Viral Genetics of HBVand HDV

Despite an effective vaccine, infection with hepatitis B virus (HBV) is mainly transmitted through the mother-to-neonate route in endemic countries. In most cases, chronic infection results and the transmission will therefore occur from generation to generation. HBV could be considered as an indirect marker of population migration. Transmitted from the mother, it might be considered as an alternate to mitochondrial DNA (mtDNA) (Ingman et al. 2000). MtDNA analyses indicate that around 59,000-100,000 years before present, earth colonization might have occurred from Africa to the Middle East, Asia then to Australia, Europe and to Americas. Several hypothesis have been proposed to link the HBV to human evolution and dispersal around the globe (reviewed in Simmonds 2001). (1) The existence of Hepadnaviridae in other primates, mammals and birds makes possible a co-speciation of HBVs during evolution. (2) The higher divergence of the South American HBV genotype F strains (HBV/F) could have given rise to HBV spreading from this area to the rest of the world during the slave trade. (3) The existence of primate-specific strains might also explain a cross species contamination such as in HIV. These hypotheses are not necessarily mutually exclusive, and each could contribute, in part, to the present day distribution of HBV among humans around the world. Regardless of the origin and evolution of HBV, nucleotide similarity approaches and evolutionary reconstructions indicate the existence of at least eight HBV genotypes (labelled HBV/A to HBV/H), the existence of many subtypes and specific gene phylogenies identify recombinant forms (Norder et al. 2004).

Hepatitis delta virus (HDV) was identified in 1977 as a foreign antigen in the serum and the liver of Italian patients infected with HBV (Rizzetto et al. 1977). The origin of HDV remains difficult to understand and the age of the HDV-HBV association needs to be clarified. This viral-like agent has been classified as a satellite of HBV, being dependant on HBV for virion assembly and propagation (Sureau et al. 1993), but not for HDV-RNA replication. The HDV genome is a circular single-stranded RNA of 1672-1697 bases with extensive intramolecular complementarity (Wang et al. 1986; Radjef et al. 2004). Part of the HDV genome might share historical homology with viroids or plant virus satellite RNA sequences (Elena et al. 1991). Interestingly, a pseudoknot ribozyme is evidenced in both genomic and antigenomic RNA strands, corresponding to the best conserved parts of the genome. However, due to a low degree of similarity between viroids and the HDV genome, this ancestral vi-roid affiliation is disputed (Jenkins et al. 2000). Furthermore, viroids are not known to code any protein, while HDV does. A double rolling circle model is involved for viral RNA replication (reviewed in the chapters by J.M. Taylor and by T.B. Macnaughton and M.M.C. Lai, this volume). The lack of fidelity and the absence of a proofreading activity of RNA polymerases give rise to heterogeneous molecules that can be considered as quasispecies, as has been described for the RNA virus world (Eigen and Biebricher 1988).

Indeed, the original paper (Wang et al. 1986) describing a full-length HDV genome sequence derived following experimental transmission to chimpanzees (Chimp A20), indicates that the RNA sequence from an HDV strain, derived from a chronically infected Italian patient-isolate, was heterogeneous. In fact, the 1679 nucleotide long sequence showed 17 ambiguous positions, mostly consisting of transitions (n=16). From the same strain successfully transmitted to the woodchuck model (where delta ribonucleoprotein is budding through the woodchuck hepatitis virus surface antigen envelope), six positions were heterogeneous in the W5 isolate (Deny et al. 1991). All the viruses characterized from this lineage show a nucleotide sequence with global similarity higher than 98% during these experimental transmissions (Wang et al. 1986; Kuo et al. 1988; Deny et al. 1991; Kos et al. 1991). This heterogeneity and evolution rate has been particularly well studied on the HDV coding gene (Imazeki et al. 1990), 1068 bases (Zhang and Hansson 1996) and on the complete genome (Lee et al. 1992; Chao et al. 1994). Lee et al. described the long-term evolution of a Lebanese HDV strain. The evidence of an acute delta

Table 1 Similarity percentage (%) for the sHD gene within (bold) and between the eight HDV major clades: mean (range)







80. (79.




78. (75.

































cg ag cagtgg gt g ag c ggaag--------------agctcgagagggaac

Fig. 1 Detection of deleted HDV genome in the serum by the use of amplification (1) and primer extension (2). 2a: Successive samples obtained from a patient before (lanes 1-4) and after (lane 5) Interferon treatment. 2b: Successive follow-up (lanes 1 and 2) of HDV in a patient harbouring two forms of deleted HDV molecules. 3: Sequence analyses indicate the 13 to 14 base deletions from the respective strains hepatitis followed by a chronic course of the infection with flares, gave the opportunity to compare the evolutionary rates of the genome during the different phases. The acute/flare phase was associated with a rapidly evolving rate of 3.0x10-2 substitutions/nucleotide/year, whereas the chronic phase seemed to be less dynamic-3.0x10-3 substitutions/nucleotide/year. Obviously, such different rates of evolution could make the hypothesis of a 'molecular clock' not as easily suitable for HDV as for other viruses (Simmonds 2001). Furthermore, treatment-induced environmental constraints also accelerate such evolutionary rates. For example, in the case of a chronically infected patient by HDV type 1, a substantial modification of the viral genome occurred with the appearance of naturally deleted HDV-RNA molecules that were detected in the plasma (Fig. 1) (Deny 1994). Under high dose alpha-interferon, both defective and natural molecules disappeared, but a resistant strain emerged (without the defective molecule) (Fig. 1) indicating that a relapse occurred under interferon treatment. In such an environment the mutation rate of the HDV coding sequence before and after treatment was 1.09x10-2 substitutions per site per year, which was higher than observed in natural infection for the delta protein coding gene: i.e., 1.14x10-3 to 1.28x10-3 mutations/ nucleotide/year (Imazeki et al. 1990; corrigendum in Lee et al. 1992) or 2.60x10-3 muta-tions/nucleotide/year (Chao et al. 1994). Interestingly, a recent paper from J.C. Wu et al., confirmed the existence of such deletions in four out of five patients chronically infected with HDV-1 or HDV-2 (Wu et al. 2005).

The evidence of mixed HDV infection has been described in Taiwan (Wu et al. 1999) and HDV genome recombination is another evolutionary pathway recently observed (Wang and Chao 2005). T.C. Wang and M. Chao described a patient who was infected with both a genotype I (HDV-1) strain and a genotype IIb (HDV-4) strain. Using a wide range of techniques, they found evidence that HDV homologous recombination occurred in 6% of clones and mapped the recombination events within homologous regions from those two strains.

0 0

Post a comment