Hepatitis C Virus (HCV) is one of the major public health challenges worldwide (Gokhale et al., 2014). According to the 2023 WHO report, 58 million people globally live with chronic HCV infection (World Health Organization, 2024). HCV can establish acute and chronic infections (Sevvana et al., 2021). Acute infections tend to be asymptomatic, and the host system can clear the virus within six months of infection through the coordinated action of the innate and adaptive immune response (Grebely et al., 2012). However, the majority (70%) of HCV-infected individuals develop chronic infections, which can manifest as liver cirrhosis, fibrosis, and hepatocellular carcinoma (Chigbu et al., 2019). One reason for the progression from acute to chronic infection is the viral escape, attributed to immune-escape mutations, from CD8+ T cell response, especially in the acute infection stage (Cox et al., 2005a; Cox et al., 2005b; Neumann-Haefelin et al., 2008b). HCV epitope mutations occur in a targeted manner, with approximately half of class I epitopes acquiring immune escape mutations during acute infection, particularly in non-structural proteins (NS3, NS5B, NS2, and E1), though the pattern and stability of these mutations depend on both viral fitness requirements and host immune responses (Timm and Walker, 2015; Walker et al., 2024; Zhang et al., 2023a). These epitope variants facilitate viral persistence by reducing their binding affinity to Major Histocompatibility Complex (MHC), also known as Human Leukocyte Antigen (HLA) in humans, which subsequently can affect presentation to and recognition by the CD8+ cytotoxic T lymphocytes (CTLs) (Mei et al., 2022).

CTLs are major effector cells that produce cytokines, facilitate B cell maturation, and kill HCV-infected cells (Hofmann et al., 2021). However, under immune response pressure, HCV epitopes undergo mutations to evade recognition by CTLs, and these escape variants can even replace the wild-type sequence, adding new consensus sequences to the existing pool of variable viral antigens (Neumann-Haefelin et al., 2008a). De novo T cell response against mutated epitopes can be incomplete and might not even occur in chronic HCV infection due to the high viral load during persistent infection (Kemming et al., 2020).

Numerous studies have shown that HCV can amplify mutations under the selection pressure from the host immune system (Walker et al., 2015; Ulsenheimer et al., 2013), leading to increased genomic diversity observed at the level of genotypes, subtypes, and quasispecies (Walker et al., 2015). To date, about 8 genotypes and 86 subtypes of HCV have been identified (Hedskog et al., 2019). Genotypes (GT) are approximately 30–35% polymorphic at the nucleotide level, and subtypes can differ by approximately 20% within the same genotype (Galal et al., 2014; Simmonds et al., 1994).

The variations/mutations in the HCV genome can affect the virus epitope recognition by the host immune system, leading to failed viral clearance (Wölfl et al., 2008; Zhang et al., 2017; Merani et al., 2011). For example, Walker et al. (2015) showed differences in CD8+ T-cell responses due to a single amino acid substitution in a conserved epitope observed in HCV GT 3a compared to a wild-type epitope observed in 1a. Similarly, a cohort study of people reinfected with HCV demonstrated that host immune responses to a specific HCV genotype did not guarantee protection against heterogeneous HCV genotypes. These responses were associated with a 49% reduction in viral rebound compared to the initial viral clearance (Islam et al., 2017).

To our knowledge, most existing studies on HCV vaccine design focus on targeting specific HCV proteins (Walker et al., 2022; Dazert et al., 2009) or a set of epitopes without considering genotype specificity (Ikram et al., 2018). Our hypothesis is that in silico methods can demonstrate how epitopes from different genotypes/subtypes at the same sequence position may exhibit varying binding energies, potentially reflecting differences in T cell responsiveness.

Therefore, this study employed sequence analysis, immunoinformatics, molecular docking, and molecular dynamics simulation approaches to examine the effect of genotype/subtype-specific polymorphisms within the Core, NS3, NS5A, and NS5B sequences on HLA binding and CTL epitopes.

The adaptive immune system, primarily CD8+ T lymphocytes (CTL), is crucial in limiting the HCV viral load from 4 to 8 weeks after infection (Neumann-Haefelin et al., 2007). However, in 70% of individuals, the immune response fails to eliminate the virus, eventually leading to chronic viremia (Chigbu et al., 2019; Neumann-Haefelin et al., 2007). A previous study suggests genetic variability is one reason for this failure (Lapa et al., 2019). Genetic variation in the HCV genome is primarily driven by low-fidelity RNA polymerase (Neumann-Haefelin et al., 2007; Lapa et al., 2019), resulting in a mutation rate of 3.5 × 10⁻⁵ to 1.2 × 10⁻⁴ substitutions per round of replication (Gokhale et al., 2014; World Health Organization, 2024; Sevvana et al., 2021). HCV genetic diversity is evident at the genotype (GT) level (GTs 1–7), with a 25–35% difference in nucleotide sequence between genotypes, and at the subtype level (defined by letters such as 1a, 2b, 3c, etc.), with a 10–20% difference within each subtype (Hedskog et al., 2019; Galal et al., 2014; Simmonds et al., 1994; Nakamura et al., 2022). Additionally, these genotypes and subtypes differ in their epidemiology, drug response, and interaction with the host immune system (Wölfl et al., 2008; Zhang et al., 2017; Merani et al., 2011; Nakamura et al., 2022; Echeverría et al., 2015). Our study aimed to identify genotype/subtype-specific polymorphisms within the Core, NS3, NS5A, and NS5B sequences on HLA binding and CTL epitopes. We examined 15, 146, 105, and 177 genotype/subtype-specific epitopes in the Core, NS3, NS5A, and NS5B proteins, respectively.

Overall, we found 66, 295, 329, and 322 genotype-specific variations in the Core, NS3, NS5A, and NS5B protein sequences, respectively. Genotype-specific variations in HCV strains impact clinical outcomes and drug responsiveness. For instance, previous clinical studies have identified HCV genotype 3 (GT3) as a risk factor for rapid liver fibrosis and hepatocellular cancer (McMahon et al., 2017). Additionally, prior studies identified epistasis in the HCV sequence as a driver of drug resistance, suggesting that HCV genetic diversity can weaken host immune responses and contribute to resistance mechanisms (Zhang et al., 2023b). These findings underscore the importance of considering immunogenic variation among genotype-specific HCV strains in vaccine design and drug development. Polymorphisms, specifically in the NS3, NS5A, and NS5B non-structural proteins, exhibited the most significant genotype/subtype-specific alterations. The NS3 protein has helicase and serine protease activity. The latter can cleave and inactivate host proteins essential for the innate immune system (Raney et al., 2010). NS5A is crucial for viral RNA replication and modulation, and it can suppress both innate and adaptive immunity, leading to chronic infection (Raney et al., 2010; Kriegs et al., 2009). NS5B, part of the HCV replication complex, can harbor mutations associated with resistance to direct-acting antivirals (Nguyen and Van Le, 2022). Therefore, we hypothesize that genotype/subtype-specific polymorphisms, especially in the non-structural proteins, might be responsible for downregulating CTL-mediated immune responses.

Our analysis showed that 11 pairs of epitopes (predicted using NetCTL) were affected by genotype/subtype-specific polymorphisms at both sequence level and in CTL scores (Table 1). For example, mutations Y43T in GT 3 h, T1619R and A1621P in GT 4d/l, and T2381A and A2382G in GT 6 k resulted in altered CTL scores compared to their paired epitopes from other genotypes/subtypes. NetCTL tool has been employed with similar settings applied in studies of pathogens such as M. tuberculosis (Bibi et al., 2021), SARS-CoV-2, and the Omicron variant (Khairkhah et al., 2020; Aasim et al., 2022). Among various prediction tools, NetCTL has demonstrated superior predictive performance for HIV epitopes compared to alternative tools such MAPP, EpiJen, WAPP, and MHC-pathway (Larsen et al., 2005), which significantly increases the confidence in the findings.

In the next step, we employed immunoinformatic techniques, namely HPEPDOCK and MD simulation, to examine interactions between subtype-specific epitopes and the restricting HLA molecules. In silico methods have widely been employed in vaccine design for their ability to screen numerous epitopes simultaneously and narrow down candidates for further in vitro and in vivo validation (Elshafei et al., 2024; Jyotisha and Qureshi, 2020). This approach enhances cost and time efficiency in the preclinical stages of vaccine development.

The HLA-peptide docking analysis identified three pairs of epitopes within NS3, NS5A, and NS5B proteins that exhibited a difference in HPEPDOCK scores exceeding one standard deviation (27.66 kcal/mol). Using standard deviation thresholds to define cutoffs is an accepted practice in structural bioinformatics and docking analyses, and has been used in previous studies (Kuhn et al., 2020; Cheng et al., 2010). Our results are supported by a previous study, showing that the CTL-mediated response in GT 3a, unlike in GT 1a, exclusively targeted non-structural HCV proteins during chronic infection (Keikha et al., 2020). In our study, the epitope LGFGAYMAK (GT 2f) in the NS3 protein exhibited a CTL score of 0.83 and a docking score of −236.67 kcal/mol, whereas the epitope LSFGAYMSK (GT 2b) had higher CTL and lower docking scores of 1.59 and − 264.33 kcal/mol, respectively. However, the opposite trend was observed for the NS5B protein, where the epitope AAKLKDFDM (GT 6d/r) resulted in a lower CTL score (0.76) and higher docking score (−207.79 kcal/mol) compared to the epitope AAKLRDCTL (GT 5a) with CTL score of 1.70, docking score of −176.31 kcal/mol. A similar phenomenon was reported by Wang et al. (2010), where variant NS3 epitopes showed a diminished CTL response compared to wild-type sequences. However, none of the variant sequences abolished HLA receptor binding, and the difference in HLA-docking complexes with either variant or wild-type epitopes was insignificant. Nonetheless, a previous study on HCV GT 1b sequences from a single-source outbreak in Irish women, using an HLA-epitope binding prediction model, reported a contrary trend (Kim et al., 2011). The study showed a 100-fold shorter dissociation time with the HLA-B*57 receptor compared to a reference sequence from GT 1a. Additionally, the interferon-γ staining assay showed no T cell response to variant epitopes, leading to chronic viremia. The authors claimed that HCV sequence variations, predominantly in the E2 and NS5 proteins, diminished virologic control by the HLA-B*57-mediated T-cell response (Kim et al., 2011).

In the next step, we used molecular dynamic (MD) simulation throughout 200 ns to explore the dynamics of the epitope-receptor complex over time. An NPT ensemble was employed to maintain constant temperature and pressure throughout the simulation. These conditions (300 K or 27°C and 1 atm or 760 mmHg) were selected to align with similar studies (Ren et al., 2022) and to reflect physiological conditions, as well as standard in vitro settings for HCV experiments (Song et al., 2010). The addition of NaCl stabilizes electrostatic interactions and ensures a neutral charge in the simulation box (Ribeiro et al., 2012). Also, a concentration of 0.15 M NaCl is a physiological salt level within mammalian cells, which more closely replicates in vitro conditions.

To highlight differences in binding energy among epitopes positioned at the same HCV sequence but originating from different genotypes, we analyzed binding energy (ΔG). Generally, more negative binding energy indicates a stronger interaction within the epitope-HLA complex, enhancing the likelihood of T-cell activation (Corse et al., 2011). There is a growing body of research linking negative ΔG values to strong protein interactions, confirmed through in vitro methods such as ELISA (Jain and Baranwal, 2019), X-ray crystallography (Li et al., 2023), and cryo-EM (Sušac et al., 2022), as well as in combination with other immunoinformatics tools (Morgan et al., 2024).

In our study, ΔG differences observed in MD simulations followed the trend seen in the docking results. For example, the epitope LSFGAYMSK in GT 6b resulted in a twice as low ΔG of −144.24 kcal/mol compared to the LGFGAYMAK epitope in GT 2f (−85.30 kcal/mol). Similarly, AAKLRDCTL from HCV GT 5a resulted in a weaker ΔG of −52.38 kcal/mol compared to AAKLKDFDM in HCV GT 6d/r (−87.78 kcal/mol). Finally, both DTASDAGSY (GT 6o) and DAGSDAGSY (GT 6 k) exhibited a similar binding energy value of −42.39 kcal/mol and 43.03 kcal/mol, respectively.

We analyzed binding energy in percentage terms to highlight differences in binding energy among epitopes located at the same HCV sequence position but from different genotypes. In MD simulation, epitopes from GT 6b, and 6d/r exhibited up to a 40% variation in binding energy when interacting with the HLA receptor. In previous studies, epitopes with binding energy (ΔG) to HLA receptors ranging from 18 to 43% lower were identified as having strong binding potential, which is critical for vaccine design (Mei et al., 2022; Kyobe et al., 2024; Chieochansin et al., 2024). In contrast, a 40% decrease in binding energy can weaken the interaction, potentially diminishing or even eliminating the T-cell response to the virus. These results also follow the same trend observed by Khairkhah et al. (2020), suggesting that immune-resistance mutations, capable of escaping both CTL responses and HLA recognition, can generate at the nucleotide level, leading to novel quasispecies and subtypes circulating within the population (Lapa et al., 2019).

In vitro study of HCV-derived T cell epitopes showed that sequence variations in HCV can modulate NK cell functions, creating potential pathways for viral immune escape (Lunemann et al., 2016). These findings underscore the importance of considering immunogenic variation among genotype-specific HCV strains in vaccine design and drug development. As we mentioned in the introduction, one primary reason for the progression from acute to chronic infection is the ability of HCV to evade the CD8+ T cell response, through immune-escape mutations, during the acute infection stage (Cox et al., 2005a; Cox et al., 2005b; Neumann-Haefelin et al., 2008b). Research by Bulteel et al. (2016) showed that spontaneous clearance of HCV is rare, occurring at a rate of 0.36 per 100 person-years, while 60–80% of infected individuals develop chronic infection (Saraceni and Birk, 2021). This indicates that even a robust immune response may not always prevent disease progression. Moreover, around 20% of patients with chronic HCV may develop liver cirrhosis within 25 years (Saraceni and Birk, 2021; Rustgi, 2007), and 30% of chronically diseased individuals are at risk of hepatocellular carcinoma (Khatun et al., 2021). Therefore, understanding how HCV’s high genetic variability affects immune recognition is crucial for predicting disease progression and determining treatment approaches (Saraceni and Birk, 2021).

Our results indicate that epitopes from genotype 6b (NS3), 6o and 6 k (NS5A), 6d/r(NS5B) exhibit higher immunogenicity compared to other genotypes, forming more energetically stable complexes with the host receptor (Table 1, light-yellow shaded rows). These findings suggest that patients infected with GT 6 may have better T cell responsiveness and broad immunogenicity. While 46.2% of all HCV cases in the world are attributed to GT1; GTs 2, 4, and 6 together make up 22.8% of cases (Messina et al., 2015). HCV genotype 6 emerged in the 2000s and has become the third most prevalent genotype in Southern Asia and surrounding regions (including China, Thailand, Indonesia, Cambodia, Malaysia, Myanmar, and Vietnam) (Irekeola et al., 2021). Here GT 6 represents 30–40% of HCV infections (Nguyen and Nguyen, 2015). According to the World Bank data, these regions are low, and low-middle-income countries (World Bank Country and Lending Groups, 2024), where access to medical treatments is limited, and many people rely on traditional remedies (Nguyen and Nguyen, 2015). While genotype 6 is less common in Western countries than GTs 1, 2, and 3, the limited preclinical and clinical data from these areas aids further investigation into GT 6 (Nguyen and Nguyen, 2015). HCV GT 6 also has distinct endemic subtypes, and its genetic diversity raises concerns about resistance to pan-genotypic direct-acting antiviral treatments. Additionally, although individuals infected with HCV GT 6 exhibit clinical symptoms similar to those of other HCV genotypes (Chao et al., 2011), numerous studies have revealed striking differences in drug therapy responses specific to GT6 (Nguyen and Nguyen, 2015; Flower et al., 2021). Notably, a previous study found that patients infected with HCV GT 6 responded better to interferon therapy compared to patients infected with genotype 1 (Hui et al., 2003). Also, GT 6 patients have been shown to experience fewer severe clinical outcomes, such as cirrhosis (Yao et al., 2020). These findings highlight the importance of considering immunogenic variation among genotype-specific HCV strains in vaccine design and drug development.

Overall, this study used sequence analysis, immunoinformatic, molecular docking, and molecular dynamics simulation to analyze a large HCV sequence dataset to identify genotype- and subtype-specific polymorphisms that can impact CD8+ T cell epitope processing and HLA-epitope interactions. In silico techniques have widely been employed in vaccine design, offering the advantage of screening numerous epitopes simultaneously and narrowing down potential candidates for subsequent in vitro and in vivo validation (Elshafei et al., 2024; Jyotisha and Qureshi, 2020). This approach contributes to preclinical steps in vaccine development by enhancing cost and time efficiency. Our results indicate that epitopes from genotype 6b (NS3), 6o and 6 k (NS5A), 6d/r(NS5B) exhibit higher immunogenicity compared to other genotypes, forming more energetically stable complexes with the host receptor. These findings suggest that patients infected with GT 6 may have better T cell responsiveness and broad immunogenicity.

The emergence of personalized medicine has facilitated more targeted therapies that account for individual, cultural, and geographical diversity (Wang and Wang, 2023). Due to the global variation in HCV genotypes and subtypes, there has been an increasing discussion about designing a genotype/subtype-targeted vaccine (Zhang et al., 2023b). This strategy could be especially beneficial since HCV subtypes differ in binding affinity to HLA receptors, leading to varying TCR responses. Prior studies on other viral infections have highlighted the effectiveness of genotype-specific vaccines (Kim et al., 2024; Mathew et al., 2023; Haslwanter et al., 2022). We propose that HCV subtype/genotype-specific vaccines may help prevent the emergence of new quasispecies, which can arise from accumulating immune escape mutations.

We identified several limitations in our study. Firstly, the analysis was conducted using in silico tools. Although we tested the viral epitopes with multiple approaches, including epitope prediction, HLA-peptide docking models, and MD simulations, each step has its own limitations. For example, although MD simulations are much superior to docking, they are constrained by timescales, as longer timescales (which may be required for biological processes) may be highly computationally demanding and cannot be handled by conventional systems (Wang et al., 2019). Additionally, MD simulations may have limited ability to sufficiently sample peptide-protein interactions and/or capture rare events such as peptide unbinding without an enhanced sampling technique (Wang et al., 2019; Aghajani et al., 2022). Therefore, further in vitro and in vivo validation is necessary to fully establish the impact of genotype/subtype-specific polymorphisms on the adaptive immune response. For instance, T cells can be co-cultured with antigen-presenting cells displaying the epitopes and T-cell activation can then be measured using cytokine release assays like ELISPOT (for IFN-γ) and flow cytometry for activation markers (CD69, CD25) to confirm immune response (De Groot et al., 2001). Secondly, this analysis focused solely on CTL epitopes, given the significant role of CTL responses in HCV control (Hofmann et al., 2021; Larrubia et al., 2015; Nelson et al., 1997). Future studies should expand the analysis by including CD4 and B-cell epitopes to examine the impact of genotype/subtype-specific polymorphisms on other immune response components. Thirdly, the last major update to the Los Alamos HCV database was in 2007, which may exclude newer HCV sequences. Additionally, certain low-income regions remain underrepresented, potentially leading to biases that underestimate genetic diversity in these areas. Despite these limitations, research using sequences from the Los Alamos HCV database has continued to yield valuable insights into HCV genetics over the past five years (García-Crespo et al., 2020; Da Silva et al., 2019; García-Crespo et al., 2021). Finally, given the constraints of comparing single docking scores, statistical methods such as Z-scores or percentile-based cutoffs were not applicable. Therefore, the one-standard deviation method was selected as a simple, quantifiable criterion that streamlined the selection process. Although we acknowledge that this method may be less robust than standard statistical tests, our primary focus is on highlighting the biologically relevant differences in genotype-specific immunogenicity of HCV.

In conclusion, our study demonstrates that differential CTL responses to HCV can emerge due to genotype/subtype-specific variants, thereby challenging HCV control and altering disease dynamics in patients infected with different subtypes. In addition to CD8+ T cell escape, amino acid substitutions in HCV proteins of specific subtypes affect HLA binding, suggesting a dual antagonizing effect on the host adaptive immune system. In this light, along with previously reported results on the immune evasion of mutated HCV epitopes (Neumann-Haefelin et al., 2008b; Spear et al., 2016; Khan et al., 2022; Osuch et al., 2022), the immune escape mechanism of the virus appears to be facilitated by viral genetic polymorphism at the level of genotypes and subtypes. This insight might be valuable for designing vaccines considering HCV genotype/subtype-specific differences to optimize host immune responses.