Open Access

Lisa Kuhnhenn

• Infections with the hepatitis B virus (HBV) or the hepatitis C virus (HCV) lead to complications like the development of cirrhosis or hepatocellular carcinoma. These complications end up in 887,000 and 500,000 deaths per year, respectively. Since the development of new direct acting antiviral agents for HCV in the past years a complete cure of an HCV infection can be achieved in the majority of the patients. In contrast, a complete cure of a chronic HBV infection still remains a challenging problem as current treatment regimens mainly suppress the viral replication and cccDNA as well as integrated DNA still persist in these patients. Several viral and host factors were described to impair the efficacy of treatment regimens or influence the course of the infection. Therefore, in this work viral factors as well as host factors were investigated in HBeAg negative chronic HBV infected patients and in chronic HCV infected patients. In the present study, it was demonstrated that mutations and/or deletions in the HBV basal core promoter (BCP), the precore and the preS domain occur in a genotype-specifc pattern in HBeAg negative HBV infected patients. While the BCP double mutation A1762T/G1764A was found with the highest prevalence in genotype E infected patients, the precore mutation G1896A occurred mostly in genotype B infected patients. Variants in the preS domain could be detected with the highest frequency in patients infected with genotype C. In patients, who had to start an antiviral therapy during the course of the disease, mutations in the precore region could be detected with a higher frequency in the samples right before treatment start in comparison to the baseline sample. While different HBV genotypes and preS mutations were not associated with HBV-DNA serum levels, precore mutations as well as BCP mutations were significantly associated with HBV-DNA levels. Furthermore, precore mutations showed lower and preS mutations higher HBsAg levels. The HBsAg serum levels varied significantly among the different genotypes. Since HBsAg levels < 1000 IU/ml have been described as a prognostic marker in several studies, the prevalence of patients with HBsAg < 1000 IU/ml was analyzed among the genotypes A - E. While most of the patients infected with HBV genotype B had HBsAg < 1000 IU/ml, only a few patients infected HBV genotype E and A had HBsAg < 1000 IU/ml. Furthermore, HBV genotype A genomes derived from patients harboring a) A1762T/ G1764A (BCP), b) G1896A/G1899A (precore), c) 15 aa deletion in preS1, d) no mutation (reference genome) were cloned and analyzed in vitro. An enhanced expression but reduced secretion of viral genomes was found in the preS-deletion- and the precore-variant. No differences in the HBsAg production and secretion were observed in the cloned precore- or BCP-variant, while the preS-deletion-variant was characterized with an elevated HBsAg release. Regarding the secretion of viral and subviral particles, a genotype-specifc pattern of the L/M/SHBs ratio was detected in the serum of patients infected with genotypes A - E. This pattern did not change in the serum of patients, who started antiviral treatment. Secreted HBsAg containing particles displayed a higher density as well as a higher filaments/spheres ratio in genotypes B and D compared to genotypes A, C and E. Population-based and deep sequencing revealed large deletions in the preS domain or preS2 start codon mutations in a certain number of the viral genomes. Theoretically, these mutations/deletions should influence the molecular weight of the expressed protein or abolish the expression of the protein at all. In contrast, LHBs/MHBs were detectable and appeared at the same molecular weight in these patient samples in comparison to patient samples without these mutations. Furthermore, in the in vitro analyses comparing the reference genome and the preS1-deletion genome, it was shown that the deletion indeed influenced the molecular weight of LHBs. Therefore, HBsAg might be expressed from a genetically different source than the released viral genomes, meaning the integrated DNA. Additionally, in the present study the prevalence of resistance associated substitutions (RASs) in the viral genes NS3, NS5A and NS5B of chronic HCV infected patients was analyzed in correlation to single nucleotide polymorphisms (SNPs) in the interferon-λ4 (IFNL4) gene of the infected patients. No significant correlation was found between IFNL4 SNPs and RASs within NS3/NS5B in the present cohort. In contrast, the frequently detected NS5A RAS Y93H could be significantly associated with beneficial IFNL4 SNPs and a high baseline viral load in HCV genotype 1-infected patients. Taken together, the present study demonstrated that viral genome mutations as well as the morphology of secreted particles occur in a genotype-dependent pattern in HBeAg negative HBV infected patients with no need of antiviral therapy. As the amount of serum qHBsAg levels varied among the different genotypes, the HBsAg cut-off < 1000 IU/ml should be adapted individually among the various genotypes. Because the composition of the secreted subviral particles varied between the different genotypes, a genotype-specific immune-response might be induced in these patients. Additionally, the results of the present study indicate that in HBeAg negative HBV infected patients with mutations or deletions in the preS domain MHBs and LHBs might be expressed from the integrated DNA and therefore from a genetically different source than the released viral genomes. Aside from that, the finding of a significant association of the NS5A RAS Y93H with beneficial IFNL4 SNPs in chronic HCV infected patients may explain a lack of a correlation or an inverse correlation of treatment response with the IFNL4 genotype in some NS5A inhibitor-containing IFN-free regimens.