A comprehensive panel of near-full-length clones and reference sequences for non-subtype B isolates of human immunodeficiency virus type 1

Abstract

Non-subtype B viruses cause the vast majority of new human immunodeficiency virus type 1 (HIV-1) infections worldwide and are thus the major focus of international vaccine efforts. Although their geographic dissemination is carefully monitored, their immunogenic and biological properties remain largely unknown, in part because well-characterized virological reference reagents are lacking. In particular, full-length clones and sequences are rare, since subtype classification is frequently based on small PCR-derived viral fragments. There are only five proviral clones available for viruses other than subtype B, and these represent only 3 of the 10 proposed (group M) sequence subtypes. This lack of reference sequences also confounds the identification and analysis of mosaic (recombinant) genomes, which appear to be arising with increasing frequency in areas where multiple sequence subtypes cocirculate. To generate a more representative panel of non-subtype B reference reagents, we have cloned (by long PCR or lambda phage techniques) and sequenced 10 near-full-length HIV-1 genomes (lacking less than 80 bp of long terminal repeat sequences) from primary isolates collected at major epicenters of the global AIDS pandemic. Detailed phylogenetic analyses identified six that represented nonrecombinant members of HIV-1 subtypes A (92UG037.1), C (92BR025. 8), D (84ZR085.1 and 94UG114.1), F (93BR020.1), and H (90CF056.1), the last two comprising the first full-length examples of these subtypes. Four others were found to be complex mosaics of subtypes A and C (92RW009.6), A and G (92NG083.2 and 92NG003.1), and B and F (93BR029.4), again emphasizing the impact of intersubtype recombination on global HIV-1 diversification. Although a number of clones had frameshift mutations or translational stop codons in major open reading frames, all the genomes contained a complete set of genes and three had intact genomic organizations without inactivating mutations. Reconstruction of one of these (94UG114.1) yielded replication-competent virus that grew to high titers in normal donor peripheral blood mononuclear cell cultures. This panel of non-subtype B reference genomes should prove valuable for structure-function studies of genetically diverse viral gene products, the generation of subtype-specific immunological reagents, and the production of DNA- and protein-based subunit vaccines directed against a broader spectrum of viruses.

FIG. 1

Phylogenetic relationships of the newly characterized viruses (highlighted) to representatives of all major HIV-1 (group M) subtypes in gag and env regions. Trees were constructed from full-length gag and env nucleotide sequences by using the neighbor-joining method (see the text for details of the method). Horizontal branch lengths are drawn to scale (the scale bar represents 0.02 nucleotide substitution per site); vertical separation is for clarity only. Values at the nodes indicate the percent bootstraps in which the cluster to the right was supported (bootstrap values of 75% and higher are shown). Asterisks denote two hybrid genomes with discordant branching orders in gag and env trees. Brackets on the right represent the major sequence subtypes of HIV-1 group M. Trees were rooted by using SIVcpzGAB as an outgroup.

FIG. 2

Diversity plots comparing the sequence relationships of the newly characterized viruses to each other and to reference sequences from the database. In each panel, the sequence named above the plots is compared to the sequences listed on the right (sequences are color coded). U455, LAI, C2220, and NDK are published reference sequences for subtypes A, B, C, and D, respectively (45). Distance values were calculated for a window of 500 bp moving in steps of 10 nucleotides. The x axis indicates the nucleotide positions along the alignment (gaps were stripped and removed from the alignment). The positions of the start codons of the gag, pol, vif, vpr, env, and nef genes are shown. The y axis denotes the distance between the viruses compared (0.05 = 5% divergence).

FIG. 3

Exploratory tree analysis. (A) Neighbor-joining trees were constructed for a 500-bp window moving in increments of 100 bp along the multiple genome alignment. Trees depicting discordant branching orders among the newly determined sequences are shown (hybrid sequences are boxed and color coded). The position of each tree in the alignment is indicated; subtypes are identified by curved brackets. Numbers at the nodes indicate the percentage of bootstrap values with which the adjacent cluster is supported (only values above 80% are shown). Branch lengths are drawn to scale. (B) Summary of the subtype assignments of the four recombinants illustrated in panel A.

FIG. 4

Recombination breakpoint analysis for 92RW009.6 and 93BR029.4. (A) Bootstrap plots depicting the relationship of 92RW009.6 to representatives of subtype A (red) and C (blue), respectively. Trees were constructed from the multiple genome alignment, and the magnitude of the bootstrap value supporting the clustering of 92RW009.6 with U455 and 92UG037.1 (subtype A) or with C2220 and 92BR025.8 (subtype C), respectively, was plotted for a window of 500 bp moving in increments of 10 bp along the alignment. Regions of subtype A or C origin are identified by very high bootstrap values (>90%). Points of crossover of the two curves indicate recombination breakpoints. The beginnings of gag, pol, vif, vpr, env, and nef open reading frames are shown. The y axis indicates the percentage of bootstrap replicates which support the clustering of 92RW009.6 with representatives of the respective subtypes. (B) Bootstrap plots depicting the relationship of 93BR029.4 to representatives of subtypes B (black) and F (magenta), respectively. Analyses are as in panel A, except that the bootstrap values supporting the clustering of 93BR029.4 with SF2, OYI, MN, LAI, and RF (subtype B) or with 93BR020.1 (subtype F), respectively, were plotted. Subtype D viruses were excluded from this analysis because of their known close relationship to subtype B viruses.

FIG. 5

Recombination breakpoint analysis of 92NG083.2 and 92NG003.1. (A and B) Diversity plots comparing the sequence relationships of 92NG003.1 and 92NG083.2 to each other and to reference sequences from the database. In both panels, the sequence named above the plots is compared to the sequences listed on the right (sequences are color coded). U455, C2220, and NDK are published reference sequences for subtypes A, C, and D, respectively (45). Distance values were calculated for a window of 300 bp moving in steps of 10 nucleotides. The x axis indicates the nucleotide positions along the alignment (gaps were stripped and removed from the alignment). The positions of the start codons of the vif, vpr, and env genes are shown. The y axis denotes the distance between the viruses compared (0.05 = 5% divergence). (C) Neighbor-joining trees depicting discordant branching orders of 92NG003.1 and 92NG083.2 in regions delineated by breakpoints identified in panels A and B (hybrid sequences are boxed and color coded). The position of each tree in the alignment is indicated; subtypes are identified by curved brackets. Numbers at the nodes indicate the percentage of bootstrap values with which the adjacent cluster is supported (only values above 80% are shown). Branch lengths are drawn to scale.

FIG. 6

Inferred structures of the four recombinant genomes characterized in this study. Regions of different subtype origin are color coded. Uncertain breakpoints are hatched. LTR sequences were not analyzed and are shown as open boxes.

FIG. 7

Phylogenetic relationships of subtype G (and “E”) viruses in vpu and env regions. Trees were constructed for the vpu (A), 5′ env (B), and 3′ env (C and D) regions to reexamine the subtype associations of previously classified subtype A, G, and “E” viruses (19). Several strains (boxed) previously thought to represent subtype A (panel B) were found to cluster in subtype G viruses in the vpu region (panel A). Exclusion of these G/A recombinants changed the topology of trees derived from the intracellular gp41 domain (panels C and D). VI525 (highlighted by an asterisk) was identified as a G/H recombinant, clustering in subtype G and H in the extracellular and intracellular portions of env, respectively. All known representatives for the different subtypes were included in the analysis, and only a few representatives for subtypes B and “E” are shown.

FIG. 8

Subtype-specific genome features. (A) Alignment of deduced Tat (region encoded by second exon) amino acid sequences. Consensus sequences were generated for available representatives of all major subtypes (question marks indicate sites at which fewer than 50% of the viruses contain the same amino acid residue). Dashes denote sequence identity with the consensus sequence, while dots represent gaps introduced to optimize the alignments. A vertical box highlights a premature Tat protein truncation (asterisk) which is present in 11 of 15 subtype D and 4 of 52 subtype B viruses (frequencies are listed in the column on the right). (B) Alignment of deduced Rev (region encoded by the second exon) protein sequences. (C) Alignment of deduced Vpu protein sequences.

FIG. 9

Generation of replication-competent proviral clones from long-PCR products. (A) Construction of a replication-competent 94UG114.1 provirus from two separately amplified genomic regions (see the text for details). (B) Replication potential of 94UG114.1 in primary PBMC cultures. Normal donor PBMCs were isolated, PHA stimulated and then infected with equal amounts (based on p24 antigen content) of 94UG114.1 and SG3 viruses derived from 293T transfections of proviral DNA. Virus production was monitored by measuring supernatant RT activity at 3-day intervals as described previously (20). Supernatants from a mock-transfected culture served as a negative control.

FIG. 10

Phylogeny of full-length pol sequences of seven major HIV-1 group M subtypes. The sequences determined in this study are highlighted. Horizontal branch lengths are drawn to scale (the scale bar represents 0.02 nucleotide substitution per site). Vertical separation is for clarity only. Values at the nodes indicate the percentage of bootstraps in which the cluster to the right was supported (bootstrap values of 80% and higher only are shown). Brackets on the right represent the major sequence subtypes of HIV-1 group M. Trees were rooted by using SIVcpzGAB as an outgroup.