancer (Comite
´
de la Dordogne), and the
Re

Bukh J (1999) In vivo analysis of the 3¢ untranslated
region of the hepatitis C virus after in vitro mutagenesis
of an infectious cDNA clone. Proc Natl Acad Sci USA
96, 2291–2295.
8 Friebe P & Bartenschlager R (2002) Genetic analysis of
sequences in the 3¢ nontranslated region of hepatitis C
virus that are important for RNA replication. J Virol
76, 5326–5338.
9 Yi M & Lemon SM (2003) 3¢ nontranslated RNA sig-
nals required for replication of hepatitis C virus RNA.
J Virol 77, 3557–3568.
10 Tanaka T, Kato N, Cho MJ, Sugiyama K & Shimo-
tohno K (1996) Structure of the 3¢ terminus of the hepa-
titis C virus genome. J Virol 70, 3307–3312.
11 Blight KJ & Rice CM (1997) Secondary structure deter-
mination of the conserved 98-base sequence at the 3¢
terminus of hepatitis C virus genome. J Virol 71, 7345–
7352.
12 Oh JW, Sheu GT & Lai MM (2000) Template require-
ment and initiation site selection by hepatitis C virus
polymerase on a minimal viral RNA template. J Biol
Chem 275, 17710–17717.
13 Schuster C, Isel C, Imbert I, Ehresmann C, Marquet R &
Kieny MP (2002) Secondary structure of the 3¢ terminus
of hepatitis C virus minus- strand RNA. J Virol 76, 8058–
8068.
14 Smith RM, Walton CM, Wu CH & Wu GY (2002)
Secondary structure and hybridization accessibility of
hepatitis C virus 3¢-terminal sequences. J Virol 76,
9563–9574.

hepatitis C virus genome. J Biol Chem 277, 28700–
28705.
21 Yanagi M, St Claire M, Shapiro M, Emerson SU,
Purcell RH & Bukh J (1998) Transcripts of a chimeric
cDNA clone of hepatitis C virus genotype 1b are infec-
tious in vivo. Virology 244, 161–172.
22 Mottola G, Cardinali G, Ceccacci A, Trozzi C,
Bartholomew L, Torrisi MR, Pedrazzini E, Bonatti S &
Migliaccio G (2002) Hepatitis virus nonstructural pro-
teins are localized in a modified endoplasmic reticulum
of cells expressing viral subgenomic replicons. Virology
293, 31–43.
23 Moradpour D, Gosert R, Egger D, Penin F, Blum HE &
Bienz K (2003) Membrane association of hepatitis C
virus nonstructural proteins and identification of the
membrane alteration that harbors the viral replication
complex. Antiviral Res 60, 103–109.
24 Aizaki H, Lee K-J, Sung VM, Ishiko H & Lai MM
(2004) Characterization of the hepatitis C virus RNA
replication complex associated with lipid rafts. Virology
324, 450–461.
Binding and replication of 3¢-end of HCV minus RNA T. Astier-Gin et al.
3886 FEBS Journal 272 (2005) 3872–3886 ª 2005 FEBS