Selecting a genetic region for molecular analysis of hepatitis B virus transmission.

The optimal genetic regions for the phylogenetic analysis of hepatitis B virus (HBV) transmission continue to be a matter of debate, with different investigators preferring different regions. However, full-length HBV sequence analysis is the gold standard for the purpose. But in developing countries, such as ours, where HBV infections are endemic, the added cost of full-length sequencing becomes a limiting factor for studying large numbers of samples. Thus, the search for a suitable genetic region of HBV is important. Initially clustering of HBV seromarkers (7, 12) and subsequently clustering of well-established mutations (e.g., HBsAgG145R or HBeAgW28Stop) were used to demonstrate intrafamilial transmission (1, 9, 11, 13, 15, 18, 19). Recently, to increase the confidence level of detection of true transmission events, phylogenetic analysis with the bootstrap resampling/ maximum-likelihood test of surface/precore (preC)/core region sequences was carried out (6, 8, 22). However, clustering of sequences from epidemiologically unrelated families was suggested to be due to a high degree of conservation of surface (S) gene sequence (8), which led to the region being considered unsuitable for transmission studies. Thus, analysis of nonoverlapping, fast-evolving regions was recommended (3, 5). Interestingly, 67% of the HBV genome is overlapping (16), leaving distal X/preC/partial core regions nonoverlapped. These regions encode important RNA structural elements, such as the epsilon signal (10), that are essential for HBV replication. Thus, one can assume that high variability in these regions might have negative selection pressure; on the other hand, variability in the HBsAg is positively selected to evade host immune pressure. Recently, using statistical models, Szmaragd and colleagues (17) found that overlapping sites have slightly higher substitution rates than nonoverlapping regions, which supports the above assumption. Further, variability of a genetic region or prevalence of certain mutations varies with the study population, infecting genotype (10), immune status, chronicity (20), clinical outcome (21), duration (associated time frame), and mode of infection (nosocomial, vertical, or intrafamilial horizontal). Thus, as recommended by Bracho and colleagues (3) for investigating the chain of recent/nosocomial fulminant cases, analysis of highly variable preC/core sequence associated with fulminant hepatitis B (14) should be preferred. Our population shows a low preC mutation prevalence (2), and S gene sequence analysis provided more phylogenetic signal, which is more appropriate for tracing horizontal transmission patterns among HBsAg-negative family members (6). High S-gene variability has been documented in previous studies among HBsAg-negative subjects (20) or among chronic virus carriers or their families (18). In fact, different specific variability levels for the S gene (genotype, subgenotype, and subtype), in addition to mutations, can provide enough confidence to prove transmission events. Actually, in one of our earlier studies, where the preC/core region of all the isolates was identical, genotype, subtype, and mutation analysis of S gene sequences proved two different sources and evidence of horizontal transmission of HBV infection in a family (4). Thus, before selecting a genetic region for investigations of transmission, it is more reasonable to consider the genetic variability of HBV among the study population, their serological profile, and the mode of probable transmission rather than to adhere to a specific genetic region, found to be more variable in a study of different population groups with different HBV genotypes, disease severity, or serological patterns.