Epidemic history and evolutionary dynamics of hepatitis B virus infection in two remote communities in rural Nigeria

Abstract

Background: In Nigeria, hepatitis B virus (HBV) infection has reached hyperendemic levels and its nature and origin have been described as a puzzle. In this study, we investigated the molecular epidemiology and epidemic history of HBV infection in two semi-isolated rural communities in North/Central Nigeria. It was expected that only a few, if any, HBV strains could have been introduced and effectively transmitted among these residents, reflecting limited contacts of these communities with the general population in the country.

Methods and findings: Despite remoteness and isolation, approximately 11% of the entire population in these communities was HBV-DNA seropositive. Analyses of the S-gene sequences obtained from 55 HBV-seropositive individuals showed the circulation of 37 distinct HBV variants. These HBV isolates belong predominantly to genotype E (HBV/E) (n=53, 96.4%), with only 2 classified as sub-genotype A3 (HBV/A3). Phylogenetic analysis showed extensive intermixing between HBV/E variants identified in these communities and different countries in Africa. Quasispecies analysis of 22 HBV/E strains using end-point limiting-dilution real-time PCR, sequencing and median joining networks showed extensive intra-host heterogeneity and inter-host variant sharing. To investigate events that resulted in such remarkable HBV/E diversity, HBV full-size genome sequences were obtained from 47 HBV/E infected persons and P gene was subjected to Bayesian coalescent analysis. The time to the most recent common ancestor (tMRCA) for these HBV/E variants was estimated to be year 1952 (95% highest posterior density (95% HPD): 1927-1970). Using additional HBV/E sequences from other African countries, the tMRCA was estimated to be year 1948 (95% HPD: 1924-1966), indicating that HBV/E in these remote communities has a similar time of origin with multiple HBV/E variants broadly circulating in West/Central Africa. Phylogenetic analysis and statistical neutrality tests suggested rapid HBV/E population expansion. Additionally, skyline plot analysis showed an increase in the size of the HBV/E-infected population over the last approximately 30-40 years.

Conclusions: Our data suggest a massive introduction and relatively recent HBV/E expansion in the human population in Africa. Collectively, these data show a significant shift in the HBV/E epidemic dynamics in Africa over the last century.

Figure 1. Phylogenetic tree of the partial S-gene sequences (378bp) from 55 Nigerian HBV isolates identified in this study (marked •) and sequences recovered from GenBank (not marked).

Sequences retrieved from GenBank are denoted by their accession numbers and the source country of the isolates. Bootstrap values of major branches are shown.

Figure 2. Phylogenetic tree of the 47 HBV/E and 2-HBV/A3 whole genome sequences from Nigeria (marked •) and GenBank references (not marked).

Sequences retrieved from GenBank are denoted by their accession numbers and the source country of the isolates. Bootstrap values of major branches are shown.

Figure 3. Median joining network of the intra-host S-gene sequence variants identified in 22 individuals infected with HBV/E.

Each node represents a single sequence variant. Each color represents a single individual. The size of the node reflects frequency of the corresponding variant in the population.

Figure 4. Phylogenetic tree of HBV sequence variants identified in 5 individuals with mixed genotype infections.

Sequences identified in this study are colored. Reference sequences retrieved from GenBank for genotypes HBV/E, HBV/G and HBV/D are shown in black. The consensus S-gene sequences for each individual are indicated with white arrows. All sequences that obtained from a single individual are shown using same color.

Figure 5. Network of individuals sharing identical sequences.

Each node represents an individual and the link connects individuals that share at least one HBV variant. Blue nodes belong to village 1 (light for females, dark for males) and yellow nodes belong to village 2 (light for females, dark for males).

Figure 6. Median joining network of the intra-host S-gene sequence variants identified in 2 individuals infected with HBV/A3.

Each node represents a single sequence variant. Each color represents a single individual. The size of the node reflects frequency of the corresponding variant in the population.

Figure 7. Bayesian skyline plot showing the epidemic history estimated from the Nigerian HBV genotype E dataset.

The middle line is the median estimate of effective population size (log10) and the grey area shows the 95% highest posterior density intervals for this estimate. The most recent time is the time of collection for the most recently collected sequences (2007) and the oldest time shown is the lower limit of the 95% highest posterior probability density for the root height for all the sequences (1959).