Journal of General Virology

Diversity of mosaic structures and common ancestry of human immunodeficiency virus type 1 BF intersubtype recombinant viruses from Argentina revealed by analysis of near full-length genome sequences

Michael M. Thomson, Elena Delgado, Isabel Herrero, María Luisa Villahermosa, Elena Vázquez-de Parga, María Teresa Cuevas, Rocío Carmona, Leandro Medrano, Lucía Pérez-Álvarez, Laureano Cuevas and Rafael Nájera

Department of Viral Pathogenesis, Instituto de Salud Carlos III, Ctra. Majadahonda-Pozuelo, Km. 2, 28220 Majadahonda, Madrid, Spain

The findings that BF intersubtype recombinant human immunodeficiency type 1 viruses (HIV-1) with coincident breakpoints in pol are circulating widely in Argentina and that non-recombinant F subtype viruses have failed to be detected in this country were reported recently. To analyse the mosaic structures of these viruses and to determine their phylogenetic relationship, near full-length proviral genomes of eight of these recombinant viruses were amplified by PCR and sequenced. Intersubtype breakpoints were analysed by bootscanning and examining the signature nucleotides. Phylogenetic relationships were determined with neighbour-joining trees. Six different recombination patterns were identified. Five viruses, each with predominantly subtype F genomes, exhibited mosaic structures that were highly similar. Two intersubtype breakpoints were shared by all viruses and seven by the majority. Of the consensus breakpoints, all nine were present in two viruses, which exhibited identical recombinant structures, and four to eight breakpoints were present in the remaining viruses. Phylogenetic analysis of partial sequences supported both a common ancestry, at least in part of their genomes, for all recombinant viruses and the phylogenetic relationship of F subtype segments with F subtype viruses from Brazil. A common ancestry of the recombinants was supported also by the presence of shared signature amino acids and nucleotides, either unreported or highly unusual in F and B subtype viruses. These results indicate that HIV-1 BF recombinant viruses with diverse mosaic structures, including a circulating recombinant form (which are widespread in Argentina) derive from a common recombinant ancestor and that F subtype segments of these recombinants are related phylogenetically to the F subtype viruses from Brazil.

Recombination is one of the major mechanisms contributing to retrovirus variability (Katz & Skalka, 1990; Temin, 1991), allowing for the rapid generation of virus variants with increased replicative capacity (Wooley et al., 1997; Gundlach et al., 2000), drug resistance (Gu et al., 1995; Kellam & Larder, 1995; Moutouh et al., 1996) or modified expression of antigenic epitopes (Tumas et al., 1993). Recombination in retroviruses occurs during reverse transcription by the alternate switching of templates between both genomic RNA strands copackaged in each virion, according to the copy-choice model (Coffin, 1996). The importance of recombination as a major factor that contributes to shaping the diversity of human immunodeficiency virus type 1 (HIV-1) genetic forms circulating in the global pandemic was not appreciated until relatively recently (Gao et al., 1996a; Carr et al., 1996; McCutchan et al., 1996). Currently, 11 intersubtype circulating recombinant forms (CRFs) have been reported (Theoretical Biology and Biophysics Group, 2001), although only CRF01_AE of Southeast Asia, CRF02_AG of West Africa, CRF03_AB of Kaliningrad, Russia and CRF07_BC and CRF08_BC of China have spread widely in epidemic proportions (McCutchan, 2000).

We reported recently the finding of BF intersubtype recombinant viruses that are widely circulating in Argentina (Thomson et al., 2000). This country has, only surpassed by Brazil, the second most numerous population of HIV-1-infected individuals in South America, with an estimated number of 130000 HIV-1-infections at the end of 1999 (UNAIDS, 2001), which were mainly concentrated in Buenos Aires city and province, where 75 % of AIDS cases have been notified (Ministry of Health, 2001). Early in the epidemic, most AIDS cases were diagnosed in homosexual men and subsequently in injecting drug users (IDUs), but, more recently, there has been an increase in infections transmitted heterosexually, particularly among women (Ministry of Health, 2001; Cahn et al., 1998; Vila-Pérez & Bianco, 1998). In our previous study, we detected BF recombinant viruses in 21 (40 %), subtype B viruses in 31 (60 %) and non-recombinant F subtype viruses in none of 52 samples collected in Buenos Aires between 1995 and 1998 (Thomson et al., 2000). Recombinant viruses were predominant among IDUs and heterosexually infected women, whereas subtype B viruses were predominant among hetero- and homosexually infected men. Coincident breakpoints in pol suggested a common ancestry of the recombinants. These results apparently contradicted other studies (Marquina et al., 1996; Campodonico et al., 1996; Fernández-Medina et al., 1999; Masciotra et al., 2000), suggesting the frequent finding of F subtype viruses (up to 40 %) and the relative scarcity (less than 5 %) of BF recombinant viruses in Argentina. This apparent discrepancy can be explained by the fact that the segments examined in these studies were, in most recombinants from Argentina, of subtype F, while our analysis included a segment of pol containing intersubtype breakpoints in all of these viruses. Consequently, our results implied that most, if not all, viruses from Argentina, identified previously as being subtype F, were probably BF recombinants. In the 93 samples collected from Buenos Aires in 1999, we have confirmed the absence of non-recombinant F subtype viruses and the high prevalence (65 % of samples) of recombinant BF subtype viruses (unpublished data).

For a more complete genetic characterization of recombinant BF viruses from Argentina, we have analysed the near full-length sequences of eight of these viruses, examining recombination points, phylogenetic relationships and the presence of characteristic amino acids and nucleotides. The results indicate that, while there is a considerable diversity of mosaic structure, all recombinant viruses examined appear to share a common ancestry.

Subjects. The eight subjects studied, five women and three men, attended hospitals in Buenos Aires and had been identified previously, by analysis of partial pol sequences, as harbouring recombinant BF viruses. Three of these subjects were reported previously by Thomson et al. (2000). Of the eight subjects, four subjects (all women) were infected through heterosexual contact and three subjects were IDUs. The risk exposure of one subject was not available. No mutual epidemiological links were known. Samples were collected in 1997 or 1999. Epidemiological data of all subjects are shown in Table 1.

Table 1. Epidemiological data of study subjects

NA, Not available.

Sample preparation, PCR amplification and sequencing. Peripheral blood mononuclear cells were separated by centrifugation on Ficoll–hypaque gradients. Samples were prepared for PCR by cell lysis and digestion with proteinase K, as described previously (Tenorio et al., 1993). A lysate of 1.2x10⁵ cells was used for each PCR. Amplification of the near full-length proviral genome (approximately 9 kb) was carried out by nested PCR in four overlapping segments of 1.8–3 kb each. Reagents and thermocycling profiles were identical for both first- and second-round PCR amplifications. Each reaction included 2.5 U Taq DNA polymerase, 0.3 U Pfu DNA polymerase, 0.2 mM each dNTP, 0.4 mM each primer, 2 mM MgCl₂, 16 mM (NH₄)₂SO₄, 67 mM Tris–HCl (pH 8.8) and 0.01 % Tween-20 in a volume of 50 µl. The thermocycling profile was as follows: initial denaturation for 3 min at 94 °C, 35 cycles of 94 °C for 3 s, 57 °C for 30 s, 72 °C for 3 min and a final extension at 72 °C for 7 min. For second-round PCR, a 2 µl aliquot of the first-round PCR mixture was used. Sequences of primers are available upon request. Amplification was checked by agarose gel electrophoresis and ethidium bromide staining. After enzymatic removal of the primers and dNTPs that remained in solution (Werle et al., 1994), purified PCR products were sequenced directly in overlapping segments of approximately 500 nt by primer walking using the ABI Prism BigDye Terminator Cycle Sequencing kit and the ABI 377 Sequencer kit (Applied Biosystems). Sequences were corrected and assembled using the BioEdit program (Tom Hall, http://www.mbio.ncsu.edu/BioEdit/bioedit.html). To exclude the possibility of PCR-mediated artefacts, breakpoints were confirmed in duplicate PCR amplifications carried out separately.

Phylogenetic analysis. Sequences were aligned using CLUSTAL X (Thompson et al., 1997) with minor manual adjustments considering protein sequences. Phylogenetic neighbour-joining trees (Saitou & Nei, 1987) were based on Kimura's two-parameter distance matrices (Kimura, 1980) with assessment of the consistency of tree topologies by bootstrapping (Felsenstein, 1985); trees were constructed with CLUSTAL X and viewed with TreeView (Rod Page, http://taxonomy.zoology.gla.ac.uk/rod/treeview.html). Analysis of recombination points was done by bootscanning (Salminen et al., 1995) using the Simplot software, version 2.5 (Stuart Ray, http://www.med.jhu.edu/deptmed/sray/download/). Sites with a gap in any of the sequences were excluded from the analysis. A 70 % bootstrap support was considered to be definitive (Hillis & Bull, 1993). Breakpoints were mapped more precisely by the inspection of subtype signature nucleotides in alignments with a set of full-length sequences of subtype reference isolates included in the 1999 compendium of the Los Alamos National Laboratory (Theoretical Biology and Biophysics Group, 1999). Signature nucleotides that discriminate between B and F1 subtypes were defined as those found in at least 50 % of reference isolates of one subtype and in less than 10 % of those of the other. Since only four full-length F1 subtype sequences are available in the Los Alamos Database, a nucleotide had to be absent in all four sequences in order for the signature to be considered subtype B.

Recombinant structures

Bootscan plots of the full-length sequences of two viruses are shown in Fig. 1 and the mosaic structures of all eight recombinant viruses, based on bootscan analysis and inspection of signature nucleotides, are depicted in Fig. 2. Six different mosaic patterns were identified, with multiple intersubtype breakpoints in each virus. Two intersubtype breakpoints were coincident in all and seven in the majority of sequences. All nine 'consensus' breakpoints were identified in two viruses (A32789 and A32989) and four to eight were found in each of the remainder six viruses. Positions (HXB-2 numeration) of consensus breakpoints representing midpoints of intersubtype signature nucleotide transitions are indicated at the bottom of Fig. 2. In A32789 and A32989, representative of the consensus mosaic structure (i.e. delimited by consensus breakpoints), there are 10 segments (designated 1–10 in 5´ to 3´ order) of alternating B and F subtypes. In both viruses, subtype B segments comprising approximately 1.6 kb of a predominantly subtype F genome are distributed along the genome as follows: (1) in the 5´ leader sequence and the 5´ segment of p17^gag; (3) across the protease–reverse transcriptase (RT) border; (5) in the polymerase (p51) domain of RT (this is the longest subtype B segment, approximately 0.7 kb); (7) in the overlap of the first coding exons of tat and rev and in the 5´ segment of vpu; and (9) in the overlap of gp41 and the second coding exon of rev. In segment 3, the bootstrap value that supports the grouping of the recombinant viruses with subtype B did not reach significant values, probably due to its short length and high sequence conservation. This resulted in few phylogenetically informative sites, although inspection of signature nucleotides revealed the presence of four subtype B and no subtype F signatures, thus supporting its tentative assignation to subtype B. A027, A047 and A050 exhibited mosaic structures highly similar to A32879 and A32989, but A027 and A047 lacked a B subtype segment in p17^gag and A050 had a short, extra B subtype segment near the 5´ end of nef.

Fig. 1. Bootscan analysis of full-length sequences of two recombinant BF viruses from Argentina. The horizontal axis represents nucleotide distance of the midpoint of the window from the 5´ end of the query sequence (nt 657 in HXB-2). The vertical axis represents the percentage of trees (using 100 bootstrap replicates) that support branching with the consensus subtype reference sequence. A 300 nt window advanced in 20 nt increments was used. Sequences were gap-stripped, transversion to transition ratio was set to 2.0, distances were calculated according to Kimura's two-parameter model and trees were constructed with the neighbour-joining algorithm.

A32878, A025 and A063, while sharing several breakpoints with the other five viruses, had different mosaic structures in substantial portions of their genomes, containing additional subtype B segments that were absent from the other viruses. In A32878, most of pol (including integrase, all of the RNase H domain and most of the polymerase domain of RT), and the entire first coding exon of tat are subtype B segments. In A025, subtype B segments include the entire p17^gag protein, the 5´ end of p24^gag, most of pol (except protease and the 3´ end of integrase) and a segment in the non-coding 3´ U3 region. In A063, B subtype segments comprise most of pol, the 3´ half of vif, most of env and a segment near the 5´ end of nef, the last coinciding with the B subtype segment at an identical site in A050. A063 is the only virus with a genome that is predominantly subtype B.

Fig. 2. Schematic representation of the mosaic structures of full-length genomes of eight recombinant BF viruses from Argentina. F subtype segments are shown in green and B subtype segments are shown in red. Segments are numbered 1–10 in the 5´ to 3´ order. LTR sequences that were not analysed are shown in white. Positions in the HXB-2 genome of consensus breakpoints, shown as vertical lines and indicated below, represent midpoints of transition segments between signature nucleotides of different subtypes. A ruler with nucleotide positions in HXB-2 is placed at the bottom.

Phylogenetic analysis of partial segments

Phylogenetic neighbour-joining trees of partial segments delimited by consensus breakpoints (i.e. those present in the majority of viruses and coinciding with the breakpoints of A32879 and A32989) (Fig. 3) confirm the subtype assignments of segments obtained by bootscanning, although, as observed in the bootscan analyses, bootstrap support of segment 3 clustering with subtype B reference viruses did not reach significant values (data not shown). The phylogenetic trees provide additional phylogenetic information: (i) subtype F segments 6 and 8 (excluding, in each tree, viruses that contain breakpoints in the corresponding segment) cluster with F subtype viruses from Brazil, with highly significant bootstrap values (Fig. 3c, d), suggesting a Brazilian ancestry of these segments; (ii) a tree of concatenated sequences obtained by joining all subtype F segments of each virus also supports the clustering of A32879, A32898, A027, A047 and A050 with each other, with 91 % bootstrap value, and with the F subtype reference virus 93BR020 from Brazil, with 100 % bootstrap support (Fig. 3f); (iii) in subtype B segment 5, located in RT, all recombinant BF viruses from Argentina, except A32878, cluster together, apart from five subtype B viruses from Argentina, with 63 % bootstrap value, which increases to 78 % when A32878 is excluded; shorter branches of recombinants relative to B subtype viruses from Argentina are consistent with a more recent common ancestry of the former (Fig. 3b); (iv) a tree of a B subtype segment in pol found only in A32878, A025 and A063 supports the phylogenetic relationship of A32878 and A063 with the BF recombinant virus 93BR029 from Brazil, but A025 branches separately (Fig. 3e). Branching of A32878 in segment 5 apart from the other viruses suggests that the B subtype sequences in pol of this virus derive from a second B subtype virus that is unrelated to the parent of segment 5 in the other viruses; the fact that A063 is phylogenetically related to A32878 in the 3´ half of pol suggests that this segment of A063, and possibly also the B subtype segments in vif and env that are unique to this virus, have an origin different from the parental segments of the other B subtype viruses that are common to all recombinants.

Fig. 3. Neighbour-joining phylogenetic trees of partial segments of recombinant BF viruses from Argentina. Viruses with intersubtype breakpoints in the segments analysed were excluded from each analysis of partial sequences. Nucleotide positions (HXB-2 numeration) that delimit the analysed fragments and the segment number according to Fig. 2 are shown above each tree. Trees are rooted with simian immunodeficiency virus isolate cpzUS. Recombinant BF viruses from Argentina are shown in boldface and F subtype viruses from Brazil are underlined. B subtype viruses from Argentina, sequenced in full-length or partial pol sequences, are marked with an asterisk and subtype reference sequences are indicated with the subtype designation followed by the name of the isolate. Bootstrap values ≥70 % (based on 100 replicates) are shown. The bootstrap value shown in parentheses (b) was obtained after excluding A32878 from the analysis. Relevant clusters including BF recombinants from Argentina and supported by ≥70 % bootstrap values are indicated in square brackets.

Phylogenetic tree of full-length sequences

In the phylogenetic tree of the full-length genome (Fig. 4), the five viruses exhibiting similar mosaic structures, A32879, A32989, A027, A047 and A050, group in a monophyletic cluster supported by 100 % bootstrap value. All viruses except A063, which clusters with subtype B sequences, branch with subtype F reference isolates, reflecting the predominance of subtype F along their genomes. When the B subtype segments of A32878, A025 and A063, found only in these viruses, are excluded, clustering of the remaining portion of their genomes with the five other isolates is supported by 100 % bootstrap values (data not shown).

Fig. 4. Unrooted neighbour-joining phylogenetic tree of full-length genomic sequences of recombinant BF viruses from Argentina. Recombinant viruses are shown in boldface and subtype reference viruses are indicated by the subtype designation followed by the name of the isolate. Bootstrap values >70 %, based on 100 replicates, of some key nodes are shown.

Recombinant structures and protein functional domains

The functional domains of the BF chimeric proteins p17^gag, RT, Tat, Rev, Vpu and gp41 of the consensus recombinant genome (which coincides with the structure of A32879 and A32989), indicating their relationship with subtype segments, are shown in Fig. 5. In the matrix p17^gag protein, the B subtype segment comprises the two N-terminal -helices of the globular domain, which is involved in membrane binding and Env incorporation into the virion probably through an interaction with the cytoplasmic tail of gp41 (Murakami & Freed, 2000; Cosson, 1996). In protease (data not shown), the presence of a short B subtype segment in the 3´ end results in one signature amino acid change at position 89, which is Leu (subtype B) instead of Met (subtype F). RT has an F/B/F structure, with the B subtype segment comprising a segment of the palm, containing Asp residues at positions 185 and 186 of the catalytic domain, the thumb and a portion of the connection subdomain. Drug resistance-associated amino acids at positions 41–118 are in the 5´ F subtype segment, whereas those at positions 151–236 are in the B subtype segment. Tat also has an F/B/F structure, with the B subtype segment, including the core and basic domains involved in binding TAR and nuclear localization. Rev exhibits the inverse pattern, B/F/B, with the basic domain, involved in RNA binding and nuclear localization (Pollard & Malim, 1998), lying in the subtype F segment. The subtype B segment of Vpu includes the membrane-spanning domain involved in virion release and ion channel formation, and one of the two -helices in the intracytoplasmic domain, which is involved in CD4 binding and degradation (Schubert et al., 1996). gp41 has a short subtype B segment in the intracytoplasmic tail, coinciding with -helix 2, which has been proposed to be involved in Env incorporation into the virion through interaction with the N-segment of the matrix protein (Murakami & Freed, 2000), which, notably, is also subtype B in the majority of BF recombinants from Argentina (excluding A027 and A047).

Fig. 5. Recombinant structures and functional domains of seven proteins of prototypical recombinant BF viruses from Argentina. B subtype segments are shown in red and F subtype segments are shown in green. Subtype segments are delimited by the midpoints of intersubtype nucleotide signature transitions.

Characteristic amino acids and nucleotides

Examination of the predicted amino acid sequences of viral proteins revealed some residues that are highly characteristic of the BF recombinants from Argentina. In Table 2, residues found in three or more recombinants, unreported in F1 subtype viruses and either unusual or unreported in B subtype viruses in the Los Alamos Database (Theoretical Biology and Biophysics Group, 2001) are shown. Of these, three are highly unusual in all group M viruses. Vif A⁶¹, identified in four of eight BF recombinants from Argentina, is absent from all 686 group M Vif sequences in the Los Alamos Database alignments. Moreover, a non-acidic amino acid at position 61 of Vif is highly unusual (present in 1.8 %) in group M viruses, whereas in the BF recombinants from Argentina, only one of eight viruses has an acidic amino acid at this position. Tat N⁶⁵, present in all eight recombinant BF viruses from Argentina, is found in only 24 (3.1 %) of 771 group M Tat sequences in the Los Alamos Database alignments, including 17 subtype B sequences derived from a single individual. gp120 S⁴⁴², identified in three BF recombinant viruses from Argentina, is unreported among B or F1/F2 subtype viruses and is present in only 11 (1.5 %) of 740 group M Env sequences in the Los Alamos Database alignments.

Table 2. Signature amino acids of recombinant BF viruses from Argentina

Amino acid positions (in parentheses) are numbered according to the proteins of the HIV-1 NL4.3 isolate. Signature amino acids of BF recombinants from Argentina are shown in boldface.

*Percentage of viruses in which amino acids shown in boldface are found in BF recombinants from Argentina (this study) and in B or F1 subtype viruses in the Los Alamos HIV Sequence Database alignments (Theoretical Biology & Biophysics Group, 2001).

Several synonymous nucleotide substitutions in coding regions and other nucleotides in non-coding regions highly characteristic of the BF recombinants from Argentina were also identified. Of these nucleotides, 12 were present in ≥50 % of the recombinants from Argentina and absent from all six F1/F2 subtype viruses and not found or highly unusual (≥2 of 64 full-length sequences) in subtype B references: A⁹³⁹, T³⁹⁵⁰, G⁴²⁵⁶, G⁵¹¹⁸, T⁷¹²⁷, T⁷²²⁰, C⁷³⁵², T⁷⁴⁶⁶, T⁸¹⁴⁴, C⁸⁹⁴⁶, C⁹⁴²⁶ and T⁹⁴⁷⁹ (positions are numbered according to the HXB-2 genome). Each recombinant virus from Argentina has 5–11 of these characteristic nucleotides. Among these, two nucleotides are either highly unusual or unreported in HIV-1 isolates: C⁹⁴²⁶ in the 3´ U3 region, found in seven of eight BF recombinants from Argentina, is present in only 2 (1 %) of 200 full-length HIV-1 sequences in the Los Alamos Database alignments, and T⁹⁴⁷⁹, at position 7 of the decanucleotide SP1-II-binding site of the 3´ U3 region (AGGGCGTGAC), found in five BF recombinants from Argentina, is absent from all 587 HIV-1 sequences spanning this segment in the Los Alamos Database alignments (in which a G is present in all cases except one). In the SP1-III site, the reverse substitution occurs: a pyrimidine (usually a T) at position 7 (HXB-2 position 9468), which is highly conserved among all group M subtypes, is replaced by a G in seven of the BF recombinants from Argentina (except A025, which is subtype B in this segment).

The crown tetrapeptide of the Env V3 loop is GPGR in five recombinants, GPGQ in two and GWGR in one. GPGR is frequent in F subtype viruses from Brazil and uncommon in other F subtype isolates. GWGR is present in A063, the only virus with gp120 of subtype B. This motif is characteristic of B subtype viruses from Brazil, found in 40 % of these viruses (Potts et al., 1993; Morgado et al., 1994, 1998) and uncommon elsewhere. In phylogenetic trees of gp120, A063 clustered with a group of subtype B viruses from Brazil, including all those containing the GWGR crown motif and excluding those with the GPGR tetrapeptide, although bootstrap values (57 %) did not reach significance (data not shown).

The results of this and our previous study on recombinant BF subtype viruses from Argentina underscore the importance of analysing full-length genome sequences to appreciate more completely the diversity of genetic forms circulating in a geographical area, particularly where diverse genetic forms of HIV-1 are cocirculating. In none of four studies in Argentina by other authors, where each study was based on the analysis of short sequences, was the real extent of the circulation of HIV-1 BF recombinant viruses recognized, since the segments analysed were, in most recombinants, subtype F (Marquina et al., 1996; Campodonico et al., 1996; Fernández-Medina et al., 1999; Masciotra et al., 2000). Only after the analysis of a segment in pol containing breakpoints in all recombinant viruses from Argentina was it realized that BF subtype recombinant viruses were circulating widely and that non-recombinant F subtype viruses could not be detected (Thomson et al., 2000); this implies that viruses classified previously as being subtype F viruses were probably recombinant BF subtype viruses. In the present study, we extend this novel finding by analysing the near full-length genome of eight recombinant BF subtype viruses from Argentina. Examination of the mosaic structures (Fig. 2) revealed six different patterns, although in a group of five viruses, with a predominantly F subtype genome, differences with each other were minor. Nine consensus breakpoints were identified: two of them were found in all eight viruses and seven in the majority. Two recombinants, A32879 and A32989, with coincident mosaic structures contained all nine consensus breakpoints, which delimit 10 segments of alternating B and F subtypes (designated 1–10). These viruses will be referred to as the prototypical BF recombinant viruses from Argentina. Partial pol and gag sequences of additional recombinant viruses from Argentina, not included in this study (Thomson et al., 2000; unpublished data), show mosaic structures that are, in most of them, coincident with those of A32879 and A32989. Sharing of several cross-over sites by all or the majority of isolates of this study suggests that they derive from a common recombinant ancestor, with subsequent recombination resulting in different mosaic patterns, as in the case of A32878, A063 and A025.

Further support for a common ancestry was obtained by phylogenetic analysis of partial segments delimited by breakpoints (Fig. 3), which also suggested a Brazilian ancestry of the F subtype segments. In the longest B subtype fragment common to all viruses (located in RT), all recombinants except A32878 formed a cluster separate from the B subtype viruses from Argentina (Fig. 3b). It is most parsimonious to assume that the remainder B subtype segments shared by all viruses probably derive from the same parental virus, considering the coincident breakpoints delimiting these segments. Clustering with F subtype viruses from Brazil was observed in two of the F subtype segments (Fig. 3c, d), as well as in concatenated sequences obtained by joining all F subtype segments of the recombinants (Fig. 3f); in the last tree, the five viruses exhibiting similar mosaic structures formed a cluster supported by a 91 % bootstrap value (Fig. 3f). The common ancestry of these five viruses was supported strongly in phylogenetic trees of full-length genomes (Fig. 4). Furthermore, when B subtype segments, found only in the remaining three viruses that showed divergent mosaic structures, were excluded from the analysis, a common origin of each virus with the other five isolates was also supported strongly in phylogenetic trees (data not shown). In two viruses, A32878 and A063, some B subtype segments appeared to derive from a virus unrelated to the parental virus of the prototypical recombinants. Clustering of these segments with the BF recombinant isolate 93BR029 from Brazil (Fig. 4e) and the presence in one virus of the GWGR V3 crown tetrapeptide, characteristically found in 40 % of B subtype viruses from Brazil (Potts et al., 1993; Morgado et al., 1994, 1998) and which is uncommon in Argentina (identified in only 1 of 24 B subtype V3 sequences from Argentina in the Los Alamos Database), suggested a Brazilian ancestry of the extra B subtype segments.

The presence of highly characteristic amino acid residues (Table 2) and nucleotides shared by all or the majority of the BF recombinants from Argentina also argues in favour of a common ancestry of these viruses. Two of these amino acids, Tat N⁶⁵ and Vpu N⁷⁷, were found in all of the recombinants. Notably, Vif A⁶¹, found in four viruses, is absent from all 686 HIV-1 group M sequences in the Los Alamos Database and Tat N⁶⁵ is found in only 1.8 % of group M sequences in the Database. Of the nucleotide substitutions that do not involve amino acid changes, the presence of a G to T mutation at position 7 of the decanucleotide SP1-II-binding site in the 3´ U3 region, found in five BF recombinants, is remarkable, as it is found in none of the 587 HIV-1 sequences in the Los Alamos Database. Similar to the SP1-II site of recombinants from Argentina, the SP1-I and SP1-III sites of other HIV-1 isolates usually have a T at position 7 instead of the consensus G of the SP1-binding sites and this substitution, in the context of the HIV-1 long terminal repeat (LTR), does not appear to diminish in vitro binding of the SP1 transcription factor (Jones et al., 1986).

Seven proteins, matrix, protease, RT, Tat, Rev, Vpu and gp41, exhibit chimeric structures in most recombinant viruses from Argentina. The spatial correlation of subtype segments with functional domains is shown in Fig. 5. Interestingly, the short B subtype segment in the cytoplasmic tail of gp41 matches with -helix 2, which has been proposed on the basis of mutational studies, to interact with residues in the N-segment of the matrix protein (Murakami & Freed, 2000; Cosson, 1996), which is also subtype B in six recombinants, for the incorporation of Env to the virions. Similarly, the F subtype of Rev, present in all viruses except A063, roughly coincides with the basic domain that interacts with the Rev responsive element (RRE) (Pollard & Malim, 1998), which is also subtype F in these viruses. In A063, both Rev (except a short segment comprising eight amino acids) and RRE are subtype B. Whether subtype coincidence of Rev–RRE and gp41–matrix interacting surfaces in the BF recombinants from Argentina is structurally favourable for increased affinity and beneficial for virus fitness remains to be determined.

The introduction of BF recombinant viruses in Argentina does not seem to be recent. Three IDUs harbouring recombinant viruses studied by us are known to have been infected by 1985 and one heterosexually infected man was infected by 1986. Masciotra et al. (2000) report a case of an IDU infected with F^env subtype virus (probably a BF recombinant) who first tested HIV-1-positive in 1987. Such an early introduction of the recombinant BF viruses in Argentina is consistent with mean intersubject genetic distances in the env V3 region (Thomson et al., 2000). Consequently, the recombinant BF viruses from Argentina might be the earliest known HIV-1 circulating intersubtype recombinant viruses to have originated outside Africa. A number of reasons point to a probable initial introduction of BF recombinants in Argentina among IDUs: (i) most HIV-1-infected IDUs harbour recombinant BF viruses [25 of 31 (81 %) samples studied]; (ii) most cases of recombinant BF virus infection with earlier dates of HIV-1 diagnosis are IDUs; (iii) the HIV-1 epidemic among heterosexually infected women, the other group in which BF recombinants are predominant, is relatively recent: up to 1990, only 2 % of the AIDS cases reported were women infected sexually, as compared to 31 % in IDUs (Ministry of Health, 2001); and (iv) all epidemics with intersubtype recombinant forms of non-African origin, CRF03_AB (Liitsola et al., 1998), CRF07_BC (Su et al., 2000), CRF08_BC (Piyasirisilp et al., 2000) and recombinant BG viruses of Spain (Thomson et al., 2001; unpublished data that show a newly characterized CRF), have been identified among IDUs.

Whether the ancestor of the recombinant viruses from Argentina originated locally or in Brazil is not known, but the absence of non-recombinant F subtype viruses in Argentina in 145 samples analysed by us suggests a probable Brazilian provenance of the recombinants. However, recombinant viruses related to those from Argentina appear to be either absent or not circulating widely in Brazil, since none of the BF recombinant sequences from Brazil reported to date (Sabino et al., 1994; Gao et al., 1996b, 1998; Morgado et al., 1994; Cornelissen et al., 1997; Brindeiro et al., 1999) nor any of the five recombinant BF viruses of this country analysed by us in gag and pol (unpublished data) exhibit a mosaic structure analogous to the prototypical recombinants from Argentina. An alternative possibility is that the individual source of the ancestor of the recombinants from Argentina resided and transmitted the recombinant form(s) in Argentina after acquiring the F subtype parental virus in Brazil.

The circulation in a geographical area of distinct but related recombinant forms has been reported previously (McCutchan et al., 1999; Janssens et al., 2000; Cornelissen et al., 2000; Motomura et al., 2000), although the degree of diversity of recombinant forms of a common ancestry identified in Argentina in this study is higher than that found in other areas. Previous reports of recombinant BF viruses from Argentina with mosaic structures that differ in partial segments from those described here (Marquina et al., 1996; Campodonico et al., 1996; Fernández-Medina et al., 1999; Masciotra et al., 2000), as well as our unpublished analysis of partial sequences, indicate that additional recombinant forms may be present in Argentina. Also, various BF recombinant forms have been identified, by us (unpublished data) and other authors, in Brazil by the analysis of partial sequences, although evidence for a common ancestry is lacking. Several factors may contribute to the high diversity of related recombinant forms in Argentina: (i) the relatively long period in which recombinant BF viruses from Argentina appear to have been in circulation, increasing the chances of recombination with other genetic forms; (ii) the cocirculation of recombinant B and BF subtype viruses in the same population; (iii) the high prevalence of HIV-1 infections among IDUs in Argentina [ranging from 20 to 92 % in different surveys (UNAIDS, 2001)], increasing the possibility of coinfections by needle sharing; and (iv) the analysis of full-length sequences of a relatively large number of isolates (compared to other studies), favouring the identification of differences in their recombinant structures.

The distribution of genetic forms in Argentina resembles, in some respects, that of Thailand, with a double epidemic: one of subtype B and another of closely related recombinant BF viruses that were probably introduced later, according to relative branch lengths in pol (Fig. 3b), and distributed unevenly among groups with different epidemiological characteristics. And again, similar to Thailand, Argentina has an apparent expansion of recombinants over the years. Of the 93 samples collected in 1999, BF recombinants increased from 50 % in individuals diagnosed before 1993 to 72 % in those diagnosed since 1993 (unpublished data). A similar temporal trend in the increase in F^env viruses (probably BF recombinants) was noticed in a previous study (Masciotra et al., 2000). Prospective studies might reveal if differences in properties of transmission contribute to unequal distribution of genetic forms among epidemiological groups and to temporal changes in the prevalence of the recombinant BF viruses from Argentina.

We have found in other countries BF recombinants related to those from Argentina. In Spain, we have identified two individuals, an Argentinian man and a Spanish woman, who was probably infected through sexual contact with a South American man harbouring recombinant BF viruses related phylogentically to the recombinant viruses from Argentina (unpublished data). In Venezuela, a woman, infected by her husband who acquired the infection in Argentina, harboured a recombinant BF virus with a breakpoint in pol identical to the A025 isolate reported here, although full-length pol sequences show differences in their mosaic structures (Delgado et al., 2001). Viruses acquired presumably in Argentina and identified in partial sequences as being subtype F viruses from Bolivia (Velarde-Dunois et al., 2000), Peru (Russell et al., 2000) and Spain (Holguín et al., 2000) might also be BF recombinants related to those reported here.

In summary, analysis of full-length HIV-1 genomic sequences reveals a considerable diversity but a common ancestry of recombinant BF viruses from Argentina and the phylogenetic relationship of these viruses with subtype F viruses from Brazil. The majority of recombinants exhibited similar mosaic structures, but some had unique patterns in part of their genomes, suggesting their generation by successive recombination of a common recombinant ancestor. With the identification of two viruses with identical mosaic structure, which was confirmed in several partial sequences (Thomson et al., 2000; unpublished data), the requirements to define an HIV-1 CRF are fulfilled (Robertson et al., 2000) and would, according to the nomenclature currently accepted, be designated CRF12_BF; [the full-length sequences of three HIV-1 isolates from Argentina and Uruguay proposed to represent a circulating BF recombinant form (CRF12_BF) was announced at the Los Alamos HIV Sequence Database website (http://hiv-web.lanl.gov) by other authors in the very recent past]. This is the first CRF identified to have originated from the Americas and, most likely, the oldest of the known CRFs to have originated outside Africa. It is of note that all known CRFs of non-African origin involve a parental subtype B virus, probably because of the circulation of B subtype viruses in the IDU populations among which they apparently originated; this parallels the predominance of subtype A viruses among the recombinant viruses of African origin. It is predictable that the cocirculation of diverse HIV-1 genetic forms that are, increasingly, brought into contact by international travel (Thomson & Nájera, 2001) will result in the identification of additional CRFs outside Africa. Analysis of full-length sequences of additional isolates from Argentina, Brazil and neighbouring countries will provide a more comprehensive picture of the spectrum of HIV-1 genetic forms circulating in South America and will contribute to the understanding of the mechanisms governing the generation of HIV-1 diversity and its impact on the progression and control of the epidemic.

We thank Horacio Salomón and Maria Ávila for providing clinical samples from Argentina, Amilcar Tanuri and Mariza Morgado for providing clinical samples from Brazil, Saladin Osmanov and José Esparza for organizing the UNAIDS program that made this study possible and Francisco Parras for his support of this study. This work was financed by grant 98BVII236 of Plan Nacional del SIDA, Ministerio de Sanidad y Consumo, Spain and by Technical Service Agreement HQ/98/457048, UNAIDS.

The eight near full-length sequences of BF recombinant viruses from Argentina have been deposited in GenBank under the accession numbers AF332867 and AF408626–AF408632

Brindeiro, R., Vanderborght, B., Caride, E., Correa, L., Oravec, R. M., Berro, O., Stuyver, L. & Tanuri, A. (1999). Sequence diversity of the reverse transcriptase of human immunodeficiency virus type 1 from untreated Brazilian individuals. Antimicrobial Agents and Chemotherapy 43, 1674–1680.

Cahn, P., Ben, G., Bloch, C., Pérez, N., San Pedro, M., González, S. & Onoto, G. (1998). Trends in HIV epidemic in Buenos Aires: analysis of 5019 patients. 12th World AIDS Conference (Geneva, Switzerland, June–July, 1998). Abstract 13178.

Campodonico, M., Janssens, W., Heyndrickx, L., Fransen, K., Leonaers, A., Fay, F. F., Taborda, M., van der Groen, G. & Fay, O. H. (1996). HIV type 1 subtypes in Argentina and genetic heterogeneity of the V3 region. AIDS Research and Human Retroviruses 12, 79–81.

Carr, J. K., Salminen, M. O., Koch, C., Gotte, D., Artenstein, A. W., Hegerich, P. A., St Louis, D., Burke, D. S. & McCutchan, F. E. (1996). Full-length sequence and mosaic structure of a human immunodeficiency virus type 1 isolate from Thailand. Journal of Virology 70, 5935–5943.

Coffin, J. M. (1996). Retroviridae: the viruses and their replication. In Fields Virology, 3rd edn, pp. 1767–1848. Edited by B. N. Fields, D. N. Knipe & P. M. Howley. Philadelphia: Lippincott–Raven.

Cornelissen, M., van den Burg, R., Zorgdrager, F., Lukashov, V. & Goudsmit, J. (1997). pol gene diversity of five human immunodeficiency virus type 1 subtypes: evidence for naturally occurring mutations that contribute to drug resistance, limited recombination patterns, and common ancestry for subtypes B and D. Journal of Virology 71, 6348–6358.

Cornelissen, M., van den Burg, R., Zorgdrager, F. & Goudsmit, J. (2000). Spread of distinct human immunodeficiency virus type 1 AG recombinant lineages in Africa. Journal of General Virology 81, 515–523.

Cosson, P. (1996). Direct interaction between the envelope and matrix proteins of HIV-1. EMBO Journal 15, 5783–5788.

Delgado, E., León-Ponte, M., Villahermosa, M. L., Cuevas, M. T., Deibis, L., Echeverría, G., Thomson, M. M., Pérez-Álvarez, L., Osmanov, S. & Nájera, R. (2001). Analysis of HIV type 1 protease and reverse transcriptase sequences from Venezuela for drug resistance-associated mutations and subtype classification: a UNAIDS study. AIDS Research and Human Retroviruses 17, 753–758.

Felsenstein, J. (1985). Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39, 783–791.

Fernández-Medina, D., Jansson, M., Rabinovich, R. D., ../LIBonatti, O. & Wigzell, H. (1999). Identification of human immunodeficiency virus type 1 subtypes B and F B/F recombinant and dual infection with these subtypes in Argentina. Scandinavian Journal of Infectious Diseases 31, 235–242.

Gao, F., Morrison, S. G., Robertson, D. L., Thornton, C. L., Craig, S., Karlsson, G., Sodroski, J., Morgado, M., Galvao-Castro, B., von Briesen, H. and others (1996a). Molecular cloning and analysis of functional envelope genes from human immunodeficiency virus type 1 sequence subtypes A through G. The WHO and NIAID Networks for HIV Isolation and Characterization. Journal of Virology 70, 1651–1667.

Gao, F., Robertson, D. L., Morrison, S. G., Hui, H., Craig, S., Decker, J., Fultz, P. N., Girard, M., Shaw, G. M., Hahn, B. H. & Sharp, P. M. (1996b). The heterosexual human immunodeficiency virus type 1 epidemic in Thailand is caused by an intersubtype (A/E) recombinant of African origin. Journal of Virology 70, 7013–7029.

Gao, F., Robertson, D. L., Carruthers, C. D., Morrison, S. G., Jian, B., Chen, Y., Barre-Sinoussi, F., Girard, M., Srinivasan, A., Abimiku, A. G., Shaw, G. M., Sharp, P. M. & Hahn, B. H. (1998). A comprehensive panel of near-full-length clones and reference sequences for non-subtype B isolates of human immunodeficiency virus type 1. Journal of Virology 72, 5680–5698.

Gu, Z., Gao, Q., Faust, E. A. & Wainberg, M. A. (1995). Possible involvement of cell fusion and viral recombination in generation of human immunodeficiency virus variants that display dual resistance to AZT and 3TC. Journal of General Virology 76, 2601–2605.

Gundlach, B. R., Lewis, M. G., Sopper, S., Schnell, T., Sodroski, J., Stahl-Hennig, C. & Uberla, K. (2000). Evidence for recombination of live, attenuated immunodeficiency virus vaccine with challenge virus to a more virulent strain. Journal of Virology 74, 3537–3542.

Hillis, D. M. & Bull, J. J. (1993). An empirical test of bootstrapping as a method for assessing confidence in phylogenetic analysis. Systematic Biology 42, 182–192.

Holguín, A., Rodés, B. & Soriano, V. (2000). Protease gene analysis of HIV type 1 non-B subtypes in Spain. AIDS Research and Human Retroviruses 16, 1395–1403.

Janssens, W., Salminen, M. O., Laukkanen, T., Heyndrickx, L., van der Auwera, Colebunders, R., McCutchan, F. E. & van der Groen, G. (2000). Near full-length genome analysis of HIV type 1 CRF02.AG subtype C and CRF02.AG subtype G recombinants. AIDS Research and Human Retroviruses 16, 1183–1189.

Jones, K. A., Kadonaga, J. T., Luciw, P. A. & Tjian, R. (1986). Activation of the AIDS retrovirus promoter by the cellular transcription factor, Sp1. Science 232, 755–759.

Katz, R. A. & Skalka, A. M. (1990). Generation of diversity in retroviruses. Annual Review of Genetics 24, 409–445.

Kellam, P. & Larder, B. A. (1995). Retroviral recombination can lead to linkage of reverse transcriptase mutations that confer increased zidovudine resistance. Journal of Virology 69, 669–674.

Kimura, M. (1980). A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. Journal of Molecular Evolution 16, 111–120.

Liitsola, K., Tashkinova, I., Laukkanen, T., Korovina, G., Smolskaja, T., Momot, O., Mashkilleyson, N., Chaplinskas, S., Brummer-Korvenkontio, H., Vanhatalo, J., Leinikki, P. & Salminen, M. O. (1998). HIV-1 genetic subtype A/B recombinant strain causing an explosive epidemic in injecting drug users in Kaliningrad. AIDS 12, 1907–1919.

McCutchan, F. E. (2000). Understanding the genetic diversity of HIV-1. AIDS 14 (Suppl. 3), S31–S44.

McCutchan, F. E., Salminen, M. O., Carr, J. K. & Burke, D. S. (1996). HIV-1 genetic diversity. AIDS 10 (Suppl. 3), S13–S20.

McCutchan, F. E., Carr, J. K., Bajani, M., Sanders-Buell, E., Harry, T. O., Stoeckli, T. C., Robbins, K. E., Gashau, W., Nasidi, A., Janssens, W. & Kalish, M. L. (1999). Subtype G and multiple forms of A/G intersubtype recombinant human immunodeficiency virus type 1 in Nigeria. Virology 254, 226–234.

Marquina, S., Leitner, T., Rabinovich, R. D., Benetucci, J., ../LIBonatti, O. & Albert, J. (1996). Coexistence of subtypes B, F, and an B/F^env recombinant of HIV type 1 in Buenos Aires Argentina. AIDS Research and Human Retroviruses 12, 1651–1654.

Masciotra, S., Livellara, B., Belloso, W., Clara, L., Tanuri, A., Ramos, A. C., Baggs, J., Lal, R. & Pieniazek, D. (2000). Evidence of a high frequency of HIV-1 subtype F infections in a heterosexual population in Buenos Aires, Argentina. AIDS Research and Human Retroviruses 16, 1007–1014.

Ministry of Health (2001). Update on AIDS in the Republic from Argentina. Buenos Aires, Argentina.

Morgado, M. G., Sabino, E. C., Shpaer, E. G., Bongertz, V., Brigido, L., Guimaraes, M. D., Castilho, E. A., Galvao-Castro, B., Mullins, J. I., Hendry, R. M. and others (1994). V3 region polymorphisms in HIV-1 from Brazil: prevalence of subtype B strains divergent from North American/European prototype and detection of subtype F. AIDS Research and Human Retroviruses 10, 569–576.

Morgado, M. G., Guimaraes, M. L., Gripp, C. B., Costa, C. I., Neves, I., Jr, Veloso, V. G., Linhares-Carvalho, M. I., Castello-Branco, L. R., Bastos, F. I., Kuiken, C., Castilho, E. A., Galvao-Castro, B. & Bongertz, V. (1998). Molecular epidemiology of HIV-1 in Brazil: high prevalence of HIV-1 subtype B and identification of an HIV-1 subtype D infection in the city of Rio de Janeiro, Brazil. Evandro Chagas Hospital AIDS Clinical Research Group. Journal of Acquired Immune Deficiency Syndromes and Human Retrovirology 18, 488–494.

Motomura, K., Kusagawa, S., Kato, K., Nohtomi, K., Lwin, H. H., Tun, K. M., Thwe, M., Oo, K. Y., Lwin, S., Kyaw, O., Zaw, M., Nagai, Y. & Takebe, Y. (2000). Emergence of new forms of human immunodeficiency virus type 1 intersubtype recombinants in central Myanmar. AIDS Research and Human Retroviruses 16, 1831–1843.

Moutouh, L., Corbeil, J. & Richman, D. D. (1996). Recombination leads to the rapid emergence of HIV-1 dually resistant mutants under selective drug pressure. Proceedings of the National Academy of Sciences, USA 93, 6106–6111.

Murakami, T. & Freed, E. O. (2000). Genetic evidence for an interaction between human immunodeficiency virus type 1 matrix and -helix 2 of the gp41 cytoplasmic tail. Journal of Virology 74, 3548–3554.

Piyasirisilp, S., McCutchan, F. E., Carr, J. K., Sanders-Buell, E., Liu, W., Chen, J., Wagner, R., Wolf, H., Shao, Y., Lai, S., Beyrer, C. & Yu, X. F. (2000). A recent outbreak of human immunodeficiency virus type 1 infection in southern China was initiated by two highly homogeneous, geographically separated strains, circulating recombinant form AE and a novel BC recombinant. Journal of Virology 74, 11286–11295.

Pollard, V. W. & Malim, M. H. (1998). The HIV-1 Rev protein. Annual Review of Microbiology 52, 491–532.

Potts, K. E., Kalish, M. L., Lott, T., Orloff, G., Luo, C. C., Bernard, M. A., Alves, C. B., Badaro, R., Suleiman, J., Ferreira, O. and others (1993). Genetic heterogeneity of the V3 region of the HIV-1 envelope glycoprotein in Brazil. Brazilian Collaborative AIDS Research Group. AIDS 7, 1191–1197.

Robertson, D. L., Anderson, J. P., Bradac, J. A., Carr, J. K., Foley, B., Funkhouser, R. K., Gao, F., Hahn, B. H., Kalish, M. L., Kuiken, C., Learn, G. H., Leitner, T., McCutchan, F. E., Osmanov, S., Peeters, M., Pieniazek, D., Salminen, M., Sharp, P. M., Wolinsky, S. & Korber, B. (2000). HIV-1 nomenclature proposal. Science 288, 55–56.

Russell, K. L., Carcamo, C., Watts, D. M., Sanchez, J., Gotuzzo, E., Euler, A., Blanco, J. C., Galeano, A., Alava, A., Mullins, J. I., Holmes, K. K. & Carr, J. K. (2000). Emerging genetic diversity of HIV-1 in South America. AIDS 14, 1785–1791.

Sabino, E. C., Shpaer, E. G., Morgado, M. G., Korber, B. T., Diaz, R. S., Bongertz, V., Cavalcante, S., Galvao-Castro, B., Mullins, J. I. & Mayer, A. (1994). Identification of human immunodeficiency virus type 1 envelope genes recombinant between subtypes B and F in two epidemiologically linked individuals from Brazil. Journal of Virology 68, 6340–6346.

Saitou, N. & Nei, M. (1987). The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4, 406–425.

Salminen, M. O., Carr, J. K., Burke, D. S. & McCutchan, F. E. (1995). Identification of breakpoints in intergenotypic recombinants of HIV type 1 by bootscanning. AIDS Research and Human Retroviruses 11, 1423–1425.

Schubert, U., Bour, S., Ferrer-Montiel, A. V., Montal, M., Maldarell, F. & Strebel, K. (1996). The two biological activities of human immunodeficiency virus type 1 Vpu protein involve two separable structural domains. Journal of Virology 70, 809–819.

Su, L., Graf, M., Zhang, Y., von Briesen, H., Xing, H., Kostler, J., Melzl, H., Wolf, H., Shao, Y. & Wagner, R. (2000). Characterization of a virtually full-length human immunodeficiency virus type 1 genome of a prevalent intersubtype (C/B´) recombinant strain in China. Journal of Virology 74, 11367–11376.

Temin, H. M. (1991). Sex and recombination in retroviruses. Trends in Genetics 7, 71–74.

Tenorio, A., EchevarrÌa, J. E., Casas, I., EchevarrÌa, J. M. & TabarÈs, E. (1993). Detection and typing of human herpesviruses by multiplex polymerase chain reaction. Journal of Virological Methods 44, 261–269.

Theoretical Biology and Biophysics Group (1999). Human retroviruses and AIDS 1999: a compilation and analysis of nucleic acid and amino acid sequences. Los Alamos National Laboratory, Los Alamos, New Mexico, USA.

Theoretical Biology and Biophysics Group (2001). HIV Sequence Database. Los Alamos National Laboratory, New Mexico, USA. http://hiv-web.lanl.gov.

Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F. & Higgins, D. G. (1997). The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Research 25, 4876–4882.

Thomson, M. M. & Nájera, R. (2001). Travel and the introduction of human immunodeficiency virus type 1 non-B subtype genetic forms into Western countries. Clinical Infectious Diseases 32, 1732–1737.

Thomson, M. M., Villahermosa, M. L., Vázquez-de Parga, E., Cuevas, M. T., Delgado, E., Manjón, N., Medrano, L., Pérez-Álvarez, L., Contreras, G., Carrillo, M. G., Salomón, H. & Nájera, R. (2000). Widespread circulation of a B/F intersubtype recombinant form among HIV-1-infected individuals in Buenos Aires, Argentina. AIDS 14, 897–899.

Thomson, M. M., Delgado, E., Manjón, N., Ocampo, A., Villahermosa, M. L., Mariño, A., Herrero, I., Cuevas, M. T., Vázquez-de Parga, E., Pérez-Álvarez, L., Medrano, L., Taboada, J. A., Nájera, R. & the Spanish Group for Antiretroviral Studies in Galicia (2001). HIV-1 genetic diversity in Galicia, Spain: BG intersubtype recombinant viruses are circulating among injecting drug users. AIDS 15, 509–516.

Tumas, K. M., Poszgay, J. M., Avidan, N., Ksiazek, S. J., Overmoyer, B., Blank, K. J. & Prystowsky, M. B. (1993). Loss of antigenic epitopes as the result of env gene recombination in retrovirus-induced leukemia in immunocompetent mice. Virology 192, 587–595.

UNAIDS (2001). Epidemiological fact sheet on HIV/AIDS and sexually transmitted infections. 2000 update. http://www.unaids.org.

Velarde-Dunois, K. G., Guimaraes, M. L., La Fuente, C., Andrade, R., Arévalo, R., Pantoja, S., Mariscal, R., Sandoval, R., Iriarte, F., Chamon, V., Melgar, M. L., Carvajal, R. & Morgado, M. G. (2000). Molecular characterization of human immunodeficiency virus type 1-infected individuals from Bolivia reveals the presence of two distinct genetic subtypes B and F. AIDS Research and Human Retroviruses 16, 1921–1926.

Vila-Pérez, M. & Bianco, M. (1998). Analysis of HIV/AIDS infection risk in young women living in Buenos Aires, Argentina. 12th World AIDS Conference. (Geneva, Switzerland, June–July, 1998). Abstract 23504.

Werle, E., Schneider, C., Renner, M., Volker, M. & Fiehn, W. (1994). Convenient single-step, one tube purification of PCR products for direct sequencing. Nucleic Acids Research 22, 4354–4355.

Wooley, D. P., Smith, R. A., Czajak, S. & Desrosiers, R. C. (1997). Direct demonstration of retroviral recombination in a rhesus monkey. Journal of Virology 71, 9650–9653.

JGV Direct table of contents

© 2002 SGM

This article is now available in the January 2001 print issue of JGV (vol. 83, 107–119). The complete issue of the journal may be seen in electronic form on JGV Online.