HCV - Virology

History

Hepatitis C virus (HCV) is a major cause of progressive liver disease with approximately
170 million people infected worldwide. HCV induces chronic infection
in up to 80% of infected individuals. The main complications of HCV infection are
severe liver fibrosis and cirrhosis, and 30-50% of individuals with cirrhosis develop
hepatocellular carcinoma (Tong 1995; Poynard 1997).
Until 1975, only two hepatitis viruses had been identified, the �ginfectious hepatitis
virus�h (hepatitis A virus, HAV) and the �gserum hepatitis virus�h (hepatitis B virus,
HBV). However, other viruses were excluded from being the cause of approximately
65% of post-transfusion hepatitis. Therefore, these hepatitis cases were
termed �gnon-A, non-B hepatitis�h (NANBH) (Feinstone 1975). Innoculation of
chimpanzees (Pan troglodytes) with blood products derived from humans with
NANB hepatitis led to persistent increases of serum alanine aminotransferase
(ALT) indicating that an infectious agent was the cause of the disease (Alter 1978;
Hollinger 1978).
Subsequently, it was demonstrated that the NANBH agent could be inactivated by
chloroform (Feinstone 1983). Moreover, it was reported that the infectious agent
was able to pass through 80 nm membrane filters (Bradley 1985). Taken together
these findings suggested that the NANBH causing agent would be a small virus
with a lipid envelope. However, the lack of a suitable cell culture system for cultivation
of the NANBH agent and the limited availability of chimpanzees prevented
further characterization of the causative agent of NANBH for several years. In
1989, using a newly developed cloning strategy for nucleic acids derived from
plasma of NANBH infected chimpanzees the genome of the major causative agent
for NANBH was characterized (Choo 1989). cDNA clone 5-1-1 encoded immunological
epitopes that interacted with sera from individuals with NANBH (Choo
1989; Kuo 1989). The corresponding infectious virus causing the majority of
NANBH was subsequently termed hepatitis C virus (HCV).

Taxonomy and genotypes

HCV is a small-enveloped virus with one single-stranded positive-sense RNA
molecule of approximately 9.6 kb. It is a member of the Flaviviridae family. This
viral family contains three genera, flavivirus, pestivirus, and hepacivirus. To date,
only two members of the hepacivirus genus have been identified and classified,
HCV and GB virus B (GBV-B), a virus that had been initially detected together
with the then-unclassified virus GB virus A (GBV-A) in a surgeon with active
hepatitis (Thiel 2005; Ohba 1996; Simons 1995). However, the natural hosts for
GBV-B and GBV-C seem to be monkeys of the Saguinus species (tamarins).
Analyses of viral sequences and phylogenetic comparisons support HCV�fs membership
of a separate genus distinct from the generas flavivirus or pestivirus (Choo
1991
The error-prone RNA-polymerase of HCV together with the high replication rate of
the virus is responsible for the large interpatient genetic diversity of HCV strains.
Moreover, the extent of viral diversification of HCV strains within a single HCVinfected
individual increases significantly over time resulting in the development of
different quasispecies (Bukh 1995).
Comparisons of HCV nucleotide sequences derived from individuals from different
geographical regions revealed the presence of six major HCV genotypes with a
large number of subtypes within each genotype (Simmonds 2004; Simmonds 2005).
The sequence divergence of genotypes and subtypes is 20% and 30%, respectively.
HCV strains belonging to the major genotypes 1, 2, 4, and 5 are found in sub-
Saharan Africa whereas genotypes 3 and 6 are detected with extremely high diversity
in South East Asia. This suggests that these geographical areas could be the
origin of the different HCV genotypes. The emergence of the different HCV genotypes
in North America and Europe and other non-tropical countries appears to represent
more recent epidemics introduced from the countries of the original HCV
endemics (Simmonds 2001; Ndjomou 2003). Besides epidemiological aspects determination
of the HCV genotype plays an important role for the initiation of anti-
HCV treatment since the response of different genotypes varies significantly with
regard to specific antiviral drug regimens, e.g., genotype 1 is most resistant to the
current therapy of the combination of pegylated interferon alfa and ribavirin
(Manns 2001; Fried 2002).

Viral structure

Structural analyses of HCV virions are very limited since the virus is difficult to
cultivate in cell culture systems, a prerequisite for yielding a sufficient number of
virions for electron microscopy. Moreover, serum-derived virus particles are associated
with serum low-density lipoproteins (Thomssen 1992), which makes it difficult
to isolate virions from serum/plasma of infected subjects by centrifugation.
Visualization of HCV virus-like particles via electron microscopy succeeded only
rarely (Kaito 1994; Shimizu 1996a; Prince 1996) and it was a point of controversy
if the detected structures really were HCV virions. Nevertheless, these studies suggest
that HCV has a diameter of 55 to 65 nm confirming size prediction of the
NANBH agent by ultra-filtration (Bradley 1985). Various forms of HCV virions
appear to exist in the blood of infected individuals: virions bound to very low density
lipoproteins (VLDL), virions bound to low-density lipoproteins (LDL), virions
complexed with immunoglobulins, and free circulating virions (Bradley 1991;
Thomssen 1992; Thomssen 1993; Agnello 1999; Andre 2002). The reasons for the
close association of a major portion of circulating virions with LDL and VLDL
remain unexplained. One possible explanation is that HCV theoretically enters
hepatocytes via the LDL receptor (Agnello 1999; Nahmias 2006). Moreover, it is
speculated that the association with LDL and/or VLDL protects the virus against
neutralization by HCV-specific antibodies.
The design and optimization of subgenomic and genomic HCV replicons in the
human hepatoma cell line Huh-7 offered for the first time the possibility to investigate
HCV RNA replication in a standardized manner (Lohmann 1999; Ikeda 2002;
Blight 2002). However, despite the high level of HCV gene expression no infec
tious viral particles are produced using this otherwise powerful tool. Thus, it cannot
be used for structural analysis of free virions.
Only recently have infectious HCV particles been achieved in cell culture by using
recombinant systems (Heller 2005; Lindenbach 2005; Wakita 2005; Zhong
2005; Yu 2007). However, even in these in vitro systems the limited production of
viral particles prevents 3D structural analysis (Yu 2007). Very recently, it was
shown by cryo-electron microscopy (cryoEM) and negative-stain transmission
electron microscopy that HCV virions isolated from cell culture have a spherical
shape with a diameter of approximately 50 to 55 nm (Heller 2005; Wakita 2005; Yu
2007) confirming earlier results that measured the size of putative native HCV particles
from the serum of infected individuals (Prince 1996). The outer surface of the
viral envelope seems to be smooth. Size and morphology are therefore very similar
to other members of the Flaviviridae family such as the dengue virus and the West
Nile virus (Yu 2007). Modifying a baculovirus system (Jeong 2004; Qiao 2004), the
same authors were able to produce large quantities of HCV-like particles (HCV-LP)
in insect cells (Yu 2007). Analysing the HCV-LPs by cryoEM it was demonstrated
that the HCV E1 protein is present in spikes located on the outer surface of the LPs.
Using 3D modeling of the HCV-LPs together with genomic comparison of HCV
and well-characterized flaviviruses it is assumed that 90 copies of a block of two
heterodimers of HCV proteins E1 and E2 form the outer layer of the virions with a
diameter of approximately 50 nm (Yu 2007). This outer layer surrounds the lipid bilayer
which itself contains the viral nucleocapsid consisting of several molecules of
the HCV core (C) protein. An inner spherical structure with a diameter of approximately
30-35 nm has been observed (Wakita 2005) suggesting the nucleocapsid that
harbours the viral genome (Takahashi 1992).

Genome organization
The genome of the hepatitis C virus consists of one 9.6 kb single-stranded RNA
molecule with positive polarity. Similar to other positive-strand RNA viruses, the
genomic RNA of hepatitis C virus serves as messenger RNA (mRNA) for the
translation of viral proteins. The linear molecule contains a single open reading
frame (ORF) coding for a precursor polyprotein of approximately 3000 amino acid
residues (Figure 1). During viral replication the polyprotein is cleaved by viral as
well as host enzymes into three structural proteins (core, E1, E2) and seven nonstructural
proteins (p7, NS2, NS3, NS4A, NS4B, NS5A, NS5B). An additional
protein (termed F [frameshift] or ARF [alternate reading frame]) has been predicted
as a result of ribosomal frameshifting during translation within the core-region of
the genomic RNA (Xu 2001; Walewski 2001; Varaklioti 2002; Branch 2005). Detection
of anti-F-protein antibodies in the serum of HCV-infected subjects indicates
that the protein is expressed during infection in vivo (Walewski 2001; Komurian-
Pradel 2004).
78 HCV - Virology
Figure 1. Genome organization and polyprotein processing.
A) Nucleotide positions correspond to the HCV strain H77 genotype 1a, accession number
NC_004102. nt, nucleotide; NTR, nontranslated region. B) Cleavage sites within the HCV precursor
polyprotein for the cellular signal peptidase the signal peptide peptidase (SPP) and the
viral proteases NS2-NS3 and NS3-NS4A, respectively.
The structural genes encoding the viral core protein and the viral envelope proteins
E1 and E2 are located at the 5�f terminus of the open reading frame followed downstream
by the coding regions for the non-structural proteins p7, NS2, NS3, NS4A,
NS4B, NS5A, and NS5B (Figure 1). The structural proteins are essential components
of the HCV virions, whereas the non-structural proteins are not associated
with virions but are involved in RNA replication and virion morphogenesis.
The ORF is flanked by 5�f and 3�f nontranslated regions (NTR; also termed untranslated
regions . UTR or noncoding regions - NCR) containing nucleotide sequences
relevant for the regulation of viral replication. Both NTRs harbour highly conserved
regions compared to the protein encoding regions of the HCV genome. The high
grade of conservation of the NTRs makes them candidates i) for improved molecular
diagnostics, ii) as targets for antiviral therapeutics, and iii) as targets for anti-
HCV vaccines.
The 5�fNTR is approximately 341 nucleotides long and it has a complex secondary
structure with four distinct domains (I-IV) (Fukushi 1994; Honda 1999). The first
125 nucleotides of the 5�fNTR spanning domains I and II have been shown to be
essential for viral RNA replication (Friebe 2001; Kim 2002). Domains II-IV build
an internal ribosome entry side (IRES) involved in ribosome-binding and subsequent
cap-independent initiation of translation (Tsukiyama-Kohara 1992; Wang
1993).
The 3�fNTR consists of three functionally different regions: a variable region, a poly
U/UC tract of variable length, and the highly conserved X tail at the 3�f-terminus of
the HCV genome (Tanaka 1995; Kolykhalov 1996; Blight 1997). The variable region
of approximately 40 nucleotides is not essential for RNA replication. However,
deletion of this sequence led to significantly decreased replication efficiency
(Yanagi 1999; Friebe 2002). The length of the poly U/UC region varies in different
Genes and proteins 79
HCV strains ranging from 30 to 80 nucleotides (Kolykhalov 1996). The minimal
length of that region for active RNA replication has been reported to be 26 homouridine
nucleotides in cell culture (Friebe 2002). The highly conserved 98-
nucleotide X tail consists of three stem-loops (SL1-SL3) (Tanaka 1996; Ito 1997;
Blight 1997) and deletions or nucleotide substitutions within that region are most
often lethal (Yanagi 1999; Kolykhalov 2000; Friebe 2002; Yi 2003). Another socalled
�gkissing-loop�h interaction of the 3�fX tail SL2 and a complementary portion
of the NS5B encoding region has been described (Friebe 2005). This interaction
induces a tertiary RNA structure of the HCV genome that is essential for HCV replication
in cell culture systems (Friebe 2005; You 2008). Finally, both NTRs appear
to work together in a long-range RNA-RNA interaction possibly resulting in temporary
genome circularization (Song 2006).
Genes and proteins
As described above, translation of the HCV polyprotein is initiated through involvement
of some domains in NTRs of the genomic HCV RNA. The resulting
polyprotein consists of ten proteins that are co-translationally or post-translationally
cleaved from the polyprotein. The N-terminal proteins C, E1, E2, and p7 are processed
by a cellular signal peptidase (SP) (Hijikata 1991). The resulting immature
core protein still contains the E1 signal sequence at its C-terminus. Subsequent
cleavage of this sequence by a signal peptide peptidase (SPP) leads to the mature
core protein (McLauchlan 2002).
The non-structural proteins NS2 to NS5B of the HCV polyprotein are processed by
two virus-encoded proteases (NS2-NS3 and NS3) with the NS2-NS3 cysteine protease
cleaving at the junction of NS2-NS3 (Santolini 1995) and the NS3 serine
protease cleaving the remaining functional proteins (Bartenschlager 1993; Eckart
1993; Grakoui 1993a; Tomei 1993).
The following positions of viral nucleotide and amino acid residues correspond to
the HCV strain H77 genotype 1a, accession number NC_004102. Some parameters
characterizing HCV proteins are summarised in Table 1.
Protein No. of aa aa position
in ref. seq.
MW of protein
Core - immature 191 1-191 23 kd
Core - mature 174 1-174 21 kd
F-protein or ARF-protein 126-161 ~ 16-17 kd
E1 192 192-383 35 kd
E2 363 384-746 70 kd
p7 63 747-809 7 kd
NS2 217 810-1026 21 kd
NS3 631 1027-1657 70 kd
NS4A 54 1658-1711 4 kd
NS4B 261 1712-1972 27 kd
NS5A 448 1973-2420 56 kd
NS5B 591 2421-3011 66 kd
Table 1. Overview of the size of HCV proteins.
aa, amino acid; MW, molecular weight; kd, kilodalton; ref. seq., reference sequence (HCV
strain H77; accession number NC_004102).
80 HCV - Virology
Core. The core-encoding sequence starts at codon AUG at nt position 342 of the
H77 genome, the start codon for translation of the entire HCV polyprotein. During
translation the polyprotein is transferred to the endoplasmic reticulum (ER) where
the core protein (191 aa) is excised by a cellular signal peptidase (SP). The Cterminus
of the resulting core precursor still contains the signal sequence for ER
membrane translocation of the E1 ectodomain (aa 174-191). This protein region is
further processed by the cellular intramembrane signal peptide peptidase (SPP)
leading to removal of the E1 signal peptide sequence (Hussy 1996; McLauchlan
2002; Weihofen 2002).
The multifunctional core protein has a molecular weight of 21 kilodalton (kd). In
vivo, the mature core molecules are believed to form homo-multimers located
mainly at the ER membrane (Matsumoto 1996). They have a structural function
since they form the viral capsid that contains the HCV genome. In addition, the core
protein has regulatory functions including particle assembly, viral RNA binding,
and regulation of RNA translation (Ait-Goughoulte 2006; Santolini 1994). Moreover,
protein expression analyses indicate that the core protein may be involved in
many other cellular reactions such as cell signalling, apoptosis, lipid metabolism,
and carcinogenesis (Tellinghuisen 2002). However, these preliminary findings need
to be analyzed further.
E1 and E2. Downstream of the core-coding region the HCV RNA genome two
envelope glycoproteins are encoded, E1 (gp35; 192 aa) and E2 (gp70; 363 aa).
During translation at the ER both proteins are cleaved from the precursor polyprotein
by a cellular SP. Inside the lumen of the ER both polypeptides experience Nlinked
glycosylation posttranslationally (Duvet 2002). Both glycoproteins E1 and
E2 harbour 5 and 11 putative N-glycosylation sites, respectively.
E1 and E2 are type I transmembrane proteins with a large hydrophilic ectodomain
of approximately 160 and 334 aa and a short transmembrane domain (TMD) of 30
aa. The TMD are responsible for the anchoring of the envelope proteins in the
membrane of the ER and their ER retention (Cocquerel 1998; Duvet 1998; Cocquerel
1999; Cocquerel 2001). Moreover, the same domains have been reported to
contribute to the formation of E1-E2 heterodimers (Op de Beeck 2000). The E1-E2
complex is involved in adsorption of the virus to its putative receptors tetraspanin
CD81 and low-density lipoprotein receptor inducing fusion of the viral envelope
with the host cell plasma membrane (Agnello 1999; Flint 1999; Wunschmann
2000). However, the precise mechanism of host cell entry is still not understood
completely. Several other host factors have been identified to be involved in viral
entry. These candidates include the scavenger receptor B type I (SR-BI) (Scarselli
2002; Kapadia 2007), the C-type lectins L-SIGN and DC-SIGN (Gardner 2003;
Lozach 2003; Pohlmann 2003), and heparan sulfate (Barth 2003).
Two hypervariable regions have been identified within the coding region of E2.
These regions termed hypervariable region 1 (HVR1) and 2 (HVR2) differ by up to
80% in their amino acid sequence (Weiner 1991; Kato 2001). The first 27 aa of the
E2 ectodomain represent HVR1, while the HVR2 is formed by a stretch of seven
amino acids (position 91-97). The high variability of the HVRs reflects exposition
of these domains to HCV-specific antibodies. In fact, E2-HVR1 has been shown to
Genes and proteins 81
be the most important target for neutralizing antibodies (Farci 1996; Shimizu
1996b). However, the combination of the mutation of the viral genome with the
selective pressure of the humoural immune response leads to viral escape via epitope
alterations. This makes the development of vaccines inducing neutralizing antibodies
challenging.
The p7 protein. The small p7 protein (63 aa) is located between the E2 and NS2
regions of the polyprotein precursor. During translation the cellular SP cleaves the
E2-p7 as well as the p7-NS2 junction. The functional p7 is a membrane protein
which is localized in the endoplasmic reticulum where it forms an ion channel
(Haqshenas 2007; Pavlovic 2003; Griffin 2003). The p7 protein is not essential for
RNA replication since replicons lacking the p7 gene replicate efficiently (Lohmann
1999; Blight 2000), however, it has been suggested that p7 plays an essential role
for the formation of infectious virions (Sakai 2003; Haqshenas 2007).
NS2. The non-structural protein 2 (p21; 217 aa) together with the N-terminal portion
of the NS3 protein form the NS2-3 cysteine protease which catalyses cleavage
of the polyprotein precursor between NS2 and NS3 (Grakoui 1993b; Santolini
1995). The N-terminus of the functional NS2 arises from the cleavage of the p7-
NS2 junction by the cellular SP. Moreover, after cleavage from the NS3 the protease
domain of NS2 seems to play an essential role in the early stage of virion morphogenesis
(Jones 2007).
NS3. The non-structural protein 3 (p70; 631 aa) is cleaved at its N-terminus by the
NS2-NS3 protease. The N-terminus (189 aa) of the NS3 protein has a serine protease
activity. However, in order to develop full activity of the protease the NS3 protease
domain requires a portion of NS4A (Faila 1994; Bartenschlager 1995; Lin
1995; Tanji 1995; Tomei 1996). NS3 together with the NS4A cofactor are responsible
for cleavage of the remaining downstream cleavages of the HCV polyprotein
precursor. Since the NS3 protease function is essential for viral infectivity it is a
promising target for design of antiviral treatments.
The C-terminal portion of NS3 (442 aa) has ATPase/helicase activity, i.e., it catalyses
binding and unwinding of the viral RNA genome during viral replication (Jin
1995; Kim 1995). However, recent findings indicate that other non-structural HCV
proteins such as the viral polymerase NS5B may interact functionally with the NS3
helicase (Jennings 2008). These interactions need to be investigated further in order
to better understand the mechanisms of HCV replication.
NS4A. The HCV nonstructural protein 4A (p4) is a 54 amino acid polypeptide that
acts as a cofactor of the NS3 serine protease (Faila 1994; Bartenschlager 1995; Lin
1995; Tanji 1995; Tomei 1996). Moreover, this small protein is involved in the targeting
of NS3 to the endoplasmic reticulum resulting in a significant increase of
NS3 stability (Wolk 2000).
82 HCV - Virology
NS4B. The NS4B (p27) consists of 217 amino acids. It is an integral membrane
protein localized in the endoplasmic reticulum. The N-terminal domain of the
NS4B has an amphipathic character that targets the protein to the ER. This domain
is crucial in HCV replication (Elazar 2004; Gretton 2005) and therefore an interesting
target for the development of anti-HCV therapeutics or vaccines, respectively.
In addition, a nucleotide-binding motif (aa 129-134) has been identified (Einav
2004). Although the function of NS4B is still unknown, it has been demonstrated
that the protein induces a membranous web that may serve as a platform for
HCV RNA replication (Egger 2002).
NS5A. The NS5A protein (p56; 458 aa) is a membrane-associated phosphoprotein
that appears to have multiple functions in viral replication. It is phosphorylated by
different cellular protein kinases indicating an essential but still not understood role
of NS5A in the HCV replication cycle. In addition, NS5A has been found to be
associated with several other cellular proteins (MacDonald 2004) making it difficult
to determine the exact functions of the protein. One important property of NS5A is
that it contains a domain of 40 amino acids the so-called IFN-ƒ¿ sensitivitydetermining
region (ISDR) that plays a significant role in the response to IFN-ƒ¿-
based therapy (Enomoto 1995; Enomoto 1996). An increasing number of mutations
within the ISDR showed positive correlation with sustained virological response to
IFN-ƒ¿-based treatment.
NS5B. The non-structural protein 5B (p66; 591 aa) represents the RNA-dependent
RNA polymerase of HCV (Behrens 1996). The hydrophobic domain (21 aa) at the
C-terminus of NS5B inserts into the membrane of the endoplasmic reticulum,
whereas the active sites of the polymerase are located in the cytoplasm (Schmidt-
Mende 2001).
The cytosolic domains of the viral enzyme form the typical polymerase righthanded
structure with �gpalm�h, �gfingers�h, and �gthumb�h subdomains (Ago 1999;
Bressanelli 1999; Lesburg 1999). In contrast to mammalian DNA and RNA polymerases
the fingers and thumb subdomains are connected resulting in a fully enclosed
active site for nucleotide triphosphate binding. This unique structure makes
the HCV NS5B polymerase an attractive target for the development of antiviral
drugs.
Using the genomic HCV RNA as a template, the NS5B promotes the synthesis of
minus-stranded RNA that then serves as a template for the synthesis of genomic
positive-stranded RNA by the polymerase.
Similar to other RNA-dependent polymerases, NS5B is an error-prone enzyme that
incorporates wrong ribonucleotides at a rate of approximately 10-3 per nucleotide
per generation. Unlike cellular polymerases, the viral NS5B lacks a proof-reading
mechanism leading to the conservation of misincorporated ribonucleotides. These
enzyme properties together with the high rate of viral replication promote pronounced
intra-patient as well as inter-patient HCV evolution.
F-protein, ARFP. In addition to the ten proteins derived from the long HCV ORF,
the F- (frameshift) or ARF- (alternate reading frame) or core+1 protein has been
Viral lifecycle 83
reported (Walewski 2001; Xu 2001; Varaklioti 2002). As the designations indicate
the ARFP is the result of a -2/+1 ribosomal frameshift between codons 8 and 11 of
the core protein-encoding region. The ARFP length varies from 126 to 161 amino
acids depending on the corresponding genotype. In vitro studies have shown that
ARFP is a short-lived protein located in the cytoplasm (Roussel 2003) primarily
associated with the endoplasmic reticulum (Xu 2003). Detection of anti-F-protein
antibodies in the serum of HCV-infected subjects indicates that the protein is expressed
during infection in vivo (Walewski 2001; Komurian-Pradel 2004). However,
the functions of ARFP in the viral life cycle are still unknown and remain to
be elucidated.
Viral lifecycle
Due to the absence of a small animal model system and efficient in vitro HCV replication
systems it has been difficult to investigate the viral life cycle of HCV. The
recent development of such systems has offered the opportunity to analyse in detail
the different steps of viral replication.
Figure 2. Current model of the HCV lifecycle.
Designations of cellular components are in red. For a detailed illustration of viral translation and
RNA replication, see Pawlotsky 2007. Abbreviations: HCV +ssRNA, single stranded genomic
HCV RNA with positive polarity; rough ER, rough endoplasmic reticulum; PM, plasma membrane.
For other abbreviations see text.
84 HCV - Virology
Adsorption and viral entry.
The most likely candidate to be a receptor for HCV is the tetraspanin CD81 (Pileri
1998). CD81 is a ubiquitous 25 kd molecule expressed on the surface of a large
variety of cells including hepatocytes and PBMC. Experimental binding of anti-
CD81 antibodies to CD81 were reported to inhibit HCV entry into Huh-7 cells and
primary human hepatocytes (Hsu 2003; Bartosch 2003a; Cormier 2004; McKeating
2004; Zhang 2004; Lindenbach 2005; Wakita 2005). Moreover, gene silencing of
CD81 using specific siRNA molecules confirmed the relevance of CD81 in viral
entry (Bartosch 2003b; Cormier 2004; Zhang 2004; Akazawa 2007). Finally, expression
of CD81 in cell lines lacking CD81 made them permissive for HCV entry
(Zhang 2004; Lavillette 2005; Akazawa 2007). However, more recent studies have
shown that CD81 alone is not sufficient for HCV viral entry and that co-factors
such as scavenger receptor B type I (SR-BI) are needed (Bartosch 2003b; Hsu
2003; Scarselli 2002, Kapadia 2007). Moreover, it appears that CD81 is involved in
a post-HCV-binding step (Cormier 2004; Koutsoudakis 2006; Bertaud 2006). These
findings together with the identification of other host factors involved in HCV cell
entry were used to generate the current model for the early steps of HCV infection
(Helle 2008).
Adsorption of HCV to its target cell is the first step of viral entry. Binding is possibly
initiated by the interaction of the HCV E2 envelope glycoprotein and the glycosaminglycan
heparan sulfate on the surface of host cells (Germi 2002; Barth 2003;
Basu 2004; Heo 2004). Moreover, it is assumed that HCV initiates hepatocyte infection
via LDL receptor binding (Agnello 1999; Monazahian 1999; Wunschmann
2000; Nahmias 2006; Molina 2007). This process may be mediated by VLDL or
LDL, reported to be associated with HCV virions in human sera (Bradley 1991;
Thomssen 1992; Thomssen 1993). After initial binding the HCV E2 glycoprotein
interacts with the SR-BI in cell culture (Scarselli 2002). SR-BI is a protein expressed
on the surface of the majority of mammalian cells. It acts as a receptor for
LDL as well as HDL (Acton 1994; Acton 1996) emphasizing the role of these compounds
for HCV infectivity. Alternative splicing of the SR-BI transcript leads to the
expression of a second isoform of the receptor SR-BII (Webb 1998), which also
may be involved in HCV entry into target cells (Grove 2007). As is the case for all
steps of viral entry the exact mechanism of the HCVE2/SR-BI interaction remains
unknown. In some studies it has been reported that HCV binding to SR-BI is a prerequisite
for the concomitant or subsequent interaction of the virus with CD81
(Kapadia 2007; Zeisel 2007). The multi-step procedure of HCV cell entry was
shown to be even more complex since a cellular factor termed claudin-1 (CLDN1)
has been newly identified involved in this process (Evans 2007). CLDN1 is an integral
membrane protein that forms a backbone of tight junctions and is highly expressed
in the liver (Furuse 1998). Inhibition assays reveal that CLDN1-
involvement occurs downstream of the HCV-CD81 interaction (Evans 2007). Recent
findings suggest that CLDN1 could also act as a compound enabling cell-tocell
transfer of hepatitis C virus independently of CD81 (Timpe 2007). Very recently,
it was reported that two other members of the claudin family claudin-6 and
claudin-9 may play a role in HCV infection (Zheng 2007; Meertens 2008).
After the complex procedure of binding to the different host factors HCV enters the
cell in a pH-dependent manner indicating that the virus is internalized via clathrinViral
lifecycle 85
mediated endocytosis (Bartosch 2003b; Hsu 2003; Blanchard 2006; Codran 2006).
The acidic environment within the endosomes is assumed to trigger HCV E1-E2
glycoprotein-mediated fusion of the viral envelope with the endosome membrane
(Blanchard 2006; Meertens 2006, Lavillette 2007).
In summary, HCV adsorption and viral entry into the target cell is a very complex
procedure that is not yet fully understood. Despite having identified several host
factors that probably interact with the viral glycoproteins the precise mechanisms of
interaction remain to be investigated more in depth. The fact that some human cell
lines are not susceptible to HCV infection despite expressing SR-BI, CD81, and
CLDN1 indicates that other cellular factors are involved in viral entry (Evans
2007).
Translation and posttranslational processes.
As a result of the fusion of the viral envelope and the endosomic membrane, the
genomic HCV RNA is released into the cytoplasm of the cell. As described above,
the viral genomic RNA possesses a nontranslated region (NTR) at each terminus.
The 5�fNTR consists of four distinct domains I-IV. Domains II-IV build an internal
ribosome entry side (IRES) involved in ribosome-binding and subsequent capindependent
initiation of translation (Fukushi 1994; Honda 1999; Tsukiyama-
Kohara 1992; Wang 1993). The HCV-IRES binds to the 40S ribosomal subunit
complexed with eukaryotic initiation factors 2 and 3 (eIF2 and eIF3), GTP, and the
initiator tRNA resulting in the 48S preinitiation complex (Spahn 2001; Otto 2002;
Sizova 1998; reviewed in Hellen 1999).
Subsequently, the 60S ribosomal subunit associates with that complex leading to
the formation of the translational active complex for HCV polyprotein synthesis at
the endoplasmic reticulum. HCV RNA contains a large ORF encoding a polyprotein
precursor. Posttranslational cleavages lead to 10 functional viral proteins Core,
E1, E2, p7, NS2-NS5B. The viral F protein (or ARF protein) originates from a ribosomal
frameshift within the first codons of the core-encoding genome region
(Walewski 2001; Xu 2001; Varaklioti 2002). Besides several other cellular factors
that have been reported to be involved in HCV RNA translation, various viral proteins
and genome regions have been shown to enhance or to inhibit viral protein
synthesis (Zhang 2002; Kato 2002; Wang 2005; Kou 2006; Bradrick 2006; Song
2006).
The precursor polyprotein is processed by at least four distinct peptidases. The cellular
signal peptidase (SP) cleaves the N-terminal viral proteins immature core
protein, E1, E2, and p7 (Hijikata 1991), while the cellular signal peptide peptidase
(SPP) is responsible for the cleavage of the E1 signal sequence from the C-terminus
of the immature core protein, resulting in the mature form of the core (McLauchlan
2002). The E1 and E2 proteins remain within the lumen of the ER where they are
subsequently N-glycosylated with E1 having 5 and E2 harbouring 11 putative Nglycosylation
sites (Duvet 2002).
In addition to the two cellular peptidases HCV encodes two viral enzymes responsible
for cleavage of the non-structural proteins NS2 to NS5B within the HCV
polyprotein precursor. The zinc-dependent NS2-NS3 cysteine protease consisting of
the NS2 protein and the N-terminal portion of NS3 autocatalytically cleaves the
junction between NS2 and NS3 (Santolini 1995), whereas the NS3 serine protease
86 HCV - Virology
cleaves the remaining functional proteins (Bartenschlager 1993; Eckart 1993;
Grakoui 1993a; Tomei 1993). However, for its peptidase activity NS3 needs NS4A
as a cofactor (Failla 1994; Tanji 1995; Bartenschlager 1995; Lin 1995; Tomei
1996).
HCV RNA replication.
The complex process of HCV RNA replication is poorly understood. The key enzyme
for viral RNA replication is NS5B an RNA-dependent RNA polymerase
(RdRp) of HCV (Behrens 1996). In addition, several cellular as well as viral factors
have been reported to be part of the HCV RNA replication complex. One important
viral factor for the formation of the replication complex appears to be NS4B which
is able to induce an ER-derived membranous web containing most of the nonstructural
HCV proteins including NS5B (Egger 2002). This web could serve as the
platform for the next steps of viral RNA replication. The RdRp uses the previously
released genomic positive-stranded HCV RNA as a template for the synthesis of an
intermediate minus-stranded RNA. In order to facilitate synthesis of minus-strand
RNA the NS3 helicase is assumed to unwind putative secondary structures of the
template RNA (Jin 1995; Kim 1995). In turn, again with the assistance of the NS3
helicase the newly synthesized antisense RNA molecule serves as the template for
the synthesis of numerous positive-stranded RNA. The resulting sense RNA could
subsequently be used as genomic RNA for HCV progeny as well as for polyprotein
translation.
Assembly and release.
After the viral proteins, glycoproteins, and the genomic HCV RNA have been synthesized
these single components have to be arranged in order to produce infectious
virions. As is the case for all other steps in the HCV lifecycle viral assembly is a
multi-step procedure involving most viral components along with many cellular
factors. Investigation of viral assembly and particle release is still in its infancy
since the development of in vitro models for the production and release of infectious
HCV occurred only recently. Previously, it was reported that core protein
molecules were able to self-assemble in vitro, yielding nucleocapsid-like particles.
Very recent findings suggest that viral assembly takes place within the endoplasmic
reticulum (Gastaminza 2008) and that lipid droplets (LD) are involved in particle
formation (Moradpour 1996; Barba 1997; Miyanari 2007; Shavinskaya 2007; Appel
2008). It appears that LD-associated core protein targets viral non-structural proteins
and the HCV RNA replication complex including positive and negative
stranded RNA from the endoplasmic reticulum to the LD (Miyanari 2007). Beside
the core protein, LD-associated NS5A seems to play a key role in the formation of
infectious viral particles (Appel 2008). Moreover, E2 molecules are detected in
close proximity to LD-associated membranes. Finally, spherical virus-like particles
associated with membranes can be seen very close to the LD. Using specific antibodies
the virus-like particles were shown to contain core protein as well as E2 glycoprotein
molecules indicating that these structures may represent infectious HCV
(Miyanari 2007). However, the precise mechanisms for the formation and release of
infectious HCV particles are still unknown.
Model systems for HCV research 87
Model systems for HCV research
For a long time HCV research was restricted due to a lack of small animal models
and efficient cell culture systems. The development of the first HCV replicon system
10 years after the identification of the hepatitis C virus offered the opportunity
to investigate the molecular biology of HCV infection in a standardized manner
(Lohmann 1999). Using total RNA derived from the explanted liver of an individual
chronically infected with HCV genotype 1b, the authors amplified and cloned
the entire HCV ORF sequence in two overlapping fragments. The flanking NTRs
were amplified and cloned separately and all fragments were assembled to a modified
full-length sequence. Transfection experiments with in vitro transcripts derived
from the full-length clones failed to yield viral replication. For this reason, the
authors generated two different subgenomic replicons consisting of the 5�f IRES, the
neomycin phosphotransferase gene causing resistance to the antibiotic neomycin,
the IRES derived from the encephalomyocarditis virus (EMCV) and the NS2-3�f
NTR or NS3-3�f NTR sequence, respectively (Figure 3).
Figure 3. Structure of subgenomic HCV replicons (Lohmann 1999).
This figure illustrates the genetic information of in vitro transcripts used for Huh-7 transfection.
A) Full-length transcript derived from the explanted liver of a chronically infected subject. B)
Subgenomic replicon lacking the structural genes and the sequence encoding p7. C) Subgenomic
replicon lacking C, E1, E2, p7, and NS2 genes. neo, neomycin phosphotransferase
gene; E-I, IRES of the encephalomyocarditis virus (EMCV).
In vitro transcripts derived from these constructs lacking the genome region coding
for the structural HCV proteins were used to transfect the hepatoma cell line Huh-7
(Lohmann 1999). The transcripts are bicistronic, i.e., the first cistron containing the
HCV IRES enables the translation of the neomycin phosphotransferase as a tool for
efficient selection of successfully transfected cells and the second cistron containing
the EMCV IRES directs translation of the HCV-specific proteins. Only some Huh-7
clones can replicate replicon-specific RNA, however, in titres of approximately 108
88 HCV - Virology
positive-stranded RNA copies per microgram total RNA. Moreover, all encoded
HCV proteins are detected predominantly in the cytoplasm of the transfected Huh-7
cells. The development of this replicon is a milestone in HCV research with regard
to the investigation of HCV RNA replication and HCV protein analyses.
In more recent years, the methodology has been improved in order to achieve significantly
higher replication efficiency. Enhancement of HCV RNA replication was
achieved by the use of replicons harbouring cell culture-adapted point mutations or
deletions within the NS genes (Blight 2000; Lohmann 2001; Krieger 2001). Further
development has led to the generation of selectable full-length HCV replicons, i.e.,
genomic replicons that also contain genetic information for the structural proteins
Core, E1, and E2 (Pietschmann 2002; Blight 2002). This improvement offered the
opportunity of investigating the influence of the structural proteins on HCV replication.
Thus it has been possible to analyse the intracellular localisation of these
proteins. However, using this methodology viral assembly and release has not been
achieved.
Another important milestone was achieved when a subgenomic replicon based on
the HCV genotype 2a strain JFH-1 was generated (Kato 2003). This viral strain
derived from a Japanese subject with fulminant hepatitis C (Kato 2001). The corresponding
replicons showed higher RNA replication efficiency than previous replicons.
Moreover, cell lines distinct from Huh-7, such as HepG2 or HeLa were transfected
efficiently with transcripts derived from the JFH-1 replicon (Date 2004; Kato
2005).
Transfection of Huh-7 and �gcured�h Huh-7.5 cells with full-length JFH-1 replicons
led for the first time to the production of infectious HCV virions (Zhong 2005;
Wakita 2005). The construction of a chimera with the core-NS2 region derived
from HCV strain J6 (genotype 2a) and the remaining sequence derived from JFH-1
improved infectivity. In addition, an alternative strategy for the production of infectious
HCV particles was developed (Heller 2005). A full-length HCV construct
(genotype 1b) was placed between two ribozymes in a plasmid containing a tetracycline-
responsive promoter. Huh-7 cells were transfected with those plasmids, resulting
in efficient viral replication with HCV RNA titres of up to 107 copies/ml cell
culture supernatant.
The development of cell culture systems that allow the production of infectious
HCV represents a breakthrough for HCV research since it is now possible to investigate
the whole viral life-cycle from virus adsorption to virion release. These studies
will help to better understand the mechanisms of HCV pathogenesis and they
should significantly accelerate the development of HCV-specific antiviral compounds.