Báo cáo khoa học: Recent contributions of in vitro models to our understanding of hepatitis C virus life cycle pdf

On the other hand, production of infectious particles in Huh7 cells has been achieved with a clone named Japanese fulminant hepatitis 1 JFH1 constituting the so-called model of HCV cellu

Recent contributions of in vitro models to our

understanding of hepatitis C virus life cycle

Morgane Re´geard, Charlotte Lepe`re, Maud Trotard, Philippe Gripon and Jacques Le Seyec

INSERM, U522, IFR 140, Hoˆpital de Pontchaillou, Rennes, France

Introduction

The hepatitis C virus (HCV) belongs to the

Flaviviri-dae family and is the only member of the Hepacivirus

genus It is a small virus with a diameter of

 50 nm, enveloped within a cell-derived lipid

mem-brane that carries viral surface glycoproteins This

envelope surrounds a capsid containing positive

ssRNA The viral genome of  9600 nucleotides

con-tains two UTR at the 5¢- and 3¢-termini and a major

ORF that encodes a unique polyprotein of  3000

amino acids (Fig 1) Translation is initiated by the

internal ribosome entry site (IRES) located in the

5¢-UTR Translated polyprotein is then co- and

post-translationally cleaved into 10 different products: three structural proteins (the core protein and the E1 and E2 envelope glycoproteins) and seven nonstruc-tural (NS) proteins (p7, NS2, NS3, NS4A, NS4B, NS5A and NS5B) The specific enzymatic functions that have been attributed to NS2⁄ 3, NS3 and NS5B are serine protease, helicase and RNA-dependent polymerase, respectively

HCV is a human pathogen and  170 million peo-ple are chronically infected worldwide [1] HCV infec-tion causes major health problems because it is a principle cause of chronic liver diseases, including cir-rhosis and hepatocellular carcinoma The natural his-tory of HCV begins with a frequently asymptomatic

Keywords

assembly; hepatitis C virus; in vitro models;

infection; replication

Correspondence

J Le Seyec, INSERM U522, Hoˆpital

Pontchaillou, Avenue Henri Le Guilloux,

Rennes, F-35033, France

Fax: +33 2 99 54 01 37

Tel: +33 2 99 54 74 07

E-mail: jacques.leseyec@univ-rennes1.fr

(Received 14 June 2007, revised 25 July

2007, accepted 26 July 2007)

doi:10.1111/j.1742-4658.2007.06017.x

Hepatitis C virus is a human pathogen responsible for liver diseases includ-ing acute and chronic hepatitis, cirrhosis and hepatocellular carcinoma Its high prevalence, the absence of a prophylactic vaccine and the poor effi-ciency of current therapies are huge medical problems Since the discovery

of the hepatitis C virus, our knowledge of its biology has been largely punctuated by the development of original models of research At the end

of the 1980s, the chimpanzee model led to cloning of the viral genome and the definition of infectious molecular clones In 1999, a breakthrough was achieved with the development of a robust in vitro replication model named

‘replicon’ This system allowed intensive research into replication mecha-nisms and drug discovery Later, in 2003, pseudotyped retroviruses har-bouring surface proteins of hepatitis C virus were produced to specifically investigate the viral entry process It was only in 2005 that infectious viruses were produced in vitro, enabling intensive investigations into the entire life cycle of the hepatitis C virus This review describes the different

in vitro models developed to study hepatitis C virus, their contribution to current knowledge of the virus biology and their future research applica-tions

Abbreviations

HCV, hepatitis C virus; HCVcc, cellular clone of HCV; HCVpp, pseudo-particles of HCV; IFN, interferon; IRES, internal ribosome entry site; JFH1, Japanese fulminant hepatitis 1; LDLR, low-density lipoprotein receptor; NS, nonstructural; SR-B1, scavenger receptor class B type 1; VSV, vesicular stomatitis virus.

acute phase of infection that leads to chronic infection

in  70–80% of cases Thereafter, 10–20% of

chroni-cally infected patients develop liver cirrhosis within

20 years and hepatocellular carcinoma after another

decade No vaccine against HCV infection is available,

and current antiviral therapies consisting of pegylated

interferon (IFN) and ribavirin injections are

character-ized by limited efficacy, substantial side effects and

high cost These clinical complications clearly

docu-ment the need for more effective therapies that depend

on a detailed understanding of HCV biology using

appropriate experimental systems Unfortunately,

research on HCV has largely been slowed by the

diffi-culties encountered in developing efficient experimental

models This review focuses on the different in vitro

models of HCV that have been developed and their

contribution to our current knowledge of the virus life

cycle

Around 10 years after discovery of the HCV genome

and after many attempts to infect chimpanzees with

transcripts from cloned isolates, consensus sequences

of genotypes 1a, 1b and 2a were constructed These

were the first viral functional sequences able to infect

chimpanzees Soon after, efforts were concentrated on

establishing cell culture models that support HCV

rep-lication by transfecting cells with cloned viral DNA or

their derived viral transcripts Although this approach

is classic in virology, it proved to be unproductive for

HCV because of the very low level of replication and

the high amount of input RNA needed for

transfec-tion These first studies precluded the difficulties of

studying HCV in vitro

Replicon system

An important breakthrough was the development of cell-culture systems based on the selection of cells that support stable replication of subgenomic HCV RNAs Lohmann et al [2] worked on an HCV consensus gen-ome of genotype 1b derived from a chronically infected patient Researchers replaced the region that encodes the core to p7 with the coding sequence of the neomycin-resistance gene and the heterologous IRES

of the encephalomyocarditis virus The resulting repli-con was bicistronic with translation of the first cistron (neomycin-resistance gene) being directed by the HCV IRES and that of the second cistron (NS2–5B) by the encephalomyocarditis virus IRES Other constructions were composed of a smaller second cistron encoding NS3 to NS5B proteins (Fig 2A) After transfection of Huh7 cells with this replicon, selection of the very few cells supporting autonomous replication was achieved

by neomycin sulfate treatment (Fig 2Ba) Viral repli-cation was sufficient to detect viral RNA by northern blot analysis Improvement of the system was obtained after the discovery of cell-culture-adaptative mutations that enhanced the replication efficiency by up to

10 000 times [3] These mutations are at the N-termi-nus of the NS3 helicase, in two distinct positions of NS4B, in the centre of NS5A and in the C-terminal region of NS5B [3–6] The significance of these muta-tions has been questioned because they have not been observed in wild-type viruses Moreover, insertion of some of these mutations into an infectious HCV clone reduced or completely abolished its in vivo infectivity

2 E 1 E

A 4 S N

B 5 S A 5 S

s n i e t o r p c l G e m o e G n i t a i s p c n

n I l e n h c

e s a t o r p e i r e S

e s a c il e A R

s r o t c a f o c e s a t o r p e i r e S

t n d e e -A R

e s a r e m y l o A R n i e t o r p h s o P

s n i t a r e t l a e a r b m e M e

i e t s C e s a t o r p

7

F

R T U

’ 3 R

T U

’ 5

l a r u t c r t s

) n 0 9 ( A R V H

n i t a l s n r T d A n i t a r u t a M

B 4 S

l a r u t c r t s o

Fig 1 Genetic organization and procession of HCV polyprotein A schematic representation of HCV genome is given at the top The HCV genome is composed of ssRNA encoding a large ORF flanked by 5¢- and 3¢-UTR Translation of the polyprotein precursor is mediated by the IRES contained in the 5¢-UTR The polyprotein is co- and post-translationally processed in 10 proteins by signal peptide peptidase (black solid arrows), by NS2 ⁄ 3 autoprotease (black and large arrow) and by NS3 ⁄ 4A protease (black doted arrows) The F protein is generated by transla-tion of an alternative reading frame but no functransla-tions have yet been attributed to this protein.

in chimpanzees [7] In parallel, the replicon system has

allowed the selection of highly permissive cell clones

(Fig 2Bb) Indeed, the subpopulation of Huh7 cells

that supports a high viral replication rate has been

cured from the replicon by long-lasting IFN treatment

Two such cell lines have been generated and named

Huh7-Lunet and Huh7.5 Another subclone, Huh7.5.1,

has been generated similarly by curing Huh7.5 cells of

replicating HCV All of these cell lines were shown to

support RNA replication to a much greater extent

than the parental cell line [8–10] The efficient

replica-tion of HCV in these cells may be explained by partial

impairment of their antiviral defence system Indeed,

Landford et al [11] suggested that some steps in the

signalling pathway for detecting dsRNA were defective

in the parental Huh7 cells Moreover, permissiveness

of HCV replication in Huh7.5 cells is probably

rein-forced by the presence of a defective mutation in the

RIG-I gene, which disturbs the antiviral immune

response [12] The antiviral effect of IFNa observed

in vivo was nevertheless reproduced in this system

[4,11] Thereafter, improvements to this model were

achieved with the efficient insertion of a reporter gene

into the viral genome facilitating measurement of the

replication activity of replicons [5,13–15] Replicons of

other genotypes (1a and 2a) have also been developed

[16,17] Genomic HCV replicons have also been gener-ated and have enabled the selection of cells with stable expression of the entire viral polyprotein (Fig 2A) However, replication efficiency was lower than that observed with subgenomic replicons and no virus pro-duction was observed in these cells [16,18,19] This defect could be due to the presence of adaptative mutations that are detrimental to viral particle assem-bly and secretion or to the lack of some critical HCV partners in Huh7 cells Some data argue for the former hypothesis On the one hand, inoculation of chimpan-zees with a Con1 sequence containing these adaptative mutations failed to establish a productive infection [7]

On the other hand, production of infectious particles

in Huh7 cells has been achieved with a clone named Japanese fulminant hepatitis 1 (JFH1) constituting the so-called model of HCV cellular clone (see below) [20] Owing to its efficient replication rate, this in vitro replicon model enabled investigation into the replica-tion process, the replicareplica-tion complex and host–virus interactions [21,22] Its exploitation has also enabled high-throughput screening of anti-HCV drugs targeting the replication of various genotype replicons [23] Although practical, only the replication step containing

in vitro adaptative mutations can be studied with this system using viral genomes containing in vitro adaptive

A 4 S 2

o N

R T U

’ 5 V

c i m o e b S n c il p r

c i m o e G n c il p

7

o N

R T U

’ 5 V

3

o N

R T U

’ 5 V

V M C S R I

d a n i t c l e 8 4 G

n i s a x l a o l c

t n m t a r t N F I

f o n i t a r o o r t c l E

h t i w s ll e 7 u H

A

s n l c u s ll e 7 u H

V C g i t a il p r s ll e 7 u H

r o c i m o e b S

n c il p r c i m o e

c i m o e b S n c il p r

B

V M C S R I

B 4 S

A 4

S S 4 B

A 4

S S 4 B

b a

V M C S R I

Fig 2 Schematic representation of the

rep-licon system Subgenomic and genomic

replicons are composed of the HCV 5¢-UTR,

the gene coding neomycin

phosphotransfer-ase (Neo R ), the encephalomyocarditis virus

IRES, the region encoding HCV proteins and

the 3¢-UTR (A) Huh7 cells are

electroporat-ed with replicon RNA Cell colonies

effi-ciently replicating the HCV replicon are

selected because of their resistance to

G418 (Ba) In parallel, Huh7 subclones highly

permissive to HCV replication can be

obtained by G418 treatment of cells

trans-fected with HCV subgenomic replicon Cells

are then treated by IFN to eliminate the

HCV replicon (Bb).

mutations Furthermore, it should be kept in mind

that, in this system, hundreds of RNA copies per cell

are present, in contrast to 5–50 copies in infected

hepatocytes Therefore, results should be ascertained

using systems closer to HCV physiology

Infection of primary cell cultures and

cell lines

Parallel to the development of this replication system,

intensive research has aimed to discover HCV

infec-tion systems Until recently, sera obtained from

infected patients or chimpanzees were the only source

of HCV infectious particles However, purification of

natural HCV particles from patient sera proved

diffi-cult because of the heterogeneity of their densities

Viral RNA is detected in fractions ranging from 1.03

to 1.25 gÆmL)1 in a sucrose gradient Low-density

HCV particles are associated with either low- or

very-low-density lipoproteins [24–28] and have been shown

by assays of chimpanzee model to be the most

infectious fraction [29,30] HCV particles of higher

densities correspond to particles associated with

immunoglobulin, free particles [28,30,31] or free

nucleocapsids [32]

Because hepatocytes are the main target of the virus

in vivo, several groups have attempted to infect

pri-mary hepatocytes or hepatic cells in vitro Thus, adult

or fetal primary human hepatocytes have been shown

to support HCV infection and replication in vitro [33–

36] Similarly, infections have also been conducted on

primary cells obtained from other mammals, including

chimpanzees and tree shrews [37,38] In these models,

the replication rate was low, between 0.01 and 0.1

RNA copies per cell, depending on the experiment

Therefore, highly sensitive assays were required to

detect viral RNA within infected cells or in cell-culture

supernatants In order to demonstrate that infection

had taken place, researchers put forward

supplemen-tary data: detection of HCV negative-strand RNA,

which only appears during ongoing replication; the

sensitivity of replication to IFNa treatment; secretion

of neosynthesized virions able to infect naive cells; and

selection of quasispecies during the culture of infected

hepatocytes To succeed in obtaining HCV infection in

primary human hepatocytes, an existing model of

hep-atitis B virus infection was used to determine optimal

infection conditions [39,40] Rumin et al pointed out

the need to reach high levels of cellular differentiation,

which they suggested may account for the ability of

these cells to support virus assembly and secretion [36]

Similarly, primary hepatocytes have recently been

cul-tivated in spheroid formation because this culture

con-dition maintains the differentiation state of the cell However, no real improvement in viral replication effi-ciency was achieved with this new model [41] It has also been noted that, for unknown reasons, sera from patients are not always infectious and no obvious cor-relation can be drawn between infectivity and viral RNA titre or with the presence of antibodies directed against structural proteins [36] Although primary human hepatocytes infected with patient sera are the most physiological in vitro model at present, the diffi-culty of obtaining cells and intrinsic technical con-straints make it hard to use in everyday experiments This may explain the limited number of studies based

on this model However, recent work using this model supported the involvement of the low-density lipopro-tein receptor (LDLR) in the entry process of HCV The soluble form of LDLR, natural LDLR ligands and antibodies directed against LDLR efficiently com-peted with HCV infection, suggesting that LDLR

is probably involved in the entry process of native HCV [42]

In parallel, some groups have focused their efforts on developing an in vitro model based on hepatic cell lines [43] Among those tested, HepG2, Huh7 and PH5CH, the latter was the most susceptible to infection and replication However, the replication rate in PH5CH remained low as viral RNA could only be detected using RT-nested PCR Recently, Aly et al immortal-ized primary human hepatocytes with human papilloma virus E6E7 genes [44] These HPV18⁄ E6E7-immortal-ized hepatocytes could maintain hepatic-specific mark-ers in long-term culture and were susceptible to HCV infection, as assayed by RT-QPCR detection of the intracellular HCV positive RNA strand However, viral replication efficiency was low compared with the infec-tion system based on primary human hepatocytes (102 and 104 copies of viral RNA per microgram of cellular RNA, respectively) [42] mAbs directed against CD81, another probable component of the receptor complex, and IFNa treatment, inhibited infection and replication, respectively Interestingly, an interferon regulatory factor-7-defective form of this cell line has been engineered by stable expression of transdominant mutant interferon regulatory factor-7 These deficient cells were more susceptible to infection Indeed, hundred more copies of HCV RNA could be detected inside these cells following infection, whatever the geno-type (1b, 2 and 3) However no production of progeny viruses was shown in this infection model

In some patients infected with HCV, analysis of HCV negative-strand RNA by RT-PCR indicated its presence in both the liver and haematopoietic cells [45] Growing evidence supports the idea that HCV is

also lymphotropic and lymphocytes may be an HCV

reservoir In fact, Cribier et al reported the in vitro

infection of primary peripheral blood mononuclear

cells with high-titre sera Despite the low replication

efficiency, HCV RNA was detected for a month in cell

culture [46] Several other laboratories have shown that

HCV could infect a B-cell line (Daudi) and T-cell lines

(MT-2 and MOLT-4) in vitro [21] In a more recent

study by Sung et al., proof of replication in this cell

type was demonstrated by the establishment of a B-cell

lymphoma cell line derived from an HCV-infected

patient with type II-mixed cryoglobulinaemia [47]

This cell line persistently replicated the HCV genome

and produced virions that were infectious in primary

human hepatocytes and lymphocytes in vitro

In summary, use of HCV-containing sera to

recon-stitute the entire life cycle of HCV in vitro has proved

to be very difficult Although infection of primary cells

has been shown with convincing data, low replication

efficiency and inherent technical difficulties have

lim-ited their use The development of the newly described

HPV18⁄ E6E7-immortalized hepatocytes might

consti-tute an easier model to conduct further analyses

Par-allel to the intensive research discussed thus far, other

groups have developed surrogate models to investigate

specific steps of HCV life cycle

HCV-like particles

HCV-like particles are generated by self-assembly of

the HCV structural proteins and are nonreplicative

The first HCV-like particle model was described by

Baumert and collaborators in 1998, with particles

pro-duced in insect cells using a recombinant baculovirus

containing the cDNA of HCV structural proteins of

genotype 1b or 1a [48] HCV-like particles were

observed by electron microscopy in intracellular

com-partments but were not secreted in the supernatant

Consequently, purification of HCV-like particles was

achieved by cell lysis followed by sucrose-gradient

purification HCV-like particles are described as being

40–60 nm in diameter and of a rather high density

( 1.17 gÆmL)1) that should correspond to the density

of the free viral particles contained in the serum of

infected patients [48] Structural characterization of

HCV-like particle envelope proteins has been

con-ducted by analysing their antigenic properties This

was done using a large panel of monoclonal and

con-formational antibodies directed against E1 and E2,

and sera from infected patients [49,50] Results

suggested that E1 and E2 located at the surface of

HCV-like particles formed E1E2 heterodimers in a

virion-like conformation

This model has essentially been used in two major fields of research: binding process and vaccination development Specific binding of HCV-like particles was obtained for various hepatic and lymphocyte cell lines and also for dendritic cells, independently of CD81 expression [51,52] The limited involvement of CD81 in the binding process was further illustrated by the poor binding inhibition achieved in the presence of mAbs directed against CD81 In contrast, heparin sul-fate seemed to mediate this interaction [53] Taking advantage of this binding model, a recent study showed that interaction of envelope glycoproteins with the surface of HepG2 cells induces gene expression modulations This suggested that HCV binding might induce changes in the cell that could favour HCV infection [54] In parallel, virus-like particles could con-stitute attractive vaccine candidates for

papillomavirus-es and retroviruspapillomavirus-es because they could mimic some properties of native viruses Concerning HCV-like par-ticles, it has been shown that they are able to induce humoral and cellular immune responses in BALB⁄ c mice, baboons and chimpanzees [55–57] Immunized chimpanzees were thereafter inoculated with HCV of homologous genotype Although vaccinated chimpan-zees became infected by HCV, the infection was controlled quickly compared with unvaccinated animals [55]

Although the structural, biophysical and antigenic properties of HCV-like particles have been character-ized and might partly mimic those of native HCV parti-cles and be close to pseudotyped partiparti-cles described later, the binding of HCV-like particle does not require CD81 participation However, CD81 at the surface of Huh7 cells has been shown to be a crucial receptor involved in the infection process of pseudotyped viruses and cellular clones of HCV (see below) Moreover, this first HCV-like particle model does not permit investiga-tion into HCV morphogenesis because viral budding was not observed in insect cells By contrast, Blanchard

et al developed another HCV-like particle model that could be used to investigate this issue HCV-like parti-cles were produced in mammalian cells (BHK-21) by expression of HCV structural proteins in a Semliki for-est virus vector [58] Budding of HCV-like particles with a diameter of 50 nm was observed using electron microscopy and occurred at the endoplasmic reticulum membrane towards the lumen Although most of these HCV-like particles seemed to display an abortive bud-ding process, it was shown that the correctly processed HCV core protein drives this event [59,60] However, due to the absence of complete budding, this model cannot be used to study the following steps of HCV assembly in eukaryotic cells

HCV pseudo particles

A few years after the development of HCV-like

parti-cles, another model was created to specifically

investi-gate the entry process of HCV This system is called

pseudo particles of HCV (HCVpp), as envelope

glyco-proteins of HCV are incorporated at the surface of

other enveloped viruses substituting their natural

enve-lope proteins The first constructed pseudotyped

viruses were vesicular stomatitis virus (VSV)⁄ HCV

pseudotypes expressing HCV E1 and⁄ or HCV E2

chimeric proteins These contain the transmembrane

and cytoplasmic domains of envelope protein G of

VSV [61] Although pseudotyped virus infectivity was

neutralized by antibodies directed against E1 and E2

or by sera from HCV-infected chimpanzees or humans,

these HCVpp exhibited a surprising tropism with

lim-ited infectivity on primary human hepatocytes, and

better infectivity on kidney cell lines of human or

non-human origin [61–63] One may speculate that the

broad tropism of these HCVpp might be influenced by

the background infectivity of VSV Moreover, some of

these viruses harboured only one of the two

glycopro-teins suggesting that E1 and E2 could independently

carry out the entry process of these HCVpp Because

E1 and E2 are thought to be assembled in

hetero-dimers at the surface of native viruses, it seems likely

that both E1 and E2 are required for the infection

process

A second generation of HCVpp (Fig 3) has been

developed using unmodified E1 and E2 envelope

glyco-proteins, which are exposed at the surface of

retro-viral particles carrying a genome with a marker gene

Retrovirus budding occurs at the plasma membrane,

although some data indicate that it might also exist

intracellularly [64] Despite the specific endoplasmic

reticulum retention of HCV envelope glycoproteins

[65,66], it has been shown that in overexpression

sys-tems, a small fraction of envelope glycoproteins could

be secreted to the surface membrane via the secretory

pathway This led to Golgi-specific modification of

gly-cosylation in envelope proteins [67,68] HCV envelope

glycoproteins at the surface of retroviral particles are

comprised mainly of correctly folded E1 and E2

assembled as heterodimers and a small fraction of E1

and E2 covalently linked in aggregates [68] In this

model, expression of both E1 and E2 should derive

from a unique expression construct for optimal

infec-tivity [67,69] Various HCVpp have been developed

with envelope proteins of genotypes 1a, 1b, 2a, 3a, 4a,

5a and 6a allowing analysis of cross- and

genotype-specific neutralization [67,70] The presence of a

mar-ker gene packaged in these HCVpp has enabled easy

evaluation of infectivity mediated by HCV glycopro-teins Using this system, the tropism of HCVpp has been studied and is, with few exceptions, liver specific [67,69–71]

Because of this model, numerous questions regard-ing the HCV entry process could be assessed On the one hand, the receptor candidates scavenger receptor class B type 1 (SR-B1) and CD81 have been evaluated

in Huh7 cells Whereas CD81 has been shown to be directly involved in the entry process, SR-B1 influ-enced HCVpp entry via its cholesterol-uptake activity [68–74] In fact, high-density lipoprotein was shown to promote HCV entry and reduce the inhibitory effect of HCV-neutralizing antibodies from infected patients [73,75] On the other hand, the route of HCVpp entry has been assessed in Huh7 cells using specific drugs that increase endosomal pH or disrupt clathrin vesi-cles, with transdominant mutants of the GTPases Rab5 and Rab7 and small interfering RNA directed against the clathrin heavy chain These results support the hypothesis that HCVpp penetrate cells using

R e rt

e r

2 E 1 E -V C V M C V

M

C G a g - o l

r e t r o e R

Ψ

R T V M

c a

b

n i c e f s n r T

T 9 K E H

n i c u o r p p V C

R e o rte r

R e rte r

s n i l e a m o t a e f o n i t c f n I

n i e t o r p r e t r o e r f o s i s l a a d a

n i s e r p e

Fig 3 Production of HCV pseudoparticles To produce recombinant retroviruses 293T are transfected with three expression vectors The first (a) is the packaging construct that encodes for retroviral Gag and Pol proteins After translation of the second vector (b), the RNA produced and which contains the sequence of a reporter gene could be encapsidated in particles via the presence of the retroviral encapsidation sequence (Y) The third vector (c) encodes HCV E1 and E2 glycoproteins Recombinant viruses collected from the supernatant are made up of a retroviral capsid containing a RNA genome with the HCV glycoproteins at their surface Their specific infectivity on hepatoma cell lines was analysed by expression of the reporter gene.

clathrin vesicles and passage in the early endosome is

necessary for fusion between the viral envelope and

an intracellular membrane [69,71,76,77] This fusion

should be carried out by a fusion peptide present in

viral envelope protein(s) Recently, data obtained with

HCVpp suggested that E1 and E2 might both contain

membrane fusion determinants, underlying a potential

difference with the fusion process of other flaviviruses

[78,79]

Although this model has technical advantages (easy

culture system and read out) and has enabled large

advances in our knowledge of HCV, it represents only

one category of HCV form derived from the sera of

infected patients: free viruses not associated with either

lipoproteins or immunoglobulins Moreover, the

chi-meric nature of the viruses and the intrinsic

character-istics of the Huh7 cell line (its nonpermissiveness to

infection with HCV-containing sera) limit absolute

extrapolation of the results to native HCV Thus

results should be reproduced in a cell-culture system

closer to physiological conditions

Cellular clone of HCV

A major breakthrough has been achieved in the in vitro

modelling of HCV propagation with the development

of a cellular clone of HCV (HCVcc) This system is

based on the utilization of a very particular HCV

molecular clone of genotype 2a obtained from a

Japa-nese patient with fulminant hepatitis (JFH1) First

results with this JFH1 clone were obtained using the

subgenomic replicon model After transfection of Huh7

cells with this replicon RNA,  4.5 · 104cfuÆlg)1

RNA were counted, whereas only 9· 102cfuÆlg)1

RNA were obtained with transfection of the con1

rep-licon of genotype 1b harbouring in vitro adaptative

mutations Moreover, no adaptative mutations seemed

to be required in the subgenomic JFH1 replicon for its

efficient replication in cell culture [17] The JFH1

sub-genome was also shown to replicate efficiently in other

human cell lines of hepatic or nonhepatic origin

(HepG2, IMY-N9, HeLa, HEK 293) and in mouse

embryonic fibroblasts [80–82] Soon after the discovery

of the extraordinary replication potential of the

subge-nomic JFH1 replicon, Wakita et al [20] described how

transfection of Huh7 cells with the whole JFH1 RNA

sequence led to the production of viruses shown to be

infectious in vitro on naive Huh7 cells and in vivo in

chimpanzees (Fig 4) Use of Huh7.5 or Huh7.5.1 cell

lines further optimized the kinetics of replication and

HCVcc secretion [10,83] Because of this model, it has

been possible to observe cell-derived HCV

parti-cles, which showed a diameter of 55 nm when

analy-sed by electron microscopy [20] In parallel, density analyses of infected-cell supernatant suggested that HCVcc are associated with lipoproteins with an hetero-geneous density peak correlated to specific infectivity ranging from 1.05 to 1.1 gÆmL)1 in sucrose gradients [10,83,84] However, in vitro association of HCVcc with lipoproteins might not be as optimal as in vivo passage of this virus in chimpanzees or in mice con-taining human liver xenografts In fact, enhanced virus infectivity is observed after in vivo passage and is cor-related with a lower density of HCV particles [85] Further analyses are needed to characterize the associ-ation between lipoprotein and HCVcc in cell culture and determine whether it reflects that seen in patient sera Most importantly, the efficiency of this model is entirely attributed to this specific JFH1 molecular clone Attempts to reproduce this model with other HCV molecular clones of various genotypes showed rather limited results with very low virus release despite demonstration of their infectivity in chimpan-zees [86–88] To expand the studies to other genotypes, intergenotypic and intragenotypic chimeras have been constructed replacing the structural proteins of the JFH1 clone with those of Con1, H77, J6 and

HCV-452 molecular clones of genotypes 1b, 1a, 2a and 3a, respectively [89,90] With the exception of the HCV-452–JFH1 chimera, efficient virus production was obtained Crossover before or inside the NS2 sequence

R T U

’ 3

C E 1 E 2

7

R T U

’ 5

3 S N

1 H F J A R V H

c V H n i c d r p

7 u H f o n i c f n I

s n il e

a r i v f o s i s l a n a n i s e r p x E

n i e t o r p r e t r o e r r o n i e t o r p

n i n i a r o o r t c l E

s n il e 7 u H

A 4 S

2 S

N N S 5 A N S 5 B

Fig 4 The HCVcc model Huh7 cell lines are electroporated with the RNA transcripts of the JFH1 genome A few days after trans-fection, viruses are secreted in the supernatant of cells replicating HCV JFH1 genome Their specific infectivity and replicative poten-tial can be assessed on Huh7 cell lines and analysed by the expres-sion of viral or reporter proteins or the quantification of intracellular viral RNA.

had no significant effect on replication efficiency but

was shown to be critical for efficient virus production

[89,90] Concerning the H77–JFH1 chimera, mutations

affecting E1, p7, NS2 and⁄ or NS3 have been detected

and have contributed to improved assembly and

release of viral particles [90] In addition to the utility

of these chimeras to decipher the HCV entry process

and assess antibody-neutralization efficiency, these

experiments point to the involvement of p7 and NS2

in HCV morphogenesis and release

This recent in vitro model reproducing the entire life

cycle of HCV in Huh7 cell lines has enabled extensive

research into various areas of HCV biology In the

field of HCV entry, much effort has been employed to

confirm results obtained using the HCVpp model For

example, HCV entry into target cells is mediated by

E2 [10], and CD81 has been shown to be critical for

HCVcc infectivity In fact, antibodies directed against

CD81 and CD81 downregulation with RNA

interfer-ence inhibited HCVcc entry into Huh7 cell lines

[10,20,83,91] Moreover, recent studies analysing the

differential permissiveness of hepatic cell lines have

shown that CD81 surface expression was a key

requirement for HCVcc infection [83,92] SR-B1 is

another putative HCV receptor that has been

exten-sively studied with HCVpp mAbs directed against

SR-B1 and oxidized low-density lipoproteins, a ligand of

high affinity for SR-B1, competed with HCVcc entry,

supporting the hypothesis that SR-B1 is involved in

HCV entry [93,94] Most recently, Claudin-1, a

compo-nent of the tight junctions, has been proposed to be

involved in the entry process of HCVpp and HCVcc

[95] In fact, exogenous complementation of human

cells expressing CD81 and SR-B1 (293T and SW13)

with Claudin-1 permitted their infection with HCVpp

or HCVcc Events late in the entry process of HCV

have also been studied and various results supported

the hypothesis drawn up using the HCVpp model:

HCV uses a clathrin-dependant pathway to enter

Huh7 cell lines in a pH-dependant fashion [76,96,97]

In the field of virus morphogenesis, the HCVcc model

enabled to incriminate p7 and NS2 in the assembly of

infectious viral particles [89,90,98] Using this model,

the association of lipoproteins with virions during their

egress has also been suggested [84,99] However, only

limited investigation could be conducted into HCV

morphogenesis due to the very rare observation of

particles in producing cells [100] Finally, this model

has been also used to study virus–host interactions As

a consequence of long-term HCVcc propagation,

resis-tant cells are selected and adaptated viruses emerge,

displaying a better infectivity partly related to a single

mutation in E2 [101]

Although the HCVcc model is dependent on the spe-cific sequence of JFH1 5¢-UTR, NS proteins and 3¢-UTR, its utilization has led to great advances in our knowledge of HCV biology Urgent information is needed to determine the genetic specificity of this JFH1 molecular clone which confers high replication efficiency and virus release Nevertheless, viruses expressing exogenous marker are becoming useful tools

to investigate both HCV biology and the potency of antiviral drugs [20,96,97,102] One example of the con-tribution of these constructs is the discovery of a pro-tective effect that cells already replicating HCV possess against HCV superinfection [102,103]

Concluding remarks

Table 1 presents all the in vitro HCV models, their potential use and limitations Increasing data about viral life cycle mechanisms have been accumulated in recent years, particularly regarding entry and replica-tion processes The recent HCVcc model has emerged

as the most useful research tool to date

Several membrane receptors are involved in the HCV entry process: CD81, SR-B1, LDLR and Claudin-1 The multiplication of candidate receptors and the potential synergistic role of CD81 and SR-B1 support the hypothesis of a multistep entry pathway that may involve different receptors Moreover, one should keep

in mind that HCV particles exist in patient sera in vari-ous forms, either free or associated with lipoproteins or immunoglobulin One can therefore speculate that, depending on its form, HCV may take advantage of different sets of receptors to enter into target cells Whereas the implications of CD81, SR-B1 and Clau-din-1 have been ascertained using the HCVpp and the HCVcc model, the role of LDLR has been assessed in primary human hepatocytes with serum-derived HCV particles Further analyses are needed to define for each form of HCV particle, the exact sequence of events and the precise implications of these candidates as attach-ment or as entry receptors In parallel, growing evidence supports an internalization of HCV by endo-cytosis followed by passage in the early endosome However, neither the exact location of the fusion pep-tide in the viral surface glycoproteins nor the fusion process has been clearly documented Given the hetero-geneity of serum-derived HCV particles, it cannot

be excluded that multiple entry pathways may lead

to productive infection with potentially different effi-ciencies

In the field of HCV replication, a large break-through was the development of subgenomic replicons that allowed examination of the viral components of

the replication complex and identification of the

cellu-lar partners of HCV replication Although discovery of

the adaptative mutations necessary for efficient in vitro

replication enabled high-throughput studies, it also

raised the as yet unsolved question of the mechanisms

of replication enhancement driven by these mutations

More recently, the JFH1 subgenome has been shown

to replicate at high levels without any adaptative

muta-tions Analysis has shown that this may be due to a

more efficient initiation of RNA synthesis by JFH1

NS5B compared with that of the Con1 clone [104]

Moreover, and despite the generation of intergenotypic

chimera, the HCVcc model is based on the

extraordi-nary replication efficiency of the unique JFH1 strain

and therefore requires minimal sequences of this

pecu-liar strain: the 5¢-UTR sequence and the C-terminus

NS2 or the NS3 sequence to the 3¢-UTR More

infor-mation is needed to better understand this greater

rep-lication efficiency and the determinants necessary for

the production and release of infectious viruses

Another limitation of both the replicon system and

the HCVcc model is their requirement for Huh7 cell

lines that possess an impaired innate antiviral

response HCVcc has also been shown to be infectious

and able to replicate efficiently in the newly described

HPV18⁄ E6E7-immortalized hepatocytes In this

experi-ment, interferon regulatory factor-7, a regulatory

fac-tor of IFN response, was downregulated It would

therefore be interesting to evaluate HCVcc infectivity

in primary human hepatocytes that possess intact antiviral responses HCVcc infection studies would then be assessed in a more physiological context and could be compared with serum-derived HCV infection Moreover, it would be of great interest to better understand all the antiviral mechanisms developed by primary human hepatocytes that may explain their lim-ited rate of infection

Acknowledgements

We thank C Gamble for her critical reading of the manuscript Our research is supported by grants from the ANRS and the Ligue Nationale contre le Cancer

CL and MT are supported by fellowships provided by the Ministe`re de l’Education et de la Recherche and the Re´gion Bretagne, respectively

References

1 WHO (1997) Hepatitis C global prevalence Wkly Epidemiol Rec 72, 341–344

2 Lohmann V, Korner F, Koch J, Herian U, Theilmann

L & Bartenschlager R (1999) Replication of

subgenom-ic hepatitis C virus RNAs in a hepatoma cell line Science 285, 110–113

3 Blight KJ, Kolykhalov AA & Rice CM (2000) Efficient initiation of HCV RNA replication in cell culture Science 290, 1972–1974

Table 1 HCV in vitro models: their possible applications and limitations.

Reproduced viral steps Area of research Limitations

intracellular host defences evasion mechanisms antiviral screening

in vitro adaptative mutations restricted to Huh7 cell lines artificially high level of replication Serum-derived HCV Entire life cycle entry process

replication mechanisms intracellular host defences antiviral screening

few cell lines support infection technical difficulties of primary cell culture

vaccination morphogenesis

no secretion of particles independence of CD81 for entry

exogenous core

no association of particles with lipoproteins

replication mechanisms intracellular host defence evasion mechanisms virus production antiviral screening

restricted to JFH1 NS sequence restricted to Huh7 cell lines

4 Guo JT, Bichko VV & Seeger C (2001) Effect of alpha

interferon on the hepatitis C virus replicon J Virol 75,

8516–8523

5 Krieger N, Lohmann V & Bartenschlager R (2001)

Enhancement of hepatitis C virus RNA replication

by cell culture-adaptive mutations J Virol 75, 4614–

4624

6 Lohmann V, Korner F, Dobierzewska A &

Bartensch-lager R (2001) Mutations in hepatitis C virus RNAs

conferring cell culture adaptation J Virol 75, 1437–

1449

7 Bukh J, Pietschmann T, Lohmann V, Krieger N,

Faulk K, Engle RE, Govindarajan S, Shapiro M,

St Claire M & Bartenschlager R (2002) Mutations that

permit efficient replication of hepatitis C virus RNA in

Huh-7 cells prevent productive replication in

chimpan-zees Proc Natl Acad Sci USA 99, 14416–14421

8 Blight KJ, McKeating JA & Rice CM (2002) Highly

permissive cell lines for subgenomic and genomic

hepatitis C virus RNA replication J Virol 76, 13001–

13014

9 Friebe P, Boudet J, Simorre JP & Bartenschlager R

(2005) Kissing-loop interaction in the 3¢ end of the

hepatitis C virus genome essential for RNA replication

J Virol 79, 380–392

10 Zhong J, Gastaminza P, Cheng G, Kapadia S, Kato T,

Burton DR, Wieland SF, Uprichard SL, Wakita T &

Chisari FV (2005) Robust hepatitis C virus infection

in vitro Proc Natl Acad Sci USA 102, 9294–9299

11 Lanford RE, Guerra B, Lee H, Averett DR, Pfeiffer B,

Chavez D, Notvall L & Bigger C (2003) Antiviral

effect and virus–host interactions in response to alpha

interferon, gamma interferon, poly(I)–poly(C), tumor

necrosis factor alpha, and ribavirin in hepatitis C virus

subgenomic replicons J Virol 77, 1092–1104

12 Sumpter R Jr, Loo YM, Foy E, Li K, Yoneyama M,

Fujita T, Lemon SM & Gale M Jr (2005) Regulating

intracellular antiviral defense and permissiveness to

hepatitis C virus RNA replication through a cellular

RNA helicase, RIG-I J Virol 79, 2689–2699

13 Moradpour D, Evans MJ, Gosert R, Yuan Z,

Blum HE, Goff SP, Lindenbach BD & Rice CM

(2004) Insertion of green fluorescent protein into

non-structural protein 5A allows direct visualization of

functional hepatitis C virus replication complexes

J Virol 78, 7400–7409

14 Murray EM, Grobler JA, Markel EJ, Pagnoni MF,

Paonessa G, Simon AJ & Flores OA (2003) Persistent

replication of hepatitis C virus replicons expressing the

beta-lactamase reporter in subpopulations of highly

permissive Huh7 cells J Virol 77, 2928–2935

15 Yi M, Bodola F & Lemon SM (2002) Subgenomic

hepatitis C virus replicons inducing expression of a

secreted enzymatic reporter protein Virology 304,

197–210

16 Blight KJ, McKeating JA, Marcotrigiano J & Rice

CM (2003) Efficient replication of hepatitis C virus genotype 1a RNAs in cell culture J Virol 77, 3181– 3190

17 Kato T, Date T, Miyamoto M, Furusaka A, Tokushige K, Mizokami M & Wakita T (2003) Effi-cient replication of the genotype 2a hepatitis C virus subgenomic replicon Gastroenterology 125, 1808– 1817

18 Ikeda M, Yi M, Li K & Lemon SM (2002) Selectable subgenomic and genome-length dicistronic RNAs derived from an infectious molecular clone of the HCV-N strain of hepatitis C virus replicate efficiently

in cultured Huh7 cells J Virol 76, 2997–3006

19 Pietschmann T, Lohmann V, Kaul A, Krieger N, Rinck G, Rutter G, Strand D & Bartenschlager R (2002) Persistent and transient replication of full-length hepatitis C virus genomes in cell culture J Virol 76, 4008–4021

20 Wakita T, Pietschmann T, Kato T, Date T, Miyamoto M, Zhao Z, Murthy K, Habermann A, Krausslich HG, Mizokami M et al (2005) Production

of infectious hepatitis C virus in tissue culture from a cloned viral genome Nat Med 11, 791–796

21 Bartenschlager R, Frese M & Pietschmann T (2004) Novel insights into hepatitis C virus replication and persistence Adv Virus Res 63, 71–180

22 Lindenbach BD & Rice CM (2005) Unravelling hepati-tis C virus replication from genome to function Nature

436, 933–938

23 De Francesco R & Migliaccio G (2005) Challenges and successes in developing new therapies for hepatitis C Nature 436, 953–960

24 Andre P, Komurian-Pradel F, Deforges S, Perret M, Berland JL, Sodoyer M, Pol S, Brechot C, Paranhos-Baccala G & Lotteau V (2002) Characterization of low- and very-low-density hepatitis C virus RNA-containing particles J Virol 76, 6919–6928

25 Nielsen SU, Bassendine MF, Burt AD, Martin C, Pumeechockchai W & Toms GL (2006) Association between hepatitis C virus and very-low-density lipopro-tein (VLDL)⁄ LDL analyzed in iodixanol density gradi-ents J Virol 80, 2418–2428

26 Prince AM, Huima-Byron T, Parker TS & Levine DM (1996) Visualization of hepatitis C virions and putative defective interfering particles isolated from low-density lipoproteins J Viral Hepat 3, 11–17

27 Thomssen R, Bonk S, Propfe C, Heermann KH, Kochel HG & Uy A (1992) Association of hepatitis C virus in human sera with beta-lipoprotein Med Micro-biol Immunol 181, 293–300

28 Thomssen R, Bonk S & Thiele A (1993) Density heter-ogeneities of hepatitis C virus in human sera due to the binding of beta-lipoproteins and immunoglobulins Med Microbiol Immunol 182, 329–334