Like other RNA viruses HCV also
Like other RNA viruses HCV also exhibits a high degree of genetic diversity, creating a major challenge for the development of both HCV vaccines and pan-genotypic therapeutics (Timm and Roggendorf, 2007). The virus has a total of seven genotypes, with >50 subtypes and millions of quasispecies (Cuypers et al., 2015). Among the seven genotypes of HCV, genotype 1 is the most prevalent worldwide, mainly present in Europe and America followed by genotype 3 (Messina et al., 2015).
The seven genotypes of HCV have been reported to differ by as much as 33%, with variations distributed throughout the genome (Okamoto et al., 1992). Most diversity however lies in the HVRs of E2 (Okamoto et al., 1991). Majority of our knowledge on HCV biology is based on genotypes 1 and 2 as reagents such as the subgenomic replicon system, HCVcc (cell culture adaptable HCV) and HCVpp (HCV pseudoparticles) have been prepared from these strains (Catanese and Dorner, 2015). However, clinical information on other strains are quite well documented and it is widely accepted that the pathogenesis and prognosis of infection by various genotypes vary widely (Alonso et al., 2016; Bochud et al., 2009; Zein, 2000), with genotype 3 being more virulent and pathogenic compared to the others (Chayama and Hayes, 2011). Genotype 3 has therefore been the focus of this study in comparison to genotype 1.
In this study, we compared E2 sequences between genotypes 1 and 3 at the nucleotide level and then extended our quest to the amino l-name and structural level. As nucleotide differences must be translated into differences at the amino acid as well, codon usage bias (CUB) where preference is given to one or two codons over the other synonymous ones was investigated. CUB stems from translational pressure governed by the conditions present in the host as well as mutational pressure (Chen, 2013). Such CUB has also been reported in HCV, this present study aimed to explore how such codon bias is maintained in spite of the wide differences at nucleotide and protein level. Furthermore using the protein modelling tools, E2 models were generated for genotypes 1 and 3 and their binding affinity to CD81 was calculated. This study brings to light an important aspect of E2-CD81 interaction; differences between genotypes 1 and 3 which once validated experimentally could be a major step towards design of better genotype specific therapeutics.
Conclusion and discussion Previous reports analysing HCV whole genome has also revealed a bias to be GC rich (van Hemert et al., 2016) in accordance with our results. That particular study compared 29 animal RNA viruses of which one was HCV. We have confirmed these results using 34 sequences of HCV across four genotypes. We have reported a distinct GC bias in the entire genome as well as in the glycoproteins E1, E2 and RNA dependent RNA polymerase (RdRp). While the exact reason for a preference for G and C remains speculative, few reports suggest that GC richness helps in immune evasion as AU rich regions elicit strong innate immune response (Vabret et al., 2012). Codon usage of HCV has been reported to be in the early 50s in a study investigating many RNA viruses (Belalov and Lukashev, 2013). Unicellular organisms such as E. coli (Botzman and Margalit, 2011; Pan et al., 1998) also exhibit such bias as use of rare codons was hypothesized to slow down the translation rate to ensure optimum protein expression and folding (Gouy and Gautier, 1982; Thanaraj and Argos, 1996). Synonymous substitutions may result in an altered structure and thereby often helps in innate immune response evasion (Kondili et al., 2016; Simmonds et al., 2004). As RNA viruses create random mutations driven by the poor proof reading activity of most RdRp, translational pressure that stems from the availability of tRNAs in the host required for the viral protein translation becomes important. The liver however has been reported to be quite abundant in its content of tRNAs compared to other tissues such as the brain (Dittmar et al., 2006). Interestingly, tRNAs for hydrophobic amino acids are more pronounced in the liver (Dittmar et al., 2006) and almost all strains of HCV maintain a similar preference for hydrophobic amino acids (data not shown). Few codons which have been shown to be preferred in the liver (Dittmar et al., 2006) were also seen to be preferred by all genotypes of HCV. These included Arg, Gly, Leu, Ser, Thr and Val. Transfer RNAs (tRNA) with a C or G in their 1st nucleotide for anticodon, obviously undergoes a stronger codon-anti codon interaction resulting in a more efficient translation. In fact binding energy of mRNA with tRNA in HCV is stronger than that in the human host (Allnér and Nilsson, 2011). Furthermore, viral infection may also regulate the activity of RNA polymerase III, thereby affecting tRNA synthesis (Fradkin et al., 1987; Hoeffler and Roeder, 1985). However in spite of many advantages of GC richness, it was intriguing to note that genotype 3 strains had a less preference for C richness compared to genotype1. Furthermore as it is well established that genotype 3 is more pathogenic than genotype1 (Chayama and Hayes, 2011), the implication of such reduced preference remains to be investigated. This codon usage can also contribute to understanding the interaction between virus and the host immune system (Kondili et al., 2016).