Tracking a Century of Global Expansion and Evolution of HIV to Drive Understanding and to Combat Disease

Review

Tracking a century of global expansion and evolution of HIV to drive understanding and to combat disease

Denis M Tebit, Eric J Arts

Since the isolation of HIV, multiple transmissions are thought to have occurred between man and other old-world Lancet Infect Dis 2011; 11: 45–56 primates. Assessment of samples from apes and human beings with African equatorial forest ancestry has traced the Published Online origin of HIV-1 to chimpanzees, and dated its most recent common ancestor to 1908. The evolution of HIV-1 has been December 1, 2010 rapid, which has resulted in a complex classifi cation, worldwide spread, and intermixing of strains; at least DOI:10.1016/S1473- 3099(10)70186-9 48 circulating recombinant forms are currently identifi ed. In addition to posing a nearly insurmountable challenge See Online/Comment for diagnosis, treatment, vaccine development, and prevention, this extreme and divergent evolution has led to Lancet DOI:10.1016/S0140- diff erences in virulence between HIV-1 groups, subtypes, or both. Coincidental changes in human migration in the 6736(10)62180-0 and Congo river basin also aff ected spread of disease. Research over the past 25 years and advances in genomic sequencing Lancet DOI:10.1016/S0140- methods, such as deep DNA sequencing, have greatly improved understanding and analysis of the thousands to 6736(10)62183-6 millions of full infectious HIV-1 genomes. Division of Infectious Disease, Case Western Reserve University, Cleveland, OH, USA Introduction capped mangabey and greater spot-nosed monkey, or a (D M Tebit PhD, Prof E J Arts PhD) 12 HIV is highly heterogeneous within infected individuals closely related species. Characterisation of SIVcpz has Correspondence to: owing to rapid turnover rates, high viral load, and an error- been complicated by the presence of lentiviruses in more Prof Eric J Arts, Division of prone reverse transcriptase enzyme that lacks proofreading than 30 species of non-human primates in sub-Saharan Infectious Disease, Case Western 1–4 13 Reserve University, 10900 Euclid activity. High variability is also the consequence of Africa. The SIVs could have crossed to man multiple Avenue, Cleveland, OH 44106, recombination, which can shuttle mutations between viral times over several decades and led to divergence (fi gure 1). USA genomes and lead to major antigenic shifts or alterations On the basis of phylogenetic analysis HIV has been [email protected] in virulence.5–7 Ultimately, therefore, the continual, classifi ed into two types—HIV-1 and HIV-2. The latter is divergent evolution of HIV-1 in man to epidemic levels separated into eight groups, of which A and B are the over the past 100 years originates from the swarms of most prominent, and HIV-1 into the groups M (main), N HIV-1 strains (or quasipecies) within each human host. (non-M, non-O), and O (outlier).14,15 Group M is further Since the isolation of HIV-1 in the early 1980s, rapid categorised into nine subtypes (A–D, F–H, J, and K), sub- development and application of various molecular tools, subtypes, and 48 CRFs (table 1, fi gure 1)16 that are such as nucleic acid isolation and purifi cation techniques identifi ed by numbers (ascending in order of discovery) in frozen, faecal, and urine samples or paraffi n- followed by letters of the parental subtypes. The origins fi xed tissue samples, have substantially improved of HIV-1 groups M and N have been traced to SIV-infected understanding of the origins and evolution of HIV-1.8,9 In Pan troglodytes troglodytes (Ptt) chimpanzees inhabiting the fi rst two decades of HIV-1 research, sequencing was the eastern equatorial forests of Cameroon, in west limited to short HIV-1 genomic regions, which could only central Africa.17–20 The same geographical region was provide crude estimates of the geographical distribution probably also the site of origin for group O HIV-1; the of HIV-1 subtypes and gene evolution. Advances in closest SIV relative was found in gorillas (Gorilla gorilla; amplifi cation and sequencing of the complete HIV-1 SIVgor), although chimpanzees are likely to have been genome, however, have enabled specifi c classifi cation of the original hosts (fi gure 1).21 Another HIV-1 group-O-like HIV-1 subtypes and recombinants. Within infected variant that is more closely related to SIVgor than to individuals analyses of quasispecies (ten to 50 clones) group O, has been identifi ed in a Cameroonian living in typically relied on bacterial cloning or sequencing or on France and has been designated to a tentative new group, single-genome amplifi cation, but thousands of clones P.15 HIV-2 originated from SIV in sooty mangabeys may now be analysed by pyrosequencing methods.10,11 (Cercocebus atys; fi gure 1).22–24 This Review summarises the emergence of new HIV By use of molecular clocks, the estimated times of the strains in the worldwide pandemic, with emphasis on most recent common ancestors of HIV-1 groups M, O, the circulating recombinant forms (CRFs), discusses the and N in central Africa are 1908 (range 1884–1924), 1920 progress made in the methods used to track the global (1890–1940), and 1963 (1948–77), respectively, and for molecular evolution of HIV, and appraises the importance HIV-2 groups A and B the dates are 1932 (1906–55) and of these new strains and methods in the future control 1935 (1907–61), respectively (fi gure 2).25–31 The HIV-2 and prevention of HIV. epidemic probably started in Guinea Bissau (fi gures 1 and 2),23,25–27,32 but Santiago and co-workers33 reported Origin and classifi cation signifi cant clustering of HIV-2 group A and groups with The likely progenitor of HIV-1, simian immunodefi ciency strains of sooty mangabey SIV in the Tai forest, virus (SIV) in chimpanzees (SIVcpz), seems to be a Côte d’Ivoire. HIV-1 and HIV-2 both spread exponentially recombinant virus derived from lentiviruses of the red early in the epidemic, but the patterns of infection have www.thelancet.com/infection Vol 11 January 2011 45 Review

Homo sapiens Pan troglodytes troglodytes (chimpanzee)

Group M B-like F-like

e B D-like

CRF05_DF

F1 Subtyp CRF19_cpx CRF03_AB Subtype D F2 CRF10_CD Subtype H 1900s K Subtype C G C-like CRF07_BC G-like CRF14_BG CRF08_BC CRF13_cpx CRF11_cpx J CRF06_cpx SIVcpz J-like CRF18_cpx CRF09_cpxA2 Group N A1

CRF02_AG CRF01_AE

A-like SIVcpz

1920s SIVgor

Group P (RBF 168) 1930s

Group O

CercocebusCercocebus atysatyys (sooty(sooty mangabey)manggabbeyy) Gorilla gorilla 0·050·05 HIV-2

Figure 1: Relations between and genetic diversity in HIV-1 groups M, N, O, and P, HIV-2, and SIVs, and patterns of cross-species transmission CRF=circulating recombinant form. cpz=chimpanzee. gor=gorilla. cpx=complex. SIV=simian immunodeﬁ ciency virus.

46 www.thelancet.com/infection Vol 11 January 2011 Review

diff ered over time: HIV-1 group M currently accounts for Circulating recombinant forms (geographic distribution) more than 30 million infections, but HIV-2 presently accounts for fewer than 1 million of all HIV infections;34–37 A-like CRF01_AE (Africa, Asia), CRF02_AG (Africa), CRF16_A2D (Kenya, Korea), CRF22_01A1 (Cameroon), CRF35_AD (Afghanistan), CRF45_cpx (central Africa) in Guinea Bissau, the epicentre of HIV-2, prevalence has 38 39 B-like CRF03_AB (eastern Europe), CRF20_BG (Cuba), CRF28,29,39_BF dropped from 8·9% in 1987, to 7·4% in 1996, to 4·4% (South America) in 2006, whereas that for HIV-1 has increased from 2·3% C-like CRF07,08,31_BC (China, Taiwan) 40 in 1996 to 4·6% in 2006. D-like CRF10_CD (east Africa), CRF19_cpx (Cuba), CRF21_A2D (Kenya), HIV-1 groups M, N, O, and P are phylogenetically CRF41_CD (pending) interspersed along SIVcpzPtt and SIVgor lineages, which F-like CRF05_DF (Belgium), CRF12,17,40_BF (South America); CRF42,44,46,47 suggests they arose from four independent ape-to-human (South America, pending) transmissions.18,41,42 The distributions in human beings G-like CRF11_cpx (central Africa), CRF14,23,24_BG (Cuba, Europe, Asia) diff er strikingly: group O infections are most concentrated H-like CRF04_cpx (Cyprus, Greece), in Cameroon, with spread being restricted mainly to J-like CRF13_cpx (central Africa), CRF18_cpx (Cuba, central Africa) neighbouring central African countries; groups N and P K-like None have been found in a small number of Cameroonians; and CRF01_AE-like CRF33,34_01B (Asia) group M strains have spread worldwide and multiple CRF02_AG-like CRF09_cpx (west and central Africa), CRF43_02G (Saudi Arabia) 27,32 subtypes have been identifi ed (fi gure 2). Group M CRF06_cpx-like CRF32_06A1 (Estonia) viruses encode protein sequences in gag, pol, and env that Undefi ned or CRF06_cpx (west Africa), CRF15_01B (Thailand), CRF17_BF, CRF25_cpx have 14–35% divergence from their closest known SIVcpz independent cluster (Cameroon), CRF26_AU (pending), CRF27_cpx (central Africa, France), relatives.43 Genetic diversity is frequent but the CRF30_cpx (Niger), CRF36_cpx (Cameroon), CRF37_cpx (Cameroon), CRF38_BF (Uruguay) consequences are unknown, although Wain and colleagues43 identifi ed a viral genetic change in the p17 Table 1: Classifi cation and distribution of various circulating recombinant forms, according to matrix protein encoded by the gag gene that might have phylogenetic clustering with pure (non-recombinant) subtypes or fi rst-generation recombinants facilitated the adaptation of SIVcpz to its human host. In particular, the aminoacid residue 30 of Gag has a conserved continued to evolve into sub-subtypes—for instance Arg or Lys in all HIV-1 groups (except subtype M-C, which A1–A4 and F1–F2 (fi gure 1),16 that form distinct lineages has a conserved Met), whereas Met30 is seen in Gag of within a given subtype but that have lesser degrees of SIVgor and SIVcpzPtt, and Lys30 is seen in SIV-infected genetic divergence. Pan troglodytes schweinfurthii.43 HIV-1 with Met30Lys Gag Sub-Saharan Africa bears the highest burden of HIV-1 in replicated better in human peripheral blood mononuclear terms of prevalence and diversity (fi gure 3).37,73–76 The cells (PBMCs) than in chimpanzee PBMCs.44 The matrix epidemics in west and central Africa seem to have stabilised protein could, therefore, modulate virus fi tness after cross- in prevalence, but these regions, along with the Congo species transmission, with primate lentiviruses being river basin, continue to be hot spots for HIV diversity subject to host selection pressure. (fi gure 3). Most, if not all, subtypes, sub-subtypes and CRFs have been reported in the DRC and Cameroon.71,77–80 Molecular epidemiology As in the DRC, HIV-1 strains in Angola are highly diverse, Group M causes most HIV-1 infections, owing to its high and classifi cation into sub-subtypes A5 and A6 might be numbers of subtypes and CRFs. These subtypes form required.81 From Cameroon, moving westwards to Nigeria, phylogenetic clusters with aminoacid diff erences of HIV-1 diversity decreases, as shown by the dominance of 25–30% in env, 20% in gag, and 10% in pol.45 Some CRF02_AG subtypes A and G (fi gure 3). This trend subtypes are linked geographically. Variation within continues with western migration to Côte d’Ivoire, Ghana, group M is greatest in the Congo river basin, which is Senegal, and Mali, where CRF02_AG predominates, with probably the site of initial zoonotic jumps and regional reports of isolated cases of CRF06_cpx (complex diversifi cation (table 2,8,9,12,15,17–22,32,42,46–69 fi gure 2).32,46,70,71 Two subtype).82–85 Finally, CRF06_cpx and the second-generation HIV-1 sequences—a 1960 sample from the current recombinants CRF02_AG/CRF06_cpx dominate the Democratic Republic of Congo (DRC) and a 1959 sample epidemic in Burkina Faso, with subtypes A and G rarely from the DRC (labelled as being from Zaire)—have being reported.86–90 helped to root the subtype A and D phylogenetic lineages, Although the HIV-1 genetic diversity is high in west respectively, and to age the epidemic in the region and central Africa, HIV-1 prevalence remains surprisingly (fi gure 2).32,46 Initial subtype distribution indicated lower than in most other regions of sub-Saharan Africa dominance of subtype B in the western world and of (fi gure 3). The highest prevalence shifted in the late 1990s subtype A in sub-Saharan Africa. Subtype A eventually from east Africa (Uganda, Kenya, and Tanzania) to the extended into the former Soviet Union (fi gure 2). In the southern African region. On average, close to 20% of the past 15 years, however, the rapid emergence of new human population in South Africa, Lesotho, Botswana, subtypes and intermixing of strains has altered the and Zimbabwe are thought to be infected with HIV-1.37 geographical distribution of subtypes.14,72 In addition, This shift provides strong evidence for the founder-eff ect some pre-existing subtypes, such as A and F, have theory (a single introduction followed by a rapid spread) www.thelancet.com/infection Vol 11 January 2011 47 Review

M–B western Europe late 1970s/1980s Thailand ~1980s

M–B USA 1972/~1970s P France 2005

O USA ~1985

M–C Ethiopia mid-1980s Israel 1990s

M–F Romania mid-1980s

O Cameroon/Gabon to France/Germany ~1985 Spain 1990s Norway ~1960

M–B South America mid-1980s M–A Russia late 1980s Eastern Europe/Asia early 1990s Haiti M–B Haiti 1966 M–A Uganda/Kenya late 1970s/early 1980s M–A P M–D Uganda/Kenya O M–F late 1970s/early 1980sM–C South Africa late 1980s India/China 1990s M–C Brazil 1990s M–B (D) M–D M–01 Thailand mid-1980s M–F Argentina 1991 M–C

M–CRF01 AE (A)

Congo river basin

HIV-1 group M HIV-2 groups A and B HIV-1 M subtype C HIV-1 subtype B HIV-1 subtype B Common ancestor of common ancestor25 common ancestor25 common ancestor28 introduced into Haiti29 introduced into USA Thai subtype E27 from Haiti29

1908 1920 1930–35 1950 1958 1963 1966 1971 1972 1985 1984–86

HIV-1 group O HIV-1 M–B (D) starts HIV-1 group N HIV-1 sub-subtype F1 HIV-1 subtype B´ common ancestor26 to diverge in human common ancestor29 introduced into introduced into Asia31 beings27 South America30

Figure 2: Estimated time line of global evolution and spread of HIV types, groups, and subtypes Enlarged parts of map show the main disease epicentres. The time line indicates the key events in the evolution of HIV-1 groups M, N, and O and of HIV-2. CRF=circulating recombinant form.

since the southern African epidemic is due almost of its divergence from the subtype B that occurs in the entirely to the spread of HIV-1 subtype C (fi gure 3).89 Americas (fi gure 2). Deng and colleagues31 reported this Independent, rapid spread of subtype C in east Asia has divergence occurred about 15 years after the B subtype also contributed to this subtype being responsible for began to spread, which roughly coincides with CRF01_AE more than 51% of all HIV-1 infections worldwide being introduced into Thailand (fi gure 2). CRF01_AE has (fi gure 4).90 Although initially absent, HIV-1 subtype C now gained dominance in Thailand; likewise, subtype C now circulates at low levels as a pure (non-recombinant) is currently dominant in most east Asian countries.94,95 subtype or a recombinant form in Kenya and Uganda. Subtype C was reportedly introduced into India from The overall HIV-1 incidence is, however, decreasing South Africa, and then into China from India (fi gure 2).95 (fi gures 2 and 3).91–93 Subtypes B and C have recombined to form CRF07_BĆ The HIV epidemic in Asia is dynamic and all subtypes and CRF08_BĆ in China (fi gure 4). The northern triangle that circulate are due to multiple founder events.16 Subtype of Burma seems to have the greatest HIV-1 diversity in B was the fi rst introduced into Asia in the mid-1980s, and Asia, with seeding of CRF01_AE and subtype B´ from was seen mainly in China, India, and Thailand. This Thailand, possibly subtypes A, B, and C from India, and strain has been named B´ (also known as Thai B) because CRF07 and CRF08_BĆ, from Yunnan province in China.

48 www.thelancet.com/infection Vol 11 January 2011 Review

A shortage of well organised needle exchange programmes Event or fi nding for injecting illicit-drug users in this region could contribute to the complexity of HIV-1 subtype 1983 Isolation of HIV-18 recombination there and further afi eld. 1984 CD4 identifi ed as HIV receptor52 AIDS reported in Africa54,55 In eastern Europe the breakdown of the former Soviet 1985 Sequence of HIV and genetic diversity reported9,56 Union has coincided with a rise of HIV-1 infections and, Development of the fi rst blood screening assays57 most notably, the spread of sub-subtype A1, which has 1986 Isolation of HIV-222 been linked to intravenous drug use, and of subtype B 1987 First antiretroviral drug, zidovudine, approved by the FDA58 (and to a lesser extent CRF03_AB), mainly through sexual 1989 Second HIV drug, didanosine, approved by the FDA58 transmission. In western Europe, as in North America 1990 First report of a divergent form of HIV-1 (group O)50 and Australia, subtype B predominates. The prevalence First full-genomic sequence of HIV-related virus in chimpanzees59 of non-B strains has, however, increased owing to the 1991 Zalcitabine approved by the FDA; zidovudine and didanosine used in combination for the fi rst time58 73,96,97 infl ux of immigrants from Africa and Asia. Portugal 1993 Transmission of zidovudine-resistant virus documented60 has the highest prevalence of HIV-2 and other non-B 1994 Detailed characterisation of two reference group O viruses49,51 subtypes in Europe, which might be related to the 1996 Discovery of CXCR4 and CCR5 as HIV co-receptors61,62 colonial war in Angola in the mid-1970s.81 Substantial protection from HIV infection by homozygous 32 bp deletion in CCR5 gene63 In South America, prevalence and diversity of HIV-1 are Introduction of combination highly active antiretroviral therapy 64 highest in Brazil and Argentina, with substantial 1997 Group O isolated from 1970s frozen Norwegian human samples65 circulation of subtypes B, C, and F, and BC and BF 1998 Sequencing of a 1959 HIV-1 sample from the DRC46 66 recombinants. Two reports showed close relations between Identifi cation of HIV-1 group N 1999 First detailed proof that HIV-1 originated from SIVcpzPtt19 the subtype C viruses in South America and those in 47,48 95 98 Description of group M/O recombinant viruses in two unlinked patients Kenya, Ethiopia, and Burundi. The subtype C epidemic Single-dose nevirapine shown to reduce transmission of HIV from mother to child67 28 is estimated to have originated in 1958 (fi gure 2), and was 2000 Description of SIVcpz env sequences similar to those in HIV-1 group N17 99 fi rst reported from Ethiopia in 1990 (fi gure 2), from 2002 SIVcpz identifi ed in wild chimpanzees18 where it spread to Israel; the virus also spread from eastern Discovery of APOBEC-3G, an anti-HIV host factor68 Africa to Brazil, and then to Argentina and Uruguay 2003 SIVcpz described as a recombinant between SIVrcm and SIVgsn12 95 (fi gure 2). Subtype C was thought to originally have been 2004 Discovery of TRIM5 α, a factor that prevents HIV from infecting monkey cells69 introduced to Brazil from Mozambique, but this theory is 2005 High SIV prevalence in wild-caught chimpanzees in equatorial Africa42 100,101 not supported by phylogenetic analysis. The subtype C 2006 Confi rmation that HIV-1 group N and M exist in chimpanzees in equatorial forests of Africa 20 epidemic is now the fastest emerging epidemic in South 2006 HIV-1 group-O-like sequences found among gorillas in southeastern Cameroon21 America: in the Rio do Sol region of southern Brazil, 2008 Sequencing of a 1960 HIV strain from a paraffi n-embedded tissue sample from the DRC32 subtype C or a BC recombinant accounted for 35% of 2009 Identifi cation of gorilla-like group O sequences in human beings, named group P15 cases in 1996, 52% in 2002,102 and nearly 59% in 2008,103 and SIVcpz infection is pathogenic in wild eastern chimpanzee sub-species53 the current estimate is nearly 70%.104 FDA=Food and Drug Administration. DRC=Democratic Republic of Congo. SIV=simian immunodefi ciency virus. The HIV-1 epidemic in the Caribbean might be the clearest refl ection of diff erent founder events, given Table 2: Some important milestones in the evolution of HIV notable immigration, travel, and trade. Longstanding links between the DRC and Haiti and between Angola and Cuba might explain the presence and age of certain Recombination HIV-1 strains in the Caribbean. Gilbert and co-workers29 CRFs and unique recombinant forms (URFs) are created applied a relaxed molecular clock to sequences derived after co-infection with at least two diff erent HIV-1 isolates. from a 1982 Haitian DNA sample and calculated that Initially, identifi cation of multiple infections was diffi cult subtype B was introduced into Haiti between 1962 and and reports were rare. New technologies, such as the 1970, probably from the DRC, and from Haiti into the heteroduplex tracking assay and real-time PCR, however, USA in 1972 (fi gure 2). Likewise, the oldest known HIV-1 improved detection. URFs comprise more than 30% of subtype B infection in the entire epidemic was identifi ed infections in regions where several HIV subtypes in a sample from 1982 or 1983, from Haitian immigrants co-circulate (fi gure 3).106 Isolates were initially classifi ed in the USA (fi gure 2). Unlike the Haitian epidemic, that into diff erent subtypes on the basis of clusters of partial in Cuba is diversifi ed, with CRF18_cpx and CRF19_cpx gag and env sequences. A shared node in a phylogenetic and subtype C all circulating, but is marked mainly by tree suggested a common ancestry. Sequencing of longer the CRF20, CRF23, and CRF24, for which subtypes B and near-full length of HIV genomes has led to clearer and G are parental (table 1).105 HIV-1 was probably understanding of lineages. CRFs and URFs have genome introduced into Cuba via troops who fought in the Angola segments derived from more than one subtype (fi gure 4), war of independence.81 Relative to Cuba and Haiti, HIV-1 but members of a CRF group share the same mosaic has recently arrived through true founder event in other HIV-1 genomic structure and are derived from at least Caribbean islands, such as Trinidad and Tobago and the three epidemiologically unlinked ancestors, whereas Dominican Republic. URFs do not. CRFs currently comprise about 20% of all www.thelancet.com/infection Vol 11 January 2011 49 Review

West Africa Proportion of HIV-1 subtypes in Africa Ethiopia/Eritrea/Somalia/Djibouti 1 000 000 CRF04–18 1 000 000

URF

CRF01_AE A1 CRF02_AG Human population K J 100 000 H 100 000 G Human population D F2 C

10 000 Total infected 10 000

02_AG G Total infected Subtype C 1000 A1 1000

Number of people (thousands) Number URFs 06_ of people (thousands) Number 100 D 100 11_ C F1 K J 01_AE 10 10

0 0 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Years Years Congo River Basin Southern Africa East/central Africa 1 000 000 1 000 000 1 000 000

Human population Human population 100 000 100 000 100 000 Human population

Total infected C 10 000 10 000 10 000 Total infected

Subtype A1 Total infected D URFs 1000 1000 1000 02_AG A1 C URFs G D Number of people (thousands) Number Number of people (thousands) Number Number of people (thousands) Number C 100 F2 100 D 100 H A1 01_AE J URFs 05_ G F1 11_ A2K 18_ 10 10 10

0 0 0 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Years Years Years

Figure 3: Evolution of HIV prevalence and genotypes in ﬁ ve regions of sub-Saharan Africa from 1990 to 200737,75,76 CRF=circulating recombinant form. URF=unique recombinant form.

50 www.thelancet.com/infection Vol 11 January 2011 Review

Worldwide prevalence of HIV-1 subtypes, CRFs, and URFs LTR gag vif env LTR CRF12_BF* pol CRF012_BF tat CRF07_BC, CRF08_BC CRF09_48 vpr vpu nef tat/rev rev CRF06_cpx CRF07_BC* CRF02_AG CRF01_AE

CRF06_cpx URFs

A1 A1

G CRF02_AG G B B D

CRF01_AE

C Subtype: colour code for subtypes in chimeric genome A B C E F G J K Unclassiﬁed

Figure 4: Mosaic structure of the most dominant circulating recombinant forms in the HIV pandemic, and the worldwide prevalence of HIV-1 subtypes, circulating recombinant forms, and unique recombinant forms LTR=long terminal repeats. CRF=circulating recombinant form. URF=unique recombinant form. *Has the same parental strain as at least one other circulating recombinant form but has a diff erent recombinant structure. known HIV infections.14,107,108 However, if CRF02_AG might not remain dominant in Argentina, and other (~2 million infections) and CRF01_AE (~1·6 million BF CRFs might emerge with greater transmissibility. To infections) are taken to be pure subtypes, on the basis of further complicate the epidemic in Argentina, subtypes A, phylogenetic reclassifi cations,73,75,109 this proportion drops B, C, and F cause around 20% of HIV infections, and to around 10%. BF CRFs cause the other 80%.30 The fi rst HIV-1 isolate to be identifi ed as a recombinant Initially, HIV-1 groups and types were thought too was MAL, a strain derived from subtypes A and D that was diverse to recombine even after potential dual infection, obtained from a 1976 blood sample from the DRC.70,71 Of but there have been three reports of M/O the 48 CRFs described so far,16 12 (25%) have genomic recombination.47,48,112 In one case, generation of the M/O structures incorporating sequences from more than two recombinant led to the gradual elimination of the HIV subtypes (complex recombinants; table 1). For parental isolates.48 The HIV-1 recombination sites example, CRF06_cpx includes genomic segments with selected during dual infection are thought to be more subtypes A, G, J, and K sequences. The origin of this CRF related to the mechanics of strand transfer during is unknown, but Burkina Faso, where CRF06_cpx is reverse transcription and selection of replication- dominant, might have been an HIV-1 epicentre from competent chimeric virus than to sequence conservation which this strain spread to neighbouring countries, such between the strains.113 Thus, given suffi cient opportunity as Niger, Mali, Côte d’Ivoire, and Nigeria, and through a through multiple dual infections within one or more founder event in Kaliningrad, Russia (table 1).87,88,106 In host species over hundreds of years, recombination south and southeast Asia (except India), CRFs and other between distantly related SIV strains with diff erent recombinants comprise about 88% of circulating strains.73 target hosts might have been possible. SIVcpz, for In particular, CRF07_BĆ and CRF08_BĆ dominate the which SIV in the greater spot-nosed monkey is one of epidemic in China to the extent that the prevalence of its parental strains, is a recombinant in the 3´ end of the subtypes C and B is very low.110 genome, which includes vpu and env genes.12 HIV-1 The well described CRF12, CRF17, CRF28, CRF29, group N arose from recombination between an SIVcpz CRF39, CRF40, CRF42, and newly identifi ed CRF44 and group-N-like strain and a progenitor of HIV-1 group M, CRF46–48 are all derived from subtypes B and F, and have which possibly infected a subspecies of chimpanzees been circulating in Argentina and other bordering before being transmitted to man. Given cross-species countries since the 1980s; the earliest identifi cation was in introductions, or possibly jumping, of SIV between children (table 1).16,111 Most BF recombinants seem to be in non-human primates, a future recombination event an early transition phase, changing from newly generated, between HIV-1 and HIV-2, groups N and O, or even transmissible URFs to more-stable CRFs. Thus, CRF12_BF with other SIV strains, cannot be ruled out.17,114 www.thelancet.com/infection Vol 11 January 2011 51 Review

Genetic assessment evolution of primate lentiviruses, samples must be The understanding of the worldwide rate and direction of collected from the remaining old-world primate HIV-1 evolution will greatly aff ect future methods and species.20,29,32 Collection from HIV-infected human beings strategies for diagnosis, therapy, and prevention. Serology is easy, but non-invasive collection of samples from by commercially available ELISA or EIA is the most monkeys in the wild has proven extremely diffi cult, widely used diagnostic method. Improvements in although sampling techniques, such as the collection of methods to identify new and diverse HIV-1 isolates urine and faecal samples from apes, have improved.20 means that the panel of necessary proteins to test has Furthermore, analyses of current samples might not substantially expanded.49–51 Fourth-generation HIV identify the most recent common ancestors of HIV-1 immunoassays now detect p24 antigen and HIV-specifi c closer to the zoonotic transmission. Generation of antibody simultaneously, and can identify most infections lentiviral sequences from stored human tissue samples, with group M and O strains.115 ELISA does, however, lack such as Bouin-fi xed, paraffi n-embedded tissue samples sensitivity and specifi city to identify serotypes. from west and central Africa, that predate the rapid DNA sequencing is highly eff ective, but given the initial expansion of AIDS is crucial. high costs and the type of equipment and advanced Finally, phylogenetic methods and algorithms have training needed, this approach was not routine in Africa been improved along with sampling and extraction during the 1990s and remains prohibitive in the techniques. For example, new Bayesian methods enabled developing world. The heteroduplex mobility assay the co-estimation of HIV-1 divergence times under relaxed off ered a more-aff oradable option for simple and rapid or strict molecular clock models, and have been essential classifi cation of HIV-1 subtypes,116–118 and was used in for dating the introduction of diff erent HIV-1 types, many areas with the help of workshops provided by groups, and even some subtypes of HIV-1 group M.25,27,32,119 WHO. However, although this method has been improved The development of second-generation DNA sequencing with the use of radiolabelled probes, heteroduplex methods has, however, quickly stretched the capabilities analyses cannot defi ne specifi c sequence diff erences of most phylogenetic algorithms. between isolates of the same or diff erent subtypes. Direct DNA sequencing is appropriate to characterise Phenotypes the infecting HIV-1 subtype or recombinant form and to The extreme diversity between HIV-1 groups or subtypes monitor regional and global HIV-1 spread. Although DNA has continually raised questions as to whether genotypes sequencing of a bulk PCR product remains less expensive are associated with any specifi c biological or phenotypic and faster to perform than a clonal DNA sequence traits. Within HIV-1 groups and subtypes, isolates use analysis, minor HIV-1 variants (frequency <20–30%) diff erent co-receptors (CCR5 and CXCR4) in association cannot be detected. The development of technologies able with CD4 for HIV-1 entry.52,120 In newly infected and to process massive DNA sequences in parallel has obvious asymptomatic individuals most viruses are those that use advantages for detecting minor HIV-1 variants. The Solexa CCR5, but as the disease progresses, variants that use the sequencing approach by Illumina (San Diego, CA, USA) CXCR4 co-receptor, or both CCR5 and CXCR4, can provides the greatest depth in clonal sequencing with the emerge and dominate in the HIV-1 population, although generation of 1 Gbp DNA sequence per lane of analyses. the pattern can diff er between subtypes. A switch to However, a drawback of this technology is poor subtyping CXCR4 usage has been associated with acceleration of and CRF identifi cation because of short sequence reads disease progression. The ability to use either or both of (ten to 50 nucleotides) and lack of linkage to reconstruct these co-receptors for viral entry is typically conferred by HIV-1 genes or genomes. The Genome Sequencer FLX discrete genetic variation in the V3 loop of the HIV-1 Titanium series (454 Life Sciences, Roche Applied Science, envelope glycoprotein. Subtype C strains predominantly Branford, CT, USA) provides longer sequence reads use CCR5 and rarely switch to CXCR4 or dual tropic use. (100–500 nucleotides), but the longest current sequencing By contrast, subtype D strains use CXCR4 receptors platform is the PacBio RS (Pacifi c Biosciences, Menlo earlier and more frequently in infection than other Park, CA, USA), which provides read lengths longer than HIV-1 subtypes, which might be why individuals infected 1000 nucleotides with low error rates. If appropriately with this subtype progress rapidly to AIDS.121,122 Such applied and with accurate tags for sequence identifi cation, diff erences in transmission and disease progression this technology can be used to reliably estimate HIV-1 could clearly impact the global distribution and prevalence population diversity within patients, provide a high- of these viral strains. throughput alternative for HIV-1 subtype identifi cation A more general question relates to the possible impact and track disease spread, readily identify dual infection of HIV-1 evolution on disease and global virus with HIV-1 subtypes or recombinants, and sequence full distribution. This topic has not been the subject of HIV-1 genomes. intense investigation because comparing the natural Even with these extreme advances in DNA sequencing history of infections by diff erent HIV-1 subtypes could technology, limitations still exist in relation to sample require 5–20 years or longer of intense follow-up and the collection. To understand the global and historic absence of treatment. Although disease progression in

52 www.thelancet.com/infection Vol 11 January 2011 Review

HIV-1 subtype B infections is at least partly understood, contrast to the pathogenic fi tness model where subtype C infections with the dominant subtype A, C, and D, as HIV-1 isolates replicated with poor effi ciency in PBMCs. well as CRF02_AG, have been poorly studied, partly Some host-pathogen models suggest that pathogen because they are highly prevalent in resource-poor spread is associated with duration of infection, settings. As antiretroviral treatment is now widely transmissibility, and opportunity for transmission.135 For available, these studies might never be ethically feasible. example, as well as being the dominant strain in the However, several studies in Uganda and Kenya have initial founder events, subtype C causes disease with a suggested progression to AIDS is faster in people with long asymptomatic period, which increases opportunity HIV-1 subtype D than in those with subtype A for transmission compared with that of other group M infections.121,123 In another study subtype C infections in subtypes. If increased opportunity for transmission Zimbabwean women progressed two to three times remains during chronic disease (5–15 years), it could slower than subtype A and D infections in Ugandan counteract high transmissibility of other subtypes soon women, on the basis of CD4 cell counts, but these fi ndings after infection (~1–3 months). Ultimately, HIV could require confi rmation in follow-up studies.124 The fi nding evolve to an attenuated state similar to the persistently that subtype D led to faster progression than subtype A in non-progressive SIVcpz infections noted in this study, though, might lend support to the theory that chimpanzees,53 but the time required for this attenuation subtype C infection leads to slow progression. might be hundreds to millions of years and is still the In other studies the fi ndings related to speed of subject of debate. progression in individuals infected with subtype C vary.125–127 Slow progression would support the reduced Conclusion virulence of HIV-1 subtype C purported on the basis of Of the past century of HIV, the latter 25 years have yielded in-vitro fi tness studies, which indicate subtype C isolates the greatest number of lessons in evolution and from human PBMCs had the lowest replicative fi tness epidemiology, which could be benefi cial in the study of when compared directly with other group M other viruses. Tracing the origin of HIV has confi rmed isolates.128 Direct correlations have been reported between that viruses can jump from one species to another after a replicative fi tness of an HIV-1 isolate and its virulence in very long period of no transmission and adapt rapidly.27 patients.129 For instance, patients infected with HIV-1 The spread of HIV-1 to large urban centres occurred harbouring a deletion in the nef gene and, therefore, with during preindependent and postcolonial times in Africa, severly impaired replication, had sustained non- when massive human emigration out of small villages in progression of disease.130 Defective viruses with reduced the dense tropical forests of Congo river basin occurred, cell-entry effi ciency have been identifi ed in patients with for example to Kinshasa (formerly Leopoldville) in the elite suppression, who maintain undetectable viral loads DRC, which is near the origin of HIV and has notable without antiretroviral treat ment.131 In-vivo virulence is also diversity in HIV subtypes.32 Constant and increasingly related to several host factors, such as immune response easy worldwide travel is a major contributor to HIV and genetic variations in host-derived HIV-1 co-receptors diversifi cation. New URFs and CRFs will continue to (eg, CCR5). Thus, correlations of disease progression and emerge and will defi nitely have major roles in the in-vitro HIV-1 fi tness might not be comparable across development of prevention and control strategies for patients. However, at the population level, low-frequency HIV. Improvements in sampling and monitoring diff erences in host polymorphisms related to HIV-1 techniques might facilitate the ability to use old archival progression are probably negated. As a consequence, clear samples to knit together the complex evolutionary past of distinctions in HIV-1 subtype fi tness could be the best the HIV epidemic. Finally, these advanced evolutionary correlate of diff erences in disease progression. studies on primate lentiviruses must be coupled with Spread of HIV-1 in the human population is determined phenotypic analyses on the actual viruses and advances by virus virulence and host-to-host transmission. For in understanding of spread in the population over time, example, prevalence of HIV-2 has been decreasing in changes in virulence, and transmissibility. west Africa over the past three decades, from 8·9% in 1987 to less than 4·4% by 2006,36,39,40 owing to an apparent decline in virulence. These eff ects could be related to Search strategy and selection criteria reduced replicative fi tness of certain HIV-2 strains in Data for this Review were identifi ed by searches of PubMed, 132,133 PBMCs. Poor HIV-2 fi tness has been noted in ex-vivo with the search terms “HIV subtypes”, “HIV origin”, “HIV dendritic cell cultures, which is a possible model of HIV evolution”, “HIV fi tness”, “HIV and recombination”, “HIV 134 transmission fi tness. By contrast, subtype C HIV-1 diagnosis”, “HIV molecular epidemiology”, and “SIV diversity”. isolates are still readily transmitted between human The references of identifi ed articles were manually searched 37 hosts, and compete with other HIV-1 group M isolates for further relevant papers and we also searched our own in various in-vitro models of transmission, such as in reference databases. We only included full-text penile, vaginal, or rectal tissue, and even Langerhans cell English-language papers. and human skin explants.112 This observation was in www.thelancet.com/infection Vol 11 January 2011 53 Review

Contributors 23 Hirsch VM, Olmsted RA, Murphey-Corb M, Purcell RH, EJA and DMT agreed on the structure and content of the Review. DMT Johnson PR. An African primate lentivirus (SIVsm) closely related did the initial search for published articles and wrote the fi rst draft. EJA to HIV-2. Nature 1989; 339: 389–92. and DMT revised the fi rst and subsequent drafts and approved the fi nal 24 Damond F, Worobey M, Campa P, et al. Identifi cation of a highly version for submission. divergent HIV type 2 and proposal for a change in HIV type 2 classifi cation. AIDS Res Hum Retroviruses 2004; 20: 666–72. Confl icts of interest 25 Wertheim JO, Worobey M. Dating the age of the SIV lineages that EJA has received speaker’s fees from the University of Washington, the gave rise to HIV-1 and HIV-2. PLoS Comput Biol 2009; 5: e1000377. University of Massachusetts, and Merck, has a patent under review, and 26 Lemey P, Pybus OG, Rambaut A, et al. The molecular population has received royalties from Diagnostic Hybrids/Quidel for HIV-1 genetics of HIV-1 group O. Genetics 2004; 167: 1059–68. yeast-based cloning and on an oligonucleotide ligation assay. DMT 27 Korber B, Muldoon M, Theiler J, et al. Timing the ancestor of the declares that he has no confl icts of interest. HIV-1 pandemic strains. Science 2000; 288: 1789–96. Acknowledgments 28 Rousseau CM, Learn GH, Bhattacharya T, et al. Extensive EJA and DMT are supported by grant R01 from the National Institutes intrasubtype recombination in South African human immunodefi ciency virus type 1 subtype C infections. J Virol 2007; of Health. 81: 4492–500. References 29 Gilbert MT, Rambaut A, Wlasiuk G, Spira TJ, Pitchenik AE, 1 Ho DD, Neumann AU, Perelson AS, Chen W, Leonard JM, Worobey M. The emergence of HIV/AIDS in the Americas and Markowitz M. Rapid turnover of plasma virions and CD4 beyond. Proc Natl Acad Sci USA 2007; 104: 18566–70. lymphocytes in HIV-1 infection. Nature 1995; 373: 123–26. 30 Aulicino PC, Holmes EC, Rocco C, Mangano A, Sen L. Extremely 2 Wei X, Ghosh SK, Taylor ME, et al. Viral dynamics in human rapid spread of human immunodefi ciency virus type 1 BF immunodefi ciency virus type 1 infection. Nature 1995; 373: 117–22. recombinants in Argentina. J Virol 2007; 81: 427–29. 3 Preston BD, Dougherty JP. Mechanisms of retroviral mutation. 31 Deng X, Liu H, Shao Y, Rayner S, Yang R. The epidemic origin and Trends Microbiol 1996; 4: 16–21. molecular properties of B´: a founder strain of the HIV-1 4 Preston BD, Poiesz BJ, Loeb LA. Fidelity of HIV-1 reverse transmission in Asia. AIDS 2008; 22: 1851–58. transcriptase. Science 1988; 242: 1168–71. 32 Worobey M, Gemmel M, Teuwen DE, et al. Direct evidence of 5 Temin HM. Retrovirus variation and reverse transcription: extensive diversity of HIV-1 in Kinshasa by 1960. Nature 2008; abnormal strand transfers result in retrovirus genetic variation. 455: 661–64. Proc Natl Acad Sci U S A 1993; 90: 6900–03. 33 Santiago ML, Range F, Keele BF, et al. Simian immunodefi ciency 6 Burke DS. Recombination in HIV: an important viral evolutionary virus infection in free-ranging sooty mangabeys (Cercocebus atys strategy. Emerg Infect Dis 1997; 3: 253–59. atys) from the Tai Forest, Côte d’Ivoire: implications for the origin 7 Hu WS, Temin HM. Genetic consequences of packaging two RNA of epidemic human immunodefi ciency virus type 2. J Virol 2005; genomes in one retroviral particle: pseudodiploidy and high rate of 79: 12515–27. genetic recombination. Proc Natl Acad Sci USA 1990; 87: 1556–60. 34 Schim van der Loeff MF, Aaby P. Towards a better understanding of 8 Barre-Sinoussi F, Chermann JC, Rey F, et al. Isolation of a the epidemiology of HIV-2. AIDS 1999; 13 (suppl A): S69–84. T-lymphotropic retrovirus from a patient at risk for acquired 35 Ayouba A, Mauclere P, Martin PM, et al. HIV-1 group O infection in immune defi ciency syndrome (AIDS). Science 1983; 220: 868–71. Cameroon, 1986 to 1998. Emerg Infect Dis 2001; 7: 466–67. 9 Wain-Hobson S, Sonigo P, Danos O, Cole S, Alizon M. Nucleotide 36 Lemey P, Pybus OG, Wang B, Saksena NK, Salemi M, sequence of the AIDS virus, LAV. Cell 1985; 40: 9–17. Vandamme AM. Tracing the origin and history of the HIV-2 10 Keele BF, Li H, Learn GH, et al. Low-dose rectal inoculation of epidemic. Proc Natl Acad Sci USA 2003; 100: 6588–92. rhesus macaques by SIVsmE660 or SIVmac251 recapitulates 37 UNAIDS. HIV Epidemic Update 2008. http://www.unaids.org/en/ human mucosal infection by HIV-1. J Exp Med 2009; 206: 1117–34. KnowledgeCentre/HIVData/GlobalReport/2008/2008_Global_ 11 Mitsuya Y, Varghese V, Wang C, et al. Minority human report.asp (accessed Sept 16, 2010). immunodefi ciency virus type 1 variants in antiretroviral-naive 38 Poulsen AG, Kvinesdal B, Aaby P, et al. Prevalence of and mortality persons with reverse transcriptase codon 215 revertant from human immunodefi ciency virus type 2 in Bissau, West Africa. mutations. J Virol 2008; 82: 10747–55. Lancet 1989; 1: 827–31. 12 Bailes E, Gao F, Bibollet-Ruche F, et al. Hybrid origin of SIV in 39 Larsen O, da Silva Z, Sandstrom A, et al. Declining HIV-2 chimpanzees. Science 2003; 300: 1713. prevalence and incidence among men in a community study from 13 Bibollet-Ruche F, Gao F, Bailes E, et al. Complete genome analysis Guinea-Bissau. AIDS 1998; 12: 1707–14. of one of the earliest SIVcpzPtt strains from Gabon (SIVcpzGAB2). 40 da Silva ZJ, Oliveira I, Andersen A, et al. Changes in prevalence and AIDS Res Hum Retroviruses 2004; 20: 1377–81. incidence of HIV-1, HIV-2 and dual infections in urban areas of 14 Robertson DL, Anderson JP, Bradac JA, et al. HIV-1 nomenclature Bissau, Guinea-Bissau: is HIV-2 disappearing? AIDS 2008; proposal. Science 2000; 288: 55–56. 22: 1195–202. 15 Plantier JC, Leoz M, Dickerson JE, et al. A new human 41 Hahn BH, Shaw GM, De Cock KM, Sharp PM. AIDS as a zoonosis: immunodefi ciency virus derived from gorillas. Nat Med 2009; scientifi c and public health implications. Science 2000; 287: 607–14. 15: 871–72. 42 Nerrienet E, Santiago ML, Foupouapouognigni Y, et al. Simian 16 HIV Databases. HIV Sequence database. http://www.hiv.lanl.gov immunodefi ciency virus infection in wild-caught chimpanzees (accessed Sept 16, 2010). from Cameroon. J Virol 2005; 79: 1312–19. 17 Corbet S, Muller-Trutwin MC, Versmisse P, et al. env sequences of 43 Wain LV, Bailes E, Bibollet-Ruche F, et al. Adaptation of HIV-1 to its simian immunodefi ciency viruses from chimpanzees in Cameroon human host. Mol Biol Evol 2007; 24: 1853–60. are strongly related to those of human immunodefi ciency virus 44 Mwaengo DM, Novembre FJ. Molecular cloning and characterization group N from the same geographic area. J Virol 2000; 74: 529–34. of viruses isolated from chimpanzees with pathogenic human 18 Santiago ML, Rodenburg CM, Kamenya S, et al. SIVcpz in wild immunodefi ciency virus type 1 infections. J Virol 1998; 72: 8976–87. chimpanzees. Science 2002; 295: 465. 45 Gao F, Robertson DL, Carruthers CD, et al. A comprehensive panel 19 Gao F, Bailes E, Robertson DL, et al. Origin of HIV-1 in the of near-full-length clones and reference sequences for non-subtype chimpanzee Pan troglodytes troglodytes. Nature 1999; 397: 436–41. B isolates of human immunodefi ciency virus type 1. J Virol 1998; 72: 5680–98. 20 Keele BF, Van HF, Li Y, et al. Chimpanzee reservoirs of pandemic and nonpandemic HIV-1. Science 2006; 313: 523–26. 46 Zhu T, Korber BT, Nahmias AJ, Hooper E, Sharp PM, Ho DD. An African HIV-1 sequence from 1959 and implications for the origin 21 Van HF, Li Y, Neel C, et al. Human immunodefi ciency viruses: SIV of the epidemic. Nature 1998; 391: 594–97. infection in wild gorillas. Nature 2006; 444: 164. 47 Takehisa J, Zekeng L, Ido E, et al. Human immunodefi ciency virus 22 Clavel F, Mansinho K, Chamaret S, et al. Human type 1 intergroup (M/O) recombination in Cameroon. J Virol 1999; immunodefi ciency virus type 2 infection associated with AIDS in 73: 6810–20. West Africa. N Engl J Med 1987; 316: 1180–85.

54 www.thelancet.com/infection Vol 11 January 2011 Review

48 Peeters M, Liegeois F, Torimiro N, et al. Characterization of a highly 71 Vidal N, Peeters M, Mulanga-Kabeya C, et al. Unprecedented degree replicative intergroup M/O human immunodefi ciency virus type 1 of human immunodefi ciency virus type 1 (HIV-1) group M genetic recombinant isolated from a Cameroonian patient. J Virol 1999; diversity in the Democratic Republic of Congo suggests that the HIV-1 73: 7368–75. pandemic originated in Central Africa. J Virol 2000; 74: 10498–507. 49 Gürtler LG, Hauser PH, Eberle J, et al. A new subtype of human 72 Robertson DL, Hahn BH, Sharp PM. Recombination in AIDS immunodefi ciency virus type 1 (MVP-5180) from Cameroon. viruses. J Mol Evol 1995; 40: 249–59. J Virol 1994; 68: 1581–85. 73 Hemelaar J, Gouws E, Ghys PD, Osmanov S. Global and regional 50 De Leys RJ, Vanderborght B, Vanden Haesevelde M, Heyndrickx L, distribution of HIV-1 genetic subtypes and recombinants in 2004. van GA, Wauters C, et al. Isolation and partial characterization of an AIDS 2006; 20: W13–23. unusual human immunodefi ciency retrovirus from two persons of 74 Serwadda D, Mugerwa RD, Sewankambo NK, et al. Slim disease: a west-central African origin. J Virol 1990; 64: 1207–16. new disease in Uganda and its association with HTLV-III infection. 51 Vanden Haesevelde M, Decourt JL, De Leys RJ, et al. Genomic Lancet 1985; 2: 849–52. cloning and complete sequence analysis of a highly divergent 75 Arien KK, Vanham G, Arts EJ. Is HIV-1 evolving to a less virulent African human immunodefi ciency virus isolate. J Virol 1994; form in humans? Nat Rev Microbiol 2007; 5: 141–51. 68: 1586–96. 76 Central Intelligence Agency. World fact book. www.cia.gov/library/ 52 Dalgleish AG, Beverley PC, Clapham PR, Crawford DH, Greaves publications/the-world-factbook/index.html (accessed Sept 16, 2010). MF, Weiss RA. The CD4 (T4) antigen is an essential component of 77 Ndembi N, Takehisa J, Zekeng L, et al. Genetic diversity of HIV the receptor for the AIDS retrovirus. Nature 1984; 312: 763–67. type 1 in rural eastern Cameroon. J Acquir Immune Defi c Syndr 2004; 53 Keele BF, Jones JH, Terio KA, et al. Increased mortality and 37: 1641–50. AIDS-like immunopathology in wild chimpanzees infected with 78 Tebit DM, Zekeng L, Kaptue L, Salminen M, Krausslich HG, SIVcpz. Nature 2009; 460: 515–19. Herchenroder O. Genotypic and phenotypic analysis of HIV type 1 54 Piot P, Quinn TC, Taelman H, et al. Acquired immunodefi ciency primary isolates from western Cameroon. syndrome in a heterosexual population in Zaire. Lancet 1984; 2: 65–69. AIDS Res Hum Retroviruses 2002; 18: 39–48. 55 Van de Perre P, Rouvroy D, Lepage P, et al. Acquired 79 Triques K, Bourgeois A, Saragosti S, et al. High diversity of HIV-1 immunodefi ciency syndrome in Rwanda. Lancet 1984; 2: 62–65. subtype F strains in Central Africa. Virology 1999; 259: 99–109. 56 Hahn BH, Gonda MA, Shaw GM, et al. Genomic diversity of the 80 Vidal N, Koyalta D, Richard V, et al. High genetic diversity of HIV-1 acquired immune defi ciency syndrome virus HTLV-III: diff erent strains in Chad, West Central Africa. viruses exhibit greatest divergence in their envelope genes. J Acquir Immune Defi c Syndr 2003; 33: 239–46. Proc Natl Acad Sci USA 1985; 82: 4813–17. 81 Bartolo I, Rocha C, Bartolomeu J, et al. Highly divergent subtypes 57 Sarngadharan MG, DeVico AL, Bruch L, Schupbach J, Gallo RC. and new recombinant forms prevail in the HIV/AIDS epidemic in HTLV-III: the etiologic agent of AIDS. Angola: new insights into the origins of the AIDS pandemic. Princess Takamatsu Symp 1984; 15: 301–08. Infect Genet Evol 2009; 9: 672–82. 58 Food and Drug Administration. Antiretroviral drugs used in the 82 Fischetti L, Opare-Sem O, Candotti D, Sarkodie F, Lee H, Allain JP. treatment of HIV; http://www.fda.gov/ForConsumers/ByAudience/ Molecular epidemiology of HIV in Ghana: dominance of ForPatientAdvocates/HIVandAIDSActivities/ucm118915.htm CRF02_AG. J Med Virol 2004; 73: 158–66. (accessed Sept 16, 2010). 83 Mamadou S, Vidal N, Montavon C, et al. Emergence of complex and 59 Huet T, Cheynier R, Meyerhans A, Roelants G, Wain-Hobson S. diverse CRF02-AG/CRF06-cpx recombinant HIV type 1 strains in Genetic organization of a chimpanzee lentivirus related to HIV-1. Niger, West Africa. AIDS Res Hum Retroviruses 2003; 19: 77–82. Nature 1990; 345: 356–59. 84 Esu-Williams E. Gender and HIV/AIDS in Africa—our hope lies in 60 Erice A, Mayers DL, Strike DG, et al. Brief report: primary infection the future. J Health Commun 2000; 5 (suppl): 123–26. with zidovudine-resistant human immunodefi ciency virus type 1. 85 Peeters M, Koumare B, Mulanga C, et al. Genetic subtypes of HIV N Engl J Med 1993; 328: 1163–65. type 1 and HIV type 2 strains in commercial sex workers from 61 Feng Y, Broder CC, Kennedy PE, Berger EA. HIV-1 entry cofactor: Bamako, Mali. AIDS Res Hum Retroviruses 1998; 14: 51–58. functional cDNA cloning of a seven-transmembrane, G protein- 86 Montavon C, Toure-Kane C, Nkengasong JN, et al. CRF06-cpx: coupled receptor. Science 1996; 272: 872–77. a new circulating recombinant form of HIV-1 in West Africa 62 Dragic T, Litwin V, Allaway GP, et al. HIV-1 entry into CD4+ cells is involving subtypes A, G, K, and J. J Acquir Immune Defi c Syndr 2002; mediated by the chemokine receptor CC-CKR-5. Nature 1996; 29: 522–30. 381: 667–73. 87 Ouedraogo-Traore R, Montavon C, Sanou T, et al. CRF06-cpx is the 63 Liu R, Paxton WA, Choe S, et al. Homozygous defect in HIV-1 predominant HIV-1 variant in AIDS patients from Ouagadougou, coreceptor accounts for resistance of some multiply-exposed the capital city of Burkina Faso. AIDS 2003; 17: 441–42. individuals to HIV-1 infection. Cell 1996; 86: 367–77. 88 Tebit DM, Ganame J, Sathiandee K, Nagabila Y, Coulibaly B, 64 Cooper DA, Merigan TC. Clinical treatment. AIDS 1996; Krausslich HG. Diversity of HIV in rural Burkina Faso. 10 (suppl A): S133–34. J Acquir Immune Defi c Syndr 2006; 43: 144–52. 65 Jonassen TO, Stene-Johansen K, Berg ES, et al. Sequence analysis 89 Rambaut A, Robertson DL, Pybus OG, Peeters M, Holmes EC. of HIV-1 group O from Norwegian patients infected in the 1960s. Human immunodefi ciency virus. Phylogeny and the origin of Virology 1997; 231: 43–47. HIV-1. Nature 2001; 410: 1047–48. 66 Simon F, Mauclere P, Roques P, et al. Identifi cation of a new 90 Ghys PD, Saidel T, Vu HT, et al. Growing in silence: selected human immunodefi ciency virus type 1 distinct from group M and regions and countries with expanding HIV/AIDS epidemics. group O. Nat Med 1998; 4: 1032–37. AIDS 2003; 17 (suppl 4): S45–50. 67 Guay LA, Musoke P, Fleming T, et al. Intrapartum and neonatal 91 Songok EM, Lwembe RM, Kibaya R, et al. Active generation and single-dose nevirapine compared with zidovudine for prevention of selection for HIV intersubtype A/D recombinant forms in a coinfected mother-to-child transmission of HIV-1 in Kampala, Uganda: patient in Kenya. AIDS Res Hum Retroviruses 2004; 20: 255–58. HIVNET 012 randomised trial. Lancet 1999; 354: 795–802. 92 Songok EM, Lihana RW, Kiptoo MK, et al. Identifi cation of env 68 Sheehy AM, Gaddis NC, Choi JD, Malim MH. Isolation of a human CRF-10 among HIV variants circulating in rural western Kenya. gene that inhibits HIV-1 infection and is suppressed by the viral Vif AIDS Res Hum Retroviruses 2003; 19: 161–15. protein. Nature 2002; 418: 646–50. 93 Zachar V, Goustin AS, Zacharova V, et al. Genetic polymorphism 69 Stremlau M, Owens CM, Perron MJ, Kiessling M, Autissier P, of envelope V3 region of HIV type 1 subtypes A, C, and D from Sodroski J. The cytoplasmic body component TRIM5alpha restricts Nairobi, Kenya. AIDS Res Hum Retroviruses 1996; 12: 75–78. HIV-1 infection in Old World monkeys. Nature 2004; 427: 848–53. 94 de Silva UC, Warachit J, Sattagowit N, et al. Genotypic 70 Vidal N, Mulanga-Kabeya C, Nzilambi N, Delaporte E, Peeters M. characterization of HIV type 1 env gp160 sequences from three Identifi cation of a complex env subtype E HIV type 1 virus from regions in Thailand. AIDS Res Hum Retroviruses 2010; 26: 223–27. the democratic republic of congo, recombinant with A, G, H, J, K, 95 Fontella R, Soares MA, Schrago CG. On the origin of HIV-1 subtype and unknown subtypes. AIDS Res Hum Retroviruses 2000; C in South America. AIDS 2008; 22: 2001–11. 16: 2059–64.

www.thelancet.com/infection Vol 11 January 2011 55 Review

96 Perez-Alvarez L, Munoz M, Delgado E, et al. Isolation and biological 117 Heyndrickx L, Janssens W, Ndumbe PM, et al. HIV-1 genetic characterization of HIV-1 BG intersubtype recombinants and other variability in Cameroon. AIDS 2000; 14: 1862–64. genetic forms circulating in Galicia, Spain. 118 Delwart EL, Shpaer EG, Louwagie J, et al. Genetic relationships J Med Virol 2006; 78: 1520–28. determined by a DNA heteroduplex mobility assay: analysis of 97 Parreira R, Padua E, Piedade J, Venenno T, Paixao MT, Esteves A. HIV-1 env genes. Science 1993; 262: 1257–61. Genetic analysis of human immunodefi ciency virus type 1 nef in 119 Gilbert MT, Haselkorn T, Bunce M, et al. The isolation of nucleic Portugal: subtyping, identifi cation of mosaic genes, and amino acid acids from fi xed, paraffi n-embedded tissues-which methods are sequence variability. J Med Virol 2005; 77: 8–16. useful when? PLoS One 2007; 2: e537. 98 Bello G, Passaes CP, Guimaraes ML, et al. Origin and evolutionary 120 Alkhatib G, Berger EA. HIV coreceptors: from discovery and history of HIV-1 subtype C in Brazil. AIDS 2008; 22: 1993–2000. designation to new paradigms and promise. Eur J Med Res 2007; 99 Ayehunie S, Johansson B, Sonnerborg A, et al. New subtype of 12: 375–84. HIV-1 in Ethiopia. Lancet 1990; 336: 942. 121 Huang W, Eshleman SH, Toma J, et al. Coreceptor tropism in 100 de Macedo Brigido LF. On the origin of South America HIV-1 C human immunodefi ciency virus type 1 subtype D: high prevalence epidemic. AIDS 2009; 23: 543–44. of CXCR4 tropism and heterogeneous composition of viral 101 Fontella R, Soares MA, Schrago CG. The origin of South American populations. J Virol 2007; 81: 7885–93. HIV-1 subtype C: lack of evidence for a Mozambican ancestry. 122 Kaleebu P, French N, Mahe C, et al. Eff ect of human AIDS 2009; 23: 1926–28. immunodefi ciency virus (HIV) type 1 envelope subtypes A and D 102 Soares EA, Martinez AM, Souza TM, et al. HIV-1 subtype C on disease progression in a large cohort of HIV-1-positive persons dissemination in southern Brazil. AIDS 2005; 19 (suppl 4): S81–86. in Uganda. J Infect Dis 2002; 185: 1244–50. 103 Dias CF, Nunes CC, Freitas IO, et al. High prevalence and 123 Belshe R, Franchini G, Girard MP, et al. Support for the RV144 HIV association of HIV-1 non-B subtype with specifi c sexual vaccine trial. Science 2004; 305: 177–80. transmission risk among antiretroviral naive patients in Porto 124 Arts EJ, Demers K, Nankya I, et al. Infection with subtype C HIV-1 Alegre, RS, Brazil. Rev Inst Med Trop Sao Paulo 2009; 51: 191–96. of lower replicative fi tness as compared to subtypes A and D leads 104 Rodriguez R, Manenti S, Romao PR, et al. Young pregnant women to slower disease progression in Zimbabwean and Ugandan living with HIV/AIDS in Criciuma, Southern Brazil, are infected women. XVI International AIDS Conference; Toronto; Aug 13–18, almost exlusively with HIV type 1 clade C. 2006 (abstr TUAA0204). AIDS Res Hum Retroviruses 2010; 26: 351–57. 125 Mehendale SM, Bollinger RC, Kulkarni SS, et al. Rapid disease 105 Cuevas MT, Ruibal I, Villahermosa ML, et al. High HIV-1 genetic progression in human immunodefi ciency virus type 1-infected diversity in Cuba. AIDS 2002; 16: 1643–53. seroconverters in India. AIDS Res Hum Retroviruses 2002; 106 Peeters M, Sharp PM. Genetic diversity of HIV-1: the moving target. 18: 1175–79. AIDS 2000; 14 (suppl 3): S129–40. 126 Choge I, Cilliers T, Walker P, et al. Genotypic and phenotypic 107 Robertson DL, Sharp PM, McCutchan FE, Hahn BH. characterization of viral isolates from HIV-1 subtype C-infected Recombination in HIV-1. Nature 1995; 374: 124–26. children with slow and rapid disease progression. AIDS Res Hum Retroviruses 2006; 22: 458–65. 108 Sharp PM, Robertson DL, Hahn BH. Cross-species transmission and recombination of ‘AIDS’ viruses. 127 Walker PR, Ketunuti M, Choge IA, et al. Polymorphisms in Nef Philos Trans R Soc Lond B Biol Sci 1995; 349: 41–47. associated with diff erent clinical outcomes in HIV type 1 subtype C-infected children. AIDS Res Hum Retroviruses 2007; 23: 204–15. 109 Abecasis AB, Lemey P, Vidal N, et al. Recombination confounds the early evolutionary history of human immunodefi ciency virus type 1: 128 Abraha A, Nankya IL, Gibson R, et al. CCR5- and CXCR4-tropic subtype G is a circulating recombinant form. J Virol 2007; 81: 8543–51. subtype C human immunodefi ciency virus Type 1 isolates have a lower level pathogenic fi tness than other dominant group M 110 Piyasirisilp S, McCutchan FE, Carr JK, et al. A recent outbreak of subtypes: Implications for the Epidemic. J Virol 2009; 83: 5592–605. human immunodefi ciency virus type 1 infection in southern China was initiated by two highly homogeneous, geographically separated 129 Troyer RM, Collins KR, Abraha A, et al. Changes in human strains, circulating recombinant form AE and a novel BC immunodefi ciency virus type 1 fi tness and genetic diversity during recombinant. J Virol 2000; 74: 11286–95. disease progression. J Virol 2005; 79: 9006–18. 111 Gomez CM, Avila M, Hierholzer J, et al. Mother-to-child HIV type 1 130 Kirchhoff F, Greenough TC, Brettler DB, Sullivan JL, transmission in Argentina: BF recombinants have predominated in Desrosiers RC. Brief report: absence of intact nef sequences in a infected children since the mid-1980s. long-term survivor with nonprogressive HIV-1 infection. AIDS Res Hum Retroviruses 2002; 18: 477–83. N Engl J Med 1995; 332: 228–32. 112 Yamaguchi J, Bodelle P, Vallari AS, et al. HIV infections in 131 Lassen KG, Lobritz MA, Bailey JR, et al. Elite suppressor-derived northwestern Cameroon: identifi cation of HIV type 1 group O and HIV-1 envelope glycoproteins exhibit reduced entry effi ciency and dual HIV type 1 group M and group O infections. kinetics. PLoS Pathog 2009; 5: e1000377. AIDS Res Hum Retroviruses 2004; 20: 944–57. 132 Kanki PJ, Travers KU, Mboup S, et al. Slower heterosexual spread 113 Baird HA, Galetto R, Gao Y, et al. Sequence determinants of of HIV-2 than HIV-1. Lancet 1994; 343: 943–46. breakpoint location during HIV-1 intersubtype recombination. 133 Marlink R, Kanki P, Thior I, et al. Reduced rate of disease Nucleic Acids Res 2006; 34: 5203–16. development after HIV-2 infection as compared to HIV-1. 114 Courgnaud V, Salemi M, Pourrut X, et al. Characterization of a Science 1994; 265: 1587–90. novel simian immunodefi ciency virus with a vpu gene from greater 134 Arien KK, Abraha A, Quinones-Mateu ME, et al. The replicative spot-nosed monkeys (Cercopithecus nictitans) provides new insights fi tness of primary human immunodefi ciency virus type 1 (HIV-1) into simian/human immunodefi ciency virus phylogeny. group M, HIV-1 group O, and HIV-2 isolates. J Virol 2005; J Virol 2002; 76: 8298–309. 79: 8979–90. 115 Kwon JA, Yoon SY, Lee CK, et al. Performance evaluation of three 135 Anderson RM, May RM. Infectious diseases of humans: dynamics automated human immunodefi ciency virus antigen-antibody and control. Oxford: Oxford University Press, 1992. combination immunoassays. J Virol Methods 2006; 133: 20–26. 116 Grez M, Dietrich U, Balfe P, et al. Genetic analysis of human immunodefi ciency virus type 1 and 2 (HIV-1 and HIV-2) mixed infections in India reveals a recent spread of HIV-1 and HIV-2 from a single ancestor for each of these viruses. J Virol 1994; 68: 2161–68.

56 www.thelancet.com/infection Vol 11 January 2011