Molecular epidemiology of hepatitis A and hepatitis B virus in central America

Abstract

The genetic variability was determined for hepatitis A virus (HAV) and hepatitis B virus (HBV) strains from cases of acute and chronic hepatitis in Central America (CA). The genetic relatedness was established for the' strains to strains from other parts of the world.

Molecular typing of HAV was performed by sequence analysis of the 5'-end of the VP1 region. HAV strains from CA were found to belong to genotype 1A. The strains were mostly similar to each other, and unrelated to subtype 1 A strains from other parts of the world. Based on their amino acid sequence four HAV strains, all related to strain CR326 sampled in Costa Rica in the 60s, were found to have been circulating in the area during the last three decades. However, on basis of nucleotide variability the isolates from two common source outbreaks in Costa Rica could be distinguished from the isolates of background sporadic cases.

The genotyping of HBV was accomplished by partial sequencing of the S gene, followed by comparison of the predicted amino acid substitutions of the obtained sequences with the substitutions known conserved for each genotype. Genotype F was found in 79% of the strains and genotypes A, D and C were represented by 14%, 6% and 1 % of the strains, respectively. Phylogenetic analysis of the small S gene showed that genotype A strains from CA belonged to two different clusters, one including strains from Europe and USA, while one comprised strains with Afro-Asian origin. One genotype C strain was related to strains from South East Asia. Interestingly, the genotype D strains were related to strains from intravenous drug addicts in Europe. Within genotype F three genetic clusters were initialle disclosed. The major cluster was formed by 18 strains from CA and one strain from Alaska. A second cluster included two strains from CA and six from South America and Europe. Two Nicaraguan strains formed the third cluster, which were that divergent that they might represent a new genotype. The deduced amino acid sequence of the small S showed that all but one of the genotype F strains from CA encoded subtype adw4. One Nicaraguan strain encoded subtype ayw4, which represented the first report of this subtype within genotype F. Noteworthy, the two subgrupes of genotype F strains correlated with the amino acid substitution at position 45, which were Thr and Leu for the Central American and South American subgroups, respectively.

The presence of pre-core stop mutations was investigated in 19 HBV strains derived from anti-HBe positive sera. The mutation A 1896 was found 11 out of 17 genotype F strains. Since the pre-core sequence in published genotype F genomes have C 1858, known to prevent the emergence of the pre-core stop mutation, position 1858 was also examined. All but one of the 17 genotype F strains had T 1858. This substitution was also found in five genotype F strains from HBeAg positive sera. The substitutions at position 1858 were found to strictly correlate with the two mentioned subgroups of genotype F.

Analysis of the complete genomes of the divergent Nicaraguan strains and one strain from California also encoding adw4, showed nucleotide divergences from 0.8-2.5 % from each other, 7.2- 10. 1 % from genotype F and 13.2-15.7% from other HBV genotypes. Since pairwise comparisons of 82 complete HBV genomes showed intra and inter-typic nucleotide differences from 0.1-7.1% and 6.8-17.1%, respectively, these three strains were found to represent a new H13V genotype, that was designated H. When comparing the primary structures of the gene products for the three strains with genotype F strains several differences were found, the majority of them in the polymerase region. At phylogenetic analysis of complete genomes and subgenomic regions the three strains always clustered together in a distinct branch from genotype F supported by 100% bootstrap. Apart from Nicaragua and California, this genotype has now also been found in Mexico.

The complete genomes were also sequenced for two HBV strains, one from Spain and one from Sweden, classified as genotype D based on their S gene sequences, although encoding d-specificity. The two strains were confirmed to belong to genotype D, and differed from other genotype D strains by 2.8-4.6%. However, their S genes encoded Lys 122, Thr 127 and Arg 160, corresponding to the putative new subtype adw3. This is also the first report on strains encoding d-specificity within genotype D known to encode ayw2, ayw3 and rarely ayw4. In conclusion HAV strains circulating in CA were found to belong to the worldwide-distributed genotype IA. Limited sequencing within the VP1 region proved useful to identify common source outbreaks of hepatitis A also in this highly endemic area, where most strains are local and only one subtype prevails.

We also found that the Amerindian genotype F of HBV predominates among Central American populations of Hispanic background, and that it may carry the pre-core stop mutant, which was not anticipated considering previously available sequence information. A new HBV genotype, designated H, was found in CA, also known to be present in Mexico and the California. Genotype H was most similar to F and has most likely both split off from the common ancestor within the New World.