Introduction
Hepatitis C Virus (HCV) is the causative agent of non-A-non-B hepatitis (
Simmonds et al., 2005). Infection is acquired from contaminated blood, mainly through transfusion, intravenous drug use, or hemodialysis. Approximately 80% of the acute infections become chronic and lead to liver cirrhosis and hepatocellular carcinoma (
EI-Serag, 2004). Recurring infections in one individual can take place, likely due to re-infection by variants of the virus arising from error-prone replication (
Micallef et al., 2006). By the year 2000, HCV had infected over 170 million people worldwide (
Wasley and Alter, 2000). To increase this public health challenge further, HCV exists as six major genotypes that differ in geographic distribution, pathology, and ability to respond to treatments (
Pang et al., 2009).
A molecular clone of HCV was first reported in 1989 by Houghton and coworkers (
Choo et al., 1989). Understanding and inhibiting HCV replication and infection have been a goal ever since. Significant progress has been made in both areas. The current standard of care consists of a combination of pegylated interferon alpha together with the purine analog ribavirin (
Shimakami et al., 2009). Interferon is a cytokine that activates many antiviral cellular responses including the JAK-STAT pathway involved in the activation of Interferon-stimulated genes (ISGs). Ribavirin is thought to indirectly target HCV replication by interacting with cellular proteins, including inosine monophosphate dehydrogenase, a key enzyme in guanine synthesis (
Theofilopoulos et al., 2005). This combination treatment has significant side effects and is only effective in approximately half of individuals infected with genotype-1 virus, a strain that is the predominant genotype in North America, Europe, and Japan (
Theofilopoulos et al., 2005). Nonetheless, ribavirin can reduce resistance in HCV (
Mo et al., 2011). In 2011, two direct acting agent (DAA) inhibitors of the genotype 1 HCV-encoded protease were approved by the FDA: Incivek (Telaprevir) and Victrelis (Boceprevir). When used in combination with interferon and ribavirin, the protease inhibitors have between a 60 and 70% cure rate (
Bacon et al., 2011;
Zeuzem et al., 2011). There is thus significant optimism in controlling HCV with DAAs. However, resistance mutations to ribavirin, the protease inhibitors, and several compounds in clinical development targeting HCV have already been found in patients (
Sarrazin et al., 2007;
Chinnaswamy et al., 2010a). Therefore, additional inhibitors are needed in combination to, or in replacement of the current treatment regime.
The HCV-encoded NS5B protein is the RNA-dependent RNA polymerase (RdRp), a validated drug target (
Chinnaswamy et al., 2010a). Intense efforts are currently focused on developing novel active site nucleoside/nucleotide inhibitors (NIs) and allosteric nonnucleoside inhibitors (NNIs) of the HCV polymerase. Given that the active site of the HCV polymerase is highly conserved, the use of NIs should prevent the rapid emergence of resistance mutations and may be effective in treatment of infections by all HCV genotypes (
Heck et al., 2008;
Burton and Everson, 2009).
Over the past decade, the drop out rate for HCV polymerase inhibitors is quite high, largely due to pharmacological issues. Therefore focused efforts are needed to develop a stronger pipeline of polymerase inhibitors. Effective assays are a necessity fordiscoveries of new inhibitors and to better understand HCV replication. Information on current assays developed for primarily genotype 1 HCV should provide guidance for assays of other HCV genotypes. The goals of this review are to first provide a context for HCV RNA synthesis and then to summarize the currently available assays to assess HCV RNA synthesis.
HCV genome organization
HCV is an enveloped virus with a positive-strand RNA genome from the
Flaviviridae family (
Moradpour et al., 2007). The HCV genome is ~9.6 kb in length and encodes a polyprotein that is cleaved by host and HCV-encoded proteases into ten structural and non-structural functional proteins (Fig. 1). The 5′ untranslated RNA sequence (UTR) contains an internal ribosome entry site that recruits host ribosomes to translate the viral genome. The 5′UTR, 3′UTR and the protein-coding region have important cis-acting replication elements that regulate HCV RNA synthesis (
Diviney et al., 2008).
After entry into the cell and uncoating, the HCV genome functions in three main roles: translation, replication and packaging into nascent virions (Fig. 2). Replication of the HCV genome takes place in modified membrane vesicles and minimally requires the HCV-encoded NS3 (protease-helicase), NS4A (NS3-coactivator), NS4B (that can reconfigure the endoplasmic reticulum membranes), NS5A (an RNA binding phosphoprotein) and NS5B (the RNA-dependent RNA polymerase) (
Egger et al., 2002;
Huang et al., 2007;
Moradpour et al., 2007). Relatively little information exists on the complex of these proteins that together constitute the HCV replicase, in large part because the complex is intimately associated with the cellular membranes and the viral replicase is present in low amounts in cells. Furthermore, it should be emphasized that the majority of the replication-associated proteins have roles that are independent of RNA synthesis, such as interference with the host innate immune responses, perturbation of normal cell cycle control, and assembly of the virions with the replicated genome (
Gale and Foy, 2005;
Munakata et al., 2007;
McLauchlan, 2009). Extensive cross-talk between the HCV RNA replication proteins and other viral and cellular processes remains the rule, rather than the exception. In addition, a large and growing list of host factors that range from phospholipids to protein kinases to micro RNA can either enhance or inhibit viral replication (
Jopling, 2005;
Shimakami et al., 2006;
Moriishi and Matsuura, 2007;
Chatterji et al., 2009;
Hsu et al., 2010;
Weng et al., 2010;
Ariumi et al., 2011).
Activities of the HCV polymerase
The key enzyme responsible for the synthesis of viral RNA and the replication of the HCV genome is NS5B, the RNA-dependent RNA polymerase. Important insights into the basic requirements for HCV RNA synthesis and the sites for inhibitor binding have been gained from studies of the recombinant HCV RNA polymerase.
As with other polymerases, the analogy of a human right hand (with the palm, thumb and finger subdomains) proposed by Tom Steitz is useful to consider polymerase architecture and function (
Steitz, 1999) (Fig. 3). The active site of NS5B is located in the palm subdomain, and it has the characteristic divalent metal binding motif that contains two consecutive aspartates (
van Dijk et al., 2004). The thumb and fingers domains are used to help regulate nucleic acid binding (Fig. 3). A fourth subdomain, a hydrophobic C-terminal tail of 21 residues that anchors the HCV polymerase to membranes, can be added to the right hand structure. These residues are dispensable for enzymatic activity in vitro, but are absolutely required for HCV replication in cells (
Tomei et al., 2000;
Brass et al., 2010).
The first characterization of RNA synthesis by the HCV RdRp reported a primer-dependent activity wherein the non-HCV template formed a partial hairpin at the 3′ end to allow the polymerase to extend by a copy-back mechanism (
Behrens et al., 1996). Lohmann et al (
1997) demonstrated that the recombinant RdRp can copy the full-length HCV plus-strand genomic RNA by the same mechanism. Although these works established the polymerase activity of HCV RdRp, a copy-back mechanism cannot be relevant to HCV RNA replication since it would result in a deletion or other covalent modification of the genome. De novo initiation was soon demonstrated for a recombinant HCV NS5B from both short synthetic RNA templates as well as from full-length and truncated HCV genomic RNA (
Oh et al., 1999;
Kao et al., 2000;
Luo et al., 2000). A stable secondary structure along with several unpaired nucleotides at the 3′ end of the template RNA are needed to promote efficient de novo RNA synthesis. De novo initiation by the polymerase takes place at the 3′-end of the RNA template and requires an initiating purine triphosphate (GTP) together with a second NTP complementary to the template residue (
Kao et al., 2000;
Sun et al., 2000;
Zhong et al., 2000;
Ranjith-Kumar et al., 2002). High levels of GTP but not other NTPs can stimulate
de novo initiation (
Lohmann et al., 1999). The initiating GTP specifically stimulates reconfiguration of the polymerase motifs in the active site, irrespective of its incorporation into the nascent RNA, and thus helps to facilitate the transition of the polymerase from initiation to elongative RNA synthesis (
Harrus et al., 2010).
The product of de novo initiation contains a non-modified 5′ triphosphate group on the first nucleotide of the RNA product since HCV lacks an RNA capping enzyme. This is important since a triphosphated RNAs are recognized by the innate immune receptor RIG-I to activate the interferon-associated responses (
Pichlmair et al., 2006). Thus, the RNA synthesis by the HCV polymerase is tied to intracellular responses to clear the virus, alert neighboring cells of the site of viral infection, and/or activate apoptosis (
Saito et al., 2008;
Lu et al., 2010). However, HCV has developed strategies to overcome recognition by host innate immune receptors, an example of which is the cleavage of the adapter protein IPS-1 (a.k.a. MAVS, Cardif, VISA) by the NS3 protease (
Gale and Sen, 2009).
Significant conformational changes are required at each stage of RNA synthesis by any polymerase, including the HCV NS5B. An excellent starting point to think about conformational changes and activities of HCV polymerase came with the elucidation of the crystal structure for the apo-protein (
Lesburg et al., 1999). Over 100 crystal structures of the HCV polymerase have been deposited into the protein data bank. Notably, the HCV polymerase differs from DNA-dependent RNA polymerases in that the thumb and fingers subdomains interact extensively via finger tip motifs to form an enclosed active site and also a template channel, the dimensions of which can accommodate a single-stranded, but not double-stranded RNA (
O’Farrell et al., 2003;
Biswal et al., 2004;
Di Marco et al., 2005;). Thus, a large conformational change within NS5B must take place during RNA synthesis when the template channel contains a partially duplexed template and nascent RNA. The Δ1 loop that extends from the fingers to the thumb subdomains is a key feature to regulate the transition from the closed to the more open polymerase conformation (Fig. 3A). Amino acids in the Δ1 loop are highly conserved in the different genotypes of HCV and make important hydrophobic contacts with the thumb domain that will define the contours of the template channel (
Chinnaswamy et al., 2008). Deletion at the tip of the loop caused the polymerase to form an open conformation and lose the ability to initiate RNA synthesis by a de novo mechanism without decreasing the ability to extend from a primed template (
Chinnaswamy et al., 2010b). Notably, the conformation of the HCV polymerase is also related to an RNA synthesis-independent activity: the binding to the master cell regulator protein, Retinoblastoma (Rb). This later activity can alter the cell’s progression in the cell cycle (
Munakata et al., 2007).
The structure of the HCV polymerase also provides valuable information for the organization of the polymerase inhibitors (Fig. 3B). As is obvious, NIs bind to the active site of the polymerase, but allosteric NNIs have now been mapped to four pockets in the 1b polymerase, two each in the thumb and palm subdomains (
Pauwels et al., 2007). Notably, the Thumb 1 site also contacts the Δ1 loop that extends from the fingers and drugs binding in this pocket will inhibit de novo initiation, but not extension from a primed template (
Tomei et al., 2003). Thus, the effects of the inhibitors acting at these sites as well as the resistance mutations arising after HCV replicons have been exposed to the inhibitors have revealed requirements for RNA synthesis by the HCV polymerase (
Chinnaswamy et al., 2010c).
There is evidence that the initiation of RNA synthesis by the HCV polymerase likely requires an oligomerized form of the protein (
Chinnaswamy et al., 2010c). This explains why de novo-initiated RNA synthesis requires a critical concentration of protein in order to form multimers while lower concentrations of the polymerase are sufficient for extension from a primer. The oligomerization model posits that the closed conformation of NS5B is maintained by homomeric interactions between NS5B subunits, likely by stabilizing the Δ1 loop and thumb domain interactions. The oligomeric interactions between the NS5B monomers are thought to involve more than one interface (
Wang et al., 2002). With regard to the RNA synthesis, a restricted template channel is still at work in a model where a dimer or multimer mediates initiation and this channel is reconfigured during elongative RNA synthesis (
Chinnaswamy et al., 2010c). Interestingly, genotypic differences have been reported in oligomeric properties of NS5B and may have clinical implications (
Clemente-Casares et al., 2011).
Assays for HCV polymerase activity
In the context of this review, the term polymerase activity is broadly defined and spans from RNA-dependent RNA synthesis by NS5B in vitro to the replication of the HCV genome. Numerous assays from the biophysical to the organismal have been developed to report on a range of activities associated with HCV RNA replication. These assays all have advantages as well as limitations. This section will be divided into three classes of assays: biochemical, cell-based, and those involving tissues/organisms, with comments on advantages and limitations.
i. Biochemical assays
Biochemical assays for RNA synthesis by NS5B have facilitated antiviral drug development and the complementary efforts to better understand the mechanism of HCV RNA synthesis. NS5B lacking the C-terminal 21 residues (referred to as Δ21) is widely used in biochemical, structural and inhibitor studies because it can be purified to homogeneity and does not require detergents in the buffer (
Tomei et al., 2000;
Ranjith-Kumar et al., 2002). Furthermore, no significant differences between the enzyme kinetics of the full-length versus the Δ21 enzymes are observed (
Tomei et al., 2000;
Wang et al., 2004). Another version that lacks the C-terminal 55 residues (referred to as Δ55) was initially used for biochemical and structural characterizations (
Bressanelli et al., 1999). However, the Δ55 enzyme lacks a portion of the sequence that lines the active site, and its use has decreased over the years.
The polymerase reaction carried by the recombinant NS5B protein is easily compatible with high-throughput screening of small molecule inhibitors in micro-plate format using automated workstations. In most cases, the RNA used to direct RNA synthesis is an annealed homopolymeric primer/template. For instance, a 5′-biotinylated oligo(G) primer is annealed to a poly(rC) template such that the enzyme incorporates radiolabeled [
3H]GTP as substrate in the reaction. The radiolabeled product can then be traced by the scintillation proximity assay (SPA) using streptavidin-coated beads. A simple alternative to the SPA detection method is to precipitate and filter the high molecular weight RNA products to separate them from the unincorporated nucleotide substrate. These two methods have been successfully employed to identify and characterize several allosteric inhibitors of NS5B (
Love et al., 2003;
Tomei et al., 2003;
Di Marco et al., 2005;
Shi et al., 2008;
Nyanguile et al., 2010). It is now well established that NNIs of NS5B do not compete with the RNA or the nucleotide substrate of the enzyme, but bind to one of the four allosteric pockets in either the thumb or the palm sub-domains (Fig. 3B). Binding of the non-competitive inhibitors to the enzyme can lock the enzyme in an inactive or partially active conformation (
Meanwell et al., 2009;
Chinnaswamy et al., 2010a;
Tomei et al., 2003). When combined with biophysical or crystallographic methods, these high-throughput HCV polymerase assays help to rapidly identify small molecule leads. For a review on non-nucleosides inhibitors, see Meanwell et al (
2009). The combination of an enzymatic approach with biophysics and crystallography has also led to successful fragment-based identification of new chemotypes that specifically inhibit NS5B. In one particular case, a weak NS5B binder was identified using surface plasmon resonance spectroscopy and then was optimized to yield an inhibitor with sub-micromolar affinity and potent enzyme inhibition properties (
Antonysamy et al., 2008). Another advantage of using high throughput polymerase assays with homopolymeric RNA substrates is the requirement for very low enzyme concentrations, typically around or below 20 nM. This parameter is important to accurately measure both the potency of inhibitors with high binding affinity or tight binding inhibitors (
Hang et al., 2009). However, the main disadvantage of homopolymeric RNA based assays is that they are not designed to characterize nucleotide analogs. Because nucleotide analogs interact with NS5B by specifically base-pairing with the template RNA, proper enzymatic assays require hetero-polymeric RNA template sequences derived from the 3′- or 5′-UTR of the HCV genome (
Carroll et al., 2003). Although these assays are typically less sensitive and require significantly higher enzyme concentrations, they provide convenient tools to measure steady-state kinetic constants such as K
m(NTP), as well as IC
50 and K
i(NTP analogs). A more comprehensive characterization of nucleotide analogs can be obtained by measuring their capacity to be incorporated into the nascent RNA and chain terminate RNA synthesis. This is usually achieved by using short synthetic RNA templates (50-mer or less) containing single sites for nucleotide incorporation, and the short RNA products are resolved by sequencing gel analysis (
Kao et al., 2000;
Carroll et al., 2003;
Olsen et al., 2004;
Dutartre et al., 2006;
Klumpp et al., 2006;
Deval et al., 2007;
Murakami et al., 2008).
Finally, it should be noted that HCV polymerase activity can also be measured from replicases extracted from the solubilized membranes of infected or replicon-transfected cells (
Tomassini et al., 2003). This biochemical assay contains all protein constituents of the authentic replicase complex, including full length NS5B bound to NS5A, NS3, and NS4B (
Hardy et al., 2003). However, this system is rarely used because of the low levels of measurable polymerase activity and the fact that NS5B is already in processive elongation mode and cannot be inhibited by non-nucleoside allosteric inhibitors (
Ma et al., 2005).
ii. Cell-based assay for HCV RNA synthesis
HCV subgenomic replicon
The development of the HCV subgenomic replicon represents a significant breakthrough in the understanding of the requirements for HCV replication as well as providing assays to screen and characterize inhibitors of HCV replication (
Horscroft et al., 2005). Inspired by the self-replicating replicons for flaviruses (
Khromykh and Westaway, 1997), Lohmann et al (
1999) developed the first generation of HCV replicons that contain the 5′ and 3′ untranslated sequences (UTRs) from the consensus genotype 1b Con1 strain of HCV along with nonstructural proteins from NS3 to NS5B. The initial replicon is bi-cistronic: the HCV 5′ UTR directs the expression of the first 12 codons of the capsid protein fused in-frame with neomycin phosphotransferase (Neo), which serves as a selectable marker. The HCV non-structural proteins are translated from an encephalomyocarditis virus (EMCV) internal ribosome entry RNA element (Fig. 4A). When transfected into a permissive host cell line (Huh-7), HCV subgenomic RNAs in neomycin-resistant colonies could replicate up to ~5,000 copies of positive-sense HCV RNA per cell and maintain autonomous replication for over one year with no obvious signs of cytotoxicity (
Pietschmann et al., 2001).
Blight et al. (
2003) developed a genotype 1a HCV replicon and found that replication enhancing adaptive mutations in the HCV non-structural genes quickly accumulated after transfection. The adaptive mutations are required for efficient replication into the Huh-7 cell line (
Pietschmann et al., 2001). A similar phenomenon was also observed by other group (
Krieger et al., 2001). Most adaptive mutations existed within the NS4B, NS5A, and NS5B coding regions. The majority of them are in a region of NS5A previously shown to decrease the effectiveness of the IFN treatment highly adaptive amino acid substitutions have been identified at more than nine positions in Con1 NS5A. Mutation S2204I in NS5A increased HCV replicon RNAs over 10,000 fold compared to the parental replicon. These adaptive mutations likely change the interactions between viral and cellular components to facilitate more efficient virus replication in the Huh-7 cell line.
Modifications to the HCV replicons have been made to facilitate high-throughput compound screening (HTS). These efforts include replacement of the
neo gene with luminescent reporters that can be easily and sensitively detected as an indication of RNA replication in transiently transfected cells (
Krieger et al., 2001;
Lohmann et al., 2001) (Fig. 4B). The reporter and selectable markers have also been engineered to require ubiquitin-mediated proteolytic cleavage (
Vrolijk et al., 2003), and the incorporation of the HIV Tat gene to activate the expression of secreted enzymes such as alkaline phosphatase, thus allowing detection of the reporters in the cell medium (
Yi, et al., 2002) (Fig. 4C). Monocistronic replicons containing a selectable or easily visualized marker (such as fluorescent proteins) fused directly to the non-structural proteins were also developed (Fig. 4D) (
Frese et al., 2002;
Moradpour et al., 2004). These designs eliminate the need for the use of a second cistron whose expression is usually driven by the internal ribsome entry sequence from the encephalomyocarditis virus (EMCV) and can eliminate factors that perturb the expression of the reporters. Similar approaches were used to generate replicons of other HCV isolates from genotypes 1a, 1b and 2a and intergenotypic chimeric replicons (
Bartenschlager and Sparacio, 2007).
Cell-based assay for HCV NS5B
While the replicons are suitable systems to study HCV replication and screens for replication-specific inhibitors, any identification of a replication inhibitor still requires validation of the target protein(s). This can be done by selecting resistance mutations and then engineering the mutation(s) of interest back into the replicon and confirming the resistance phenotype. This reverse genetics process can be laborious. Therefore, a number of cell-based assays specific for RNA synthesis by the HCV polymerase have been developed.
Lee et al (
2010) constructed BHK-NS5B-FRLuc reporter cell line which carries stably transfected NS5B, and a bicistronic reporter gene that are regulated by RNA polymerase II and functional NS5B polymerase, respectively. The dependence of NS5B to synthesize RNA that can be translated to produce reporters is useful for inhibitor identification. However, the amount of RNA produced by NS5B may be limited. Ranjith-Kumar et al. (
2011) reported an assay, named the 5BR assay, which can amplify the signal in response to RNA synthesis. This assay is based upon NS5B synthesizing RNAs using cellular RNAs as templates and having the RNAs serve as potent agonists for innate immune RNA sensors (RIG-I or MDA5). Activation of the sensors will lead to luciferase production driven from specific promoters, such as the one from the IFNβ gene (Fig. 4E). The advantages of this assay are that it can work in multiple cell lines as opposed to hepatocytes for subgenomic replicon and JFH1 infection assays and it is functional for NS5Bs from all six major HCV genotypes. Several NNIs identified in HCV replicons could inhibit the readout from the 5BR assay. Since the assay does not employ other non-structural proteins, it could be used to validate whether compounds identified to be effective in the replicon assays, directly targets the HCV polymerase. Resistance mutations in NS5B can also be studied. One of the assay’s limitations is that some truncated RNAs may still activate innate immune receptor RIG-I. Furthermore, the assay uses transiently expressed NS5B and hence calculation of IC
50 for anti-HCV agents will depend on the level of expression of NS5B. However, this assay could be modified by using polymerases expressed from integrated transgenes and should be suitable for high-throughput screening efforts to identify HCV polymerase inhibitors.
Infectious 2a HCV
HCV replicon systems do not allow virion production and thus cannot be used to study the early or late stages of HCV infection such as receptor binding, viral entry, uncoating, assembly, virion maturation, and release. Instead, these processes were studied using HCV pseudoparticles in a recently established replication-independent infectious virion system (
Hsu et al., 2003;
Triyatni et al., 2011). An infectious HCV clone from genotype 2a (JFH1) has become widely used tool to study HCV replication (
Wakita et al., 2005;
Zhong et al., 2005;
Lindenbach et al., 2006). The JFH-1 virus was isolated from a Japanese patient with fulminant hepatitis, can replicate efficiently in Huh-7 cells and generate authentic HCV particles. The production of infectious HCV genotype 1a (Hutchinson strain) in cultured hepatoma cells is also available (
Yi et al., 2006). However, compared to JFH-1, the viral proteins accumulated more slowly with infectious 1a strain. In terms of drug development, an issue with the 2a HCV is of less interest for inhibitor development in some parts of the world. Furthermore, many NNIs have been limited efficacy across HCV genotypes. Thus, while JFH-1 has become a valuable tool in studying HCV replication and infection, it needs additional modifications to assess effects of inhibitor for the replication of other HCV genotypes.
Significant efforts have been made to improve the JFH-1 infectivity titer and expand its use. JFH-1 chimeras that contain the structural region of a genotype 2a isolate [J6 (CF)] fused to cDNAs from other HCV genotypes are now available (
Lindenbach et al., 2005;
Pietschmann et al., 2006). The titers of the most efficient intragenotypic chimera Jc1 is 100-fold higher than parental JFH-1 (
Pietschmann et al., 2006). JFH-1 variants with adaptive mutations are also available (
Yi, et al., 2006). Finally, more permissive cell lines derived from Huh7 cells (e.g. Huh7.5, Huh7-Lunet/CD81 high) with impairment in the innate immune response as well as increased HCV receptor expression have increased the replication of JFH-1 and its derivatives (
Sumpter et al., 2005;
Koutsoudakis et al., 2007;
Kato et al., 2009). Monitoring the levels of JFH-1 RNAs and infectious titers using reporter constructs or quantitative RT-PCR can provide valuable information on HCV replication. Some HCV particles can also infect primary hepatocytes, although the utility of these cells is limited by their availability and their donor-to-donor variability (
Buck, 2008).
Tissues and animal systems to study HCV replication
Replicon and virus-based cell-culture systems have been invaluable tools to study HCV replication and to identify inhibitors. However, interpretation of results from these cell-lines needs to be done judiciously since the genetic make-up of the host can significantly alter virus replication as well as response to treatments (
Ge et al., 2009;
Thomas et al., 2009). For example, rs12979860 is a single nucleotide polymorphism (SNP) near the Interferon 28B gene (encoding Interferon-lambda-3) that is strongly associated with clearance of HCV and the response to treatment with interferon-α and ribavirin (
Thomas et al., 2009). Cell lines with different alleles of this SNP will permit HCV infections to different levels because of the differences arising in the innate immunity pathway as a result of this genetic variation (
Bensadoun et al., 2011). While additional interplay of this SNP and the pathways affected and their relationship to HCV infection and treatment remain to be elucidated, it clearly demonstrates that some effects on HCV replication must be studied in more complex systems.
Animal models have traditionally been used to evaluate the responses to both pathogens and treatments. Rodents, dogs, and monkeys are invaluable in providing understanding of the pharmacological properties of drugs, but these systems cannot be infected by HCV.
Chimpanzee is the only other species beside humans that can be successfully infected by HCV. In fact, the molecular clone of HCV was isolated from this animal after inoculation of a non-A-non-B hepatitis serum from a human patient (
Choo et al., 1989;
Bukh, 2004). Chimpanzees infected through intrahepatic or peritoneal inoculation develop acute hepatitis and viremia at stages similar to that of humans, and a majority of the infected individuals (~60%) develop chronic infections. But, for unknown reasons, the viral titers do not reach high levels (
Bukh, 2004). Similar to the scenario in human infection, neutralizing antibodies that develop in the course of acute infection do not seem to play a role in spontaneous resolution of the virus (
Prince et al., 1999). Chronic infections can lead to hepatocellular carcinoma, but no evidence of fibrosis and cirrhosis is seen in this animal model (
Boonstra et al., 2009). In brief, even though chimpanzees serve well for studying replication and therefore antiviral development against HCV, the somewhat attenuated response indicate that additional features of the disease must be examined in humans (
Bukh, 2004;
Grakoui et al., 2001). In addition, very significant issues are associated with the use of chimpanzees for studies of HCV, including that they are an endangered species, high cost associated with their use and maintenance, and that the use of chimpanzees for research has been banned by the European Union and is been actively considered by government of USA as well.
The chimeric-liver mouse is currently the best small animal model for testing antivirals against HCV. These mice are produced by crossing a mouse with severe combined immunodeficiency (SCID) with one that has been genetically engineered to have an active urokinase-type plasminogen activator (uPA) (
Mercer, et al., 2001). The uPA gene activates apoptotic pathways in the mouse liver while the SCID genetic background allows xenotransplants. The progeny mice can receive human liver tissue that can subsequently be infected with HCV. However, repopulation of the mouse liver with human hepatocytes was variable and require selective depletion of the mouse liver prior to xenotransplantation (
Rhim et al., 1994;
Tateno et al., 2004). However, severe mitochondrial aberrations in the human hepatocytes were observed using this strategy (
Douglas, et al., 2010).
The trimera mouse model was developed by first irradiating normal mice and then transplanting with bone marrow from SCID mice (
Ilan et al., 2002). Human liver tissue can then be transplanted under the ear or under the kidney capsule to allow studies on HCV infection (
Ilan et al., 2002). HCV viremia was observed for about a month, but later declined due to fibrosis and cirrhrosis of the liver transplant. However, only low levels of viremia (<10
5 copies/ml) were achieved in this model even though 85% of the trimeric mice had active virus replication. The low levels may be due to insufficient human hepatocyte proliferation in the adjacent non-hepatic tissues (
Ilan et al., 2002).
The xenograft mouse models discussed above can be used to study antivirals but the use of immunocompromised mice significantly limits modeling of HCV disease pathology. Transgenic mice carrying the entire polyprotein or the core coding regions of HCV were developed to study the immune response and carcinogenic potential of HCV structural proteins but not specifically to study HCV replication (
Wakita, et al., 1998). Recently a humanized mouse model has been created to express the human occludin and CD81 genes in their liver (
Dorner et al., 2011). This system can allow complete HCV infection in immunocompetent mice. While promising, the usefulness of this system in efficacy studies of polymerase inhibitors remains to be better evaluated. Finally, two non-traditional model systems deserve mention. First, a zebra fish model harboring a HCV sub-genomic replicon was established (
Ding et al., 2011). Ribavirin added in the water of the fish inhibited NS5B-mediated replication. Since many of the polymerase inhibitors are water insoluble, alternate routes such as intraperitoneal routes may be needed for drug delivery in this otherwise well characterized model (
Zon and Peterson 2005). Second, the Tree shrew (
Tupaia belangeri) has been found to exhibit low viremia after inoculation with HCV (
Amako et al., 2010). Tree shrews can be easily maintained in laboratories, but a lack of information on their genetics and more complete evaluation and possible manipulation of this novel system will be needed before it can be considered valuable for analysis of HCV replication and infection.
Summary
Since the initial identification of HCV in 1989, much progress has been made in developing assays that have contributed to a better understanding of HCV infection. Virus-specific drugs are becoming available for the genotype 1 HCV. However, we are still at an early stage in meeting the challenges to better understand HCV infection and to developing additional inhibitors for all genotypes of HCV. The assays described in this review, especially the cell-based and biochemical ones, should serve as models for assays to meet the needs imposed by hepatitis C worldwide.
Higher Education Press and Springer-Verlag Berlin Heidelberg