Identification of Hepatitis B virus putative intergenotype recombinants by using fragment typing.

Abstract

Eight hundred and thirty-seven human Hepatitis B virus (HBV) genomes were categorized into pure genotypes and potential intergenotypes, according to their fragment types which were determined based on similarity and phylogenetic analyses of 13 contrived fragments of 250 bp against the corresponding fragments of the consensus sequences of genotypes A-H. Twenty-five intergenotypes, including 171 genomes, were revealed from the potential intergenotype recombinants by phylogenetic analysis of the precisely derived mosaic fragments. Among these, four new intergenotypes were discovered. Many genomes were revealed as putative intergenotype recombinants for the first time. About 87 % of the putative recombinants were B/C (120) and A/D (29) hybrids. The other recombinants comprised A/B/C, A/C, A/E, A/G, C/D, C/F, C/G, C/U (U for unknown genotype) and B/C/U hybrids. Genotypes A and C showed a higher recombination tendency than did other genotypes. The results also demonstrated region priority and breakpoint hot spots in the intergenotype recombination. Recombination breakpoints were found to be concentrated mainly in the vicinity of the DR1 region (nt 1640-1900), the pre S1/S2 region (nt 3150-100), the 3'-end of the C gene (nt 2330-2450) and the 3'-end of the S gene (nt 650-830). These results support the suggestion that intergenotype recombinants may result from co-infection with different genotypes.

Full text

Page 1

Identification of Hepatitis B virus putative
intergenotype recombinants by using fragment
typing
Jie Yang,1,2Ke Xing,1Riqiang Deng,1Jinwen Wang1
and Xunzhang Wang1
Correspondence
Xunzhang Wang
wxz@zsu.edu.cn
1State Key Laboratory for Biocontrol, School of Life Science, Sun Yat-sen (Zhongshan)
University, Guangzhou 510275, People’s Republic of China
2Department of Infectious Diseases, Nanfang Hospital, Guangzhou 510515, People’s Republic
of China
Received 9 December 2005
Accepted 30 March 2006
Eight hundred and thirty-seven human Hepatitis B virus (HBV) genomes were categorized into
pure genotypes and potential intergenotypes, according to their fragment types which were
determined based on similarity and phylogenetic analyses of 13 contrived fragments of 250 bp
against the corresponding fragments of the consensus sequences of genotypes A–H. Twenty-five
intergenotypes, including 171 genomes, were revealed from the potential intergenotype
recombinants by phylogenetic analysis of the precisely derived mosaic fragments. Among these,
four new intergenotypes were discovered. Many genomes were revealed as putative intergenotype
recombinants for the first time. About 87% of the putative recombinants were B/C (120) and
A/D (29) hybrids. The other recombinants comprised A/B/C, A/C, A/E, A/G, C/D, C/F, C/G,
C/U (U for unknown genotype) and B/C/U hybrids. Genotypes A and C showed a higher
recombination tendency than did other genotypes. The results also demonstrated region priority
and breakpoint hot spots in the intergenotype recombination. Recombination breakpoints were
found to be concentrated mainly in the vicinity of the DR1 region (nt 1640–1900), the pre S1/S2
region (nt 3150–100), the 39-end of the C gene (nt 2330–2450) and the 39-end of the S gene
(nt 650–830). These results support the suggestion that intergenotype recombinants may
result from co-infection with different genotypes.
INTRODUCTION
Human Hepatitis B virus (HBV), which is the prototype
member of the family Hepadnaviridae, is a circular, partially
double-stranded DNA virus of approximately 3200 bp with
four overlapping ORFs encoding the polymerase (P), core
(C), surface (S) and X proteins (Lee, 1997). Based on an
intergroup divergence of 8% or more in the complete
nucleotide sequence, HBV sequences have been classified
into genotypes A–F (Okamoto et al., 1988; Norder et al.,
1992, 1994; Magnius & Norder, 1995). Genotypes G and H
were suggested later (Stuyver et al., 2000; Arauz-Ruiz et al.,
2002). Generally, different HBV genotypes have distinct
geographicaldistributions.Forexample,genotype Aisprev-
alent in northern and central Europe and is also found in
North and South America and Africa. Genotypes B and C
are common in south and east Asia. Genotype D spreads
worldwide, but predominates in the Mediterranean area.
Genotype E is found mainly in the western part of sub-
Saharan Africa, while genotype F is found in aboriginal
populations of the Americas (Norder et al., 1993). Genotype
G is found in USA and France (Stuyver et al., 2000).
GenotypeHismainlyrecordedinNorthandSouthAmerica
(Arauz-Ruiz et al., 2002). It has been assumed that the HBV
genotype may influence the biology of HBV and clinical
disease in hosts (Mayerat et al., 1999). Genotype C induces
more severe liver disease than genotype B in Asia (Bowyer
et al., 1997; Kao et al., 2000), while genotype A is associated
with chronic infection more frequently than genotype D in
Europe (Mayerat et al., 1999).
HBV has a high rate of diversity. The mutation rate of the
genomeismuchhigherthaninotherDNAviruses(Okamoto
et al., 1987; Georgi-Geisberger et al., 1992) because HBV
replicates similarly to retroviruses through reverse tran-
scription of an RNA intermediate, which lacks a proof-
reading function (Summers & Mason, 1982). Homologous
recombination between different genotypes is a source of
variation. Accumulating data show that DNA recombina-
tion is frequent in HBV infections. Recombination between
HBV genotypes B and C (B/C hybrid) is prevalent in south
and east Asia (Morozov et al., 2000; Sugauchi et al., 2002,
2003; Luo et al., 2004). A/D hybrids have been described in
Italy and South Africa (Morozov et al., 2000; Owiredu et al.,
2001). Other HBV intergenotype hybrids have also been
0008-1752 G 2006 SGMPrinted in Great Britain2203
Journal of General Virology (2006), 87, 2203–2215
DOI 10.1099/vir.0.81752-0

Page 2

discovered, such as C/D hybrids in China (Cui et al., 2002;
Wang et al., 2005), an A/E hybrid in Cameroon (Kurbanov
et al., 2005) and a C/G hybrid in Thailand (Suwannakarn
et al., 2005). Intergenotype recombination may lead to
different biological characteristics of the virus and a differ-
ent clinical outcome in patients. For instance, recombina-
tion of genotype B with genotype C (B/C hybrid) might
affect the severity of clinical disease (Sugauchi et al., 2002).
Then, many questions are raised: how many kinds of inter-
genotypic recombinants exist in published HBV data; where
are the recombination breakpoints located; and are there
any rules for the recombination? These questions may lead
to an understanding of the mechanism of HBV intergeno-
type recombination. In this paper we report 25 intergeno-
types comprising 171 putative intergenotype recombinants
found in 837 complete HBV genome sequences.
METHODS
HBV data. Complete human HBV genome sequences were collected
and updated every month from GenBank, EMBL and DDBJ by
BLASTN searches using the complete HBV genome sequences of geno-
types A–H. Unique accession numbers were collected by eliminating
duplicates in a text file. Sequences were then downloaded from
GenBank using the Batch Entrez tool on the NCBI (National Center
for Biotechnology Information) website (www.ncbi.nlm.nih.gov).
After removing identical entries, patents, artificial mutants, partial
sequences and animal HBV sequences, genome sequences with
lengths between 3000 and 3300 bp were changed so that they started
with the traditional hypothetical EcoRI site. Finally, 837 genomes
consisting of 146, 152, 292, 142, 40, 42, 13 and 10 of genotypes
A–H, respectively, were obtained in March 2006 (see Table 2 for a
complete list of accession numbers).
Creation of consensus sequences for genotypes A–H. All the
genome sequences were classified into genotypes A–H by using
BLASTN (Altschul et al., 1990) and phylogenetic analysis against 24
representatives of genotypes A–H which were selected at random,
three for each genotype, from HBV full-length sequences of geno-
types A–H, according to Norder et al. (2004). Consensus sequences
of all genotypes except genotype B were then created from all the
full-length sequences of the corresponding genotypes. Genotype B
consists of subgenotypes Bj (j for Japan) and Ba (a for Asia)
(Sugauchi et al., 2002). Because the majority of genotype B sequences
are of subgenotype Ba, which is considered to be a B/C hybrid with
genotype C clustered in the preC/C region, the consensus sequence
for genotype B was created using full-length subgenotype Bj sequences.
Primary fragment typing. Consensus sequences of genotypes A–H
were aligned using CLUSTAL W (Thompson et al., 1994). Starting
with the traditional hypothetical EcoRI site, the alignment was split
into 13 subalignments of 250 bp (the last one 257 bp) and the frag-
ments from the 13 subalignments were formatted as a database for
BLASTN. Genotypes of the 13 fragments of each genome (called ‘frag-
ment types’ in this report) were determined by BLASTN of the 837
complete HBV genome sequences against this database. According
to the BLASTN result, each genome was interpreted with a string of
13 letters (from A to H), representing the types of the corresponding
fragments. Genomes of the pure genotypes are denoted with 13 sym-
bols of only one letter, for example AAAAAAAAAAAAA for genotype
A. Genomes of potential intergenotype recombinants are therefore
denoted with 13 symbols of different letters, for example ADADAA-
AAAAAAD, letters that differ from the dominant letter implying
potential mosaic fragments in the genome. A Perl script was written
to analyse and record the fragment types of 837 HBV complete
genome sequences.
Calibration of the fragment types. Genomes of pure fragment
type were considered to have little probability of being intergenotype
recombinants and needed no further consideration. Only genomes
of impure fragment type were probably recombinants. Therefore,
potential mosaic fragments from impure fragment type genomes
were collected according to their position with respect to each
250 bp region and aligned with correlated fragments from represen-
tative genomes with minor manual adjustments. The alignments
were fed into the PHYLIP software package, version 3.62 (Felsenstein,
1993). Genetic distances were estimated by the Kimura two-parameter
model and phylogenetic trees were reconstructed by the neighbour-
joining method; the reliability of topologies was estimated by
performing bootstrap resampling and reconstruction with 1000
replicates, then the CONSENSE program in the PHYLIP package was
used to compute major-rule consensus trees. Potential mosaic frag-
ments were retyped according to the phylogenetic trees. Fragments
not clustered with any of the A–H genotypes were retyped with ‘N’,
meaning uncertain, and the fragments with a deletion longer than
125 bp were marked with ‘-’. So the fragment types of some geno-
mes were modified, changing some originally impure fragment type
genomes to pure genotypes and others to different impure fragment
types.
Identification of putative recombinants. For the genomes of
impure fragment types, recombination breakpoints were determined
with the SimPlot program (Lole et al., 1999) and by bootscanning
analysis, according to Robertson et al. (1995), Lole et al. (1999) and
Sugauchi et al. (2002). The SimPlot program version 3.5.1 was
obtained from http://sray.med.som.jhmi.edu/SCRoftware/ and was
used to identify phylogenetically informative sites supporting alter-
native tree topologies. This was performed by considering four
sequences at a time: one putative recombinant sequence, two con-
sensus sequences of the parental genotypes and one consensus
sequence of another genotype as an outgroup. Each informative site
supports one of three possible phylogenetic relationships among the
four taxa. Bootscanning and cluster analysis maximizing x2were
used to identify the breakpoints. P values for the subsequent division
of the sequence into genotypes were calculated by using Fisher’s
exact test. After determination of recombination breakpoints,
mosaic fragments were derived precisely from the recombinants,
according to the breakpoints. Mosaic fragments with the same
breakpoints were collected and aligned with corresponding frag-
ments of genotype A–H representatives. Phylogenetic trees were
reconstructed based on these alignments and compared with the
trees reconstructed based on the correlated complete genome
sequences. A 75% bootstrap value cut-off was used to identify the
mosaicism of each fragment.
RESULTS
HBV genomes: classification and creation of
consensus sequences
Genotypes of most HBV genomes could be determined by
using BLASTN against the 24 representatives of genotypes
A–H shown in Table 1 and by analysis of the BLASTN
results based on a nucleotide sequence identity cut-off
(>92%). Others required further phylogenetic analysis. Of
the 837 HBV genomes, 146, 152, 292, 142, 40, 42, 13 and 10
were categorized in genotypes A–H, respectively. Of the 837
HBV genomes, 631 were found to have a standard genotype
2204Journal of General Virology 87
J. Yang and others

Page 3

length, including 96, 128, 215, 105, 33, 37, 9 and 8 genomes
for genotypes A–H, respectively, and these were chosen for
consensus sequence creation. Consensus sequences for
genotypes except genotype B were then created from
alignments of genome sequences of the corresponding
genotypes. For the reason explained previously, the
consensus sequence of genotype B was obtained by the
alignment of 28 Bj subgenotype sequences which were
selected according to Sugauchi et al. (2002). The resulting
consensus sequences were used as reference sequences for
genotypes A–H in the subsequent fragment typing and
SimPlot bootscanning analyses.
Potential intergenotype identification using
fragment typing
Using BLASTN against the consensus fragments which were
derived by splitting the alignment of consensus genome
sequences into 13 subalignments of 250 bp, 633 genomes
were identified to be pure genotypes A–H, and 204 to be
potential intergenotypes, implying intergenotype recombi-
nants. Fragment types of the 204 potential intergenotype
genomes were determined further by phylogenetic analysis
of potential mosaic fragments. Phylogenetic trees were
reconstructed based on alignments of the potential mosaic
fragments together with correlated fragments from repre-
sentative genomes. Mosaic fragments that were not
supported by the phylogenetic trees were modified accord-
ing to the phylogenetic results, changing some originally
impure fragment type genomes to pure types and others to
differentfragmenttypes.Thefragmenttypesofthe837HBV
genome sequences determined with the above methods are
shown in Table 2.
Identification of putative recombinants
By SimPlot analysis, the recombination breakpoints of the
potential recombinants were determined precisely. Mosaic
fragments were derived from the intergenotype recombi-
nants according to the breakpoints. A phylogenetic tree was
inferred to determine further the genotype of each precisely
derived mosaic fragment using a 75% bootstrap value cut-
off. For example, eight genomes with the fragment type
‘ADADAAAAAAAAD’ and four genomes with the fragment
type ‘DDADAAAAAAAN-’ belonged to genotype A in the
whole genome, and all had a mosaic fragment between nt
209and526. Aphylogenetictreewastheninferred,basedon
the alignment of these mosaic fragments and corresponding
fragments of representative genomes of genotypes A–H. All
the 12 mosaic fragments clustered with the corresponding
representative fragments of genotype D with a bootstrap
value of 93% in the tree (Fig. 1a). A similar tree was also
inferred based on alignment of the corresponding whole
genomes. All 12 genome sequences clustered with the
representativegenomesofgenotypeAwithabootstrapvalue
of100%(Fig. 1b).Phylogenetic treesbasedonothermosaic
fragments were also reconstructed to demonstrate the
mosaicism using a 75% bootstrap value cut-off (data not
shown). The mosaicism of the precisely derived fragments
had much stronger bootstrap support than did that of the
contrived fragments.
Finally, 166 genomes were revealed as putative intergeno-
typerecombinantssupportedbyphylogenetictrees.Accord-
ing to the fragment types, these genomes were classified into
23 intergenotypes. For convenience of discussion, the 23
intergenotypeswerenamedintergenotype1–23(IG-1toIG-
23 as indicated in Table 2 and Table 3). Fig. 2 shows the
SimPlot bootscanning results and Table 3 lists the break-
points of representative genome sequences of the 23 inter-
genotypes. The bootstrap values for each mosaic fragment
are also listed in Table 3. Sequences in the same intergeno-
type had similar breakpoints. The 59 terminal breakpoints
and the 39 terminal breakpoints of the 117 genomes in IG-6
were located within nt 1730–1846 and 2244–2485 respec-
tively.Otherintergenotypeshadbreakpoints withvariations
of only tens of base pairs
In addition to the above 166 genomes, constituting the
23 intergenotypes, there are five genomes that need
further investigation and are currently categorized to two
uncertain intergenotypes. Genomes AF241407–AF241409
and AB231908 were originally from Vietnam. These four
genomes exhibited a complicated fragment type of ‘NAGG-
GECCCCCCN’ in the fragment typing stage, where N
represents an uncertain genotype of the 250 bp contrived
fragment, i.e. the fragment did not cluster with any of the
representative fragments of the eight genotypes in the tree.
In the following analysis stage with precisely derived mosaic
fragments, the mosaic areas corresponding to A, G and E
werenotsupportedbybootstrapvalueshigherthanthe75%
cut-off.SimPlotanalysisalsoshowedthatthecorresponding
area between nt 2866 and 1800 was significantly different
Table 1. Twenty-four full-length representatives of genotypes A–H
Subgenotypes of some representatives are indicated in parentheses.
GenotypeABCDEFGH
Representative 1
Representative 2
Representative 3
Standard length
(bp)
M57663 (A1)
X70185 (A2)
X51970 (A3)
3221
D50521 (B1)
AY217358 (B2)
D00331 (B3)
3215
AF411411 (C1)
AF223960 (C2)
X75656 (C3)
3215
Y07587 (D1)
AB090268 (D2)
AB048703 (D4)
3182
AB091255
X75657
X75664
3212
AY090458 (F1)
AB036910 (F2)
X75658 (F2)
3215
AF405706
AF160501
AB056515
3248
AB059660
AY090454
AY090460
3215
http://vir.sgmjournals.org 2205
Identification of HBV intergenotype recombinants

Page 4

Table 2. Fragment type analysis of the 837 HBV complete genome sequences
Accession numbers and countries are listed alphabetically according to the fragment types. Continuous accession numbers were combined for parity: for instance, ‘AB116076–94’ stands
for 19 genomes from AB116076 to AB116094. The latterly identified 25 intergenotypes are indicated in the second column (IG, intergenotype). C/Gi, Genotype C and gibbon hybrid sug-
gested by Simmonds & Midgley (2005).
Fragment type IGGenotypeNo. CountryAccession no.
AAAAAAAAAAAAAA122Bangladesh, Belgium, China, Cameroon,
Canada, Congo, France, Gambia, Germany,
India, Japan, Korea, Malawi, Nepal, Poland,
Philippines, Russia, Somalia, UAE,
South Africa, Tanzania, Uganda, USA
AB014370, AB064314, AB076678–79, AB116076–94, AB126580,
AB194950–52, AB205118, AF090838–42, AF143298–306, AF143308,
AF297621–25, AF418676–78, AF462041, AF536524, AF537371–72,
AJ012207, AJ309369–71, AJ344115, AP007263, AY034878, AY128092,
AY152726, AY161138–39, AY161142–44, AY233274–90, AY373428–29,
AY373432, AY707087, AY738139–43, AY902775, AY934763–74,
DQ020002–03, E10905, L1399, M57663, S50225, U87742, U87746,
V00866, X02763, X51970, X70185, Z35717, Z72478–79
AB010289–92, AB014366, AB073838, AB073842–58, AB106884–85,
AB205121, D00329, D23677–79, D50521–22
AB014360–99, AB026811–15, AB031262, AB033550–53, AB033556–57,
AB042282–85, AB049609–10, AB050018, AB074047, AB074755–56,
AB105172–74, AB106895, AB111112–14, AB111116–25, AB111946,
AB112063, AB112065–66, AB112348, AB112408, AB112471–72,
AB113875–79, AB115417–18, AB117758, AB176642, AB182589,
AB198076–84, AB205123–25, AF068756, AF182802–05, AF223954–61,
AF241410–11, AF286594, AF330110, AF363961–63, AF384371–72,
AF411408–12, AF458664–65, AF461357, AF461361, AF461363, AF473543,
AF533983, AJ748098, AY040627, AY050574, AY066028, AY123041,
AY123424, AY148342, AY167090–92, AY167095–96, AY167099, AY206374,
AY206376, AY206378–82, AY206384–86, AY206388–89, AY206392–93,
AY217371–78, AY220699–702, AY247030–32, AY306136, AY596107–08,
AY641558–63, AY800390, D00630, D12980, D16666–67, D23680–84,
D28880, D50489, D50517–20, DQ089756–88, DQ246215
AB078031–33, AB090268–70, AB104709–12, AB109475–79, AB110075,
AB116266, AB119251–56, AB120308, AB126581, AB188241–45,
AB205126–28, AF043593–94, AF121239–42, AF151735, AF280817,
AF418679–80, AJ131956, AJ344116, AY090452–53, AY161150–60,
AY161162–63, AY233291, AY236162–63, AY341335, AY373430–31,
AY661792–93, AY721605–09, AY721611–12, AY741794–98, AY796030–32,
AY945307, DQ111986, L27106, M32138, X02496, X59795, X80924–26,
X97848–49, Y07587, Z35716, DQ298161–65, DQ304547–51, DQ304556–57
AB091255–56, AB106564, AB194947–48, AB201287–90, AB205129,
AB205188–92, AB219529–34, AP007262, AY738144–47, AY739674–75,
AY935700, DQ060822–29, DQ060830, X75657, X75664
AB036905–20, AB064316, AB086397, AB116549–52, AB116654, AB166850,
AF223962–65, AY090455–56, AY090458, AY090459, AY090461, AY179734–35,
AY311369–70, NT_033927, X69798, X75658, X75663
BBBBBBBBBBBBBB 32Japan
CCCCCCCCCCCCCC 272Cambodia, China, Indonesia, Japan, Korea,
Malaysia, Myanmar, Sweden, Thailand, USA,
Vietnam
DDDDDDDDDDDDDD 108Australia, Belgium, China, Egypt, France,
Germany, Greece, India, Iran, Italy, Japan,
Mongolia, Papua New Guinea, Poland,
Russia, South Africa, Spain, Sweden,
Turkey, UK, USA
EEEEEEEEEEEEEE 40Angola, Congo Benin, Japan, Cameroon,
Ghana, Namibia, Madagascar, Sweden, UK
FFFFFFFFFFFFFF 41Argentina, Bolivia, Costa Rica, El Salvador,
Japan, Germany, Panama, Nicaragua,
Sweden, USA, Venezuela
2206
Journal of General Virology 87
J. Yang and others

Page 5

GGGGGGGGGGGGG
HHHHHHHHHHHHH
G
H
11
10
Belgium, Japan, Germany, USA
Japan, Nicaragua, USA
AB056513–15, AB064310–13, AF160501, AF405706, AP007264, DQ207798
AB059659, AB059660, AB059661, AB064315, AB179747, AB205010, AP007261,
AY090454, AY090457, AY090460
AF418684–89, AY161148–49
AF418690–92, AY161147
AF418674–75, AY161140–41, AY161145–46, AF418682–83
AF297619, AF297620
AF297621
AB031266–67, AB033554–55, AB073821–37, AB073839–41, AB100695–51,
AB117759, AB205119–20, AB205122, AB212625–26, AB219426–29,
AF100308–09, AF121243–51, AF282917–18, AF461360, AF461362, AF479684,
AJ131133, AY033072–73, AY163869–70, AY167089, AY167093–94,
AY167097–102, AY206373, AY206375, AY206377, AY206380, AY206383,
AY206390–91, AY217355–70, AY220697–98, AY220703–04, AY293309,
AY518556, AY596102–12, AY766463, AY800389–92, D00330–31, M54923,
X97850–51, X98072–75, X98077
U87747
AY057947
AB031265
AF233236
D16665
AF461043, AY800249, AY817509–10, AY8175–15
AY057948, AY817511
AY236161
X65258
X65259
X68292
AF418681, AY161161
AB194949
AB214516
AB056516
DQ078791
AB219430
AB231908, AF241407–09
AB231909
AB048704–05
AB048701–03
AB033558–59, AJ132335, AJ344117, AY233292–96, AY236160, AY236164,
AY902768–70, AY902772–74, AY902776–77, DQ111987, U95551, V01460,
X65257, X72702, X85254
ADADAAAAAAAAD
DDADAAAAAAAN-
AAAAAAADDNAAA
DDAAAAAAAAAND
AAAAAADNAAAAA
BBBBBBBCCNBBB
IG-1
IG-2
IG-3
IG-4
IG-5
IG-6
A
A
A
A
A
B
8
4
8
2
1
India
India
India
South Africa
South Africa
Cambodia, China, Indonesia, Japan,
Switzerland, Thailand, Vietnam
117
BBBBBBBCCNBBA
CCCCCCCCCCCCA
CCCCNBBCCCCCC
CCCCCBBCCCCCC
CCCCCCCBBCCCC
DDDCCCCCCCCCC
DDDDDDCCCCCCC
DDAAAAADDDDDD
DDADDDAADDDDD
DDDAAAANDDDDD
DDDADAAADDDDD
DDADDDDDDDDDD
AAAEAAEAAAAAA
FFFFFFCFFFFFF
AGGGGGGGGGGGG
GGGGGGGCCCGGG
BBBBBAACCNBBB
NAGGGECCCCCCN
BBBGGECCCNBBB
CNCNCCCCCCCCC
DDDDDDDDDNDDD
DDDDDDNDDDDDD
IG-7
IG-8
IG-9
IG-10
IG-11
IG-12
IG-13
IG-14
IG-15
IG-16
IG-17
IG-18
IG-19
IG-20
IG-21
IG-22
IG-23
IG-24
IG-25
C/Gi
B
C
C
C
C
C
C
D
D
D
D
D
A
F
G
G
B
C
B
C
D
D
1
1
1
1
1
8
2
1
1
1
1
2
1
1
1
1
1
4
1
2
3
South Africa
China
Vietnam
China
Japan
China
China
Italy
Italy
Italy
Italy
India
Cameroon
Bolivia
USA
Thailand
Philippines
Vietnam
Vietnam
Australia
Australia
France, Italy, Japan, Mongolia,
South Africa, USA
25
Total837
Table 2. cont.
Fragment typeIGGenotypeNo.Country Accession no.
http://vir.sgmjournals.org
2207
Identification of HBV intergenotype recombinants