A novel variant genotype C of hepatitis B virus identified in isolates from Australian Aborigines: complete genome sequence and phylogenetic relatedness

Hepatitis B virus (HBV) is an aetiological agent of acute and chronic liver disease including fatal fulminant hepatitis, cirrhosis and hepatocellular carcinoma, which is one of the most common human cancers and causes of death worldwide (Lee, 1997 ). It is estimated that there are currently more than 350 million carriers worldwide. There have been no studies on DNA sequences of HBV from Australian Aborigines, although the antigen known as the hepatitis B surface antigen (HBsAg) was discovered in Australian Aborigines (Blumberg et al., 1965 ). It is widely accepted that Australian Aborigines are a high-risk population for HBV infection. In contrast, a low prevalence of HBV marker is found in the non-Aboriginal population (Patterson et al., 1993 ; Campbell et al., 1989 ). Several studies have shown, by determining the prevalence of serological markers of HBV in different Aboriginal groups around mainland Australia and the Torres Strait islands (Holman et al., 1987 ; Campbell et al., 1989 ), that the prevalence of HBsAg varies from approximately 3% to as high as 35%. Most HBV infection in Aborigines occurs in early life, either by vertical transmission from a mother who is an HBsAg carrier, by horizontal transmission in households that include an HBsAg carrier or as a result of tattooing or intravenous drug use in young adults (Campbell et al., 1989 ).

The genotypic classification of HBV has been extended to seven genotypes, types AG, on the bass of S gene sequence and complete genome sequences (Norder et al., 1992 ; Okamoto et al., 1988 ; Stuyver et al., 2000 ). These genotypes reflect the geographical distribution of HBV. Genotype A predominates in north-western Europe, North America and Africa. Genotypes B and C are found in Asia. Genotype D is the most widespread worldwide and is the predominant genotype of the Mediterranean region. Genotype E is found in East Africa and genotype F is found mainly in the New World (Magnius & Norder, 1995 ; Norder et al., 1993 , 1994 ). Genotype G was found recently in isolates from America and France (Stuyver et al., 2000 ) but it has not yet been characterized. There is, however, no information regarding HBV genotype distribution in Oceania (consisting mainly of Australia). In this study, we analysed the complete genome sequences of HBV strains isolated from Australian Aborigines living mainly in Queensland and compared them with published sequences by molecular evolutionary analysis.

Samples.
Five serum samples positive for HBsAg were collected from Australian Aborigines from different parts of Queensland between 1989 and 1994 and used as the source of HBV DNA for sequencing. The clinicopathological features of these five patients are shown in Table 1. Two were positive for antibody to HBeAg (anti-HBe) but negative for HBeAg. There was no blood-transfusion history for any of the patients. From the results of liver biopsies, one patient was found to have liver cancer and three to have chronic hepatitis.

Table 1. Clinicopathological features of Australian Aborigines

Forty-eight HBV complete nucleotide sequences, representing each of genotypes AG, and six complete nucleotide sequences of HBV isolated from non-human primates (chimpanzee, gorilla, gibbon and woolly monkey) were obtained from international DNA databases (DDBJ/EMBL/GenBank). The accession numbers, genotypes, serotypes, countries of origin and references for genotype C and D strains are listed in Table 2. The most divergent strains in each basic genotype were included, the accession numbers of which were M57663 for genotype A, D00331 and M54923 for genotype B, X75665 and X75656 for genotype C and X65259 and X68292 for genotype D. Open reading frame analysis was performed with the same sequences as were used for comparison of the complete nucleotide sequence.

Table 2. HBV strains of genotypes C and D used in this study

HBV DNA extraction.
Serum samples were stored at -80 °C until assay. DNA was extracted from 100 µl of serum by using a DNA extractor kit (Sumitomo, Tokyo).

HBV DNA amplification.
The complete genome of HBV was amplified as two overlapping fragments by using the sense primer HB8F and antisense primer HB6R to yield a 3200 bp amplicon (fragment A) and the sense primer HB7F and antisense primer HB7R to yield a 462 bp amplicon (fragment B). For fragment A, nested PCR was performed to amplify 11 overlapping fragments by using the primers listed in Table 3. The primers used in this study were based on the most conserved regions derived from known sequences published in DDBJ/EMBL/GenBank. The amplification reaction was performed in a 96-well cycler (GeneAMP9600, Perkin-Elmer Cetus) and the PCR was initiated by the hot-start technique. The first round of PCR for fragment A was undertaken for 35 cycles (94 °C for 1 min, 55 °C for 1 min and 72 °C for 1·5 min) followed by an extension reaction at 72 °C for 5 min. The first round of PCR for fragment B and the second round of PCR for fragment A were performed for 30 cycles (94 °C for 1 min, 55 °C for 1 min and 72 °C for 1 min) followed by extension at 72 °C for 5 min. The PCR products were analysed by electrophoresis on 2·0% agarose gels, stained with ethidium bromide and visualized on a UV transilluminator.

Table 3. HBV DNA oligonucleotide primers used for full-length sequencing

Standard precautions to avoid contamination during PCR were taken. A negative control serum was also included in each run to ensure specificity.

Sequencing of the full-length sequences of HBV.
Twelve overlapping fragments of PCR-amplified HBV DNA were sequenced directly by the dideoxy method using the Taq Dye Deoxy Terminator cycle sequencing kit with a fluorescent 373A DNA sequencer (Applied Biosystems).

Serotyping for HBsAg.
Serotypes of samples from five Aborigines were deduced by using monoclonal antibodies directed against the a, d, y, w and r determinants of the surface antigen with the HBsAg subtype kit from the Institute of Immunology Co. (Tokyo, Japan) and confirmed on the basis of sequence data. The serotypes of the 54 HBV strains used for comparison were obtained from published articles or deduced from the registered sequence.

Phylogenetic analysis.
The nucleotide sequences of the five Australian Aboriginal HBV strains were compared with those of the 54 reference strains. Sequences were aligned by using the CLUSTAL W software and confirmed by visual inspection. Genetic distances were estimated by using the six-parameter method and phylogenetic trees were constructed by the neighbour-joining (NJ) method (Saitou & Nei, 1987 ). To confirm the reliability of the phylogenetic tree analysis, bootstrap resampling and reconstruction were carried out 1000 times. These analyses were carried out using the ODEN program of the National Institute of Genetics (Mishima, Japan) (Ina, 1994 ). The HBV genotype was assigned according to classification systems reported previously (Norder et al., 1992 ; Okamoto et al., 1988 ; Stuyver et al., 2000 ).

Genetic distances and phylogenetic relatedness
The five strains AustDF, AustKW, AustGC, AustRC and AustSJ were analysed and found to have nucleotide lengths of 3182, 3182, 3182, 3215 and 3194 bp. Phylogenetic analysis of the complete genome sequences of these five strains compared with those of 54 global HBV strains from DDBJ/EMBL/GenBank showed that three, AustDF, AustGC and AustKW, were classified into genotype D with bootstrap values of 100% (Fig. 1a). Furthermore, the three strains were shown to be most closely related in terms of evolutionary distance (0·01441, 0·01334 and 0·02613 substitutions per position, respectively) to a strain isolated from a healthy blood donor in Papua New Guinea (accession no. AB033559; Okamoto et al., 1988 ) that was reported previously. The remaining two strains, AustSJ and AustRC, were classified into a novel variant genotype C (mean nucleotide difference from genotype C 6·7%; range 5·97·4%) that has not been reported previously, with bootstrap values of 100% (Fig. 1a). The mean percentage nucleotide differences, obtained by pair-wise analysis, in the complete genome sequence between the two Australian HBV strains (AustSJ and AustRC) and all genotypes, including 48 HBV DNA sequences from DDBJ/EMBL/GenBank, are shown in Table 4.

(76K): Fig. 1. Phylogenetic tree analysis of the HBV strains isolated from five Australian Aborigines compared with 54 reference strains. Genetic distances were estimated by the six-parameter method and phylogenetic trees were constructed by the NJ method. The outgroup consisted of the Woolly Monkey strain (WMHBV). Accession numbers are indicated on the tree. Arrows show the Australian Aboriginal strains. Bootstrap values are shown along each main branch. The lengths of the horizontal bars indicate the number of nucleotide substitutions per site. The regions included in the analysis were: the complete genome (a), the ORF of the small-S gene (only the HBsAg region) (b), the ORF of the large-S gene (including preS1, preS2 and HBsAg) (c), the ORF of the P gene (d), the ORF of the Core gene (including the pre-core region) (e) and the ORF of the X gene (f).

Table 4. Mean percentage nucleotide differences for the small-S gene and complete nucleotide sequence between two Australian strains and other HBV genotypes

Based on the results of small-S gene analysis, AustSJ and AustRC were classified into a novel genotype that was separate from genotype C, with significant bootstrap values, similar to the relationship between genotypes A and D or A and G (Fig. 1b). The mean percentage nucleotide differences, obtained by pair-wise analysis, between the small-S genes of the two Australian HBV strains (AustSJ and AustRC) and all genotypes, including 48 HBV DNA sequences from DDBJ/EMBL/GenBank, are shown in Table 4. The mean percentage nucleotide differences between the Australian strains (AustSJ and AustRC) and strains belonging to the each of the seven genotypes were greater than 4%. To confirm the relationship, we also analysed the nucleotide sequences of the small-S genes of 91 HBV strains retrieved from DDBJ/EMBL/GenBank, including sequences of 81 strains of genotype C (references not shown), and another phylogenetic tree was constructed (Fig. 2). AustSJ and AustRC were again classified into a novel genotype distinctly separate from the genotype C group with significant bootstrap values.

(31K): Fig. 2. Phylogenetic analysis of the HBV strains isolated from two Australian Aborigines compared with 91 strains on the basis of small-S gene sequence, using the same methods as outlined in the legend to Fig. 1.

In the genotype D group, strains AustDF, AustGC and AustKW were again clustered with a strain isolated from Papua New Guinea (AB033559; Okamoto et al., 1988 ) based on small-S gene analysis (Fig. 1b). These analyses were also performed for each open reading frame, large-S (pre-S1 to S gene), P, Core and X (Fig. 1cf). The results were consistent with those obtained for complete genome sequence analysis. Based on analysis of the Core and X gene sequences (Fig. 1e, f), two Australian strains (AustSJ and AustRC), which were classified into a variant genotype C, were clustered with isolates from New Caledonia (X75665) and Polynesia (X75656) that had been reported previously (Norder et al., 1994 ).

Serotypic relatedness and characterization of the deduced amino acid sequence
The serotype of all five samples was ayw, deduced by using a serum reaction kit for HBsAg. Two of the five strains, however, AustSJ and AustRC, were classified further into the ayw3 serotype, based on the presence of Arg¹²², Lys¹⁶⁰ and Thr¹²⁷ residues determined by sequence data analysis. The three remaining strains, AustDF, AustGC and AustKW, were classified into the ayw1 serotype based on the presence of Arg¹²², Lys¹⁶⁰, Pro¹²⁷ and Tyr¹³⁴ residues (Fig. 3). There was no specific mutation in the a determinant, encompassing residues 124147 of HBsAg of the five Australian strains (Fig. 3). Within the pre-S1 region, an 11 amino acid deletion at codons 313, which is a specific deletion for genotype D, was found in AustDF, AustKW and AustGC, and a 7 amino acid deletion at codons 17 was found in AustRC, which was the pathogenic agent of a liver cancer (data not shown). Within the pre-S2 region, one substitution, of residue Thr⁴⁶, was found in the variant genotype C strain sequences and this was not present in the other genotype sequences (data not shown). Within the X region, three substitutions were found in the variant genotype C strain sequences that were not present in the other genotype sequences (data not shown). No specific substitutions were found in the pre-core and core regions (data not shown). Within the P region, the YMDD motif was conserved in all of the Australian strains, but it was shown that a Val⁵⁵⁵→Leu mutation occurred downstream of the YMDD motif, as in genotype F strains (data not shown).

(45K): Fig. 3. The amino acid sequences deduced from the small-S genes of the different HBV genotypes using 18 human HBV strains and one chimpanzee HBV strain. Dots indicate nucleotides identical to those of the genotype A sequence.

On the basis of phylogenetic analysis of the complete genome sequences of five HBV strains from Australian Aborigines and comparison with the complete genome sequences of global strains from international databases, a novel genotype C variant strain was identified that belonged to the ayw serotype and has not been reported previously in the genotype C group. Phylogenetically, based on nucleotide divergence, HBV strains are considered to be genotypically different if individual virus strains differ by more than 8% from other HBV strains over the complete genome sequence (Okamoto et al., 1988 ). We conclude that these strains belong to a novel variant genotype C (Variant C), based on analysis of the complete genome sequence, according to standard HBV genotyping. Interestingly, based on analysis of the small-S gene region in two Australian strains (AustSJ and AustRC), they were classified into a novel genotype that was distinctly separate from the genotype C group on the basis of sequence comparison, genetic distances and phylogenetic analysis performed using bootstrap values. Based on analysis of complete genome sequences, subtypes have been reported of the following genotypes: A (accession no. M57663), B (M54923 and D00331), C (X75665 and X75656) and D (X65259 and X68292). Based on small-S gene sequences, however, the nucleotide divergence of these individual virus strains between each basic genomic group was less than 4%. The fact that, in this study, the genetic distance was more than 4% from the basic group based on small-S gene analysis is important when the evolutionary process, virus pathogenicity and clinical outcome are considered. It should be noted that the patient who was infected with this variant genotype C virus strain had liver cancer. Further investigation is necessary to determine the relationship between variant genotype C and clinical characteristics.

The genotypic classification of HBV is likely to correlate with the geographical origin of strains. Genotype D, although it has been found to be the most widespread strain, predominates in the Mediterranean region. This genotype was also found in parts of Asia and in aboriginal populations in Papua New Guinea. In our study, three (AustGC, AustDF and AustKW) of five strains isolated from Australian Aborigines were classified into genotype D. Furthermore, they were clustered with a strain isolated from Papua New Guinea (AB033559) with closer genetic distances, based on the complete genome sequence and ORF analysis. Australia was first settled from South-east Asia more than 40000 years ago (Roberts et al., 1990 ) in a migration to the ancient continent of Sahul, which comprised Australia and Papua New Guinea (White & OConnell, 1982 ). Seven thousand years ago, Papua New Guinea and Australia became islands separated by the Torres Strait. Analysis of α-globin haplotypes indicated a close association between Australian Aborigines and Papua New Guinea highlanders (Roberts-Thomson et al., 1996 ). In this study, a common ancestral source population of these aboriginal populations was indicated on the basis of molecular evolutionary analysis of HBV. Further investigations of the route of infection of HBV using phylogenetic analysis should lead to very interesting results regarding the affinity between Australian Aborigines and Papua New Guinea populations. On the other hand, AustSJ and AustRC, which were classified into a genotype C variant, clustered with isolates from New Caledonia (X75665) and Polynesia (X75656), exhibiting close genetic distances based on analysis of the X and Core genes. Further investigation is required to determine the relationship between these Polynesian peoples and Australian Aborigines.

The amino acid residues specifying d/y and w/r have been shown to be present at positions 122 and 160 of HBsAg (Okamoto et al., 1988 ). Nine serological groups have been designated: adw2, adw4, adr, adrq, ayw1, ayw2, ayw3, ayw4 and ayr (Couroucé et al., 1976 ; Couroucé-Pauty et al., 1978 ). It has been shown that these nine different HBsAg serotypes may belong to either one or several HBV genotypes (Magnius & Norder, 1995 ; Norder et al., 1993 ). Most ayw serotypes are grouped in genotype D. In the present study, however, we identified two ayw strains isolated from Australian Aborigines that were grouped in a variant genotype C. There have been no previous reports on the sequence variability of the S gene for strains of genotype C with the ayw serotype. The serotype of the strains isolated from Australian Aborigines was ayw; however, studies regarding the serotypic distribution of HBV using a large number of subjects in Australian Aborigines may need to consider that patients with variant genotype C or genotype D of ayw serotype may present different clinical and virological features. The serotypic subtype of HBsAg determined by the serological reaction of the product of the HBV S gene does not always agree with molecular evolutionary classification at the gene level. Further investigation is required, since these strains occur not only in Australian Aborigines but in other parts of the world.

In conclusion, we report for the first time interesting characteristics of the complete genome sequences of HBV strains isolated from the HBsAg-positive serum of Australian Aborigines. These results will contribute to the investigation of the worldwide spread of HBV, the relationship between serotype and genotype and the ancient common origin of aboriginal Australians. It will be interesting to try to understand the migration of Australian Aborigines by analysing the spread of HBV.

This study was supported by grants from the Ministry of Education, Science and Culture of Japan (11691222).

Footnotes

The nucleotide sequence data reported in this paper will appear in the DDBJ/EMBL/GenBank databases under the accession numbers AB048701AB048705.