Positive selection signals of hepatitis B virus and their association with disease stages and viral genotypes

- We applied various methods to reveal positive selection signals in HBV genome.
- Positively selected codons are associated with disease stages and viral genotypes.
- Immune surveillance, anti-viral treatment, and viral replication fitness are driving forces of HBV adaptive evolution.
- Selections at nucleotide level are also identified to be associated with disease stage.
- A few mutations are under the process of fixation in viral adaptive evolution.

The hepatitis B virus (HBV) is a global health problem that causes different types of liver diseases. The high mutation rate of HBV, which results from a lack of proofreading activity of the viral polymerase, leads to the actively adaptive evolution of mutant strains under various selection pressures. This study focuses on the positive selection signals in the whole HBV genome and the association of these selection signals with the disease stages and/or viral genotypes. A total of 486 complete HBV genomes from HBV-infected individuals of different illness categories (i.e., acute, chronic, and severe hepatitis) were analyzed. To obtain a panoramic view of the selection signals, codon-based maximum likelihood analysis, three-dimensional (3D) mapping, and allele frequency comparison were conducted on genotypes B and C HBV from subjects with different stages of hepatitis. A total of 95 selected codons were resolved, and a significantly higher number of positive selection signatures were found in the chronic and severe hepatitis groups compared with the acute groups. Many of the selected codons were associated with either a unique disease stage or a specific genotype. The conservation analysis of the selection signals in the viral core protein (HBcAg) illustrated the occurrence of selected codons in the highly diversified regions. The allele-frequency-based analysis identified eight additional nucleotide substitutions, and the frequencies of these mutations were found to increase with disease progression. Moreover, we found that three substitutions, including A1762T, G1764A, and A2739C, were nearly fixed. The mapping of all of the selected codons and nucleotide substitutions to the functional domains of the viral proteins suggested that more than 60% of the mutations were subject to selection forces from host immune surveillance, antiviral therapy, and replication fitness.