Infections with non-B HIV-1 subtypes are rare in the United States, but comprise a significant percentage of infections among U. S. military personnel. Risk behavior while on overseas deployment correlates with non-B infection in this population. Extensive genetic characterization will be required to define HIV-1 diversity, and to effectively evaluate requirements for HIV-1 vaccines and other prevention strategies in this group. From 1997 to 2000, 520 recent seroconverters, identified through routine HIV-1 testing in the U. S. active military force, volunteered for a prospective study. V3 loop serology or partial genome sequencing identified 28 non-B subtype infections; 14 were studied by full genome sequencing and phylogenetic analysis. Five strains were CRF01_AE. Four of these clustered with CM240 from Thailand, and one clustered with African CRF01_AE. Four strains were CRF02_AG, prevalent in West and West Central Africa. Two strains were subtype C. One strain was a unique recombinant between CRF01_AE and subtype B, and another was a complex unique recombinant between subtype A and D. The final strain was a member of a complex circulating recombinant first identified in Senegal, CRF09_cpx, incorporating subtypes A, F, G, and an unclassified genome. This diversity of non-B subtype HIV-1 strains, encompassing three globally prevalent non-B strains and including rare or even possibly unique strains, illustrates the breadth of U. S. military exposure while deployed and sets the bar higher for breadth of cross-subtype protection to be afforded by an HIV-1 vaccine.

In West/West Central Africa, CRF02_AG is the most prevalent HIV-1 strain and circulates in the milieu of rare subtypes, circulating recombinant forms (CRFs), and unique recombinant forms (URFs). The molecular complexity of HIV-1 epidemics in this region and the need to extensively sample large populations, such as in the case of vaccine trials, pose seemingly conflicting requirements between full-genome sequencing and high-throughput low-resolution assays. Here we describe the development and evaluation of a multiregion hybridization assay (MHAcrf02) for the efficient genotyping of CRF02_AG in West/West Central Africa. Subtype A, G, and CRF02_AG-specific fluorescent probes were designed flanking five recombination breakpoints in CRF02_AG and were used in real-time PCRs. A panel representing West/West Central African HIV-1 genetic diversity was evaluated by MHAcrf02. The sample set, previously characterized by full-genome sequencing, included CRF02_AG and CRF02_AG-containing recombinants (n = 28), other subtypes, CRFs, and URFs (n = 34). DNA from peripheral blood mononuclear cells, cocultures, and plasmids was used as template. When the patterns of probe reactivity were evaluated. CRF02_AG was identified with a 100% specificity and sensitivity. In conclusion, MHAcrf02 will permit more efficient characterization of HIV-1 in West/West Central Africa, where CRF02_AG is an important strain. Together with other regional genotyping assays MHAcrf02 will contribute to the development of a global picture of HIV-1 diversity and geographic distribution, providing a strong foundation for intervention, including vaccine development.

ABSTRACT Phylogenetic analyses demonstrate significant diversity in worldwide circulating strains of human immunodeficiency virus type 1 (HIV-1). Detailed studies have revealed a complex pattern of intersubtype recombinations, as well as evidence of sub-subtypes circulating in various populations. In this study, we characterized an HIV-1 strain that had previously been identified as a distinct subcluster within the subtype A radiation based on partial sequence data. These viruses were of particular interest given that we recently found that their prevalence was significantly higher in dually infected individuals compared to women who were singly infected with HIV-1. Five viruses isolated from commercial sex workers in Dakar, Senegal, were full-length PCR amplified and sequenced. Phylogenetic analyses indicated that, whereas three of these viruses were closely related and clustered overall within the HIV-1 subtype A radiation, they were distinct from previously characterized sub-subtype A1 and A2 viruses. The clustering pattern was maintained in the individual gag, pol, and env regions of the genome. Distance calculations between these viruses, which we termed A3, and other reference sub-subtype A1 and A2 viruses fell in the range of distances between previously characterized sub-subtype groups. In addition, we found evidence of two A3-containing recombinants in our cohort. These recombinants are mosaics composed of sequence from both sub-subtype A3 and CRF02_AG, the major circulating recombinant form in this West African population. Based on phylogenetic analyses, we propose that the group of viruses found in the Dakar sex worker cohort, previously referred to as HIV-1 A subcluster 2, be referred to as HIV-1 sub-subtype A3.

Two HIV-1 intersubtype recombinant forms are circulating widely in populations and have become important strains in the pandemic: CRF01_AE in Southeast Asia and CRF02_AG in West and West Central Africa, respectively. Several other circulating recombinant forms (CRF) have also been identified, but with fewer numbers of infections and/or more limited geographic spread. Here we expand knowledge of HIV-1 CRF using clinical samples, principally from West Africa, that were difficult to classify by partial genome sequencing. DNA was extracted from primary patient peripheral blood mononuclear cells (PBMC). The virtually complete HIV-1 genome was amplified by polymerase chain reaction (PCR) and directly sequenced. Additional strains were characterized by partial envelope sequencing. Phylogenetic analysis was used to identify and map intersubtype recombination breakpoints. Four virtually complete genome sequences and two partial envelope sequences represent CRF09_cpx, a newly identified complex recombinant HIV-1 whose principal focus seems to be in West Africa. This recombinant includes segments of subtypes A, F, G, and unclassified genetic material. It shares unique unclassified regions with the early Zaire strain Z321. There are similarities in structure, but considerable genetic distances, between CRF09_cpx and CRF02_AG IbNG. In conclusion, it is possible that this CRF shared common ancestors with both Z321 and CRF02_AG in the course of the pandemic, perhaps arising by recombination between earlier forms of these strains. Although newly identified, at least one infection with CRF09_cpx has already occurred outside of Africa.

CRF01_AE and subtype B have dominated the HIV-1 epidemic in Thailand since 1989. We reported a new circulating recombinant form of HIV-1, CRF15-01B, as well as other unique CRF01_AE/B recombinants among prevalent HIV infections in Thailand. We sought to study this challenging molecular picture through assessment of subtypes among recent HIV-1 seroconverters in northern Thai drug users. A total of 847 HIV-1 seronegative drug users (342 IDU and 505 non-IDU) were enrolled, from 1999 to 2002, in a prospective study; 39 HIV-1 incident cases were identified and characteristics were collected. The overall HIV-1 incidence rate was 2.54/100PY, but it was 10.0/100PY among male IDU. HIV was strongly associated with injection history; 38 of 39 seroconverters gave a history of IDU. A near full-length genome of HIV-1 was recovered by PCR amplification and sequenced from peripheral mononuclear cell extracted DNA of 38 seroconverters. Phylogenetic analysis revealed that 33 (86.8%) were CRF01_AE and 5 (13.2%) were CRF01_AE/B recombinants. These recombinants had different structure but shared some common breakpoints, indicating an ongoing recombination process. Recombinant infection increased with year of sampling (0 to 57.1 %). The molecular epidemiology of HIV-1 among drug users in northern Thailand has thus entered a new era. CRF01_AE remains predominant while pure subtype B is becoming rare, and now a substantial component of the epidemic. These findings support the need for CRF01_AE and subtype B components in clade-matched vaccine strategies for Thai phase III trials. Ongoing molecular surveillance of circulating HIV-1 strains is imperative for the evaluation of HIV vaccine efficacy.

The impact of HIV-1 genetic diversity on candidate vaccines is uncertain. One approach to minimize genetic diversity in the evaluation of HIV-1 vaccines is to match the vaccine sequence to the predominant subtype in a vaccine cohort. Over two million Ethiopians are infected with HIV-1, and the predominant subtype is thought to be subtype C. Understanding the phylogenetic relationships between sequences from Ethiopia and within subtype C can help decide what sequence(s) should comprise a candidate vaccine. To that end, nearly full genome sequencing was used to characterize HIV-1 from volunteers who emigrated from Ethiopia. DNA extracted from peripheral blood mononuclear cells (PMBC) was amplified using primers in the long terminal repeats to generate nearly full-length genomes. Amplicons were directly sequenced with dye terminators and automated sequencers. Sequences were phylogenetically analyzed by neighbor joining. The six new Ethiopian sequences were all subtype C, consistent with previous partial and full genome analysis. Together with two other Ethiopian sequences, the new sequences formed a geographic cluster when the complete genome was analyzed. However, subgenomic trees showed only a weak geographic cluster, or none, with respect to Ethiopian strains. Although immunological responses must be considered, from a phylogenetic perspective, there is no compelling support for use of Ethiopian subtype C sequences, compared to other subtype C, as vaccine prototype strains.

OBJECTIVE:To further define the genetic diversity of HIV-1 in Kenya using approaches that clearly distinguish subtypes from inter-subtype recombinants. DESIGN:Near full genome sequencing and analysis were used, including sensitive new tools for detection and mapping of recombinants. METHODS:Purified peripheral blood mononuclear cell DNA from 41 HIV-1 positive blood donations collected from six hospitals across southern Kenya was used to amplify near full-length genomes by nested PCR. These were sequenced on an ABI 3100 automated sequencer and analyzed phylogenetically. RESULTS:Among 41 near full-length genomes, 25 were non-recombinant (61%) and 16 were recombinant (39%). Of the 25 pure subtypes, 23 were subtype A, one was subtype C and one was subtype D. Most recombinants consisted of subtype A and either subtype C or subtype D; a few contained A2, a recently identified sub-subtype. Two A2/D recombinants had identical breakpoints and may represent a circulating recombinant form. A third A2/D recombinant had the same structure as a previously described Korean isolate, and these may constitute a second A2-containing circulating recombinant form. CONCLUSIONS:In Kenya, 93% of HIV-1 genomes were subtype A or A-containing recombinant strains. Almost 40% of all strains were recombinant. Vaccine candidates tested in Kenya should be based on subtype A strains, but the methods used for evaluation of breakthrough infections during future vaccine trials should be capable of identifying non-A subtypes, the A2 sub-subtype, and recombinants.

The impact of HIV-1 genetic diversity on candidate vaccines is uncertain. To minimize genetic diversity in the evaluation of HIV-1 vaccines, vaccine products must be matched to the predominant subtype in a vaccine cohort. To that end, full genome sequencing was used to detect and characterize HIV-1 subtypes and recombinant strains from individuals in Rakai District, Uganda. DNA extracted from peripheral blood mononuclear cells (PMBC) was PCR amplified using primers in the long terminal repeats (LTRs) to generate nearly full length genomes. Amplicons were directly sequenced with dye terminators and automated sequencers. Sequences were phylogenetically analyzed and recombinants were detected and mapped with distance scan and bootscan. Among 46 sequences, 54% were subtype D, 15% were subtype A, and 30% were recombinant. All recombinants were individually unique, and most combined subtypes A and D. Subtype D comprised more than 70% of all the HIV-1 genomes in Rakai when both pure subtypes and recombinants were considered. Candidate vaccines based on HIV-1 subtype D would be appropriate for evaluation in Rakai District, Uganda.

Four full-length genome subtype C sequences from South Africa, three of which are being used for vaccine development, were characterized. Three isolates were obtained from recently infected individuals in KwaZulu/Natal: Du151, Du422, and Du179. A fourth isolate, CTSc2, was obtained from an individual residing in Cape Town. All four strains used the CCR5 coreceptor, although Du179 also used CXCR4. The four isolates clustered within subtype C, but the three Du isolates formed a subcluster with a bootstrap value of 100%, with CTSc2 outside the subcluster. None of the strains showed evidence of intersubtype recombination, as expected from the predominance of subtype C in South Africa. All 4 isolates had a 16-amino acid truncation on the 3' end of the Rev protein, identified in other subtype C isolates. Like many other subtype C strains, Du151, Du422, and Du179 had three NF-kappa B-binding sites in the LTR; however, CTSc2 had only two.

In Thailand, the HIV-1 epidemic started abruptly in 1988 with the introduction of subtype B and subtype E, now called the circulating recombinant form (CRF), CRF01_AE. These two strains appeared independently in distinct high-risk populations [1] : subtype B among injecting drug users (IDU) and CRF01_AE among those who were heterosexually exposed [2,3]. HIV-1 subtype B is still common among infected Bangkok IDU, but CRF01_AE was found in 80% of IDU surveyed in 1995–1998 [4]. Dual infection with HIV-1 subtype B and CRF01_AE was observed by 1994 [5], providing the opportunity for recombination between these two subtypes in the Thailand epidemic. Whereas recombination between CRF01_AE and subtype B has occurred in an experimental dual infection of a chimpanzee [6], such a recombinant in humans has not yet been described. Here, we identify an AE/B inter-subtype recombinant of HIV-1 found in a multiply exposed individual in Thailand. The full-length genome of this recombinant has been analysed and characterized. In 1997, screening assays for the subtype of the virus of a 40-year-old Thai man (NP1623) provided evidence of subtype discordance in different parts of the genome. A V3 loop peptide enzyme immunoassay [7] classified the serum of NP1623 as CRF01_AE and an envelope heteroduplex mobility assay [8] confirmed that designation. A restriction fragment length polymorphism analysis from the gag leader region [9], however, indicated that NP1623 was infected with subtype B, as did a differential polymerase chain reaction (PCR) assay in gp41 [10]. These results suggested discordance between the subtype of gp120 and the subtype of the rest of the virus. Peripheral blood mononuclear cells (PBMC) were separated by Ficoll gradient and co-cultivated with phytohemagglutinin-stimulated donor PBMC. Full genomes of HIV-1 were amplified from the cultured PBMC DNA by nested PCR, with endpoint dilution of the DNA template in the first round. The DNA template was fully sequenced on both strands using BigDye terminator reaction kits and an ABI 373 DNA sequencer. A multiple alignment of the NP1623 full-length sequence with reference sequences of all HIV subtypes was generated. Bootscan analysis of the full genome sequence revealed that the virus had a recombinant structure with three segments (Fig. 1). Neighbor-joining phylogenetic analyses with parsimony bootstrap were performed on the segments of the genome and are shown in the upper panel of Fig. 1. The structure of the virus from NP1623 is as follows: subtype B from the beginning of gag until mid-vpu, where the subtype shifts to CRF01_AE. It then changes back to subtype B in the C5 region of gp120 and remains subtype B through gp41 and nef. The subtype B segments of the genome cluster most closely with the ‘Thai B’ sample, RL42. These same breakpoints were also confirmed in an independent amplification and sequencing of envelope directly from patient PBMC (Fig. 1, lower diagram).Fig. 1.: Recombinant analysis of full-length genome. Bootscan analysis of the full-length genome of NP1623 isolate using CRF01_AE (CM240), subtype B’ (RL42) as parental subtypes and subtype C (ETH2220) as the outgroup. The arrowheads mark the recombination breakpoints and the numbers refer to the location of the breakpoints corresponding to nt on HXB2 [11]. The upper panel shows the phylogenetic trees of sub-regions, which were constructed by neighbor joining. Parsimony bootstrap values are indicated at the nodes. The panel at the bottom shows the deduced structure of NP1623, both from culture and from primary peripheral blood mononuclear cells (PBMC), with respect to the HIV-1 genome structure. The full-length sequence from virus culture and the envelope sequence from primary PBMC of NP1623 are available under GenBank accession nos. AF362994 and 362995, respectively.The initial separation of CRF01_AE and subtype B in different risk groups in Thailand may have delayed the onset of significant numbers of dual infections, each of which can potentially lead to recombination, for almost a decade. The most significant factor in this respect may not be individuals who are exposed both heterosexually and through injecting drug use, as is the case reported here, but may rather be the growing proportion of CRF01_AE among an IDU population that, initially, was almost exclusively infected with subtype B. Indeed, we cannot discern whether the patient studied here was dually exposed by injecting drug use, was exposed to each strain by a different route, or was singly infected with the recombinant strain itself. The HIV-1 epidemic in southeast Asia is becoming more complex with respect to HIV-1 diversity. Recent reports include subtypes B, ‘Thai B', C, D, and CRF01_AE and a B/C recombinant in southern China. A CRF01_AE/subtype C recombinant has been detected in Thailand. The CRF01_AE/B recombinant reported here may be a harbinger of more recombinants. Intensified monitoring is particularly important in the light of the ongoing and projected HIV-1 vaccine trials. A significant fraction of recombinant strains among incident infections could necessitate an approach such as that used here, with multiple genetic regions analysed and discordances followed up with full genome sequencing, to evaluate the relative effectiveness of vaccines against different HIV-1 subtypes. Acknowledgements The authors would like to thank Puangmalee Buapunth and the staff of the Joint Clinical Research Center in Bangkok for their invaluable assistance. The views and opinions expressed herein do not necessarily reflect those of the US Army or of the Department of Defense. Sodsai Tovanabutraab Victoria Poloniscd Mark De Souzacd Rapee Trichavarojd Penprapa Chanbancherde Bohye Kimb Eric Sanders-Buellb Sorachai Nitayaphanf Arthur Brownd Merlin R. Robbg Deborah L. Birxg Francine E. McCutchanb Jean K. Carrb

The genetic diversity of group M HIV-1 is highest in west central Africa. Blood samples from four locations in Cameroon were collected to determine the molecular epidemiology of HIV-1. The C2-V5 region of envelope was sequenced from 39 of the 40 samples collected, and 7 samples were sequenced across the genome. All strains belonged to group M of HIV-1. The circulating recombinant form CRF02 AG (IbNG) was the most common strain (22/39, 56%). Two of these were confirmed by full genome analysis. Four samples (4/39, 10%) clustered with the sub-subtype F2 and one of these was confirmed by full genome sequencing. Recombinant forms, each different but containing subtype A, accounted for the next most common form (7/39, 18%). Among these recombinants, those combining subtypes A and G were the most common (4/7, 57%). Also found were 3 subtype A, 2 subtype G, and 1 subtype B strain. Many recombination break points were shared between IbNG and the other AG recombinants, though none of these other AG recombinants included IbNG as a parent. This suggests that there was an ancestral AG recombinant that gave rise to CRF02 AG (IbNG), the successful circulating recombinant form, and to others that were less successful and are now rare.

OBJECTIVE:To improve our understanding of the genetic complexity of HIV-1 subtype A by increasing the number of subtype A isolates that have been sequenced in their entirety. METHODS:Nine HIV-1-seropositive patients from Africa living in Sweden contributed peripheral blood mononuclear cells (PBMC) for this study. Sequencing of the C2-V3 region of env had shown them to be subtype A. DNA from virus cultures was used for the amplification of virtually full-length proviral sequences, and the resulting fragment was sequenced. RESULTS:Six of the nine viral isolates were subtype A throughout the genome, or non-recombinant, and all of these were from east Africa. One virus from the Ivory Coast had the AG(IbNG) genetic form, a recombinant form common in west Africa. Two of the isolates were novel recombinants: one was an A/C recombinant and the other was A/D. Analysis of gag reveals three subclusters within the A subtype: one containing the AG(IbNG) subtype viruses, one containing the AE(CM240) viruses and one containing the non-recombinant A viruses. These genetic clusters have different geographical distributions in Africa. CONCLUSION:The prevailing view of HIV-1 subtype A forming a uniform band across the center of sub-Saharan Africa needs revision. In all probability, the most common subtype in west Africa and west central Africa is the AG recombinant, AG(IbNG), whereas in east central Africa it is the non-recombinant subtype A.