Hepatitis C virus molecular evolution: Transmission, disease progression and antiviral therapy

Abstract

Hepatitis C virus (HCV) infection represents an important public health problem worldwide. Reduction of HCV morbidity and mortality is a current challenge owned to several viral and host factors. Virus molecular evolution plays an important role in HCV transmission, disease progression and therapy outcome. The high degree of genetic heterogeneity characteristic of HCV is a key element for the rapid adaptation of the intrahost viral population to different selection pressures (e.g., host immune responses and antiviral therapy). HCV molecular evolution is shaped by different mechanisms including a high mutation rate, genetic bottlenecks, genetic drift, recombination, temporal variations and compartmentalization. These evolutionary processes constantly rearrange the composition of the HCV intrahost population in a staging manner. Remarkable advances in the understanding of the molecular mechanism controlling HCV replication have facilitated the development of a plethora of direct-acting antiviral agents against HCV. As a result, superior sustained viral responses have been attained. The rapidly evolving field of anti-HCV therapy is expected to broad its landscape even further with newer, more potent antivirals, bringing us one step closer to the interferon-free era.

Core tip: Hepatitis C virus (HCV) infection remains as an important public health problem worldwide. Viral molecular evolution determines, in many ways, the outcome of HCV infection. Here, we present up-to-date information about the role of HCV molecular evolution in virus ransmission, disease progression and antiviral therapy.

INTRODUCTION

Globally, hepatitis C virus (HCV) infection affects approximately 180 million people[1], in addition to three million new infections occurring annually[2,3]. The prevalence of HCV infection varies greatly from region to region, and as a result, the burden of disease exhibits important differences worldwide[2]. HCV is one of the leading causes of chronic liver disease associated with end-stage cirrhosis and hepatocellular carcinoma[4], with approximate 20% of chronically infected patients developing cirrhosis, and about 10% progressing to cancer[5].

HCV is a small single-stranded, positive polarity, enveloped virus belonging to the Hepacivirus genus within the Flaviviridae family. The RNA genome (about 9.6 kb in length) contains a single open reading frame (ORF) encoding for a polyprotein which is flanked by a 5’- and 3’-untranslated regions (UTR). The 5’-UTR contains an internal ribosomal entry site (IRES) which is essential for viral replication. The polyprotein is processed into three structural proteins (C, E1, E2) and seven nonstructural proteins (p7, NS2, NS3, NS4A, NS4B, NS5A, NS5B) (Figure 1)[6,7]. The HCV core is a highly conserved protein that makes up the viral nucleocapsid and plays role in pathogenesis[8]. E1 and E2 are highly glycosylated proteins that participate in cell entry (Figure 2)[9]. E2 contains three hypervariable regions (HVR), known as HVR1-3[10], which are under continuous selection pressure exerted by the host immune system. P7 is a 63-amino acid polypeptide that serves as a signal sequence for the translocation of NS2 into the lumen of the endoplasmic reticulum for further cleavage. P7 is also essential for particle assembly and release of infectious virions[11,12]. NS2 is a transmembrane protein required for viral replication while NS3 is the HCV protease and NTPase/helicase[13,14]. The HCV protease disrupts the interferon (IFN) and toll-like receptor-3 (TLR3) signaling pathways by cleaving host proteins including the caspase recruitment domain of the mitochondrial antiviral signaling protein (MAVS)[15,16], and TIR-domain-containing adapter-inducing interferon-β (TRIF) (Figure 2)[17]. NS4A acts as a cofactor for the NS3 protease and NS4B is a small hydrophobic protein required for recruitment of other viral proteins[18,19]. NS5A is a hydrophilic phosphoprotein needed for viral replication[20,21]. Lastly, NS5B is the HCV RNA dependent RNA polymerase (RdRp)[22], which lacks proofreading and error correction mechanisms, resulting in a highly error prone replication process[23].

Figure 1.

Hepatitis C virus genome and nucleotide variability. A schematic representation of the viral genome is depicted. The degree of nucleotide variability along the viral genome is also shown. The target molecules for anti-HCV therapy are noted, and the antiviral agents are indicated. A selection of FDA approved and in development compounds are shown. Roman numerals in brackets indicate the current clinical phase of development. HCV: Hepatitis C virus; FDA: Food and Drug Administration; PI: Protease inhibitors.

Figure 2.

Hepatitis C virus replication cycle. The replicative cycle of hepatitis C virus (HCV) is displayed. HCV interaction with its cell receptors is shown. Upon entry, the HCV genome is released into the cytoplasm and subsequently translated and translocate into the RE. The membranous web is used as scaffold for viral replication. Interferon and TLR3 signaling pathways are disrupted by the HCV NS3/4A protease by cleaving MAVS and TRIF (upper right window). Assembled virions are released via the constitutive secretory pathway. HCV: Hepatitis C virus; MAVS: Mitochondrial antiviral signaling protein; TRIF: Toll-like receptor-domain-containing adapter-inducing interferon-β.

HCV virions bind to the host cellular receptors via E2[6]. The initial viral attachment is mediated by heparin sulfate proteoglycans on the hepatocyte surface[24]. Multiple cellular receptors such as the scavenger receptor class B type I[25], CD81[26,27], claudin-1[28], and occludin[29], in addition to several entry factors including the receptor tyrosine kinases, the epidermal growth factor receptor[30], the ephrin receptor A2[30] and the Niemann-Pick C1-like 1 cholesterol absorption receptor[31] have been identified (Figure 2). Once bound to the cell, HCV particles are then internalized by pH-dependent and clathrin mediated endocytosis[32,33]. Upon entry, the viral genome is released from the nucleocapsid into the cytoplasm and subsequently translated. The NS4B then induces the formation of membranous webs that serve as scaffolds for viral replication. After genome amplification and protein expression, progeny virions are assembled and released by the constitutive secretory pathway (Figure 2)[7,23].

The HCV mutation rate generates a high degree of intrahost genetic diversity[34], allowing for rapid adaptation[35]. This characteristic molecular plasticity of HCV is a key biological property that enables rearrangement of the intrahost viral population under different selection pressures, such as the immune response and antiviral therapy, warranting viral persistence[36,37]. HCV molecular evolution plays a pivotal role in virus transmission, progression of disease and outcome of therapy. Therefore, understanding the mechanisms driving the molecular evolution of HCV is likely to help to implement measures aimed to control HCV-related disease.

GENETIC HETEROGENEITY AND MOLECULAR EVOLUTION

Viral genotypes

Seven major HCV genotypes and several subtypes have been identified (Figure 3)[38]. HCV exhibits a complex taxonomic structure[38]. Genetic diversity between HCV genotypes is about 30%, while subtypes differ by about 15%[39,40]. HCV genotypes show a characteristic distribution in different geographical regions[41]. Genotypes 1, 2 and 3 exhibit a worldwide distribution. Genotypes 1 and 2 are endemic in West Africa while genotype 3 is endemic to the Indian subcontinent. Genotypes 4 and 5 are primarily found in Africa, and genotype 4 is particularly endemic in Egypt and Central Africa. Genotype 6 is endemic of Asia whereas the distribution of genotype 7 has not been fully assessed[41-45]. Genotype 2 is the oldest lineage, followed by genotypes 3, 5 and 6, while genotypes 1 and 4 emerged more recently. Globally distributed genotypes are referred as epidemic, and their rapid dissemination over the last century is primarily attributed to modes of transmission including the use of intravenous drugs, nosocomial transmission and blood transfusions[46]. Endemic genotypes are usually highly divergent and limited to well defined geographic regions[47]. These characteristic distribution patterns facilitate the identification of the site of origin and tracking of the genetic history of different HCV lineages. In general, high degree of genetic variability among HCV strains evolving in relatively small geographical regions suggests long-term evolution. Conversely, strains exhibiting lower genetic heterogeneity have shorter genetic histories and are often associated with recent introduction and higher transmission rates[47].

Figure 3.

Hepatitis C virus genotypes and subtypes. Representative strains belonging to all seven genotype and all different subtypes are presented.

Molecular evolution

Different molecular mechanisms including mutation, genetic drift, recombination and natural selection shape the molecular evolution of HCV.

Insertion of point mutations by the RdRp is the primary element contributing to the high genetic variability of HCV. The HCV mutation rate in vivo is about 2.5 × 10^-5 per nucleotide per genome replication[48]; however, higher estimates have been obtained[49]. Selection determines if a given mutation would be fixed in the viral population. Extrinsic selective pressures, such as the host immune response and antiviral treatment, and intrinsic functional constraints determine the accumulation of mutations in specific subgenomic regions[50]. These constraints define the tolerance to mutations in different regions resulting in an uneven distribution of genetic variability along the genome (Figure 1)[51,52]. Despite the high degree of genetic variability, the HCV antigenic diversity is significantly convergent[51], implying a large but restricted genetic space. This reduction in antigenic diversity is also likely due to structural and functional constrains[53] that restrict the sequence space and favors homoplasy[51]. The homoplastic nature of HCV antigenicity contrasts with the extensive nucleotide variability and represents an important venue for vaccine development[51].

Amino acid variability plays an important role in HCV infection. For instance, amino acid variability in the NS3/4 protease coding region affects its catalytic efficiency[54], allowing HCV to explore a broad range of protease genetic configurations[55]. Interestingly, minor HCV variants can display improved catalytic activities when compared to the master sequence, including those bearing resistant mutations to antiviral drugs[55]. Thus, these functional differences between variants can affect various aspects of the HCV replication cycle.

Genetic bottlenecks are an important force driving the molecular evolution and transmission of HCV[56,57], by inflicting a strong selective pressure during the acute phase of infection[57]. The intensity of the selective pressure imposed by genetic bottlenecks is such that only a handful of HCV variants manage to establish infection[58,59], commonly resulting in a profound founder effect (Figure 4). Upon transmission, HCV nucleotide diversity is reduced by purifying selection and subsequent bottlenecks[57]. Interestingly, the antibody response does not seem to significantly affect the composition of the transmitted/founder HCV population[58]. Rapid evolution after the occurrence of genetic bottlenecks usually results in the emergence of rare fit variants that quickly dominate the next population (Figure 4)[57,59].

Figure 4.

Hepatitis C virus molecular evolution. The staging of molecular evolution intrahost is illustrated. All four stages of infection are portrayed. The phylogenetic relatedness between different populations throughout time is also represented. The main molecular processed driving the molecular evolution are shown in their respective stage.

Genetic drift is also a contributor to the molecular evolution of HCV[60-62]. Genetic drift refers to the stochastic fluctuations in frequencies of variants in the viral population. Overall, large populations are less stochastic and undergo less genetic drift than small populations. While virus populations are often large, genetic drift still has been shown to be an important factor in viral evolution, suggesting that the genetic processes are mostly deterministic. Viral populations undergoing severe genetic bottleneck events frequently experience founder effects, which significantly reduce population size, allowing a small number of variants to establish infection, favoring genetic drift. In HCV infection, the extent of genetic drift is reduced as the time of infection progresses[60-63]. This might be explained by the large population size frequently observed in chronically infected patients in comparison to acute cases. Additionally, occurrence of genetic drift might be affected by fluctuations in the viral load throughout the infection[64]. Anti-HCV therapy might also prompt genetic drift upon relapsing since viral populations are reduced considerably prior to reestablishment as a result of treatment failure.

Genetic recombination is another mechanism that participates in HCV genetic heterogeneity. Recombination generates genetic variability by rearranging genomic molecules during RNA elongation when the polymerase switches from donor to acceptor molecules, resulting in a nascent RNA with traits from both parental viruses[65]. HCV recombination also occurs via endoribonuclease digestion of parental molecules within base-unpaired regions, followed by ligation[66]. For recombination to occur, co-infection or superinfection in the same cell by two parental viruses is required[67]. HCV genetic recombination is considered a rare event[68], and therefore, while co-infection is possible, it remains relatively infrequent. Superinfection exclusion during HCV infection has been reported[69-71]. Thus, simultaneous infection is feasible but sequential infection is severely impaired[72], suggesting that infected cells become refractory to succeeding infections; and therefore, limiting recombination. Nonetheless, naturally occurring inter-genotype, intra-genotype (inter-subtype) and intra-strain HCV recombinants have been reported[73-76]. HCV recombination has important clinical and epidemiological implications that affect molecular epidemiology, surveillance of emergent new lineages and drug resistance strains[67].

HCV exists as an ensemble of closely related but genetically divergent variants, commonly referred as “quasispecies”[77]. Nevertheless, the existence of viral quasispecies has been extensively debated[77-79]. In HCV molecular epidemiology, the term quasispecies has become a synonymous of intrahost genetic variation[79]. Analysis of the HCV intrahost genetic variation is important for genetic relatedness, drug resistance, molecular evolution and epidemiological studies[80,81]. Rapid[34,82], and slow[83,84] HCV divergence has been reported in different settings. Rapid divergence of HCV represents an important challenge for genetic relatedness studies since molecular epidemiological links can be lost between related cases in relatively short periods of time[34].

HCV molecular evolution involves a series of complex processes characterized by temporal variations that constantly reshape the architecture of the viral population[81,83,85,86]. HCV evolves in an specific staging fashion through incremental changes between communities and random mutations accompanied by fluctuation in the frequency of coexisting viral subpopulations (Figure 4)[85]. In stage 1, HCV infection is established in the new host after occurrence of the initial genetic bottleneck, resulting in a strong founder effect before the onset of the adaptive immune response. Stage 2 is characterized by small incremental evolution steps through different communities, due to selective pressures imposed by the immune response. Stage 3 features a genetic diversification leading to the emergence of new subpopulations owned to the decline of previously dominant populations. In stage 4, HCV reaches settlement under strong negative selection[85]. These temporal variations are likely to affect HCV transmission since different viral variants are available at different time points during infection (Figure 4). The degree of “transmissibility” may exhibit important differences among different viral variants; thus, determining the capacity of each population to successfully establish infections in the new host[87]. This staging is also likely to play an important role in clinical outcome and assist in therapy management since viral populations may be differentially sensitive to treatment at particular stages[88].

Compartmentalization

While the liver is the main site for HCV replication, compartmentalization affecting other organs has been suggested to occur in both, natural infections in humans and experimentally infected chimpanzees[81,89-93]. However, limited information on HCV variability in extrahepatic sites is available. Different HCV populations have been found in a variety of tissues and cell types, suggesting the existence of extrahepatic reservoirs[94-96]. The concept of compartmentalization assumes that the distribution of viral variants in a non-random manner varies between tissues with an specific evolution rate for each compartment[97,98]. Compartmentalization in the same organ has also been observed as nontumor and tumor hepatic cells exhibit different populations, and are likely subjected to different selective pressures[99]. Compartmentalization has been shown in immunocompromised[97,100], immunocompetent[101], and transplanted individuals[102,103], further supporting its role in HCV persistence and pathogenesis.

Compartmentalization has also been proposed to play role in HCV molecular evolution[86]. In this evolutionary model, co-existing lineages represent genetically distinct subpopulations of infected liver cells (Figure 5). However, only one subset of viral subpopulations commonly circulates in serum at any given time. This may be explained by the fluctuation of neutralizing antibodies over time which in turn modify the relative frequency of lineages and differences in replication and shedding rates of virus populations[86].

Figure 5.

Hepatitis C virus compartmentalization. Generation of multiple HCV subpopulations is color coded. Releasing of different subpopulation to circulation and infection of PBMC is also depicted. Antibodies elicited against different HCV subpopulation is also shown throughout time. However, the generation of neutralizing antibodies is delayed and gives opportunity for arisen of new viral subpopulations and subsequent infection of PBMC. In turn, infection in PBMC facilitates vertical transmission of HCV. PBMC: Peripheral blood mononuclear cells; HCV: Hepatitis C virus.

Different mutation patterns in the HCV 5’UTR have been reported to occur in compartmentalized populations. Importantly, minor nucleotide changes occurring in this region can directly affect HCV translation efficiency. Thus, the effect of mutations in the HCV IRES might be cell type dependent[93,104], resulting in different replication rates. Therefore, HCV compartmentalization is not only determined by cell entry but also replication. NS5A and NS5B compartmentalized variants with different functional properties have also been observed[105,106]. These findings highlight the importance of compartmentalization in functionality of specific viral proteins.

Detection of HCV negative RNA strands has been reported in peripheral blood mononuclear cells (PBMC), lymph nodes, bone narrow and brain. Genetic differences between plasma and PBMC HCV variants have been found to persist for long periods of time[101,102,107]. Compartmentalization can be explained by the distribution of tissue specific receptor on different cell types[95]. Compartmentalization in PBMC is facilitated by high levels of CD81 expression[108]. The existence of HCV variants with different CD81 binding capacities has also been suggested[108]. Higher conservation in the CD81-2 HCV binding region has been observed in PBMC-derived variants in comparison to serum-derived strains[108]. Additionally, PBMC- associated HCV variants have also been detected after resolution of HCV infection, suggesting that compartmentalization can also play role in occult HCV infection[108]. In B cells, the interaction between E2 and CD81, as a result of compartmentalization, leads to cell activation, which in turns protects lymphocytes from activation-induced cell death and regulates their function[94,109]. Therefore, E2-CD81 engagement may contribute to HCV associated B cell lymphoproliferative disorders and insufficient neutralizing antibody production[109]. In human T lymphocytes, HCV infection is mediated by the CD5 receptor, a member of the scavenger receptor cysteine-rich family[110]. However, the factors determining lympho- and hepatotropism are not well known. While compartmentalization remains controversial[111-113], the potential implications in persistence, cell tropism, drug resistance, and vaccine development warrant further research[52].

VIRUS TRANSMISSION

HCV transmission primarily occurs via parenteral routes. Epidemic of recreational injection drugs and unsafe injections resulted in a large number of HCV infections during the 20^th century[114,115]. Importantly, differences in transmission rates have resulted in distinctive HCV prevalence and genotype distribution worldwide[116-118].

HCV viral transmission is a dynamic process that has been subjected to significant changes during the last century. Certain risk factors, such as use of illegal intravenous drugs and risk behaviors among homosexual men, significantly facilitate virus transmission and create optimal conditions for rapid HCV molecular evolution[34]. The modes of transmission in a given epidemiological settings affect the intra and interhost genetic variability within the transmission network[34,83,119]. In high risk population groups, higher transmission rates occur among acute cases leading to spread of more infectious variants[87]. However, the patterns of HCV transmission can change over time[120], and this in turn affects HCV transmissibility.

HCV transmission networks are difficult to identify for several reasons. The long incubation period makes it difficult to link related cases to a common source of infection[34]. Additionally, acute HCV infections are usually asymptomatic, making identification of cases challenging[121]. The lack of laboratory methods and appropriate molecular surveillance systems capable of distinguishing acute from chronic infections further complicates identification of incident cases[122]. Recognition of HCV transmission is of critical importance in implementing measures aimed to prevent virus spread. Transmissions of fast evolving viruses, such as HCV, are difficult to trace since strains from epidemiological linked cases are genetically related but rarely identical. Thus, interpretation of viral phylogenies is inherently uncertain and should be used cautiously[123]. Phylogenetic branching not always correspond to directly linked transmission events, and available sequences may not represent all individuals belonging to the transmission network. In consequence, local epidemic sequences can group together in the absence of direct transmission[123], and therefore, phylogenetic linkage among strains cannot show, beyond any doubt, direct transmission between possibly linked cases, despite of inclusion of local unrelated controls[124]. However, local control sequences do help in those instances where direct transmission did not occur by showing sufficient phylogenetic separation between suspected cases[124,125]. Hence, proper epidemiological investigations should be performed in conjunction to adequately assess relatedness between HCV cases. Additionally, in high risk groups re-infection with closely related strains may preclude the use of paraphyletic clustering to relate cases[124].

Estimating the time of infection based purely on genetic data has been reported by using evolutionary molecular clock models[125-127]. However, confidence limits associated with molecular clock estimates can vary significantly, making interpretation difficult. Different factors can affect the accuracy and reliability of estimates of molecular divergence including sampling, temporal and anatomical distribution of sampling, genome region sequenced, super- or re-infection, and evolutionary models and algorithms used[124]. Thus, interpretation and reliability of such approaches requires further validation.

HCV evolution is also affected by other instances such as co-infection with other viruses or pregnancy[35]. This might be associated with the implicit alteration of the immune response in the mother. During pregnancy, viral levels are commonly increased and CD8+ T cell cytotoxicity reduced along with loss of HLA escape mutants, with the consequently emergence of enhanced fitness strains, and further reversion during the inter-pregnancy periods[102]. Thus, maternal cellular immunity impairment and emergence of more fit viruses might facilitate HCV perinatal transmission. Compartmentalization of HCV in PBMC has been proposed to play role in perinatal transmission, acting as Trojan horses for viral entry[128,129]. Moreover, PBMC-infected cells have been shown to be a risk factor for HCV vertical transmission[130,131]. In this setting, children infected perinatally bear infection with more fit viruses[101,132]. Transmission of more fit wild-type or revertant viruses could favor persistent infection in infants[128].

Tracking of HCV infection depends on sequence information originated from different subgenomic regions. The 5’-UTR region has been widely used for detection owned to its degree of conservation across genotypes while the NS5B region is the target of election for HCV genotyping[133,134]. However, these two regions do not contain sufficient sequence information to establish genetic relatedness between clinical isolates. Genetic relatedness studies primarily rely on information obtained from the HVR1[34,81,83,119]. Importantly, the rapid divergence in this region represents a challenge for molecular epidemiological studies, which can lead to loss of genetic links between related isolates[34]. Sequencing of multiple and longer subgenomic regions has been proposed as an alternative to overcome the limitations imposed by the rapid molecular evolution of HCV[34]. The NS5A has also been used to establish relatedness among HCV cases[85], which can aid to restore links between isolates owned to a lower nucleotide substitution rate. Despite the usefulness of different subgenomic regions for the characterization of clinical isolates, whole genome sequencing should be the ultimate goal for HCV molecular characterization.

The selective forces that shape the HCV molecular evolution are complex in nature and required sophisticated methods to be assessed. Until recently, molecular approaches used to detect low frequencies variants and assess the intrahost viral genetic variation, such as standard cloning techniques[57], limiting dilution[135] and single genome amplification[59], were cumbersome, time-consuming, and expensive[136]. Moreover, these conventional methods are difficult to implement and provide limited insights in the composition of the intrahost population[133]. The relatively recent advent of next generation sequencing platforms has allowed the detection of rare variants present at much lower frequencies[136-139], facilitating the characterization of the HCV intrahost viral population in remarkable detail[136,140,141].

DISEASE PROGRESSION

HCV molecular evolution is intimately associated with disease outcome and progression[142,143]. HCV-related liver disease gradually advances from chronic hepatitis to liver cirrhosis and to hepatocellular carcinoma (HCC)[144]. The rate of disease progression differs considerably among HCV cases, and the cause for these differences remains poorly understood. However, viral genotypes have been associated with pathogenesis of HCV-related infection, in addition to dictating the path of treatment[145]. Impairment of different pathways has also been associated with the infecting genotypes, which in turns leads to distinct pathological settings. Overall, HCV genotype 1 is associated with more aggressive disease, increased insulin resistance, poor response to therapy, higher risk of cirrhosis and hepatocellular carcinoma development, while genotype 3 is associated with increased steatosis and fibrosis[145,146]. Thus, the identification and characterization of HCV types and subtypes provides insight into the different outcome of HCV infection and responsiveness to therapy[145].

HCV core has been associated with disease progression by inducing apoptosis through the extrinsic and intrinsic pathways[147]. The signaling pathway converges into a common apoptotic pathway where the caspase enzyme cascade is triggered. As a consequence, intracellular components and DNA are degraded, leading to cells death and ultimately causing liver damage[148]. Apoptosis of liver cells plays an important role in the pathogenesis of end stage liver disease[149]. Interestingly, HCV core induces both pro- and anti-apoptotic effects[150-154]. HCV suppresses apoptosis by preventing the release of cytochrome C in the mitochondria, resulting in inactivation of caspase-9, -3 and -7[155]. When HCV core binds to the downstream death domain of the Fas-associated death domain (FADD) protein and FLICE (FADD-like interleukin-1beta-converting enzyme)-like inhibitory protein results in anti-apoptotic state[156]. HCV core also binds to p53 resulting in either inhibition or activation of apoptosis[157,158]. This dual behavior of HCV core in regulation of apoptosis is of the utmost importance in progression to HCC. The association between different amino acid residues in the HCV core region and disease progression has been reported[159]. Particularly, amino acid 70 in the HCV core has been closely associated with progression to HCC[160-163]. A detailed analysis of the quasispecies nature of the HCV core region showed that mutants bearing this residue is linked to advance of liver disease. Sequence variations in the core region have also been reported in tumor cells[164]. Analyses of the full-length of the core gene from patients with HCC showed that mutations in the core protein may affect the course of HCV infection[165-168]. HCV infection is associated with an increased risk of developing diabetes mellitus[169,170]. Additionally, amino acid substitutions in the HCV core from genotype 1b isolates have been suggested as predictor markers for insulin resistance in diabetic patients without signs of cirrhosis[171]. Isolates from patients with severe insulin resistance bear higher proportions of Glu70 (His70) and/or Met91. Therefore, genetic variability occurring in the HCV core is likely to affect its interaction with host proteins and thereby disease progression.

Steatosis is commonly observed in HCV cases and contributes to progressive hepatic injury. Development of hepatocellular steatosis is more frequently observed in cases infected with HCV genotype 3[172,173]. However, not all patients infected with genotype 3 develop steatosis[174]. Therefore, infection with genotype 3 is only one of the elements in the multi factorial nature of steatosis. Other studies have also implicated HCV genotype 3 in rapid progression to fibrosis[175], further supporting the role of genotyping in disease progression.

The complexity of the viral population in the NS5A region has been shown to be associated with HCC[176]. However, the specific mechanisms by which NS5A promotes disease progression are not understood. The NS5A inhibits the activated protein kinase (PKR), an important mediator of the IFN anti-viral response, apoptosis and cell proliferation[177-179], which can lead to liver damage. Additionally, NS5A can act as transcriptional activators affecting cellular signaling[180]. The NS5A encompasses the interferon-sensitivity determining region (ISDR) which is associated with response to antiviral therapy[181]. Thus, the NS5A interaction with host proteins is pleiotropic in nature, making difficult the characterization of the participant elements in pathogenesis.

In individuals progressing to end-stage liver disease, complexity of the quasispecies population at baseline is greater and tends to reduce over time, while non-progressors course with gradual increase in complexity[182]. Patients progressing to end stage liver disease, who likely exert less immune pressure on the HCV population, exhibit reduction in the complexity of the viral population[182]. High genetic complexity of the intrahost HCV population and liver injury has also been observed in immune-competent children associated with vertical transmission in comparison to immune-suppressed patients[183]. The dynamics of the HCV quasispecies are also associated with progression to fibrosis after liver transplantation[184-187]. Recurrence of HCV infection is universal after transplantation; however, clinical outcome is variable[188,189]. Several factors affect the outcome of recurrent HCV disease upon transplantation including HCV genotype[190,191], virus load[192], and co-infection with cytomegalovirus[193,194]. Co-infection with human immunodeficiency (HIV) virus has also been associated with progression to severe liver disease[195]. Co-infected patients with unfavorable prognosis are presented with a relatively stable quasispecies population which can be a reflection of the limited immune pressure commonly observed in these patients[184].

MOLECULAR ASPECTS OF THERAPY AND DRUG RESISTANCE

The “ideal” outcome of the anti-HCV treatment is a sustained virologic response (SVR), defined as undetectable viral RNA for six months after completion of treatment[196]. Until recently, the only option for HCV therapy was a combination of pegylated interferon-α (PEG-IFNα) and ribavirin (RBV). The efficacy of this regime ranged between 20%-80%, depending on race, disease stage, infecting genotype and distinct single nucleotide polymorphisms located in the IFN-λ3 promoter gene[197,198]. Poor outcome was the common feature among patients infected with HCV genotype 1 undergoing IFNα-RBV therapy[196]; thus, the infecting genotype was an important determinant for response to treatment[199]. Therefore, efforts aimed to find predictive markers for SVR were pursued by several groups. Besides infecting genotype and viral titer, other viral molecular factors used to predict IFN therapy outcome rendered controversial results. For instance, the 39 amino acids located in the interferon sensitivity determining region (ISDR) in the N5A gene were proposed to be predictive of SVR[181,200,201]. Thus, four or more substitutions in this small region correlated with responsiveness in Asian patients infected with HCV genotype 1b. However, the association between ISDR and SVR in other races rendered controversial results[202,203].

Major advances in the anti-HCV therapy have been made during the past few years[204]. This has resulted in significant improvement in SVR. The landscape of treatments for HCV is expected to change drastically with more efficient IFN-free antiviral therapies being approved in the near future[205]. Insights into the molecular structure of different HCV proteins have greatly fostered the development of new drugs, commonly referred as direct-acting antiviral agents (DAA). Experience with anti-retrovirals has provided a valuable framework for the development of DAA against HCV. In theory, every step of the HCV replication cycle is a potential target for antiviral therapy; however, good results with HIV protease inhibitors suggested the HCV protease as an ideal candidate. Nonetheless, both NS3 and NS5B have been targets for anti-HCV therapy; with two-thirds of drugs directed against these proteins currently in Phase II and III trials, and few already approved by the United States Food and Drug Administration (FDA). In addition, DAA directed to the NS5A protein have also been developed (Figure 1, Table 1).

Table 1.

Amino acid substitutions within the direct-acting antiviral agents target associated with resistance to different direct-acting antiviral agents agents

Drug target	DAA agents	RAV
NS3/4A protease
	Telaprevir	V36M/A1, T54A/S1, R155K/T1, A156S/T/V1
	Boceprevir	V36M/A1, T54A/S1, V55A1, R155K/T1, A156S/T/V1, V158I, V170A, M175L
	Faldaprevir	R155K1, A156T/V, D168V/E1
	Simeprevir	V36M, F43S, Q80K1, S122A/R, R155K1, A156T/V, D168V/E1, I170T
	ABT-450	V36M/A/G, V55I, R155K1, A156T/V, D168A/V1
	Danoprevir	V36M/A, V55I, R155K1, D168E/T1
	Vaniprevir	R155G1/K/N/S/T, A156V, D168V1/G/A
NS5B polymerase
	Mericitabine	S282T, L159F, L320F
	BI 207127	P495L1/S/A/T/Q, P496S, V499A
	Lomibuvir/VX-222	L419S1, R422K1, M423TV
	Setrobuvir	M414T/L, G554D, D559G
NS5A
	Daclatasvir	M28V/A/T, Q30R/E/H, L31V/M1, Q54H/N/Y, H58D, Q62R/E, A92K/T, Y93H/N/C1

¹Amino acid position number represents the key mutations that are clearly associated with virologic failure. A: Alanine; C: Cysteine; D: Aspartate; E: Glutamate; F: Phenylalanine; G: Glycine; H: Histidine; L: Leucine; I: Isoleucine; K: Lysine; M: Methionine; N: Asparagine; P: Proline; Q: Glutamine; R: Arginine; S: Serine; T: Threonine; V: Valine; Y: Tyrosine; DAA: Direct-acting antiviral agents; RAV: Resistant associated variants.

HCV NS3 protease inhibitors

From a chemical point of view, HCV NS3/4A protease inhibitors can be divided into three main categories: (1) linear peptidomimetics with an alpha-ketoamide group that binds the active site, covalently blocking the enzymatic activity (first class); (2) linear non-covalent peptidomimetic inhibitors (second class); and (3) macrocyclic non-covalent peptidomimetic inhibitors (third class)[206,207].

In May 2011, telaprevir and boceprevir, both belonging to the first class of protease inhibitors, were approved by the FDA for treatment of patients chronically infected with HCV genotype 1. In large Phase III studies, SVR rates of 67%-75% among treatment-naive and 59%-64% in treatment-experienced patients were achieved with triple regimens (PEG-IFNα, RBV plus either telaprevir or boceprevir) in comparison to the standard therapy with IFNα/RBV[109,208-210]. SVR rates depend on the patient previous response to dual therapy and fibrosis stage[211], being considerably lower in patients with null response and liver cirrhosis[210,212]. Additionally, these drugs still require the leading phase with IFNα/RBV; and therefore, are prone to develop drug resistance[207]. Two features contribute to this observation: (1) the low-fidelity of the HCV RdRp and large progeny generated per day; and (2) the complex structure of the HCV protease active site. The estimated error rate of the HCV RpRd is about 10-fold higher than the HIV reverse transcriptase[213], and the virion production rate about 100-fold higher, originating resistant mutants as minor populations in treatment naïve patients[133,214,215]. While some studies have reported frequencies of HCV resistance mutants < 1%[133,214,215], others have shown higher proportions[216]. These resistant associated variants (RAV) can rapidly emerge and come to prominence after few days of treatment and are responsible for treatment failure in many patients[217-219].

Boceprevir and telaprevir share extensive cross-resistance. RAV most frequently associated with telaprevir monotherapy include R155K, A156S/T/V, V36M, and T54A/S (Table 1). Similar variants are observed with boceprevir monotherapy, in addition to V170A and V55A (Table 1). Similar profiles of resistance in vitro have been observed for each substitution resulting in low-to-moderate resistance for V36M, T54A/S, V55A, R155K, A156S, and V170A, and high resistance for A156T/V and double mutants R155K/V36M[218,220]. Substitutions associated with resistance to protease inhibitors generally reduce the catalytic efficiency of the HCV protease. For this reason, these mutants are rarely detected as the dominant variant in the intrahost HCV population in the absence of the antiviral drug. Recent studies have shown that non-responders to protease inhibitors carried a dominant strain of RAV at the time of virus breakthrough or relapse, but the wild-type strain re-emerged as the dominant variant after completion of therapy[221,222]. RAV have been reported for most antiviral agents in development, with widespread cross-resistance for both linear and macrocyclic groups[218,223]. For example, certain substitutions involving residues such as R155 as well as R155/A156 often confer resistance against most protease inhibitors at a moderate fitness cost to the virus. Thus, in the absence of antiviral drugs, RAV replicate with moderate efficiency compared to wild-type viruses[222,224].

Telaprevir and boceprevir have restricted spectrum of action over HCV genotypes. While telaprevir has some clinical effect against genotype 2, and boceprevir seems to be effective against genotype 3, their use is prescribed off-label for these genotypes[223]. Drug resistance to protease inhibitors displays a low genetic barrier (i.e., the number of nucleotide changes required to generate RAV) which represents a particular problem in HCV subgenotype 1a infection, as the genetic barrier is lower compared to subgenotype 1b[225].

Second and third classes of NS3 protease inhibitors, which do not form covalent bounds with their targets, have several advantages over the first-class compounds[226]. These NS3 protease inhibitors include linear non-covalent molecules such as faldaprevir, asunaprevir, sovaprevir, GS-9451 and macrocyclic inhibitors such as simeprevir, danoprevir, ABT-450, GS-9256; and vaniprevir (Table 1)[207,227-234]. Most of these new DAA are currently in Phase II or III clinical trials with the exception of simeprevir which has been already approved by FDA. These antiviral agents have showed high rates of SVR, comparable or higher than boceprevir or telaprevir triple-combination regimens, in HCV genotype 1 patients when used in combination with PEG-IFN and RBV. These NS3 HCV protease inhibitors tend to have a broader spectrum of action over different HCV genotypes; however, these DAA are not pan genotypic antivirals owned to different inhibitory efficacies across genotypes[227,235,236]. Simeprevir exhibits high antiviral activity against genotypes 2, 4, 5, and 6 whereas no effect has been observed with genotype 3[236]. First class protease inhibitors genetic barrier is low and extensive cross-resistance in comparison to second and third class protease inhibitors[218]. In particular, mutations R155K/T/Q have been shown to confer broad cross-resistance while mutations D168/E/G/H/T/Y confer resistance specifically to non-covalent peptidomimetic inhibitors, regardless of linear or macrocyclic structure[218,233,237-239]. Notably, D168 is one of the few active site residues not entirely conserved among HCV genotypes, which is replaced by glutamine in isolates belonging to genotype 3, partly explaining why HCV genotype 3 is “naturally resistant”[240]. Moreover, substitution Q80K is present in about 20% of genotype 1a sequences and confers resistance to Simeprevir[237]. This mutant has been associated with breakthrough infection in treatment-naïve subjects[220].

NS5B polymerase inhibitors

Another strategy to inhibit viral replication is the development of polymerase inhibitors which interfere with viral replication by binding to the NS5B RdRp. NS5B inhibitors can be divided into two distinct categories: nucleosides inhibitors and non- nucleoside inhibitors.

Nucleoside analogue inhibitors mimic the natural substrates of the polymerase causing direct chain termination[241,242]. The NS5B active site is well-conserved across HCV genotypes as amino acid substitutions in this location are generally poorly tolerated and result in remarkable loss of fitness[223,243]. Among nucleoside analogue, mericitabine has been shown to increase SVR rates in patients infected with genotype 1 and 4[244]. Sofosbuvir, recently approved by the FDA, has demonstrated extraordinary efficacy for treatment of adults infected with genotypes 1 and 4 in combination with IFN/RBV, and for the treatment of adults infected with genotypes 2 and 3 in IFN-free regimens[106,245,246]. Importantly, several factors usually associated with poor outcomes in HCV infected patients had no effect on the efficacy of sofosbuvir; thus, only male gender and the presence of cirrhosis are predictive of unfavorable outcome. Substitution S282T conferring resistance to sofosbuvir has been observed in vitro but not in vivo[223]. However, mutants L159F/L320F in the NS5B polymerase conferring low-level resistance to mericitabine and sofosbuvir has been reported recently[223,247]. Overall, nucleoside polymerase inhibitors tend to exhibit good activity against a broad range of genotypes and have a high genetic barrier to resistance which makes them a promising class of anti-HCV agents.

Non-nucleoside inhibitors are chemically and functionally much more diverse than nucleoside inhibitors. Non-nucleoside inhibitors usually bind to several discrete sites on the HCV polymerase, which results in conformational protein changes before the elongation complex is formed[241,242]. A limitation of this mechanism of action is that binding sites are less conserved among genotypes compared to the active site. As a consequence, lower cross-genotypic activity and higher probability of RAV development is observed. Despite efficacy against HCV genotype 1, resistance occurs at a low fitness cost[218].

Several non-nucleoside inhibitors targeting at least four distinct allosteric binding sites are in development. Two of these binding sites, named “thumb I” and “thumb II”, are located on the polymerase thumb domain, whereas the other two sites, “palm I” and “palm II” are close to the active site cavity and involve primarily amino acids from the palm domain. Certain non-nucleoside inhibitors develop drug resistant variants carrying mutations at different positions, e.g., BI 207127 targeting thumb I at P495, P496 and V499; lomibuvir targeting thumb II at L419, R422 and M423 as well as setrobuvir affecting palm I at M411, G554 and D559[220].

NS5A inhibitors

Disruption of the recruiting capabilities of NS5A can also be targeted for antiviral drug development. Recently, identification of lead compounds capable of inducing a rapid decline in viral load and emergence of RAV, characterized by amino acid substitutions located in the NS5A protein, confirmed their specificity[248,249]. These agents feature broadly genotypic coverage but low genetic barrier to resistance[207,250].

Daclatasvir is a replication complex inhibitor currently in Phase III. Daclatasvir is active at picomolar concentrations in vitro displaying activity over a broad range of HCV genotypes[249,251]. The resistance profile of daclatasvir has been linked to the N-terminus of NS5A[249]. The most remarkable RAV associated to daclatasvir are Y93H/C/N, which also confer cross-resistance to other NS5A inhibitors[249]. Interestingly, Y93 is found near the protein dimer interface, leading to speculate that NS5A inhibitors might affect the monomer/dimer equilibrium[252]. Monotherapy with these drugs rapidly selects for the outgrowth of resistant variants, similar to other HCV protease inhibitors[220,251,253,254].

Interferon-free therapy regimens

IFN-free combination trials are ongoing with different DAA. The current challenge is the development of an oral regimen including compounds with different mechanisms of action, synergistic interactions and pan-genotypic activity. Drugs in an IFN-free therapy should not display overlapping resistance profiles. In principle, this is achievable since most DAA target different viral proteins or binding sites in the same protein. Several DAA are expected to successfully complete phase III trials in anticipation of licensing. Initially, different IFN-based regimens (sofosbuvir, faldaprevir and simeprevir) will be readily available for treatment of HCV genotype 1. In the near future, combination of antiviral agents lacking cross-resistance with good safety profile will be the new recommended therapy regime. Recent clinical studies have shown SVR > 90% in patients administered with DAA cocktails including protease and polymerase inhibitor[207]. For example, Sofosbuvir is currently being evaluated in IFN-free combinations with simeprevir and daclatasvir. In turn daclatasvir efficacy is being assessed in combination with asunaprevir and/or non-nucleoside polymerase inhibitor BMS791325[207,255].

The rapid development of DAA has broadened the landscape of anti-HCV drugs which will probably include extensive number of possible ingredients for an effective combinatory regimen. Results from recent clinical trials have firmly established the concept that a permanent cure can be achieved with IFN-free combinations of DAA[96,256].

CONCLUSION

HCV molecular evolution in many ways dictates virus transmission, progression to severe liver disease and therapy outcome. The molecular mechanisms exploited by the virus to achieve persistence are numerous and complex in nature. The sophisticated evolutionary process to which HCV is subjected during transmission and upon infection further limits our understanding of the participating elements of HCV pathogenesis. The remarkable HCV mutation rate represents a challenging task for vaccine development and molecular epidemiology. In this new era of advance sequencing technologies, the implementation of enhanced molecular surveillance is of the utmost importance to accurately monitor circulation of viral strains. Comprehensive molecular studies are also required to uncover the participant elements responsible for virulence.

The rapidly evolving field of DAA is heading to the development of IFN-free regimens with superior SVR and pan genotypic activity and less prone to side effects. However, emergence of RAV remains an important concern. Importantly, and despite the emergence of HCV RAV, which are frequently developed upon anti-HCV treatment, high SVR rates are attainable. Detection of minor variants should be conducted before and during treatment with new drugs to monitor development of resistance and better manage patients.

The major advances on therapy have been the results of many efforts focused on discovering the nuances associated with HCV replication. However, the development of successful vaccines imposed a more daunting task. Thus, the path to HCV vaccination should be as exciting, if not more, than the road to efficacious treatment.